Hx, a novel fluorescent, minor groove and sequence specific recognition element: Design, synthesis, and DNA binding properties of p - anisylbenzimidazole-imidazole/pyrrole-containing polyamides

Article abstract of DOI:10.1021/bi102028a

With the aim of incorporating a recognition element that acts as a fluorescent probe upon binding to DNA, three novel pyrrole (P) and imidazole (I)-containing polyamides were synthesized. The compounds contain a p-anisylbenzimidazolecarboxamido (Hx) moiety attached to a PP, IP, or PI unit, giving compounds HxPP (2), HxIP (3), and HxPI (4), respectively. These fluorescent hybrids were tested against their complementary nonfluorescent, non-formamido tetraamide counterparts, namely, PPPP (5), PPIP (6), and PPPI (7) (cognate sequences 5′-AAATTT-3′, 5′-ATCGAT-3′, and 5′-ACATGT-3′, respectively). The binding affinities for both series of polyamides for their cognate and noncognate sequences were ascertained by surface plasmon resonance (SPR) studies, which revealed that the Hx-containing polyamides gave binding constants in the 10⁶ M^-1 range while little binding was observed for the noncognates. The binding data were further compared to the corresponding and previously reported formamido-triamides f-PPP (8), f-PIP (9), and f-PPI (10). DNase I footprinting studies provided additional evidence that the Hx moiety behaved similarly to two consecutive pyrroles (PP found in 5-7), which also behaved like a formamido-pyrrole (f-P) unit found in distamycin and many formamido-triamides, including 8-10. The biophysical characterization of polyamides 2-7 on their binding to the abovementioned DNA sequences was determined using thermal melts (δT_M), circular dichroism (CD), and isothermal titration calorimetry (ITC) studies. Density functional calculations (B3LYP) provided a theoretical framework that explains the similarity between PP and Hx on the basis of molecular electrostatic surfaces and dipole moments. Furthermore, emission studies on polyamides 2 and 3 showed that upon excitation at 322 nm binding to their respective cognate sequences resulted in an increase in fluorescence at 370 nm. These low molecular weight polyamides show promise for use as probes for monitoring DNA recognition processes in cells.

Full text of DOI:10.1021/bi102028a

ARTICLE
pubs.acs.org/biochemistry
Hx, a Novel Fluorescent, Minor Groove and Sequence Specific
Recognition Element: Design, Synthesis, and DNA Binding Properties
of p-Anisylbenzimidazole-imidazole/pyrrole-Containing Polyamides
Sameer Chavda,† Yang Liu,‡ Balaji Babu,† Ryan Davis,† Alan Sielaff,† Jennifer Ruprich,† Laura Westrate,†
Christopher Tronrud,† Amanda Ferguson,† Andrew Franks,† Samuel Tzou,† Chandler Adkins,# Toni Rice,†
Hilary Mackay,† Jerome Kluza,§ Sharjeel A Tahir,§ Shicai Lin,§ Konstantinos Kiakos,§ Chrystal D. Bruce,#
W. David Wilson,‡ John A. Hartley,§ and Moses Lee*,†
†Division of Natural and Applied Sciences and Department of Chemistry, Hope College, Holland, Michigan 49423, United States
‡Department of Chemistry, Georgia State University, Atlanta, Georgia 30302, United States
§Cancer Research UK Drug-DNA Interactions Research Group, UCL Cancer Institute, Paul O’Gorman Building, 72 Huntley Street,
London WC1E 6BT, UK
#Department of Chemistry, Erskine College, Due West, South Carolina 29639, United States
S Supporting Information
b
ABSTRACT: With the aim of incorporating a recognition
element that acts as a fluorescent probe upon binding to DNA,
three novel pyrrole (P) and imidazole (I)-containing polyamides
were synthesized. The compounds contain a p-anisylbenzimida-
zolecarboxamido (Hx) moiety attached to a PP, IP, or PI unit,
giving compounds HxPP (2), HxIP (3), and HxPI (4), respec-
tively. These fluorescent hybrids were tested against their
complementary nonfluorescent, non-formamido tetraamide
counterparts, namely, PPPP (5), PPIP (6), and PPPI (7)
(cognate sequences 50-AAATTT-30, 50-ATCGAT-30, and 50-
ACATGT-30, respectively). The binding affinities for both series
of polyamides for their cognate and noncognate sequences were
ascertained by surface plasmon resonance (SPR) studies, which
revealed that the Hx-containing polyamides gave binding con-
stants in the 106 Mꢀ1 range while little binding was observed for
the noncognates. The binding data were further compared to the
corresponding and previously reported formamido-triamides
f-PPP (8), f-PIP (9), and f-PPI (10). DNase I footprinting studies provided additional evidence that the Hx moiety behaved
similarly to two consecutive pyrroles (PP found in 5ꢀ7), which also behaved like a formamido-pyrrole (f-P) unit found in distamycin
and many formamido-triamides, including 8ꢀ10. The biophysical characterization of polyamides 2ꢀ7 on their binding to the
abovementioned DNA sequences was determined using thermal melts (ΔTM), circular dichroism (CD), and isothermal titration
calorimetry (ITC) studies. Density functional calculations (B3LYP) provided a theoretical framework that explains the similarity
between PP and Hx onthe basisof molecular electrostatic surfaces and dipole moments. Furthermore, emission studieson polyamides
2 and 3 showed that upon excitation at 322 nm binding to their respective cognate sequences resulted in an increase in fluorescence at
370 nm. These low molecular weight polyamides show promise for use as probes for monitoring DNA recognition processes in cells.
yrrole (P) and imidazole (I)-containing polyamide analogues
of the naturally occurring distamycin A (1) bind in the minor
expression has been widely explored.7 However, attempts to
inhibit the transcription of genes that code for a disease in cell
lines have been met with moderate success which presumably
points toward poor cellular uptake of some of these polyamides
and their limited ability to achieve nuclear localization.8,9 Some
success in this area has been achieved through the attachment of
P
groove of DNA in a stacked, antiparallel 2:1 (ligand/DNA)
motif.1ꢀ3 Early studies have led to the establishment of a set of
pairing rules for DNA base pair recognition; a P/P stacking
recognizes an A/T or T/A base pair, an I/P stacking binds to a
G/C base pair, while a P/I stacking recognizes a C/G base pair.
In addition, a stacked I/I pairing targets a T/G mismatch.4ꢀ6 As a
result of the high degree of sequence specificity that these
molecules exhibit, their potential use as regulators of gene
Received: December 21, 2010
Revised:
March 9, 2011
Published: March 10, 2011
r
2011 American Chemical Society
3127
dx.doi.org/10.1021/bi102028a Biochemistry 2011, 50, 3127–3136
|

Biochemistry
ARTICLE
Figure 1. Structures of distamycin A (1); target fluorescent Hx polyamides: HxPP (2), HxIP (3), HxPI (4); their nonfluorescent nonformamido
tetraamide counterparts PPPP (5), PPIP (6), PPPI (7); and the formamido-triamides f-PPP (8), f-PIP (9), and f-PPI (10).
specific chemical groups that facilitate cell uptake, and these
pendant groups include fluorescein (FITC)9aꢀc and isophthalic
acid.9d,e Over the past decade, the need for the development of
DNA recognition elements that exhibit fluorescence, in addition
to high sequence specificity and binding affinity, upon binding to
their target DNA sequences has become evident. Such com-
pounds would provide an intrinsic probe for monitoring the
cellular and nuclear uptake of the molecule, movement through
the cell, and ultimately binding to nuclear DNA.
