A fragment-based approach towards the discovery of N-substituted tropinones as inhibitors of Mycobacterium tuberculosis transcriptional regulator EthR2

Article abstract of DOI:10.1016/j.ejmech.2019.02.023

Tuberculosis (TB) caused by the pathogen Mycobacterium tuberculosis, represents one of the most challenging threat to public health worldwide, and with the increasing resistance to approved TB drugs, it is needed to develop new strategies to address this

Full text of DOI:10.1016/j.ejmech.2019.02.023

Accepted Manuscript
A fragment-based approach towards the discovery of N-substituted tropinones as
inhibitors of Mycobacterium tuberculosis transcriptional regulator EthR2
Hugues Prevet, Martin Moune, Abdalkarim Tanina, Christian Kemmer, Adrien
Herledan, Rosangela Frita, Alexandre Wohlkönig, Marilyne Bourotte, Baptiste
Villemagne, Florence Leroux, Marc Gitzinger, Alain R. Baulard, Benoit Déprez, René
Wintjens, Nicolas Willand, Marion Flipo
PII:
S0223-5234(19)30137-0
DOI:
https://doi.org/10.1016/j.ejmech.2019.02.023
EJMECH 11115
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 19 December 2018
Revised Date: 4 February 2019
Accepted Date: 7 February 2019
Please cite this article as: H. Prevet, M. Moune, A. Tanina, C. Kemmer, A. Herledan, R. Frita, A.
Wohlkönig, M. Bourotte, B. Villemagne, F. Leroux, M. Gitzinger, A.R. Baulard, B. Déprez, René.
Wintjens, N. Willand, M. Flipo, A fragment-based approach towards the discovery of N-substituted
tropinones as inhibitors of Mycobacterium tuberculosis transcriptional regulator EthR2, European
Journal of Medicinal Chemistry (2019), doi: https://doi.org/10.1016/j.ejmech.2019.02.023.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
A fragment-based approach towards the discovery of N-substituted tropinones as inhibitors of
Mycobacterium tuberculosis transcriptional regulator EthR2.
Hugues Prevet1, Martin Moune2, Abdalkarim Tanina3, Christian Kemmer4, Adrien Herledan1,
Rosangela Frita2, Alexandre Wohlkönig5, Marilyne Bourotte4, Baptiste Villemagne1, Florence Leroux1,
Marc Gitzinger4, Alain R. Baulard2, Benoit Déprez1, René Wintjens3, Nicolas Willand*,#,1, Marion
Flipo#,1
# These authors contributed equally to this work
*nicolas.willand@univ-lille.fr
1. Univ. Lille, Inserm, Institut Pasteur de Lille, U1177 - Drugs and Molecules for living Systems,
F-59000 Lille, France
2. Institut Pasteur de Lille, Univ. Lille, CNRS, Inserm, CHU Lille, U1019-UMR8204-CIIL-Centre
d'Infection et d'Immunité de Lille, F-59000 Lille, France
3. Unité Microbiologie, Bioorganique et Macromoléculaire (CP206/04), département R3D,
Faculté de Pharmacie, Université Libre de Bruxelles, B-1050 Brussels, Belgium
4. Bioversys AG, Hochbergerstrasse 60C, 4057 Basel, Switzerland
5. Structural Biology Brussels, Vlaams Instituut voor Biotechnology (VIB), B-1050 Brussels,
Belgium.
Declarations of interest: none
Abstract
Tuberculosis (TB) caused by the pathogen Mycobacterium tuberculosis, represents one of the most
challenging threat to public health worldwide, and with the increasing resistance to approved TB
drugs, it is needed to develop new strategies to address this issue. Ethionamide is one of the most
widely used drugs for the treatment of multidrug-resistant TB. It is a prodrug that requires activation
by mycobacterial monooxygenases to inhibit the enoyl-ACP reductase InhA, which is involved in
mycolic acid biosynthesis. Very recently, we identified that inhibition of a transcriptional repressor,
termed EthR2, derepresses a new bioactivation pathway that results in the boosting of ethionamide
activation. Herein, we describe the identification of potent EthR2 inhibitors using fragment-based
screening and structure-based optimization. A target-based screening of a fragment library using
thermal shift assay followed by X-ray crystallography identified 5 hits. Rapid optimization of the
tropinone chemical series led to compounds with improved in vitro potency.
Keywords
ethionamide, EthR2, fragment-based drug design, tropinone, tuberculosis
Abbreviations
cyHex, cyclohexane; DCE, 1,2-dichloroethane; DCM, dichloromethane; DMAP, 4-
dimethylaminopyridine; DMF, dimethylformamide; DMSO, dimethylsulfoxide; DRC, dose-response
curve; ETH, ethionamide; EtOH, ethanol; EtOAc, ethyl acetate; HAC, heavy atoms count; MeCN,
acetonitrile; MeOH, methanol; Mtb, Mycobacterium tuberculosis; PTSA, p-toluenesulfonic acid; SAR,

