High DNA affinity of a series of peptide linked diaromatic guanidinium-like derivatives

Article abstract of DOI:10.1021/jm300296f

In this paper we report the design and synthesis of a new family of asymmetric peptide linked diaromatic dications as potent DNA minor groove binders. These peptide-linked compounds, with a linear core, displayed a much larger affinity than other guanidinium-like derivatives from the same series with curved cores. As a first screening, the DNA affinity of these structures was evaluated by means of thermal denaturation experiments, finding that the nature of the cation (guanidinium vs 2-aminoimidazolinium) significantly influenced the binding strength. Their binding affinity was assessed by implementing further biophysical measurements such as surface plasmon resonance and circular dichroism. In particular, it was observed that compounds 6, 7, and 8 displayed both a strong binding affinity and significant selectivity for AT oligonucleotides. In addition, the thermodynamics of their binding was evaluated using isothermal titration calorimetry, indicating that the binding is derived from favorable enthalpic and entropic contributions.

Full text of DOI:10.1021/jm300296f

Article
pubs.acs.org/jmc
High DNA Affinity of a Series of Peptide Linked Diaromatic
Guanidinium-like Derivatives
Padraic S. Nagle,† Fernando Rodriguez,†,§ Binh Nguyen,‡,∥ W. David Wilson,‡ and Isabel Rozas*,†
†School of Chemistry, Trinity Biomedical Sciences Institute, University of Dublin, Trinity College, Pearse St., Dublin 2, Ireland
‡Department of Chemistry, Georgia State University, Atlanta, Georgia, United States
S
* Supporting Information
ABSTRACT: In this paper we report the design and synthesis of a new family of asymmetric peptide linked diaromatic dications
as potent DNA minor groove binders. These peptide-linked compounds, with a linear core, displayed a much larger affinity than
other guanidinium-like derivatives from the same series with curved cores. As a first screening, the DNA affinity of these
structures was evaluated by means of thermal denaturation experiments, finding that the nature of the cation (guanidinium vs
2-aminoimidazolinium) significantly influenced the binding strength. Their binding affinity was assessed by implementing further
biophysical measurements such as surface plasmon resonance and circular dichroism. In particular, it was observed that
compounds 6, 7, and 8 displayed both a strong binding affinity and significant selectivity for AT oligonucleotides. In addition, the
thermodynamics of their binding was evaluated using isothermal titration calorimetry, indicating that the binding is derived from
favorable enthalpic and entropic contributions.
structure to optimally fit into the groove. In fact, studies on
many synthetic derivatives such as furamidine and furimidazole
have shown that the curvature degree plays an important role in
the binding.7 Thus, on the one hand, it was observed that
furamidine exhibits a too open molecular curvature that
prevents the furan ring from optimizing its contacts with the
groove floor.8 On the other hand, Boykin et al. reported a
thiophenebenzimidazole analogue of furamidine whose binding
strength was increased as a result of a decreased curvature,
thereby allowing the central moiety of the molecule to establish
optimized contacts with the minor groove.9 Moreover, it was
found that linear derivatives, which differ from the classical
minor groove binder criteria, rendered very good binding
affinity by hydrogen-bonding to water molecules. It seems that
these linear compounds form HBs with water molecules to link
to the DNA base pairs at the floor of the minor groove, forming
the corresponding DNA complex; hence, these water molecules
provide the curvature required.10,11 In addition, Wilson et al.
have reported a compound that deviates from the curvature
theory by forming a “seesaw” complex with the minor groove.12
Additionally, different oligoamides have been prepared that are
capable of highly specific DNA binding;13 further research has
also resulted in the preparation of hairpins polyamides that fit
snuggly into the groove.14 The degree of polyamide curvature is
a relevant issue when studying the facility of such compounds
INTRODUCTION
■
The DNA minor groove is an important target site for many
synthetic compounds, and for that reason, during the past 5
years, we have been designing and preparing symmetric and
asymmetric bis-cationic diaromatic derivatives as DNA minor
groove binders. Thus, we have been able to develop new
molecular species with good affinity toward DNA and the
ability to recognize specific sequences.1−3
Minor groove binders are finding application in different
biological fields. Thus, a correlation between DNA binding
affinity at AT sites and trypanocidal activity for a series of
derivatives was found, supporting the hypothesis that the
antitrypanosomal activity could occur by the formation of a
DNA complex.4 In this sense, a number of bis(2-amino-
imidazolinium) diaromatic derivatives that bind strongly to the
DNA minor groove have been reported by Dardonville, Rozas,
and co-workers with very good antitrypanosomal and anti-
plasmodial activity.3 In addition, Motoshima et al. have
reported a series of bis-amide derivatives related to furamidine
that show very good antimalarial activity.5 Moreover, Suckling
and Khalaf have shown in several studies that DNA minor
groove binders related to distamycin are good and efficient
antibacterial agents.6 In particular, they have shown that
peptide derivatives (amide isosters) show good DNA binding
affinity and antibiotic activity.
It is generally accepted that DNA minor groove binders,
which interact with DNA by means of hydrogen bonds (HBs)
and van der Waals and ionic forces, should have a curved
Received: March 1, 2012
Published: April 12, 2012
© 2012 American Chemical Society
4397
dx.doi.org/10.1021/jm300296f | J. Med. Chem. 2012, 55, 4397−4406

Journal of Medicinal Chemistry
Article
Figure 1. Design of peptide-linked dicationic derivatives 6, 7, and 8.
to bind into the DNA minor groove, and it has been recently
studied.15
model for the bis-guanidinium derivative (6) was prepared and
optimized using DFT methods (B3LYP/6-31+G*, mimicking
water solvation with PCM). Two models were calculated, one
with the isolated peptide-linked molecule and the other using
hydrogen bonded water molecules to elongate the curvature of
the dicationic structure (Figure 2). In the case of the isolated
As mentioned, we have previously reported a number of
symmetric and asymmetric bis-guanidinium or -2-aminoimidazoli-
nium diaromatic compounds that strongly bind to DNA with
significant AT sequence selectivity.1 Moreover, their interactions
with DNA were investigated by various biophysical techniques such
as thermal denaturation, circular dichroism (CD), surface plasmon
resonance (SPR), and isothermal titration calorimetry (ITC), and
their mode of binding was confirmed by linear dichroism (LC) as
being minor groove binders.2 In particular, those molecules with a
NH or a CO linker (Figure 1) gave very interesting results, the
NH-linked ones displaying the largest binding affinity in each series
(symmetric bis-guanidinium (1) and bis-2-aminoimidazolinium (2)
and asymmetric guanidinium/2-aminoimidazolinium (3)) with
good AT selectivity, whereas from the two CO-linked analogues 4
and 5, the asymmetric guanidinium/2-aminoimidazolinium (5)
displayed unusual CG binding.
Taking into account these compounds and considering that
many known minor groove binders (i.e., netropsin or
distamycin) contain peptide bonds (-CONH-) that allow hydro-
gen bonding to DNA bases, we present here the preparation and
DNA affinity evaluation of a series of symmetric and asymmetric
peptide linked derivatives carrying guanidinium-like cations
(compounds 6, 7, and 8; see Figure 1). The reasoning behind
this design is that the combination of a HB acceptor such
as CO with a HB donor such as NH could provide extra
points of contact, resulting in an increased binding affinity
for DNA.