A large number of molecules are known to display an increase
in fluorescence upon binding to DNA with a prime example
being the dye 40,6-diamidino-2-phenylindole (DAPI).10 How-
ever, the usefulness of DAPI as a tracking element is hampered
due to the multiple binding modes it possesses in the minor
The bis-benzimidazole fluorophores, such as bisbenzimide or
Hoechst 33258, are known to selectively recognize A/T-rich
sequences in the DNA minor groove and are capable of crossing
certain cellular and nuclear membranes.19 To this end, Bruice
and co-workers have developed hairpin tripyrrole/imidazole-
Hoechst conjugates and have shown their capability of passing
through NIH 3T3 cell membranes and inhibiting DNA tran-
scription factor (TF) binding.19 Although a number of prede-
termined DNA sequences can be targeted through the design of
hairpin polyamides through obeying the established pairing
rules, their poor cellular uptake still remains a drawback.
On the basis of the sequence specificity and high affinity of
formamido-containing polyamides for the minor groove as
ascertained from our previous studies,20 we envisioned the
incorporation of a p-anisylbenzimidazolecarboxamido (Hx) moi-
ety of Hoechst 33258 attached to imidazoleꢀpyrrole, pyrroleꢀ
pyrrole, or pyrroleꢀimidazole heterocyclic units (2, 3, and 4
respectively; Figure 1). Because of the aforementioned specificity
and affinity of Hoechst 33258 for A/T-rich sequences, we
anticipated that the Hx component of the hybrid polyamide
should behave similarly to a non-formamido, terminal diheter-
ocyclic unit (PP), thus, binding over two A/T base pairs. In
effect, the Hx component also behaves similarly to a formamido-
pyrrole (f-P) unit found in formamido-containing polyamides
since a stacked f-P unit also recognizes two A/T base pairs similar
to a PP unit but with enhanced binding affinity.20b,f
groove,11,12 major groove,13 and as an intercalator.14 The
incorporation of fluorescent dyes such as thiazole orange,15
fluorescein,16 and bodipy17 into hairpin polyamides is well
documented. The results from the DNA binding studies of such
fluorophore-polyamide hybrids have demonstrated that such
compounds can induce fluorescence while maintaining a high
degree of specificity and affinity upon binding to their cognate
DNA sequences. However, nuclear localization of such mol-
ecules has been subject to a number of restrictions, which include
DNA sequence, position of imidazole residues, dye choice,
position of attachment, and hairpin linker composition.8,9aꢀ9c
In addition, the behavior of the nonfluorescing hairpins did not
always mimic the behavior of its fluorescing counterpart thus
limiting the usefulness of this approach. Thus, there is clearly
a need for the development of polyamide fluorophores that
have increased cell and nuclear penetration abilities. To address
this issue, a number of hairpin polyamides have been reported
wherein the fluorescing component is incorporated as a hetero-
cyclic recognition element.18
Because of the absence of an N-terminus formamido group as
found in distamycin (1), and with reference to the aforemen-
tioned pairing rules, polyamides 2, 3, and 4 (Figure 1) are
anticipated to form an overlapped stacked dimer and bind in
a 2:1 fashion to their cognate sequences 50-AAATTT-30,
50-ATCGAT-30, 50-ACATGA-30 respectively.20 Three control
nonfluorescing polyamides PPPP (5),21,22 PPIP (6), and PPPI (7)
3128
dx.doi.org/10.1021/bi102028a |Biochemistry 2011, 50, 3127–3136

Biochemistry
ARTICLE
were also synthesized which should target the same sequences.
The binding affinities, minor groove and sequence specificity and
fluorescence properties of 2ꢀ7 were studied against their cognate
sequences and a negative control sequence 50-ACGCGT-30 using
thermal melts (ΔTM), circular dichroism (CD), surface plasmon
resonance (SPR), isothermal titrationcalorimetry (ITC), DNaseI
footprinting, and fluorescence emission studies. In addition, the
binding properties of the Hx compounds were also compared to
their formamidoꢀtriamide counterparts, f-PPP 8,20f f-PIP 9,20c,f
f-PPI 10.20f
control. The SPR binding experiments were performed in 0.01 M
cacodylic acid (CCA) solution at pH 6.25 containing 0.001 M
EDTA (disodium ethylenediamine tetraacetate), 0.1 M NaCl,
and 0.005% v/v surfactant P20 (Biacore AB) was used to reduce
the nonspecific binding of polyamides to the fluidics and sensor
chip surface. A 0.01 M N-[2-hydroxyethyl]piperazine-N0-[2-
ethanesulfonic acid] (HEPES) solution at pH 7.4 containing
0.15 M NaCl, 0.003 M EDTA, and 0.005% v/v surfactant P20
was used during the DNA immobilization process. All buffers
were degassed and filtered prior to experiments. DNA sequences
were obtained from Integrated DNA Technologies (San Diego,
CA) with HPLC purification and were used without further
purification. The lyophilized 50 biotin-labeled DNA hairpin
constructs were dissolved in the appropriate amount of DI
H2O to create 1.0 mM DNA stock solutions. Further dilutions
were made using 0.01 M HEPES buffer during the DNA
immobilization step. DNA concentrations were determined
spectrophotometrically using molar absorptivity coefficients
calculated for each individual DNA sequence by using the nearest
neighboring method for the singled-stranded DNA method.
Stock solutions of 4 and 5 were prepared by dissolving the solid
compound in the necessary amount of distilled H2O to create a
stock solution with a concentration of approximately 1.0 ꢁ 10ꢀ3 M.
Stock solutions were kept frozen at 4 °C until experimental
use to minimize any degradation of the compound. Samples for
SPR experiments were prepared by a series of dilutions from the
stock solution using a 0.01 M cacodylic acid buffer solution. The
SPR experiments were conducted using a four-flow cell BIAcore
2000 biosensor instrument (GE Life Sciences). DNA hairpin
constructs labeled with biotin at the 50 end were immobilized onto
a streptavidin-coated sensor chip (BIAsensorSA) as previously
reported.20a,b,24 The 50-biotin-labeled DNA hairpins were immo-
bilized on three ofthe four flow cells. Thefourth was left blank and
used as a control. All SPR experiments were performed at 25 °C
and used 0.01 M CCA as the running buffer. The amount of DNA
immobilized was approximately 400 response units (RU). This
was achieved by continuously injecting ∼20 μL of an approxi-
mately 50 nM DNA solution at a rate of 2 μL/min onto the sensor
chip surface until a relative response of 400 units was reached.
Binding data were obtained by injecting known concentrations
and were analyzed with one or two site binding models as
previously described.20a,b,24
’ EXPERIMENTAL SECTION
Thermal Denaturation (ΔTM) Studies. DNA oligomers were
purchased from Operon with the following sequences: A3T3_10;
50-CGAAATTTCCCTCTGGAAATTTCG-30; TCGA_10, 50-
50-GAATCGATTGCTCTCAATCGATTC-30; ACGCGT;
50-GAACGCGTCGCTCTCGACGCGT-TC-30; CATG_10;
50-GAACATGTTGCTCTCAACATGTTC-30 and CTAG_10;
50-GAACT-AGTTGCTCTCAACTAGTTC-30. Thermal dena-
turation studies were performed by following an earlier published
procedure.23 Experiments for polyamides 2ꢀ7 were performed
at a concentration of 3 μM ligand and 1 μM DNA. All experi-
ments were run in PO40 buffer (10 mM sodium phosphate,
1 mM EDTA, pH 6.2). For the salt-effect experiments, NaCl was
added accordingly to obtain Naþ concentrations of 25, 200,
and 400 mM. Oligonucleotide samples were reannealed prior
to denaturation studies by heating at 70 °C for 1 min then
allowing them to cool to rt. Experiments were typically per-
formed between 25 and 95 °C, with a heating rate of 0.5 °C
minꢀ1 while continuously monitoring the absorbance at 260 nm
(digitally sampled at 200 ms intervals). All melts were performed
in 10-mm path length quartz cells. TM values were determined as
the maximum of the first derivative.