ACCEPTED MANUSCRIPT
structure-activity relationships; TB, tuberculosis; tBuOK, potassium tert-butoxide; TEA, triethylamine;
TLC, thin layer chromatography; Tm, melting temperature; TSA, thermal shift assay.

ACCEPTED MANUSCRIPT
1. Introduction
Tuberculosis (TB), a disease caused by Mycobacterium tuberculosis (Mtb), is the leading cause of
death worldwide from a single infectious agent. According to the WHO, 10.0 million people
developed tuberculosis in 2017, and 1.3 million HIV-negative people died from TB the same year.[1]
Ethionamide (ETH) is a second-line antitubercular drug that is bioactivated by the monooxygenase
EthA. EthR, a mycobacterial transcriptional regulator, represses the expression of EthA and thus
downregulates ETH bioactivation. A rational design based on the X-ray structure of EthR led to the
identification of BDM31343, the first EthR inhibitor able to boost three times ETH activity in mice.[2]
The optimization of this compound led to the discovery of BDM41906 that boosts 10 times ETH
activity on Mtb infected macrophages in the low nanomolar range (EC50 = 60 nM).[3, 4] BDM41906
also showed boosting activity in a mouse model of Mtb infection.[5] In parallel, Tatum et al.
described the identification of new series of EthR inhibitors using in silico structure-based
screenings.[6]
In recent years, fragment-based approaches have emerged as an alternative approach to high-
throughput screening to identify inhibitors of Mtb targets.[7, 8] These approaches rely on the
screening of small libraries of molecules (~1000 compounds) with low molecular weight (< 300 Da)
and high solubility. Usually, although hits identified by this approach display relatively weak biological
activities, they are key to reveal compound-target interactions. Fragments can then be optimized
using growing, merging or linking strategies to obtain lead compounds with higher affinity for the
target. For the past five years, fragment-based approaches have been used successfully to discover
and optimize EthR inhibitors.[9-13]
More recently, we have demonstrated that drug resistance to ETH due to mutations in the known
bioactivation pathway can be circumvented by derepressing an alternative bioactivation pathway.
This can be achieved by the use of a drug-like molecule (SMARt-420) that targets a new
transcriptional repressor, termed EthR2, allowing ETH bioactivation through this new pathway. This
allowed to restore ETH activity on drug-resistant strains but also to increase ETH activity on drug-
sensitive strains.[14] The X-ray structure of EthR2, liganded with SMARt-420, revealed that the
protein belongs to the TetR-type family of repressors. It is an all-α helix protein with a DNA-binding
domain at the N-terminal region, showing a Helix-Turn-Helix motif. The C-terminal region presents a
dimerization domain and a ligand-binding domain. EthR2 (PDB ID: 5N7O [15]) shares only 14% of
sequence identity with EthR (PDB ID: 1U9N), which translates in a rmsd of 3.4 Å over 184
superimposed Cα atoms.
In the search for novel chemotypes of EthR2 inhibitors, we initiated a fragment-based approach
combining fragment screening, co-crystallization of the hits and chemical modulation of the most
potent chemical family. A thermal shift assay was used as a primary assay to identify compounds able
to bind to EthR2. Thermal shift assay has been used successfully for several years in drug discovery to
identify protein ligands in a high-throughput mode.[16, 17] The principle of this assay is based on
monitoring the thermal unfolding of a protein using a hydrophobicity sensitive dye that will undergo
a significant increase in quantum yield when bound to the hydrophobic regions of the unfolded
protein. Ligands that contribute to the stabilization of a protein in a holo conformation give rise to
the increase of the melting temperature of the protein. This variation of Tm (noted ΔTm) is directly
correlated to the affinity of the ligand for the protein. A mammalian assay, previously developed to
characterize the mode of action of SMARt-420, was further used as a functional secondary assay to
measure the potency of the new inhibitors to disrupt the interaction between the transcriptional
repressor EthR2 and its DNA operator. This synthetic mammalian gene circuit uses a chimeric

