Sequence specific and high affinity recognition of 5′-ACGCGT-3′ by rationally designed pyrrole-imidazole H-pin polyamides: Thermodynamic and structural studies

Article abstract of DOI:10.1016/j.bmc.2008.09.034

Imidazole (Im) and Pyrrole (Py)-containing polyamides that can form stacked dimers can be programmed to target specific sequences in the minor groove of DNA and control gene expression. Even though various designs of polyamides have been thoroughly investigated for DNA sequence recognition, the use of H-pin polyamides (covalently cross-linked polyamides) has not received as much attention. Therefore, experiments were designed to systematically investigate the DNA recognition properties of two symmetrical H-pin polyamides composed of PyImPyIm (5) or f-ImPyIm (3e, f = formamido) tethered with an ethylene glycol linker. These compounds were created to recognize the cognate 5′-ACGCGT-3′ through an overlapped and staggered binding motif, respectively. Results from DNaseI footprinting, thermal denaturation, circular dichroism, surface plasmon resonance and isothermal titration microcalorimetry studies demonstrated that both H-pin polyamides bound with higher affinity than their respective monomers. The binding affinity of formamido-containing H-pin 3e was more than a hundred times greater than that for the tetraamide H-pin 5, demonstrating the importance of having a formamido group and the staggered motif in enhancing affinity. However, compared to H-pin 3e, tetraamide H-pin 5 demonstrated superior binding preference for the cognate sequence over its non-cognates, ACCGGT and AAATTT. Data from SPR experiments yielded binding constants of 1.6 × 10⁸ M^-1 and 2.0 × 10¹⁰ M^-1 for PyImPyIm H-pin 5 and f-ImPyIm H-pin 3e, respectively. Both H-pins bound with significantly higher affinity (ca. 100-fold) than their corresponding unlinked PyImPyIm 4 and f-ImPyIm 2 counterparts. ITC analyses revealed modest enthalpies of reactions at 298 K (ΔH of -3.3 and -1.0 kcal mol^-1 for 5 and 3e, respectively), indicating these were entropic-driven interactions. The heat capacities (ΔC_p) were determined to be -116 and -499 cal mol^-1 K^-1, respectively. These results are in general agreement with ΔC_p values determined from changes in the solvent accessible surface areas using complexes of the H-pins bound to (5′-CCACGCGTGG)₂. According to the models, the H-pins fit snugly in the minor groove and the linker comfortably holds both polyamide portions in place, with the oxygen atoms pointing into the solvent. In summary, the H-pin polyamide provides an important molecular design motif for the discovery of future generations of programmable small molecules capable of binding to target DNA sequences with high affinity and selectivity.

Full text of DOI:10.1016/j.bmc.2008.09.034

Bioorganic & Medicinal Chemistry 16 (2008) 9145–9153
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
0
0
Sequence specific and high affinity recognition of 5 -ACGCGT-3
by rationally designed pyrrole-imidazole H-pin polyamides:
Thermodynamic and structural studies
a
a
b
b
b
b
Hilary Mackay , Toni Brown , Peter B. Uthe , Laura Westrate , Alan Sielaff , Justin Jones ,
a
c
c
d
e
e
James P. Lajiness , Jerome Kluza , Caroline O’Hare , Binh Nguyen , Zach Davis , Chrystal Bruce ,
W. David Wilson , John A. Hartley , Moses Lee
d
c
a,b,
*
a
Department of Chemistry, Hope College, 35 E. 12th Street, P.O. Box 9000, Holland, MI 49422, USA
Department of Chemistry, Furman University, Greenville, SC 29613, USA
Cancer Research UK Drug-DNA Interactions Research Group, UCL Cancer Institute, Paul O’Gorman Building, 72 Huntley Street, London WC1E 6BT, USA
Department of Chemistry, Georgia State University, Atlanta, GA 30303, USA
b
c
d
e
Department of Chemistry, Erskine College, Due West, SC 29639, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 9 May 2008
Revised 29 August 2008
Accepted 10 September 2008
Available online 13 September 2008
Imidazole (Im) and Pyrrole (Py)-containing polyamides that can form stacked dimers can be programmed
to target specific sequences in the minor groove of DNA and control gene expression. Even though various
designs of polyamides have been thoroughly investigated for DNA sequence recognition, the use of H-pin
polyamides (covalently cross-linked polyamides) has not received as much attention. Therefore, experi-
ments were designed to systematically investigate the DNA recognition properties of two symmetrical H-
pin polyamides composed of PyImPyIm (5) or f-ImPyIm (3e, f = formamido) tethered with an ethylene
glycol linker. These compounds were created to recognize the cognate 5 -ACGCGT-3 through an over-
lapped and staggered binding motif, respectively. Results from DNaseI footprinting, thermal denaturation,
circular dichroism, surface plasmon resonance and isothermal titration microcalorimetry studies demon-
strated that both H-pin polyamides bound with higher affinity than their respective monomers. The bind-
ing affinity of formamido-containing H-pin 3e was more than a hundred times greater than that for the
tetraamide H-pin 5, demonstrating the importance of having a formamido group and the staggered motif
in enhancing affinity. However, compared to H-pin 3e, tetraamide H-pin 5 demonstrated superior bind-
ing preference for the cognate sequence over its non-cognates, ACCGGT and AAATTT. Data from SPR
Keywords:
Binding
Biophysical
DNA
0
0
H-pin
Sequence selectivity
Polyamide
SPR
ITC
8
ꢁ1
10
ꢁ1
experiments yielded binding constants of 1.6 ꢀ 10 M and 2.0 ꢀ 10
M
for PyImPyIm H-pin 5 and
f-ImPyIm H-pin 3e, respectively. Both H-pins bound with significantly higher affinity (ca. 100-fold) than
their corresponding unlinked PyImPyIm 4 and f-ImPyIm 2 counterparts. ITC analyses revealed modest
ꢁ1
enthalpies of reactions at 298 K (
D
H of ꢁ3.3 and ꢁ1.0 kcal mol for 5 and 3e, respectively), indicating
these were entropic-driven interactions. The heat capacities (
D
C
p
) were determined to be ꢁ116 and
ꢁ1
ꢁ1
ꢁ
499 cal mol
K
, respectively. These results are in general agreement with
from changes in the solvent accessible surface areas using complexes of the H-pins bound to (5 -
D
C
p
values determined
0
2
CCACGCGTGG) . According to the models, the H-pins fit snugly in the minor groove and the linker com-
fortably holds both polyamide portions in place, with the oxygen atoms pointing into the solvent. In sum-
mary, the H-pin polyamide provides an important molecular design motif for the discovery of future
generations of programmable small molecules capable of binding to target DNA sequences with high
affinity and selectivity.
Ó 2008 Elsevier Ltd. All rights reserved.
1
. Introduction
search. Such compounds have the potential to be used as gene con-
trol agents by inhibition of native transcription factors at the target
1
The development of polyamide analogs of distamycin (1, Fig. 1)
that can target specific sequences of DNA is an area of active re-
site. Monomer polyamides, which bind DNA in a 2:1 (ligand:DNA)
motif, have been shown to successfully target the desired sequen-
1
c,2
ce.
A concern in this strategy is that the monomers may slip into
extended staggered motifs, thus, creating a reading frame different
to that intended. The monomer may also bind in a mixed 2:1 and
*
Corresponding author. Tel.: +1 616 395 7190; fax: +1 616 395 7923.
E-mail address: lee@hope.edu (M. Lee).
