Design, Synthesis, and Binding Affinity Evaluation of Hoechst 33258 Derivatives for the Development of Sequence-Specific DNA-Based Asymmetric Catalysts

Article abstract of DOI:10.1021/acscatal.6b00495

To date, the concept of DNA-based asymmetric catalysis has been successfully applied to various synthetic transformations by way of hybrid catalysts involving either an intercalator or an integrated ligand anchored through supramolecular interactions. We

Full text of DOI:10.1021/acscatal.6b00495

Research Article
pubs.acs.org/acscatalysis
Design, Synthesis, and Binding Affinity Evaluation of Hoechst 33258
Derivatives for the Development of Sequence-Specific DNA-Based
Asymmetric Catalysts
Karen Amirbekyan,† Nicolas Duchemin, Erica Benedetti, Rinah Joseph, Aude Colon,
,‡
‡
†
§
§
§
§
§
#
#
#
Shiraz A. Markarian, Lucas Bethge, Stephan Vonhoff, Sven Klussmann, Janine Cossy,
,§,¶
,†
Jean-Jacques Vasseur, Stellios Arseniyadis,* and Michael Smietana*
†
Institut des Biomolec
́
ules Max Mousseron, UMR 5247 CNRS, Universite
́
de Montpellier, ENSCM Place Eugen
̀
e Bataillon, 34095
Montpellier, France
‡
Department of Physical Chemistry, Yerevan State University, 1 Alex Manoogian, Yerevan 0025, Armenia
§
Laboratoire de Chimie Organique, Institute of Chemistry, Biology and Innovation (CBI) - ESPCI ParisTech/CNRS
(
UMR8231)/PSL* Research University, 10 rue Vauquelin, 75231 CEDEX 05 Paris, France
#
NOXXON Pharma AG, Max-Dohrn-Strasse 8-10, 10589 Berlin, Germany
¶
School of Biological and Chemical Sciences, Queen Mary University of London, Joseph Priestley Building, Mile End Road, London
E1 4NS, United Kingdom
*
S Supporting Information
ABSTRACT: To date, the concept of DNA-based asymmetric catalysis has been successfully applied to various synthetic
transformations by way of hybrid catalysts involving either an intercalator or an integrated ligand anchored through
supramolecular interactions. We report here a new anchoring strategy based on the well-known groove-binder Hoechst 33258.
The interaction between calf thymus DNA (ct-DNA) and poly[d(A-T) ] with a series of Hoechst 33258-derived ligands was
2
studied by UV−vis absorption spectroscopy, thermal melting analysis, fluorescence emission, CD spectroscopy, mass
spectrometry, and molecular docking. The results clearly show that a groove-binding anchoring strategy can be envisioned for
DNA-based asymmetric catalysis, offering additional mechanistic insight on how the intrinsic chirality of DNA can be transferred
to a reaction product. Most importantly, this new anchoring strategy offers interesting compartmentalization possibilities and
provides a new way to reverse the enantioselectivity outcome of a given reaction.
KEYWORDS: DNA-based asymmetric catalysis, minor groove binder, Hoechst 33258, Friedel−Crafts, biohybrid catalysis
INTRODUCTION
chirality from natural DNA to a Cu(II) complex all the way to a
■
Diels−Alder reaction product was reported in 2005 by Roelfes
The use of artificial metalloenzymes, which combine transition-
metal catalysis and biocatalysis, has recently emerged as a
promising concept to evolve biomolecular scaffolds into
3
and Feringa. Since then, the concept has been extended to
many carbon−carbon, carbon−heteroatom, and carbon−
2
,4
1
halogen bond-forming reactions. Motivated by the potential
significance of DNA in asymmetric catalysis, our group
demonstrated the decisive role of the nature of the helix on
performing new catalytic reactions and/or new functions.
While early studies were devoted to the development of
artificial metalloproteins and peptides, DNA-based artificial
metalloenzymes have recently drawn considerable interest
because of their unique features that comprise a chemically
stable chiral double helix associated with many programmable
Received: February 18, 2016
Revised: March 29, 2016
2
secondary structures. The first example of a direct transfer of
©
XXXX American Chemical Society
3096
DOI: 10.1021/acscatal.6b00495
ACS Catal. 2016, 6, 3096−3105

ACS Catalysis
Research Article
the enantioselectivity outcome by studying the influence of L-
and D-DNA interacting with a Cu(II) bipyridine complex in5
Friedel−Crafts alkylations and Michael addition reactions.
More recently, we even extended the concept of DNA-based
asymmetric catalysis to continuous-flow applications by
developing a fully recyclable trivial-to-use cellulose-supported
eventually evaluated for their catalytic activity. Several
techniques including UV−vis absorption spectroscopy, thermal
melting analysis, fluorescence emission, CD spectroscopy, mass
spectrometry, and molecular docking were employed to
monitor their prospective interaction with ct-DNA and
poly[d(A-T) ] and various DNA hairpins. Most importantly,
2
6
DNA-based catalyst. Despite these efforts, the implementation
we wished to address three key questions: (i) How much
binding affinity can be attained? (ii) Will the modifications
sustain the AT sequence-selectivity? (iii) What is the
correlation between DNA affinity and enantioselectivity
outcome? We report here the results of our endeavor.
of DNA-based artificial metalloenzymes toward concurrent
cascade reactions requires a precise and programmable
anchoring of the transition metal into the biomolecular scaffold.
Interestingly, all the anchoring strategies reported so far involve
either intercalation of the coordination complexes into natural
7
RESULTS AND DISCUSSION
double-strand DNA (ds DNA), second coordination sphere
■
8
interactions, or covalent attachment of the ligand directly onto
The first series of bifunctional ligands we wished to prepare
contained a minor groove-binding moiety bound to a metal-
chelating agent through a flexible benzylamine-type linkage. To
access these types of compounds, we therefore needed to
introduce an aldehyde moiety to the Hoechst 33258 scaffold
9
the DNA. While the latter allows for precise positioning of the
metal complex, it requires the synthesis of modified nucleotides
or nucleotide analogues and their incorporation into synthetic
oligonucleotides, which hampers any attempt of compartmen-
talization.
(
Scheme 1). This was achieved by first converting the phenol
Sequence specific targeting of ds-DNA is a well-known
10
Scheme 1. Synthesis of New Hoechst 33258-Derived Ligands
and Representative Structures
strategy to control the regulation of gene expression, site-
11
12
directed mutagenesis, and gene repair. In particular, the
DNA minor groove is the center of action of a large number of
drugs such as the antiviral agents netropsin and distamycin, the
anti-Pneumocystis carinii drug, pentamidine, and DNA stain
Hoechst 33258 (H33258). These molecules interact in AT-rich
13
regions of B form duplex DNA, where they can block the
1
4
15
transcription or the action of DNA topoisomerases. Studies
by X-ray crystallography and NMR analysis carried out with
oligonucleotides of various sequences revealed that four
consecutive AT-base pairs are required for H33258 to bind
16
the minor groove of dsDNA. Quantitative DNase I foot
printing studies actually showed that AATT represents the
17
strongest affinity binding site for H33258, while biophysical
studies performed on poly[d(A-T) ], poly[d(A)·d(T)] and
2
poly[d(G-C) ] revealed the existence of multiple binding
2
13b
modes between the drug and synthetic DNAs.