ACCEPTED MANUSCRIPT
transcription factor, composed of the bacterial transcriptional regulator EthR2 as DNA binding
domain that is linked to the virus protein 16 (VP16) activation domain, to monitor transcriptional
changes in mammalian cells.[14] Monitoring of the activity of the reporter human placental secreted
alkaline phosphatase allows to discriminate if EthR2 can bind to its cognate DNA-operator or if small-
molecules may alter or inactivate EthR2 DNA-binding ultimately leading to deactivation of reporter
gene expression.
2. Results and discussion
2.1 Screening and X-ray crystallography
An in-house 960-fragment library was screened on the transcriptional regulator EthR2 using a
dedicated thermal shift assay (Figure 1). This fluorescence-based assay was initially developed to
successfully detect compounds able to bind to EthR and was then adapted to EthR2.[14, 18] Each
member of the library was screened at 1 mM and 50 fragments displaying thermal stabilization
(ΔTm) greater than 2 °C were selected for dose-response experiments. Among them, 12 fragments
were able to stabilize the transcriptional regulator EthR2 in a dose-dependent manner, giving a hit-
rate of 1.25%. These 12 compounds were selected for co-crystallization experiments with EthR2 and
the X-ray crystallographic structures of 5 fragment hits (compounds 1-5, Table 1) bound to EthR2
were obtained.
Figure 1. Screening of a 960-fragment library on EthR2 leading to the discovery of the tropinone series.
X-ray diffraction revealed that the 5 molecules bind to the same pocket as SMARt-420. Compound 1
occupies at full occupancy only one of the two monomers of EthR2. The tropinone scaffold is facing
the polar area formed by Asp168 and Glu70 with the carbonyl group pointing towards Glu70 and the
tertiary amine being H-bonded to Asp168. The bicyclic aliphatic moiety is in contact with Trp100,
Ala130 and Ile113. The chlorophenethyl group substituting the nitrogen atom points towards the
hydrophobic portal of the ligand binding domain, making close contacts with Leu167 and Met77, and
more specifically, a T-shaped π-π stacking occurs between the phenyl ring and Trp100. Compound 2
binds identically the two monomers. The exocyclic nitrogen is H-bonded to Glu70 and the
heteroaromatic imidazole ring is engaged in a π-stacking interaction with Trp100. Finally, the
isopropyl sulfide substituent occupies the hydrophobic pocket lined by Met74, Met77, Trp100 and
Leu167. Compound 3 binds the two monomers in slightly different ways. In both cases the pyridine
ring is π-stacked to Trp100 and the urea function is H-bonded to Glu70. However, the propyl
substituent binds two different hydrophobic subpockets, one delimited by Leu114, Met172, Trp123
and Phe126, one delimited by Ile113, Phe126 and Val129. Compound 4 also adopts a different
binding mode, as two molecules bind only one monomer. The guanidine functions of the two distinct