The proposed peptide-linked dicationic molecules 6, 7, and 8
would be linear structures, but as we have discussed before, in
principle this should not be a problem for their binding into the
DNA minor groove. Additionally, the mono-guanidinium (9)
and mono-2-aminoimidazolinium (10) peptide derivatives have
also been prepared. Finally, to investigate the binding of all
these compounds to DNA, different biophysical techniques
have been used such as thermal denaturation and CD to
measure their binding strength, SPR to obtain information on
their selectivity, and ITC to find the relevant thermodynamic
quantities and binding constants.
Figure 2. DFT (B3LYP/6-31+G*, PCM-water) optimized structures
of (a) bis-guanidinium peptide linked derivative showing a curvature
radius of 11.8 Å and (b) the same compound complexed with three
water molecules showing a radius of 14 Å. In both cases an
intramolecular interaction between the peptide CO group and one
of the aromatic H atoms was found. Graphics and radius
determination were done with Gaussview 5.0 (C atoms are
represented in gray, N atoms in blue, O atoms in red, and H atoms
in white).
dication (Figure 2a), using as reference a NH2 of each of the
guanidinium groups and the N and C atoms of the peptide
linker, a curvature of a 11.8 Å radius was obtained.
In the model containing the water molecules, two of them
were used to complex each of the guanidinium cations while a
third water molecule was hydrogen bonded to the peptide
CO group in order to avoid a double “intramolecular” HB
with one of the end side water molecules, thus avoiding the
bending of the molecule. In this case, the same points were
used as reference to calculate the curvature radius plus the O
atoms of each of the water molecules at both ends of the
system, and a value of 14 Å was obtained (see Figure 2b). The
water molecules on both ends of the molecule have extended
RESULTS
■
Computational Study. By use of the crystal structure of
a peptide-linked bis-isouronium analogue16 as a template, a
4398
dx.doi.org/10.1021/jm300296f | J. Med. Chem. 2012, 55, 4397−4406

Journal of Medicinal Chemistry
Article
a
Scheme 1. Synthesis of Asymmetric Mono- and Bis-guanidinium-like Hydrochloride Salts 8, 9, and 10
a
Reagents and conditions: (i) N,N′-bis(tert-butoxycarbonyl)thiourea, HgCl2, NEt3, DMF, room temp; (ii) N,N′-bis(tert-butoxycarbonyl)imidazoline-
2-thione, HgCl2, NEt3, DMF, room temp; (iii) TFA, DCM, room temp followed by Amberlyte resin, H2O, room temp.
the arc, and for that reason the radius is larger. However, this is
still a reasonable curvature radius to fit into the minor groove.
From this simple computational model it was evident that
these peptide-linked dications (as isolated structures as much as
networked with water molecules) will adopt a reasonable
curvature to fit into the DNA minor groove.
of the aromatic diamine was first performed yielding only
compound 11, which was subsequently reacted with the 2-
aminoimidazolinium precursor producing the asymmetric Boc-
protected derivative 13 (Scheme 1). The Boc-protected
guanidine derivative substituted in position para to the aromatic
CO group was never obtained because of the deactivating effect
of such a carbonyl group.
Chemistry. On the basis of our previous experience and
taking into account the encouraging synthetic results obtained
in our group,17,18 the guanidylation and 2-aminoimidazolylation
processes to achieve the present target molecules were per-
formed following our established methodology. This consists of
the reaction of the peptide-linked (mono)deactivated aromatic
diamine or the mono-Boc-protected derivative (both commer-
cially available) with N,N′-bis(tert-butoxycarbonyl)thiourea
(guanidine precursor) or N,N′-bis(tert-butoxycarbonyl)-
imidazoline-2-thione14 (2-aminoimidazoline precursor) assisted
by mercury(II) chloride and in the presence of an excess of
triethylamine in DMF (Scheme 1). Deprotection of the Boc
intermediates thus obtained with trifluoroacetic acid/dichloro-
methane and further treatment with Amberlyte resin in water lead
to the hydrochloride salts of the target molecules (Scheme 1).14,15
In the preparation of the Boc-protected mono-guanidinium-like
compound 11, reaction occurs preferentially at the amino group
in para position with respect to the peptide NH group which is
more activated. The synthesis of 9 and 10 was previously
reported by us using different strategies.3
In the case of the asymmetric derivative, the preparation of
the Boc-protected compound 13 required a careful strategy.
Considering the nature of the peptide linker, those derivatives
with the guanidinium or the 2-aminoimidazolinium moiety
in position para to the aromatic NH group are preferentially
formed. Thus, it could be possible to first guanidylate the
diamine and then 2-aminoimidazolydate it or the other way
around, first introducing the 2-aminoimidazoline ring and then
the guanidinium precursor. However, in our experience, early
introduction of the 2-aminoimidazoline moiety usually results in
very poor yields due to decomposition during chromatographic
purification in silica support. For that reason, Boc-guanidylation
Total yields obtained in the preparation of the two mono-
guanidinium-like derivatives were 55% (compound 9) and 60%
(compound 10), and for the asymmetric guanidinium/2-
aminoimidazolinium 8 a 35% yield was achieved. Boc
deprotection reactions worked with very good yields around
92−95%.
Thermal Denaturation Assays: Relative Binding
Affinity. The interaction of compounds 3 and 5−10 with
unspecific salmon sperm DNA was evaluated by means of
thermal denaturation experiments, enabling a quick qualitative
evaluation for the relative DNA binding affinities as we have
previously shown.1,3 In addition, AT specific interactions were
investigated by performing these thermal denaturation experi-
ments with poly(dA·dT)2 in compounds 6−8. The melting
temperature obtained for all these molecules with unspecific
salmon sperm DNA and the AT homopolymer poly(dA·dT)2 is
presented in Table 1 together with previous results reported by
us for NH-linked and CO-linked derivatives and with the AT
specific poly(dA)·poly(dT) oligonucleotide, for the sake of
comparison.
The results presented in Table 1 show that all molecules
displayed a stabilizing effect on natural DNA where compounds
9 and 10 show minimal affinity for mixed sequence DNA.
Furthermore, compound 10 reveals an extra stabilizing
effect on natural DNA due to the hydrophobic nature of the
2-aminoimidazoline cation. These results were some-
what surprising given the fact that the aromatic rings are
virtually planar thus allowing for optimized contacts within
the groove.
