Understanding the guanidine-like cationic moiety for optimal binding into the DNA minor groove

Article abstract of DOI:10.1002/cmdc.201402264

Based on our previous positive results with bis-guanidine-like diaromatic compounds as DNA minor groove binders, we propose a new family: bis-2-amino-1,4,5,6-tetrahydropyrimidines. According to calculated parameters, these dicationic systems would have a more suitable size and lipophilicity for binding into the minor groove than previous series. Moreover, their DFT-optimised structures and docking into an AT oligomer model show that they would bind in the minor groove with good strength and without energy penalty. Hence, we prepared compounds 4 a-c and evaluated their binding to ssDNA and poly(dA-dT)₂ by thermal denaturation experiments. The results showed that 4 a (CO) and 4 d (NH) were the best DNA binders. Compared to the previous series, 4 a-d are better binders than bis-guanidiniums but poorer than bis-2-aminoimidazolinium derivatives. Moreover, circular dichroism experiments using ssDNA and poly(dA-dT)₂ confirmed binding into the minor groove. Based on our computational design as well as biophysical studies, we have been able to determine that the optimal interaction of guanidine-like dications in the minor grove occurs with bis-2-aminoimidazolinium systems.

Full text of DOI:10.1002/cmdc.201402264

CHEMMEDCHEM
FULL PAPERS
DOI: 10.1002/cmdc.201402264
Understanding the Guanidine-Like Cationic Moiety for
Optimal Binding into the DNA Minor Groove
Patrick O’Sullivan and Isabel Rozas*[a]
Based on our previous positive results with bis-guanidine-like
diaromatic compounds as DNA minor groove binders, we pro-
pose a new family: bis-2-amino-1,4,5,6-tetrahydropyrimidines.
According to calculated parameters, these dicationic systems
would have a more suitable size and lipophilicity for binding
into the minor groove than previous series. Moreover, their
DFT-optimised structures and docking into an AT oligomer
model show that they would bind in the minor groove with
good strength and without energy penalty. Hence, we pre-
pared compounds 4a–c and evaluated their binding to ssDNA
and poly(dA-dT)2 by thermal denaturation experiments. The re-
sults showed that 4a (CO) and 4d (NH) were the best DNA
binders. Compared to the previous series, 4a–d are better
binders than bis-guanidiniums but poorer than bis-2-aminoimi-
dazolinium derivatives. Moreover, circular dichroism experi-
ments using ssDNA and poly(dA-dT)2 confirmed binding into
the minor groove. Based on our computational design as well
as biophysical studies, we have been able to determine that
the optimal interaction of guanidine-like dications in the minor
grove occurs with bis-2-aminoimidazolinium systems.
Introduction
DNA is a valid target in a multitude of infectious diseases, in-
cluding those caused by bacteria, viruses and parasites.[1] In
particular, the minor groove of DNA is a key recognition site of
many enzymes;[2] hence, compounds that specifically bind with
the minor groove are very promising for the treatment of dis-
eases at DNA level and, accordingly, these compounds have
been widely studied as antibacterial, antiprotozoal, antitumor
and antiviral agents.[3–7]
plained that this activity results not only from hydrophobic in-
teractions, but also from very specific hydrogen bonds (HBs)
established by an amide group present in these molecules.[16]
These are only but a few of the therapeutic applications of
DNA MGBs, which represent one of the most interesting bio-
molecular families of compounds.
During the last ten years, we have been working on the
preparation of symmetric and asymmetric bis-guanidinium-like
diaromatic compounds (1–3, Figure 1) that strongly bind to
the minor groove of DNA with significant AT sequence selectiv-
ity and show antiprotozoal activity.[17,18] Furthermore, their in-
teractions with DNA have been investigated by various bio-
physical techniques such as thermal denaturation, circular di-
chroism (CD), surface plasmon resonance (SPR) and isothermal
titration calorimetry (ITC), and their mode of binding has been
confirmed by linear dichroism (LC) as being DNA MGBs.[19]
It is well known that MGBs, which in general are cationic at
physiological pH values, interact with DNA by means of HBs,
electrostatic interactions with the phosphates on the DNA
backbone, as well as by strong hydrophobic interactions, dis-
placing the spine of hydration.[20,21] Additionally, it is accepted
that a curved structure to optimally fit into the groove is very
beneficial.[22,23,24] Our compounds (1–3, Figure 1) have a diaro-
matic core connected by a linker (X) of different nature that
determines the curvature of the whole structure orientating
the guanidine-like cations on both ends of the molecules to
optimally interact within the groove.
The US Food and Drug Administration (FDA) has approved
two minor-groove binders (MGBs) for therapeutic use: pentam-
idine for the treatment of African trypanosomiasis,[8] Pneumo-
cystis carinii pneumonia and leishmaniasis,[9] and berenil for
animal trypanosomiasis.[10] Additionally, MGBs that have
reached clinical trials for cancer include SJG-136/SG2000,
which is in phase I for advanced solid tumours[11] and recruiting
phases I/II for acute lymphoblastic leukaemia and chronic lym-
phocytic leukaemia.[12] Analogues of an imidazoline-pyrrole
based polyamide MGB have showed efficacy in human papillo-
ma virus strains in vitro.[13] Additionally, in a very interesting
paper, Dervan has described how a sequence-specific DNA-
binding polyamide inhibits the expression of a number of an-
drogen-receptor-regulated genes in prostate cancer cells.[14]
Moreover, distamycin-related MBGs have been found to act as
antibacterial agents;[15] in particular, Suckling and co-workers
have shown a series of bis-alkenyl MGBs with potent activity
against Gram-positive bacteria (Staphylococcus aureus and
Streptococcus faecalis), and in a detailed NMR study has ex-
The crystal structure of one of our DNA MGBs bound to an
AT oligonucleotide was resolved by Glass et al.[25] (PDB code:
3FSI; Figure 2) and, recently, a more refined structure (PDB
code: 4OCD) has been published by Acosta-Reyes et al.[26]
From these studies it is clear that the most important interac-
tions of these MGBs are established by means of the cationic
[a] P. O’Sullivan, Prof. I. Rozas
School of Chemistry, Trinity Biomedical Sciences Institute
Trinity College, University of Dublin, 152-160 Pearse St., Dublin 2 (Ireland)
E-mail: rozasi@tcd.ie
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cmdc.201402264.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 2065 – 2073 2065

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
Figure 1. Rationale behind the design of bis-2-amino-1,4,5,6-dihydropyrimidinium derivatives 4.
date to be incorporated in new MGBs likely increasing their
ability to bind into the DNA minor groove.