ACCEPTED MANUSCRIPT
fragments make ionic interactions with Asp168 and Glu70. Ser134 is H-bonding the closest molecular
entity. The two benzothiazole rings are facing the hydrophobic surface area formed by Ile164,
Leu167, Trp100, Met77, Met137 and Val73. Finally compound 5 occupies both monomers of EthR2 in
the upper part of the ligand binding domain. Indeed, the piperazine ring is H-bonding Asp168 and
Ser134, while facing Trp100 and the two methionine residues (Met77 and Met137). The meta-
chlorobenzyl group is sandwiched between Leu167 and Leu141.
Figure 2: A/ X-ray structure representation of EthR2 as a dimer. The ligand-binding domain is highlighted in black for each
monomer. B/ X-ray structure representations of the ligand-binding pocket of EthR2 filled with compound 1 (PDB ID
6HRW), compound 2 (PDB ID 6HRX), compound 3 (PDB ID 6HRY), compound 4 (PDB ID 6HRZ), compound 5 (PDB ID 6HS0).
Surface of the ligand binding domain is highlighted. Hydrogen bonds with Asn179 and Asn176 are represented with
dotted lines. Colors legend: magenta (compound) or dark green (EthR2) = carbon, dark blue = nitrogen, red = oxygen,
yellow = sulfur, light green = chlorine. Images were generated with Pymol.
The ability of these 5 fragments to disrupt the interaction between EthR2 and its DNA-operator was
assessed using a EthR2-based reporter gene assay hosted in mammalian cells.[14] In this assay,
compounds 3 and 4 were inactive at the highest tested concentration (150 µM) and compound 1
(pIC50 = 6.0) was found to be more potent than compounds 2 and 5 (pIC50 = 5.1 and 5.5 respectively).
In addition, compound 1 exhibited the peculiarity of presenting a less flat structure given the
presence of the tropinone core, which can improve its physicochemical properties, especially the
solubility by avoiding stacking. Therefore, compound 1 was selected for further optimization. To
check the selectivity of this compound for EthR2 vs EthR, compound 1 was tested in a mammalian
reporter assay using EthR [19] and was totally inactive at the highest concentration tested (30 µM).

ACCEPTED MANUSCRIPT
ΔTm (°C)
b
ID structure cpd Structure
screening
conditionsa
3.4
pIC50
BDM_14272
BDM_72201
1
2
6.0
2.5
2.2
5.1
< 4
BDM_72719
3
BDM_72170
BDM_71847
4
5
2.8
3.1
< 4
5.5
Table 1. Biological activities of compounds 1-5. aΔTm data were obtained using thermal shift assay with compounds
tested at 1 mM. bpIC50 = -log(IC50); IC50 represents the concentration of compound that allows the inhibition of 50% of
the interaction between EthR2 and its promotor using a reporter gene assay in mammalian cells.
2.2 Chemistry
A direct synthetic route to obtain N-substituted nortropinone derivatives was developed previously
in our lab and we adapted this synthesis to explore structure-activity relationships (SAR) around
compound 1.[20]
2.2.1. Modifications of the tropinone core
First, the importance of the ethylene bridge on the tropinone scaffold was evaluated by either
removing it (piperidone analogue 7) or extending it by homologation with the introduction of a
pseudopelletierine ring. Compounds 7 and 9 were synthesized in two steps according to a procedure
adapted from the literature.[20] Commercially available N-methyl-4-piperidone or pseudopelletierine
were reacted with iodomethane in acetone. The tertiary ammonium iodide intermediates 6 and 8
were then refluxed with 4-chlorophenethylamine in a 1/1 mixture of water and ethanol, using
potassium carbonate as base (Scheme 1). These two steps reactions afforded compounds 7 and 9 in
good yields.
a
b
57%
80%
6
7
a
b
69%
54%
9
8
Scheme 1. Synthesis of piperidone and pseudopelletierine derivatives 7 and 9.
Reagents and conditions: (a) CH3I (1.1 eq), acetone, rt, 1 h, (b) 4-Cl-C6H4-(CH2)2-NH2 (1 eq), K2CO3 (2.1 eq), EtOH/H2O (1/1),
reflux, 4 h.
We also tested the introduction of the ethylene bridge in α-positions of the carbonyl group.
Compound 11 bearing a 3-azabicyclo[3.2.1]octan-8-one ring and compound 12 bearing a 3-