Given the high affinity of Hoechst 33258 toward DNA, we
became interested in designing structural modifications of
H33258 that would confer interesting asymmetric induction
capabilities in conjunction with high sequence specificity.
Hence, by taking advantage of the free phenolic group of the
molecule, we planned to synthesize new bifunctional DNA
ligands bearing both a minor groove-binding moiety and a
chelating site able to coordinate to a transition metal; both
linked to either a rigid or a flexible spacer (Figure 1). Two
classes of bifunctional substrates bearing either a benzylic or a
propargylic amine were thus devised, synthesized, and
to the corresponding trifluoromethansulfonate by subjecting
Hoechst 33258 to N-phenyl-bis(trifluoromethane)sulfonimide
and diisopropylethylamine. The resulting triflate derivative 1
was then engaged in a Stille coupling with tributyl(vinyl)tin in
the presence of LiCl and a catalytic amount of Pd(PPh ) Cl in
3
2
2
order to introduce a vinyl moiety, which was subsequently
cleaved using osmium tetroxide, NaIO and 2,6-lutidine to
4
afford the desired Hoechst-aldehyde precursor 2. The synthesis
of the amine partner began by first monoprotecting
commercially available ethylene diamine or 1,3-diamino-
propane as the corresponding mono tert-butyl carbamates 5a
and 5b (Scheme 2). These latter compounds were then
subjected to a reductive amination with 2-pyridinecarboxalde-
Figure 1. Sequence-specific catalysts for DNA-based asymmetric
catalysis.
3
097
DOI: 10.1021/acscatal.6b00495
ACS Catal. 2016, 6, 3096−3105

ACS Catalysis
Research Article
Scheme 2. Synthesis of Diamines Bearing a Pendant
Pyridine Ring
Scheme 3. Synthesis of Propargylamines Bearing a Pendant
Pyridine Ring
tional ligands 4a−f in good to excellent yields ranging from
6
1% to 89% (Figure 2).
The affinity of these newly synthesized Hoechst derivatives
with DNA was then investigated using various techniques
including UV−vis absorption spectroscopy, thermal melting
fluorescence, CD spectroscopy, mass spectrometry (see
Supporting Information), and molecular docking studies. The
results show a strong influence of the type of linkage involved
between the Hoechst residue and the metal-binding domain.
UV−vis Absorption Studies. With the objective to study
the interaction between our bifunctional ligands and DNA and
to assess how the structural modifications influenced the DNA
binding affinity, we performed a series of UV−vis absorption
studies and compared the results with the ones obtained with
Hoechst 33258 (see Supporting Information). The association
constant (expressed in molar base pairs) of Hoechst 33258 with
hyde to afford the secondary amines 6a and 6b in quantitative
yield. Finally, a nucleophilic substitution with the appropriate
benzyl chloride followed by the deprotection of the primary
amine afforded the desired amines 8a−e in moderate to
excellent yields ranging from 52% to quantitative. With the two
coupling partners in hand, the last step of the synthesis involved
a reductive amination between aldehyde 2 and the correspond-
ing amines. This was achieved under standard conditions,
affording the desired bifunctional Hoechst 33258-derived
amine-type ligands 3a−e in only five steps (longest linear
sequence) and yields ranging from 43% to 66% (Figure 2).
In an alternative design, the DNA-binding moiety was linked
to the metal-binding domain through a more rigid alkyne type
spacer. In this case, the synthesis involved a Sonogashira
coupling between Hoechst triflate 1 and several diversely
substituted propargyl amines 10a−f. The latter were obtained
following the reaction sequence reported in Scheme 3. Hence,
starting from commercially available propargyl amine, a
reductive amination with 2-pyridine carboxyaldehyde followed
by a nucleophilic substitution with the appropriate benzyl
chloride derivative readily afforded the desired propargylic
amines, which were finally engaged in the key Sonogashira
coupling to afford the desired “Hoechst-alkyne” type bifunc-
ct-DNA and poly[d(A-T) ] are in good agreement with the
2
13b
literature data.
The absorption spectra of the Hoechst
derivatives were recorded in the presence of an increasing
concentration of ct-DNA, thus providing characteristic dose-
dependent curves convenient to describe their interaction
(Table 1). The first obvious result obtained from these
titrations is that in the 300−420 nm region, the Hoechst-
amines exhibit stronger hypochromism (up to 65%) and red
shifts than their alkyne analogues.
This was particularly marked for compound 3a, which
induced an extensive broadening in the 310−425 nm region as
−
1
well as a significant red-shift (916 cm ) (Figure 3A).
Interestingly, this red shift (341−352 nm) was larger than
a
Figure 2. Library of Hoechst 33258-derived ligands. Yields correspond to the final coupling step.
3
098
DOI: 10.1021/acscatal.6b00495
ACS Catal. 2016, 6, 3096−3105

ACS Catalysis
Research Article
Table 1. Spectroscopic Data and Association Constants Obtained for the Probe/ct-DNA Mixtures Studies from UV−vis
Titration (Cacodylate Buffer, pH 7.4, 25 °C)
λmax free
(nm)
λmax bound
(nm)
ΔE
% hypochromism(wavelength,
nm)
εfree (M−1cm−1
at λmax free
)
εbound (M−1cm−1
at λmax bound
)
(cm 1)
−
K /10 , M
a
6
−1
compound
3
3
3
3
3
3
3
4
4
4
4
4
4
4
a
341
345
344
347
340
338
343
355
352
354
350
355
347
349
352
357
352
356
349
346
348
358
358
361
354
360
354
357
916
974
661
728
758
684
419
236
476
548
323
391
570
640
65 (341)
41 (345)
62 (344)
38 (347)
45 (340)
55 (338)
60 (343)
8 (355)
25500
22300
24500
21100
19500
21250
25000
28750
25750
23400
22450
18500
22570
20600
8920
13160
9310
3.24 (±0.3)
2.85 (±0.3)
2.40 (±0.2)
1.90 (±0.2)
1.00 (±0.1)
1.20 (±0.1)
1.60 (±0.1)
0.15 (±0.02)
0.75 (±0.1)
0.68 (±0.1)
0.65 (±0.05)
0.54 (±0.03)
0.35 (±0.03)
0.28 (±0.01)
a
a
b
b
c
d
e
a
b
b
c
d
e
f
a
13080
10730
9560
10000
26450
21370
20590
19310
16280
20540
18950
17 (352)
12 (354)
14 (350)
12 (355)
10 (347)
9(349)
a
a
Data corresponding to binding with ct-DNA in the presence of one equivalent of Cu2+.