Circular Dichroism (CD) Studies. CD studies were performed
using previously reported procedures20b,24 and were conducted
at ambient temperature in a 1-mm path length quartz cell using
PO40 buffer. Buffer and stock DNA were added to the cuvette to
give a final DNA concentration of 9 μM. Each polyamide (in 500
μL in double distilled H2O) was titrated in 1 molar equivalents
into the relevant DNA (160 μL of 9 μM DNA) until saturation
was achieved. Each run was performed over 400ꢀ220 nm. The
CD response at the λmax of the induced peak was plotted against
the mole ratio of ligand/DNA.
2
2
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
r ¼ ðK1 Cfree þ 2 K1 K2 Cfree Þ=ð1 þ K1 Cfree þ K1 K2 Cfree
Þ
Emission Studies. The experiments were performed using a
SPEX Fluoro-Log 2 spectrofluorometer to measure the emission
spectrum on the binding of Hx-polyamides 2 and 3 to their
respective cognate sequences. A 20 μM solution of DNA in PO40
buffer was introduced to a 1 mm path length cuvette. The
polyamides were subsequently added to the DNA solutions in
1 mol equivalent aliquots past the point of saturation. The
emission spectrum was recorded upon excitation of the Hx
moiety in the complex at 322 nm; the λmax of emission spectrum
was measured at 370 nm. KaleidaGraph was then used to analyze
and plot the data.
Biosensor-Surface Plasmon Resonance (SPR). SPR mea-
surements were performed with a four-channel BIAcore T100
optical biosensor system (Biacore, GE Healthcare Inc.).
50-Biotin-labeled DNA hairpin duplex samples were immobilized
onto streptavidinꢀcoated sensor chips (BIAcore SA) as pre-
viously described.20 Three flow cells were used to immobilize
the DNA oligomer samples, while a fourth cell was left blank as a
where r represents the moles of bound compound per mole of
DNA hairpin duplex, K1, K2 are macroscopic binding constants,
and Cfree is the free compound concentration in equilibrium with
the complex.
Isothermal Titration Calorimetry (ITC). ITC analysis was
performed at 25 °C using a VP-ITC microcalorimeter (MicroCal,
Northampton, MA) aspreviouslyreported.25 Afterthe systemwas
equilibrated to 25 °C, and after an initial delay of 300 s, the
polyamide (50 μM) was titrated 50 times (3 μL for 7.2 s, repeated
every 240 s) into 2 μM DNA. The solution was constantly mixed
at 300 rpm. The reference cell contained PO40 or PO420 as
appropriate for the experiment. Origin 7.0 was used to analyze the
data, where the integration of each titration was plotted against
ligand concentration. A linear fit was then employed, and this was
subtracted from the reaction integrations to normalize for non-
specific heat components. A two-site, nonsequential model was
then fitted to the curve and used to determine the binding
constant, Keq.25
3129
dx.doi.org/10.1021/bi102028a |Biochemistry 2011, 50, 3127–3136

Biochemistry
ARTICLE
Scheme 1. Synthesis of Polyamides 2ꢀ7a
a Conditions and reagents (i) 15, 16, or 17; 5% Pd/C, H2, MeOH, rt 12 h; (ii) 14 in SOCl2/THF (1:1), 90°C, 1.5 h; (iii) add Hx acid chloride to amine,
Et3N, DCM, 0 °C then stir 12 h rt; (iv) N-methylpyrrole-2-carboxylic acid, EDCI, DMAP, DCM, stir rt 3 days; (v) add 1-methylpyrrole-2-carbonyl
chloride to amine, Et3N, DCM, 0 °C to rt, 12 h.
DNase1 Footprinting Studies with DNase I and Micro-
coccal Nuclease. Radiolabeled DNA fragments of 131 base pairs
each, designed to contain the cognate sites of the Hx compounds
and their nonfluorescent tertaamide counterparts were generated
by polymerase chain reaction, as described previously.26 The
amplified fragments were purified on a Bio-Gel P6 column
followed by agarose gel electrophoresis and isolated using the
Mermaid Kit (MP biomedicals) according to the manufacturer’s
instructions.
DNase I digestions were conducted in a total volume of 8 μL.
In each case, the labeled DNA fragment (2 μL, 200 counts sꢀ1
)
was incubated for 30 min at room temperature in 4 μL of
TN binding buffer (10 mM Tris Base, 10 mM NaCl, pH 7)
containing the required drug concentration. Cleavage by
DNase I was initiated by addition of 2 μL of DNase I solution
(20 mM NaCl, 2 mM MgCl2, 2 mM MnCl2, DNase I 0.02U,
pH 8.0) and stopped after 3 min by snap freezing the samples
on dry ice.
Micrococcal nuclease digestions were conducted in a total
volume of 8 μL. The labeled DNA fragment (2 μL, 200 counts
s
binding buffer (10 mM Tris Base, 10 mM NaCl, pH 7) contain-
ing the required drug concentration. Cleavage by MNase (NEB)
was initiated by addition of 2 μL of MNase (5U) diluted in 1ꢁ
micrococcal nuclease reaction buffer (50 mM Tirs-HCl, 5 mM
CaCl, pH 7.9) followed by incubation at 37 °C and stopped after
5 min by snap freezing the samples on dry ice.
The nuclease-digested samples were subsequently lyophilized
to dryness and resuspened in 5 μL of formamide loading dye
(95% formamide, 20 mM EDTA, 0.05% bromophenol blue, and
0.05% xylene cyanol). Following heat denaturation for 5 min at
90 °C, the samples were loaded on a denaturing polyacrylamide
(10%) gel (Sequagel, National Diagnostics, UK) containing urea
(7.5 mM). Electrophoresis was carried out for 2 h at 1650 V
(∼70 W, 50 °C) in 1ꢁ TBE buffer. The gel was then transferred
onto Whatman 3MM and dried under a vacuum at 80 °C for 2 h.
ꢀ1) was incubated for 30 min at room temperature in 4 μL TN
3130
dx.doi.org/10.1021/bi102028a |Biochemistry 2011, 50, 3127–3136

Biochemistry
ARTICLE
The gel was exposed overnight to Fuji medical X-Ray film and
developed on a Konica Medical Film Processor SRX-101A.