It has been shown that structural variations that change the
molecular properties of MGBs (lipophilicity or hydrogen bond-
ing ability)[17,27] can result in a modification of their biological
responses. Therefore, in order to improve the DNA binding of
our diaromatic dications and considering that, in the past we
have mostly modified the nature of the linker (X) within the di-
aromatic core, we present now the modelling, preparation and
DNA affinity evaluation of a series of symmetric derivatives car-
rying a modified cation with increased size and lipophilicity.
Thus, a series of bis-2-amino-1,4,5,6-tetrahydropyrimidinium
derivatives (4, Figure 1) have been prepared and, to investigate
their binding to DNA, different biophysical techniques such as
DNA thermal denaturation and CD have been used to measure
the binding strength and mode of binding of these com-
pounds.
Results and Discussion
Computational design
To understand the difference between the bis-guanidinium-like
diaromatic systems previously prepared and those presently
proposed, we have theoretically calculated, with the help of
the Marvin package,[28] different physicochemical parameters
such as logP or polar surface area (PSA) for all the compounds
now proposed (4a–d) and their corresponding analogues pre-
viously prepared (1a–d and 3a–d), and the results obtained
are presented in Table 1.
Figure 2. Crystal structure of 4,4’-bis(2-aminoimidazoline)diphenylamine (3d)
complexed with the self-complementary nucleotide d(CTTAATTCGAATTAAG)2
(PDB: 3FSI).[25] A hydrogen bond between an imidazolinium (NꢀH) and a thy-
mine (C=O) is indicated. Protons are not shown.
moieties which can form HBs and ionic interactions with the
bases.
Regarding lipophilicity, as expected and as reflected by the
calculated logP values, those derivatives with the 2-amino-
1,4,5,6-tetrahydropyrimidinium cations are more lipophilic than
those with the 2-aminoimidazolinium groups and more than
the bis-guanidinium derivatives (see Table 1). The size of the
cationic groups increases also in the same order as reflected
by the van der Waals surface area of guanidinium (94.08 ꢁ2), 2-
aminoimidazolinium (141.51 ꢁ2) and 2-amino-1,4,5,6-tetrahy-
dropyrimidinium (170.58 ꢁ2). The PSA parameter reflects the
ability of a molecule to form hydrogen bonds (HBs), and for
that reason both cyclic guanidinium-like containing systems (3
Looking at our previous dicationic series 1–3, we have ob-
served an increment in the strength of binding to the minor
groove of DNA from the bis-guanidinium derivatives (1,
Figure 1), passing through the asymmetric guanidinium/2-ami-
noimidazolinium (2, Figure 1) to the bis-2-aminoimidazolinium
compounds (3, Figure 1), suggesting that an increase in bulk
and lipophilicity in the cationic moieties will favour the binding
into the minor groove. Thus, the 2-amino-1,4,5,6-dihydropyri-
midinium ring, which is bulkier and more lipophilic than guani-
dinium or 2-aminoimidazolinium, seems like a suitable candi-
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 2065 – 2073 2066

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
planar while in the case of compounds 4b and 4c they are
perpendicular. In order to fit into the DNA minor groove, we
have previously determined that these aromatic rings have to
turn to become almost coplanar paying an energetic penalty
of around 5.3–9.0 kcalmolꢀ1.[15] Accordingly, compounds 4a
and 4d would have to pay a smaller energy penalty than com-
pounds 4b and 4c since they are already almost coplanar
before fitting into the DNA minor groove. Hence, we could
predict that compounds 4a and 4d would be the ones better
binding to DNA.
Table 1. Theoretical logP and PSA values calculated with the Marvin
package[19] for the three series of bis-guanidinium-like derivatives studied
(1, 3 and 4).
Cation[a] (Compd)
Linker (X)
CO
logP
PSA
Gua (1a)
Imi (3a)
Thp (4a)
Gua (1b)
Imi (3b)
Thp (4b)
Gua (1c)
Imi (3c)
Thp (4c)
Gua (1d)
Imi (3d)
Thp (4d)
1.38
2.04
2.16
1.42
2.08
2.20
2.01
2.67
2.79
1.36
2.02
2.14
140.87
89.91
89.91
133.03
82.07
82.07
123.80
72.84
72.84
135.83
84.87
84.87
O
Next, using Autodock we have docked these four structures
into a model of a DNA minor groove using as template the
crystallographic structure published for the bis-2-aminoimida-
zolinium diaromatic system with a NH linker (3d) and the
d(CTTAATTCGAATTAAG)2 nucleotide (see a superimposition of
the four ligands dock with the AT oligomer in Figure 4).[25] The
CH2
NH
[a] Gua=guanidinium,
Imi=2-aminoimidazolinium,
Thp=2-amino-
1,4,5,6-tetrahydropyrimidinium.
and 4) show similar PSA values since the HB donor/acceptor
pattern is exactly the same for 2-aminoimidazolinium and 2-
amino-1,4,5,6-tetrahydropyrimidinium (see Table 1). These
values are smaller than those of the guanidinium-containing
derivatives 1, because these compounds can form extra HB in-
teractions.
Additionally, we have calculated the possible orientation of
the cationic moieties with respect to the diaromatic core for
the bis-cationic derivatives 4a–d (Figure 3) by means of densi-
ty functional theory (DFT, B3LYP/6-31G* level) calculations. The
2-amino-1,4,5,6-tetrahydropyrimidinium moiety present an en-
velope-type conformation with the 2-NH group as well as the
C6–N1–C2–N3–C4 moiety in the same plane and the CH2 in
position 5 out of the plane. In all cases, different conformations
of minimum energy were found differing mostly in the orienta-
tion of the cationic rings, well towards the “front” of the aro-
matic ring to which they are attached, or towards the “back”
of such aromatic ring. In all four cases, a conformer with the
two cations towards the same side of the diaromatic core was
found indicating a good disposition to fit into the minor
groove (Figure 3).
