Synthesis and Biological Screening of New Lawson Derivatives as Selective Substrate-Based Inhibitors of Cytochrome bo3 Ubiquinol Oxidase from Escherichia coli

Article abstract of DOI:10.1002/cmdc.201900707

The respiratory chain of Escherichia coli contains two different types of terminal oxidase that are differentially regulated as a response to changing environmental conditions. These oxidoreductases catalyze the reduction of molecular oxygen to water and contribute to the proton motive force. The cytochrome bo₃ oxidase (cyt bo₃) acts as the primary terminal oxidase under atmospheric oxygen levels, whereas the bd-type oxidase is most abundant under microaerobic conditions. In E. coli, both types of respiratory terminal oxidase (HCO and bd-type) use ubiquinol-8 as electron donor. Here, we assess the inhibitory potential of newly designed and synthesized 3-alkylated Lawson derivatives through L-proline-catalyzed three-component reductive alkylation (TCRA). The inhibitory effects of these Lawson derivatives on the terminal oxidases of E. coli (cyt bo₃ and cyt bd-I) were tested potentiometrically. Four compounds were able to reduce the oxidoreductase activity of cyt bo₃ by more than 50 percent without affecting the cyt bd-I activity. Moreover, two inhibitors for both cyt bo₃ and cyt bd-I oxidase could be identified. Based on molecular-docking simulations, we propose binding modes of the new Lawson inhibitors. The molecular fragment benzyl enhances the inhibitory potential and selectivity for cyt bo₃, whereas heterocycles reduce this effect. This work extends the library of 3-alkylated Lawson derivatives as selective inhibitors for respiratory oxidases and provides molecular probes for detailed investigations of the mechanisms of respiratory-chain enzymes of E. coli.

Full text of DOI:10.1002/cmdc.201900707

Accepted Article
Title: Synthesis and Biological Screening of new Lawson Derivatives
as selective substrate-based Inhibitors of Cytochrome bo3
Ubiquinol Oxidase from Escherichia coli
Authors: Isam Elamri, Melanie Radloff, Katharina F. Hohmann,
Vijaykumar D. Nimbarte, Hamid R. Nasiri, Michael Bolte,
Schara Safarian, Hartmut Michel, and Harald Schwalbe
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemMedChem 10.1002/cmdc.201900707
Link to VoR: http://dx.doi.org/10.1002/cmdc.201900707
A Journal of
www.chemmedchem.org

10.1002/cmdc.201900707
ChemMedChem
FULL PAPER
Synthesis and Biological Screening of new Lawson Derivatives as selective
substrate-based Inhibitors of Cytochrome bo3 Ubiquinol Oxidase from
Escherichia coli
Isam Elamri+,[a] Melanie Radloff+,[b] Katharina F. Hohmann,[a] Vijaykumar D. Nimbarte,[a] Hamid
R. Nasiri,[a] Michael Bolte, [c] Schara Safarian,[b] Hartmut Michel,[b] and Harald Schwalbe* [a]
[a] I. Elamri, K. F. Hohmann, V. D. Nimbarte. H. R. Nasiri. Prof. Dr. H. Schwalbe. Center for Biomolecular Magnetic Resonance Institute of Organic Chemistry and Chemical Biology Goethe-University Frankfurt am Main
(Germany) Max-von Laue-Straße 7, 60438 Frankfurt am Main (Germany) Email: schwalbe@nmr.uni-frankfurt.de
[b] M. Radloff, Dr. S. Safarian, Prof. Dr. H. Michel Department of Molecular Membrane Biology Max Planck Institute of Biophysics Max-von-Laue-Straße 3, 60438 Frankfurt am Main, Germany
[c] M. Bolte, Institute for Inorganic Chemistry, Goethe-University Frankfurt am Main (Germany).
[+] These authors contributed equally to this work.
Abstract
The respiratory chain of Escherichia coli contains two different types of terminal oxidases that are differentially regulated as a response
to changing environmental conditions. These oxidoreductases catalyze the reduction of molecular oxygen to water and contribute to
the proton motive force. The cytochrome bo3 oxidase (cyt. bo3) acts as the primary terminal oxidase under atmospheric oxygen levels,
while the bd-type oxidase is most abundant under microaerobic conditions. In E. coli, both types of respiratory terminal oxidases (HCO
and bd-type) utilize ubiquinol-8 as electron donor. Here, we assess the inhibitory potential of newly designed and synthesized
3-alkylated Lawson derivatives via the L-proline-catalyzed three-component reductive alkylation (TCRA). The inhibitory effects of these
Lawson derivatives on the terminal oxidases of E. coli (cyt. bo3 and cyt. bd-I) were tested potentiometrically. Four compounds were
able to reduce the oxidoreductase activity of cyt. bo3 by more than 50% without affecting the cyt. bd-I activity. Moreover, two inhibitors
for both cyt. bo3 and cyt. bd-I oxidase could be identified. Based on molecular docking simulations, we propose binding modes of the
new Lawson inhibitors. The molecular fragment benzyl enhances the inhibitory potential and selectivity for cyt. bo3 while heterocycles
reduce this effect. This work extends the library of 3-alkylated Lawson derivatives as selective inhibitors for respiratory oxidases and
provides molecular probes for detailed investigations of the mechanisms of respiratory chain enzymes of E. coli.
Introduction
antibacterial,[14,15] antiseptic or cytotoxic[16,17] activities. The
Naphthoquinones are among the most abundant compounds in
nature.[1] The growing interest to develop novel quinone
respiratory chain of E. coli performs redox chemistry by
exclusively utilizing quinols as electron carriers. Key enzymes are
derivatives for targeting diverse biomolecular processes
the quinol oxidizing cytochrome bd-I, bd-II and bo3 oxidases. At
originates from their role as essential metabolites in living
atmospheric oxygen conditions (250 µM O2), the cyt. bo3 oxidase
organisms, e.g. phylloquinone (Vitamin K1) in green plants or
ubiquinols and menaquinols in bacteria.[2] The naphthalene-
is the primary terminal oxidase (Km 2.4 µM). Under limited oxygen
concentrations cytochrome bd-type oxidases are essential for
based 1,4-diketone compounds are interacting with a broad
maintaining aerobic respiration due to their higher affinity for
spectrum of biological targets through two different mechanisms,
oxygen (Km 3 – 500 nM)[18]. The stoichiometry of the four electron
by either forming covalent bonds due to their electrophilic
reduction of O2 through oxidation of membrane bound ubiquinol-8
[3,4]
properties or reducing oxygen.
Derivatization of the 1,4-
(UQ8H2) to ubiquinone-8 (UQ8) by cyt. bo3 is shown in the
naphthoquinones scaffold into novel inhibitors opens potential
avenues to increase the possibility of interaction with biological
following equation:
[5–7]
molecules, in particular also to gain target specificity
Consequently, the inhibitory properties of 1,4-naphthoquinones
and their derivatives have been elucidated in multiple studies
The free energy of this redox reaction is utilized to pump four
protons from the cytoplasm to the periplasm while four substrate
protons from the cytoplasm are used for the reduction of dioxygen
presenting
e.g.
antimalarial,[8–10]
anticancer,[11–13]
1
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201900707
ChemMedChem
FULL PAPER
to water.[18,19] Based on their electron donors, members of the
heme-copper oxidases superfamily (HCO) are divided into two
classes. Quinol-type oxidases (QOX) use membrane-soluble
quinol molecules as two-electron donors while cytochrome c type
oxidases (COX) receive their electrons from membrane
associated cytochrome c proteins.[20–23] The E. coli cyt. bo3
oxidase is the best studied member of the QOX subfamily and is
characterized by four-subunits. Subunit I, II and III are
Synthesis
We synthesized a new set of alkylated hydroxynaphthoquinones
and probed their activity and selectivity against cyt. bo3 and
bd-I-type oxidases from E. coli. We aimed to synthesize a
systematic set of new 3-alkylated hydroxynaphthoquinones
(HNQ) using L-proline-catalyzed three-component reductive
alkylation (TCRA) (Scheme 1). TCRA was described previously
by Ramachary[23,33,34] as an efficient procedure to substitute cyclic
ß-keto-lactones with different aldehyde derivatives. Gribble et
al.[35] have shown that the reaction yield can be significantly
improved upon addition of two equivalents of corresponding
aldehyde. Alkylation of HNQ scaffold using the TCRA method
showed high alkylation rates, despite the expected
mesomerization in presence of the aromatic benzene ring and the
2-hydroxyl group, yielding up to 80% of desired 3-alkylated
Lawson-derivatives. However, in case of benzothiazol-2-
carbaldehyde derivative (Scheme 1, HNQ-11), the reaction yield
was considerably lower. Analysis of the reaction utilizing TLS-MS
spectrometry showed that proline-intermediate [II] was the
dominating part in the reaction mixture. Attempts to force the
equilibrium to the olefin-intermediate [III], either by increasing the
quantity of the reduction agent (Hantzsch ester) or of the catalyst
(L-proline), did not prove to be a successful approach. However,
we were able to isolate the proline-intermediate [II] from this
reaction using column chromatography as HNQ-11 (see
Scheme 1). This confirmed the hypothesized feasibility of the
TCRA alkylation mechanism for reductive alkylation of
homologues to the core subunits of the mitochondrial A-type
[22,24]
cytochrome c oxidase.