When the second amino group was functionalized by another
cationic group, the affinity rose significantly with a 5- to 10-fold
4399
dx.doi.org/10.1021/jm300296f | J. Med. Chem. 2012, 55, 4397−4406

Journal of Medicinal Chemistry
Article
Table 1. Summary of Thermal Denaturation Results
Obtained for All Peptide-Linked Structures Prepared by
Rozas and Co-Workers in the Present and Previous Studies
Scheme 2. Structures of the Hairpins Used for the SPR
Studies
d
difference (with a ΔTm(SS) between 10 and 12.5). This is
justified because of both (i) the addition of an extra point of
contact in terms of a HB donor on the guanidine or 2-
aminoimidazoline functionality and (ii) the additional ionic
interaction between the dicationic molecule and the negatively
charged DNA. Furthermore, the order of binding affinity found
is 7 > 8 > 6, and this seems to indicate that the hydrophobic
nature of the 2-aminoimidazoline functionality plays a role in
increasing the binding affinity.
Considering these promising results, the experiments were
repeated by replacing the mixed sequence DNA for the AT
homopolymer as a method to investigate sequence selectivity.
The minor groove of the AT polymer is narrower compared to
natural DNA, allowing for a tighter fit and for the formation of
closer contacts with the walls of the minor groove. The results
from Table 1 show the preferred binding to AT sequences with
up to 2.6-fold difference. Moreover, when the guanidine
functionality was replaced by a 2-aminoimidazoline group, the
affinity for AT sequences rose dramatically from a 1.8- to a 2.6-
fold difference. This can be attributed to the presence of an extra
ethylene bridge which has the ability to form van der Waals
contacts with the walls of the minor groove. From the results
presented, the presence of the extra bridge is necessary for
increasing the sequence selectivity of these particular compounds.
a
DNA melting temperature in phosphate buffer (10 mM) is 69 °C.
b
P/D = 10. Poly(dA·dT)2 Tm in phosphate buffer (10 mM) is 47 °C
c
(as in ref 1). P/D = 3. Poly(dA)·poly(dT) Tm in MES buffer
(10 mM) is 43 °C (as in ref 3). The increment in DNA thermal melting
(ΔTm, °C) is presented in unspecific salmon sperm DNA as well as in
specific AT oligonucleotides such as poly(dA·dT)2 and poly(dA)·poly(dT).
d
Figure 3. Sensorgrams for the interaction of compound 7 with the AATT hairpin (upper left), with (AT)4 hairpin (upper right), with (CG)4 hairpin
(bottom right) and in the bottom left the fits of 7 with AATT (squares), with (AT)4 (circles), and with (CG)4 (diamonds).
4400
dx.doi.org/10.1021/jm300296f | J. Med. Chem. 2012, 55, 4397−4406

Journal of Medicinal Chemistry
Article
Table 2. Binding Constants (M−1) Obtained from SPR
Experiments for Compounds 6 and 7 with AATT, (AT)4,
and (CG)4 Hairpins
hairpins (see the corresponding for compound 7 in Figure 3, bottom
left) show very small increments, indicating reduced binding.
The binding constants (K) for compounds 6 and 7 to the
three DNA hairpin duplexes shown in Scheme 2 were
determined by plotting the steady-state response values versus
the concentration of the free compounds and fitting as
described in the Experimental Section. The results obtained
are presented in Table 2. SPR errors in K for AATT and (AT)4
are 20% based on repeat experiments for compounds 6 and 7.
For the lower binding constants for 3, 5, and proflavine, the
errors increase to 30% for all DNAs tested.
In general, larger binding constants are obtained for the
interactions with AATT and (AT)4 than with (CG)4 and only a
single strong AT binding site is observed for these compounds.
This clearly indicates a strong AT sequence specificity for the
binding of these peptide-linked dications which probably are
pure minor groove binding agents. Moreover, from a com-
parison of the binding constants of compounds 6 and 7 with
those of the intercalator proflavine (see Table 2), the high
selectivity toward AT sequences is more evident confirming
even further the pure minor groove binder character of these
compounds.
a
K (×105 M−1
)
compd
AATT
(AT)4
(GC)4
sequence specificity
b
71.0
93.7
18.0
47.7
<1
AT
b
7
3
5
<1
AT
c
b
6.4
<1
AT
c
c
2.7
3.6
AT
proflavine
5.7
12.2
6.5
none
a
The corresponding binding constants for asymmetric dications NH-
linked 3 and CO-linked 5 and for a classical intercalator proflavine are
b
shown for comparison. There is not enough signal-to-noise ratio to
c
get a binding constant for (GC)4. Reference 2.
Surface Plasmon Resonance (SPR). SPR is a quantitative
method to evaluate the binding of small molecules to specific
oligonucleotides and provides accurate binding constants. In
the present study, the hairpins shown in Scheme 2 (AATT, AT,
and CG) have been used to determine the binding of some of
the peptide-linked dications.
Circular Dichroism Studies. Given the demonstrated
affinity of these molecules to poly(dA·dT)2 sequences as
indicated by the thermal denaturation results, circular
dichroism (CD) experiments were carried out. These have
proved useful for determining binding mode and can provide
information on conformational changes in the polynucleotides
as well as offer a means to evaluate the binding stoichiometry.19
A CD spectrum monitors the induced chiral environment of
a molecule on binding to DNA. Thus, it can be used as a
Sensorgrams for compound 7 showing strong binding [to
AATT and (AT)4] and weak binding [to (CG)4] are presented
in Figure 3. All the sensorgrams recorded for the binding to AT
hairpins (for compounds 6 and 7 and also those previously
reported by us in ref 2 for compounds 3 and 5) show rapid kinetics
of association and dissociation from low to high concentrations,
indicating that these compounds can enter and leave the minor
groove with a relatively low energetic barrier. However, the
sensorgrams obtained for compounds 6 and 7 binding to CG
Figure 4. CD titration spectra of poly(dA·dT)2 [37.5 μM in base] on addition of the peptide linked compounds 6 (upper, left), 7 (upper, right), and
8 (bottom, left) with a Bp/D ratio from 22.4 to 0.56. Also shown are CD saturation curves for 6, 7, and 8 (bottom, right); the dark line indicates
where saturation occurs. Experiments were run in phosphate buffer (pH 7) at T = 25 °C.
4401
dx.doi.org/10.1021/jm300296f | J. Med. Chem. 2012, 55, 4397−4406

Journal of Medicinal Chemistry
Article
Figure 5. ITC results obtained for the bis-2-aminoimidazoline compound 7 (left) and for the guanidine/2-aminoimidazoline derivative
8 (right) in poly(dA·dT)2. These plots show corrected experimental data, and the experiments were done in phosphate buffer at
25 °C.
diagnostic for the mode of binding where compounds that bind
to the minor groove are in proximity to the chiral sugar,
resulting in a large induced CD signal (ICD).20 In particular,
minor groove binders, because of their proximity to chiral sugar
molecules, typically exhibit strong ICD signals, whereas
intercalators show little or no ICD signals.
