Comparative Structural Study of Metal-Mediated Base Pairs Formed outside and inside DNA Molecules

Article abstract of DOI:10.1021/acs.inorgchem.0c01210

The formation of copper(II)-mediated base pairs involving pyridine-2,6-dicarboxylate derivatives and canonical nucleosides has proven to be a smart approach to introduce copper(II) ions at specific locations of DNA duplexes. However, the structural characteristics of these metalized base pairs have not yet been revealed, and their effect on DNA structures is difficult to assess. Herein, for the first time, we report on the different structural details of copper-mediated base pairs formed by themselves and in DNA duplexes. The individual base pairs [Cu(mcheld)(N3-Cyt)(H2O)]·3H2O (1Cu_Cyt), [Cu(mcheld)(N7-Ade)(H2O)2]·2H2O (1Cu_Ade), [Cu(mcheld)(N7-Gua)(H2O)] (1Cu_Gua), and [Cu(mcheld)(N1-7CAde)(H2O)]·H2O (1Cu_7CAde) were obtained from the reaction of the metal complex [Cu(mcheld)(H2O)2] (1Cu) (mcheld = 4-methoxypyridine-2,6-dicarboxylic acid) with model nucleosides (Cyt = N1-methylcytosine, Ade = N9-ethyladenine, Gua = N9-propylguanine, 7CAde = N9-propyl-7-deazaadenine). The crystal structure of the five complexes was determined by means of single-crystal X-ray diffraction. Furthermore, the formation of the 1Cu_Cyt and 1Cu_Gua base pairs in the middle of DNA duplexes, duplex DNA15 (917 atoms) and DNA10 (649 atoms), respectively, was studied using highly demanding ab initio computational calculations. These theoretical studies aimed to validate, from a structural point of view, whether base pairs of the kind 1Cu_nucleosides can be included in a DNA double helix and how this situation affects the double-helical structure. The results indicate that the 1Cu_Cyt and 1Cu_Gua base pairs can be formed in a DNA molecule without significant structural constraints. In addition, the double-helix DNA structure remains virtually unchanged when it contains these Cu(II)-mediated base pairs.

Full text of DOI:10.1021/acs.inorgchem.0c01210

pubs.acs.org/IC
Article
Comparative Structural Study of Metal-Mediated Base Pairs Formed
outside and inside DNA Molecules
́
Alicia Dominguez-Martin, Simona Galli, Jose A. Dobado, Noelia Santamaría-Díaz,
Antonio Perez-Romero, and Miguel A. Galindo*
́
Cite This: https://dx.doi.org/10.1021/acs.inorgchem.0c01210
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: The formation of copper(II)-mediated base pairs
involving pyridine-2,6-dicarboxylate derivatives and canonical nucleo-
sides has proven to be a smart approach to introduce copper(II) ions
at specific locations of DNA duplexes. However, the structural
characteristics of these metalized base pairs have not yet been
revealed, and their effect on DNA structures is difficult to assess.
Herein, for the first time, we report on the different structural details
of copper-mediated base pairs formed by themselves and in DNA
duplexes. The individual base pairs [Cu(mcheld)(N3-Cyt)(H2O)]·
3H2O (1Cu_Cyt), [Cu(mcheld)(N7-Ade)(H2O)2]·2H2O
(1Cu_Ade), [Cu(mcheld)(N7-Gua)(H2O)] (1Cu_Gua), and [Cu-
(mcheld)(N1-7CAde)(H2O)]·H2O (1Cu_7CAde) were obtained from
the reaction of the metal complex [Cu(mcheld)(H2O)2] (1Cu)
(mcheld = 4-methoxypyridine-2,6-dicarboxylic acid) with model
nucleosides (Cyt = N1-methylcytosine, Ade = N9-ethyladenine, Gua = N9-propylguanine, 7CAde = N9-propyl-7-deazaadenine).
The crystal structure of the five complexes was determined by means of single-crystal X-ray diffraction. Furthermore, the formation
of the 1Cu_Cyt and 1Cu_Gua base pairs in the middle of DNA duplexes, duplex DNA15 (917 atoms) and DNA10 (649 atoms),
respectively, was studied using highly demanding ab initio computational calculations. These theoretical studies aimed to validate,
from a structural point of view, whether base pairs of the kind 1Cu_nucleosides can be included in a DNA double helix and how this
situation affects the double-helical structure. The results indicate that the 1Cu_Cyt and 1Cu_Gua base pairs can be formed in a
DNA molecule without significant structural constraints. In addition, the double-helix DNA structure remains virtually unchanged
when it contains these Cu(II)-mediated base pairs.
tions, which include LTN-M-LTN and LTN-M-nucleobase
pairs (M = metal ion).7,8Among the LTNs, some nucleobase
analogues are capable of mimicking either Watson−Crick or
Hoogsteen pairing in both the absence and presence of metal
ions.9−12 Remarkably, the use of LTNs has led to short
oligonucleotides featuring specific transition-metal ions situ-
ated at precise locations inside a DNA double helix, with
applications ranging from sensing to electron transport
processes.13−17 The formation of metal-mediated base pairs
inside DNA is normally determined in solution using
spectroscopic and spectrometric methodologies. However,
unraveling the structural organization of these systems can
be a complicated task due to the difficulty in either obtaining
INTRODUCTION
■
The development of DNA-based nanotechnology has received
deep attention due to the remarkable properties of nucleic
acids.1 DNA is very soluble in water, and its synthesis is well
established, making it an ideal biopolymer for research
purposes. The code programmability of DNA derives from
the natural self-assembly capabilities of adenine−thymine (A-
T) and guanine−cytosine (G-C) base pairs. The DNA
structure and conformation can be rationally preprogrammed
on the basis of the self-recognition abilities of DNA
sequences,2 leading to tailored structures that range from the
nanometer to the micrometer scale.3 In addition, the DNA
genetic code can be further expanded if novel ligand-type
nucleobases (LTNs), i.e. specific ligands replacing canonical
nucleobases, are employed.4 In this regard, different LTNs
have been demonstrated to self-assemble inside DNA in the
presence of metal ions, forming the so-called metal-mediated
base pairs.5 This strategy, first proposed by Shionoya’s research
group6 and first achieved by Schultz,4 has led to the formation
of smart metal-mediated base pairs with different combina-
Received: April 23, 2020
https://dx.doi.org/10.1021/acs.inorgchem.0c01210
© XXXX American Chemical Society
Inorg. Chem. XXXX, XXX, XXX−XXX
A

Inorganic Chemistry
pubs.acs.org/IC
Article
crystals suitable for single-crystal X-ray diffraction, or perform-
ing structural high-resolution NMR studies. Notably, in some
cases, solid-state crystallographic investigations and solution
NMR studies have been carried out for DNA molecules
containing metal-mediated base pairs.18 Alternatively, theoreti-
cal computational procedures can help in revealing the
structural aspects of metal-mediated base pairs. Such
calculations are commonly performed for individual base
pairs,19,20 with few cases involving the DNA framework.21,22
The ligand pyridine-2,6-dicarboxylic acid (Cheld, a planar
tridentate ligand) and its carboxyamide derivative23 were
among the pioneering ligands to be employed in the formation
of metal-mediated base pairs. Cheld has been widely used in
the preparation of metalated double-stranded DNA (ds-DNA)
molecules via the formation of metal-mediated base pairs.
Initially, Cheld and pyridine were employed to introduce
Cu(II) ions into DNA duplexes by forming unsymmetrical [3
+ 1] metal−base pairs. This approach led to the formation, in
solution, of DNA duplexes containing Cheld-Cu-nucleobase
mismatches, which showed higher thermal stability in
comparison to the unmetalated counterparts.4 The stabiliza-
tion of these mismatches followed the trend adenine > cytosine
> guanine, with no data for thymidine mismatches. These
results demonstrated that the ligand Cheld can be employed to
form metal-mediated base pairs with canonical bases within ds-
DNA molecules.
EXPERIMENTAL SECTION
■
Materials and Methods. 1-Methylcytosine (Cyt),29 N9-ethyl-
adenine (Ade),30 and diethyl-4-hydroxypyrimidine-2,6-carboxylic acid
(decheld)31 were synthesized according to literature procedures. 6-
Amino-7-deazapurine, 6-chloro-2-amino-7-deazapurine, adenine, so-
dium hydride (60% oil dispersion), anhydrous N,N-dimethylforma-
mide, 1-iodopropane, and the oligonucleotides d(Ade)15, d(7CAde)15,
d(Cyt)15 and d(Gua)10 were purchased from Sigma-Aldrich. 1H NMR
spectra were recorded with a Bruker AMX instrument working at 300
MHz. 13C NMR spectra were recorded with a Bruker Neo 500 MHz
spectrometer. Elemental analyses were carried out with a Fisons-Carlo
Erba Model EA 1108 analyzer. CD spectra were recorded on a
JASCO J-815 spectrometer. High-resolution electrospray mass
spectrometry was performed with a Waters LCT Premier XE mass
spectrometer. Infrared (IR) spectra were registered with a Bruker
Tensor-27 FT-IR spectrometer.
Synthesis of the Ligands. Diethyl-4-methoxypyrimidine-2,6-
dicarboxylic Acid (demcheld). To a solution of decheld (3.00 g,
12.50 mmol) in anhydrous N,N-dimethylformamide (150 mL) was
added sodium hydride (0.50 g, 12.50 mmol) under N2. The reaction
mixture was stirred for 30 min at room temperature. After this time,
iodomethane (0.79 mL, 12.50 mmol) was added and the mixture was
stirred under N2 for 48 h. The solvent was then removed under
reduced pressure, and the product was diluted in dichloromethane
(150 mL) and washed with aqueous sodium hydrogen carbonate
(10% m/m, 150 mL) and brine (150 mL). The organic layers were
collected, dried over magnesium sulfate, filtered off, and concentrated
in vacuo, giving the title compound as a white powder (3.00 g, 94%
1
yield). H NMR (300 MHz, DMSO-d6): δ (ppm) 7.72 (s, 2H; CH),
The crystal structure of a DNA duplex containing one
Cheld-Cu-pyridine metallo-base pair was determined,24 as well
as the crystal structure of the complex itself,25,26 demonstrating
the predictable coplanar arrangement between Cheld and
pyridine in both cases. However, no further studies have been
performed to fully reveal the structure and conformation of
Cheld-Cu-nucleoside systems, where noncoplanar arrange-
ments between the units can be anticipated. In this regard, it is
important to compare the structure of metal-mediated base
pairs when they are prepared inside and outside a double helix,
in order to determine the effect of the DNA scaffold on the
final arrangement. In this respect, Cheld-Cu-nucleoside
systems are important representative examples.
To the best of our knowledge, only the crystal structure of
the Cheld-Cu-adenine ternary complex27 and the interaction of
other iminodiacetic acid based copper(II) complexes toward
adenine derivatives28 have been reported. Though the
molecular structure of Cheld-Cu-nucleoside systems could be
foretold in principle on the basis of these previously described
systems, detailed, experimentally derived structural information
at the molecular and crystal-packing level is still needed to
accurately understand the structural implications that the
Cheld-Cu-nucleoside systems can have in DNA molecules.