3
0
968-0896/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2008.09.034

9
146
H. Mackay et al. / Bioorg. Med. Chem. 16 (2008) 9145–9153
H C
3
H N
2
N
CH3
H
NH2
NH
N
O
HN
HN
N
H C
O
H
N
O
3
NH
N
HN
O
O
N
N
H
N
HN
N
H C
CH3
3
CH3
O
N
N
O
CH3
H C
O
N
3
H3C
CH3
Distamycin, 1
N
CH3
PyImPyIm, 4
H
HN
H C
3
NH
N
O
N
N
H3C
O
CH3
N
H
N
NH
HN
N
O
CH3
N
HN
N
O
H C
O
N
H
N
3
N
N
H C
O
CH3
HN
H C
3
3
N
O
f-ImPyIm, 2
N
CH3
H
CH3
N
O
NH
O
HN
N
O
H
N
N
O
N
H C
HN
N
3
CH3
O
O
N
R
O
O
N
N
O
O
CH3
O
CH3
N
H C
3
H C
N
N
3
NH
N
H
N
NH
N
H
O
N
N
N
N
O
O
O
5
NH
HN
HN
NH 3a, R =
(CH2)6
CH3
H
N
3
3
3
3
b
c
d
e
(CH2)7
(CH2)8
(CH2)9
(CH2)2
H3C
N
H C
3
N
CH3
CH3
O
(CH ) O (CH )
2 2 2 2
Figure 1. Structure of polyamides: distamycin 1, f-ImPyIm 2, f-ImPyIm H-pins 3a-e, PyImPyIm 4, and the polyethylene glycol PyImPyIm H-pin 5.
1
:1 fashion, thereby reducing the overall affinity. To overcome
+
f
+
these issues, many strategies have been employed to tether both
monomers together, such as the ‘hairpin’ design which links the
+
+
f
monomers ‘head’-to-‘tail’ using a flexible
c
-butyric acid moi-
Hairpin polyamides have demonstrated excellent binding
affinities and their binding characteristics have been extensively
1
c,2,4
ety.
C-Terminus =
Imidazole
+
Formamido = f
1
,2
reviewed.
An alternative motif is the ‘H-pin’ which links the
two monomers via the N1-nitrogen of the central heterocyclic
units. To date H-pins linked via alkyl chains of varying lengths
=
Pyrrole
=
5
(
C3–C12) have been investigated as duplex DNA binders that inter-
Figure 2. Diagram to show the ‘overlapped’ (A), and ‘staggered’ (B), H-pin motifs of
polyamide minor-groove binding to duplex DNA. The wiggly line connecting the
two polyamide units represents the linker.
act with the minor-groove in the overlapped motif (the two polya-
mides units are stacked directly over each other and the linker
directly joins two heterocycles across the polyamides, Fig. 2A).
5
,6
However, weak binding affinities were reported and the com-
pounds were not tested against DNAs that enable a different bind-
ing motif.
The authors’ laboratory has previously reported that the poly-
amide monomer formamido-imidazole-pyrrole-imidazole (f-Im-
PyIm, 2, Fig. 1) binds as a dimer in a staggered binding motif
0
3 sequence is of significant interest because it is present in the
2
,4–6
core sequence of the MluI cell-cycle box (MCB) transcriptional ele-
ment found in the promoter of the human Dbf4 (huDbf4 or ASK ,
activator of S-phase kinase) gene. Dbf4 is the regulatory subunit
of Cdc7 (cyclin dependent 7) kinase, and high levels of this kinase
8
(
(
Fig. 2B), with exceptional affinity, to its cognate DNA sequence
5 -ACGCGT-3 ). In this motif, the polyamides units are stacked
have been implicated for development of various cancers. As part
0
0
7
of our strategy to develop MCB-targeted polyamides, the synthesis
of a series of H-pin polyamides composed of two f-ImPyIm mono-
mers tethered via the central pyrrole using either a hydrocarbon
in an off-centered or staggered fashion and the linker joins two
0
adjacent heterocycles in a diagonal arrangement. The 5 -ACGCGT-

H. Mackay et al. / Bioorg. Med. Chem. 16 (2008) 9145–9153
9147
chain of length C6–C9 or an ethylene glycol linker (3a–e, Fig. 1)
was carried out. These studies were designed to discover a suit-
CGCG-non-hairpin: 5 -C C A C G C G T G G-3
9
,
,
able covalent linker on the H-pins that would enable the com-
pounds to bind tightly and specifically to the cognate sequence
through the staggered binding motif. It was determined via DNaseI
footprinting, and analysis of molecular models, that the C8 and
C9 hydrocarbon linked H-pins yielded the optimal length to allow
binding in a staggered fashion. However, the alkyl linked com-
pounds readily formed aggregates in aqueous solutions. The ethyl-
ene glycol linked H-pin 3e overcame the solubility obstacle and
was found to bind with exceptionally high affinity to the cognate
3 -G G T G C G C A C C-5
,
ACGCGT:
ACCGGT:
AAATTT:
5 - G A A C G C G T C G
C
T
,
C
3
- C T T G C G C A G C T
,
5 - G A A C C G G T C G C
T
1
0
ꢁ1
,
sequence (K = 2 ꢀ 10
M
by SPR). H-pin 3e interacts with DNA
C
3
- C T T G G C C A G CT
in a similar manner to that of its monomer counterpart 2, suggest-
ing that the linker does not affect the DNA sequence recognition
9
process. The outstanding binding affinity of 3e was somewhat
,
5 - C G A A A T T T C C
C
compromised by the reduced selectivity of this compound in com-
T
9
parison to monomer 2. It must be noted that the Lown group had
3 - G C T T T A A A G G C
T
conducted some experiments on an H-pin related to 3c; however,
the H-pin was only tested against DNA through the overlapped mo-
Figure 3. The DNA sequences used in this study.
5
tif. The possibility that using such linkers would allow enough
flexibility to bind in a staggered binding motif was not anticipated
2
,4–6
nor investigated.
hydrolysis of intermediate 9. H-pin 3e was formed in 22% yield by
following a similar strategy from the nitro-containing compound
10 with 4-formamido-1-methylimidazole-2-carboxylic acid 11.9
In an effort to design compounds with improved sequence spec-
ificity and to expand the repertoire of polyamide–DNA binding de-
signs, the authors’ laboratory reported the binding characteristics
of a monomer tetraamide, PyImPyIm (4, Fig. 1). The non-form-
amido compound 4 was anticipated to bind in an overlapped fash-
ion, extending the core reading frame from two heterocycles (as
with f-ImPyIm 2) to four. Thus, compound 4 was expected to be
more selective than 2 for the cognate sequence (5 -ACGCGT-3 ).
This hypothesis was verified by DNaseI footprinting which demon-
strated no other footprints at non-cognate sites up to concentra-
,11
1
0
2.2. DNaseI footprinting
To investigate the sequence selectivity of monomer 4 and H-pin
5, DNaseI footprinting experiments were performed using a 130 bp
0
0
0
32
5 -[ P]-radiolabeled DNA fragment containing the following se-
0
0
0
0
0
0
quences 5 -ACGCGT-3 (a); 5 -ACCGGT-3 (b); 5 -ACGTGT-3 (c);
0
0
5 -AGCGCT-3 (d). The autoradiogram given in Figure 4 shows that
0
0
tions as high as 50
determined from SPR studies, was reduced by about 23-fold in
l
M. However, the binding affinity of 4,
a footprint appears at the 5 -ACGCGT-3 cognate site from 0.5
for monomer 4 and 0.1 M for H-pin 5. The only other protection
site observed with both compounds was at 5 -ACGTGT-3 . Here
binding occurred from 5 M, with H-pin 5 showing stronger bind-
ing than monomer 4 at this site. Some affinity for 5 -ACGTGT-3
was expected in view of its high homology with the 5 -ACGCGT-
lM
l
8
ꢁ1
0
0
comparison to its f-containing counterpart 2 (1.9 ꢀ 10 M
,
SPR).1 Thus, the gain in selectivity was at the expense of binding
0
l
0
0
affinity.