Figure 3. UV−vis absorption spectra of compound 3a (A) and 4a (B) in the absence (black) and in the presence of increasing concentration of ct-
DNA. The arrows show the decrease in absorbance at 341 and 355 nm (for 3a and 4a, respectively). Experimental conditions: 4 μM initial
concentration of ligands, [ct-DNA] = 0−32 μM, 100 mM NaCl, 10 mM sodium cacodylate, pH 7.4, 25 °C. Insets: Scatchard plots for the binding of
compounds 3a and 4a with ct DNA obtained from UV−vis titration data.
the one observed with Hoechst 33258 under otherwise
identical conditions (340−345 nm, see Supporting Informa-
tion). In contrast, as exemplified with compound 4a, alkyne
derivatives displayed low hypochromism (as low as 8% for
compound 4a) upon binding to ct-DNA (Figure 3B).
On the basis of these UV−vis titration curves, the binding
constants of all our newly synthesized compounds were
determined using the neighbor exclusion model of McGhee
and von Hippel as described in the Experimental Section. In all
compounds bearing a diaminoethyl linker such as compounds
3c, 3d, and 3e. Meanwhile, it is also interesting to point out the
influence of the naphthalene moiety in 3b, which appears to
enhance the binding affinity in the amine series.
In contrast, all the ligands bearing an alkyne moiety were
characterized by low affinities toward ct-DNA, thus showcasing
that van der Waals interactions are not sufficient to counter-
balance flexibility and electrostatic interactions. While having a
low affinity toward ct-DNA, it could be noted that 4b appears
to be the most effective binder among the “Hoechst-alkyne”
derivatives.
cases, the Scatchard plots (r/C vs r) revealed an obvious
f
downward curvature of the plot which arises from the binding
of a single molecule. This excluded the binding of other ligands
nearby along the DNA polymer, a phenomenon also referred to
Another important factor was the degree of methoxy-
substitution on the aromatic ring, which also appeared to
enhance the DNA-binding affinity. As shown in Table 1,
compound 3e, which contains three methoxy groups attached
to the aromatic ring, exhibited a higher affinity compared to
compounds 3c and 3d, which contain one and two,
respectively. This can be the result of the electron-donating
properties of the methoxy substituents and perhaps favorable
steric factors promoting the binding. This trend was also
observed with compound 4c bearing a pentamethyl-substituted
aromatic ring. In addition, these results show that para-
methoxy-substituted derivatives have a slightly higher affinity
than the corresponding meta-substituted compounds (4e vs 4f).
The binding of 3a, 3b, and 4b was also studied in the
presence of one equivalent Cu(NO ) for consistency purposes
18
as neighbor exclusion. As can be seen from Table 1, the
highest affinity toward ct-DNA was obtained with compound
3
a. The Scatchard plot analysis (Figure 3A, inset), also allowed
6
us to determine the association constant K [(3.24 ± 0.3) × 10
M ] and the n parameter (6.8 ± 0.2), while standard Gibbs
energy for the binding of compound 3a with ct-DNA was found
a
−
1
−1
equal to −36.4 kJ mol .
These results highlight the importance of the flexibility
brought by the diaminopropyl linker and most probably a
favorable interaction of the secondary and tertiary amine
moieties with DNA. This hypothesis is corroborated by the
high binding affinity of the other “Hoechst-amine” type ligands,
such as compound 3b, and, to a lesser extent, all the
3
2
3
099
DOI: 10.1021/acscatal.6b00495
ACS Catal. 2016, 6, 3096−3105

ACS Catalysis
Research Article
and in order to evaluate the influence of the copper ions on the
binding of the ligands to ct-DNA. As can be seen from Table 1,
the presence of Cu2+ ions reduces the binding constant of the
ligands by 12, 21, and 9% for 3a, 3b, and 4b, respectively. The
effect of copper ions on ligand−DNA binding was reported
earlier, and this decrease can be explained by the electrostatic
chromism effect in absorbance with even some hyperchromism
effect detected for several compounds such as 4a, 4e, and 4f. As
a consequence, we limited the K determination to the amine-
a
type derivatives.
The high binding constants toward poly[d(A-T) ] displayed
2
by compounds 3a and 3e confirmed their high affinity toward
AT base pairs and the steric and electronic effects that promote
the binding. Similarly, comparing compounds 3c, 3d, and 3e, all
bearing one or multiple methoxy groups on the benzylic ring,
indicate that the methoxy substituents do not affect the AT-
base recognition of the ligands.
2+
19
interactions and coordination of the ligands by Cu .
Binding Study with Poly[d(A-T) ]. The use of UV−vis
2
absorbance spectroscopy allows us to unveil the binding
properties of a given ligand toward specific base pairs in a DNA
1
3b,16f,20
sequence.
In this context, we performed a series of
UV−vis absorbance spectroscopy analyses of all our Hoechst-
derivatives in the presence of the double strand alternating
copolymer poly(deoxyadenylic-thymidylic) acid sodium salt,
Another important outcome of this study was the fact that
these compounds were characterized by a low binding constant
with poly[d(A-T) ] compared to ct-DNA, which we believe is
2
poly[d(A-T) ], in order to ascertain a possible sequence
related to their high GC recognition ability. Hence, the absence
of GC base pairs causes a significant decrease in binding
constants (and binding site size). This GC-specificity will be
later confirmed by molecular docking studies.
Fluorescence Measurements. Fluorescence emission
measurements were performed to describe the sequence-
specificity of compound 3a, which was found to be the best
2
specificity. As a general trend, the binding of these derivatives
to poly[d(A-T) ] was characterized by a hypochromic effect
2
and a red shift similar to the ones observed with ct-DNA, albeit
to a lesser extent (up to 35%). The high affinity of compound
3
a with poly[d(A-T) ] was also confirmed (Figure 4); however,
2
binder to ct-DNA and poly[d(A-T) ] following the UV−vis
2
experiments. For this purpose, a 1 μM solution of ligand 3a was
titrated individually with five different 28-mer DNA hairpins
with sequences having the following structure: 5′-d-
(
CGCGXCGCGTTTTCGCGXCGCG)-3′, where X repre-
sents the part of the hairpin which varies (X = ATAT,
TAAT, TATA, TTAA, and AATT). On the basis of the
fluorescence measurements, we were able to determine the
strongest and the weakest binding sites for 3a. Figure 5 presents
the fluorescence spectra of compound 3a titrated with DNA
hairpins with X = AATT (strongest binding site) and X =
TATA (weakest binding site). Fluorescence titration spectra for
the three other sequences are given in the Supporting
Information. Fluorescence titration results for all five sequences
were analyzed by plotting the relative fluorescence increase (F
Figure 4. UV−vis absorption spectra of compound 3a without (black
curve) and in the presence of increasing concentration of poly[d(A-
T) ]. Inset: Scatchard plots for the binding of compounds 3a with
2
poly[d(A-T) ] obtained from UV−vis titration data.