Ultimately, determining the crystal and molecular structure
of metal-mediated base pairs will provide fundamental
information that can be used in the design of novel metal−
DNA systems, potentially finding applications in different
areas, from catalysis to charge transport.
4.38 (q, J = 7.1 Hz, 4H; CH2), 3.98 (s, 3H; CH3), 1.34 (t, J = 7.1 Hz,
6H; CH3). HRMS (ESI): m/z calcd for C12H16NO5 [M + H]+,
254.1028; found, 254.1037.
4-Methoxypyridine-2,6-dicarboxylic Acid (mcheld). demcheld
(2.00 g, 8.37 mmol) was dissolved in 0.5 M aqueous sodium
hydroxide (25 mL), and the solution thus obtained was refluxed
overnight. Then, the solution was acidified with concentrated
hydrochloric acid until the pH reached 3.6 and was concentrated
under reduced pressure to give the title compound as a white
precipitate, which was filtered off and dried under vacuum (1.21 g,
1
58% yield). H NMR (300 MHz, DMSO-d6): δ (ppm) 7.36 (s, 2H;
CH), 3.80 (s, 3H; CH3). IR (ν/cm−1): 3564.7, 3429.8, 3084.3, 2923,
1722, 1631.5, 1601, 1574.7, 1465.8, 1439.5, 1410.2, 1372.5, 1312.1,
1234, 1165.4, 1112.3, 1047.7, 892.7, 865.1, 814.9, 783, 696.7, 579.6.
HRMS (ESI): m/z calcd for C8H8NO5 [M + H]+, 198.0402; found,
198.0406.
N9-Propyl-7-deazaadenine (7CAde). To a solution of 6-amino-7-
deazapurine (1.00 g, 7.23 mmol) in anhydrous N,N-dimethylforma-
mide (200 mL) was added NaH (0.10 g, 2.48 mmol) under argon.
The reaction mixture was stirred for 30 min until H2 evolution ceased.
1-Iodopropane (0.22 mL, 2.26 mmol) was then added, and the
reaction mixture was stirred at room temperature for 24 h under
argon. The solvent was removed under reduced pressure. The crude
residue was dissolved in dichloromethane (200 mL) and washed with
an aqueous solution of sodium hydrogen carbonate (10% m/m, 150
mL) and brine (150 mL). The organic layer was collected, dried over
magnesium sulfate, filtered off, and concentrated in vacuo. The
precipitate was recrystallized from dichloromethane to give the title
1
compound in the form of a powder (0.70 g, 55% yield). H NMR
(300 MHz, DMSO-d6): δ (ppm) 8.04 (s, 1H; CH), 7.13 (d, J = 3.4
Hz, 1H; CH), 6.87 (s, 2H; NH2), 6.51 (d, J = 3.4 Hz, 1H; CH), 4.05
(t, J = 7.1 Hz, 2H; CH2), 1.90−1.61 (m, 2H; CH2), 0.81 (t, J = 7.4
Hz, 3H; CH3). 13C NMR (500 MHz, DMSO): δ (ppm) 157.84 (C6),
151.89 (C2), 150.01 (C4), 124.52 (C8), 102.77 (C5), 98.68 (C7),
45.76 (C10), 23.70 (C11), 11.54 (C12). IR (ν/cm−1): 1680, 1639.6,
1612.6, 1568.2, 1502.7, 1467.9, 1415.8, 1363.8, 1334.8, 1305.9,
1215.2, 1084.1, 1047.4, 1022.3, 960.62, 925.9, 879.6, 792.8, 752.3,
733, 700, 673.2. HRMS (ESI): m/z calcd for C9H13N4 [M + H]+,
177.1140; found, 177.1113.
In this context, we report herein the preparation and solid-
state characterization of complexes of the type Cheld-Cu-
nucleoside as well as, utilizing ab initio theoretical calculations,
their organization inside DNA duplexes, in an effort to get
some insight into the molecular arrangement of these systems
when they are prepared by themselves and in the context of a
DNA molecule.
B
https://dx.doi.org/10.1021/acs.inorgchem.0c01210
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
N9-Propyl-6-chloro-2-amino-7-deazapurine (Cl-7CGua). 6-
Chloro-2-amino-7-deazapurine (1.00 g, 5.90 mmol) was dissolved in
anhydrous N,N-dimethylformamide (100 mL), and NaH (0.24 g, 6.00
mmol) was added under argon. The reaction mixture was stirred until
H2 evolution ceased. 1-Iodopropane (0.58 mL, 6.00 mmol) was then
added, and the solution was stirred for 24 h at room temperature
under argon. The solvent was removed under reduced pressure. The
crude precipitate was dissolved in dichloromethane (200 mL) and
washed with aqueous sodium hydrogen carbonate (10% m/m, 150
mL) and brine (150 mL). The organic layer was collected, dried over
magnesium sulfate, filtered, and concentrated in vacuo. The sample
was then recrystallized from dichloromethane to give the title
aqueous solution (10 mL) of 1Cu (0.03 g, 0.10 mmol) dropwise. The
solution was heated to 65 °C and left at this temperature with stirring
for 30 min. A very thin precipitate was formed and eliminated by
filtration. The clear solution was then left to crystallize at room
temperature by slow solvent evaporation. After a few days, blue single
crystals suitable for X-ray diffraction were collected (0.02 g, 47%
yield). Anal. Calcd for C16H18CuN6O7(H2O): C, 39.39; H, 4.13; N,
17.23. Found: C, 39.71; H, 4.22; N, 17.12. The detection of higher
water content by elemental analysis than by X-ray diffraction (vide
infra) could be due to the presence of surface-absorbed moisture.
[Cu(mcheld)(N1-7CAde)(H2O)2]·H2O (1Cu_7CAde). To an aqueous
solution (15 mL) of 1Cu (0.02 g, 0.07 mmol) was added an aqueous
solution (10 mL) of the ligand (0.01 g, 0.07 mmol) dropwise. The
solution was heated to 60 °C and left at this temperature with stirring
for 30 min. A thin precipitate was formed and eliminated by filtration.
The clear solution was then left to crystallize at room temperature by
slow solvent evaporation. After a few days, blue single crystals suitable
for X-ray diffraction were recovered (0.01 g, 32% yield). Anal. Calcd
for C17H19CuN5O6(H2O): C, 43.35; H, 4.49; N, 14.87. Found: C,
43.18; H, 4.92; N, 14.70.
Single-Crystal X-ray Diffraction Structure Determination. X-
ray diffraction data for 1Cu_7CAde were collected using a Bruker X8
Proteum diffractometer equipped with a Cu (λ = 1.54178 Å) sealed
rotating anode X-ray tube, a Bruker AXS Smart 6000 CCD detector,
and an Oxford Cryostream 700 plus cooling apparatus. X-ray
diffraction data for 1Cu, 1Cu_Ade, and 1Cu_Gua were collected
on a Bruker D8 Venture diffractometer equipped with either a Cu (λ
= 1.54178 Å) or a Mo (λ = 0.71073 Å) X-ray tube, a Bruker AXS
Photon 100 detector, and a Kryoflex II cooling apparatus. For all four
of the complexes, data reduction was performed with the software
APEX3,32 while data correction for absorption was carried out using
the software SADABS.33 X-ray diffraction data for 1Cu_Cyt were
collected using a Bruker SMART APEX I diffractometer equipped
with a Mo (λ = 0.71073 Å) X-ray tube and a Kryoflex cooling
apparatus. In this case, data reduction was performed with the
software SAINT V6.36A, while data correction for absorption was
carried out using SADABS.33 The structures were solved by direct
methods as implemented in SHELXS-97,34 which allowed the
location of most of the atoms of the asymmetric unit. All the
remaining non-hydrogen atoms were located from difference Fourier
maps calculated from successive full-matrix least-squares refinement
cycles on F2 using SHELXL-2018/3.34 The electron densities around
the propyl group in 1Cu_Gua and 1Cu_7CAde were ill-defined: the
central carbon atom of the propyl group was found to be disordered
into two positions. Isotropic thermal displacement parameters were
refined for the three carbon atoms of the propyl group, and no
hydrogen atoms were located on them. Anisotropic thermal
displacement parameters were assigned to all the other non-hydrogen
atoms. The hydrogen atoms were located at idealized positions using
HFIX instructions and described with isotropic thermal displacement
parameters fixed at 1.2 or 1.5 times those of the atom to which they
were bound. The main crystallographic information and experimental
and data treatment details are provided in Table S1 in the Supporting
Information.
1
compound in the form of a powder (0.85 g, 68% yield). H NMR
(300 MHz, DMSO-d6): δ (ppm) 7.17 (d, J = 3.5 Hz, 1H; CH), 6.61
(s, 2H; NH2), 6.28 (d, J = 3.4 Hz, 1H; CH), 3.97 (t, J = 7.1 Hz, 2H;
CH2), 1.73 (m, 2H; CH2), 0.82 (t, J = 7.3 Hz; 3H; CH3). HRMS
(ESI): m/z calcd for C9H12ClN4 [M + H]+, 211.0750; found,
211.0732.
N9-Propyl-7-deazaguanine (7CGua). Cl-7CGua (0.40 g, 1.90
mmol) was refluxed in a hydrochloric acid (25 mL, 1 M) and
ethanol (5 mL) mixture for 2 h. After the mixture was cooled to room
temperature, the pH was adjusted to ca. 7 using sodium hydroxide
and the suspension was cooled in an ice bath for 1 h. The precipitate
was collected by filtration and dried in vacuo to afford the title
compound in the form of crystals (0.05 g, 15% yield). 1H NMR (300
MHz, DMSO-d6): δ (ppm) 10.22 (s, 1H; NH), 6.71 (d, J = 3.2 Hz,
1H; CH), 6.19 (d, J = 3.3 Hz, 2H; NH2), 6.17 (s, 1H; CH), 3.85 (t, J
= 7.0 Hz, 2H; CH2), 1.68 (m, 2H; CH2), 0.82 (t, J = 7.3 Hz, 3H;
CH3). 13C NMR (500 MHz, DMSO-d6): δ (ppm) 159.1 (C6), 152.9
(C2), 150.5 (C4), 120.4 (C8), 101.4 (C7), 100.3 (C5), 45.81 (C10),
23.7 (C11), 11.5 (C12). IR (ν/cm−1): 3402.7, 3173.1, 2874.1,
1660.8, 1608.7, 1539.3, 1508.4, 1433.2, 1410.1, 1369.6, 1334.8, 1304,
1213.3, 1180.5, 1072.5, 889.2, 850.7, 785.1, 717.8, 677.1, 601.8, 555.5.
HRMS (ESI): m/z calcd for C9H13N4ONa [M + H]+, 193.1089;
found, 193.1086.
Synthesis of the Complexes. [Cu(mcheld)(H2O)2] (1Cu). To an
aqueous solution of Cu(CH3COO)2 (0.09 g, 0.50 mmol) was added
an aqueous solution of mcheld (0.10 g, 0.50 mmol) dropwise with
stirring. The resulting solution was heated to 90 °C and left at this
temperature with stirring for 30 min. A light blue precipitate
appeared, which was filtered off and washed with water, ethanol, and
ether. The filtrate was left to crystallize by slow evaporation from the
aqueous solution. After a few days, blue single crystals suitable for X-
ray diffraction were collected (0.07 g, 50% yield). Anal. Calcd for
C8H9CuNO7: C, 32.60; H, 3.08; N, 4.75. Found: C, 32.55; H, 3.75;
N, 4.81.