CD titration experiments were carried out to evaluate the
binding mode and the subsequent binding stoichiometry for
compounds 6, 7, and 8 to random sequenced DNA and
poly(dA·dT)2 homopolymer (Figure 4). The free compounds
have no CD signal by itself; however, in the presence of DNA a
large ICD signal is formed at 302, 308, and 303 nm for 6, 7, and
8, respectively. The magnitude of these ICD signals is related to
the chiral environment as previously mentioned, and thus,
molecules interacting in the minor groove will induce a larger
signal change.21
Titrations with natural DNA (spectra in Supporting
Information) result in the formation of an ICD signal at
280 nm accompanied by small changes in the DNA spectrum
on addition of the compounds. In contrast, when the
experiments were repeated in the presence of the poly(dA·dT)2
homopolymer, large ICD signals were observed in the region
from 300 to 310 nm, suggesting that the compounds interact
strongly with the minor groove. Moreover, small changes in the
DNA signal at 260 nm were observed when the experiments
were carried out with natural DNA whereas large ICD signals
were observed in the AT homopolymer CD spectra, implying
that significant changes in the minor groove of the AT
homopolymer occurred on addition of the compound. As the
minor groove for AT sequences is much narrower than that of
natural DNA with a 32% GC content, the binding of these
molecules might cause more significant changes in the structure
and hence in the DNA spectrum. Moreover, it could be
postulated that the binding of these molecules causes significant
widening of the groove, as evidenced from the mentioned
changes in the DNA spectrum. However, interpretation of this
region is difficult to make because the compound absorption
occurs nearby.
Moreover, the ICD signal for the asymmetric guanidine/2-
aminoimidazoline compound 3 (NH linker) was considerably
larger than that of its analogue 8 (peptide linker), although the
DNA binding of the latter was more favorable than the former.2
The larger induced CD signal may suggest that the diphenyl
torsional angle is larger when the amide linker is replaced by
the -NH- group.8
The binding stoichiometry was evaluated using saturation
plots (Figure 4, bottom right), and it supports a Bp/D binding
ratio of 4, which is a good agreement with our previous
compounds.
Isothermal Titration Calorimetry. Calorimetric titrations
allow for the evaluation of the binding enthalpy and equilibrium
association constant in one experiment.22 In Figure 5 the
titration results representing exothermic reactions for all
binding experiments for compounds 7 and 8 are shown.
Furthermore, the equilibrium association constant (K; ΔG =
nRT ln K) thus obtained and the binding enthalpy (H) allow
the calculation of the binding entropy (S) from the relationship
ΔG = ΔH − TΔS. A summary of the thermodynamic
parameters is presented in Table 3. The binding thermody-
namics reveal favorable enthalpic and entropic contributions
where the enthalpy varies by 1 kcal mol−1, while the entropic
contribution varies from 13.9 to 20.1 cal mol−1 K−1.
4402
dx.doi.org/10.1021/jm300296f | J. Med. Chem. 2012, 55, 4397−4406

Journal of Medicinal Chemistry
Article
Table 3. Thermodynamic Parameters, Binding Stoichiometry, and Binding Constants Calculated for Compounds 7 and 8 by
ITC with Poly(dA·dT)2 (Phosphate Buffer)
compd
binding stoichiometry
K (×105 M−1
)
ΔH0 (kcal mol−1
)
ΔS0 (cal mol−1 K−1
)
7
8
0.187 0.001
0.193 0.005
4.9 0.8
−1.78 0.04
−2.55 0.09
20.1
15.9
1.54 0.63
Remarkably, the binding enthalpy contribution for these
compounds becomes more negative following the order of
8 < 7, suggesting that replacing the guanidine functionality for
the 2-aminoimidazoline group results in the formation of more
favorable interactions. Considering that these molecules exhibit
distinct structural similarities and accounting for the main
difference being the ethylene bridge on the 2-aminoimidazoline
functionality, the resulting decrease in enthalpy could be
attributed to the extra van der Waals contacts from these cyclic
functionalities.
Moreover, we have observed that in the crystal structure of a
related compound with an O linker,24 the angle between the
phenyl rings is 72°, and in the computed bis-guanidinium NH
and CO derivatives this angle was found to be 161° and 157°,
respectively; i.e., they are almost perpendicular to each other.
This indicates that compounds 1 and 4 need to undergo a
torsion to fit within the minor groove, and we have previously
evaluated the energy required for this type of torsion to be
around 6 kcal mol−1 (for asymmetric guanidine/2-amino-
imidazoline derivatives with NH or CO as linkers2). On the
contrary, the phenyl rings on the peptide-linked derivatives are
coplanar, and hence, such torsion would not be required prior
to binding into the minor groove and this binding would be
energetically more favored.
The dicationic structures with a linear Ph-CO-NH-Ph core
presented in this study show a favored curvature by means of
the guanidinium cations and exhibit a better length and fitting
than the previously prepared bis-guanidine-like minor groove
binders with only a NH or a CO linker. The crystal structure of
the NH-linked bis-2-aminoimidazoline diaromatic molecule
complexed to an AT oligonucleotide25 shows that this linker
clearly points outside the groove. In a similar manner, the
linkers from compounds 1 and 4 (NH and CO, respectively)
are expected to point out of the groove; however, the NH of
the amide linker is pointing inside the groove (see Figure 2),
thus increasing the probability of extra HBs forming with the
inner side of the groove improving the binding.
DISCUSSION AND CONCLUSIONS
■
In the present study, we have focused on a series of amide
linked molecules that have demonstrated the strongest binding
affinity in our current series of guanidine and 2-aminoimidazo-
line molecules. Thus, this study served to address the following
questions: (i) What, if any, is the relevance for the radius of
curvature in binding, and (ii) what is the importance of a
functionalized linker with a HB donor and acceptor?
Many detailed studies have focused on the design of mole-
cules that complement the concave structure of the DNA minor
groove. Nevertheless, some linear molecules have attracted
attention because of their unpredictable higher affinity for the
DNA minor groove where a water molecule allows the
completion of the required curvature. This allowed us to
realize the importance of water molecules that serve as HB
bridges, thus allowing the formation of an appropriate radius of
curvature of subsequent molecules. These interactions have
found new scope for investigations for targeting the minor
groove.
As a result of calculations performed on bis-guanidinium
derivatives 1 (NH) and 4 (CO), we have observed that the
curvature radius of these compounds (8.7 and 8.2 Å, re-
spectively; see Supporting Information) strongly differs from
that obtained for compound 6 (11.8 Å), which could explain
the differences in binding. Furthermore, berenil, which has a
radius of curvature of 14 Å, demonstrates significant affinity for
the minor groove of AT sequences. Given that the radius of 8
is 11.8 or 14 Å when water molecules are incorporated (as
calculated computationally with B3LYP/6-31+G* and PCM-
water), their potential as minor groove targeting drugs is
evident. Dickerson and co-workers have done a detailed
quantitative analysis of the appropriate curvature and position
of functional groups for optimum binding to the DNA minor
groove.23 On the basis of the helical curvature of B-form DNA
and known minor groove binders, they conclude that a 20−25 Å
radius of curvature is optimal for minor groove binding.
However, there is some flexibility in this range due to some
possible induced fit conformational changes in both the binding
compound and DNA.
By use of our already established methodology, three new
peptide-linked compounds have been prepared: an asymmetric
guanidinium/2-aminoimidazolinium diaromatic derivative and
the mono-guanidinium and mono-2-aminoimidazolinium de-
rivatives compounds 8, 9, and 10, respectively.