2
−
F )/F ) vs [ligand:DNA] ratio. The highest increase in
0 0
fluorescence was observed for AATT, whereas the lowest was
observed for TATA (Figure 6). In all cases, however, the
saturation limit was reached at a 1:1 ligand/DNA ratio, which
means that the ligand binds to the DNA in a monomeric form.
In addition, all the fluorescence experiments revealed that the
affinity of 3a increased depending on the base sequences
5
−1
its binding constant with poly[d(A-T) ] (8.5 ± 0.2) × 10 M
2
was lower than the one observed with ct-DNA. In addition,
compound 3a was shown to have an affinity 13.5 times lower
than the unmodified Hoechst toward poly[d(A-T) ] (Table 2).
Interestingly, alkyne-type derivatives expressed a very low
2
affinity toward poly[d(A-T) ] showcased by a slight hypo-
2
(
TATA < TTAA < TAAT < ATAT < AATT), which was in
agreement with the sequence specificity exhibited by Hoechst
Table 2. Spectroscopic Data and Association Constants
16e,17b
3
3258.
Thermal Melting Study. Thermal melting studies were
undertaken to describe the relative binding affinity and the
Obtained for the Probe/Poly[d(A-T) ] Mixture Study from
2
UV−vis Titration (Cacodylate Buffer, pH 7.4, 25 °C)
21
5
−1
compound
% hypochromism (wavelength/nm)
K /10 , M
sequence selectivity of the ligands toward DNA. Indeed,
DNA denaturation is usually associated with a hyperchromic
effect in the absorbance, which is temperature-dependent. All
compounds that bind to DNA give rise to a stabilization of the
DNA visualized by an increase of the melting temperature
a
3a
3b
3c
3d
3e
4a
4b
4c
4d
4e
4f
35 (341)
23 (344)
25 (340)
26 (338)
30 (343)
8.5(±0.3)
5.2(±0.2)
5.5(±0.1)
7.2(±0.1)
7.8(±0.1)
n.d.
denoted T . Conversely, a decrease in Tm is indicative of a
m
a
distortion induced by the DNA binder. On the basis of
fluorescence results which revealed that these derivatives have a
higher affinity toward the AATT sequence, we performed a
melting analysis with the 20-mer hairpin 5′-d-
−4 (355)
7 (352)
1 (350)
2 (355)
n.d.
n.d.
n.d.
a
(
CGAATTCGTTTTCGAATTCG)-3′. Melting measurements
−7 (347)
−5 (349)
n.d.
a
were performed by monitoring the absorbance of the hairpins
n.d.
solutions at 260 nm in the absence (the T value of the free
Hoechst 33258
45 (340)
114.8 (±5)
m
sequence is 67.1 °C) and the presence of one equiv of
compounds 3a−d (benzyl amine-type derivatives), compounds
a
Negative values refer to hyperchromism.
3
100
DOI: 10.1021/acscatal.6b00495
ACS Catal. 2016, 6, 3096−3105

ACS Catalysis
Research Article
Figure 5. Fluorescence emission spectra for 3a titrated with AATT (A) and with TATA (B): 1 μM solution of 3a in 10 mM sodium cacodylate-100
mM NaCl, pH 7.4 buffer at 25 °C, excitation wavelength was set at 341 nm, which is λmax for the free ligand. The changes in ligand:DNA ratio for
both cases fall in the same region (1:0−1:1.04). Under these concentration conditions, 3a is fully bound to the DNA at the highest concentration
shown in the figures.
Figure 6. Relative fluorescence intensity increase of the compound 3a as a function of [ligand]/[DNA] ratio for five different DNA sequences.
4
b and 4e (propargyl alkyne type derivatives), and Hoechst
conformational mobility of 3a and 3b and opened up the space
for the binding of the pyridine and the methoxybenzyl moieties
linked to the Hoechst residue. This linkage most probably
promotes the H-bonding between the ligands and the DNA
bases. Meanwhile, the results substantiate the role played by the
methoxy substituents on the benzyl ring.
3
3258 as a reference. Melting temperatures were determined
from the first derivatives of the melting curves and summarized
in Table 3. As a general trend, compounds 3a and 3b showed
Table 3. Thermal Melting Studies of the Designed Hoechst
Derivatives with 20-mer DNA Hairpin
Circular Dichroism Measurements. In order to character-
ize the type of interaction between our ligands and DNA, we
performed a set of circular dichroism spectroscopy experi-
a
5′‑d(CGAATTCGTTTTCGAATTCG)-3′
compound
Tm (°C)
ΔTm (°C)
22
ments. Indeed, compared to intercalative interactions, minor
groove binders induce larger CD signals. Moreover, neither ct-
DNA nor the free compounds exhibit CD signals above 300
nm. Monitoring induced CD, binding mode, and saturation
limit for selected compound 3a were performed by CD titration
experiments as a function of compound concentration.
Addition of 3a to ct-DNA resulted in significant positive
induced CD signals on complex formation with DNA at the
wavelength between 300 and 400 nm. The CD spectra of ct-
DNA and its complexes with 3a at dye/DNA base pairs mixing
ratios varying from 0 to 0.15 are shown in Figure 7. Notably,
ICD peaks positions of 3a matched the absorption peak
positions of bound probes (Table 1), thus indicating that these
ICD peaks correspond to the same chromophore, which is
responsible for the red-shifted absorption spectra shown Figure
3
3
3
3
3
4
4
a
b
c
74.3
74.1
70.2
70.6
70.8
68.9
68.3
74.5
7.2
7.0
3.1
3.5
3.7
1.8
1.2
7.4
d
e
b
e
Hoechst 33258
a
ΔT = T (complex) − T (free DNA). The listed values are for a
:1 [ligand]/[DNA] ratio and an average of two independent
experiments with a reproducibility of ±0.5 °C.
m
m
m
1
the highest DNA-stabilizing effect among all the Hoechst
derivatives (ΔT = 7.2 and 7.0 °C, respectively) almost
m
identical to the stabilization induced by Hoechst 33258 and in
good agreements with their binding constant values obtained
by UV−vis titration experiments. Compounds with an
ethylamine type linker (3e, 3d, and 3c) showed a moderate
increase in DNA stability (ΔTm = 3.7, 3.5, and 3.1 °C,
respectively). Conversely, compound 4e did not show any
significant enhancement. These results confirm the importance
of the aminopropyl linker compared to the aminoethyl one as
well as a necessary flexibility for efficient groove binding.
Introducing the aminopropyl-type linker clearly increased the
3
.