[Cu(mcheld)(N3-Cyt)(H2O)]·3H2O (1Cu_Cyt). To a warm aqueous
solution (15 mL) of 1-methylcytosine (0.09 g, 0.07 mmol) was added
an aqueous solution (10 mL) of 1Cu (0.02 g, 0.07 mmol) dropwise.
The solution was heated to 65 °C and left at this temperature with
stirring for 30 min. When the solution appeared clear, it was left to
crystallize by slow solvent evaporation at room temperature. After a
few days, blue single crystals suitable for X-ray diffraction were formed
(0.01 g, 40% yield). Anal. Calcd for C13H14CuN4O7·3H2O: C, 34.25;
H, 4.42; N, 12.29. Found: C, 34.55; H, 4.50; N, 12.54.
[Cu(mcheld)(N7-Ade)(H2O)]·2H2O (1Cu_Ade). To a warm aque-
ous solution (15 mL) of N9-ethyladenine (0.02 g, 0.10 mmol) was
added an aqueous solution (15 mL) of 1Cu (0.03 g, 0.10 mmol)
dropwise. The solution was heated to 65 °C and left at this
temperature with stirring for 30 min. The precipitate was collected by
filtration and dried in vacuo to afford the title compound in the form
of a powder (0.04 g, 87% yield). Anal. Calcd for C15H16CuN6O6·
6.6H2O: C, 32.03; H, 5.30; N, 14.94. Found: C, 31.84; H, 4.93; N,
15.45. The precipitate was recrystallized by room-temperature slow
evaporation in DMF, giving blue single crystals suitable for X-ray
diffraction after a few days. The detection of higher water content by
elemental analysis than by X-ray diffraction (vide infra) could be due
to the presence of surface-absorbed moisture.
Computational Details. Theoretical models were built from
scratch using the Avogadro software.35 Sodium counterions or
protons were added to the phosphate groups, thus resulting in
neutral model systems. After a partial geometry optimization of these
molecules, using semiempirical methods (PM3) within the ORCA
4.2.1 computational program,36,37 a suitable geometry was obtained
for further full optimization by quantum mechanics ab initio
calculations using the HF-3c method.38 The latter method includes
the so-called MINIX basis set and a correction of the basis set
superposition error (BSSE) by means of the geometrical counterpoise
gCP correction algorithm,39 as well as the atom-pairwise London
dispersion energy from the D3 dispersion correction scheme with the
Becke−Johnson damping scheme (D3BJ).40−42 The solvent effect was
included by the so-called “implicit solvent model”. Thus, the solute
was placed in a cavity of roughly molecular shape, and the solvent was
described by a continuum that interacts with the charges on the cavity
[Cu(mcheld)(N7-Gua)(H2O)] (1Cu_Gua). To an aqueous solution
(15 mL) of N9-propylguanine (0.02 g, 0.10 mmol) was added an
C
https://dx.doi.org/10.1021/acs.inorgchem.0c01210
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
a
Scheme 1. Synthesis of the Model Nucleosides N9-Propyl-7-deazaadenine (7CAde) and N9-Propyl-7-deazaguanine (7CGua)
a
Legend: (a) sodium hydride, 1-iodopropane, anhydrous N,N-dimethylformamide, rt, 24 h; (b) hydrochloric acid (1 M), reflux, 2 h.
surface. Specifically, the conductor-like polarizable continuum model
(CPCM) was used for the computations of the DNA structure
comprising only three base pairs (DNA3), using water as a solvent,
which is an efficient way of accounting for solvent effects in quantum
chemical calculations.43 The solvent effect was not included for the
calculations of larger DNA15 and DNA10 models, composed of 971
and 649 atoms, respectively, due to the extremely costly computa-
tional demand.
ray diffraction according to the formula [Cu(mcheld)(H2O)2]
(1Cu).
1Cu crystallizes in the monoclinic space group P21/n (Table
S1). The asymmetric unit contains one copper(II) ion, one
mcheld molecule, and two water molecules, all in general
positions. The copper(II) ion is five-coordinated and adopts a
distorted-square-pyramidal geometry (τ = 0.08)45 (Figure S1),
as already observed in [Cu(Cheld)(H2O)2].46 In 1Cu, the
mcheld ligand56 shows the μ,κ3-N10,O2,O4 coordination mode
(Cu−N10 1.900(2) Å, Cu−O2 2.056(1) Å, Cu−O4 2.043(2)
Å) and occupies three basal positions. Two water molecules
complete the coordination sphere, occupying the basal position
opposite to the pyrimidine nitrogen atom (Cu−O2w 1.939(2)
Å) and the apical position (Cu−O1w 2.174(2) Å). The latter
water molecule forms a slightly elongated Cu−O bond, in
agreement with the Jahn−Teller distortion that typically occurs
in five- or six-coordinated Cu(II) complexes.47 Intermolecular
hydrogen bonds56 are present among neighboring [Cu-
(mcheld)(H2O)2] complexes, involving the carboxylic oxygen
atoms of mcheld and the water molecules (O1w···O2w
2.925(2) Å, O1w···O3 2.706(2) Å, O1w···O5 2.698(2) Å,
O2w···O2 2.693(2) Å, O2w···O4 2.973(2) Å, O2w···O5
2.712(2) Å) (Figure S2). Overall, these interactions bring
about the formation of a 3D supramolecular network. The
methylation of the hydroxyl group of mcheld avoids the
formation of intermolecular hydrogen bonds via this
position.48,49 As shown below, this feature is important
regarding the coordination of the [Cu(mcheld)] fragment to
nucleobases and nucleosides, since it favors the formation of
hydrogen bonds exclusively between the carboxylic groups of
mcheld and the donor or acceptor groups of the nucleobase
moiety.55
RESULTS AND DISCUSSION
■
Synthesis of 4-Methoxypyridine-2,6-dicarboxylic
Acid (mcheld) and Model Nucleosides. The synthetic
route to prepare 4-methoxypyridine-2,6-dicarboxylic acid
(mcheld) consists of three steps (Scheme S1). The synthesis
is initiated from the commercially available 4-hydroxypyridine-
2,6-dicarboxylic acid, which is converted to diethyl-4-
hydroxypyridine-2,6-carboxylate (decheld) in good yields, as
described previously.31 The successive reaction steps consist of
the methylation of the hydroxyl functional group of decheld,
using sodium hydride and iodomethane, to obtain diethyl 4-
methoxypyridine-2,6-dicarboxylate (demcheld) in 94% yield.
Finally, the demcheld ester is converted into the carboxylic
acid mcheld using sodium hydroxide in water, affording it in
58% yield.
The model nucleosides N9-propyl-7-deazaadenine (7CAde)
and N9-propyl-7-deazaguanine (7CGua) were prepared by
following an adaptation of previous synthesis protocols
(Scheme 1).30,44 The synthesis of 7CAde was achieved in
good yield via N-alkylation of the commercially available 7-
deazaadenine using sodium hydride and 1-bromopropane, in
anhydrous N,N-dimethylformamide. In the case of N9-propyl-
7-deazaguanine (7CGua), the synthesis starts from 6-chloro-2-
amino-7-deazapurine and the N9 position is blocked by
introducing a propyl chain via N-alkylation, to form N9-propyl-
6-chloro-2-amino-7-deazapurine. Then, a hydroxyl group
displaces the chlorine atom by nucleophilic aromatic
substitution in acidic conditions (1 M hydrochloric acid) to
obtain 7CGua.
Synthesis and Crystal Structure of mcheld-Cu-
Nucleoside Complexes. The interaction of 1Cu with a
number of model nucleobases (Scheme 2) was studied in the
solid state and solution by means of single-crystal X-ray
diffraction and CD spectroscopy, respectively.
The general procedure to obtain complexes of the type
[Cu(mcheld)(model nucleoside)] consisted of adding a warm
aqueous solution of 1Cu to a warm solution of the appropriate
model nucleobase and leaving the mixture to react at 65 °C for
30 min with stirring. The precipitate was then filtered off and
left to crystallize by slow evaporation (in DMF or water; see
Synthesis and Crystal Structure of [Cu(mcheld)-
(H2O)2] (1Cu). The reaction of copper(II) acetate and mcheld
was carried out in water at 90 °C, leading to a light blue
precipitate. The latter was recovered by filtration and
recrystallized in water, isolating blue single crystals which
were characterized by elemental analysis and single-crystal X-
D
https://dx.doi.org/10.1021/acs.inorgchem.0c01210
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
Scheme 2. Model Nucleosides Employed in the Present
a
Work
a
For easy reading, the labels of the modified nucleobases avoid any
reference to their alkyl functionalization.57
the Experimental Section). In all cases except for the Thy and
7CGua nucleobases, single crystals were formed after a few days
and were characterized by X-ray diffraction as [Cu(mcheld)-
(N3-Cyt)(H2O)]·3H2O (1Cu_Cyt), [Cu(mcheld)(N7-Ade)-
(H2O)2]·2H2O (1Cu_Ade), [Cu(mcheld)(N7-Gua)(H2O)]
(1Cu_Gua) and [Cu(mcheld)(N1-7CAde)(H2O)]·H2O
(1Cu_7CAde).
Figure 1. (a) Molecular structure of the complex [Cu(mcheld)(N3-
Cyt)(H2O)]·3H2O (1Cu_Cyt) with thermal ellipsoids set at the 30%
probability level. (b) Portion of the crystal structure of 1Cu_Cyt,
showing pairs of 1Cu_Cyt complexes connected by hydrogen bonds
(dashed cyan lines) and π−π interactions (dashed purple line). The
uncoordinated water molecules O4w (and the corresponding
hydrogen bonds) have been omitted for clarity. The O2w and O3w
noncoordinated water molecules have been represented as space-
filling spheres.
The complex 1Cu_Cyt crystallizes in the triclinic space
group P1 (Table S1). The asymmetric unit consists of one
̅
metal ion, one mcheld ligand, one Cyt nucleoside, and four
water molecules, all in general positions. The copper(II) ion is
six-coordinated and adopts a distorted-octahedral geometry.
The mcheld ligand shows the μ,κ3-N10,O2,O4 coordination
mode (Figure 1-a) (Cu−N10 1.893(3) Å, Cu−O2 1.997(2) Å,
Cu−O4 2.027(2) Å), occupying three equatorial positions,
while the Cyt ligand shows the μ,κ2-N3,O21 coordination
mode (Figure 1a) (Cu−N3 1.956(3) Å, Cu−O21 2.711(3) Å),
completing the equatorial plane with the pyrimidine nitrogen
atom N3 and locating the ketonic oxygen atom O21, with a
loose interaction, at one of the axial positions. One water
molecule completes the coordination sphere (Cu−O1w
2.407(2) Å) at the remaining axial position, opposite to
O21. Again, the longer axial Cu−O distances are reasonably
the result of the Jahn−Teller effect. An intramolecular
hydrogen bond is found between the coordinated water
molecule O1w and the nitrogen atom N4 of the Cyt amino
group (Figure 1a) (N4···O1w 2.822(2) Å). Intermolecular
hydrogen bonds56 are present between N4 and one non-
coordinated water molecule (N4···O2w 2.863(4) Å) and
among noncoordinated water molecules (O3w···O2w 2.821(5)
Å, 2.826(3) Å, O3w···O3w 2.796(5) Å), as well as between
water molecules and coordinated and dangling oxygen atoms
of the carboxylate groups of mcheld (O4w···O2 2.886(4) Å,
O4w···O4 2.866(4) Å, O1w···O5 2.775(4) and 2.841(4) Å).