Cytochrome bd-type oxidases show
no sequence homology to members of the heme-copper
superfamily.[2,25,26] Recently, several studies reported the ability of
cyt. bo3 to utilize low potential quinones, menaquinones (MKs)
and demethylmenaquinones (DMKs) in their reduced form as
electron donors under anaerobic conditions.[27–32] Despite the high
structural similarity to these quinone substrates, the inhibitory
effects of naphthoquinone derivatives on the cytochrome bo3
(cyt. bo3) oxidase have not been performed in the past. Only
non-selective Aurachin and hydroxyquinoline N-oxide derivatives
(e.g. 2-heptyl-4-quinolinol 1-oxide, HQNO) have been used to
investigate the functional and structural properties of quinol
binding and oxidation.
Here, we describe synthesis and in vitro analysis of new
3-alkylated hydroxynaphthoquinones using L-proline-catalyzed
three-component reductive alkylation. Within this set of
compounds, we have identified new inhibitors with target
selectivity for cyt. bo3 of E. coli.
Scheme 1. Synthesis of 3-alkylated hydroxynaphthoquinone derivatives (HNQ-1-12) via Three-component reductive alkylation TCRA.
Hydroxynaphthoquinones (1 equiv), Hantzsch ester (1 equiv), L-proline (0.5 equiv). Solvent: DCM or MeOH for poorly soluble aldehyde derivatives. In
some cases, the crude product was stirred for 4h with LiOH (4 equiv) in H2O/MeOH to remove the excess of oxidized Hantzsch ester (saponification).
The general procedure of the TCRA method is described in experimental section. Characterization of data is shown in the Supporting Information.
Figures S2-S36.
2
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201900707
ChemMedChem
FULL PAPER
2-hydroxynaphthoquinone, as shown in Scheme 2. The
multiple-component cascade reaction starts with activation of the
aldehyde to the corresponding iminium intermediate [I], which can
undergo a nucleophilic attack from the double bond of the vinyl
alcohol of Lawson 0 under very mild conditions, generating the
substituted naphthalenetrione-intermediate in situ [II]. Due to high
acidity of [II] a reductive elimination reaction occurs, giving a
3-alkylated olefin [III], which undergoes the redox reaction with
Hantzsch ester, forming the corresponding hydrogenated product
[IV] (Scheme 2a). The hypothesized mechanism of this reaction
was supported by the inadvertently isolated intermediate HNQ-11
and the crystal structure of HNQ-6 (Scheme 2b and 2c).
higher than 8.3 mM or even 10.3 mM in buffer. Therefore, all
twelve HNQ compounds are feasible for further functional studies.
Scheme 3.a) Solubility profile of HNQ-2 in buffer (20 mM NaPi, 50 mM
NaCl, 0,02% DDM, pH 8.0) at room temperature and detection
wavelength λ = 274 nm. b) Absorbance spectra of HNQ-2 for all
concentrations used in the assay.
Table 1: Experimental solubility data of HNQ-1 – HNQ-12 in buffer
(20 mM NaPi, 50 mM NaCl, 0,02% DDM, pH 8.0) at room temperature.
Stock concentrations and concentrations for particular measurements
are shown in Supporting Information. Table S1.
Compound
Detection
Wavelength [nm]
Solubility
[µg/mL]
335±24
334±7
627±21
267±19
1669±69
400±44
> 2677
> 3135
> 2677
840±151
800±85
1640±94
Solubility
[mM]
HNQ-1
HNQ-2
HNQ-3
HNQ-4
HNQ-5
HNQ-6
HNQ-7
HNQ-8
HNQ-9
HNQ-10
HNQ-11
HNQ-12
273
274
470
275
273
274
479
275
273
490
275
273
1.02±0.08
1.26±0.03
2.36±0.08
0.87±0.06
5.5 ±0.3
0.78±0.09
> 8.3
> 10.3
> 8.3
2.1±0.4
1.84±0.19
4.6±0.3
Scheme 2.a) Hypothesized mechanism for catalytic cycle of TCRA to
alkylated Lawson. b) Isolated HNQ-11 confirming intermediate [II]. c) Atomic
model of the X-ray structure of HNQ-6 confirming the alkylated Lawson.
Crystallographic data of HNQ-6 are shown in Supporting Information Figure
S1.
Enzyme activity assay
Solubility assay
The inhibitory effect of a set of HNQs was screened in micromolar
range to identify selective cyt. bo3 and cyt. bd-I inhibitors. Oxygen
reduction was measured via a Clark-type electrode. Turnover
rates of cyt. bo3 and bd-I experiments are presented as relative
activities (Figure 1). For experimental overview and further
information see Supporting Information (Figures S38-40), Tables
S2-3). Activity of cyt. bo3 was strongly (>50 %) reduced by
HNQ-1, -2, -6, -7 and -12. HNQ-1 and HNQ-12 are the most
effective cyt. bo3 inhibitors, yet lack selectivity as they also inhibit
cyt. bd-I, but in a lower extend compared to HQNO. HQNO is
described as a noncompetitive[37] or uncompetitive[38,39] inhibitor.
To meet the requirements for further functional investigations,
solubilities of the HNQ compounds were determined by an
absorbance-based thermodynamic equilibrium solubility assay.[36]
As shown in Scheme 3a, plotting absorbance over concentration
at the wavelength of the maximum peak of the absorbance
spectra, (HNQ-2 at λ=274 nm), provides a saturation point at
1.26±0.03 mM (solubility graphs for all HNQ compounds are
shown in Supporting Information. Figure S37). Table 1 indicates
that solubility of most evaluated compounds ranges between
0.87 mM and 5.48 mM, whereas three of them show solubility
3
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201900707
ChemMedChem
FULL PAPER
Overall, the E. coli bd-I type oxidase was less sensitive to the
majority of examined compounds. Merely the two most
hydrophobic substances showed an inhibitory effect on the
cyt. bd-I type oxidase (HNQ-1: 30 %; HNQ-12: 56 %).
affects the selectivity compared to HNQ-6. HNQ-10 contains a
resorcin residue and shows a low inhibition of 31 %. This suggests
that addition of two hydroxy groups lowers the inhibitory effect on
cyt. bo3 compared to the simply benzylated HNQ-2. It is
conceivable that the negative inductive effect (–I) of the
OH-groups lowers the aromatic Pi-Pi system in HNQ-10 since this
group possibly interacts with the amino acids of the cyt. bo3 quinol
binding site.
No or a low inhibitory effect (< 30 %) was observed for
HNQ-3, -4, -8, -9 and -11 with both oxidases. With respect to
cyt. bo3 selectivity HNQ-2, -6 and -7 are the most promising
candidates for further rounds of chemical optimization.
A similar effect was observed for the nitrogen containing moieties
in HNQ-3, -4, -5, -8 and -11. HNQ-11 is characterized by a bulky
residue composed of a benzthiazole and proline that lacks the
methylene spacer to the HNQ backbone. HNQ-8 contains a
7-azaindol group. We could not show an influence on either
cyt. bo3 or bd-I type oxidase activity with these compounds. In
HNQ-3 a pyridine group was added that seems to impede an
interaction with both enzymes since the activity remains
unaffected. Changing this group to N,N-dimethylanilin, as in
HNQ-4, improved the inhibitory effect on both enzymes (15 % rest
activity for cyt. bo3 and cyt. bd-I) but did not improve selectivity for
cyt. bo3.
Regarding
HNQ-5
a
modification
of
the
N,N-dimethylanilin to a 1-H-indazole residue results in a higher
selectivity for cyt. bo3. HNQ-5 shares a bulky residue similar to
HNQ-8 composed of two rings including two nitrogens as
heteroatoms. However, in HNQ-5 the benzylic-part of the
1-H-indazole group is linked to the hydroxynaphthoquinone
backbone instead of the C5 heterocycle of HNQ-8.