Figure 4. Rotated views of the optimised compounds 4a–d (stick models)
docked into the d(CTTAATTCGAATTAAG)2 nucleotide.
diaromatic portions of the ligands have been rotated to corre-
spond to the coplanar system observed within the original
crystal (see Figure 1). The best pose adopted for each of the
4a–d compounds showed very similar docking scores inde-
pendent of the linker X (around ꢀ8.4 kcalmolꢀ1) indicating
that such a linker does not play any role in binding to the
minor groove.
A superimposition of the four bis-2-amino,1,4,5,6-tetrahydro-
pyrimidinium derivatives (4a–d) docked within the
d(CTTAATTCGAATTAAG)2 nucleotide is presented in Figure 4
showing that despite the large
Regarding the diaromatic core, it is interesting to notice that
for compounds 4a and 4d the two phenyl rings are almost co-
size of the dicationic moieties,
there are no steric clashes, and
these compounds fit well into
this model of the minor groove.
In the crystal structure of 3d
within this AT oligomer (see
Figure 1), HBs are observed be-
tween the amino groups of the
imidazolinium system and some
of the base pairs of the oligo-
mer; however, in the case of
4a–d, even though two of the
NH of the 2-amino-1,4,5,6-tetra-
Figure 3. Examples of the ‘front’–‘front’ conformers of minimum energy calculated for 4a–d at B3LYP/6-31G level.
hydropyrimidine cations point
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 2065 – 2073 2067

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
towards the nucleic bases, no HBs were observed.
the ketone reduced before the pyrimidine ring yielding com-
pound 4c (X=CH2). It is well known that benzophenone un-
dergoes facile reduction under a range of conditions including
a LiAlH4/AlCl3 mixture,[31] catalytic I2 in H3PO2,[32] Zn in AcOH[33]
and H2, Pd/C;[34] thus, conversion of 7a into 4c was not surpris-
ing.
Summarising, our computational modelling indicates that
considering size and lipophilic properties, structural arrange-
ment of the cations and fitting into a model of the minor
groove, the proposed 2-amino-1,4,5,6-tetrahydropyrimidinium
derivatives 4a–d compare well with previously prepared
minor-groove binders (MGBs; 1–3a–d) and should bind into
the DNA minor groove with good strength.
The synthesis of 4d (X=NH) proved to be less trivial. Al-
though 6d was commercial available, it did not form 7d under
the mentioned conditions (see conditions c) in Scheme 1) due
to competition with the NH linker. Therefore, different amino-
protecting groups (pivaloyl, carboxybenzyl) were explored, but
these protected analogues also failed to react with 2-amino-
pyrimidine. Initial reactions of Boc-protected 6d’ (Scheme 1)
with 2-aminopyrimidine only yielded the corresponding depro-
tected diamine, and systematic alteration of the conditions
(i.e., using K3PO4 as base, Pd(Ph3)4 as Pd source, dioxane or
THF as solvents) did not produce the expected results. Howev-
er, employing 6d’ at lower temperature (808C) and using
higher catalyst loading (10 mol% Pd2dba3, 15 mol% Xantphos)
provided 7d’. Interestingly, under acidic hydrogenation condi-
tions, Boc-deprotection was preceded by hydrogenation of the
pyrimidine ring isolating both 4d and Boc-protected 4d’
(Scheme 1). Next 4d’ was dissolved in methanol and stirred at
room temperature in excess of 3m HCl affording 4d.
Chemistry
As informed by our computational modelling, we have first
prepared compounds 4a–c following a methodology recently
developed in our group (Scheme 1).[29] Initially, dianilines 5a,
b (X=CO and O, respectively) were converted to the corre-
sponding dibromides 6a, b by means of a Sandmeyer reac-
tion.[30] Then, Buchwald–Hartwig conditions were utilised to
convert these dibromides into bis-2-aminopyrimidines 7a,
b using Pd2dba3 (dba=dibenzylideneacetone) and Xantphos.
Palladium-catalysed reduction under acidic conditions in a H2
atmosphere at room temperature afforded the corresponding
salt 4b (Scheme 1). Attempts to isolate 4a in this way failed as
To access compound 4a (X=CO), we returned to our stan-
dard HgCl2 promoted guanidylation and Boc deprotection
methodology used in the previous syntheses of N-arylguani-
dines and N-aryl-2-aminoimidazolines by means of reaction
with N,N’-bis-tert-butoxycarbonylthiourea and N,N’-bis-tert-bu-
toxycarbonylimidazolidine-2-thione respectively.[35] Thus, we
first prepared the appropriate Boc-thione 9 from commercially
available 8 (Scheme 2). Then, HgCl2 promoted reaction of 9
with 4,4’-diaminobenzophenone, which is the poorest dianiline
nucleophile of the four starting diamines, yielded the Boc-pro-
tected derivative 10 in poor yield. After a straightforward pu-
rification, HCl-mediated Boc deprotection furnished salt 4a in
high purity (Scheme 2).
Scheme 1. General synthesis of compounds 4b–d, d’. Reagents and condi-
tions: a) HBr/H2O, D, overnight; NaNO2, H2O, 08C, 15 min; CuBr, HBr, 08C!
708C, 3 h; b) Boc2O, cat. DMAP, THF, D, overnight; c) Pd2dba3 (2 mol%), Xant-
phos (3 mol%), NaOtBu, 2-aminopyrimidine, toluene, D, overnight; d) H2, Pd/
C (10%), 1m HCl, CH3OH, overnight; e) MeOH, excess 3m HCl.
Scheme 2. General synthesis of compound 4a. Reagents and conditions:
a) NaH, THF, 08C, 45 min; Boc2O, 08C!RT, overnight; b) 4,4’-diaminobenzo-
phenone, HgCl2, Et3N, DMF/CH2Cl2, overnight; c) 4m HCl/dioxane, D, 8 h.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 2065 – 2073 2068

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
DNA thermal denaturation
2-amino-1,4,5,6-tetrahydropyrimidine cation, we were expect-
ing that this new family 4 would result in better binding to the
DNA minor groove. In the present study, we have used both
ssDNA and poly(dA-dT)2, and the results obtained (Table 2)
seem to indicate that derivatives 4, even though better than
the bis-guanidiniums, are not as good DNA binders as the bis-
2-aminoimidizolidiniums 3.