Overall, all the hydrogen-bond donor and acceptor atoms,
from both mcheld and the Cyt moiety, are involved in
hydrogen bonds (Figure 1b). The O1w···O5 interactions
further stabilized by the formation of π−π interactions
(distance between the centroids 3.596(2) Å, α = 0°, β = γ =
27.25°) between the pyridine rings of mcheld (Figure 1b). The
angle between the rms planes of Cyt and the [Cu(mcheld)]
fragment is 77.64(11)°, highlighting the high deviation from
coplanarity of the two ligands, likely enforced by the
coordination of N3 and O21 to the same metal center.
The complex [Cu(mcheld)(N7-Ade)(H2O)2]·2H2O
(1Cu_Ade) crystallizes in the monoclinic space group P21/n
(Table S1). The asymmetric unit features one copper(II) ion,
one mcheld ligand, one Ade nucleoside, and four water
molecules, all in general positions. The copper(II) ion is six-
coordinated in a distorted-octahedral geometry. The ligand
mcheld shows the μ,κ3-N10,O2,O4 coordination mode (Figure
2) (Cu−N10 1.953(4) Å, Cu−O2 2.229(3) Å, Cu−O4
2.310(4) Å), occupying three meridional positions within the
octahedron. Ade is coordinated to the copper(II) ion through
define a graph set motif of the type R2 (8) (Figure S3).50 In
4
addition, each O2w···O3w···O3w···O2w segment (Figure 1b),
running along the [001] crystallographic direction and lying on
an inversion center, connects six [Cu(mcheld)(N3-Cyt)-
(H2O)] complexes. The crystal structure of 1Cu_Cyt is
Figure 2. Molecular structure of the complex [Cu(mcheld)(N7-
Ade)(H2O)2]·2H2O (1Cu_Ade) with thermal ellipsoids set at the
30% probability level. The intramolecular hydrogen bond is depicted
as a cyan dashed line.
E
https://dx.doi.org/10.1021/acs.inorgchem.0c01210
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
N7 (Cu−N7 1.960(4) Å), occupying the remaining meridional
position. Two water molecules occupy the axial positions
(Cu−O1w 2.032(3) Å, Cu−O2w 2.072(4) Å) (Figure 2). The
observed N7-Ade coordination mode is reinforced by an
intramolecular hydrogen bond, involving the Ade N6 exocyclic
amino group as a donor and one carboxylate group of mcheld
as an acceptor (N6···O2 2.846(6) Å) (Figure 2). This
molecular recognition pattern has been previously observed
in related aromatic amine analogues27,51,52 and reasonably
influences the low value of the angle between the rms planes of
mcheld and the nucleobase (23.98(12)°).
The crystal structure is further stabilized by intermolecular
hydrogen bonds56 involving coordinated and uncoordinated
water molecules, the aminic nitrogen atom N6 of Ade, and the
carboxylate groups of mcheld (N6···O5 2.964(5) Å, O1w···O3
2.678(4) Å, O1w···O5 2.743(5) Å, O2w···O4w 2.749(4) Å,
O2w···N3 2.819(6) Å, O3w···N1 2.834(5) Å, O3w···O4w
2.736(5) Å, O4w···O3 2.814(5) Å) (Figure 3). Interestingly,
Figure 4. Molecular structure of [Cu(mcheld)(N7-Gua)(H2O)]
(1Cu_Gua) with thermal ellipsoids set at the 30% probability level.
The intramolecular hydrogen bond interaction is highlighted by a
cyan dashed line. For clarity, an ordered model has been adopted for
the Gua propylic residue.
guanine endocyclic N1 nitrogen atom and the carboxylate O3
atom of mcheld (N2···O5 2.988(7) Å, N1···O3 2.827(7) Å)
(Figure S7). Overall, a 3D network of intermolecular
hydrogens bond is formed. The [Cu(mcheld)] fragment and
the Gua moiety are not coplanar, with an angle between their
rms planes of 78.94(10)°. This non-negligible deviation from
planarity can be due to the repulsive interactions that take
place between the Gua keto group and the mcheld carboxylate
groups, which also allows for the formation of the intra-
molecular hydrogen bond. No π−π stacking interactions are
observed in the crystal structure.
Figure 3. Portion of the crystal structure of 1Cu_Ade. The hydrogen
bonds (dashed cyan lines) involving the coordinated water molecules
O2w lead to the formation of 1D double-strand motifs running along
[001]. These motifs are connected by additional intermolecular π−π
interactions (purple lines), contributing to the formation of 2D
supramolecular layers (a detail of the antiparallel π−π interactions is
provided in the inset).
The interaction of complex 1Cu with Ade and Gua
derivatives occurs via the Hoogsteen-position N7 atom, as is
evidenced by the complexes 1Cu_Ade and 1Cu_Gua,
respectively. This metal−nucleoside binding mode is expected,
since the N7 position is usually preferred by metal ions that
interact with purine nucleobases/nucleosides, especially
guanine.53 The fact that no interaction was observed via the
Watson−Crick position (the N1 atom) prompted us to use a
different strategy to evaluate this alternative binding mode that
could also occur when Cheld-Cu complexes are introduced in
a DNA molecule. Thus, 7-deazapurine analogues were
employed, namely N9-propyl-7-deazaadenine (7CAde) and
N9-propyl-7-deazaguanine (7CGua), and their interactions
toward 1Cu were evaluated. This strategy has been already
proven to be successful to study the interactions of metal ions
toward Watson−Crick positions.11,12
upon consideration of the hydrogen bonds involving the
coordinated water molecule O2w, a 1D double-strand motif
running along the [001] direction can be envisaged (Figure 3).
The strands interact through antiparallel π−π stacking
interactions between the five- and six-membered rings of
Ade (distance between the centroids 3.552(3) Å, α = 2.30°, β
= 19.93°, γ = 22.26°) (Figure 3 and Figure S4). Further
hydrogen bond interactions define a 3D supramolecular
network, in which two graphs set motifs of the kinds
R3 (10) and R3 (10) can be identified (Figure S5).
4
2
The complex 1Cu_7CAde was obtained upon reaction of
1Cu with 7CAde in water. This complex crystallizes in the
triclinic space group P1 (Table S1). The asymmetric unit
̅
The complex 1Cu_Gua crystallizes in the monoclinic space
group Cc (Table S1). The asymmetric unit consists of one
copper(II) ion, one mcheld ligand, one Gua nucleoside, and
one water molecule, all in general positions. The copper(II)
ion adopts a distorted-square-pyramidal geometry (τ = 0.20)45
(Figure 4). The mcheld ligand shows the μ,κ3-N10,O2,O4
coordination mode (Cu−N10 1.908(4) Å, Cu−O2 2.061(4)
Å, Cu−O4 2.046(4) Å), occupying three basal positions of the
pyramid. The nucleobase Gua is coordinated to the copper(II)
ion through N7 (Cu−N7 1.979(4) Å), occupying the fourth
basal position. The apical position is occupied by a water
molecule (Cu−O1w 2.155(4) Å). An intramolecular hydrogen
bond is present between the water molecule and the keto
group oxygen atom O61 of the nucleobase (Figure 4) (O1w···
O61 2.698(5) Å). Moreover, intermolecular hydrogen bonds56
are observed between the water molecule and the carboxylic
atom O4 of mcheld (O1w···O4 2.746(6) Å), which leads to
the formation of a 1D supramolecular chain of collinear metal
ions 5.3 Å apart running along [010] (Figure S6). Hydrogen
bonds are also present between the guanine N2 amino group
and the carboxylate O5 atom of mcheld and between the
contains one copper(II) ion, one mcheld ligand, one 7CAde
nucleoside, and one water molecule, all in general positions.
The copper(II) ion shows a distorted-square-pyramidal
coordination (τ = 0.10)45 (Figure 5). The mcheld ligand
shows the μ,κ3-N10,O2,O4 coordination mode (Cu−N10
1.922(3) Å, Cu−O4 2.128(2) Å, Cu−O2 2.038(2) Å),
occupying three basal positions. In this case, the nucleoside
7CAde is coordinated to the copper(II) ion through the
Watson−Crick position N1, occupying the fourth basal
position of the pyramid (Cu−N1 1.974(3) Å) (Figure 5).
The apical position is occupied by a water molecule (Cu−O1w
2.210(2) Å) (Figure 5).
The mcheld ligand and the 7CAde nucleoside are not
coplanar, with an angle between their rms planes of 45.47(7)°,
considerably larger than the angle observed when the
[Cu(mcheld)] fragment binds via the Ade-N7 atom (vide
supra). Intra- and intermolecular hydrogen bonds are present
between the 7CAde-N6 amino protons and the carboxylate O2
F
https://dx.doi.org/10.1021/acs.inorgchem.0c01210
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
thus indicating that the DNA helix conformation is not
significantly altered. Therefore, nothing but weak interactions
between 1Cu and the ss-DNA homopolymers can be expected
in solution. This is in contrast with the reported formation of
individual mcheld-Cu-nucleobase base pair mismatches within
double-stranded DNA scaffolds.4 In that case, the copper(II)
complex was covalently attached to the backbone of the DNA
molecule and only one mismatch was introduced within the
helix. On the other hand, in our study, the 1Cu fragments are
unrestricted in solution and the Cu···N(nucleoside) interaction
does not seem to be strong enough to yield a Cu(II)-mediated
base pair in solution. This fact, together with the hindering
effect of several copper(II) ions close to each other, reasonably
prevents the formation of stable Cu-ss-DNA hybrid systems in
solution.
Ab Initio Computational Studies: Structural Features
of 1Cu_nucleoside Pairs inside DNA Molecules. As was
mentioned above, the binding of complex 1Cu toward
canonical nucleosides has been previously reported in solution
at the oligomer level in double-stranded DNA (except for
thymidine),4 while the crystal and molecular structures of the
resulting 1Cu_nucleoside monomers have been determined in
this work. These results prompted us to evaluate the structural
organization of 1Cu_nucleoside pairs inside DNA double-
helix structures and compare the structural features derived
therefrom with those of the monomers described herein, as we
believe that the DNA structure should have an important
influence on the arrangement of these copper−base pairs.
Furthermore, we aimed at evaluating the effect of Cu(II)-
mediated nucleobase formation on the DNA structure that
contains it.
Figure 5. Molecular structure of the complex [Cu(mcheld)-
(N1-7CAde)(H2O)]·H2O (1Cu_7CAde) with thermal ellipsoids set
at the 30% probability level. The intramolecular hydrogen bond is
highlighted by a dashed cyan line. The uncoordinated water molecule
(and the corresponding hydrogen bonds) has been omitted for clarity.