Changes of ICD intensities of 3a at 350 nm are illustrated in
Figure 7, inset. The CD signals of ct-DNA around 245 and 280
nm are decreasing in intensity with a saturation limit for
compound 3a reached at ligand/DNA ratio of 0.15 in good
agreement with UV−vis absorption titration data. This is
indicative of a binding density value n = 6.8 corresponding to a
dye/DNA ratio equal to 1/6.8 = 0.147. Hence, CD spectros-
copy measurements revealed the minor groove binding
3
101
DOI: 10.1021/acscatal.6b00495
ACS Catal. 2016, 6, 3096−3105

ACS Catalysis
Research Article
shown in Figure 10, and the length of the hydrogen bonds
determined with PyMOL software are given in Table 4.
DNA-Based Asymmetric Catalysis. To demonstrate the
efficacy of our bifunctional Hoechst-derived ligands, amine-type
3a and 3b and alkyne-type 4b derivatives were selected and
engaged in the copper-catalyzed Friedel−Crafts alkylation of
α,β-unsaturated 2-acyl imidazole 11 with 5-methoxyindole 12.
The experiments were conducted at 5 °C using a stock solution
of a Cu(II)-ligand complex (in MOPS 20 mM, pH 6.5) and
DNA (2 mM in MOPS 20 mM, pH 6.5). The results are
reported in Table 5. 4,4′-Dimethyl-2,2′-bipyridine (bmbpy)
was also used as a control.
Figure 7. Circular dichroism spectra of representative compound 3a
with ct-DNA. Mixing ratios of ligand to ct-DNA base pairs are 0, 0.03,
0
.06, 0.09, 0.12, and 0.15. Color arrows indicate positive (green) and
As a general trend, the ligand concentration had a non-
negligible impact on the enantioselectivity of the reaction; the
best ee values were obtained at a 0.1 mM concentration (Table
negative (red) induced changes. The experiments were conducted in
cacodylate buffer at 25 °C. Inset picture is the change of ICD
intensities of 3a at 350 nm during titration with ct-DNA. Ligand/DNA
base pairs ratio was varied from 0 to 0.15.
5
, entries 1 and 2). Interestingly, this phenomenon was not
observed when using an intercalating ligand such as 4,4′-
dimethyl-2,2′-bipyridine. In the presence of ct-DNA, we were
pleased to observe a significant induction, in particular with 3b,
albeit not in the range of the one obtained when using dmbpy.
In contrast, with low binding alkyne-type ligands such as 4b,
the ee value decreased sharply to 7%, thus confirming that the
affinity of the ligand had a significant impact on the
enantioselectivity outcome of the reaction (Table 5, entries
character of the representative compound 3a and confirmed the
saturation limit obtained from UV−vis measurements.
Molecular Modeling Studies. To better understand the
binding of 3a to DNA and to reveal the most probable
conformations in which the ligand binds to DNA, we
performed molecular docking studies. The initial DNA duplex
s e l e c t e d f o r d o c k i n g w a s d o d e c a m e r 5 ′ - d -
1
−4). Interestingly, in the presence of the AATT hairpin, the
enantioselectivity decreased from +80% to +31% with dmbpy.
Although the exact reason for this sharp decrease is still unclear,
we were pleased to observe an unexpected inversion of
selectivity with the groove binder ligands 3a (from +29% to
(
CGCGAATTCGCG)-3′ (PDB ID: 1BNA). Docking results
are presented in Figure 8. A detailed view of the diamine part of
compound 3a in docked conformation is shown in Figure 9.
For compound 3a, the docking program identified nine
possible conformations with values of affinity comprised
−
16%) and 3b (from +47% to −26%) (Table 5, entries 5−8)
−1
which was also observed with the TATA hairpin (Table 5,
entries 9−12).
between −54.8 and −47.88 kJ mol . Docking results revealed
the possibility of hydrogen-bonding interactions between the
inner facing benzimidazole N−H group and AT bases of DNA
in the floor of the minor groove of DNA, which is typical for
benzimidazole-containing drugs of the Hoechst 33258 family.
Additionally, docking results showed the possibility of H-bond
forming between the nitrogen atom of the methylpyridine and
the G-NH moiety, which support the G-recognition ability of
compound 3a obtained from UV−vis absorption measure-
ments. The H-bond acceptor property of the pyridine nitrogen
These results clearly validate the groove-binder anchoring
strategy in the context of DNA-based asymmetric catalysis.
Even though it is evident that a simple affinity/enantioselec-
tivity correlation cannot be categorical, we have confirmed that
derivatives such as 3 were the strongest binding ligands and
that AATT was the best binding site. A precise tuning of the
binder/DNA complex (optimal binding position and orienta-
23
2
tion of the ligand, flexibility of the linker, (A/T) sites that
n
24
atom is a well-established phenomenon. Hydrogen bonds are
modulate the minor groove width and most probably flanking
Figure 8. Molecular models for 3a docked into the d(CGCGAATTCGCG) sequence. (A) Conformation of compound 3a (obtained from docking
2
results) in which it has a higher affinity toward the DNA dodecamer. View from the minor (B) and the major groove (C) of compound 3a docked in
the minor groove of DNA. These images are the ones of the lowest-energy conformation and have been drawn with the PyMOL software. The
transparency of the molecular surface of DNA dodecamer is shown with 60% transparency. Note that the surface of 3a and the surface of minor
groove wall of DNA are in close contact.
3
102
DOI: 10.1021/acscatal.6b00495
ACS Catal. 2016, 6, 3096−3105

ACS Catalysis
Research Article
Figure 9. Structural properties of compound 3a in docked conformation: (A,B,C) Angles between bonds of diamine moiety of compound 3a. (D)
Dihedral angles N3CICIICIII and N5CIVCVCVI. The angles are shown as dashed arc lines. Three different pictures (A−C) of the same moiety
were drawn for clarity. Angles between bonds were determined from docked conformation of 3a using the PyMOL software.
Table 5. Friedel−Crafts Alkylation Using Hoechst-Derived
Ligands 3a, 3b, and 4b
a
a
entry
ligand
DNA
conversion (%)
ee (%)
1
dmbpy
3a
ct-DNA
ct-DNA
ct-DNA
ct-DNA
ODN-1
ODN-1
ODN-1
ODN-1
ODN-2
ODN-2
ODN-2
ODN-2
>99
>99
>99
94
(+) 80
(+) 29
(+) 47
(+) 7
2
3
3b
4
4b
5
dmbpy
3a
>99
>99
78
(+) 31
(−) 16
(−) 26
(+) 4
Figure 10. Detailed view from the minor groove of the hydrogen-bond
interactions between compound 3a (shown as dashed arrows) and
DNA bases. Data are based on zoomed-in (30×) view of docking
results as presented in Figure 8. For clarity purposes, only the bases
which participate through H-bonding interaction with the ligand are
shown. Shown hydrogen bonds are the following: between nitrogen
atoms of inner-facing benzimidazole N−H groups: I. N4 hydrogen
atom of 3a and N3 atom of adenine 18, II. N4 hydrogen atom of 3a
with O2 atom of T7 (I and II are equally probable H bonds), III. N6
hydrogen atom of 3a and O2 atom of T8, IV. H-bond between N9 of
6
7
3b
8
4b
>99
>99
72
9
dmbpy
3a
(+) 43
(−) 19
(−) 23
(−) 5
10
11
3b
>99
>99
1
2
4b
a
Determined by supercritical fluid chromatography (SFC) analysis.