Thermal melting denaturation experiments were carried out
for these compounds and compared with previous results
obtained for related peptide-linked and CO- or NH-linked
diaromatic bis-guanidinium-like derivatives. Mono-guanidi-
nium-like compounds showed poor binding; however, the
bis-guanidinium-like peptide-linked derivatives showed very
strong binding to DNA with selectivity for the AT sequences,
indicating minor groove preference. Compound 7 shows the
highest affinity toward AT sequences with a ΔTm value 1.1 and
1.3 times that of 8 and 6, respectively, illustrating that the 2-
aminoimidazoline ring promotes a higher affinity for AT
sequences. Given that the only difference between both cations
is the extra ethylene bridge, these results can indicate that the
extra van der Waals contact with the minor groove increases
both the selectivity and affinity of these molecules significantly.
Finally, other biophysical experiments on the peptide-linked
compounds 6, 7, and 8 were carried out using SPR, CD, and
ITC techniques, all of them indicating strong binding to DNA
in particular to the minor groove. The binding constants
obtained with SPR for compounds 6 and 7 and for the
guanidine/2-aminoimidazoline dications with a NH or a CO
linker (3 and 5) clearly show that the peptide-linked derivatives
bind more strongly despite the presumed linearity of their core.
Despite the fact that the SPR and ΔTm values were obtained at
The nature of the strong interaction with AT sequences can
also be explained by the fact that the CO group from the amide
linker can form a weak HB interaction with the ortho-H with
respect to the linker (bond distance of 2.0 Å), thus making the
amide linker and the resulting aromatic moiety coplanar al-
lowing for a better fit into the narrow groove of AT sequences.
4403
dx.doi.org/10.1021/jm300296f | J. Med. Chem. 2012, 55, 4397−4406

Journal of Medicinal Chemistry
Article
150.9 MHz for 13C NMR. Shifts are referenced to the internal solvent
signals. HRMS spectra were measured using methanol as the carrier
solvent. Melting points are uncorrected. Infrared spectra were
recorded on a spectrometer equipped with Universal ATR sampling
accessory. Elemental analyses (C, H, N) of the target compounds,
which were performed at the Microanalysis Laboratory (School of
Chemistry and Chemical Biology, University College Dublin, Ireland),
are within 0.4% of the calculated values, confirming ≥95% purity.
Fractional moles of water frequently found in amidinium-like salts
could not be prevented despite 24−48 h of drying in vacuum.
General Procedure for the Boc-Guanidylation or Boc-2-
Aminoimidazolylation of Aromatic Amines: Method 1. Each of
the corresponding amines was treated in DMF at 0 °C with 1.1 equiv
of mercury(II) chloride, 1.0 equiv of N,N′-di(tert-butoxycarbonyl)-
imidazolidine-2-thione (for the 2-aminoimidazoline precursors) or
N,N′-di(tert-butoxycarbonyl)thiourea (for the guanidine precursors),
and 3.1 equiv of NEt3. The resulting mixture was stirred at 0 °C for 1 h
and for the appropriate duration at room temperature. Then the
reaction mixture was diluted with EtOAc and filtered through a pad of
Celite to get rid of the mercury sulfide formed. The filter cake was
rinsed with EtOAc. The organic phase was washed with water (2 ×
30 mL), washed with brine (1 × 30 mL), dried over anhydrous
Na2SO4, and concentrated under vacuum to give a residue that was
purified by silica gel (guanidine precursors) or neutral alumina column
flash chromatography (2-aminoimidazoline precursors), eluting with
the appropriate hexane/EtOAc mixture.
different salt concentrations (92 and 0 mM NaCl, respectively),
qualitative comparison of these binding constants with the ΔTm
values using poly(dA·dT)2 and poly(dA)·poly(dT) (see Table 1)
shows similar trends, the best agreement found between ΔTm and
the AATT binding constants indicating a possible preference for
the minor groove.
Recently, Terame and co-workers26 have reported the
interactions of a guanidine derivative (amiloride) and DNA,
suggesting that there may be a possible repulsion between the
methyl group in the thymine base and the cationic guanidinium
group in the major groove and that this may lead the molecule
to bind in the minor groove where no such repulsion is possible. A
similar reason could be behind the minor groove binding of our
molecules. Moreover, they also hint that the interaction of the
guanidinium cation and the phosphate backbone could be possible
in the minor groove, favoring these compounds to lay into the
DNA minor groove.
In the ITC experiments, each of the compounds displayed
larger binding constants than previous molecules in this series.2
Comparison of the binding constants indicated that the order
of binding strength 7 > 8 correlates with the thermal
denaturation results. The large entropy observed for 7 seems
to be responsible for the increased binding constant. However,
the order of the enthalpies is reversed, which could indicate that
the guanidine functionality (present in compound 8) forms
stronger hydrogen bonds than the 2-aminoimidazoline one.
This is in agreement with the hypothesis that the bis-2-
aminoimidazoline 7 has more freedom inside the narrow minor
groove than 8. However, further investigation is required to
confirm this.
4-Amino-N-[4-N′,N″-di(tert-butoxycarbonyl)guanidinophenyl]-
benzamide (11). Following method 1 and with purification by flash
chromatography in silica (hexane/ethyl acetate, 2:1), a white solid was
1
obtained (58%): mp 136−138 °C; H NMR (CDCl3) δ 1.40 (s, 9H,
(CH3)3), 1.52 (s, 9H, (CH3)3), 5.77 (brs, 2H, NH2), 6.61 (d, 2H, J =
8.4 Hz, Ar.), 7.46 (d, 2H, J = 8.4 Hz, Ar.), 7.69−7.72 (m, 4H, Ar.),
9.82 (brs, 1H, NH), 9.95 (brs, 1H, NH), 11.47 (brs, 1H, NH); 13C
NMR (CDCl3) δ 27.6, 27.9 ((CH3)3), 78.7, 83.3 (C(CH3)3), 112.5,
120.0, 120.9, 123.1, 123.2, 129.3, 136.8, 144.4 (Ar.), 152.1, 153.0
(CO), 162.7 (CN), 165.1 (CO). νmax (film)/cm−1: 3474, 3373, 2984
(NH), 1718, 1647, 1624, 1608 (CO, CN).
In summary, the presence of a planar amide linker between
the phenyl rings seems to provide the bis-cataionic ligands with
the orientation ideal to better fit into the minor groove and to
establish very favorable interactions.
4-(N-tert-Butoxycarbonylamino)-N-{4-[N′,N″-di(tert-butoxy-
carbonyl)-2-aminoimidazolino]phenyl}benzamide (12). Follow-
ing method 1 and with purification by flash chromatography in
alumina (hexane/ethyl acetate, 2:5), a white solid was obtained (65%):
EXPERIMENTAL SECTION
■
Computational Methods. The systems have been optimized
using the Gaussian 0927 package at the B3LYP28 computational level
with the 6-31+G*29 basis sets. Frequency calculations have been
performed at the same computational level to confirm that the
resulting optimized structures are energetic minima. Effects of water
solvation have been included by means of the SCRF-PCM approaches
implemented in the Gaussian 09 package including dispersion,
repulsion, and cavitation energy terms of the solvent in the
optimization. The electron density of the complexes has been analyzed
within the atoms in molecules (AIM) theory30 to confirm the
formation of hydrogen bonds.