Molecular Modeling Studies. The Gaussian 09 software
package was used to perform geometry optimizations, generate
electrostatic potential surfaces, and calculate dipole moments of
Hx and PP.27a Specifically, the calculations were conducted on
the carboxamide derivative of Hx-acid 14 (Hx-carboxamide), and
4-(1-methylpyrrole-2-carboxamido)-1-methylpyrrole-2-carboxa-
mide (PP-carboxamide). All calculations were performed at the
B3LYP level with the 6-31G(d) basis set. WebMO was used as
the interface to prepare all calculations, which were conducted on
the MU3C cluster computer at Hope College.27b
13 over 5% Pd/C in methanol furnished the amine intermedi-
ates, which were immediately coupled to p-anisylbenzimidazole-
2-carbonyl chloride which, in turn, was synthesized by refluxing
the corresponding carboxylic acid 1428 in a 1:1 mixture of thionyl
chloride and dry THF. The desired polyamides 2ꢀ4 were
isolated in ∼20% yield. The low yields can be attributed to the
instability of the amine intermediates.29 For the synthesis of 5
(PPPP), an EDCI-mediated coupling of N-methylpyrrole-2-
carboxylic acid to the reduced form of the nitro triamide 15
was performed. The other two nonfluorescent control non-
formamido tetraamides 6 and 7 were synthesized utilizing
Schotten-Baumann conditions from the reduction of their
respective nitro-triamides 16 and 17 followed by coupling to
N-methylpyrrole-2-carbonyl chloride (which in turn was synthe-
sized from N-methylpyrrole-2-carboxylic acid and SOCl2).
Thermal Denaturation Studies. The binding of the Hx-
substituted polyamides 2ꢀ4 along with their nonfluorescing
non-formamido tetraamides 5ꢀ7 was tested using the DNA
sequences shown in Figure 2 in thermal denaturation experi-
ments. An excess of polyamides (3 μM) to DNA (1 μM) was
used in these studies, and in all cases the concentration of DNA
used was identical. The DNA sequences examined contained the
following sequences: 50-AAATTT-30 (A3T3_10), the cognate
site for distamycin A and that anticipated for 2, 5, and 8,
50-ATCGAT-30 (TCGA_10) (anticipated to be the preferred
binding sequence for 3, 6 and 9) and 50-ACATGT- 30
(CATG_10) along with 50-ACTAGT-30 (CTAG_10) (anti-
cipated to be the preferred binding sequences for 4, 7,
and 10). In addition, a GC-rich sequence 50-ACGCGT- 30
(ACGCGT) was used as a negative control. These thermal
denaturation experiments screened the preferred binding se-
quences of the fluorescent hybrid polyamides 2ꢀ4 along with
their nonfluorescing counterparts 5ꢀ7 by measuring their ability
to stabilize each double-stranded DNA upon binding. ΔTM
values were determined from the differences in melting tem-
perature of the DNAꢀpolyamide complexes and duplex DNA
alone and are shown in Table 1. For comparative purposes,
the reported ΔTM value for distamycin A against its cognate
sequence A3T3_10 is 14 °C.25
’ RESULTS AND DISCUSSION
Chemistry. The three Hx-substituted polyamides 2, 3, and 4
were synthesized as described in Scheme 1. Reduction of the
nitro group of the diheterocyclic intermediates O2N-PP-N,N-
dimethyl-3-aminopropylamine 11, O2N-IP-N,N-dimethyl-3-amino-
propylamine 12, and O2N-PI-N,N-dimethyl-2-aminoethylamine
Figure 2. Proposed sequence specific recognition of polyamides 2ꢀ7
to their target (cognate) DNA sequences. Molecules are shown binding
in an antiparallel, stacked motif. A = HxPP (2), B = PPPP (5), C = HxIP
(3), D = PPIP (6), E = HxPI (4), and F = PPPI. A3T3_10 is also referred
as 50-AAATTT-30; likewise TCGA_10 is 50-ATCGAT-30; ACGCGT is
50-ACGCGT- 30; CATG_10 is 50-ACATGT- 30; and CTAG_10 is
50-ACTAGT-30.
The data presented in Table 1 indicate that the Hx-substituted
polyamides 2ꢀ4 bind more strongly to their cognate sequences
(ΔTM .10 °C) relative to their nonfluorescent counterparts,
non-formamido tetraamides 5ꢀ7, and the formamido-triamides
8ꢀ1020c,f (both ΔTM ,10 °C). For HxPP (2) and HxIP (3)
(ΔTM 19 and 15 °C respectively), this is a greater than 3-fold
Table 1. Thermal Denaturation Data and SPR Binding Constants for Polyamides 2ꢀ10
A3T3_10 ΔTM (°C)a;
[Keq Mꢀ1 b
TCGA_10 ΔTM (°C);
ACGCGT ΔTM (°C);
CATG_10 ΔTM (°C);
CTAG_10 ΔTM (°C);
[Keq Mꢀ1
]
[Keq Mꢀ1
]
[Keq Mꢀ1
]
[Keq Mꢀ1
]
]
HxPP (2)
HxIP (3)
19; [1 ꢁ 106]c
5; [4 ꢁ 106]d
18; [5 ꢁ 106]
5; [3 ꢁ 106]
0; [2 ꢁ 104]
4; [3 ꢁ 106]
9; [3 ꢁ 106]
3; [2 ꢁ 105]
6; [4 ꢁ 105]
10; [<103]
0;
- [<1 ꢁ 103]
- [1 ꢁ 104]
15; [3 ꢁ 106]
14; [2 ꢁ 106]
2;
4; [4 ꢁ 104]
0;
0;
5;
HxPI (4)
19; [7 ꢁ 106]
17;
PPPP (5)
PPIP (6)
0; [<1 ꢁ 105]
0;
0; [<1 ꢁ 104]
PPPI (7)
8; [3 ꢁ 106]
8;
f-PPP (8)20f
f-PIP (9)20f
f-PPI (10)20f
0; [2 ꢁ 105]
1; [1 ꢁ 105]
a ΔTM = TM (bound DNA) ꢀ TM (free DNA). b Keq = (K1K2)1/2, determined at 25 °C. c Cognate sequences are typed in bold face. d Denotes a 1:1
binding mode.
3131
dx.doi.org/10.1021/bi102028a |Biochemistry 2011, 50, 3127–3136

Biochemistry
ARTICLE
Figure 3. Circular dichroism results for polyamides 2ꢀ7 with cognate and noncognates DNA sequences. In each experiment, 9 μM DNA in PO40
was used, and the polyamides were titrated at ratios of 1, 2, 3, 4, 5, 6, and 8 mol equiv of polyamides relative to DNA. A ratio of 0:1 indicates DNA alone.
Note that the CD (mdeg) scale on the right-hand-side of the figure is applicable to the spectra marked with an asterisk.
increase in melting temperature relative to that given by PPPP
(5) and PPIP (6) with A3T3_10 (ΔTM 5 °C) and TCGA_10
(ΔTM 4 °C), respectively. The ΔTM values for f-PPP (8) and
f-PIP (9) were reported at 9 and 0 °C, respectively.20c,f These
results indicate that the Hx group contributes favorable stabiliz-
ing properties when it is bound to its cognate DNA sequence.
In addition, no ΔTM was observed for the GC-rich sequence
ACGCGT, which contains a flanking G/C base pair toward the 50
and 30 termini of the oligonucleotide. This suggests that the Hx
group, when paired with a pyrrole unit, prefers A/T base pairs
over G/C base pairs.
HxPI (4) gave a greater than 2-fold increase in melting
temperature (ΔTM > 10 °C) over 7 upon binding to the
degenerate cognate sites 50-ACATGT-30 and 50-ACTAGT-30.