0
The above studies have demonstrated the fine balance between
high binding affinity and high selectivity. The present study aims
to address this balance by combining the high selectivity of tetraa-
mide monomers with the outstanding binding affinity of the H-pin
motif. The target compound 5 (Fig. 1) consists of two PyImPyIm (4)
monomers tethered via the central Py heterocycles with an ethyl-
ene glycol linker joining two pyrrole moieties in a diagonal
arrangement (Fig. 2A). This structural design is different from the
0
3 . These results demonstrated that linkage of the two monomers
did not affect selectivity and increased affinity for the cognate site.
2.3. Thermal denaturation
Thermal denaturation analyses were performed to support the
0
selectivity of PyImPyIm H-pin 5 for its cognate DNA sequence (5 -
0
‘
directly joined’ H-pins, as in Figure 2A, reported previously by
Lown’s and Dervan’s groups. The ethylene glycol linker will aid
in the dissolution of the final molecule, as shown with H-pin 3e,
ACGCGT-3 ). The data shown in Table 1 indicate that 5 has incred-
5
6
m
ible affinity for the cognate sequence with a DT of 16 °C. The
two non-cognate sequences ACCGGT and AAATTT show no melt
9
and will allow each two polyamides to stack in the overlapped motif
for recognition of the cognate sequence 5 -ACGCGT-3 . Accordingly,
whatsoever, strongly corroborating the footprinting results. This
0
0
m
DT value is less than that previously reported for H-pin 3e
9
the synthesis of H-pins 3e and 5 as well as their DNA binding prop-
which showed a value >30 °C, an anticipated result due to the
absence of the formamido groups. For comparison, the binding
affinity of distamycin to its cognate AAATTT was found to be
14 °C.
9
erties, minus those previously reported for 3e, are reported here-
in. The cognate and non-cognate oligonucleotides utilized in the
biophysical chemistry studies are given in Figure 3.
2
.4. Circular dichroism
2
2
. Results and discussion
Circular dichroism studies were carried out to provide evidence
.1. Synthesis
for the binding mode and to further probe selectivity for the target
sequence. H-pin 5 was found to bind in the minor groove of its cog-
nate sequence ACGCGT, as indicated by the strong induced CD
H-pin 5 was formed in 28% yield by the reaction of the central
1
0
core 6a and PyIm carboxylic acid 7 using a carbodiimide coupling
procedure with EDCI (Scheme 1). PyIm-acid 7 was obtained by
reacting pyrrole-2-carbonyl-chloride10 with intermediate 8b
formed by reduction of 8a using Pd-C catalyzed hydrogenation)
in a Schotten-Bauman coupling. The acid 7 was obtained after
7
,12
band at 335 nm (Fig. 5A).
The appearance of an isodichroic point
at ꢂ315 nm provided evidence that the H-pin binds to the oligonu-
cleotide via a single mechanism, presumably interacting in the
minor groove as a covalently tethered and stacked dimer.
(

9
148
H. Mackay et al. / Bioorg. Med. Chem. 16 (2008) 9145–9153
H C
3
N
CH
3
i
HN
N
O
CH3
NH2 O N
2
N
N
O
ii
vi
5
O
3e
O
H
NH
1
1
OH
N
O
N
H C
O
O
N
3
H C
3
N
NH
NO
2
NH2
O
N
6
NH
i
H C
3
N
CH3
CH3
8
8
a
b
= NO2
= NH2
R
O
OH
O
H
N
H
N
N
N
N
CH3
v
iv
iii
O
O
N
N
N
N
N
O
O
O
CH3
CH3
CH3
CH3
CH
3
9
7
Scheme 1. Reagents and conditions (i) 10% Pd/C, cold MeOH, H
2
, 55 PSI, 18 h, RT; (ii) 7, EDCI, DMAP, dry DMF, 6 days, RT; (iii) 10% Pd/C, cold MeOH, H
O, 24 h, reflux; (vi) 11, EDCI, DMAP, dry DMF, 50 °C.
2
, 18 h, RT; (iv) pyrrole-
2
-carbonyl-chloride, dry DCM, dry TEA, 18 h, 0 °C-RT; (v) NaOH (2 M), H
2
By contrast, titration of H-pin 5 to the non-cognate ACCGGT and
AAATTT sequences did not produce an appreciable induced band
at concentrations up to 400 nM. The SPR binding profile of PyIm-
PyIm H-pin is identical to that of the monomer PyImPyIm 4,10 sup-
porting earlier evidence that the linker played a minor role in DNA
7
(
Fig. 5B and C). Consistent with the binding of f-ImPyIm (2), PyIm-
1
0
9
9
PyIm (4) and H-pins (3a–e) to ACGCGT, absence of a defined in-
duced band or an isodichroic point often suggests that binding of
the polyamide to these oligonucleotides is weak, and non-selective
at best. It is worthy to note that an alternative binding mode begins
to emerge on addition of 5 to the non-cognate sequences as evi-
denced by shifting of the isodichroic point to the right as the con-
centration of ligand is increased. Overall, the CD results
corroborate both the footprinting and thermal denaturation re-
sults; that H-pin 5 binds selectively to the cognate sequence.
CD experiments were also conducted using H-pin 3e and
ACGCGT and a non-cognate ACCGGT with varying concentrations
of added salt (12.5, 25, 50, 75, 100, 200, and 300 mM). In all cases
salt concentration had virtually no effect on the signal intensity
and shape of the induced band for the DNA/3e complex (data not
shown), indicating insignificant changes in the conformation of
the complex. This suggests that the electrostatic effects on the
binding of compound 3e and the DNA appear as secondary and that
the binding is mainly driven by non-electrostatic factors similar to
that observed for distamycin.13
sequence selectivity. The linker was crucial, however, in enhanc-
ing the binding affinity. A binding constant (K) for PyImPyIm H-
8
ꢁ1
pin 5 for ACGCGT was determined to be 1.6 ꢀ 10 M and the data
best fit a 1:1 stoichiometry. Even though H-pin 5 bound with about
100-fold lower affinity than the f-ImPyIm H-pin 3e to the same se-
1
0
ꢁ1
6
quence (K = 2.0 ꢀ 10
M
), it represented a 25-fold greater bind-
ꢁ1
ing constant (K = 7.1 ꢀ 10 M ) than the PyImPyIm monomer 4.
The results are summarized in Table 1. Judging from the sensor-
gram for the binding of PyImPyIm H-pin 5 to ACGCGT (Fig. 6)
9
and H-pin 3e with the same DNA sequence, it is evident that
the enhanced binding affinity of the H-pins over their monomer
counterparts is largely driven by a slow dissociation rate of the li-
gand from DNA. Achieving this result is beneficial from a biological
perspective because compounds with this DNA binding property
are generally more effective in interfering the binding of pro-
teins/transcriptional factors to promoter sites over compounds
1
4
with rapid off-rates.