1
mp 216−218 °C; H NMR (DMSO-d6) δ 1.30 (s, 18H, 2(CH3)3),
1.51 (s, 9H, (CH3)3), 3.77 (s, 4H, CH2 Im), 6.83 (d, 2H, J = 8.5 Hz,
Ar.), 7.59 (d, 2H, J = 8.5 Hz, Ar.), 7.63 (d, 2H, J = 8.5 Hz, Ar.), 7.91
(d, 2H, J = 8.5 Hz, Ar.), 9.70 (brs, 1H, NH), 9.96 (brs, 1H, NH),
10.14 (brs, 1H, NH); 13C NMR (DMSO-d6) δ 28.8, (2(CH3)3), 29.4
((CH3)3), 44.2 (CH2 Im), 80.8 (C(CH3)), 82.7 (C(CH3)), 118.4,
122.0, 122.1, 129.4, 129.8, 135.1, 140.2, 143.8, (Ar.), 145.6, 151.0,
(CO), 153.9 (CN), 165.7 (CO). νmax (film)/cm−1: 3364, 3276 (NH),
1731, 1710, 1655, 1594 (CO, CN).
4-[N′,N″-Di(tert-butoxycarbonyl)-2-aminoimidazolino]-N-[4-
N′,N″-di(tert-butoxycarbonyl)guanidinophenyl]benzamide
(13). Following method 1 from 11 and with purification by flash
chromatography in alumina (hexane/ethyl acetate, 1:1), a white solid
Compounds, DNA, and Buffers. The syntheses of compounds 1,
2, 3, 4, and 5 have been previously described by us.3 The identity of all
compounds was determined by 1H and 13C NMR, IR, and HRMS and
their purity assessed by elemental analysis. Natural salmon sperm
DNA and poly(dA·dT)2 were purchased and used as received. The
concentration of the DNA solutions was determined spectrophoto-
metrically using an extinction coefficient of 6600 cm−1 M−1. Phosphate
buffer contained 10 mM Na2HPO4/NaH2PO4 adjusted to pH 7 and
was prepared using Millipore water.
1
was obtained (65%): mp 112−114 °C; H NMR (DMSO-d6) δ 1.28
(s, 18H, (CH3)3 Im), 1.47 (s, 9H, (CH3)3), 1.52 (s, 9H, (CH3)3), 3.79
(s, 4H, CH2 Im), 6.93 (d, 2H, J = 8.5 Hz, Ar.), 7.49 (d, 2H, J = 9.0 Hz,
Ar.), 7.76 (d, 2H, J = 9.0 Hz, Ar.), 7.88 (d, 2H, J = 8.5 Hz, Ar.), 9.97
(brs, 1H, NH), 10.08 (brs, 1H, NH), 11.47 (brs, 1H, NH); 13C NMR
(DMSO-d6) δ 27.5 ((CH3)3 Im), 27.7, 27.9 ((CH3)3 Gu), 43.0 (CH2),
81.7 (C(CH3)3 Gu), 82.4 (C(CH3)3 Gu), 83.1 (C(CH3)3 Im), 119.5,
120.4, 123.3, 127.4, 128.4, 131.9, 136.5, 140.2 (Ar.), 152.1, 152.2,
153.0, 162.8, 164.9, 170.4 (CO, CN). νmax (film)/cm−1: 3386, 3252,
3155 (NH), 1764, 1731, 1717, 1668, 1645, 1634 (CO, CN).
Chemistry. All the commercial chemicals were obtained from usual
suppliers and were used without further purification. Dry solvents were
prepared using standard procedures, according to Vogel, with
distillation prior to use. Chromatographic columns were run using
silica gel 60 (230−400 mesh ASTM) or aluminum oxide (activated,
Neutral Brockman I STD grade 150 mesh). Solvents for synthetic
purposes were used at GPR grade. Analytical TLC was performed
using silica gel plates or aluminum oxide plates. Visualization was by
UV light (254 nm). NMR spectra were recorded in a spectrometer
General Procedure for the Synthesis of the Hydrochloride
Salts: Method 2. Each of the corresponding Boc-protected
precursors (0.5 mmol) was treated with 15 mL of a 50% solution of
trifluoroacetic acid in DCM for 3 h. After that time, the solvent was
eliminated under vacuum to generate the trifluoroacetate salt. This salt
was dissolved in 20 mL of water and treated for 24 h with IRA400
1
operating at 400.13 and 600.1 MHz for H NMR and at 100.6 and
4404
dx.doi.org/10.1021/jm300296f | J. Med. Chem. 2012, 55, 4397−4406

Journal of Medicinal Chemistry
Article
Amberlyte resin in its Cl− form. Then the resin was removed by
filtration and the aqueous solution washed with DCM (2 × 10 mL).
Evaporation of the water afforded the pure hydrochloride salt. Absence
of the trifluoroacetate salt was checked by 19F NMR.
Dihydrochloride Salt of 4-(2-Aminoimidazolino)-N-(4-
guanidinophenyl)benzamide (8). Following method 2 starting
from compound 13, a white solid was obtained (94%): mp 220 °C
(dec); 1H NMR (D2O) δ 3.66 (s, 4H, CH2), 7.24 (d, 2H, J = 7.0 Hz,
Ar.), 7.41 (d, 2H, J = 8.5 Hz, Ar.), 7.51 (d, 2H, J = 8.5 Hz, Ar.), 7.89
(d, 2H, J = 8.5 Hz, Ar.); 13C NMR (D2O) δ 42.7 (CH2), 122.6, 123.6,
123.8, 125.1, 132.2, 133.4, 135.2, 135.9 (Ar.), 158.7 (CN), 168.4
(CO); HRMS (ESI+) m/z 338.1729 calcd [M+ + H]; found 338.1734.
Anal. (C17H21Cl2N7O·2.7H2O) C, H, N.
maximum response per ligand bound (RUmax) was calculated as
previously described.17 The binding constants were obtained from
fitting RUobs vs free ligand concentration using RUobs = RUmax(K1L +
2K1K2L2)/(1 + K1L + K1K2L2) (L= ligand concentration in the flow
solution).
Circular Dichroism (CD). CD spectra were collected using a
JASCO J-800 spectrometer with a 1 cm cell and a scan speed of
50 nm min−1. The spectra from 400 to 200 nm were averaged over six
scans. A buffer baseline was collected in the same cuvette and subtracted
from each of the averaged scans. The DNA concentration was 37.5 μM in
base. Titrations were carried out by sequentially adding a solution of the
relevant compound (0.5 mM) (with a Bp/D ratio from 22.4 to 0.56) to
the buffered DNA concentration solution at 25 °C.