For 50-ACTAGT-30, the ΔTM for f-PPI (10) was reported at
1 °C.20f As f-PPI (10), HxPI (4) was found to bind to the
noncognate 50-AAATTT-30 site (ΔTM 18 °C at A3T3_10 versus
19 and 17 °C at 50-ACATGT-30 and 50-ACTAGT-30, re-
spectively) indicating tolerance of AT/TA over CG/GC base
pairs. In addition HxPI (4) gave a ΔTM value of 14 °C upon
binding to the noncognate alternative GC/CG containing
50-ATCGAT-30 site while in comparison, it gave a ΔTM value
of 5 °C for binding to 50-ACGCGT-30. This suggests that the
benzimidazole (of Hx)/P pairing has preference for A/T base
pairs at the 50 and 30-flanks of the hexamer sequence. The poor
binding of 4 to ACGCGT provides additional support for the
binding preference. Overall, the TM data for polyamides 2ꢀ4
suggest that the p-anisylbenzimidazolecarboxamido (Hx) moiety
binds to two AT or TA base pairs with a higher affinity than that
of a PP heterocyclic unit and an f-P unit. Although it should be
stated that the shift in TM upon binding is a complicated function
of the enthalpy of DNA melting alone, the ligand binding affinity
and stoichiometry, the ligand binding enthalpy, and the ligand
concentration.
Circular Dichroism Studies. CD spectroscopy was used to
probe the binding of polyamide 2ꢀ7 in the minor groove of
double-stranded DNA. CD assays were conducted by titrating
each ligand to DNA solutions comprised of the DNA sequences
tested in the thermal denaturation studies (one cognate and
three control DNA sequences). In all cases, a fixed DNA
concentration of 9 μM was used as were the ratios of polyamides
(1, 2, 3, 4, 5, 6, and 8 molar equivalents) titrated into the solution.
The results given in Figure 3 show that both Hx polyamides 2
and 3 bind effectively to their respective cognate sequences
50-AAATTT-30 and 50-ATCGAT-30 as shown by the strong
DNA-induced ligand bands at ∼330 nm along with the appear-
ance of a clear isodichroic point in the CD overlaid spectra. These
results are corroborated by the aforementioned ΔTM values for 2
and 3 and their respective cognate sequences (Table 1). Further-
more, HxIP (3) gave no induced CD band when titrated to its
noncognate sequences 50-AAATTT-30 and 50-ACGCGT-30. A
weak induced CD band was observed for the binding of 3 with
the other noncognate sequence 50-ACATGT-30. Interestingly,
the non-Hx tetramer PPIP 6 gave comparable CD results to its
Hx counterpart 4. The non-Hx tetramer 6 (PPIP) also shows the
weakest induced CD. Consistent with thermal denaturation
studies, the results suggest that the Hx moiety may be binding
with higher affinity than that of two consecutive pyrrole moieties.
On the other hand, HxPP (2) produced induced CD bands when
titrated to the noncognates 50-ACATGT-30 and 50-ATCGAT-30
(Figure 3). However, this is more or less in correlation with its
counterpart PPPP (5), which demonstrates binding to these
noncognate sequences (Figure 3). In contrast to 2 and 3, HxPI
(4) shows a moderate DNA-induced band upon binding to its
cognate 50-ACATGT-30 although a ΔTM value of 19 °C was
observed (Table 1). For the noncognates 50-AAATTT-30,
50-ACGCGT-30, and 50-ATCGAT-30, 4 appears to be binding
in a differential fashion to the non-Hx tetramer PPPI (7) as
3132
dx.doi.org/10.1021/bi102028a |Biochemistry 2011, 50, 3127–3136

Biochemistry
ARTICLE
shown by the two CD spectra marked by an asterisk in Figure 3.
This suggests a mole ratio of ligand/DNA of >33. In the complex,
the induced CD spectra changed drastically indicating a signifi-
cant change in the conformation of the DNA, possibly due to
formation of a higher ordered structure.30,31 Despite this un-
expected binding behavior of 4, the thermal denaturation data
indicate that 4 binds with higher specificity to cognate
50-ACATGT-30 than 7. In summary, comparison of the Hx-
substituted molecules with the control tetramers suggests there is
no loss of binding affinity or specificity upon incorporation of the
Hx moiety into these polyamides, in turn, indicating that these
molecules are reasonable analogues of pyrrole/imidazole poly-
amides. In addition, this data also suggest that the Hx moiety can
act as two consecutive pyrrole (PP) heterocyclic moieties and
f-P units20c,f in accordance with the aforementioned pairing rules.
SPR and ITC Studies. The binding constants (Keq) of the Hx-
containing compounds were determined using a surface plasmon
resonance (SPR) biosensor assay. The studies were conducted
using similar DNA sequences as those used in the DNA melting
and CD titration studies. The sensorgrams recorded from the
SPR experiments for the Hx compounds are given in Figure 4.
According to the sensorgram in Figure 4A, it is evident that Hx-
PP (2) showed preference for 50-AAATTT-30 (Keq = 1 ꢁ
106 Mꢀ1, Table 1) and avoided a G/C-containing DNA sequence,
50-AATGAT-30. It would not even tolerate a single G/C mis-
matched sequence. Similarly, the sensorgram for Hx-IP 3, given in
Figure 4B, demonstrated a preference for its cognate 50-ATC-
GAT-30 (Keq = 4 ꢁ 106 Mꢀ1) and avoided a mismatched
sequence, 50-AATGAT-30. HxPI (4) showed a preference for its
cognate 50-ACATGT-30 (Keq = 7 ꢁ 106 Mꢀ1) as exhibited by the
slow off-rates in the sensorgram (Figure 4C). The binding
constant of compounds 2 and 3 were ascertained from fitting
the binding data to a two-site model and the values are reported in
Table 1. In comparison, the nonfluorescing non-formamido
tetraamides 5ꢀ7 gave lower binding constants in the 105 Mꢀ1
range (Table 1), in agreement with the biophysical data discussed
previously. Likewise, the reported binding constants for forma-
mido-triamides 8ꢀ1020c,f were also lower that their Hx-counter-
parts but higher than the non-formamido tetraamides. This
finding is significant because thus far structural modifications of
the N-terminus formamido group in polyamides have resulted in
lowering affinity. The Hx group stands alone as a heterocyclic
moiety that imparts higher binding affinity to the cognate
sequence compared to the formamido-pyrrole (f-P) unit.
In further thermodynamic studies on the Hx-containing
compounds, polyamides 2ꢀ4 were titrated to their cognate
DNA sequences and monitored by isothermal titration calorim-
etry (ITC) studies at 25 °C (data not shown).25 In comparison,
titration of distamycin to 50-AAATTT-30, done under identical
conditions gave a negative enthalpy, with a ΔH comparable to
the reported value.28 The results provided evidence that the
enthalpy of binding for all three compounds 2ꢀ4 were essentially
zero, suggesting the interactions were driven by entropy. Using
the binding constants determined from SPR studies, the free
energies of binding were calculated from the equation (ΔG =
ꢀRT ln Keq, R = 1.987 cal molꢀ1 Kꢀ1, and T = 298 K). The ΔG
for polyamide 2 for 50-AAATTT-30, 3 for 50-ATCGAT-30, 4 for
50-ACATGT-30 are 8.2, 8.8, and 9.3 kcal molꢀ1
.