An ordered model has been adopted for the 7CAde propylic residue.
and O3 atoms of the mcheld ligand (N6···O2 2.768(4) Å and
N6···O3 2.960(3) Å, respectively). Also, the water molecule
O1w forms intermolecular hydrogen bonds56 with the donor
atoms of two different adjacent complexes (O1w···O4
2.790(4) Å, O1w···O5 2.717(3) Å, O2w···N3 2.956(4) Å),
leading to a staircase-like 3D structure (Figure 6). No π−π
stacking interactions are present.
Interestingly, no complexes were either isolated or detected
in solution when 1Cu was reacted with thymidine and 7-
deazaguanine derivatives (Scheme 2). This is probably due to
the experimental conditions adopted. Unfortunately, although
several attempts were performed at higher pH (9.5) to fully
deprotonate such nucleosides, no complexes were isolated and
further efforts still need to be done in this regard.
CD Spectroscopy Interaction Studies between Com-
plex 1Cu and Single-Stranded DNA Homopolymers.
The interaction of 1Cu toward single-stranded DNA (ss-
DNA) molecules was studied in an effort to evaluate the
possible formation of a contiguous array of 1Cu_nucleobase
base pairs. To this aim, we used circular dichroism (CD) to
unveil the interactions, at the oligomer level, between 1Cu and
different synthetic ss-DNA homopolymers containing the
nucleoside analogues studied herein: namely, d(Ade)15,
d(7CAde)15, d(Cyt)15, and d(Gua)10. The 1Cu complex is
CD inactive, and its UV−vis absorbance spectrum shows a
broad band in the range 240−280 nm. Therefore, any induced
circular dichroism (ICD) signal caused by the interaction of
1Cu with the DNA homopolymers should be developed in this
region. Unfortunately, in all cases, the addition of 1Cu did not
bring about significant changes in the CD spectra (Figure S8),
The study of a DNA structure, in the solid state or solution,
can be very difficult and expensive due to the difficulty in
obtaining single crystals suitable for X-ray diffraction measure-
ments or powdered batches in enough quantity to accomplish
nuclear magnetic resonance studies (in the present case the
latter studies are made even more difficult by the presence of
the paramagnetic Cu(II) ions). More affordable spectroscopic
or spectrometric studies (for instance, UV−vis absorption and
fluorescence spectroscopy, CD, ESI-MS, MALDI-TOF) do not
provide information on the three-dimensional structure. For
this reason, theoretical computational studies can be a very
important tool.
Figure 6. (A) In the crystal structure of 1Cu_7CAde, hydrogen bonds (dashed cyan lines) involving the coordinated water molecule O1w and the
exocyclic 7CAde amino group lead to the formation of staircase motifs. (B) Detail of the symmetry-related hydrogen bond interactions present in
each “step” of the staircase. Red dotted lines represent hanging contacts.
G
https://dx.doi.org/10.1021/acs.inorgchem.0c01210
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
Figure 7. Ab initio HF-3c optimized geometries for the structures DNA3-Na (left) and DNA3-H (right) hosting the 1Cu_Cyt complex.
Table 1. Comparison of the Relevant Bond Distances (Å) and Dihedral Angles (deg) of the 1Cu_Cyt Base Pair in the DNA3
a
Optimized Models with the Experimental Values
b
c
d
exptl 1Cu_Cyt
DNA3-Na
DNA3-H
DNA3-H-SE
Cu−N10
Cu−O2
Cu−O4
Cu−Oaxial
Cu−N3
Cu−O21
interplanar angle
C1′Gua···C1′Cyt
C1′1Cu···C1′Cyt
C1′Gua···C1′Cyt
1.893
1.997
2.027
2.407
1.956
2.711
77.7
1.902 (0.009)
1.900 (−0.097)
1.900 (−0.127)
2.527 (0.120)
1.939 (−0.017)
2.844 (0.133)
47.6 (−30.1)
10.509
1.924 (0.031)
1.908 (−0.089)
1.9 (−0.127)
2.504 (0.097)
1.965 (0.009)
2.957 (0.246)
36.6 (−41.1)
10.656
1.966 (0.073)
1.923 (−0.074)
1.916 (−0.111)
2.488 (0.081)
1.977 (0.021)
2.948 (0.237)
34.9 (−42.8)
10.668
8.759
10.760
9.284
10.682
9.566
10.688
a
b
c
Deviations with respect to the latter are shown in parentheses. Na indicates the use of sodium(I) as a cation. H indicates the use of a proton as a
d
cation. SE indicates that the “solvent effect” has been included in the calculation.
We have employed computational calculations to study the
structure of a 10-mer and a 15-mer DNA duplex model
containing one Cu(II)-mediated base pair covalently attached
to the DNA structure. Although they could be considered
simple, these models enable us to study a realistic structure.
Within the available theoretical models, molecular mechanics is
not a valid method in the presence of metal ions such as
Cu(II), while semiempirical models, although parametrized,
can result in significant errors in the estimation of bond
distances and angles.
Therefore, we strongly believe that the most useful methods
to approach a DNA molecule of ca. 1000 atoms and containing
metal ions are the ab initio and QMM methodologies. The
latter approach, though faster, has the problem of validation. In
our case, we do not have a reliable starting DNA structure to
be validated. Thus, we decided not to use an approach that
could lead to artifacts but rather to carry out a theoretical−
computational structural study using the ab initio HF-3c
method for DNA molecules holding the 1Cu_Cyt or
1Cu_Gua base pair located in the middle of the double
helix. We have selected these Cu(II)−base pairs, as they
exhibited the longest deviation from coplanarity between the
[Cu(mcheld)] fragment and the Cyt or Gua bases in their
crystal structure (see Synthesis and Crystal Structure of
mcheld-Cu-Nucleoside Complexes), and thus they will
potentially have a greater effect on the DNA double-helix
conformation.
the 1Cu_Cyt or 1Cu_Gua complexes. To the best of our
knowledge, this is the first time that a duplex DNA molecule
containing metal-mediated base pairs has been studied at this
computational level.
Single Model Complex. Initially, we carried out a
theoretical study of the 1Cu_Cyt molecular structure, starting
from the experimentally determined one, to validate the
theoretical ab initio HF-3c method employed in our following
studies. After this preliminary study, we compared the
computational results (1Cu_Cyt-S) with the experimental
results, giving special attention to the interaction of the Cyt
base with the copper(II) ion. Subsequently, we performed
additional studies taking into account solvent effects
(1Cu_Cyt-SE), using water as the solvent, and removing the
coordinated water molecule (1Cu_Cyt-NW) to evaluate the
consequence of the absence of the sixth coordination position
observed experimentally.
For 1Cu_Cyt-S, the results show a good agreement among
the calculated and observed Cu−O and Cu−N distances
(Table S2), with the greatest deviations observed for the Cu−
O2 and Cu−O4 distances and for the angle between the
[Cu(mcheld)] fragment and Cyt base (by 10°; see Figure S9).
When the calculations were repeated with the solvent effects
(1Cu_Cyt-SE) or without the coordinated water molecule
(1Cu_Cyt-NW) taken into account, the results were very
similar, and no significant differences were observed.
Model Systems of Three Base Pairs. After validating the
theoretical model of 1Cu_Cyt with the experimental
observations, we set out to study how the scaffolding of a
DNA molecule would affect the conformation of the 1Cu_Cyt
base pair. In an effort to work with the simplest DNA model,
we studied a DNA duplex structure comprising only three base
To this purpose, we have followed a bottom-up strategy,
starting with the study of the 1Cu_Cyt complex, followed by
the study of a 3-mer DNA molecule containing the 1Cu_Cyt
base pair in the center of the structure, and finally performing
calculations for 10-mer and 15-mer DNA molecules containing
H
https://dx.doi.org/10.1021/acs.inorgchem.0c01210
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
pairs (DNA3), constructed with the sequences 5′-(G 1Cu G)-
3′ and 5′-(CCC)-3′. This DNA3 molecule represents a good
starting point before studying larger DNA molecules (vide
infra) because it takes into account the influence of the base
pairs above and below the 1Cu_Cyt base pair, which has been
positioned in the middle.
For these calculations, the counteranions (namely the
phosphate groups) were treated with either sodium cations
(DNA3-Na) or protons (DNA3-H) and the results obtained
with the two strategies were compared. We also evaluated the
influence of the solvent effect on the latter model (DNA3-H-
SE).
In all cases, the copper(II) ion is five-coordinated in a
square-pyramidal geometry. The mcheld ligand shows the μ,κ3-
N10,O2,O4 coordination mode, occupying three equatorial
positions, while the Cyt ligand acts as a monodentate ligand via
N3 coordination, completing the basal plane. Remarkably, the
axial position of the copper(II) coordination sphere now
involves a neighboring nucleotide, and the copper(II) ions
bind the guanine-O6 atom (Figure 7). This situation is similar
to that observed in the crystal structure of a related copper-
mediated base pair in DNA.20
The results show that in all cases there is a good agreement
among the experimental and calculated bond distances
involving the copper(II) ions (Table 1). The greatest
deviations are again observed for the Cu−O2 and Cu−O4
distances and the interplanar angle between the [Cu(mcheld)]
fragment and the cytosine base. Both DNA3 models show a
significantly lower interplanar angle in comparison to the
complex alone, which can be explained as a result of the
influence of the DNA scaffold: the presence of the base pairs
above and below the 1Cu_Cyt base pair reasonably limits the
rotation around the Cu−N3 bond. In this regard, the best
agreement was found with the DNA3-Na model, which exhibits
an interplanar angle closer to the observed angle.
Figure 8. Close view of the HF-3c optimized geometry for DNA3-Na
showing the position of two sodium ions (purple) bridging
contiguous phosphate groups.
designed to contain only guanosine and cytidine bases (Table
2).
Table 2. DNA Duplexes Employed in the Ab Initio
Calculations
duplex
sequence
DNA15
5′-(CCC CCC C C C CCC CCC)-3′
3′-(GGG GGG G 1Cu G GGG GGG)-5′
5′-(CCC C 1Cu C CCC C)-3′
DNA10
The glyosidic C1′···C1′ distances between opposite
nucleobases are within the range 10.5−10.8 Å observed for
natural base pairs, while they are significantly smaller for the
1Cu_Cyt base pair (<9.56 Å), especially in the DNA3-Na
model, which displays a distance of 8.76 Å.
3′-(GGG G G G GGG G)-5′
First, we studied the structure of the 15-mer duplex DNA15,
compensating the negative charge of the phosphate groups
with the presence of sodium cations (DNA15-Na). It is worth
noting that this study required an extensive computational
effort, as up to 28 sodium cations were considered in the ab
initio calculations.
We also performed the same calculation using protons as
counterions (DNA15-H) to reduce the computational cost, and
the results were compared to those obtained with DNA15-Na
to evaluate the consequences of this approach. Finally, we
studied duplex DNA10 by treating the phosphate groups with
protons (DNA10-H). The solvent effect was not taken into
account in any case, due to the great calculation effort that this
would have required.
The optimized geometry of duplex DNA15-Na contains 15
base pairs and completes an entire turn, holding a right-handed
B-like conformation (Figure 9). The 1Cu_Cyt base pair is
placed at site 8, and copper(II) is six-coordinated and adopts a
distorted-octahedral geometry, as seen for the 1Cu_Cyt
complex. The mcheld ligand shows the expected μ,κ3-
N10,O2,O4 coordination mode (Cu−N10 1.916 Å, Cu−O2
1.900 Å, Cu−O4 1.906 Å), occupying three basal positions.