ODN-1 = 5′-d(CGAATTCGTTTTCGAATTCG)-3′, [ODN-2] =
′-d(CGTATACG TTTTCGTATACG)-3′].
[
5
3
a and NH2 group of G16.
Table 4. Ligand−DNA Hydrogen-Bond Distances in the 3a−
d(CGCGAATTCGCG) Complex
fluorescence and CD experiments showed that compounds
with an alkyne type linkage showed low affinity toward ct-DNA
2
3
a atom
DNA atom
O2 (T7)
distance (Å)
and poly[d(A-T) ] compared to their corresponding amine
2
H of N4
H of N4
H of N6
N9
3.1
2.9
3.2
3.0
analogues. This was most probably preconditioned by the low
flexibility of the triple bond, which impedes favorable
interactions between the pyridine and the methoxybenzyl
groups with the DNA bases. Also, it is probably the absence of
one charged nitrogen atom in the alkyne-type derivatives
(compared with the amine-type derivatives), which reduces the
strength of the electrostatic interactions between the ligand and
the DNA phosphate backbone and thus weakens the affinity of
the ligand toward DNA. Meanwhile, designed derivatives
bearing a diaminoalkyl linker and especially the diaminopropyl
tether appeared to have a high affinity toward both ct-DNA and
N3 (A18)
O2 (T8)
H of NH2 (G16)
bases) is now essential to attain higher levels of enantiose-
lectivity.
CONCLUSION
■
The design and synthesis of compounds exhibiting a high
affinity toward DNA sequences has been successfully achieved
on the basis of structural modifications brought to the well-
known minor groove binder Hoechst 33258. Analysis based on
UV−vis titration and thermal denaturation, as well as
poly[d(A-T) ], which confirms the potential of our minor-
2
groove-binder anchoring strategy. The results presented in this
work clearly show that a rational design of Hoechst derivatives
3
103
DOI: 10.1021/acscatal.6b00495
ACS Catal. 2016, 6, 3096−3105

ACS Catalysis
Research Article
capable of binding to mixed DNA sequences is possible and
that a groove binder anchoring strategy for DNA-based
asymmetric catalysis is valuable to achieve sequence-selective
reactions. The inversion of enantioselectivity observed with
AT-rich sequences extends the tools available until now to
control the enantioselectivity outcome of a DNA-catalyzed
reaction. The success of Hoechst 33258-amine type
derivatives having mixed DNA sequence recognition capa-
bilities such as 3a and 3b offers exciting prospects for DNA-
based asymmetric catalysis.
model spectrophotometer in 1 cm quartz cell at 25 °C. Prior to
all measurements, a buffer spectra was recorded for blank
correction. A solution of calf thymus DNA (80 μM) in buffer
(
10 mM sodium cacodylate, 100 mM NaCl, pH 7.4) was added
to the cell prior to the experiment followed by the solution of
a (0.48 mM in MeOH containing 5% (v/v) of 0.1 mM HCl)
3
2
5
in 5 μL increments. The resulting solution was incubated for 10
min to achieve the equilibrium binding for the DNA complex.
Each spectrum presents the average of three scans from 220 to
4
1
20 nm with a scan speed 60 nm·min− and a response time of
s.
1
EXPERIMENTAL SECTION
Ligands and DNA Oligonucleotides. Hoechst 33258, ct-
■
Fluorescence Emission Spectroscopy. Fluorescence
spectra were recorded on a Jasco 815 model spectrophotometer
using 1 cm path length quartz cells. Excitation and emission
bandwidths were fixed at 2 and 10 nm, respectively. A solution
of ligand 3a at a concentration of 1 μM was prepared in a 10
mM sodium cacodylate/100 mM NaCl buffer (pH 7.4) and
hairpin DNA aliquots (5 μL increments) were added from a
concentrated stock (52 μM in a cacodylate buffer). The spectra
were collected after allowing an equilibration time of 10 min.
Compound 3a was excited at 341 nm, which is λmax for the free
ligand as identified by UV−vis spectroscopy. Emission spectra
were monitored from 350 to 600 nm. All the fluorescence
titrations were performed at 25 °C.
DNA, and poly[d(A-T) ] were purchased from Sigma-Aldrich
2
(
U.S.A.). For the preparation of the Hoechst derivatives
solutions, all the compounds were carefully weighed and
dissolved in a methanol solution containing 5% (v/v) of 0.1
mM HCl. Concentrations of Hoechst 33258, ct-DNA,
poly[d(A-T) ], and DNA hairpins were determined spectro-
2
scopically using the following extinction coefficients: ε
4
=
338
2 000 M− cm for Hoechst 33258, ε260 = 13 200 M cm
1
−1
−1
−1
for ct-DNA concentration expressed as base pairs, ε = 13 600
2
62
−
1
−1
M
cm for poly[d(A-T) ] concentration as base pairs. For
2
DNA hairpins, the extinction coefficients provided by the
supplier were used. All the solutions used for spectroscopic
titrations were prepared in sodium cacodylate buffer containing
Molecular Docking Studies. The crystal structure of the
synthetic DNA dodecamer 5′-d(CGCGAATTCGCG)-3′ has
been refined to a residual error of R = 17.8% at 1.9 Å resolution
1
0 mM sodium cacodylate and 100 mM NaCl at pH 7.4. Milli-
Q water was used for the preparation of all the solutions. The
binding studies were carried out at 25 °C.
(
two-sigma data). The molecule forms slightly more than one
UV-Spectrophotometry. UV−vis scanning and melting
measurements were performed using Varian Cary 300 Bio
spectrophotometer equipped with Peltier temperature control
system using 1 cm path length quartz cells. A solution of probe
27
complete turn of a right-handed double-stranded B helix.
AutoDock Vina was used for docking experiments. PDB files
2
8
for the ligand structures were created using Open Babel
29
(4 μM, 1 mL, cacodylate buffer) was titrated by adding 5 μL
graphical user interface 2.3.1. PDBQT molecular structure
increments of nucleic acid (1 mM) while monitoring the
absorbance in the 200−420 nm region. The total volume of
nucleic acid during the titration did not exceed 10% of the
initial volume, and spectra were corrected for small changes in
volume during the titration. The absorption data for all
derivatives were analyzed according to neighbor exclusion
model of McGhee and von Hippel using eq 1. The binding-
constant values were estimated from nonlinear regression
analysis. All data were treated using OriginPro8 software.
files necessary for AutoDock Vina were created using
30
AutoDock Tools graphical user interface. Water molecules
were deleted, and hydrogen atoms were added before creating
PDBQT files. Kollman all-atom charges for DNA and
̈
Gasteiger−Hukel charges for the ligand were computed.