2.6. Isothermal titration calorimetry
2
.5. Surface plasmon resonance
ITC experiments were performed at 25 °C to probe the thermody-
namics of the H-pins binding to DNA. The H-pins again demon-
strated binding to the cognate sequence as evidenced by the
exothermic enthalpy of binding of H-pin 5 to ACGCGT (Fig. 7A);
the same result observed with f-ImPyIm (2) and f-ImPyIm H-pin
3e (Fig. 7B). No heat change was observed to either ACCGGT or
AAATTT. The ITC experiment was repeated at elevated temperatures
p
(30to 50 °C) in order to discernthe heatcapacity(DC ) of thebinding
To obtain a more accurate measure of binding affinity, selectiv-
ity and to probe the stoichiometry of binding, SPR-biosensor exper-
iments were performed. From the sensorgrams in Figure 6, it can
be observed that H-pin 5 (Fig. 6C) retains excellent selectivity for
the cognate sequence. No binding to either non-cognate AAATTT
or ACCGGT sequence (Fig. 6A and B, respectively) was observed
9

H. Mackay et al. / Bioorg. Med. Chem. 16 (2008) 9145–9153
9149
observed for PyImPyIm 4 and f-ImPyIm H-pin 3e (ꢁ116 and
ꢁ
1
ꢁ1
ꢁ
499 cal mol
K , respectively).
Using the binding constant of H-pin
5
for ACGCGT of
H value of
3.3 kcal mol from ITC, the DG and TDS at 298 K were calculated
8
ꢁ1
1
ꢁ
.6 ꢀ 10 M (determined by SPR at 25 °C), and the
D
ꢁ
1
ꢁ
1
to be ꢁ11.2 and 7.9 kcal mol , respectively (Table 1). This finding
is consistent with the binding of f-ImPyIm H-Pin 3e to ACGCGT in
which the binding event is also mainly driven by entropy
ꢁ1
ꢁ1
(T
D
S = 13.0 kcal mol
,
D
H = ꢁ1.0 kcal mol ). The results could
also be explained by considering the attribution and additivity of
binding energies of the interaction of small molecules to large bio-
molecules. In this instance, A–B is a ligand containing two iden-
tical units of A and B covalently linked together. The relationship
1
5
between binding free energy of A–B with DNA, A–B would there-
D
G
fore be the sum of , and , in which DG is the ‘coop-
D
G
A
,
D
G
B
D
G
C
C
erative free energy.’ For the linked A–B, the free energy gain
should be favorable, since the lower entropic cost is less (compar-
ing with unlinked ligands).
2
.7. Molecular modeling
Based on the aforementioned results on the binding of PyIm-
0
0
PyIm H-pin 5 in the minor groove of 5 -ACGCGT-3 , in a similar
manner as that reported for the stacked dimer formed from un-
1
0
linked PyImPyIm 4, a representation of the complex is depicted
in Figure 8A. A computer generated model of the complex of
PyImPyIm H-pin 5 bound to a self-complementary duplex (5 -
0
0
dCCACGCGTGG-3 )
2
was subsequently produced and shown in
Figure 8B. The complex was generated using SYBYL and the
model was structurally optimized. The model provides several
insights on the complex formation. In common with the DNA
9
binding properties of f-ImPyIm H-pin 3e, the cross-linked and
stacked polyamide 5 fits snugly within the minor groove and
forms specific contacts with the floor and wall of the groove.
The ethylene glycol linker provides sufficient length in order
for the polyamides to stack in a staggered manner. Finally, the
oxygen atoms in the linker point out from the complex to max-
imize interactions with water molecules, thereby improving sol-
ubility and binding affinity.
0
32
Figure 4. DNase I footprinting of monomer 4 and H-pin 5 using a 130 bp 5 -[ P]-
0
0
0
0
0
0
radiolabeled DNA fragment showing the sites 5 -ACGCGT-3 (a), 5 -ACCGGT-3 (b),
With the models of H-pin 5 and 3e bound to 5 -ACGCGT-3 , the
heat capacity ( ) arising from changes in hydrophobicity could
be estimated using solvent accessible surface area (SASA) calcula-
0
0
0
0
5
-ACGTGT-3 (c), 5 -AGCGCT-3 (d). DNA is undigested control and G+A is the
sequencing lane. Thin boxes indicate the DNaseI cleavage protection for the cognate
site.
DC
p
1
0
tions as previously reported. The calculated values of
p
DC for
H-pins 5 and 3e were determined using Eqs. (2)–(4). For H-pin 5,
Table 1
the results obtained from these calculations were ꢁ481 ± 85,
(kcal mol 1
ꢁ
)
and SPR derived binding constants,
K
eq (M
ꢁ1
)
for
D
T
m
(°C),
D
H
ꢁ1 ꢁ1
c
ꢁ649 ± 47, ꢁ595 ± 78 cal mol
K . Similarly, for H-pin 3e, the val-
compounds 2, 3e, 4, and 5 with the three DNA sequences
ꢁ
1
ꢁ1
ues were ꢁ369 ± 82, ꢁ480 ± 47, ꢁ470 ± 85 cal mol
K , respec-
ACGCGT
ACCGGT
AAATTT
tively. The calculated values are generally close to the
experimental values providing support for the structure and con-
formation of the complexes of PIPI H-pin 5 and f-IPI H-pin 3e
and ACGCGT. The SASA calculated data of the H-pins with
CCACGCGTGG are given in Figure 8C and D, respectively.
PyImPyIm H-pin (5)
D
D
2
T
m
(°C)
16.0 ± 0.3
0
0
—
H (kcal molꢁ1) at
ꢁ3.3
—
5 °C
K (Mꢁ1)a
8
ndb
1.6 ꢀ 10
nd
PyImPyIm (4)
D
D
K
T
m
3.0
0.1
—
nd
0.1
—
nd
(
ref.10)
H
ꢁ3.2
6
7.1 ꢀ 10
3
. Conclusions
f-ImPyIm (2)
ref.7c)
D
D
K
T
m
7.8
7.6
1.9 ꢀ 10
1.1
—
2.2 ꢀ 10
0.9
(
H
—
8
5
4
A successful strategy for systematically developing polyamides
5.3 ꢀ 10
that exhibit an excellent balance between sequence selectivity and
binding affinity is reported herein. This strategy utilizes the H-pin
molecular design that combines the high selectivity of tetraamide
monomers with the strong binding affinity of the H-pin motif. As a
result, we have produced molecules, e.g., H-pins 3e and 5, which
showed excellent affinity and sequence specificity. The develop-
ment of the next generation of H-pins is currently underway and
the goal is to produce molecules with even higher affinity while
simultaneously addressing cellular uptake/nuclear localization is-
f-ImPyIm H-pin
3e)
D
D
K
T
H
m
> 30
14.1 ± 0.3 9.3 ± 0.3
(
ꢁ1.0
—
—
5 ꢀ 10
1
0
8
5
2.0 ꢀ 10
1.1 ꢀ 10
a
b
c
The uncertainty is estimated to be 10%.7a
nd, not determined because no response units (RU) observed.