However, in contrast to the studied trimer DNA3-Na, where
the Cyt base acts as a monodentate ligand via N3 coordination,
in duplex DNA15-Na the Cyt base shows a μ,κ2-N3,O21
Interestingly, the presence of the 1Cu_Cyt complex seems
to affect more the conformation of the strand that holds the
1Cu metal fragment. This is highlighted by the shorter P···P
distance (5.94 Å) of contiguous phosphate groups attached to
the sugar unit of [Cu(mcheld)] vs the other P···P distances in
the duplex (>6.39 Å). This is situation is more pronounced in
the case of the DNA3-Na model, where the phosphate groups
are separated by only 4.2 Å and seem to be stabilized by two
nearby sodium cations (Na···Na 3.766 Å) that bridge both
phosphate groups (Na···O 2.068−2.132 Å) (Figure 8).
It is worth mentioning that, since a water molecule was
found to be involved in the coordination sphere of copper(II)
in the 1Cu_Cyt complex, the DNA3-H model was also studied
in the presence of a water molecule placed near the metal ion.
In this situation, the results also showed the formation of the
Cu(II)−Gua(O6) bond (Figure S10).
Full DNA Model Systems. Finally, we performed a
theoretical study for 15-mer (DNA15) and 10-mer (DNA10)
duplexes containing the 1Cu_Cyt and 1Cu_Gua base pairs,
respectively, in the middle of the double helixes. In an effort to
reduce the influence of arbitrary DNA sequences surrounding
the copper-mediated base pair, the model sequences have been
I
https://dx.doi.org/10.1021/acs.inorgchem.0c01210
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
in the DNA3-Na model (43.63°). This evidence shows that our
original assumptions were not entirely correct; larger DNA
molecules are not necessarily more constrained. Indeed, the
rotation around the Cu−N3 bond inside the helix is allowed,
leading to arrangements far from coplanarity. As a con-
sequence, the Cu−O21 distance in our optimized DNA15-Na
model is very similar to that observed in 1Cu_Cyt, indicating
that this interaction could also occur in a duplex structure. This
unexpected result indicates that the adjacent base pairs are not
the only ones that influence the organization of the 1Cu_Cyt
base pair, but instead, the overall structure of a large DNA
molecule has an effect. This could also be explained by the fact
that the energy required to rotate the Cu−N3 bond and
diminish the deviation from planarity of the complex can be
dampened by a higher number of slight deviations from the
optimal bond geometry of the higher number of bonds present
in larger DNA structures.
In the duplex DNA15-Na, the C1′···C1′ distances between
opposite 2′-deoxyriboses of natural base pairs vary in the range
10.5−10.7 Å, which is consistent with a standard B-form
conformation. However, as foreseen by our previous DNA3
models, at the copper−base pairs level (site 8) the C1′···C1′
distance is considerably shorter (8.3 Å), causing a narrowing of
the double helix. The distortion caused by the 1Cu_Cyt base
pair in the duplex structure is analogous to what was observed
for DNA3-Na. Actually, two sodium(I) cations are also very
close to each other (Na···Na 3.741 Å) and bridge two
contiguous phosphate groups (Na···O 2.093−2.128 Å). In
addition, stacking interactions occur between the mcheld
ligand and the contiguous guanosine base to which the
copper(II) ions coordinate through O6 coordination. The rest
of the DNA structure presents a characteristic B-form
conformation, without other remarkable changes. Therefore,
the presence of 1Cu_Cyt appears to only distort neighboring
base pairs and is suitable to be formed in a DNA structure
without major disruption of the structure.
Figure 9. (left) The geometry-optimized duplex DNA15-Na. (right)
Closer view of the 1Cu_Cyt base pair formed in the duplex structure
showing the apical Cu−Gua-O6 and the Cu−Cyt-μ-N3,O21 bonds.
Dots surrounding the double helix represent sodium cations which
balance the negative charge of the phosphate DNA backbone. Color
scheme: black, DNA structure different from the Cu(II)-mediated
base pairs; green, neighboring nucleobases to 1Cu_Cyt base pair.
coordination mode (Cu−N3 1.964, Cu−O21 2.703 Å), as seen
in the X-ray crystal structure of the 1Cu_Cyt complex (Table
3).
The axial position of the copper(II) coordination sphere
involves a neighboring nucleotide: i.e., the copper(II) ion
binds a neighboring Gua-O6 atom (Cu−O6 2.444 Å). It
should be noted that this distance is considerably shorter than
the sum of the Cu/O van der Waals radii. This situation has
been also observed for the DNA3-Na model, which displays a
longer distance (2.527 Å), as well as for the crystal structure of
the copper-mediated Dipic-Py base pair in DNA (Dipic =
pyridine-2,6-dicarboxylate; Py = pyridine), although in that
case the Cu−O6 distance was considerably longer (3.1 Å).24
The fact that our ab initio calculation reveals the formation of
the Cu−O axial bond with an adjacent guanidine nucleoside,
for both DNA3-Na and DNA15-Na models, as reported for the
DNA crystal structure, constitutes a validation that supports
our theoretical method to study these copper-mediated base
pairs in DNA duplexes.
The structure of DNA15-H is shown in Figure 10. The
optimized copper(II) ion is in a five-coordinated square-
pyramidal geometry. The mcheld ligand shows the expected
μ,κ3-N10,O2,O4 coordination mode (Cu−N10 1.927 Å, Cu−
O2 1.910 Å, Cu−O4 1.901 Å), occupying three basal positions.
However, in contrast to the molecular structures of 1Cu_Cyt
and DNA15-Na, where the Cyt ligand shows a μ,κ2-N3,O21
coordination mode (vide supra), in duplex DNA15-H the Cyt
base acts as a monodentate ligand via N3 coordination,
Interestingly, the optimized geometry of DNA15-Na reveals a
larger interplanar angle (54.43°) between the [Cu(mcheld)]
fragment and the cytidine base in comparison to that observed
Table 3. Comparison of the Relevant Bond Distances (Å) and Dihedral Angles (deg) in the Crystal Structures of the 1Cu_Cyt
a
Base Pair and the DNA3-Na, DNA15-Na and DNA15-H Optimized Models
b
c
exptl 1Cu_Cyt
DNA3-Na
DNA15-Na
DNA15-H
Cu−N10
Cu−O2
Cu−O4
Cu−Oaxial
Cu−N3
Cu−O21
interplanar angle
C1′Gua···C1′Cyt
C1′1Cu···C1′Cyt
C1′Gua···C1′Cyt
1.893
1.997
2.027
2.407
1.956
2.711
77.7
1.902 (0.009)
1.900 (−0.097)
1.900 (−0.127)
2.527 (0.120)
1.939 (−0.017)
2.844 (0.133)
47.6 (−30.1)
10.509
1.916 (0.023)
1.900 (−0.097)
1.906 (−0.121)
2.444 (0.037)
1.964 (0.008)
2.703 (−0.008)
54.4 (−23.3)
10.547
1.927 (0.034)
1.910 (−0.087)
1.901 (−0.126)
2.470 (0.063)
1.965 (0.009)
2.968 (0.257)
38.9 (−38.9)
10.681
8.759
10.760
8.347
10.694
9.263
10.624
a
b
c
Deviations with respect to the observed crystal structure are shown in parentheses. Na indicates the use of sodium(I) as a cation. H indicates the
use of a proton as a cation.
J
https://dx.doi.org/10.1021/acs.inorgchem.0c01210
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
helix contains 10 base pairs, holds a right-handed B-like form,
and nearly completes a turn, with all the complementary
cytidine and guanosine bases assembled via Watson−Crick
hydrogen bonds along with stacking interactions. The
[Cu(mcheld)] fragment is placed at site 5 (starting from the
5′-end), in the middle of the duplex, and binds guanosine 5 via
the N7 atom, as observed in 1Cu_Gua (vide supra). The
copper(II) ion adopts a square-planar geometry, as previously
described in related systems,24 with no additional axial
coordination observed. The mcheld ligand shows the expected
μ,κ3-N10,O2,O4 coordination mode, and the copper(II) ion is
involved in shorter bonds (Cu−N10 1.925 Å, Cu−O2 1.884 Å,
Cu−O4 1.903 Å) than in the crystal structure of 1Cu_Gua.
The guanosine moiety is coordinated to the copper(II) ion
through N7 with a distance (Cu−N7 1.975 Å) very close to
that present in the crystal structure. However, the spatial
assembly of the units has considerably changed in comparison
to the free 1Cu_Gua monomer. In duplex DNA10-H, the
[Cu(mcheld)] fragment and the guanosine base self-assemble
with a lower deviation from coplanarity, with an interplanar
angle of 19.6°, instead of the 78.94(10)° angle observed in the
crystal structure. Surprisingly, this arrangement appears to be
possible despite the short distance observed between the
guanosine keto group and the mcheld carboxylic atom (O61···
O2 2.866 Å), which would be expected to generate a
significant repulsion. This fact could be due to the strong
influence that the overall duplex structure must have on the
copper−base pair, indicating that the formation of the double
helix may restrict the free rotation around the Cu−N7 bond,
hampering the separation of the keto and carboxylic groups
any further.
Figure 10. (left) The geometry-optimized duplex DNA15-H. (right)
Closer view of the 1Cu_Cyt base pair formed in the duplex structure
showing the Cu-Gua-O6 and Cu-Cyt-N3 bonds Color scheme: black,
DNA structure different from the Cu(II)-mediated base pairs; green,
neighboring nucleobases to 1Cu_Cyt base pair.
completing the basal plane (Cu−N3 1.965 Å). This
organization can be explained as a result of a smaller
interplanar angle between the [Cu(mcheld)] fragment and
the cytidine base (38.88°) and consequently a greater Cu−
O21 distance (2.968 Å). The axial position of the copper(II)
coordination sphere also involves a neighboring nucleotide:
i.e., the copper(II) ion binds a neighboring Gua-O6 atom
(Cu−O6 2.470 Å), as observed in DNA15-Na. In this case, the
C1′···C1′ distance between opposite 2′-deoxyriboses at the
copper−base pairs level (site 8) is larger (9.2 Å), although the
distance for natural base pairs remains within the expected
10.5−10.7 Å range. Importantly, the rest of the DNA15-H
structure does not show any major differences vs DNA15-Na.
These results indicate that our ab initio calculations
performed for the DNA molecules containing one 1Cu_Cyt
base pair can be successfully performed using sodium(I) ions
or protons as counterions to compensate for the negative
charge of the phosphate groups. The choice of one strategy or
another will depend on the objective of the study. The DNA15-
Na model is more accurate by comparison with the observed
molecular structure of 1Cu_Cyt, at the expense of a
considerably higher computational cost. Replacing Na+ ions
with protons (DNA15-H) requires less demanding and more
affordable computational resources. However, this model can
lead to variations in the angles and bond distances of the
1Cu_Cyt base pair, although the overall DNA structure is little
affected. Therefore, these considerations must be taken into
account depending on the aim of the calculation: namely, to
study the overall DNA structure containing metal-mediated
base pairs or to evaluate the arrangement of a specific base pair
in the DNA structure. Although experimental structural data
are required to evaluate the conformation in solution, namely
high-resolution NMR data, our result indicates that the 1Cu-
Cyt base pair will exist inside the double helix with a large
interplanar angle, as observed in the DNA15 models.