Dihydrochloride Salt of 4-Amino-N-(4-guanidinophenyl)-
benzamide (9). Following method 2 starting from compound 11, a
white solid was obtained (95%): mp 220 °C (dec); 1H NMR (D2O), δ
8.10 (d, 2H, J = 7.5 Hz, Ar.), 8.17 (d, 2H, J = 7.5 Hz, Ar.), 8.23 (d, 2H,
J = 7.5 Hz, Ar.), 8.42 (d, 2H, J = 7.5 Hz, Ar.); 13C NMR (D2O) δ
109.9, 117.4, 122.1, 123.3, 126.3, 128.9, 130.9, 135.8 (Ar.), 155.7
(CN), 168.0 (CO); HRMS (ESI+) m/z 270.1355 calcd [M+ + H];
found 270.1342. Anal. (C14H17Cl2N5O·1.0H2O) C, H, N.
Isothermal Titration Calorimetry (ITC). ITC experiments were
performed using a MicroCal VP-ITC instrument (MicroCal Inc.,
Northampton, MA) interfaced with a computer equipped with VP-
2000 viewer instrument control software. ITC data were analyzed with
Origin 7.0 software. In ITC experiments, an amount of 1.5 μL of 2.48
mM compound solution in 10 mM phosphate buffer was injected
every 300 s for a total of 29 injections into a solution of DNA in the
calorimeter cell at 2 mM. The observed heat for each injection (peak)
was measured by area integration of the power peak with respect to
time. ITC data were fit according to a standard model that assumes a
single set of equivalent binding sites.
Dihydrochloride Salt of 4-Amino-N-[4-(2-aminoimidazolino)-
phenyl]benzamide (10). Following method 2 starting from
compound 12, a white solid was obtained (92%); mp 210 °C (dec);
1H NMR (D2O) δ 3.69 (s, 4H, CH2), 7.25−7.31 (m, 4H, Ar.), 7.52
(d, 2H, J = 8.5 Hz, Ar.), 7.81 (d, 2H, J = 8.0 Hz, Ar.); 13C NMR
(D2O) δ 42.7 (CH2), 123.0, 123.6, 126.6, 129.2, 131.3, 131.7, 136.4,
138.9 (Ar.), 158.1 (CN), 168.4 (CO); HRMS (ESI+) m/z 296.1511
calcd [M+ + H]; found 296.1510. Anal. (C16H19Cl2N5O·1.3H2O)
C, H, N.
ASSOCIATED CONTENT
* Supporting Information
■
S
Combustion analysis data of the new target compounds, cirular
dichroism spectra, and atomic coordinates of the B3LYP/6-
31+G*-PCM(water) optimized systems (compounds 6, 6 plus
three water molecules, 1, and 4). This material is available free
of charge via the Internet at http://pubs.acs.org.
Thermal Denaturation−UV Spectroscopy. Thermal denatura-
tion experiments were conducted with a Varian Cary 300 Bio
spectrophotometer equipped with a 6 × 6 multicell temperature-
controlled block. Temperature was monitored with a thermistor
inserted into a 1 mL quartz cuvette containing the same volume of
water as in the sample cells. Absorbance changes at 260 nm were
monitored from a range of 20−90 °C with a heating rate of 1 °C/min
and a data collection rate of five points per °C. The salmon sperm
AUTHOR INFORMATION
Corresponding Author
*Phone: +353 1 896 3731. Fax: +353 1 671 2826. E-mail:
rozasi@tcd.ie.
■
DNA was purchased from Sigma Aldrich (extinction coefficient ε260
=
6600 cm−1 M−1 base). A quartz cell with a 1 cm path length was filled
with a 1 mL solution of DNA polymer or DNA−compound complex.
The DNA polymer (150 μM base) and the compound solution
(15 μM) were prepared in a phosphate buffer [0.01 M Na2HPO4/
NaH2PO4], adjusted to pH 7) so that a compound to DNA base ratio
of 0.1 was obtained. The thermal denaturation temperatures of the
duplex or duplex−compound complex obtained from the first
derivative of the melting curves are reported.
Present Addresses
§
́
Departamento de Quımica, Universidad de La Rioja, E-26006
Logrono, Spain.
̃
∥Department of Biochemistry and Molecular Biophysics,
Washington University School of Medicine, St. Louis, MO,
United States.
Notes
Surface Plasmon Resonance (SPR). SPR experiments were
conducted as previously described31 with a BIAcore 2000 instrument
(Biacore AB) using degassed MES buffer (10 mM 2-(N-morpholino)-
ethanesulfonic acid, 1 mM EDTA, 92 mM NaCl, 0.0005% v/v of
surfactant P20, pH 6.25) at 25 °C. The 5′-biotin labeled DNA hairpins
were purchased from Midland Certified Reagent Co., Inc. (Midland,
TX), with HPLC purification. The DNA hairpin sequences included
5′-biotin-CGAATTCGTCTCCGAATTCG-3′, 5′-biotin-
C G T T A A C G T C T C C G T T A A C G - 3 ′ , a n d 5 ′ - b i o t i n -
CGCGCGCGTTTTCGCGCGCG-3′, referred to in the text as
AATT, TTAA, and CG, respectively. The DNA hairpins were
immobilized on a streptavidin-derivatized gold chip (SA chip from
BIAcore) by manual injection of a 25 nM hairpin DNA solution with a
flow rate of 1 μL/min until the response units (RUs) reach about
375−415. Flow cell 1 was left blank for reference subtraction, while
flow cells 2, 3, and 4 were immobilized with three different DNA
hairpins. Typically, a series of different concentrations of ligand was
injected onto the chip at 25 °C with a flow rate of 20 μL/min for a
period of 5 min followed by a 5 min dissociation. After the dissociation
process, the chip surface was regenerated with a 20 μL injection of
200 mM NaCl and 10 mM NaOH solution, injection tube rinsing, and
multiple 1-min buffer injections. The observed steady−state responses,
RUobs, are proportional to the amount of ligand bound, and the
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This research was supported by Science Foundation Ireland
(SFI-CHE275). P.S.N. thanks SFI for generous funding. F.R.
thanks the Consejeria de Educacion Cultura y Deporte de la
Comunidad Autonoma de La Rioja for his grant. We are
indebted to Prof. Amir R. Khan (School of Biochemistry and
Immunology, Trinity College Dublin, Dublin, Ireland) for
providing us with access to and advice for the ITC experiments.
ABBREVIATIONS USED
■
DNA, deoxyribonucleic acid; AT, adenine−thymine base pair;
HB, hydrogen bond; CD, circular dichroism; SPR, surface
plasmon resonance; ITC, isothermal titration calorimetry; P/D,
phosphate to drug ratio; Bp/D, base pair to drug ratio
REFERENCES
(1) Nagle, P. S.; Rodriguez, F.; Kahvedzic,
J. Med. Chem. 2009, 52, 7113−7121.
■
̌
́
A.; Quinn, S.; Rozas, I.
4405
dx.doi.org/10.1021/jm300296f | J. Med. Chem. 2012, 55, 4397−4406

Journal of Medicinal Chemistry
Article
(2) Nagle, P. S.; Rodriguez, F.; Quinn, S. J.; O’Donovan, D. H.; Kelly,
J. M.; Nguyen, B.; Wilson, W. D.; Rozas, I. Org. Biomol. Chem. 2010, 8,
5558−5567.