Figure 1. Screening of the inhibition potentials of tested
hydroxynaphthoquinones (HNQ) on oxidase reduction activity of purified
E. coli cyt.bo3 (white) and bd-I (light grey) oxidase. Inhibition assay using
HNQs at 250 µM in presence of 200 µM ubiquinone-1 and 5 mM
dithiothreitol. Oxygen reduction activity was calculated from oxygen
consumption rates at 30 nM enzyme concentration at RT for each
experiment. 100% activity = adding corresponding volume of EtOH. KCN
at 20 µM and HQNO at 250 µM represent positive inhibition controls for
cyt. bo3. Presented data are mean values of three independent
experiments ± standard deviation.
a)
b)
In order to understand the correlation between inhibition and the
chemical nature of the alkylated residue at position 3 of the HNQ
backbone, we systematically grouped our compounds based on
the substituents and comparatively evaluated their effects
(Figure 2). An overview of the effect of structural modifications on
the inhibition potential of oxygen reduction activity is given in
Table 2. For residues composed of a benzylic ring in combination
with an alkoxy or hydroxy group a decrease in inhibition is
noticeable depending on position and arrangement of the
heteroatoms (from HNQ-12 over -2, -6, -7 to -10) (Figure 2A).
Benzylated HNQ-2 showed a strong inhibitory effect on cyt. bo3,
while not influencing the cyt. bd-I activity at all. Compounds
HNQ-6 and -7 share a positive mesomeric effect (+M) due to the
aromatic character of their residues. The benzodioxole of HNQ-6,
enhanced the inhibition of cyt. bo3 activity (<30 % rest activity for
cyt. bo3) while also showing an increased selectivity for the
HCO type oxidase. Changing the dioxolan ring to a hydroxy and
a methoxy group in HNQ-7 slightly reduces the inhibitory effect on
cyt. bo3 (<50 % residual activity for cyt. bo3) and also negatively
Figure 2. Overview of relative inhibition of benzylic and alkoxy or hydroxy
containing residues (a) and nitrogen containing residues (b) on oxidase
reduction activity of E. coli cyt bo3 with relevant residues in position 3.
Interestingly, while HNQ-5 shows a selective inhibitory effect on
cyt. bo3, HNQ-8 does not affect oxygen reductase activities at all
(40 % inhibition HNQ-5; 0 % inhibition HNQ-8). Concerning the
poor inhibitory effect of the heteroaromatic compound HNQ-3 it is
reasonable that the heteroaromatic character of the pyridine
residue hinders interaction with cyt. bo3.
4
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201900707
ChemMedChem
FULL PAPER
Figure 3. a) Proposed binding site of conjugate HNQ-2, HNQ-6 and HNQ-7 within the subunit-1 of cytochrome bo3 ubiquinol oxidase (PDB: 1FFT);
b) 2D-interactions of HNQ-2 with ASN-A-124 (Pi-donor in cyan color), PHE-A-146 (Pi-Pi-stacking interactions in pink color), ILE-A-120 and PRO-A-117
(Pi-alkyl interactions in light pink color); 2D-interaction of HNQ-7 with PHE-A-146 (Pi-sigma interactions in violet color), SER-A-145, ILE-A-120 (amide-Pi
stacked in violet color), GLY-A-121, PHE-A-118 (Carbon-hydrogen donor in cyan color) and with PRO-A-117 (Pi-alkyl interactions in pink color);
2D-interactions of HQNO with VAL-A-114 (Pi-sigma interactions in violet color) and with ILE-A-109, PHE-A-113, PRO-A-117, VAL-A-153, VAL-A-156
(Pi-alkyl interactions in light pink color); 2D-interactions of Aurachin C with ARG-A-71 (Pi-cationic interactions in yellow color), VAL-A-114 (Pi-sigma
interactions in violet color), MET-A-110 (Pi-sulfur interactions in yellow color), and ILE-A-109, PHE-A-113, PRO-A-117, PHE-A-146, VAL-A-153,
VAL-A-156, ILE-A-109 (Pi-alkyl interactions in light pink color), and with GLY-A-72 (carbon-hydrogen bond interactions in light green color).
Our
experiments
revealed
that
alkylation
of
To explore molecular binding interactions with cyt. bo3 we
subsequently performed in-depth docking analyses for HNQ-2,
HNQ-6 and HNQ-7, respectively.
hydroxynaphthoquinones with nitrogenous heterocycles e.g.
benzamine (HNQ-4), indazole (HNQ-5), azaindole (HNQ-8) and
benzthiazole (HNQ-11) is not expedient to inhibit neither cyt. bo3,
nor cyt. bd-I activity.
Table 2: Effect of structural hydroxynaphthoquinone modification on the
inhibition potential of purified E. coli cyt. bo3 and bd-I oxidase.
Substituted benzenes groups showed an improved inhibitory
effect since this seems to show a certain level of tolerance
towards electron-donating groups e.g. alkoxy in HNQ-6 or
HNQ-12. This effect diminishes by introducing a hydroxy group
(HNQ-7) or regarding HNQ-10 a double hydroxylation. HNQ-9
represents the only tested metallocene. This feature did not
improve the inhibitory effect nor selectivity compared to metal-free
HNQs.
Effect on inhibition
Structural Inhibitor Modification
HNQ
bo3
bd
2-Hydroxy group instead of 2-Methyl
group of quinone substrates
Alkylation with long chain or bulky
residues
1-12
+++
- - -
1,12
2
+++
+++
++
Simple benzylation*
- - -
Benzylation with aromatic substituents
that exhibit a positive mesomeric
effect
6,7
++
- - -
Benzylation with aromatic substituents
that exhibit a negative inductive effect
values for HQNO, HNQ-12 and HNQ-6 were Alkylation with nitrogen heterocycles
To gain further insights into inhibition efficiencies, the apparent ki
10
- - -
- - -
- - -
- - -
(kiapp
determined.[37,40]. The assay was performed without preincubation
of enzyme and inhibitor and with fixed ubiquinone-1
)
3-5,
8,11
*the attempt to increase the inhibition effect on cyt. bo3 through introducing a
phenoxygroup (HNQ-12) containing two phenyl groups was successful but at
expence of selectivity since inhibition of bd-oxidase was also noticeably
increased
a
concentration of 200 µM. For HQNO kiapp was found to be in a
micromolar range at 3,2 ± 0,7 µM. In case of HNQ-12 kiapp was
determined on a similar scale at 5,7 ± 0,6 µM. The kiapp for the
third tested compound HNQ-6 was found to be two orders of
magnitude higher at 225 ± 9 µM (see Supporting Information
Figure S41-43). Previous work determined a kiapp for HQNO at
0.7 µM via steady state kinetic experiments.[40]
Molecular docking studies
Ubiquinol cyt. bo3 oxidase crystal structure is composed of
tetramer subunits chain-A, chain-B, chain-C, chain-D, chain-E
and chain-F, respectively with a sequence length of 663 amino
acids. Most frequently used inhibitors for cyt. bo3 are the
5
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201900707
ChemMedChem
FULL PAPER
non-competitive HQNO (2-n-heptyl-4-hydroxyquinoline N-oxide)
and different Aurachin C derivatives.[37,40–43] Previous work by
Choi et al.[40] explored E.coli cyt. bo3 with the natural substrate
ubiquinol-8 and known inhibitors Aurachin C derivatives as well
as HQNO (2-n-heptyl-4-hydroxyquinoline N-oxide) suggested that
there is only one binding site for ubiquinol, a high affinity
single-site (QH), which disprove the current consensus model of
two binding sites (QH, QL). Furthermore, it was confirmed that
inhibitors such as HQNO bind also on the QH site displacing only
the quinone part of the endogenous substrate due the high affinity
of the isoprene chain to the protein. [40]
Interactions of HNQ-2: We found that for HNQ-2, the
3-hydroxynaphthalene-1,4-dione central scaffold is involved in the
Pi-Pi stacking (pink color) interaction with PHE-A-146. In addition,
unconventional Pi-donor hydrogen binding interactions (cyan
color) were observed with ASN-A-124. The phenyl ring with
methylene spacer attached to central ring gives an opportunity to
rotate freely and found to be involved in the Pi-Pi stacking
interactions (pink color) with the PHE-A-146 and some Pi-alkyl
interactions (light Pink color) with ILE-A-120 and PRO-A-117
respectively (see Figure 3 and Supporting Information Figure
S44).