The detection limit for SPR in our hands is about 1 ꢀ 10
2
ꢁ1
M .
reaction. FromFigure 7C,
from the slope of the line to be ꢁ116 cal mol
p
DC for PyImPyIm H-pin 5 was determined
ꢁ1
ꢁ1
K
, compared to that

9
150
H. Mackay et al. / Bioorg. Med. Chem. 16 (2008) 9145–9153
7
6
4
3
1
5
0
5
0
5
0
1
00
A
B
8
6
4
0
0
0
20
0
-
-
15
30
100
80
2
50
50
50
275
275
275
300
325
350
375
375
375
400
400
400
Wavelength (nm)
60
4
2
0
0
0
7
6
4
3
1
5
0
5
0
5
1
00
C
8
6
4
2
0
0
0
0
0
-
-
15
30
2
300
325
350
Wavelength (nm)
0
5
10
15
20
25
30
Time (minute)
7
6
4
3
1
5
Figure 6. SPR thermograms of H-pin 5 with non-cognate sequences AAATTT (A)
and ACCGGT (B) and cognate sequence ACGCGT (C). The ligand concentrations from
the lowest to highest response are 0.1, 0.2, 0.4, 0.6, 0.8, 1, 2, 4, 6, 8, 10, 20, 40, 60, 80,
0
5
0
5
100, 200, and 400 nM. All sensorgrams were plotted in the same scale and
concentration range for comparison. The binding constant with ACGCGT
8
ꢁ1
(
1.6 ꢀ 10
M ) was obtained from kinetic fits of the 4 highest concentration
sensorgrams (four fit curves are overlaid with the top four sensorgrams).
were provided by the Mass Spectrometry Laboratory, University of
South Carolina, Columbia. Reaction progress was assessed by thin-
layer chromatography (TLC) using Merck silica gel (60 F254) on alu-
minum plates unless otherwise stated. Visualization was achieved
-
-
15
30
2
300
325
350
Wavelength (nm)
with UV light at 254 nm and/or 366 nm, I
hydrin spray.
2
vapor staining and nin-
Figure 5. CD spectra of H-pin 5 with (A) cognate sequence ACGCGT and non-
cognate sequences ACCGGT (B) and AAATTT (C).
4
.1.1. Synthesis of 9
N-Methylpyrrole-2-carbonyl chloride
1
0,11
(2 mmol) was dis-
solved in DCM (anhydrous, 2 mL) and added drop-wise to a solu-
sues, which are key stumbling blocks in the development of polya-
mides as potential therapeutic agents.
16
1
7
tion of 8 (1 mmol) in DCM (2 mL) and TEA (0.28 mL, 2 mmol)
and cooled to 0 °C with stirring. The reaction was allowed to warm
to RT overnight. Reaction mixture transferred to sep. funnel and
water (ꢂ15 mL) added. The aqueous layer was then basified with
NaOH, and the organic layer collected. Aqueous layer extracted
4
4
. Experimental
.1. Synthesis
with DCM (2 ꢀ 25 mL) and organic layers collected, dried (Na
and solvent removed in vacuo. The residue was then purified by
column chromatography (silica, gradient, 100:0–96:4 % CHCl
MeOH) to yield 9 as an off-white solid (74 mg, 24 %), mp. 182.8–
186.2 °C: R 0.35 (99:1 % v/v CHCl /MeOH); IR (neat) 730, 754,
1025, 1058, 1116, 1176, 1245, 1275, 1320, 1375, 1411, 1471,
2 4
SO )
Solvents and organic reagents were purchased from Aldrich or
Fisher, and in most cases were used without further purification.
DCM (P ), and DMF (BaO) were distilled prior to use. Melting
3
/
2 5
O
points (mp) were performed using a Mel-temp instrument and
are uncorrected. Infrared (IR) spectra were recorded using a Midiac
f
3
m
1
ꢁ1
1
FT-IR instrument as films on NaCl discs. H-NMR spectra were ob-
3
1545, 1649, 1704, 2362, 2927 cm ; H NMR (400 MHz, CDCl ) d
tained using a Varian Unity Inova 400 and 500 instruments. Chem-
ical shifts (d) are reported at 20 °C in parts per million (ppm)
8.19 (s, 1H), 7.56 (s, 1H), 6.79 (t, J = 2 Hz, 1H), 6.66 (dd, J = 1.6,
4 Hz, 1H), 6.15 (dd, J = 2.4, 2 Hz, 1H), 4.44 (q, J = 7.1 Hz, 2H), 4.02
(s, 3H), 3.99 (s, 3H), 1.45 (t, J = 7.2 Hz, 3H) ppm; MS (ES ) m/z
+
4
downfield from internal tetramethylsilane (Me Si). High-resolu-
tion mass spectra (HRMS) and Low-resolution mass spectra (LRMS)
(rel. intensity) 277 ([M+H], 100%), 299 ([M+Na], 55%).

H. Mackay et al. / Bioorg. Med. Chem. 16 (2008) 9145–9153
9151
Figure 7. (A) Thermogram on the binding of H-pin 5 with ACGCGT at 25 and 40 °C. (B) ITC studies on the binding of H-pin 3e to ACGCGT at 25, 30, 40, and 50 °C. (C) Plots of
H on the binding of 5 (N) and 3e (j) to ACGCGT over a temperature range for determining values of
D
p
DC .
the product precipitated. The suspension was filtered to yield com-
pound 7 as an off-white solid (59 mg, 88 %), mp. 141.2–141.8 °C: R
A
B
f
0
9
2
.17 (50:10:1, CHCl
64, 1038, 1064, 1121, 1190, 1367, 1440, 1456, 1646, 1680,
888, 2960, 3308 cm
3 4
:MeOH:NH OH); IR m (neat) 745, 847, 926,
ꢁ1 1
3
; H NMR (400 MHz, CDCl ) d 7.46 (s, 1H),
5
' - A C G C G T - 3'
6.90 (m, 1H), 6.71 (s, 1H), 6.04–6.02 (m, 1H), 3.95 (s, 3H), 3.84 (s,
3H) ppm; MS (ES ) m/z (rel. intensity) 249 ([M+H], 100%), 271
+
+
(
12 4 3
[M+Na], 18%); HRMS calcd for C11H N O , 249.0987; obsd,
+
2
49.0982.
3' - T G C G C A - 5'
4
.1.3. Synthesis of PyImPyIm H-Pin 5
Glycol linked intermediate 6 (79 mg, 0.095 mmol) was dis-
solved in MeOH (75 mL) and reduced for ꢂ18 h using 10% Pd/C
in the presence of H . PyIm-acid 7 (59 mg, 0.24 mmol, 2.5 equiv)
2
EDCI (55 mg, 0.29 mmol), DMAP (5.8 mg, 0.05 mmol) and dry
DMF (2 mL) were reacted with the resulting amine, stirring at RT
for 6 days, protected from light in an Ar atmosphere. The DMF
was removed via Kugelröhr distillation (0.005 atm.) and the resi-
due purified via column chromatography (silica, gradient, 0:100–
B
1
H
1
1
00:0 % v/v CHCl
OH) to yield H-pin 5 as a pale brown solid (32 mg, 28 %), mp.
27 °C: R 0.62 (CHCl :MeOH:NH OH, 50:10:1); IR (neat) 1685,
536, 1468, 1410, 1248, 1118, 752 cm H NMR (400 MHz,
3 3
/MeOH then 100:10:1–70:10:1, CHCl :MeOH:N-
4
f
3
4
m
ꢁ
1
1
;
DMSO) 10.40 (br s, 2H), 10.39 (br s, 2H), 9.83 (br s, 2H), 7.74 (br
C
D
t, J = 5.6 Hz, 2H), 7.55 (s, 2H), 7.50 (s, 2H), 7.47 (s, 2H), 7.14 (s,
2
H), 7.12 (m, 2H), 6.99 (s, 2H), 6.05 (t, J = 3.2 Hz, 2H), 4.46 (t,
0
0
Figure 8. (A) A model of the complex of H-pin 5 with 5 -ACGCGT-3 . (B) Computer
generated depiction of the complex of H-pin 5 with 5 -ACGCGT-3 using SYBYL
following a 1000-step structural optimization. (C) Solvent accessible surface area of
PyImPyIm H-pin 5 bound to (dCCACGCGTGG)
J = 3.2 Hz, 4H), 3.97 (s, 6H), 3.93 (s, 6H), 3.89 (s, 6H) 3.67–3.66
0
0
(
(
(
m, 4H), 3.48–3.47 (m, 4H), 2.40 (s, 4H), 2.18 (s, 12H) ppm; MS
ES ) m/z (rel. intensity) 1214 ([M+H], 40%), 607 (100%), 405
60%); HRMS calcd for C56H N O10, 1213.5880; obsd, 1213.5837.