PyMOL software was used to visualize the docking results.
All modeling studies were carried out on a windows
workstation. The following parameters were used during
docking with AutoDock Vina: search space center x = 10.9,
center y = 22.0, center z = 8.9 (these coordinates allow for the
grid box center to be in the minor groove), search space sizes:
size x = 70, size y = 65, size z = 80, spacing 1 Å. Both the ligand
and the bound DNA were permitted torsional flexibility in the
docking process.
26
n−1
r/C = K [1 − nr(1 − nr)]/[1 − (n − 1)r]
(1)
f
a
In eq 1, C is the concentration of the free probe, n is the
f
binding density, and r is the ratio of concentration of bound
probe (C ) to that of DNA. Based on absorbance titration, the
b
values of C were calculated using eq 2, where A is the
f
obs
DNA-Based Asymmetric Catalysis Experiments. To a 2
mM DNA solution in a 20 mM MOPS buffer (600 μL, pH 6.5)
was added the stock solution of [Cu(Hoechst-derived ligand)
absorbance at the peak position of the ligand-DNA mixture, Ct
is the total concentration of the probe, ε and ε are the
f
b
extinction coefficients of the free and bound probes,
respectively. The value of εb was estimated from a half
reciprocal plot of Aobs versus 1/[DNA]. The percent
hypochromism was calculated as equal to (1 − ε /ε )·100%.
(NO ) ]. The resulting DNA solution was cooled to 5 °C. To
3 2
the mixture was then added a 0.5 M solution of enone in
MeCN (1.2 μL), followed by the 5-methoxyindole (12). The
reaction was mixed by inversion at 5 °C in a cold room. After
b
f
Cf = (Aobs − ε C )/(ε − ε )
(2)
b
t
f
b
1
−3 days, the mixture was warmed to room temperature and
extracted with Et O (3 × 2 mL). The combined organic layers
were washed with brine (2 × 5 mL), dried over Mg SO ,
The free energy of binding (ΔG°) was calculated using eq 3.
ΔG° = −RTln Ka
2
2
4
(3)
gravity filtered, and concentrated under reduced pressure to
give the crude product, which was subjected to supercritical
fluid chromatography (SFC) analysis.
Circular Dichroism Spectroscopy (CD). Circular dichro-
ism spectroscopy experiments were performed on a Jasco 815
3
104
DOI: 10.1021/acscatal.6b00495
ACS Catal. 2016, 6, 3096−3105

ACS Catalysis
ASSOCIATED CONTENT
Research Article
(
7
10) (a) Maher, L. J.; Dervan, P. B.; Wold, B. Biochemistry 1992, 31,
■
* Supporting Information
0−81. (b) Giovannangeli, C.; Perrouault, L.; Escude, C.; Thuong, N.;
S
Helene, C. Biochemistry 1996, 35, 10539−10548. (c) Giovannangeli,
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acscatal.6b00495.
C.; Helene, C. Antisense Nucleic Acid Drug Dev. 1997, 7, 413−421.
(
11) Wang, G.; Seidman, M. M.; Glazer, P. M. Science 1996, 271,
8
(
02−805.
12) Broitman, S.; Amosova, O.; Dolinnaya, N. G.; Fresco, J. R. J.
Biol. Chem. 1999, 274, 21763−21768.
13) (a) Zimmer, C.; Wahnert, U. Prog. Biophys. Mol. Biol. 1986, 47,
Experimental procedures as well as spectroscopic and
analytical data (PDF)
(
AUTHOR INFORMATION
Corresponding Authors
E-mail for S.A.: s.arseniyadis@qmul.ac.uk.
E-mail for M.S.: msmietana@um2.fr.
31−112. (b) Loontiens, F. G.; Regenfuss, P.; Zechel, A.; Dumortier,
L.; Clegg, R. M. Biochemistry 1990, 29, 9029−9039. (c) Dervan, P. B.
Science 1986, 232, 464−471.
■
*
(
(
14) Henderson, D.; Hurley, L. H. Nat. Med. 1995, 1, 525−527.
15) (a) Chen, A. Y.; Yu, C.; Gatto, B.; Liu, L. F. Proc. Natl. Acad. Sci.
*
Notes
U. S. A. 1993, 90, 8131−8135. (b) Beerman, T. A.; McHugh, M. M.;
Sigmund, R.; Lown, J. W.; Rao, K. E.; Bathini, Y. Biochim. Biophys.
Acta, Gene Struct. Expression 1992, 1131, 53−61.
The authors declare no competing financial interest.
(16) (a) Teng, M.; Usman, N.; Frederick, C. A.; Wang, A. H. J.
ACKNOWLEDGMENTS
We thank the Minister
■
Nucleic Acids Res. 1988, 16, 2671−2690. (b) Squire, C. J.; Baker, L. J.;
Clark, G. R.; Martin, R. F.; White, J. Nucleic Acids Res. 2000, 28, 1252−
1258. (c) Pjura, P. E.; Grzeskowiak, K.; Dickerson, R. E. J. Mol. Biol.
1987, 197, 257−271. (d) Fede, A.; Billeter, M.; Leupin, W.; Wuthrich,
K. Structure 1993, 1, 177−186. (e) Breusegem, S. Y.; Clegg, R. M.;
Loontiens, F. G. J. Mol. Biol. 2002, 315, 1049−1061. (f) Breusegem, S.
Y.; Sadat-Ebrahimi, S. E.; Douglas, K. T.; Clegg, R. M.; Loontiens, F.
G. J. Mol. Biol. 2001, 308, 649−663.
̀
e de l’Enseignement Super
́
ieur et de la
Recherche, the Agence Nationale de la Recherche (D-CYSIV
Project; ANR-2015-CE29-0021-01) and the Erasmus Mundus
Backis Program for their support.
REFERENCES
1) (a) Du
■
(
1
̈
rrenberger, M.; Ward, T. R. Curr. Opin. Chem. Biol. 2014,
(
17) (a) Harshman, K. D.; Dervan, P. B. Nucleic Acids Res. 1985, 13,
825−4835. (b) Abu-daya, A.; Brown, P. M.; Fox, K. R. Nucleic Acids
Res. 1995, 23, 3385−3392.
18) Dedon, P. C. Curr. Protoc. Nucleic Acid Chem. 2000, Unit 8.2.1−
.2.8.
19) Barone, G.; Terenzi, A.; Lauria, A.; Almerico, A. M.; Leal, J. M.;
Busto, N.; Garcia, B. Coord. Chem. Rev. 2013, 257, 2848−2862.
20) Bailly, C.; Colson, P.; Henichart, J. P.; Houssier, C. Nucleic Acids
Res. 1993, 21, 3705−3709.