Finally, considering the nature of the DNA used in these ex-
periments, the four compounds 4a–d exhibit a stronger bind-
ing to the poly(dA-dT)2 oligomer
Thermal denaturation experiments are a very useful tool to
provide information on the strength of binding of a compound
to DNA. Accordingly, we have performed thermal denaturation
experiments using unspecific salmon sperm DNA (Tm =68.08C)
and poly(dA-dT)2 (Tm =46.38C) with the four bis-2-amino-
1,4,5,6-tetrahydropyrimidiniums 4a–d, and the results (DTm)
are presented in Table 2 and Figure 5.
than to the unspecific ssDNA, in-
Table 2. DNA binding affinity (DTm, 8C) for bis-guanidinium-like diaromatic compounds 1 (bis-guanidine), 2
(guanidine/2-aminoimidazoline), 3 (bis-2-aminoimidazoline) and 4 (bis-2-amino-1,4,5,6-tetrahydrocyclohexane).
dicating that the four com-
pounds have selectivity for AT
sequences and are likely to be
DNA MGBs.
X
DTm (4)
DTm (1)
DTm (2)
ssDNA[b]; poly(dA)-poly(dT)[b]
DTm (3)
[a]
[a]
ssDNA; poly(dA-dT)2
poly(dA-dT)2
poly(dA-dT)2
CO (a)
O (b)
CH2 (c)
NH (d)
9.3ꢁ0.3; 14.2ꢁ0.2
6.2ꢁ0.2; 8.0ꢁ0.2
3.8ꢁ0.4; 5.5ꢁ0.1
7.2ꢁ0.3; 12.3ꢁ0.1
27.6
22.1
15.0
29.6
8.0; 32.0
7.0; 25.3
6.0; 17.2
12.0; 35.1
–
27.1
18.8
38.5
Circular dichroism experiments
Circular dichroism (CD) experi-
ments can be used to determine
binding and conformational
changes of polynucleotides such
as DNA. It has been shown that
a peak at 245 nm indicates helic-
ity of B-DNA and a peak at
275 nm relates to p–p stacking
of the DNA bases. If a compound
intercalates into DNA, the CD
spectrum does not change sig-
nificantly because the p–p stack-
ing interactions are in the same
region as the p–p stacking of
DNA base pairs. However, com-
pounds that bind into the minor
groove of DNA can induce
a new peak in the spectrum,
arising from the coupling of
[a] Taken from Ref. [19]; [b] Taken from Ref. [18].
Figure 5. Thermal melting plots for 150 mm of a) ssDNA (c) and b) poly(dA-dT)2 (c), both in the presence of
15 mm 4a (a), 4b (c), 4c (a) and 4d (c).
In terms of the linker X, previous biophysical studies with
derivatives 1–3 showed that compounds with a NH linker ex-
hibit the strongest DNA binding whereas those with a CO
linker displayed weaker binding. However, in the present study
and independently of the DNA used (unspecific or AT specific),
compound 4a (X=CO) has shown to be the best binder fol-
lowed by 4d (X=NH) (Table 2 and Figure 5). These results are
in agreement with those obtained in the computational study,
which indicated that the diaromatic moiety of compounds 4a
and 4d was almost coplanar outside the minor groove imply-
ing a lower energy penalty than those compounds with a per-
pendicular diaromatic core such as 4b and 4c.
Regarding the cationic systems, in previous studies with bis-
guanidines (1), asymmetric guanidine/2-aminoimidazolines (2)
and bis-2-aminoimidazolines (3) (see Figure 2), compounds of
type 1 were the weakest DNA binders while compounds 3
were the strongest ones (Table 2). Even though the oligonucle-
otides used in those studies were not always the same, a linear
correlation was found among all sets of values.[36] Based on our
assessment of size, lipophilicity and hydrogen bonding of the
electronic transition moments of the ligand and DNA bases in
an achiral environment. Thus, a strong induced CD signal (ICD)
indicates minor-groove binding.[37,38]
All the symmetric dications studied here are achiral with no
inherent CD signals; however, in the presence of DNA, they
may induce a signal. Thus, CD titrations were performed for
4a–d by increasing the compound to ssDNA Bp/D (base pair
per drug/ligand) ratio from 100 to 10, and the results are pre-
sented in Figure 6. Interpretation of CD spectra in the region
of DNA absorbance is difficult due to contributions arising
from both changes in DNA conformation, resulting from ac-
commodation to the ligand, and changes in transitions of the
ligand due to the new environment. For this reason, we have
analysed the ICD spectra and focussed on the region of the
spectrum >260 nm. In the particular cases shown in Figure 6,
the maximum absorption occurred at around 320 nm for 4a
and 4d. Upon increasing addition of these compounds to
ssDNA, a growth in this 320 nm band is observed confirming
that 4a and 4d are both MGBs. In the case of the other two
dications (4b and 4c), no ICD signal at 320 nm is observed;
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 2065 – 2073 2069

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
acting far from the chiral sugar
backbone, may have
a low
signal yet bind very tightly,
whereas molecules interacting in
the minor groove will induce
a greater CD signal, even if they
actually have a weak affinity.
Considering the ICD results as
well as the DTm values using
both unspecific (ssDNA) and AT-
specific DNA [poly(dA-dT)2],
compounds 4b and 4c seem to
be MGBs. Despite not producing
a clear ICD, they show AT selec-
tivity in the thermal denatura-
tion experiments and an in-
crease in the DNA CD band. In
the case of 4a and 4d, we can
clearly conclude from our results
that they bind into the DNA
minor groove with a medium to
strong potency.
Conclusion
Figure 6. ICD spectra obtained for 4a--d titrated with ssDNA in a concentration of 150 mm varying the Bp/D ratio
from 100 to 10 through nine additions.
Considering the encouraging re-
sults previously obtained with
symmetric and asymmetric bis-
guanidinium-like
diaromatic
compounds as DNA minor groove binders, we pro-
posed a new family, bis-2-amino-1,4,5,6-tetrahydro-
pyrimidinium diaromatic derivatives, based on
a series of computational studies. First, we assessed
the size, lipophilicity and hydrogen-bonding charac-
teristics of these compounds, then, we optimised
their structures at B3LYP/6-31G* level to understand
their potential fitting into the minor groove and, fi-
nally, we docked them into an AT-oligomer model. All
these studies indicate that the proposed 2-amino-
Figure 7. ICD spectra obtained for 4a–d titrated with poly(dA-dT)2 in a concentration of
37.5 mm varying the Bp/D ratio from 19.2 to 2.5.