(26) Rajendran, A.; Zhao, C.; Rajendar, B.; Thiagarajan, V.; Sato, Y.;
Nishizawa, S.; Teramae, N. Biochim. Biophys. Acta 2010, 1800, 599−
610.
(27) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci,
B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian,
H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada,
M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.;
Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.;
Montgomery, J. A., Jr.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.;
Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi,
R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar,
S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, J. M.; Klene, M.; Knox,
J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.;
Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.;
Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.;
Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels,
(3) Rodríguez, F.; Rozas, I.; Kaiser, M.; Brun, R.; Nguyen, B.; Wilson,
W. D.; García, R. N.; Dardonville, C. J. Med. Chem. 2008, 51, 909−
923.
(4) Dardonville, C.; Barrett, M. P.; Brun, R.; Kaiser, M.; Tanious, F.;
Wilson, W. D. J. Med. Chem. 2006, 49, 3748−3752.
(5) Motoshima, K.; Hiwasa, Y.; Yoshikawa, M.; Fujimoto, K.; Tai, A.;
Kakuta, H.; Sasaki, K. ChemMedChem 2007, 2, 1527−1532.
(6) Khalaf, A. I.; Anthony, N.; Breen, D.; Donoghue, G.; Mackay, S.
P.; Scott, F. J.; Suckling, C. J. Eur. J. Med. Chem. 2011, 46, 5343−5355
and references therein..
(7) Neidle, S.; Kelland, L. R.; Trent, J. O.; Simpson, I. J.; Boykin,
D. W.; Kumar, A.; Wilson, W. D. Bioorg. Med. Chem. 1997, 7, 1403−
1408.
(8) Liu, Y.; Kumar, A.; Depauw, S.; Nhili, R.; David-Cordonnier, M.-
H.; Lee, M. P.; Ismail, M. A.; Farahat, A. A.; Say, M.; Chackal-Catoen,
S.; Batista-Parra, A.; Neidle, S.; Boykin, D. W.; Wilson, W. D. J. Am.
Chem. Soc. 2011, 133, 10171−10183.
(9) Mallena, S.; Lee, M. P. H; Bailly, C.; Neidle, S.; Kumar, A.;
Boykin, D. W.; Wilson, W. D. J. Am. Chem. Soc. 2004, 126, 13659−
13669.
̈
A. D.; Farkas, O.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J.
Gaussian 09, revision A.1; Gaussian, Inc.: Wallingford, CT, 2009.
(28) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648−5652. (b) Lee,
C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785−789.
(29) Frisch, M. J.; Pople, J. A.; Binkley, J. S. J. Chem. Phys. 1984, 80,
3265−3269.
(30) Bader, R. F. W. Atoms in Molecules: A Quantum Theory;
Clarendon Press: Oxford, U.K., 1990.
(31) (a) Nguyen, B.; Tanious, F. A.; Wilson, W. D. Methods 2007, 42,
150−161. (b) Tanious, F. A.; Nguyen, B.; Wilson, W. D. Methods Cell
Biol. 2008, 84, 53−77.
(10) Munde, M.; Ismail, M. A.; Arafa, R.; Peixoto, P.; Collar, C. J.;
Liu, Y.; Hu, L.; David-Cordonnier, M.-H.; Lansiaux, A.; Bailly, C.;
Boykin, D. W.; Wilson, W. D. J. Am. Chem. Soc. 2007, 129, 13732−
13743.
(11) Nguyen, B.; Neidle, S.; Wilson, W. D. Acc. Chem. Res. 2009, 42,
11−21.
(12) Nguyen, B.; Lee, M. P. H; Hamelberg, D.; Joubert, A.; Bailly, C.;
Brun, R.; Neidle, S.; Wilson, W. D. J. Am. Chem. Soc. 2002, 124,
13680−13681.
(13) Recent reviews: (a) Kumar, V. A. Eur. J. Org. Chem. 2002,
2021−2032. (b) Nielsen, P. E. Curr. Med. Chem. 2001, 8, 545−550.
(c) Nielsen, P. E. Curr. Opin. Biotechnol. 2001, 12, 16−20.
(d) Uhlmann, E.; Peyman, A.; Breipohl, G.; Will, D. W. Angew.
Chem., Int. Ed. 1998, 37, 2796−2823. (e) Nielsen, P. E. Acc. Chem. Res.
1999, 32, 624−630.
(14) Pilch, S. D.; Poklar, N.; Gelfand, C. A.; Law, S. M.; Breslauer,
K. J.; Barid, E. E.; Dervan, P. B. Proc. Nat. Acad. Sci. U.S.A. 1996, 93,
8306−8311.
(15) Lajiness, J.; Sielaff, A.; Mackay, H.; Brown, T.; Kluza, J.; Nguyen,
B.; Wilson, W. D.; Lee, M.; Hartley, J. A. Med. Chem. 2009, 216−226.
(16) Kahvedzic, A.; Rozas, I. Unpublished results.
(17) Dardonville, C.; Goya, P.; Rozas, I.; Alsasua, A.; Martin, I.;
Borrego, M. J. Bioorg. Med. Chem. 2000, 8, 1567−1577 and references
therein..
(18) (a) Rodriguez, F.; Rozas, I.; Ortega, J. E.; Meana, J. J.; Callado,
L. F. J. Med. Chem. 2007, 50, 4516−4527. (b) Rodriguez, F.; Rozas, I.;
Ortega, J. E.; Erdozain, A. M.; Meana, J. J.; Callado, L. F. J. Med. Chem.
2008, 51, 3304−3312. (c) Rodriguez, F.; Rozas, I.; Ortega, J. E.;
Erdozain, A. M.; Meana, J. J.; Callado, L. F. J. Med. Chem. 2009, 52,
601−609.
(19) Lyng, R.; Rodger, A.; Norden, B. Biopolymers 1992, 32, 1201−
1214.
(20) Rodger, A.; Norden, B. Circular Dichroism and Linear Dichroism;
Oxford University Press: New York, 1997.
(21) (a) Nguyen, B.; Tardy, C.; Bailly, C.; Colson, P.; Houssier, C.;
Kumar, A.; Boykin, D. W.; Wilson, W. D. Biopolymers 2002, 63, 281−
297. (b) Liu, Y.; Kumar, A.; Boykin, D. W.; Wilson, W. D. Biophys.
Chem. 2007, 131, 1−14.
(22) Buurma, N. J.; Haq, I. Methods 2007, 42, 162−172.
(23) Goodsell, D.; Dickerson, R. E. J. Med. Chem. 1986, 29, 727−733.
̌
́
(24) Nagle, P. S.; Kahvedzic, A.; McCabe, T.; Rozas, I. Struct. Chem.
2012, 23, 315−323.
(25) Glass, L. S.; Nguyen, B.; Goodwin, K. D.; Dardonville, C.;
Wilson, W. D.; Long, E. C.; Georgiadis, M. M. Biochemistry 2009, 48,
5943−5952.
4406
dx.doi.org/10.1021/jm300296f | J. Med. Chem. 2012, 55, 4397−4406