CONCLUSIONS
■
The formation of copper-mediated base pairs containing a
pyridine-2,6-dicarboxylate derivative (mcheld) and model
nucleosides (Ade, Cyt, Gua, and 7CAde) was studied in the
solid state and solution by means of single-crystal X-ray
diffraction and CD spectroscopy, respectively. The former
revealed the crystal and molecular structures of four mcheld-
Cu-nucleoside systems, namely 1Cu_Cyt, 1Cu_Ade,
1Cu_7CAde, and 1Cu_Gua, which have been reported to
occur individually in double-stranded DNA molecules.4
Unfortunately, no single crystals suitable for X-ray diffraction
could be isolated for the 1Cu_7CGua compound. The
versatility of these nucleoside analogues is evidenced in the
case of adenine derivatives, where the coordination to the
metal ion can be directed via either the Hoogsteen N7 position
(as in 1Cu_Ade), or the Watson−Crick N1 atom (as in
1Cu_7CAde). Although the interaction metal−N1-Gua has not
been observed in the present work, its formation cannot be
ruled out; indeed, it is expected for the 1Cu_7CGua complex
and it has been previously reported for silver(I) ions.54 The
complex 1Cu_Cyt demonstrates that the binding of 1Cu could
occur via both keto and amino groups simultaneously. The
presence of a keto group in the nucleobase moiety induces a
nearly perpendicular disposition of the [Cu(mcheld)] frag-
ment vs the Gua or Cyt residues, as observed in 1Cu_Gua and
1Cu_Cyt, respectively, which could be attributed to the
reciprocal repulsion of the keto and carboxylate groups.
Interestingly, this high deviation from coplanarity between the
[Cu(mcheld)] moiety and the nucleosides is not present in the
1Cu_Ade and 1Cu_7CAde derivatives, most likely due to the
cooperation of the coordination bonds along with an
In an effort to assess the effect of the 1Cu_Gua base pair in
the DNA structure, we also performed a calculation using
protons as counterions (DNA10-H), as the computational cost
was considerably lower (Figure S10). The DNA10-H double
K
https://dx.doi.org/10.1021/acs.inorgchem.0c01210
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
Authors
intramolecular hydrogen bond interaction involving the
exocyclic amino group NH2. These results are in agreement
with the stabilization trend observed for the Cheld-Cu-
nucleobase system inside DNA duplexes (adenine > cytosine
> guanine),4 since a lower deviation from coplanarity of the
units can facilitate the formation of intermolecular π−π
interactions and intramolecular hydrogen bonds between
adenine and Cheld.
́
Alicia Dominguez-Martin − Departamento de Quimica
Inorganica, Universidad de Granada, 18071 Granada, Spain;
orcid.org/0000-0001-8669-6712
Simona Galli − Dipartimento di Scienza e Alta Tecnologia,
Universita dell’Insubria, 22100 Como, Italy; orcid.org/
́
̀
0000-0003-0335-5707
́
́
Jose A. Dobado − Grupo de Modelizacion y Diseno Molecular,
̃
́
́
Departamento de Quimica Organica, Universidad de Granada,
18071 Granada, Spain
Importantly, the ab initio calculations performed for
duplexes DNA15 and DNA10, which embrace the 1Cu_Cyt
and 1Cu_Gua base pairs in the middle of the duplexes,
respectively, revealed that these base pairs can be formed
inside DNA duplexes without significant distortion of the
natural base pair arrangement. As a matter of fact, the
geometry-optimized structures for duplexes DNA15 and
DNA10 show that the 1Cu_nucleoside base pairs remain in
good harmony with the rest of the canonical Watson−Crick
base pairs. The ab initio methodologies also revealed that the
phosphate groups can be treated with Na+ ions or protons,
depending on the purposes of the computational calculations.
A better agreement among the 1Cu_Cyt calculated and
experimental bond distances was found for the DNA15-Na
model, using Na+ ions as counterions. However, the calculation
performed on the DNA15-H model, using protons as
counterions, revealed some small deviations for the relevant
bond distances at the Cu(II) ion, but the overall DNA
structure can be correctly studied using this approach.
These results demonstrate that the organization of Cu−
nucleobase base pairs can be studied in the solid state and in
the context of a double helix using ab initio methodologies to
evaluate the potential effects derived from the DNA scaffold,
which may drive both metal coordination and the supra-
molecular assembly. These results can be extended to other
metal-mediated base pairs.
́
Noelia Santamaría-Díaz − Departamento de Quimica
Inorganica, Universidad de Granada, 18071 Granada, Spain
Antonio Perez-Romero − Departamento de Quimica
Inorganica, Universidad de Granada, 18071 Granada, Spain
́
́
́
́
Complete contact information is available at:
https://pubs.acs.org/10.1021/acs.inorgchem.0c01210
Author Contributions
The manuscript was written through contributions of all
authors. All authors have given approval.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
Financial support from the Spanish MINECO (CTQ2017-
89311-P), FEDER/Junta de Andalucia-Consejeria de Econo-
mia y Conocimiento/Proyecto (A-FQM-465-UGR18), Junta
■
́
́
́
́
de Andalucia (FQM-2293, FQM-283, FQM-174), and Unidad
de Excelencia de Quimica aplicada a Biomedicina y
Medioambiente (UGR) are acknowledged. S.G. acknowledges
Universita dell’Insubria for partial funding. We also thank the
“Centro de Servicios de Informatica y Redes de Comunica-
ciones” (CSIRC) (UGRGrid), University of Granada, for
providing computing time.
́
̀
́
REFERENCES
ASSOCIATED CONTENT
■
■
(1) Jones, M. R.; Seeman, N. C.; Mirkin, C. A. Programmable
Materials and the Nature of the DNA Bond. Science 2015, 347,
1260901.
sı
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acs.inorgchem.0c01210.
(2) Rothemund, P. W. K. Folding DNA to Create Nanoscale Shapes
and Patterns. Nature 2006, 440, 297−302.
Crystallographic information and relevant X-ray dif-
fraction experimental details, additional figures of the
crystal structures, circular dichroism spectra, additional
figures for ab initio optimized geometries, and Cartesian
coordinates (Å) for geometry-optimized molecules
(PDF)
(3) Wilner, O. I.; Willner, I. Functionalized DNA Nanostructures.
Chem. Rev. 2012, 112, 2528−2556.
(4) Meggers, E.; Holland, P. L.; Tolman, W. B.; Romesberg, F. E.;
Schultz, P. G. A Novel Copper-Mediated DNA Base Pair. J. Am.
Chem. Soc. 2000, 122, 10714−10715.
(5) Jash, B.; Muller, J. Metal-Mediated Base Pairs: From Character-
̈
ization to Application. Chem. - Eur. J. 2017, 23, 17166−17178.
(6) Tanaka, K.; Shionoya, M. Synthesis of a Novel Nucleoside for
Alternative DNA Base Pairing through Metal Complexation. J. Org.
Chem. 1999, 64, 5002−5003.
Accession Codes
CCDC 1835231−1835235 contain the supplementary crys-
tallographic data for this paper. These data can be obtained
free of charge via www.ccdc.cam.ac.uk/data_request/cif, or by
emailing data_request@ccdc.cam.ac.uk, or by contacting The
Cambridge Crystallographic Data Centre, 12 Union Road,
Cambridge CB2 1EZ, UK; fax: +44 1223 336033.
(7) Takezawa, Y.; Shionoya, M. Metal-Mediated DNA Base Pairing:
Alternatives to Hydrogen-Bonded Watson−Crick Base Pairs. Acc.
Chem. Res. 2012, 45, 2066−2076.
(8) Takezawa, Y.; Muller, J.; Shionoya, M. Artificial DNA Base
̈
Pairing Mediated by Diverse Metal Ions. Chem. Lett. 2017, 46, 622−
633.
(9) Polonius, F.-A.; Muller, J. An Artificial Base Pair, Mediated by
̈
AUTHOR INFORMATION
Hydrogen Bonding and Metal-Ion Binding. Angew. Chem., Int. Ed.
2007, 46, 5602−5604.
■
Corresponding Author
(10) Megger, D. A.; Fonseca-Guerra, C.; Hoffmann, J.; Brutschy, B.;
́
́
Miguel A. Galindo − Departamento de Quimica Inorganica,
Universidad de Granada, 18071 Granada, Spain;
orcid.org/0000-0003-4355-4313; Email: magalindo@
ugr.es
Bickelhaupt, F. M.; Muller, J. Contiguous Metal-Mediated Base Pairs
̈
Comprising Two Ag(I) Ions. Chem. - Eur. J. 2011, 17, 6533−6544.
́
(11) Santamaría-Díaz, N.; Mendez-Arriaga, J. M.; Salas, J. M.;
Galindo, M. A. Highly Stable Double-Stranded DNA Containing
L
https://dx.doi.org/10.1021/acs.inorgchem.0c01210
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
Sequential Silver(I)-Mediated 7-Deazaadenine/Thymine Watson-
(30) Nowick, J. S.; Chen, J. S.; Noronha, G. Molecular Recognition
in Micelles: The Roles of Hydrogen Bonding and Hydrophobicity in
Adenine-Thymine Base-Pairing in SDS Micelles. J. Am. Chem. Soc.
1993, 115, 7636−7644.
Crick Base Pairs. Angew. Chem., Int. Ed. 2016, 55, 6170−6174.
́
(12) Mendez-Arriaga, J. M.; Maldonado, C. R.; Dobado, J. A.;
Galindo, M. A. Silver(I)-Mediated Base Pairs in DNA Sequences
Containing 7-Deazaguanine/Cytosine: Towards DNA with Entirely
Metallated Watson-Crick Base Pairs. Chem. - Eur. J. 2018, 24, 4583−
4589.
̈
(31) Vermonden, T.; Branowska, D.; Marcelis, A. T. M.; Sudholter,
E. J. R. Synthesis of 4-Functionalized Terdendate Pyridine-Based
Ligands. Tetrahedron 2003, 59, 5039−5045.
(32) APEX3 Software,V2016.1; Bruker AXS: Madison, WI, USA,
2016.
(13) Toomey, E.; Xu, J.; Vecchioni, S.; Rothschild, L.; Wind, S.;
Fernandes, G. E. Comparison of Canonical versus Silver(I)-Mediated
Base-Pairing on Single Molecule Conductance in Polycytosine
DsDNA. J. Phys. Chem. C 2016, 120, 7804−7809.
(33) Sheldrick, G. M. SADABS 2016/2, Program for Empirical
̈
Absorption Correction of Area Detector Data; University of Gottingen:
(14) Jash, B.; Scharf, P.; Sandmann, N.; Fonseca Guerra, C.; Megger,
̈
Gottingen, Germany. 2016.
D. A.; Muller, J. A Metal-Mediated Base Pair That Discriminates
̈
between the Canonical Pyrimidine Nucleobases. Chem. Sci. 2017, 8,
1337−1343.
(15) Al-Mahamad, L. L. G.; El-Zubir, O.; Smith, D. G.; Horrocks, B.