Emission Studies. Fluorescence emission studies were per-
formed on 2 and 3 to confirm the pattern of induced fluorescence
of the Hx fluorophore upon binding to dsDNA. 20 μM of DNA
was used in these studies and the ratios of polyamides titrated
were 1, 2, 3, 4, 5, and 6 mol equiv. The Hx-compound was excited
Figure 5. Emission studies on 2 (A) and 3 (B). Aliquots of polyamide
solutions were titrated into the 20 μM solution of DNA in PO40 buffer at
ratios of 1, 2, 3, 4, 5, and 6 mol equiv of polyamides relative to DNA. A
ratio of 0:1 denotes DNA alone. HxPP (2) was titrated to its cognate
sequence 50-AAATTT-30 and HxIP (3) was titrated to its cognate
sequence 50-ATCGAT-30.
Figure 4. Representative SPR sensorgrams for the interaction of (top)
Hx-PP, (middle) Hx-IP, and (bottom) Hx-PI with cognate and non-
cognate DNA sequences. The compound concentrations from bottom
to top are 0ꢀ1 μM in all plots and the experiments were performed as
described in the Experimental Section.
3133
dx.doi.org/10.1021/bi102028a |Biochemistry 2011, 50, 3127–3136

Biochemistry
ARTICLE
at 322 nm, based on results from UVꢀvis spectroscopy of the
Hx-polyamide compound. Figure 5 illustrates that upon binding
to their cognate sequences, both Hx-PP and Hx-IP showed
strong fluorescence with a λmax of 370 nm when bound to
their respective 50-AAATTT-30 and 50-ATCGAT-30 cognate
sequence. This behavior of enhancement of emission is consis-
tent with the binding of Hoechst 33258 itself32 as well as DNA
intercalating agents such as ethidium bromide.33 The enhanced
emission can be attributed to the reduction in interaction
between the photoexcited bound Hx fluorophore and solvent.34
DNase I and Micrococcal Nuclease Footprinting Studies.
Footprinting experiments were performed on 2ꢀ7 using a 131
bp 50-[32P]-radiolabeled DNA fragments containing the follow-
ing sequences: (a) 50-AAATTT-30; (b) 50-TTGCAA-30; (c) 50-
TCTAGA-30 (an alternative cognate for 4); (d) 50-ACTAGT-30
(another alternative cognate for 4); and (e) 50-TCTTGA-30.
Autoradiogram A of a micrococcal nuclease footprinting gel
(Figure 6A) shows that HxPP (2) and PPPP (5) exhibit
sequence specificity for 50-AAATTT-30. Micrococcal nuclease
was used rather than DNase I in this case because of the poor
Figure 6. DNA footprinting using micrococcal nuclease (A) or DNase I (B, C) of compounds 2 and 5 (A), 3 and 6 (B) and 4 (C) using three different
131 bp 50-32P-radiolabeled DNA fragments containing the corresponding cognate sites as denoted on the autoradiograms. DNA is the undigested
control, and G þ A is the sequencing lane.
Figure 7. Molecular electrostatic surfaces of Hx-carboxamide (A) and PP-caboxamide (B), and dipole moments of Hx-carboxamide (C) and
PP-caboxamide (D). The structures depicted as electrostatic surfaces or dipole moment images are identical including the hybridization of each atom in
the molecules.
3134
dx.doi.org/10.1021/bi102028a |Biochemistry 2011, 50, 3127–3136

Biochemistry
ARTICLE
digestion of DNase I at AT-rich sequences. For 2, a footprint
begins to appear at 50-AAATTT-30 at 1 μM, while for 5, a
footprint starts to appear at 5 μM, which is consistent with the
higher binding affinity of HxPP over PPPP observed in the
thermal denaturation studies (Table 1). For HxIP (3) and PPIP
(6), autoradiogram B (Figure 6B) shows the appearance of a
footprint at 50-TTGCAA-30 at 4 μM for 3, while a weak footprint
appears at 6 μM for 6, consistent with results from SPR studies
indicating that HxIP (3) binds slightly more strongly than PPIP
(5) at the 50-TTCGAA-30 site. For HxPI (4), autoradiogram C
(Figure 6C) shows binding at 20 μM at the 50-ACTAGT-30 and
50-TCTTGA-30 cognate sites. The binding to the 50-TCTAGA-30
site occurs at a slightly higher concentration (30ꢀ40 μM)
suggesting the ability of HxPI to discriminate an A/T base pair
in the context of the respective sequences. Overall, these data
indicate that the Hx moiety clearly behaves similarly to two
consecutive pyrrole heterocyclic moieties, and this, in turn, is in
accordance with the thermal denaturation and SPR data.
graphs for HxIP (3) are given in this section. This material is
available free of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
*Tel.: (616) 395-7190; fax: (616) 395-7923; e-mail: lee@hope.
edu.
Funding Sources
The authors thank the NSF [CHE 0809162, CHE 0550992,
CHE 0520704 (MU3C computer cluster) and CHE 0922623
(NMR)], Cancer Research UK (C2259/A9994 to J.A.H.), and
the Georgia Research Alliance for their generous support.
’ ABBREVIATIONS USED
(ΔTM), thermal melts; (CD), circular dichroism; (ITC), isother-
mal titration calorimetry; (SPR), surface plasmon resonance
Molecular Modeling Studies. Electrostatic potential surfaces
of each of the molecules studied and depicted in Figure 7A and B
show distinct electropositive regions where the molecule binds
to the minor groove. Qualitatively, the Hx-carboxamide surfaces
resemble those of PP-carboxamide, and the dipole moments for
both molecules are also in alignment (Figure 7C,D). These
results provide a theoretical framework for supporting the
resemblance of Hx in recognizing two consecutive A/T base
pairs in the same manner as PP. It is worthy to note that the
magnitude of dipole moment for Hx is significantly larger than
that for PP (5.845 versus 2.813 Debye) suggesting a possible
contribution toward the higher observed binding affinity of HxIP
(3) for its cognate sequence over PPIP (6).
’ REFERENCES
(1) Dwyer, T. J., Geierstanger, B. H., Bathini, Y., Lown, J. W., and
Wemmer, D. E. (1992) Design and binding of a distamycin A analogue
to d(CGCAAGTTGGC).d(GCCAACTTGCG): synthesis, NMR stud-
ies, and implications for the design of sequence-specific minor groove
binding oligopeptides. J. Am. Chem. Soc. 15, 5911–5919.
(2) Mrksich, M., Wade, W. S., Dwyer, T. W., Geierstanger, B. H.,
Wemmer, D. E., and Dervan, P. B. (1992) Antiparallel side-by-side
dimeric motif for sequence specific recognition in the minor groove of
DNA by the designed peptide 1-methylimidazole-2-carboximide ne-
tropsin. Proc. Natl. Acad. Sci. U.S.A. 89, 7586–7590.
(3) Pelton, J. G., and Wemmer, D. E. (1989) Structural character-
ization of a 2:1 distamycin A. d(CGCGAAATTGGCC)2 complex by
two dimensional NMR. Proc. Natl. Acad. Sci. U.S.A. 86, 5723–5727.
(4) Kielkopf, C. L., Baird, E. E., Dervan, P. B., and Rees, D. C. (1998)
Structural basis for GC recognition in the DNA minor groove. Nat.
Struct. Biol. 5, 104–109.