In-depth docking analysis has been performed to elucidate the
interactions of the most active agents HNQ-2, HNQ-6 and HNQ-7
and HQNO, Aurachin C1-10 as reference inhibitors respectively
at the QH site of cyt. bo3 ubiquinol oxidase (PDB: 1FFT) by using
Autodock 4.2.[21] Interestingly, we observed that despite structural
differences, the quinolinone ring of Aurachin C1-10 interacts with
the same two residues like the hydroquinone ring of ubiquinol-2,
namely Arg 71 and Asp 75. It is likely that the binding sites for
Aurachin C and ubiquinol-2 are partially overlapping. Further, by
comparing docking of all three inhibitor classes it could be found
that the binding mode of HNQ-derivatives is more similar to that
of HQNO than of Aurachin C. Aurachin C interacts with residues
in the substrate binding loop (see Figure 3 and Supporting
Interactions of HNQ-6: Molecular docking insight of HNQ-6
suggests that the central naphthalene ring is involved in Pi-sigma
interactions (violet color), Pi-Pi stacking interactions (pink color)
and Pi-alkyl interactions (light pink color) with PHE-A-146 and
ILE-A-120 respectively. In addition to this conventional hydrogen
interactions (green color) and unconventional Pi-donor hydrogen
binding interactions (cyan color) were observed in 1,4-dione
carbonyl and hydroxyl moiety of HNQ-6 with ASN-A-124 and
PRO-A-117 respectively. The 1,3-benzodioxole residues
attached to the HNQ-ring gives an opportunity to rotate freely and
found to be involved in the Pi-alkyl interactions (light pink color)
with the LEU-A-122 and unconventional Pi-donor hydrogen
binding interactions (cyan color) with PHE-A-118 (see Figure 3
and Supporting Information Figure S44).
Information
Figure
S47).
Our
designed
set
of
hydroxynaphthoquinones inhibitors were inspired from the native
substrate ubiquinol-8 (ring B in Figure 4) and HQNO (ring A in
Figure 4). During the development of diverse set of ligands
several modifications were introduced into the basic scaffold.
The first derivative HNQ-1 exhibits a long aliphatic side chain,
which resembles to the side chain of HQNO and substrate
ubiquinol-8. The results of enzyme assays suggest that the
scaffolds with relatively long aliphatic side chain are non-selective
in nature. To investigate it further we performed intensive
molecular docking studies and found that the aliphatic side chains
are involved in steric hindrance and some unconventional
intermolecular bumps were observed in HNQ-1 and HQNO. This
might be the reason for weak binding of HNQ-1 and HQNO into
the hydrophobic pocket of the cyt. bo3 ubiquinol oxidase
(PDB: 1FFT). Further docking analysis were performed on the
most promising derivatives HNQ-2, HNQ-6 and HNQ-7, which
showed selective inhibition of cyt. bo3. The binding energies
Interactions of HNQ-7: Hydroxynaphthoquinone HNQ-7 is
involved in the Pi-Pi stacking (pink color) and Pi-alkyl (light pink
color) with ILE-A-128, PHE-A-146, SER-A-145. The methoxy
hydroxy ring is involved in the unconventional hydrogen binding
interactions (light green color) in between methoxy group and
GLY-A-121 and PHE-A-118. In addition to this pi electron cloud of
substituted phenyl ring is involved in Pi-Pi stacking interaction with
PRO-A-117 and GLY-A-121 respectively (see Figure 3 and
Supporting Information Figure S44).
Table 3. Binding energies for best fives conformations of HNQ-derivatives
and HQNO and Aurachin C1-10 for cyt. bo3. For remaining
HNQ-derivatives, see Supporting Information Table S4.
Binding energies kcal/mol
HQNO
-5.72
-4.7
-5.41
-4.65
-5.32
-4.01
-5.24
-3.96
-5.19
-3.65
Aurachin
C1-10
HNQ-2
HNQ-6
HNQ-7
-6.69
-6.95
-6.44
-6.68
-6.81
-6.22
-6.58
-6.65
-6.35
-6.63
-6.49
-6.28
-6.18
-6.22
-6.05
presented in Table
2 support results of enzyme activity.
HNQ-2, -6, -7 showed a lower binding energy in comparison with
HQNO and Aurachin C1-10.
As outlined in Figure 3, the comparison of the binding pockets of
HNQ-2, HNQ-6 and HNQ-7 to the reference inhibitor HQNO
6
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201900707
ChemMedChem
FULL PAPER
revealed that the designed set of ligands are binding into the
similar pocket but with slight difference mode of interactions
pattern with regard to the involved amino acids. Furthermore, the
methylene-bridge of all alkylated HNQ-derivatives showed no
interactions, which facilitate presumably to the flexibility of the
compounds needed to enter the binding pocket. Indeed, HNQ-11
lacks this methylene spacer and shows relatively lowest inhibitory
potential of all tested HNQs. In addition, the key finding here
suggesting that in all scaffolds PHE-A-146 is the major interaction
partner with the hydroxynaphthoquinone ring. In addition, for the
most selective cyt. bo3 inhibitor HNQ-2, an interesting set of Pi-Pi
interactions in between PHE-A-146 and the benzyl-ring was
observed. Moreover, this benzylic Pi-electron cloud is also playing
crucial role in interactions with ILE-A-120 and PRO-A-117, which
creates space to accommodate the HNQ-2 within the binding
pocket of ubiquinol oxidase cyt. bo3. These key findings from the
molecular modelling studies are in good agreement with the
enzyme inhibition assays.
To confirm the selectivity of designed alkylated HNQ derivatives
for cyt. bo3 oxidase we performed docking analyses with cyt. bd
oxidase. We used the atomic structure of the bd-oxidase (5DOQ)
from Geobacillus thermodenitrificans[44] (determined by X-ray
diffraction at 3,05 Å). Based on this crystal-structure (5DOQ) of
the menaquinol oxidase we performed docking in the quinol
binding loop.[45] We observed that HQNO and Aurachin C show a
higher affinity for cyt. bd oxidase based on the binding energies
compared to cyt. bo3 (see Supporting Information Table S5). In
contrary, our set of tested HNQ-derivatives is more selective for
cyt. bo3 oxidase. Our in silico experiments are hence
complementary to the biochemical data presented in this work.
Detailed interactions of HQNO, Aurachin C, and HNQ-derivatives
are discussed in Supporting Information (Figure S48-52). The
long side chain seems to be the reason for the higher affinity of
HQNO and Aurachin C1-10 towards the menaquinol binding
domain of the cytochrome bd oxidase. By computational docking,
it becomes obvious that the side chain is involved in alkyl and
Pi-alkyl type of interactions with ILE-A-146, ILE-A-22, TRP-A-374,
ILE-A-370 and VAL-A-187. The central scaffold of
Aurachin C1-10 is involved in a Pi-sigma type interaction with
ALA-A-377 and a Pi-Pi T-shaped interaction with HIS-A-186.
In conclusion, the current work extended the library of cyt. bo3
oxidase inhibitors with the beneficial naphthoquinone class.
Twelve
new
3-alkylated
synthesized
hydroxynaphthoquinones
using L-proline-catalyzed
derivatives were
three-component reductive alkylation (TCRA). The successfully
tested one-step reaction was suitable to achieve a high alkylation
efficiency in most cases. Furthermore, due to isolation of one
intermediate HNQ-11 the hypothesized alkylation mechanism
was confirmed. Solubility of all synthesized derivatives
was determined spectrophotometrically by a thermodynamic
equilibrium assay in 96-well plates to meet the requirements for
functional enzyme tests. Hence in these inhibition
Figure 4. a) Primary design of HNQ-1 inspired from the combination of
substrate
ubiquinol-8
and
inhibitor
HQNO.
b)
Modified
hydroxynaphthoquinones HNQ-2, HNQ-6, and HNQ-7.
Conceptually, we consider our results as an extension of the
conclusion that has been shown in the study of Mogi et al., by
which the ubiquinol oxidation sites in cytochromes cyt. bo3 and
cyt. bd were examined with modified Aurachin C analogues.[41] It
was demonstrated that an increase in the chain length of the alkyl
chain decreases the inhibitory effect towards cyt. bo3 in contrast
to cyt. bd, suggesting a smaller binding site of cyt. bo3 compared
to cyt. bd. In the current study, we observed that alkylations of
hydroxynaphthoquinone scaffold with alkyl chains or bulky
substituents are positively influencing the inhibition of both,
cyt. bd and cyt. bo3 (e.g. HNQ-1 and HNQ-12). We further found
that alkylation with cyclic residues like benzyl groups (HNQ-2,
HNQ-6, HNQ-7) increases the inhibitory potency only for cyt. bo3.
experiments, this
class
shows
a
higher
selectivity
for E. coli cyt. bo3 compared to previously studied inhibitors
Aurachin C or HQNO. A purposed interaction mode with cyt. bo3
was
demonstrated. Simple benzylated HNQ-2 is the
most
favorable cyt. bo3 inhibitor regarding selectivity and efficacy within
our library. Furthermore, combination of enzyme activity
assay and In-depth docking analysis provided detailed insights
into the binding site of new alkylated HNQ derivatives,
the positive role of the methylene bridge as well as characteristic
of substituents to improve interactions with the amino acids within
the binding pocket of targeted enzyme.
7
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201900707
ChemMedChem
FULL PAPER
2H) ppm. 13C NMR (150 MHz, CDCl3): δ = 183.2, 180.7, 152.1,
137.9, 134.0, 132.0, 131.8, 128.4, 128.2, 127.4, 125.9, 125.3,
125.1, 122.0, 28.2 ppm. (MS-ESI): m/z = calcd for C17H12O3 [M-
H]-: 263.08; found 263.09.