73 22
+
2
. Red is polar and white is non-polar.
(
D) SASA of f-ImPyIm H-pin 3e bound to (dCCACGCGTGG) .
2
4
.1.4. Synthesis of f-ImPyIm H-pin 3e
The procedure was similar to that used for the synthesis of H-
4
.1.2. Synthesis of 7
PyIm-ethyl ester (9, 74 mg, 0.27 mmol) was dissolved in MeOH
3 mL). NaOH (0.16 mL, 0.32 mmol) and water (3 mL) were then
added and the reaction mixture heated under reflux with stirring
for 24 h. The cooled solution was then washed with CHCl
20 mL) and the aqueous layer acidified with HClaq (6 M) until
Pin 5, except nitro-starting material 6 (63 mg, 0.077 mmol) and
f-Im-acid 11 [11] (28 mg, 0.16 mmol) were used. The product
was isolated as a grey/brown solid (19 mg, 22%), mp. 142–
(
3
145 °C: R
f 3 4
0.37 (79:20:1 v/v CHCl /MeOH/NH OH); IR m (neat)
2924, 2862, 1655, 1535, 1463, 1437, 1380, 1256 cm ; 1H NMR
ꢁ
1
(

9
152
H. Mackay et al. / Bioorg. Med. Chem. 16 (2008) 9145–9153
(
4
4
DMSO-d
6
) d 1.69 (t, J = 6.5 Hz, 6H), 2.25 (s, 12H), 3.03 (q, J = 6.5 Hz,
H), 3.47 (s, 4H), 3.65 (t, J = 5.2 Hz, 4H), 3.93 (s, 6H), 3.94 (s, 6H),
.43 (t, J = 6.9 Hz, 4H), 7.17 (s, 2H), 7.43 (s, 2H), 7.48 (s, 2H), 7.50
4.4. Circular dichroism
CD studies were performed on an OLIS DSM20 spectropho-
tometer using the three oligonucleotides detailed above. Experi-
ments were conducted at ambient temperature in a 10 mm
(
(
s, 2H), 8.21 (s, 2H), 10.04 (s, 2H), 10.32 (s, 2H), 10.32 (s, 2H); MS
ES ) m/z (rel. intensity) 1056 ([M+H], 50%), 529 ([M+2H] , 20%).
+
2+
pathlength cuvette, using a 2.4 nm bandpass. A 9
l
M solution
+
4
.2. DNaseI footprinting
of DNA in PO 5 (10 mM phosphate, 50 mM Na , 1 mM EDTA,
4
pH 6.2) was titrated with a solution of H-pin 5 (0.5 mM in
water), in aliquots of 1 molar equivalent, past the point of satu-
ration. CD response was recorded every 1 nm over a wavelength
range of 250 to 400 nm.
4
.2.1. Preparation of the DNA substrate, radiolabeling and
purification
A 130 bp fragment was amplified by using PCR as fellow. The
0
0
forward primer 5 -GTCGTTAGGAGAGCTCACTTG-3 (4 ng) was
radioactively labeled by treatment with
3
2
c
-[ P] (3
lL) and 1
lL
4.5. Surface plasmon resonance
T4 polynucleotide kinase (Invitrogen) following the standard pro-
tocols. PCR was performed in thermophilic DNA polybuffer con-
SPR-biosensor experiments were conducted in degassed MES
buffer (200 mM Na , 10 mM [2-(N-morpholino)ethanesulfonic
+
taining dNTPs (50
l
L, 125
l
M), MgCl
2
(1 mM), Flexi Taq
3
2
polymerase (1 U) and P-labeled forward primer, reverse primer
acid], 1 mM EDTA, 0.00005 v/v of 10% surfactant P20 – BIACORE,
0
0
0
0
5
-CTCCAGAAAGCCGGCAACTCAG-3 and the templates 5 -ATGCTC
pH 6.25) at 25 °C. The 5 -biotin-labeled DNA hairpins were pur-
CAGAAAGCCGGCACTCAGTCTACAAACGCGTCATCTTGATCACCGGTG
TTCACAGAAATTTCTCTAGATCTACACGTAACTCTAGTAGCGCTCTTCA
chased from Midland Certified Reagents (Midland, TX) with HPLC
purification with the following sequences: 5 -biotin-ACGCGT, 5 -
0
0
0
0
0
0
AGCAAGTGGAGCTCTCCTAACCGACTTT-3 (20 ng) and 5 -AAAGTCG
GTTAGGAGAGCTCCACTTGCTTGAAGAGCGCTACTAGAGTTACGTGTA
GATCTAGAGAAATTTCTGTGAACACCGGTGATCAAGATGACGCGTTT
biotin-GAACGCGTCCTCTGACGC-GTTC; 5 -biotin-AAATTT, 5 -bio-
0
0
tin-CGAAATTTCCTCTGAAATTTCG; 5 -biotin-ACCGGT, 5 -biotin-
GAACCGGTCCTCTGACCGGTTC. The experiments were conducted
using a BIACORE 2000 instrument (Biacore, AB). The DNA hairpins
were immobilized on a streptavidin-derivatized gold chip (SA chip
from BIAcore) by manual injection of 25 nM hairpin DNA solution
0
GTAGACTGAGTGCCGGCTTTCTGGAGCAT-3 (20 ng). Polymerase
chain reaction was carried out as follows: an initial denaturation
step for 3 min at 95 °C and [1 min at 94 °C, 1 min at 63 °C, and
ꢁ1
1
min at 72 °C] for 35 cycles. The PCR products were purified by
with a flow rate of 1 lL min until the response units reached
2
% agarose gel electrophoresis. Finally, DNA was isolated by using
ꢂ400 RUs. Flow cell 1 was left blank, while flow cells 2, 3, and 4
were immobilized with the three different DNA hairpins. Typically,
a series of different concentrations of ligand was injected onto the
the Mermaid Kit (Q-biogene) according the manufacturer’s
instructions.
ꢁ1
chip with a flow rate of 20
lowed by a dissociation period of 20 min. At the end of every cycle,
the chip surface was regenerated with a 10 L injection of 500 mM
lL min for a period of 12.5 min, fol-
4
.2.2. DNase I footprinting experiments
DNase I digestions were conducted in a total volume of 8
l
L.
l
ꢁ1
The labeled DNA fragment (200
lL, 200 count s ) was incubated
NaCl/25 mM NaOH aqueous solution, injection tube rinsing, and
multiple 1 min buffer injections. For H-pin 5 and ACGCGT, the glo-
bal kinetic fit (1:1) was conducted on four sensorgrams of highest
concentrations (80, 100, 200, 400 nM), and the binding constant
3
0 min in 4
pH 7) containing the desired drug concentration. Cleavage by
DNase I was initiated by addition of 2 L DNase I solution (2 L,
, DNaseI 0.1 U mL , pH
) and was stopped after 3 min by cooling the samples on dry
lL TN binding buffer (10 mM Tris Base, 10 mM NaCl,
l
l
ꢁ1
2
8
0 mM NaCl, 2 mM MgCl
2
, 2 mM MnCl
2
was reported as the ratio of k
a d
/k .