’ CONCLUSIONS
The results described in this study indicate that the Hx
heterocyclic moiety is a novel DNA recognition element that
targets A/T base pairs. Hx-polyamide compounds bind DNA in
a highly sequence selective manner, and upon binding they
strongly emit blue light upon excitation at 322 nm. The Hx
moiety when incorporated into polyamides provides a direct
means to measure polyamide binding in the minor groove of
DNA in a test tube or in cells by fluorescence. In addition, the Hx
moiety behaves similarly to a two consecutive pyrrole unit such
that an Hx/PP pairing found in HxPP (2) recognizes two
consecutive A/T base pairs. An Hx/IP pairing present in HxIP
(3) recognizes an A/T.C//G.A/T and an Hx/PI pairing found in
HxPI (4) recognizes C.A/T//A/T.G sequence, respectively.
Overall, the Hx polyamides bind in a similar fashion to the
nonfluorophore counterparts: non-formamido tetraamides 5ꢀ7
and formamido-triamides 8ꢀ10 but with a higher affinity for
their cognate sequences. Given its simple molecular design and
ease of synthesis, the Hx moiety represents a significant advance-
ment in the polyamide field. Sequence targeted Hx polyamides
such as those synthesized in this article provide a direct and
simple way to monitor binding, without the incorporation of a
bulky fluorescent tag. The application of such fluorophores can
be vital in studying the use of these compounds in living systems.
(5) Lown, J. W., Krowicki, K., Bhat, U. G., Skorobaty, A., Ward, B.,
and Dabrowiak, J. C. (1986) Molecular recognition between oligopep-
tides and nucleic acids: novel imidazole-containing oligopeptides related
to netropsin that exhibit altered DNA sequence specificity. Biochemistry
25, 7408–7416.
(6) Yang, XꢀL., Hubbard, R. B., IV, Lee, M., Tao, ZꢀF., Sugiyama,
A., and Wang, H-J. (1999) Imidazoleꢀimidazole pair as a minor groove
recognition motif for T:G mismatched base pairs. Nucleic Acids Res.
27, 4183–4190.
(7) Melander, C., Burnett, J. M., and Gottesfield, J. (2004) Regula-
tion of gene expression with pyrroleꢀimidazole polyamides. J. Biotech-
nol. 112, 195–220.
(8) Edelson, B. S., Best, T. P., Olenyuk, O., Nickols, N. G., Doss,
R. M., Foister, S., Heckel, A., and Dervan, P. B. (2004) Influence of
structural variation on nuclear localization of DNA-binding polyamide-
fluorophore conjugates. Nucleic Acids Res. 32, 2802–2818.
(9) (a) Hsu, C. F., and Dervan, P. B. (2008) Quantitating the
concentration of Py-Im polyamideꢀfluorescein conjugates in live cells.
Bioorg. Med. Chem. Lett. 18, 5851–5855. (b) Nishijima, S., Shinohara, K.,
Bando, T., Minoshima, M., Kashiwazaki, G., and Sugiyama, H. (2010)
Cell permeability of Py-Im-polyamide-fluorescein conjugates: Influence
of molecular size and P/I content. Bioorg. Med. Chem. 18, 978–983. (c)
Crowley, K. S., Phillion, D. P., Woodard, S. S., Schweitzer, B. A., Singh,
M., Shahany, H., Burnette, B., Hippenmeyer, P., Heitmeier, M., and
Bashkin, J. K. (2003) Controlling the intracellular localization of
fluorescent polyamide analogues in cultured cells. Bioorg. Med. Chem.
Lett. 13, 1565–1570. (d) Nicholas, N. G., Jacobs, C. S., Farkas, M. E., and
Dervan, P. B. (2007) Improved nuclear localization of DNA-binding
polyamides. Nucleic Acids Res. 35, 363–370. (e) Jacobs, C. S., and
Dervan, P. B. (2009) Modifications at the C-terminus to improve
’ ASSOCIATED CONTENT
S
Supporting Information. The synthesis and character-
b
ization of compounds 2ꢀ7 as well as representative DNA melt
3135
dx.doi.org/10.1021/bi102028a |Biochemistry 2011, 50, 3127–3136

Biochemistry
ARTICLE
pyrrole-imidazole polyamide activity in cell culture. J. Med. Chem.
52, 7380–7388.
(10) Kapuschinski, J., and Yanagi, K. (1979) Selective staining by
40,6-diamidine-2-phenylindole of nanogram quantities of DNA in the
presence of RNA on gels. Nucleic Acids Res. 6, 3535–3542.
(11) Larsen, T. A., Goodsell, D. S., Cascio, D., Grzeskowiak, K., and
Dickerson, R. E. (1989) The structure of DAPI bound to DNA. J. Biomol.
Struct. Dyn. 7, 477–491.
(12) Trotta, E., and Paci, M. (1998) Solution structure of DAPI
selectively bound in the minor groove of a DNA T.T mismatch-
containing site: NMR and molecular dynamics studies. Nucleic Acids
Res. 26, 4706–4713.
(13) Kim, S. K., Eriksson, S., Kubista, M., and Norden, B. (1993)
Interaction of 40,6-diamidino-2-phenylindole (DAPI) with poly[d(G-C)2]
and poly[d(G-m5C)2]: evidence for major groove binding of a DNA
probe. J. Am. Chem. Soc. 115, 3441–3447.
(14) (a) Wilson, W. D., Tanious, F. A., Barton, H. K., Jones, R. L., H.,
J., Strekowski, L., and Boykin, D. W. (1989) Binding of 40,6-diamidino-2-
phenylindole (DAPI) to GC and mixed sequences in DNA: intercalation
of a classical groove-binding molecule. J. Am. Chem. Soc. 111,
5008–5010. (b) Wilson, W. D., Tanious, F. A., Barton, H. K., Jones,
R. L., Fox, K., Wydra, L., and Strekowski, L. (1990) DNA sequence
dependent binding modes of 40,6-diamidino-2-phenylindole (DAPI).
Biochemistry 29, 8452–8461.
comparative binding properties to AT- and GC-rich DNA sequences.
Eur. J. Org. Chem. 2002, 3604–3615. (b) Thomas, M., Rao, A. R.,
Varshney, U., and Bhattacharya, S. (2001) Unusual DNA binding
exhibited by synthetic distamycin analogues lacking the N-terminal
amide unit under high salt conditions. J. Biomol. Struct. Dyn. 18, 858–871.
(22) Rucker, V. C., Dunn, A. R., Sharma, S., Dervan, P. B., and Gray,
H. B. (2004) Mechanism of sequence-specific fluorescent detection
of DNA by N-methylimidazole, N-methylpyrrole and β-alanine linked
polyamides. J. Phys. Chem. B 108, 7490–7494.
(23) Buchmueller, K. L., Taherbhai, Z., Howard, C. M., Bailey, S. L.,
Nguyen, B., O’Hare, C., Hochhauser, D., Hartley, J. A., Wilson, D. W.,
and Lee, M. (2005) Design of a hairpin polyamide, ZT65B, for targeting
the Inverted CCAAT Box (ICB) site in the multidrug resistant (MDR1)
gene. ChemBiochem 6, 2305–2311.
(24) Lacy, E. R., Nguyen, B., Minh Le, N., Cox, K. K., O’ Hare, C.,
Hartley, J. A., Lee, M., and Wilson, D. (2002) Recognition of T G
3
mismatched base pairs in DNA by stacked imidazole-containing poly-
amides-surface plasmon resonance and circular dichroism studies.
Nucleic Acids Res. 30, 1834–1841.
(25) Buchmueller, K. L., Bailey, S. L., Matthews, D. A., Taherbhai,
Z. T., Register, J. K., Davis, Z. S., Bruce, C. D., O’Hare, C., Hartley, J. A.,
and Lee, M. (2006) Physical and structural basis for the strong
interactions of the ꢀImPy- central pairing motif in the polyamide
f-IPI. Biochemistry 45, 13551–13565.