1,4,5,6-tetrahydropyrimidinium
derivatives
4a–d
should bind into the DNA minor groove with good
strength.
nonetheless, addition of these two compounds changes the
Accordingly, we prepared compounds 4a–c following
absorption of the 275 nm band.
a methodology recently developed in our group consisting on
the transformation of dianilines (5a, b) into the corresponding
dibromides (6a, b) by means of a Sandmeyer reaction, and
then using Buchwald–Hartwig conditions (Pd2dba3 and Xant-
phos) to convert these dibromides into bis-2-aminopyrimidines
(7a, b) which were reduced using Pd catalysed hydrogenation
under acidic conditions to 4b (X=O) and, because of the re-
duction of the CO linker, to 4c (X=CH2). Preparation of 4d
(X=NH) required the Boc-protection of the NH linker of the di-
bromide 6d, and then, using Pd2dba3 and Xantphos, the corre-
sponding Boc-protected bis-pyrimidine (7d’) was obtained,
which after acidic hydrogenation yielded 4d. Finally, we pre-
pared 4a (X=CO) by reaction of the appropriate Boc-thione 9
with 4,4’-diaminobenzophenone in the presence of HgCl2,
The binding mode of 4a and 4d (Bp/D ratio: 12.9–2.5) was
further investigated by measuring their induced CD in the
presence of poly(dA-dT)2 (37.5 mm). The corresponding ICD
plots (Figure 7) display a clear induced peak at 310 nm. These
data are in agreement with the ICD spectra of 4a and 4d with
ssDNA, implying that the binding mode of these compounds
with both types of DNA is the same. Molecules that bind pref-
erentially to AT-rich DNA are usually MGBs and, taking into ac-
count the similarity of ICD spectra with previous known
MGBs,[17–19] we can conclude that these compounds bind into
the minor groove.
The magnitude of the induced CD signal is related to the
chiral environment. Hence, compounds that intercalate, inter-
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 2065 – 2073 2070

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
purity was confirmed using a purity channel. The stationary phase
consisted of an ACE 5 C18-AR column (150ꢂ4.6 mm), and the
mobile phase used the following gradient system, eluting at
1.0 cm3minꢀ1: aq formate buffer (30 mm, pH 3.0) for 10 min, linear
ramp to 85% MeOH buffered with the same system over 25 min,
held at 85% buffered MeOH for 10 min.
yielding the Boc-protected derivative 10 that was deprotected
using HCl to produce 4a.
The four compounds were used for DNA thermal denatura-
tion experiments using salmon sperm DNA and poly(dA-dT)2
to evaluate their DNA binding and AT selectivity, respectively.
Thus, we found that the four compounds 4a–d were AT selec-
tive, and 4a (X=CO) and 4d (X=NH) are the stronger binders
to ssDNA with DTm values of 9.3 and 7.28C, and to poly(dA-
dT)2 with DTm =14.2 and 12.38C, respectively. Compared to the
bis-guanidinium (1), asymmetric guanidinium/2-aminoimidazo-
linium (2) and bis-2-aminoimidazolinium derivatives (3), these
NMR spectra and HPLC plots for final compounds 4a–d and N-Boc-
protected derivative 4d’, CD spectra, B3LYP/6–31G optimised struc-
tures and preparation and characterisation of intermediates is
available in the Supporting Information.
General method A: The corresponding diamine (3.289 mmol) was
heated in HBr (48% in H2O, 5 mL) at 1008C for 12 h until a solution
was formed. Upon cooling to 08C, a solution of NaNO2 (480 mg,
6.958 mmol) in H2O (5 mL) was added dropwise. This solution was
in turn added dropwise to a solution of CuBr (1.400 g, 9.676 mmol)
in HBr (48% in H2O, 5 mL), and the solution was heated slowly to
708C over 2 h and stirred at 708C for up to 3 h until evolution of
gas ceased. To dissolve organic solids that precipitated, it was nec-
essary to add Et2O (0.5 mL) as the mixture reached 508C. The mix-
ture was cooled and diluted with EtOAc (20 mL) and H2O (20 mL).
The layers were separated, and the aqueous layer was extracted
with EtOAc (2ꢂ20 mL). The organic phases were combined and
washed with saturated NaHCO3 (2ꢂ20 mL), H2O (2ꢂ20 mL) and
brine (2ꢂ20 mL). The organic layer was dried over MgSO4, filtered,
and the solvent was evaporated and purified on silica gel to yield
the appropriate dibromide.
bis-2-amino-1,4,5,6-tetrahydropyrimidinium
derivatives 4a–d
are better binders than 1, similar to derivatives 2, but poorer
binders than 3.
Finally, to understand the mode of binding of these new de-
rivatives, circular dichroism (CD) experiments were carried out
using salmon sperm DNA and poly(dA-dT)2. In the case of 4b
and 4c, an induced CD signal for DNA was observed indicating
that these compounds bind to DNA and their DTm values in
the AT oligomer seem to indicate minor groove binding. For
4a and 4d, a strong positive induced CD signal was detected
both for ssDNA and the AT oligomer, an indication of binding
into the minor groove.
All the data presented here indicate that DNA binding for
these dicationic systems is optimal for compounds with linkers
CO or NH, which have logP values around 2.04 and a van der
Waals surface around 141.5 ꢁ2. The size and properties of the
cationic moieties of these guanidine-like diaromatic minor-
groove binders have an optimal limit, that of the 2-aminoimi-
dazolinium cation.
General method B: Toluene (1.5 mL) was added to a vacuum-dried
mixture of Pd2(dba)3 (21.7 mg, 0.023 mmol), Xantphos (27.4 mg,
0.047 mg), NaOtBu (170 mg, 1.776 mmol), 2-aminopyrimidine
(169 mg, 1.776 mmol) and the appropriate dibromide (0.592 mmol)
under an atmosphere of Ar at 908C, and the mixture was stirred
for 16 h. The flask was cooled and diluted with EtOAc (50 mL), fil-
tered through Celite and separated with H2O (20 mL). The aqueous
layer was extracted again with EtOAc (2ꢂ20 mL), and the com-
bined organic layers were washed with brine, dried over Na2SO4.
The solvent was removed in vacuo and purified by silica gel chro-
matography in a hexane/EtOAc mixture.
Experimental Section
Chemistry
General materials and methods: All chemicals were obtained
from Sigma–Aldrich or Fisher and were used without further purifi-
cation. Phosphate buffer solutions contained 10 mmK2HPO4/
KH2PO4 adjusted to pH 7 and were prepared using Millipore H2O.