21) (a) Wilson, W. D.; Tanious, F. A.; Fernandez-Saiz, M.; Rigl, C.
T. Methods Mol. Biol. 1997, 90, 219−240. (b) Akashi, S. Med. Res. Rev.
006, 26, 339−368. (c) Rettig, M.; Kamal, A.; Ramu, R.; Mikolajczak,
J.; Weisz, K. Bioorg. Med. Chem. 2009, 17, 919−928.
22) (a) Norden, B.; Kubista, M.; Kurucsev, T. Q. Q. Rev. Biophys.
992, 25, 51−170. (b) Karlsson, H. J.; Eriksson, M.; Perzon, E.;
Akerman, B.; Lincoln, P.; Westman, G. Nucleic Acids Res. 2003, 31,
227−6234.
23) Clark, G.; Boykin, D.; Czarny, A.; Neidle, S. Nucleic Acids Res.
997, 25, 1510−1515.
24) (a) Li, T.; Zhou, P.; Mattei, A. CrystEngComm 2011, 13, 6356−
360. (b) Du, Y.; Du, X.; George, S. J. Phys. Chem. C 2007, 111, 219−
26.
25) Rioz-Martínez, A.; Roelfes, G. Curr. Opin. Chem. Biol. 2015, 25,
0−87.
26) McGhee, J. D.; von Hippel, P. H. V. J. Mol. Biol. 1974, 86, 469−
89.
27) Drew, H. R.; Wing, R. M.; Takano, T.; Broka, C.; Tanaka, S.;
Itakura, K.; Dickerson, R. E. Proc. Natl. Acad. Sci. U. S. A. 1981, 78,
9, 99−106. (b) Hoarau, M.; Hureau, C.; Gras, E.; Faller, P. Coord.
4
Chem. Rev. 2016, 308, 445−459.
(
(
3
2) Bos, J.; Roelfes, G. Curr. Opin. Chem. Biol. 2014, 19, 135−143.
3) Roelfes, G.; Feringa, B. L. Angew. Chem., Int. Ed. 2005, 44, 3230−
(
8
(
232.
(
4) (a) Coquiere, D.; Feringa, B. L.; Roelfes, G. Angew. Chem., Int. Ed.
2
007, 46, 9308−9311. (b) Dijk, E. W.; Boersma, A. J.; Feringa, B. L.;
(
́
Roelfes, G. Org. Biomol. Chem. 2010, 8, 3868−3873. (c) Dijk, E. W.;
Feringa, B. L.; Roelfes, G. Tetrahedron: Asymmetry 2008, 19, 2374−
(
2
377. (d) Boersma, A. J.; Coquiere, D.; Geerdink, D.; Rosati, F.;
Feringa, B. L.; Roelfes, G. Nat. Chem. 2010, 2, 991−995. (e) Boersma,
2
A. J.; Feringa, B. L.; Roelfes, G. Angew. Chem., Int. Ed. 2009, 48, 3346−
3
348. (f) Boersma, A. J.; Klijn, J. E.; Feringa, B. L.; Roelfes, G. J. Am.
(
1
Chem. Soc. 2008, 130, 11783−11790. (g) Oelerich, J.; Roelfes, G.
Chem. Sci. 2013, 4, 2013−2017. (h) Shibata, N.; Yasui, H.; Nakamura,
S.; Toru, T. Synlett 2007, 2007, 1153−1157. (i) Boersma, A. J.;
Megens, R. P.; Feringa, B. L.; Roelfes, G. Chem. Soc. Rev. 2010, 39,
6
(
1
(
6
2
(
8
2
2
083−2092. (j) Garcia-Fernandez, A.; Roelfes, G. Met. Ions Life Sci.
012, 10, 249−268. (k) Park, S.; Ikehata, K.; Watabe, R.; Hidaka, Y.;
Rajendran, A.; Sugiyama, H. Chem. Commun. 2012, 48, 10398−10400.
(
6
l) Li, Y.; Wang, C.; Jia, G.; Lu, S.; Li, C. Tetrahedron 2013, 69, 6585−
590.
5) Wang, J.; Benedetti, E.; Bethge, L.; Vonhoff, S.; Klussmann, S.;
(
Vasseur, J.-J.; Cossy, J.; Smietana, M.; Arseniyadis, S. Angew. Chem., Int.
(
4
(
Ed. 2013, 52, 11546−11549.
(
6) Benedetti, E.; Duchemin, N.; Bethge, L.; Vonhoff, S.; Klussmann,
S.; Vasseur, J.-J.; Cossy, J.; Smietana, M.; Arseniyadis, S. Chem.
Commun. 2015, 51, 6076−6079.
2
(
(
179−2183.
(
7) (a) Petrova, G. P.; Ke, Z.; Park, S.; Sugiyama, H.; Morokuma, K.
28) Trott, O.; Olson, A. J. J. Comput. Chem. 2010, 31, 455−461.
Chem. Phys. Lett. 2014, 600, 87−95. (b) Rosati, F.; Boersma, A. J.;
Klijn, J. E.; Meetsma, A.; Feringa, B. L.; Roelfes, G. Chem. - Eur. J.
29) O’Boyle, N. M.; Banck, M.; James, C. A.; Morley, C.;
Vandermeersch, T.; Hutchison, G. R. J. Cheminf. 2011, 3, 33.
30) Morris, G. M.; Huey, R.; Lindstrom, W.; Sanner, M. F.; Belew,
R. K.; Goodsell, D. S.; Olson, A. J. J. Comput. Chem. 2009, 30, 2785−
791.
2
(
009, 15, 9596−9605.
8) (a) Boersma, A. J.; de Bruin, B.; Feringa, B. L.; Roelfes, G. Chem.
(
Commun. 2012, 48, 2394−2396. (b) Draksharapu, A.; Boersma, A. J.;
2
Browne, W. R.; Roelfes, G. Dalton Trans. 2015, 44, 3656−3663.
(
c) Draksharapu, A.; Boersma, A. J.; Leising, M.; Meetsma, A.;
Browne, W. R.; Roelfes, G. Dalton Trans. 2015, 44, 3647−3665.
9) (a) Gjonaj, L.; Roelfes, G. ChemCatChem 2013, 5, 1718−1721.
b) Park, S.; Zheng, L.; Kumakiri, S.; Sakashita, S.; Otomo, H.; Ikehata,
K.; Sugiyama, H. ACS Catal. 2014, 4, 4070−4073. (c) Fournier, P.;
Fiammengo, R.; Jas
(
(
̈
chke, A. Angew. Chem., Int. Ed. 2009, 48, 4426−
429. (d) Oltra, N. S.; Roelfes, G. Chem. Commun. 2008, 45, 6039−
041.
4
6
3
105
DOI: 10.1021/acscatal.6b00495
ACS Catal. 2016, 6, 3096−3105