4
.6. Isothermal titration microcalorimetry
ice. The samples were then lyophilized dry. DNA was resuspended
in 4 L formamide loading dye (95% formamide, 20 mM EDTA,
.05% bromophenol blue, and 0.05% cyanol blue) and denatured
l
ITC experiments were carried out on a MicroCal VP-ITC (North-
ampton, MA) at 25, 30, and 35 °C. DNA (2 M in PO 5) was titrated
with H-pin 5 (0.5 mM in PO 5) in 3 L aliquots every 300 s (50
injections). Data was processed with MicroCal Origin 7.0 as previ-
0
l
4
by heating the sample to 90 °C and cooled on ice prior to loading
onto a conventional denaturing polyacrylamide (10%) gel contain-
ing urea (7.5 M). Electrophoresis was performed for 2.5 hours
4
l
7
d,18
ously described
fit was then employed and this was subtracted from the reaction
integrations to normalize for non-specific heat components.
was calculated from Eq. 1 and using the binding constant of
using a one-site model to fit the curve. A linear
(
70 W, 50 °C) in TBE Buffer (89 mM Tris base, 89 mM boric acid,
2
3
.5 mM Na
2
EDTA, pH 8.3). The gel was transferred onto Whatman
D
G
MM paper and dried under vacuum at 80 °C for 2 h. The gel was
exposed overnight to X-Ray film (Super RX, Fuji).
8
ꢁ1
1
.6 ꢀ 10 M that was obtained from SPR studies.
G ¼ ꢁRT ln Keq
Where R is 1.987 cal mol
4
.3. Thermal denaturation
D
ð1Þ
ꢁ1
ꢁ1
K
and T is measured in K.
The synthetic DNA hairpins used in these studies were obtained
0
from Operon (Huntsville, AL): ACGCGT, 5 -GAACGCGTCGCTCTCGA
0
4.7. Molecular modeling
CGCGTTC; ACCGGT, 5 -GAACCGGTCGCTCTCGACCGGTTC; AAATTT,
0
5
-CGAAATTTCCCTCTGGAAA-TTTCG. Data was obtained using a
0
The 3 D model depiction of f-ImPyIm-EG-8 (3e) /5 -CCAC
Cary Bio 100 spectrophotometer and cells with a 10 mm path-
length. Experiments were performed in PO 0 (10 mM sodium
phosphate, 1 mM EDTA, pH 6.2) with 1 M oligonucleotide and
M ligand. Oligonucleotide samples were reannealed prior to
0
GCGTGG-3 complex was generated using SYBYL 7.0 on a Silicon
4
9
,19
Graphics workstation and reported previously.
The model for
l
H-pin 5 with the same oligonucleotide was generated in the same
manner. B-form double helical DNA was generated in SYBYL, and
the complex was constructed using the published unlinked PyIm-
3
l
denaturation studies by heating at 70 °C for 1 min then allowing
cooling to RT. The temperature was programmed to ramp from
1
0
PyIm (4)/CGCG complex by building the ethylene glycol linker
joining the two polyamide moieties through the central pyrrole
group at N1. The complex in water was then structurally optimized
in a 1000-step minimization.
2
0
5 to 95 °C at a rate of 0.5 °C/min recording the Abs260 every
.5 °C. The data was analyzed using KaleidaGraph (Synergy Soft-
m
ware, Reading PA) and the T values determined as the maximum
of the first derivative.

H. Mackay et al. / Bioorg. Med. Chem. 16 (2008) 9145–9153
9153
4
.8. Solvent accessible surface areas (SASA)
1993, 115, 9892; (c) Mrksich, M.; Dervan, P. B. J. Am. Chem. Soc. 1994, 116, 3663;
d) Olenyuk, B.; Jitianu, C.; Dervan, P. B. J. Am. Chem. Soc. 2003, 125, 4741.
(a) Lacy, E. R.; Le, N. M.; Price, C. A.; Lee, M.; Wilson, W. D. J. Am. Chem. Soc.
002, 124, 2153; (b) Buchmueller, K. L.; Staples, A. M.; Howard, C. M.; Horick, S.
M.; Uthe, P. B.; Le, N. M.; Cox, K. K.; Nguyen, B.; Pacheco, K. A. O.; Wilson, W. D.;
Lee, M. J. Am. Chem. Soc. 2005, 127, 742; (c) 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, W. D.; Lee, M. Nucleic Acids Res/ 2005, 33, 912;
(d) Buchmueller, K. L.; Bailey, S. L.; Matthews, D. A.; Taherbhai, Z. T.; Register, J.
K.; Davis, Z. S.; Bruce, C. D.; O’Hare, C. C.; Hartley, J. A.; Lee, M. Biochemistry
(
7
.
The SASA calculation procedures have been previously descri-
2
bed.7
c,20
In brief, ions were removed and carbon, carbon-bound
hydrogen, and phosphorus atoms were assigned as non-polar and
the rest polar. The PyImPyIm/CGCG complex (a total of 780 atoms:
6
30 from DNA, 150 from ligands) consists of 284 polar (p) atoms
(
(
244 from DNA, 40 from ligands) and 496 non-polar (np) atoms
386 from DNA, 110 from ligands). The solvent accessible surface
2006, 45, 13551.
8
.
(a) Verma, R.; Patapoutian, A.; Gordon, C. B.; Campbell, J. L. Proc. Natl. Acad. Sci.
U.S.A. 1991, 88, 7155; (b) Wu, X.; Lee, H. Oncogene 2002, 21, 7786; (c) Yamada,
M.; Sato, N.; Taniyama, C.; Ohtani, K.; Arai, K.; Masai, H. J. Biol. Chem. 2002, 277,
27668; (d) Hess, G. F.; Drong, R. F.; Weiland, K. L.; Slightom, J. L.; Sclafani, R. A.;
Hollingsworth, R. E. Gene 1998, 28, 133.
21
area (SASA) was calculated with GRASP using a probe radius of
1
.7683 Å and Cornell et al. radii.22
The heat capacity change (cal mol
ꢁ1
ꢁ1
K ) arisen from the polar/
non-polar area change was calculated with three different models
9. O’Hare, C. C.; Uthe, P. B.; Mackay, H.; Blackmon, K. N.; Jones, J.; Brown, T.;
Eq. 2,2 Eq. 3, Eq. 4.
3
24
25
Nguyen, B.; Wilson, W. D.; Lee, M.; Hartley, J. A. Biochemistry 2007, 46,
11661.
D
A ¼ Acomplex ꢁ ðAfree dna þ Aligand
1
þ Aligand
2
Þ
10. Brown, T.; Mackay, H.; Turlington, M.; Sutterfield, A.; Smith, T.; Sielaff, A.;
Westrate, L.; Bruce, C.; Kluza, J.; O’Hare, C.; Hartley, J. A.; Nguyen, B.; Wilson,
W. D.; Lee, M. Bioorg. Med. Chem. 2008, 16, 5266.
D
D
C
C
p-sasa ¼ ð0:32 ꢃ 0:04Þ
p-sasa ¼ ð0:45 ꢃ 0:02Þ
D
AÞnp ꢁ ð0:14 ꢃ 0:04Þ
D
A
p
ð2Þ
11.
Mulder, K.; Brown, T.; Nolan, P.; Smith, T.; Lee, M. Heterocycl. Commun. 2008,
4, 137.