Deuterated solvents for NMR use were purchased from Apollo. 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). Solvents for
synthetic purposes were used at general purpose reagent (GPR)
grade. Analytical thin-layer chromatography (TLC) was performed
using Merck silica gel 60 F254 silica gel plates. Visualisation was per-
formed using UV light (254 nm). NMR spectra were recorded in
Bruker DPX-400 Avance spectrometers operating at 400.13 MHz
and 600.1 MHz for 1H NMR and 100.6 MHz and 150.9 MHz for
13C NMR. Shifts are referenced to the internal solvent signals. NMR
data were processed using MestReNova software. HRMS spectra
were measured on a Micromass LCT electrospray TOF instrument
with a Waters 2690 autosampler with MeOH as carrier solvent.
Melting points were determined using a Stuart Scientific Melting
Point SMP1 apparatus and are uncorrected. Infrared spectra were
recorded on a PerkinElmer Spectrum One FT-IR Spectrometer
equipped with a Universal ATR sampling accessory. HPLC purity
analysis was carried out using a Varian ProStar system equipped
with a Varian Prostar 335 diode array detector and a manual injec-
tor (20 mL). UV detection was performed at 245 nm, and peak
General method C: Pd/C (10%, 82 mg) and aq HCl (1m, 0.5 mL)
was added to a solution of the appropriate bis-2-aminopyrimidine
diaromatic derivative (0.136 mmol) in degassed CH3OH (1 mL). The
mixture was stirred vigorously under an atmosphere of H2 for 16 h
at RT, diluted with CH3OH, filtered and concentrated to directly
yield the bis(2-amino-1,4,5,6-tetrahydropyrimidinium)dichloride
salts.
4,4’-Bis(2-amino-1,4,5,6-tetrahydropyrimidine)benzophenone di-
hydrochloride (4a): HCl (4m in dioxane, 1.84 mL, 7.35 mmol) was
added to Boc-protected derivative 10 (238 mg, 0.31 mmol), and
the mixture was stirred at 558C for 8 h before evaporating to dry-
ness. The product was dissolved in H2O (2.0 mL) and washed with
CH2Cl2 (3ꢂ5 mL). After evaporation of the aqueous layer, the prod-
uct was purified on reverse-phase silica, eluting in 100% H2O as
a white solid (94 mg, 69%): mp: 1838C; 1H NMR (400 MHz, CDCl3):
d=2.04 (quin., J=5.8 Hz, 4H), 3.44 (t, J=5.8, 8H), 7.41 (d, J=
8.6 Hz, 4H), 7.87 ppm (d, J=8.6 Hz, 2H); 13C NMR (100 MHz, D2O):
d=19.2, 38.4, 122.7, 132.0, 133.8, 140.0, 151.7, 198.3 ppm; IR (ATR):
nmax =2982, 2891, 1624, 1228, 1147, 1032, 963, 825, 762 cmꢀ1
;
HRMS (ESI): m/z [M+H]+ calcd for C21H25N6O: 377.2084, found:
377.2088; RP-HPLC (254 nm): tR =20.7 min, purity >99%.
4,4’-Bis(2-amino-1,4,5,6-tetrahydropyrimidine)diphenylether di-
hydrochloride (4b): Method C, employing 4,4’-bis(2-aminopyrimi-
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 2065 – 2073 2071

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
dine)diphenylether 7b (42.5 mg, 0.119 mmol), yielded 4b as a gum
were taken at an interval of 1 nm, scanning at 200 nmminꢀ1. Data
are an average of three accumulations. The ICD spectra were ob-
tained by subtracting the CD spectrum of the compound with
DNA from the spectrum of DNA in buffer alone. The data were
smoothed using an FFT filter with a cut-off frequency of 0.25.
(44 mg, 85%), which eluted in reverse-phase silica as a yellow solid
1
(44 mg, 85%): mp: 1838C; H NMR (400 MHz, D2O): d=1.99 (quin.,
J=5.7 Hz, 4H), 3.36–3.39 (m, 8H), 7.15 (d, J=8.7, 4H) 7.30 ppm (d,
J=8.7 Hz, 4H); 13C NMR (100 MHz, D2O): d=19.3, 38.2, 120.0, 126.6,
128.0, 152.5, 155.6 ppm; IR (ATR): nmax =2982, 2891, 1711, 1624,
1490, 1317, 1249, 1228, 1154, 1032, 824 cmꢀ1; HRMS (ESI): m/z
[M+H]+ calcd for C20H25N6O: 365.2084, found 365.2075; RP-HPLC
(254 nm): tR =21.3 min, purity >99%.
Computational methods
Structure optimisation: All compounds proposed (4a–d) were op-
timised using the Gaussian09 package[39] at the B3LYP[40] computa-
tional level with the 6–31G* basis set.[41] Frequency calculations
have been performed at the same computational level to confirm
that the resulting optimized structures are energetic minima.
4,4’-Bis(2-amino-1,4,5,6-tetrahydropyrimidine)diphenylmethane
dihydrochloride (4c): Method C, employing bis-pyrimidine 7a
(50 mg, 0.271 mmol), yielded 4c as a grey gum after purification in
reverse-phase silica, eluting in 100% H2O (41 mg, 83%): 1H NMR
(400 MHz, D2O): d=1.82 (quin., J=5.8 Hz, 4H), 3.20 (t, J=5.8 Hz,
8H), 3.89 (s, 2H), 7.06 (d, J=8.2 Hz,4H), 7.22 ppm (d, J=8.2, 4H);
13C NMR (100 MHz, D2O): d=19.3, 38.3, 40.1, 126.1, 130.1, 132.2,
140.9, 152.5 ppm; IR (ATR): nmax =2983, 1711, 1613, 1241, 1165,
1077, 1030 cmꢀ1; HRMS (ESI): m/z [M+H]+ calcd for C21H27N6:
363.2292, found 363.2294; RP-HPLC (254 nm): tR =21.8 min, purity
>99%.
Docking
Receptor preparation: The PDB file for DNA–ligand crystal 3FSI
was obtained from the Protein Data Bank. The solvent, protein and
metal ions were removed using AutoDockTools 1.5.4.[42] The coordi-
nates of the binding site around the bis-aminoimidazoline ligand
was found using the AutoGrid function and the ligand was re-
moved. The Gasteiger charge for the DNA octamer was computed
as ꢀ12 and polar hydrogens were added. The macromolecule was
saved as a.pdbqt file.