Experimental Section
General
Procedure:
Synthesis
of
3-
alkylated
hydroxynaphthoquinone derivatives (HNQ1-12) via TCRA
methode: A mixture of 2-hydroxy-1, 4-naphthoquinone (1 equiv),
Hantzsch ester (1 equiv), L-Proline (0,5 equiv) and corresponding
aldehyde (2 equiv) dissolved in dried CH2Cl2 or MeOH was heated
under N2 atmosphere 18 h to reflux. After cooling to room
temperature the obtained solution was concentrated in vacuum
and the alkylated product was purified by silica-gel column
chromatography. In some cases, the purification was
impracticable in the presence of oxidized Hantzsch ester
exhibiting identical Rf value to those of desired product. Therefore,
a mixture of Methanol/Water, 3:1 was added to the mixed fraction
(after HPLC purification). The precipitated solid was filtrated out
and LiOH (3-5 equiv) was added (reaction was controlled by TLC).
The Solution was than stirred for 2 h and Methanol was removed.
The aqueous layer was covered with DCM and acidified with a
saturated NaHCO3-Solution. The two layers were separated and
the aqueous layer was further extracted with DCM. The organic
layers were combined and the solvent was evaporated under
reduced pressure to yield finally after the saponification the pure
product.
2-Hydroxy-3-methylen-(4-N-pyridyl)-1,4-naphthoquinone
(HNQ-3) Aldehyde: 4-Pyridincarboxyaldehyde. Reaction solvent:
CH2Cl2. Column chromatography with c-hexane/EtOAc, 1:1.
Orange solid (55%). Rf = 0.3 (c-hexane/EtOAc, 4:1). 1H NMR (600
MHz, CDCl3): δ 8.49 (d, 3J = 5.9 Hz, 2H), 8.1 (dd, 3J = 7.6 Hz, 4J
= 0.7 Hz, 1H), 8.0 (dd, 3J = 7.5 Hz, 4J = 0.9 Hz, 1H) 7.90 (td, 3J =
4
3
4
7.7 Hz, J = 1.2 Hz, 2H) 7.85 (td, J = 7.7 Hz, J = 1.2 Hz, 2H),
7.34 (d, 3J = 5.9, 2H), 3.90 (s, 2H) ppm. 13C NMR (150 MHz,
DMSO-d6): δ = 183.9, 181.0, 156.9, 149.0, 134.5, 133.2, 131.9,
130.2, 125.7, 124.0, 120.3, 28.1 ppm. ESI-MS (m/z): calcd for
C16H11NO3, [M-H] -: 264.07; found 264.09.
2-Hydroxy-3-methylen-(4-N,N-dimethylanilid)-1,4
naphthoquinone
Dimethylaminobenzaldehyde. Reaction solvent: CH2Cl2. Column
chromatography with c-hexane/EtOAc, 2:1. Orange solid
(HNQ-4)
Aldehyde:
4-N,N-
(28.4 %). Rf = 0.4 (c-hexane/EtOAc, 4:1). 1H NMR (600 MHz,
CDCl3): δ 8.01 (d, 3J =7.2 Hz, 1H), 7.95 (d, 3J =7.4 Hz, 1H), 7.63
(m, 1H), 7.56 (m, 1H), 7.2 (d, 3J =8.6 Hz, 2H), 6.58 (d, 3J =8.2 Hz,
2H), 3.77 (s, 2H), 2.79 (s, 6H) ppm. 13C NMR (150 MHz, CDCl3):
δ = 183.5, 180.8, 151.7, 148.2, 133.8, 131.8, 128.8, 128.4, 125.8,
125.0, 122.7, 111.9, 39.8, 27.1 ppm. ESI-MS (m/z): [M-H] - calcd
for C19H17NO3, 306.12; found 306.13.
2-Hydroxy-3-tridecyl-1,4-naphthoquinone (HNQ-1)
Aldehyde: Tridecanal. Reaction solvent: CH2Cl2. Column
chromatography with c-hexane/EtOAc, 4:1. Yellow solid (65%).
Rf = 0.70 (c-hexane/EtOAc, 1:1).1H NMR (600 MHz, CDCl3): δ
8.12(d, 3J= 7.9 Hz, 1.0 Hz, 1H), 8.07 (dd, 3J = 7.6 Hz, 4J = 1.1 Hz,
1H), 7.75 (td, 3J = 7.7 Hz, 4J = 1.5 Hz, 1H), 7.67 (td, 3J = 7.7 Hz,
4J =1.1, 1H), 7,30 (br s, 1H, OH), 2.59 (t, 3J = 7.7 Hz, 2H), 1.53-
1.25 (m, 22H), 0.87 (t, 3J = 6.9 Hz, 3H) ppm. 13C NMR (150 MHz,
CDCl3): δ = 184.8, 181.6, 153.2, 135.0, 133.1, 132.9, 129.6, 127.0,
126.2, 125.0, 32.0, 29.9, 29.8, 29.7, 29.6, 29.5, 28.4, 23.6, 22.8,
14.3 ppm. (MS-ESI): m/z = calcd for C23H32O3 [M-H]-: 355.24;
found 355.27.
2-Hydroxy-3-[(indazol-5-yl)methyl]-1,4-naphthoquinone
(HNQ-5) Aldehyde: Indazol-5-carboxyaldehyde. Reaction
solvent: CH2Cl2. Column chromatography with c-
hexane/EtOAc, 2:1.
Orange solid (66%). Rf = 0.16 (c-
1
hexane/EtOAc, 2:1). H NMR (600 MHz, DMSO-d6): δ 12.9
(br,1H), 8.0 -7.9 (m, 3H), 7.8 (td, 3J = 7.9 Hz, 4J = 1.1 Hz, 1H),
7.8 (td, 3J = 7.7 Hz, 4J = 1.0 Hz, 1H), 7.6 (s, 1H), 7.42 (d, J =
8.9 Hz,1H), 7.3 (dd, 3J = 8.5 Hz, 4J = 1.2 Hz, 1H), 3.9 (s, 2H)
ppm. 13C NMR (150 MHz, CDCl3): δ = 184.3, 181.2, 155.7,
134.5, 133.2, 131.9, 131.2, 130.1, 127.4, 125.8, 125.7, 123.0,
122.9, 119.3, 109.8, 31.1, 29.9, 28.4 ppm. ESI-MS (m/z): [M-
H] - calcd for C18H12N2O3, 303.08, found 303.11.
2-Hydroxy-3 -methylene-(benzyl)-1,4-naphthoquinone (HNQ-
2)
Aldehyde: Benzaldehyde. Reaction solvent: CH2Cl2. Column
chromatography with c-hexane/EtOAc, 5:1. Yellow solid (55%). Rf
= 0.45 (c-hexane/EtOAc, 5:1). 1H NMR (600 MHz, CDCl3): δ 8.04
(dd, 3J = 7.6 Hz, 4J = 0.7 Hz, 1H), 7.98 (dd, 3J = 7.5 Hz, 4J = 0.9
Hz,1H), 7.66 (td, 3J = 7.5 Hz, 4J = 1.2 Hz, 1H), 7.59 (td, 3J = 7.6
2-Hydroxy-3-[(1,2-methylenedioxybenz-5yl)methyl]-1,4-
naphthoquinone (HNQ-6) Aldehyde: Piperonal.
Reaction
4
3
Hz, J = 1.2 Hz, 1H), 7.37 (br s, 1H, OH), 7.32 (d, J = 7.4 Hz,
2H), 7.18 (t, 3J = 7.4 Hz, 2H), 7.10 (t, 3J = 7.2 Hz, 1H), 3.87 (s,
solvent: CH2Cl2. Column chromatography with c-hexane/EtOAc,
2:1. Orange solid (80%). Rf = 0.37 (c-hexane/EtOAc, 2:1). 1H
8
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201900707
ChemMedChem
FULL PAPER
3
3
NMR (600 MHz, CDCl3): δ 8.04 (d, J = 7.6 Hz, 1H), 7.98 (d, J
=7.4 Hz, 1H), 7.67 td, 3J = 7.5 Hz, 1H), 7.60 (td, 3J = 7.5 Hz, 1H),
6.82 (br, 1H), 6.78 (d, 3J = 8.2 Hz, 1H), 6,62 (d, 3J = 8.0 Hz, 1H),
5.81 (s, 2H), 3.78 (s, 2H) ppm. 13C NMR (150 MHz, CDCl3): δ =
184.4, 181.7, 152.9, 147.5, 146.0, 135.1, 133.0, 132.8, 132.6,
129.4, 126.9, 126.1, 123.1, 122.1, 109.8, 108.2, 100.8, 28.9 ppm.