R.; Houlton, A. A Coordination Polymer for the Site-Specific
Integration of Semiconducting Sequences into DNA-Based Materials.
Nat. Commun. 2017, 8, 720.
(34) Sheldrick, G. M. A Short History of SHELX. Acta Crystallogr.,
Sect. A: Found. Crystallogr. 2008, A64, 112−122.
(35) (a) Hanwell, M. D.; Curtis, D. E.; Lonie, D. C.;
Vandermeerschd, T.; Zurek, E.; Hutchison, G. R. Avogadro: An
Advanced Semantic Chemical Editor, Visualization, and Analysis
Platform. J. Cheminf. 2012, 4, 17. (b) Avogadro: an open-source
molecular builder and visualization tool. Version 1.XX; http://avogadro.
cc/.
(16) Liu, S.; Clever, G. H.; Takezawa, Y.; Kaneko, M.; Tanaka, K.;
Guo, X.; Shionoya, M. Direct Conductance Measurement of
Individual Metallo-DNA Duplexes within Single-Molecule Break
Junctions. Angew. Chem., Int. Ed. 2011, 50, 8886−8890.
(17) Xu, E.; Lv, Y.; Liu, J.; Gu, X.; Zhang, S. An Electrochemical
Study Based on Thymine-Hg-Thymine DNA Base Pair Mediated
Charge Transfer Processes. RSC Adv. 2015, 5, 49819−49823.
(36) Neese, F. The ORCA Program System. Wiley Interdiscip. Rev.:
Comput. Mol. Sci. 2012, 2, 73−78.
(37) Neese, F. Software Update: The ORCA Program System,
Version 4.0. Wiley Interdiscip. Rev.: Comput. Mol. Sci. 2018, 8, e1327.
(38) Sure, R.; Grimme, S. Corrected Small Basis Set Hartree-Fock
Method for Large Systems. J. Comput. Chem. 2013, 34, 1672−1685.
(39) Kruse, H.; Grimme, S. A Geometrical Correction for the Inter-
and Intra-Molecular Basis Set Superposition Error in Hartree-Fock
and Density Functional Theory Calculations for Large Systems. J.
Chem. Phys. 2012, 136, 154101.
(18) Muller, J. Nucleic Acid Duplexes with Metal-Mediated Base
̈
Pairs and Their Structures. Coord. Chem. Rev. 2019, 393, 37−47.
(19) Fortino, M.; Marino, T.; Russo, N. Theoretical Study of Silver-
Ion-Mediated Base Pairs: The Case of C-Ag-C and C-Ag-A Systems. J.
Phys. Chem. A 2015, 119, 5153−5157.
(40) Becke, A. D.; Johnson, E. R. A Density-Functional Model of the
Dispersion Interaction. J. Chem. Phys. 2005, 123, 154101.
(41) Johnson, E. R.; Becke, A. D. A Post-Hartree−Fock Model of
Intermolecular Interactions. J. Chem. Phys. 2005, 123, 024101.
(42) Johnson, E. R.; Becke, A. D. A Post-Hartree-Fock Model of
Intermolecular Interactions: Inclusion of Higher-Order Corrections. J.
Chem. Phys. 2006, 124, 174104.
(43) (a) Marenich, A. V.; Cramer, C. J.; Truhlar, D. G. Universal
Solvation Model Based on Solute Electron Density and on a
Continuum Model of the Solvent Defined by the Bulk Dielectric
Constant and Atomic Surface Tensions. J. Phys. Chem. B 2009, 113,
6378−6396. (b) Tomasi, J.; Benedetta, M.; Cammi, R. Quatum
Mechanical Continuum Solvation Models. Chem. Rev. 2005, 105 (8),
2999−3094.
(44) Harnden, M. R.; Jarvest, R. L.; Bacon, T. H.; Boyd, M. R.
Synthesis and Antiviral Activity of 9-[4-Hydroxy-3-(Hydroxymethyl)-
but-l-Yl]Purines. J. Med. Chem. 1987, 30, 1636−1642.
(45) Addison, A. W.; Rao, T. N.; Reedijk, J.; Van Rijn, J.; Verschoor,
G. C. Synthesis, Structure, and Spectroscopic Properties of Copper-
(II) Compounds Containing Nitrogen-Sulphur Donor Ligands; the
Crystal and Molecular Structure of Aqua[l,7-Bis(N-Methylbenzimi-
dazol-2′-Yl)- 2,6-Dithiaheptane]Copper(II) Perchlorate. J. Chem. Soc.,
Dalton Trans. 1984, 7, 1349−1356.
(46) Hong-Ling, G.; Jian-Zhong, C. Hydrothermal Synthesis and
Crystal Structures of Complex [Cu(CAM)(H2O)2]. Chin. J. Inorg.
Chem. 2007, 23, 1072−1074.
(47) Janes, R.; Moore, E. A.; Open University. Metal-Ligand Bonding;
Royal Society of Chemistry: London, 2004; pp 21−27.
(48) Mirzaei, M.; Eshtiagh-Hosseini, H.; Karrabi, Z.; Notash, B.;
(20) Li, G.; Liu, H.; Chen, X.; Zhang, L.; Bu, Y. Multi-Copper-
Mediated DNA Base Pairs Acting as Suitable Building Blocks for the
DNA-Based Nanowires. J. Phys. Chem. C 2011, 115, 2855−2864.
(21) Swasey, S. M.; Rosu, F.; Copp, S. M.; Gabelica, V.; Gwinn, E.
G. Parallel Guanine Duplex and Cytosine Duplex DNA with
Uninterrupted Spines of Ag I -Mediated Base Pairs. J. Phys. Chem.
Lett. 2018, 9, 6605−6610.
̈
(22) Schonrath, I.; Tsvetkov, V. B.; Zatsepin, T. S.; Aralov, A. V.;
Muller, J. Silver(I)-Mediated Base Pairing in Parallel-Stranded DNA
̈
Involving the Luminescent Cytosine Analog 1,3-Diaza-2-Oxophenox-
azine. JBIC, J. Biol. Inorg. Chem. 2019, 24, 693−702.
(23) Zimmermann, N.; Meggers, E.; Schultz, P. G. A Second-
Generation Copper(II)-Mediated Metallo-DNA-Base Pair. Bioorg.
Chem. 2004, 32, 13−25.
(24) Atwell, S.; Meggers, E.; Spraggon, G.; Schultz, P. G. Structure
of a Copper-Mediated Base Pair in DNA. J. Am. Chem. Soc. 2001, 123,
12364−12367.
(25) Ghosh, S. K.; Ribas, J.; Bharadwaj, P. K. Metal-Organic
Framework Structures of Cu(II) with Pyridine-2,6- Dicarboxylate and
Different Spacers: Identification of a Metal Bound Acyclic Water
Tetramer. CrystEngComm 2004, 6, 250−256.
(26) Das, B.; Baruah, J. B. Assembling of Copper(II) Dipicolinate
Complexes. Polyhedron 2012, 31, 361−367.
(27) Del Pilar Brandi-Blanco, M.; Choquesillo-Lazarte, D.;
́
́
́
Domínguez-Martín, A.; Matilla-Hernandez, A.; Gonzalez-Perez, J.
́
́
M.; Castineiras, A.; Niclos-Gutierrez, J. Molecular Recognition Modes
̃
between Adenine or Adeniniun(1+) Ion and Binary MII(Pdc)
Chelates (M = Co-Zn; Pdc = Pyridine-2,6-Dicarboxylate(2-) Ion).
J. Inorg. Biochem. 2013, 127, 211−219.
́
Bauza, A.; Frontera, A.; Habibi, M.; Ardalani, M.; Shamsipur, M.
(28) Domínguez-Martín, A.; Choquesillo-Lazarte, D.; Dobado, J. A.;
Synthesis, Structure, Solution and DFT Studies of a Pyrazine-Bridged
Binuclear Cu(II) Complex: On the Importance of Noncovalent
Interactions in the Formation of Crystalline Network. J. Mol. Struct.
2015, 1079, 78−86.
́
́
́
Vidal, I.; Lezama, L.; Gonzalez-Perez, J. M.; Castineiras, A.; Niclos-
̃
́
Gutierrez, J. From 7-Azaindole to Adenine: Molecular Recognition
Aspects on Mixed-Ligand Cu(II) Complexes with Deaza-Adenine
Ligands. Dalt. Trans. 2013, 42, 6119−6130.
̇
(49) Vural, H.; Uca̧ r, I. A Mixed Experimental and Theoretical Study
(29) Greco, E.; Aliev, A. E.; Lafitte, V. G. H.; Bala, K.; Duncan, D.;
Pilon, L.; Golding, P.; Hailes, H. C. Cytosine Modules in Quadruple
Hydrogen Bonded Arrays. New J. Chem. 2010, 34, 2634−2642.
on Chelidamate Copper(II) Complex with 4-Methylpyrimidine. J.
Coord. Chem. 2016, 69, 3010−3020.
M
https://dx.doi.org/10.1021/acs.inorgchem.0c01210
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
pubs.acs.org/IC
Article
(50) Grell, J.; Bernstein, J.; Tinhofer, G. Graph-Set Analysis of
Hydrogen-Bond Patterns: Some Mathematical Concepts. Acta
Crystallogr., Sect. B: Struct. Sci. 1999, B55, 1030−1043.
(51) Yenikaya, C.; Bu
̈
yu
̈
kkidan, N.; Sari, M.; Kesl
̈
̧
i, R.; Ilkimen, H.;
Bulbul, M.; Buyukgu
̈
̈
̈
̈
̈
ngor, O. Synthesis, Characterization, and
Biological Evaluation of Cu(II) Complexes with the Proton Transfer
Salt of 2,6-Pyridinedicarboxylic Acid and 2-Amino-4-Methylpyridine.
J. Coord. Chem. 2011, 64, 3353−3365.
̇
(52) Yenikaya, C.; Poyraz, M.; Sarı, M.; Demirci, F.; Ilkimen, H.;
Buyukgungor, O. Synthesis, Characterization and Biological Evalua-
̈ ̈ ̈
̈
tion of a Novel Cu(II) Complex with the Mixed Ligands 2,6-
Pyridinedicarboxylic Acid and 2-Aminopyridine. Polyhedron 2009, 28,
3526−3532.
(53) Lippert, B. Multiplicity of Metal Ion Binding Patterns to
Nucleobases. Coord. Chem. Rev. 2000, 200−202, 487−516.
(54) Kondo, J.; Tada, Y.; Dairaku, T.; Hattori, Y.; Saneyoshi, H.;
Ono, A.; Tanaka, Y. A Metallo-DNA Nanowire with Uninterrupted
One-Dimensional Silver Array. Nat. Chem. 2017, 9, 956−960.
(55) For the sake of comprehension, mcheld numbering in 1Cu
follows that adopted in the mcheld-Cu-nucleobase base pairs.
(56) Hydrogen bonds with D···A distances larger than 3.0 Å have
been omitted. The reader is referred to the CIF file.
(57) The difference in the length of the alkyl chain is merely due to
the access to specific starting materials at the time of the synthesis.
N
https://dx.doi.org/10.1021/acs.inorgchem.0c01210
Inorg. Chem. XXXX, XXX, XXX−XXX