4,4’-Bis(2-amino-1,4,5,6-tetrahydropyrimidine)phenylaniline tri-
hydrochloride (4d): Method C, using N-Boc-protected deriva-
tive 7d’ (72 mg, 0.16 mmol), afforded 4d as a yellow solid after pu-
rification on reverse-phase silica, eluting in 100% H2O, (43 mg,
57%): mp: 2688C, decomposition; 1H NMR (400 MHz, D2O): d=
1.75–1.90 (m, 4H), 3.17–3.21 (m 8H), 7.05 ppm (s, 8H); 13C NMR
(100 MHz, D2O): d=19.3, 38.2, 115.5, 118.7, 126.7, 132.2,
142.6 ppm; IR (ATR): nmax =2981, 2891, 1710, 1625, 1491, 1318,
1153, 823 cmꢀ1; HRMS (ESI): m/z [M+H]+ calcd for C20H25N7:
364.2244, found: 364.2248; RP-HPLC (254 nm): tR =20.0 min, purity
>98%.
Ligand preparation: The optimised structure of 4a was opened in
GaussView 5.0 and its dihedral angles were changed to match the
crystal structure of the bis-aminoimidazoline ligand from 3FSI. The
linker atom in this skeleton was modified to create structures of li-
gands 4b–d. In AutoDock 1.5.4, Gasteiger charges and polar hydro-
gens were added as before and the rotation of all bonds in each
molecule was made rigid.
Docking: The ligands were docked into binding site of the DNA
octamer using AutoDockVina 1.1.2.[43]
DNA thermal denaturation
Thermal melting experiments were conducted with a Varian Cary
300 Bio spectrophotometer equipped with a 6ꢂ6 multicell temper-
ature-controlled block. Temperatures were monitored with a ther-
mistor inserted into a 1 cm quartz cuvette containing the same
volume of water as in the sample cells. Absorbance changes at
260 nm were monitored in the range of 30–908C with a heating
rate of 1 8Cminꢀ1 and a data collection rate of five points per 8C.
The salmon sperm DNA and poly(dA-dT)2 were purchased from
Sigma–Aldrich (extinction coefficient e260 =6600mꢀ1 cmꢀ1 for both
oligomers). A quartz cell with a 1-cm path length was filled with
a solution of DNA polymer or DNA–compound complex (1 mL). For
thermal denaturation, the DNA polymer (150 mm base) and the
compound solution (15 mm) were prepared in a phosphate buffer
(0.01m Na2HPO4/NaH2PO4, pH 7) so that a compound to DNA base
ratio of 0.1 was obtained. The thermal melting temperatures of the
duplex or duplex–compound complex obtained from the first de-
rivative of the melting curves are reported.
Acknowledgements
This work has been funded by Science Foundation Ireland (SFI-
RFP project: CH3060). P.O’S. thanks SFI for generous support.
Keywords: 2-amino-1,4,5,6-tetrahydropyrimidine
·
circular
dichroism DNA thermal denaturation
·
· guanidine-like
dications · minor groove binders · molecular design
[1] P. G. Baraldi, A. Bovero, F. Fruttarolo, D. Preti, M. A. Tabrizi, M. G. Pavani,
R. Romagnoli, Med. Res. Rev. 2004, 24, 475–528.
[2] X. Cai, P. J. Gray, D. D. von Hoff, Cancer Treat. Rev. 2009, 35, 437–450.
[3] M. P. Barrett, C. G. Gemmell, C. J. Suckling, Pharmacol. Ther. 2013, 139,
12–23.
[4] R. K. Arafa, M. A. Ismail, M. Munde, W. D. Wilson, T. Wenzler, R. Brun,
D. W. Boykin, Eur. J. Med. Chem. 2008, 43, 2901–2908.
[5] P. B. Dervan, Bioorg. Med. Chem. 2001, 9, 2215–2235.
[6] S. Lang, A. I. Khalaf, D. Breen, J. K. Huggan, C. J. Clements, S. P. MacKay,
C. J. Suckling, Med. Chem. Res. 2014, 23, 1170–1179.
Circular dichroism
[7] X. Zhang, S. C. Zhang, D. Sun, J. Hu, A. Wali, H. Pass, F. Fernandez-
Madrid, M. R. Harbut, N. Tang, PloS ONE 2011, 6, e25822.
[8] M. F. Paine, M. Z. Wang, C. N. Generaux, D. W. Boykin, W. D. Wilson, H. P.
De Koning, C. A. Olson, G. Pohlig, C. Burri, R. Brun, G. A. Murilla, J. K.
Thuita, M. P. Barrett, R. R. Tidwell, Curr. Opin. Investig. Drugs 2010, 11,
876–883.
[9] M. Sands, M. A. Kron, R. B. Brown, Rev. Infect. Dis. 1985, 7, 625–634.
[10] W. Mulugeta, J. Wilkes, W. Mulatu, P. A. O. Majiwa, R. Masake, A. S. Pere-
grine, Acta Trop. 1997, 64, 205–217.
CD spectra were collected with a JASCO J-800 spectrometer at dif-
ferent ratios of compound to DNA polymer [ssDNA or poly(dA-
dT)2] at 258C in phosphate buffer at pH 7.0. Titrations were carried
out by adding the compound (1.5 mm to 15 mm over 10 additions)
in a DNA polymer solution (ssDNA: 150 mm or poly(dA-dT)2:
37.5 mm) in a 1 cm quartz cuvette and scanned over a desired
wavelength range. Data points for ssDNA were taken at an interval
of 0.5 nm, scanning at 50 nmminꢀ1. Data points for poly(dA-dT)2
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 2065 – 2073 2072

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
[11] S. J. Gregson, P. W. Howard, J. A. Hartley, N. A. Brooks, L. J. Adams, T. C.
Jenkins, L. R. Kelland, D. E. Thurston, J. Med. Chem. 2001, 44, 737–748.
[12] Clinical trial identifier: NCT02034227.
[13] G. He, E. Vasilieva, G. D. Harris Jr., K. J. Koeller, J. K. Baskin, C. M. Dupur-
eur, Biochemie 2014, 102, 83–91.
[32] L. D. Hicks, J. K. Han, A. J. Fry, Tetrahedron Lett. 2000, 41, 7817–7820.