ESI-MS (m/z): [M-H] - calcd for C18H12O5, 307.07, found 307.08.
2-Hydroxy-3-[(3,5-dihydroxybenzyl)methyl]-1,4-
naphthoquinone (HNQ-10) Aldehyde:
3,5-
Dihydroxybenzaldehyde Reaction solvent: MeOH. Column
chromatography with CH2Cl2: Acetone 49:1. Red solid (66%). Rf
= 0.56 (CH2Cl2/MeOH, 3:1). 1H NMR (600 MHz, DMSO-d6): δ
8.90 (br. s, 2H), 7.91-7.70 (m, 2H), 7.76 (m, 1H), 7.70 (m, 1H),
6.16 (s, 2H), 5.96 (s, 1H), 3.18 (s, 2H) ppm. 13C NMR (150 MHz,
MeOD): δ = 185.3; 159.0; 145.4; 136.07; 132.7; 127.0; 122.0;
115.6; 108.4; 100.9; 29.8 ppm. ESI-MS (m/z): [M-H] - calcd for
C17H12O5 295.07; found 294.94.
2-Hydroxy-3-[(4-hydroxy-3-methoxyphenyl)methyl]-1,4-
naphthoquinone (HNQ-7) Aldehyde: Vanillin. Reaction solvent:
CH2Cl2. Column chromatography with c-hexane/EtOAc, 2:1.
Yellow solid 37%). Rf = 0.16 (c-hexane/EtOAc, 2:1). 1H NMR (600
MHz, DMSO-d6): δ 8.7 (br. s, 1H), 8.0 (dd, 3J = 7.7 Hz, 4J = 0.9
Hz, 1H); 7.97(dd, 3J = 7.7 Hz, 4J = 1.4 Hz, 1H), 7.82 (td, 3J = 7.6
Hz, 4J = 1.5 Hz, 1H); 7.77 (td, 3J = 7.5 Hz, 4J = 1.5 Hz, 1H); 6.85
(s, 1H); 6.64 (m, 2H); 3.73 (s, 3H) ppm. 13C NMR (150 MHz,
DMSO-d6): δ = 181.8, 178.7, 153.0, 144.8, 142.2, 132.0, 130.7,
129.5, 127.7, 127.5, 123.3, 123.2, 120.5, 118.3, 112.8, 110.6,
53.2, 25.5 ppm. ESI-MS (m/z): [M-H] - calcd for C18H14O5, 309.08;
found 309.12.
2-Hydroxy-3-[(2-thiazol(2-3)benzyl)methyl)pyrrolidine-2-
carboxylic acid-1,4-naphthoquinone (HNQ-11) Aldehyde:
Benzothiazole-2-carboxaldehyde. Reaction solvent: MeOH.
Column chromatography with CH2Cl2/MeOH 14:1.Red solid
(60%).Rf = 0.55 (CH2Cl2/MeOH, 14:1).1H NMR (600 MHz, CDCl3):
δ 7.99 (d, 3J = 7.9 Hz, 2H), 7.96 (d, 3J = 7.9 Hz, 1H), 7.80 (d, 3J =
8.1 Hz, 1H), 7.61 (td, 3J = 7.4 Hz, 4J = 1.3 Hz, 1H), 7.54 (td, 3J =
7.7 Hz, 4J = 1.3 Hz, 1H), 7.43 (t, 3J = 7.6 Hz, 1H), 7.34 (t, 3J = 7.6
Hz, 1H), 4.55 (t, 3J = 8.4 Hz, 1H), 4.47 (d, 3J = 14.7 Hz, 1H), 4.14
(d, 3J = 14.6 Hz, 1H) 3.55 (m, 1H), 2.94 (q, 3J = 9.7 Hz), 2.43 (m,
1H), 2.0 – 1.93 (m, 3H), ppm. 13C NMR (150 MHz, CDCl3): δ =
183.1, 181.8, 163.1, 152.8, 135.7, 133.9, 132.8, 132.4, 131.5,
126.6, 126.4, 125.9, 125.8, 123.6, 121.8, 64.7, 54.9, 52.9, 31.6,
23.4 ppm. MALDI: [M-CO2]- calcd for C23H18N2O5S, 390.1; found
390.7.
2-Hydroxy-3-[(1H-pyrrolo(2,3-b)pyridine-3-yl)methyl]-1,4-
naphthoquinone (HNQ-8) Aldehyde: 7-azaindol-3-
carboxyaldehyde. Reaction solvent: CH2Cl2. Column
chromatography with c-hexane/EtOAc, 2:1. Orange solid (51%).
Rf = 0.23 (c-hexane/EtOAc, 1:7).1H NMR (600 MHz, DMSO-d6):
δ 11.34 (s, 1H), 8.16 (dd, 3J = 4.7 Hz, 4J = 1.4 Hz, 1H), 8.04 (dd,
3J = 7.8 Hz, 4J = 1.2 Hz, 1H), 7.97 (dd, 3J = 7.5 Hz, 4J = 1.1 Hz,
2H); 7.80 (td, 3J = 7.4 Hz, 4J = 1.4 Hz, 1H), 7.74 (td, 3J = 7.5 Hz,
4J = 1.3 Hz, 1H), 7.23 (d, 3J = 1.8 Hz, 1H), 7.03 (dd, 3J = 7.9 Hz,
4.7 Hz, 1H), 3.9 (s, 2H) ppm. 13C NMR (150 MHz, DMSO-d6): δ
= 184.2, 181.2, 155.5, 148.3, 142.2, 134.4, 133.0, 131.9, 130.1,
126.7, 125.7, 125.6, 123.8, 122.5, 119.2, 114.8, 110.6, 18.7
ppm. ESI-MS (m/z): [M-H] + calcd for C18H12N2O3, 305.08; found
305.16.
2-Hydroxy-3-[(4-phenoxybenzyl)methyl]-1,4-naphthoquinone
(HNQ-12) Aldehyde: 3-phenoxy-benzaldhyde. Reaction solvent:
CH2Cl2. Column chromatography with c-hexane/EtOAc, 3:1.
orange solid (45%). Rf = 0.57 (c-hexane/EtOAc, 3:1). 1H NMR
(600 MHz, DMSO-d6): δ 8.05 (dd, 3J = 7.7 Hz, 4J = 1.1 Hz, 1H),
8.0 (dd, 3J = 7.5 Hz, 4J = 1.1 Hz, 1H ), 7.67 (td, 3J = 7.7 Hz, 4J =
3
1.4, 1H), 7.60 (td, J = 7.7 Hz, 4J = 1.1 Hz, 1H), 7.35 (s, 1H),
7.27 (d, 3J = 8.7 Hz, 2H), 7.21 (m, 2H), 6.98 (t, 3J = 7.3 Hz, 1H),
6.88 (dd, 3J = 8.7 Hz, 4J = 0.9 Hz, 2H), 6.82 (d, 3J = 8.5 Hz, 2H),
3.83 (s, 2H) ppm. 13C NMR (150 MHz, DMSO-d6): δ = 179.2,
176.5, 152.2, 150.4, 147.7, 129.8, 128.5, 127.8, 127.6, 125.3,
124.4, 124.2, 121.7, 121.0, 117.8, 113.7, 113.5, 23.1 ppm. ESI-
MS (m/z): [M-H] - calcd for C23H16O4, 355.1; found 355.05.
2-Hydroxy-3-[(ferreocenyl)methyl]-1,4-naphthoquinone
(HNQ-9) Aldehyde: Ferrocenecarboxaldehyde Reaction solvent:
CH2Cl2. Column chromatography with CH2Cl2/MeOH, 49:1.
Green solid (22%). Rf = 0.25 (CH2Cl2/MeOH, 49:1). 1H NMR (600
MHz, CDCl3): δ 8,02 (d, 3J = 7.5 Hz, 1H), 7.95 (d, 3J = 7.1 Hz, 1H)
7,64-7.56 (m, 2H), 7.32 (br. s, 7.32), 4,08 (m, 9H), 3,97(s, 1H),
3,57 (s, 2H) ppm. 13C NMR (150 MHz, CDCl3): δ = 184.5, 181.9,
152.7, 135.0, 133.0, 129.5, 127.0, 126,1. 123.5, 85.8, 69.1, 68.8,
Characterization data for all synthesized compounds can be
found in the Supporting Information
Solubility assay
Thermodynamic equilibrium solubility of HNQ compounds was
determined in sodium phosphate buffer containing 20 mM NaPi,
50 mM NaCl, 0,02% DDM at pH 8.0. All steps were performed at
-
67.4, 23.2 ppm. ESI-MS (m/z): [M-H] calcd for C23H22FeO3,
372.03; found 371.05.
room temperature following the method by Bharate
&
9
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201900707
ChemMedChem
FULL PAPER
Vishwakarma 2015.[36] Measurements were done in triplicates.