(15) Fechter, E. J., Olenyuk, B., and Dervan, P. B. (2005) Sequence-
specific fluorescence detection of polyamide-thiazole orange conjugates.
J. Am. Chem. Soc. 127, 16685–16691.
(26) Westrate, L., Mackay, H., Brown, T., Nguyen, B., Kluza, J.,
Wilson, W. D., Lee, M., and Hartley, J. A. (2009) Effects of the
N-terminal acylamido group of imidazole- and pyrrole-containing poly-
amides on DNA sequence specificity and binding affinity. Biochemistry
48, 5679–5688.
(27) (a) Frisch, M. J. et al. Gaussian 09, Revision A.1, (2009),
Gaussian, Inc., Wallingford CT, USA. (b) Schmidt, J. R., Polik, W. F.
WebMO Enterprise, version 10.0. WebMO LLC: Holland, MI, USA.
Available from http://www.webmo.net (accessed October 2010).
(28) Bathini, Y., and Lown, J. W. (1990) Convenient routes to
substituted benzimidazoles and imidazolo[4,5-b]pyridines using nitro-
benzene as oxidant. Synth. Commun. 20, 955–963.
(16) Best, T. P., Edelson, B. S., Nickols, N. G., and Dervan, P. B.
(2003) Nuclear localization of pyrroleꢀimidazolepolyamideꢀfluorescein
conjugates in cell culture. Proc. Natl. Acad. Sci. U.S.A. 100, 12063–12068.
(17) Belitski, J. M., Leslie, S. J., Arora, P. S., Beerman, T. A., and
Dervan, P. B. (2002) Cellular uptake of N-methylpyrrole/N-methyli-
midazole polyamide-dye conjugates. Bioorg. Med. Chem. 10, 3313–3318.
(18) Chenoweth, D. M., Viger, A., and Derven., P. B. (2007)
Fluorescent sequence-specific dsDNA binding oligomers. J. Am. Chem.
Soc. 129, 2216–2217.
(19) White, C. M., Satz, A. M., Bruice, T. C., and Beerman, T. A.
(2001) Inhibition of transcription factor-DNA and gene expression by
a microgenotropen. Proc. Natl. Acad. Sci. U.S.A. 98, 10590–10595.
(20) (a) Liu, Y., and Wilson, W. D. (2010) Quantitative analysis of
small moleculeꢀnucleic acid interactions with biosensor surface and
surface plasmon resonance detection. Methods Mol. Biol. 613, 1–23. (b)
Lacy, E. R., Minh Le, N., Price, C. A., Lee, M., and Wilson., W. D. (2002)
Influence of a terminal formamido group on the sequence recognition of
DNA by polyamides. J. Am. Chem. Soc. 124, 2153–2163. (c) Minh Le, N.,
Sielaff, A., Cooper, A. J., Mackay, H., Brown, T., Kotecha, M., O’Hare.,
C., Hochauser, D., Lee, M., and Hartley, J. A. (2006) Binding of f-PIP,
a pyrrole- and imidazole-containing triamide, to the inverted CCAAT
box-2 of the topoisomerase IIR promoter and modulation of gene
expression in cells. Bioorg. Med. Chem. Lett. 16, 6161–6164. (d)
Buchmueller, K. L., Staples, A. M., Uthe, P. B., Howard, C. M., Pacheco,
K. A. O., Cox, K. K., Henry, J. A., Bailey, S. L., Horick, S. M., Nguyen, B.,
Wilson, D. W., and Lee, M. (2005) Molecular recognition of DNA base
pairs by the formamido/pyrrole and formamido/imidazole pairings
in stacked polyamides. Nucleic Acids Res. 33, 912–921. (e) Brown, T.,
Taherbhai, Z., Sexton, J., Sutterfield, A., Turlington, M., Jones, J.,
Stallings, L., Stewart, M., Buchmueller, K., Mackay, H., O’Hare, C.,
Kluza, J., Nguyen, B., Wilson, D. W., Lee, M., and Hartley, J. A. (2007)
Synthesis and biophysical evaluation of minor-groove C-terminus
modified pyrrole and imidazole triamide analogs of distamycin. Bioorg.
Med. Chem. 15, 474–483. (f) Buchmueller, K., Staples, A. M., Uthe, P. B,
Howard, C. M., Pacheco, K. A. O., Cox, K. K., Henry, J. A., Bailey, S. L.,
Horick, S. M., Nguyen, B., Wilson, D. W., and Lee, M. (2005) Extending
the language of DNA molecular recognition by polyamides: unexpected
influence of pyrrole and imidazole arrangement on binding affinity and
specificity. J. Am. Chem. Soc. 127, 742–750.
(29) Yields for tri- and diamides reported previously in the authors’
laboratory have typically ranged from 20ꢀ40% despite the development
of specific handling techniques in an attempt to minimize the losses at
each stage in the construction of each polyamide Brown, T., Mackay, H.,
Turlington, M., Sutterfield, A., Smith, T., Sielaff, A., Westrate, L., Bruce,
C., Kluza, J., O’Hare, C., Nguyen, B., Wilson, D. W., Hartley, J. A., and
Lee, M. (2008) Modifying the N-terminus of polyamides: PyImPyIm
has improved sequence specificity over f-ImPyIm. Bioorg. Med. Chem.
16, 5266–5276.
(30) Lee, M., Rhodes, A. L., Wyatt, M. D., D'Incalci, M., Forrow, S.,
and Hartley, J. A. (1993) In vitro cytotoxicity of GC sequence directed
alkylating agents related to distamycin. J. Med. Chem. 36, 863–870.
(31) Lee, M., Preti, C. S., Vinson, E., Wyatt, M. D., and Hartley, J. A.
(1994) GC sequence specific recognition by an N-formamido,
C-terminus-modified and imidazole-containing analog of netropsin.
J. Med. Chem. 37, 4073–4075.
(32) (a) Loontiens, F. G., Regenfuss, P., Zeche1, A., Dumortier, L.,
and Clegg, R. M. (1990) Binding characteristics of Hoechst 33258 with
calf thymus DNA, Poly[ d(A-T)], and d(CCGGAATTCCGG): multi-
ple stoichiometries and determination of tight binding with a wide
spectrum of site affinities. Biochemistry 29, 9029–9039. (b) Loontiens,
F. G., McLaughlin, L. W., Diekmann, S., and Clegg, R. M. (1991)
Binding of Hoechst 33258 and 40,6-diamidino-2-phenylindole to self-
complementary decadeoxynucleotides with modified exocyclic base
substituent. Biochemistry 30, 182–189.
(33) Olmsted, J., and Kearns, D. R. (1977) Mechanism of ethidium
bromide fluorescence enhancement on Binding to nucleic acids. Bio-
chemistry 16, 3647–3654.
(34) Barawkar, D. A., and Ganesh, K. N. (1995) Fluorescent
d(CGCGAATTCGCG): characterization of major groove polarity
and study of minor groove interactions through a major groove
semantophore conjugate. Nucleic Acids Res. 23, 159–164.
(21) (a) Thomas, M., Varshney, U., and Bhattacharya, S. (2002)
Distamycin analogues without leading amide at their N-termini-
3136
dx.doi.org/10.1021/bi102028a |Biochemistry 2011, 50, 3127–3136