[33] M. L. di Vona, V. Rosnati, J. Org. Chem. 1991, 56, 4269–4273.
[34] T. Kurita, F. Aoki, T. Mizumoto, T. Maejima, H. Esaki, T. Maegawa, Y. Mon-
guchi, H. Sajiki, Chem. Eur. J. 2008, 14, 3371–3379.
[35] F. Rodriguez, I. Rozas, J. E. Ortega, A. M. Erdozain, J. J. Meana, L. F. Calla-
do, J. Med. Chem. 2008, 51, 3304–3312.
[36] Correlations found: [2 poly(dA)-poly(dT)]=0.82 [1 poly(dA-dT)2]+1.02;
R2 =0.998. [3 poly(dA-dT)2]=1.10 [2 poly(dA)-poly(dT)]+0.39; R2 =
0.998. [3 poly(dA-dT)2]=1.35 [1 poly(dA-dT)2]+1.90; R2 =0.994.
[37] B. Nguyen, C. Tardy, C. Bailly, P. Colson, C. Houssier, A. Kumar, D. W.
Boykin, W. D. Wilson, Biopolymers 2002, 63, 281–297.
[14] N. G. Nickols, P. B. Dervan, Proc. Natl. Acad. Sci. USA 2007, 104, 10418–
10423.
[15] R. W. Bꢃrli, J. A. Kaizerman, J.-X. Duan, P. Jones, K. W. Johnson, M. Iwa-
moto, K. Truong, W. Hu, T. Stanton, A. Chen, S. Touami, M. Gross, V.
Jiang, Y. Ge, H. E. Moser, Bioorg. Med. Chem. Lett. 2004, 14, 2067.
[16] N. G. Anthony, D. Breen, G. Donoghue, A. I. Khalaf, S. P. Mackay, J. A. Par-
kinson, C. J. Suckling, Org. Biomol. Chem. 2009, 7, 1843–1850.
[38] Y. Liu, A. Kumar, D. W. Boykin, W. D. Wilson, Biophys. Chem. 2007, 131,
1–14.
ˇ ´
[17] P. S. Nagle, F. Rodriguez, A. Kahvedzic, S. Quinn, I. Rozas, J. Med. Chem.
2009, 52, 7113–7121.
[18] F. Rodrꢄguez, I. Rozas, M. Kaiser, R. Brun, B. Nguyen, W. D. Wilson, R. N.
Garcꢄa, C. Dardonville, J. Med. Chem. 2008, 51, 909–923.
[19] P. S. Nagle, F. Rodriguez, S. J. Quinn, D. H. O’Donovan, J. M. Kelly, B.
Nguyen, W. D. Wilson, I. Rozas, Org. Biomol. Chem. 2010, 8, 5558–5567.
[20] M. L. Kopka, C. Yoon, D. Goodsell, P. Pjura, R. E. Dickerson, Proc. Natl.
Acad. Sci. USA 1985, 82, 1376–1380.
[21] S. Neidle, Nat. Prod. Rep. 2001, 18, 291–309.
[22] S. Neidle, L. R. Kelland, J. O. Trent, I. J. Simpson, D. W. Boykin, A. Kumar,
W. D. Wilson, Bioorg. Med. Chem. 1997, 7, 1403–1408.
[23] Y. Liu, A. Kumar, S. Depauw, R. Nhili, M.-H. David-Cordonnier, M. P. Lee,
M. A. Ismail, A. A. Farahat, M. Say, S. Chackal-Catoen, A. Batista-Parra, S.
Neidle, D. W. Boykin, W. D. Wilson, J. Am. Chem. Soc. 2011, 133, 10171–
10183.
[39] Gaussian 09, Revision D.01, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E.
Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Men-
nucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian,
A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara,
K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O.
Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro,
M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Ko-
bayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyen-
gar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B.
Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,
O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin,
K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg,
S. Dapprich, A. D. Daniels, ꢅ. Farkas, J. B. Foresman, J. V. Ortiz, J. Cio-
slowski„ D. J. Fox, Gaussian, Inc., Wallingford CT, 2009.
[24] S. Mallena, M. P. H. Lee, C. Bailly, S. Neidle, A. Kumar, D. W. Boykin, W. D.
Wilson, J. Am. Chem. Soc. 2004, 126, 13659–13669.
[40] a) A. D. Becke, J. Chem. Phys. 1993, 98, 5648–5652; b) C. Lee, W. Yang,
R. G. Parr, Phys. Rev. B 1988, 37, 785–789; c) S. H. Vosko, L. Wilk, M.
Nusair, Can. J. Phys. 1980, 58, 1200–1211; d) P. J. Stephens, F. J. Devlin,
C. F. Chabalowski, M. J. Frisch, J. Phys. Chem. 1994, 98, 11623–11627.
[41] W. J. Hehre, R. Ditchfield, J. A. Pople, J. Chem. Phys. 1972, 56, 2257–
2263.
[25] L. S. Glass, B. Nguyen, K. D. Goodwin, C. Dardonville, D. W. Wilson, E.
Long, M. M. Georgiadis, Biochemistry 2009, 48, 5943–5952.
[26] F. J. Acosta-Reyes, C. Dardonville, H. de Koning, M. Natto, J. A. Subirana,
J. L. Campos, Acta Crystallogr. Section D 2014, 70, 1614–1621.
[27] J. A. Parkinson, F. J. Scott, C. J. Suckling, G. Wilson, Med. Chem. Commun.
2013, 4, 1105–1108.
[42] G. M. Morris, R. Huey, W. Lindstrom, M. F. Sanner, R. K. Belew, D. S. Good-
sell, A. J. Olson, J. Comput. Chem. 2009, 30, 2785–2791.
[28] Calculator plugins were used for structure property prediction and cal-
culation, Marvin 6.1.0, 2013, ChemAxon (http://www.chemaxon.com).
[29] J. W. Shaw, D. H. Grayson, I. Rozas, Arkivoc 2014, 161–174.
[30] A. Barker, C. C. Barker, J. Chem. Soc. 1954, 870–873.
[31] K. Ofosu-Asante, L. M. Stock, J. Org. Chem. 1990, 55, 3409–3410.
[43] O. Trott, A. J. Olson, J. Comput. Chem. 2010, 31, 455–461.
Received: June 26, 2014
Published online on August 1, 2014
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 2065 – 2073 2073