HNQ compounds were dissolved in a particular amount of
methanol or DCM (HNQ-9) needed for complete dissolving.
Required volumes of these methanol stock solutions were
transferred in 1.5 mL reaction tubes (0.1 – 600 µL dependent on
compound and stock concentration) to get a concentration row.
After 24 - 48h complete evaporation of methanol or DCM was
accomplished. Remaining solid compounds were dissolved by
adding 100 / 200 µL of sodium phosphate buffer to the reaction
tubes, which were shaken at 200 rpm for one day for equilibration.
Next day, centrifugation of the reaction tubes for 30 min at
13 000 rpm. 50 / 100 µL of supernatant were pipetted into a 96er
well plate (UV-Star, Greiner) and absorbance spectra were
recorded for each well with a TECAN Spark plate reader (Tecan
Trading AG, Switzerland). By plotting absorbance at λmax against
concentration [mM] of the compound, saturation point of graph
indicates the thermodynamic equilibrium solubility. Data analysis
and visualization with SigmaPlot Pro 12.5 (Systat Software GmbH,
Germany).
Production of cytochrome bd-I and bo3 from E. coli
Production of cytochrome bd-I from E. coli
Production of cytochrome bd-I oxidase from E. coli in E. coli C43
(DE3) Δbo3 cells transformed with the pet17b-cydABX-StrepII
plasmid.[46] 1 ml of 50 % Glycerol stock was added to 50 mL
LB-Carb-Kan (50 µg/ml carbenicillin; 50 mg/ml kanamycin) and
incubated at 175 rpm at 37°C for 8 h. Pre- culture was transferred
to 1 L LB-Carb-Kan to grow over night. Production culture of 2,5 L
LB-Kan was inoculated with 70 ml of overnight culture
supplemented with 0,025 mM IPTG to start basal production from
the beginning. After reaching OD600 0,7 heterologous bd-I oxidase
production was started by adding IPTG to a final concentration of
0,25 mM. Incubation at 37 °C for 4 h and lower temperature to
30 °C for another 16h. Cell harvesting was carried out by
centrifugation with Avanti J-26 XP at 4°C at 8000 x g. Cell
disruption via microfluidizer for six cycles at 80 psi in 50 mM NaPi
(pH 8,0) and 100 mM NaCl supplemented with 1 mM MgCl,
recombinant DNase I (Sigma) and proteaseinhibitor Aminoethyl-
benzene-sulfonyl fluoride (Pefabloc, Roche). Low-velocity
centrifugation at 5000 x g at 4 °C for 30 min before high-velocity
centrifugation of the supernatant at 220 000 x g at 4°C for 90 min.
Membrane pellets were resuspended in 50 mM NaPi (pH 8,0),
100 mM NaCl containing buffer and stored at -80 °C.
Oxygen reductase activity measurements
Oxygen reductase activity was measured as oxygen consumption
rate of purified protein by a OX-MR Clark-type oxygen electrode
linked to a PA 2000 picoammeter and to an ADC-216 for analog
to digital conversion. Recording of data with SensorTrace Basic
2.1 software (all Unisense, Denmark).
Streptactin purification of E. coli bd-I oxidase
Solubilization of isolated membranes in 50 mM NaPi (pH 8,0),
100 mM NaCl with 1 % n-Dodecyl β-D-maltoside (β-DDM) to the
mass ratio of 1:5 detergent:membrane protein at 4 °C for 120 min
on orbitalshaker. Removal of unsolubilized material by 70 000 x g
for 30 min. The supernatant was filtrated through a 0,2 µm syringe
filter and Avidin was added to a final concentration of 2 mg/mL.
Affinity chromatography was done via peristaltic pump with
prepacked 5 mL StrepTrap HP column (GE Healthcare)
equilibrated with 20 mM NaPi (pH 8,0), 100 mM NaCl, 0,02 %
DDM at a flow rate of 3 mL/min. Washing step was carried out
with the same buffer for 8 CV. Elution with the same buffer
containing 10 mM Desthiobiotin (IBA Lifesciences). Sample
polishing by dialysis with Slide-A-Lyzer (CutOff 10K) dialysis
casettes (Thermo Fisher Scientific) in 4 L 50 mM NaPi (pH 8,0),
100 mM NaCl with 1% n-Dodecyl β-D-maltoside (β-DDM) over
night. Verification of E. coli bd-I type oxidase and purity of the
product was realized by SDS-page and native page gel
electrophoresis, see Supporting information for results.
Measurements were performed at room temperature in stirred
2 ml - glass vials with a total reaction volume of 600 µL. Oxygen
consumption was initiated by adding 30 nM of the respective
enzyme to the equilibrated mixture containing 20 mM NaPi
(pH 8,0), 50 mM NaCl, 0,02 % DDM, 5 mM Dithiothreitol (DTT)
and 200 µM Ubiquinone-1 (2,3-Dimethoxy-5-methyl-6-(3-methyl-
2-buten-1-yl)-1,4-benzoquinone). Equilibration was done for
10 minutes prior to enzyme addition. Inhibition experiments with
250 µM of respective HNQ from 20 mM stock solution in EtOH
before equilibration. Data analysis and visualization with Origin
Lab Pro 9.5 (Additive GmbH, Germany). HQNO (2-heptyl-4-
quinolinol 1-oxide) was purchased from biomol GmbH, Hamburg.
Determination of apparent ki (kiapp) was performed under the same
conditions but with a range of inhibitor concentrations from
1000 µM to 0,2 µM. The kiapp value was adopted as the EC50 from
the sigmoidal DoseResp Fit. Further information can be found in
Supporting Information Figure S40-43.
Production of cytochrome bo3 from E. coli
The protocol for the bo3 – oxidase production from (Rumbley et al.
1997) was used. In brief strain GO195 transformed with pIRHisA
10
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201900707
ChemMedChem
FULL PAPER
[1]
[2]
[3]
J. R. Widhalm, D. Rhodes, Hortic. Res. 2016, 3, DOI
plasmid was a kindly gift from Bob Gennis Lab. The purified
membranes were a kindly gift from Hao Xie.
10.1038/hortres.2016.46.
B. Søballe, R. K. Poole, Microbiology 1999, 145, 1817–
1830.
From an overnight pre-culture in 50 ml LB-Amp-Kan (100 µg/mL
ampicillin; 50 µg/mL kanamycin) 10 mL was transferred into 2.5 L
LB medium supplemented with 3 % Lactic acid and 500 µM
CuSO4. Harvest in mid-logarithmic phase. Membranepreparation
via two French Press cycles at 20 000 psi in 50 mM KPi (pH 8,3),
5 mM MgSO4 with 4 mg/mL DNAse and Pefabloc. Followed by
15 min centrifugation at 17 000 x g and high spin centrifugation of
the supernatant at 180 000 x g for 3 h. Membranes were stored
at -80 °C prior to Ni-NTA affinity purification.
H. R. Nasiri, K. Hohmann, M. G. Hatemler, A. Plodek, F.
Bracher, H. Schwalbe, Med. Chem. Res. 2017, 26, 1170–
1175.
[4]
[5]
[6]
Y. Kumagai, Y. Shinkai, T. Miura, A. K. Cho, Annu. Rev.
Pharmacol. Toxicol. 2012, 52, 221–247.
J. L. Bolton, M. A. Trush, T. M. Penning, G. Dryhurst, T. J.
Monks, Chem. Res. Toxicol. 2000, 13, 135–160.
A. V. Pinto, S. L. De Castro, Molecules 2009, 14, 4570–
4590.
Ni-NTA purification of E. coli bo3 oxidase
[7]
[8]
[9]
O’Brien, Chem. Biol. Interact. 1991, 80, 1–41.
B. Prescott, 1969, 431, 181–182.
A slightly modified protocol to (Rumbley et al. 1997) was used for
cytochrome bo3-oxidase. Solubilization of isolated membranes in
50 mM NaPi with 1 % Triton X-100 and 1,25 % Octylglucoside at
4 °C for 1 h on orbitalshaker. Removal of unsolubilized material
by 70 000 x g for 30 min. 4 mL Ni-NTA bed were equilibrated with
50 mM NaPi (pH 8,3), 0,1 % Triton X-100 and 25 mM Imidazole
before solubilized protein was added. Washing step with 4 CV of
the equilibration buffer, followed by high salt washing step with
250 mM NaPi (pH 8,3), 0,1 % Triton X-100 and 25 mM Imidazole
for 4 CV. Third washing with 50 mM NaPi (pH 8,3), 0,1 % DDM
and 25 mM Imidazole for 4 CV to change the detergent from
Triton X-100 to DDM and lower the salt concentration. For elution
a linear gradient of imidazole up to 300 mM was used. His-tagged
bo3 oxidase eluted between 100 mM and 200 mM. The pooled
fractions were concentrated using a Amicon ultra concentrator
with 50 kDa cutoff (Millipore) and dialyzed overnight with Slide-A-
Lyzer (CutOff 10K) dialysis casettes (Thermo Fisher Scientific) in
4 L 50 mM NaPi (pH 8,0), 100 mM NaCl with 1 % n-Dodecyl β-D-
maltoside (β-DDM). Verification of E. coli bo3 oxidase and purity
of the product was realized by SDS-page and native page gel
electrophoresis, see Supporting information for results.