1
DA
np ꢁ ð0:26 ꢃ 0:03Þ
DA
p
12. (a) Lyng, R.; Rodger, A.; Norden, B. Biopolymers 1992, 32, 1201; (b) Lyng, R.;
þ ð0:17 ꢃ 0:07Þ
D
A
OH
ð3Þ
ð4Þ
Rodger, A.; Norden, B. Biopolymers 1991, 31, 1709.
1
1
3. Lah, J.; Vesnaver, G. J. Mol. Biol. 2004, 3, 73.
D
C
p-sasa ¼ ð0:382 ꢃ 0:026Þ
D
A
np ꢁ ð0:121 ꢃ 0:077Þ
DA
p
4. (a) Paal, K.; Baeuerle, P. A.; Schmitz, M. L. Nucleic Acids Res. 1997, 25, 1050; (b)
Ptashne, M.; Gann, A. Genes and Signals, first ed.; Cold Spring Harbor
Laboratory: NY, 2001.
Acknowledgments
15. Jencks, W. P. Proc. Natl. Acad. Sci. U.S.A. 1981, 78, 4046.
6. (a) Nickols, N. G.; Jacobs, C. S.; Farkas, M. E.; Dervan, P. B. ACS Chem. Biol.
007, 2, 561; (b) Nickols, N. G.; Jacobs, C. S.; Farkas, M. E.; Dervan, P. B.
1
2
Support from the National Science Foundation (CHE-0550992),
the National Institutes of Health, (GM61587 to WDW), Glaxo-
smithkline summer undergraduate research fellowship (J.P.L.),
Cancer Research UK (C2259/A3083), and Hope College is gratefully
acknowledged.
Nucleic Acids Res. 2007, 35, 363; (c) Edelson, B. S.; Best, T. P.; Olenyuk, B.;
Nickols, N. G.; Doss, R. M.; Foister, S.; Heckel, A.; Dervan, P. B. Nucleic Acids
Res. 2004, 20, 2802; (d) Best, T. P.; Edelson, B. S.; Nickols, N. G.; Dervan, P.
B. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 12063; (e) Belitsky, J. M.; Leslie, S.
J.; Arora, P. S.; Beerman, T. A.; Dervan, P. B. Bioorg. Med. Chem. 2002, 10,
3313; (f) Anthony, N. G.; Fox, K. R.; Johnston, B. F.; Khalaf, A. I.; Mackay, S.
P.; McGroarty, I. S.; Parkinson, J. A.; Skellern, G. G.; Suckling, C. J.; Waigh, R.
D. Bioorg. Med. Chem. Lett. 2004, 14, 1353; (g) Anthony, N. G.; Johnston, B.
F.; Khalaf, A. I.; Mackay, S. P.; Parkinson, J. A.; Suckling, C. J.; Waigh, R. D. J.
Am. Chem. Soc. 2004, 15, 11338.
References and notes
1
.
(a) Melander, C.; Burnett, R.; Gottesfeld, J. M. J. Biotechnol. 2004, 112, 195; (b)
Baraldi, P. G.; Bovero, A.; Fruttarolo, F.; Preti, D.; Tabrizi, M. A.; Pavani, M. G.;
Romagnoli, R. Med. Res. Rev. 2004, 24, 475; (c) Dervan, P. B.; Edelson, B. S. Curr.
Opin. Struct. Biol. 2003, 13, 284; (d) Neidle, S. Nat. Prod. Rep. 2001, 18, 291; (e)
Reddy, B. S.; Sharma, S. K.; Lown, J. W. Curr. Med. Chem. 2001, 8, 475.
Dervan, P. B.; Bürli, R. W. Curr. Opin. Chem. Biol. 1999, 3, 688.
(a) Pelton, J. G.; Wemmer, D. E. J. Am. Chem. Soc. 1990, 112, 1393; (b) Chen, X.;
Ramakrishana, B.; Rao, S. T.; Sundaralingam, M. Nat. Struct. Biol. 1994, 1, 169.
(a) Dervan, P. B. Bioorg. Med. Chem. 2001, 9, 2215; (b) Dervan, P. B.; Doss, M. R.;
Marques, M. A. Curr. Med. Chem. Anti-Cancer Agents 2005, 5, 373; (c) Walker, W.
L.; Kopka, M. L.; Goodsell, D. S. Biopolymers 1997, 44, 323; (d) Jones, G.;
Palumbo, B. M. Advances in DNA Sequence-Specific Agents; JAI Press Inc.:
Greenwich, 1998.
17. (a) Krowicki, K.; Lown, J. W. J. Org. Chem. 1987, 52, 3493; (b) Wurtz, N. R.;
Turner, J. M.; Baird, E. E.; Dervan, P. B. Org. Lett. 2001, 3, 1201; (c) Nishiwaki, E.;
Tanaka, S.; Lee, H.; Shibuya, M. Heterocycles 1988, 27, 1945.
18. Indyk, L.; Fisher, H. F. Methods Enzymol. 1998, 295, 350.
19. Isaacs, R. J.; Rayens, W. S.; Spielmann, H. P. J. Mol. Biol. 2002, 319, 191.
20. Nguyen, B.; Stanek, J.; Wilson, W. D. Biophys. J. 2006, 90, 1319.
21. Nicholls, A.; Sharp, K. A.; Honig, B. Proteins 1991, 11, 281.
22. Cornell, W. D.; Cieplak, P.; Bayly, C. I.; Gould, I. R.; Merz, K. M., Jr.; Ferguson, D.
M.; Spellmeyer, D. C.; Fox, T.; Caldwell, J. W.; Kollman, P. A. J. Am. Chem. Soc.
1995, 117, 5179.
23. (a) Spolar, R. S.; Ha, J. H.; Record, M. T., Jr. Proc. Natl. Acad. Sci. U.S.A. 1989, 86,
8382; (b) Livingstone, J. R.; Spolar, R. S.; Record, M. T., Jr. Biochemistry 1991, 30,
4237; (c) Spolar, R. S.; Livingstone, J. R.; Record, M. T., Jr. Biochemistry 1992, 31,
3947; (d) Spolar, R. S.; Record, M. T., Jr. Science 1994, 263, 777.
24. (a) Murphy, K. P.; Bhakuni, V.; Xie, D.; Freire, E. J. Mol. Biol. 1992, 227, 293; (b)
Murphy, K. P.; Freire, E. Adv. Protein Chem. 1992, 43, 313; (c) Gomez, J.; Freire, E.
J. Mol. Biol. 1995, 252, 337; (d) Gomez, J.; Hilser, V. J.; Xie, D.; Freire, E. Proteins
1995, 22, 404; (e) Habermann, S. M.; Murphy, K. P. Protein Sci. 1996, 5,
1229.
2
3
.
.
4.
5.
6.
(a) Al-Said, N.; Lown, J. W. Synth. Commun. 1995, 25, 1059; (b) Chen, Y-H.; Yang,
Y.; Lown, J. W. J. Biomol. Struct. Dyn. 1996, 14, 341; (c) Chen, Y-H.; Liu, J-X.;
Lown, J. W. Bioorg. Med. Chem. Lett. 1995, 5, 2223; (d) O’Hare, C. C.; Mack, D.;
Tandon, M. S.; Sharma, K.; Lown, J. W.; Kopka, M. L.; Dickerson, R. E.; Hartley, J.
A. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 72.
(a) Dwyer, T. J.; Geierstanger, B. H.; Mrksich, M.; Dervan, P. B.; Wemmer, D. E. J.
Am. Chem. Soc. 1993, 115, 9900; (b) Mrksich, M.; Dervan, P. B. J. Am. Chem. Soc.
25. Ren, J.; Jenkins, T. C.; Chaires, J. B. Biochemistry 2000, 39, 8439.