T. S. Lin, L. Y. Zhu, S. P. Xu, A. A. Divo, A. C. Sartorelli, J.
Med. Chem. 1991, 34, 1634–1639.
[10]
D. C. Schuck, S. B. Ferreira, L. N. Cruz, D. R. Da Rocha,
M. S. Moraes, M. Nakabashi, P. J. Rosenthal, V. F.
Ferreira, C. R. S. Garcia, Malar. J. 2013, 12, 1–6.
K. W. Wellington, RSC Adv. 2015, 5, 20309–20338.
A. Kacmaz, N. G. Deniz, S. G. Aydinli, C. Sayil, E. Onay-
Ucar, E. Mertoglu, N. Arda, Open Chem. 2019, 17, 337–
345.
[11]
[12]
[13]
[14]
and P.-K. L. Bhasin D., Somsundaram N. Chettiar,
Jonathan P. Etter, May Mok, Bioorganic Med. Chem. 2013,
21, 1–22.
H. R. Nasiri, M. G. Madej, R. Panisch, M. Lafontaine, J. W.
Bats, C. R. D. Lancaster, H. Schwalbe, J. Med. Chem.
2013, 56, 9530–9541.
[15]
[16]
[17]
V. K. Tandon, H. K. Maurya, M. K. Verma, R. Kumar, P. K.
Shukla, ChemInform 2010, 41, no-no.
T. Hayashi, F. T. Smith, K. Lee, J.Med.Chem 1987, 11,
2005–2008.
V. C. Niculescu, N. Muresan, A. Salageanu, C. Tucureanu,
G. Marinescu, L. Chirigiu, C. Lepadatu, J. Organomet.
Chem. 2012, 700, 13–19.
[18]
[19]
[20]
[21]
K. Kita, K. Konishi, Y. Anraku, J. Biol. Chem. 1984, 259,
3368–3374.
Molecular docking procedure
The optimization of all the geometries of scaffolds was performed
with Gaussian 09 semi-empirical PM3 force-field method (see
Supporting Information Figure S44).[47] The crystal structures of
PDB (RSCB PDB: 1FFT and 5DOQ) was obtained from the RSCB
protein data bank.[21] Docking studies were performed using
Autodock 4.2 software.[48] The visualization and analysis of
interactions were performed using Pymol, version 0.99.[49]
K. C. Minghetti, R. B. Gennis, Bichemical Biophys. Res.
Commun. 1988, 243–248.
J. and V. Borisov, V.B., Gennis, R.B., Hemp, Biochim.
Biophys. Acta 2011, 11, 1398–1413.
J. Abramson, S. Riistama, G. Larsson, A. Jasaitis, M.
Svensson-Ek, L. Laakkonen, A. Puustinen, S. Iwata, M.
Wikström, Nat. Struct. Biol. 2000, 7, 910–917.
L. J. Lemieux, M. W. Calhoun, J. W. Thomas, W. J.
Ingledew, R. B. Gennis, J. Biol. Chem. 1992, 267, 2105–
2113.
[22]
[23]
D. B. Ramachary, C. F. Barbas, Chem. - A Eur. J. 2004,
11
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201900707
ChemMedChem
FULL PAPER
DOI 10.1002/chem.200400597.
[46]
[47]
J. Hoeser, S. Hong, G. Gehmann, R. B. Gennis, T.
[24]
V. Chepuri, L. Lemieux, D. C. T. Au, R. B. Gennis, J. Biol.
Chem. 1990, 265, 11185–11192.
Friedrich, FEBS Lett. 2014, 588, 1537–1541.
M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B.
Mennucci, G. A. Petersson, et al., Gaussian 09, Revis.
B.01, Gaussian, Inc., Wallingford CT 2009.
[25]
[26]
I. G. Y. B.J. Wallace, Growth (Lakeland) 1977, 461, 75–83.
G. B. Cox, N. A. Newton, F. Gibson, A. M. Snoswell, J. A.
Hamilton, Biochem. J. 1970, 117, 551–562.
Z. Ding, C. Sun, S. M. Yi, R. B. Gennis, S. A. Dikanov,
Biochemistry 2019, 58, 4559–4569.
[27]
[28]
[48]
[49]
W. Forli, S. Halliday, R. Belew, A. Olson, Citeseer 2012.
L. Schrödinger, Thomas Hold. 2015, DOI 10.1007/s13398-
014-0173-7.2.
M. Bekker, G. Kramer, A. F. Hartog, M. J. Wagner, C. G. de
Korster, K. J. Hellingwerf, M. J. Teixeira de Mattos,
Microbiology 2007, 153, 1974–1980.
[29]
P. Sharma, M. J. Teixeira De Mattos, K. J. Hellingwerf, M.
Bekker, FEBS J. 2012, 279, 3364–3373.
Acknowledgment
[30]
[31]
A. Nitzschke, K. Bettenbrock, PLoS One 2018, 13, 1–18.
B. J. Wallace, I. G. Young, BBA - Bioenerg. 1977, 461, 84–
100.
We thank the students: Sayaka Nagasaki, Vanessa Bause,
Giuseppe Faudone, Fabian Gohl, Patrick P. Rolo for their help in
this project. The Work was supported by DFG in graduate college:
CLIC, by EU in ITN: EUROPOL and by state of Hesse for BMRZ.
[32]
J. Rendon, E. Pilet, Z. Fahs, F. Seduk, L. Sylvi, M. H.
Chehade, F. Pierrel, B. Guigliarelli, A. Magalon, S.
Grimaldi, Biochim. Biophys. Acta - Bioenerg. 2015, 1847,
739–747.
Keywords: Inhibitor design, membrane proteins, respiratory
chain, cytochrome bo3, bd.
[33]
[34]
[35]
[36]
[37]
D. B. Ramachary, M. Kishor, Org. Biomol. Chem. 2010,
DOI 10.1039/c003588b.
D. B. Ramachary, S. Jain, Org. Biomol. Chem. 2011, DOI
10.1039/c0ob00611d.
E. E. Kim, E. O. Onyango, L. Fu, G. W. Gribble,
Tetrahedron Lett. 2015, 56, 6707–6710.
S. S. Bharate, R. A. Vishwakarma, Bioorganic Med. Chem.
Lett. 2015, 25, 1561–1567.
L. L. Yap, M. T. Lin, H. Ouyang, R. I. Samoilova, S. A.
Dikanov, R. B. Gennis, Biochim. Biophys. Acta - Bioenerg.
2010, 1797, 1924–1932.
[38]
M. Sato-Watanabe, T. Mogi, H. Miyoshi, H. Iwamura, K.
Matsushita, O. Adachi, Y. Anraku, J. Biol. Chem. 1994,
269, 28899–28907.
[39]
[40]
[41]
[42]
[43]
[44]
S. M. Musser, M. H. B. Stowell, Hung Kay Lee, J. N.
Rumbley, S. I. Chan, Biochemistry 1997, 36, 894–902.
S. K. Choi, L. Schurig-Briccio, Z. Ding, S. Hong, C. Sun, R.
B. Gennis, J. Am. Chem. Soc. 2017, 139, 8346–8354.
H. Miyoshi, K. Takegami, K. Sakamoto, T. Mogi, H.
Iwamura, J. Biochem. 1999, 125, 138–142.
R. Welter, L. Q. Gu, L. Yu, C. A. Yu, J. Rumbley, R. B.
Gennis, J. Biol. Chem. 1994, 269, 28834–28838.
M. Sato-Watanabe, T. Mogi, H. Miyoshi, Y. Anraku,
Biochemistry 1998, 37, 5356–5361.
S. Safarian, C. Rajendran, H. Müller, J. Preu, J. D. Langer,
S. Ovchinnikov, T. Hirose, T. Kusumoto, J. Sakamoto, H.
Michel, Science (80-. ). 2016, 352, 583–586.
A. Harikishore, S. S. M. Chong, P. Ragunathan, R. W.
Bates, G. Grüber, Mol. Divers. 2020, DOI 10.1007/s11030-
020-10034-0.
[45]
12
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201900707
ChemMedChem
FULL PAPER
Entry for the Table of Contents
13
This article is protected by copyright. All rights reserved.