ACCEPTED MANUSCRIPT
azabicyclo[3.3.1]nonan-9-one were therefore synthesized. These two compounds were obtained in
two steps according to Scheme 2. 4-chlorophenethylamine reacted with paraformaldehyde in
anhydrous MeOH under basic conditions to give compound 10 that reacted with cyclopentanone or
cyclohexanone using methyltrichlorosilane as activator to yield to compounds 11 and 12 respectively
[21, 22].
( )n
b
a
( )n
90%
n = 0
n = 1
11
12
10
Scheme 2. Synthesis of analogues 11 and 12.
Reagents and conditions: (a) HCHO (3.25 eq), K2CO3 (1 eq), anhydrous MeOH, rt, 48 h, (b) CH3SiCl3 (2 eq), anhydrous
MeCN, rt, 3-8 h.
2.2.2. Modifications of the ethylene linker
To evaluate the importance of the chain length between the nitrogen atom and the phenyl ring, the
phenethyl group was replaced by a benzyl group. Compound 14 was synthesized in two steps as
described previously from commercially available tropinone (Scheme 3). 8,8-dimethyl-3-oxo-8-
azonia-bicyclo[3.2.1]octane iodide (IDABO, compound 13) was prepared by reacting iodomethane
with tropinone in acetone. IDABO was then refluxed with 4-chlorobenzylamine under basic
conditions to give the corresponding N-alkylated nortropinone derivative 14.
a
b
92%
48%
13
14
Scheme 3. Synthesis of analogue 14.
Reagents and conditions: (a) CH3I (1.1 eq), acetone, rt, 1 h, (b) 4-Cl-C6H4-CH2-NH2 (1 eq), K2CO3 (2.1 eq), EtOH/H2O (1/1),
reflux, 4 h.
The introduction of a methyl (compound 15) to evaluate the impact of the steric hindrance in
position alpha of the tertiary amine on the activity was carried out by reacting 1-(4-
chlorophenyl)propan-2-one with nortropinone hydrochloride under reducing conditions (Scheme 4)
[23].
a
68%
15
Scheme 4. Synthesis of analogue 15.
Reagents and conditions: (a) TEA (1.1 eq), NaHB(OAc)3 (1.4 eq), AcOH (1 eq), molecular sieves, DCE, rt, 4 h.
An amide function was introduced to evaluate the importance of the basic nitrogen for the activity.
Compound 17 was synthesized in two steps by activating 4-chlorophenylacetic acid using thionyl

ACCEPTED MANUSCRIPT
chloride and benzotriazole. Compound 16 was then coupled with nortropinone under microwave
irradiation (Scheme 5).
a
b
74 %
81%
17
16
Scheme 5. Synthesis of analogue 17.
Reagents and conditions: (a) SOCl2 (1 eq), benzotriazole (3 eq), DCM, rt, 24 h, (b) Nortropinone hydrochloride (1.1 eq),
DMAP (1.1 eq), CH3Cl, µW 100 °C, 15 min.
2.2.3. Replacement of the ketone function
The ketone function was replaced by bioisosteres as gem-difluoroolefine, cyclic and acyclic acetals.
Compounds 18-20 were synthesized in one step from compound 1. Compound 18 was obtained by
reaction of compound 1 with 2-difluoromethylsulfonylpyridine using potassium tert-butoxide as base
[24]. The cyclic acetal 19 was synthesized under microwave irradiation using ethane-1,2-diol and
trimethylsilyl chloride in MeCN [25]. The reaction of compound 1 with trimethylorthoformate and p-
toluenesulfonic acid in MeOH led to acyclic acetal 20.
18
a
30%
b
83%
1
19
c
60%
20
Scheme 6. Synthesis of tropinone derivatives 18-20.
Reagents and conditions: (a) i. t-BuOK (2.2 eq), 2-difluoromethylsulfonylpyridine (1 eq), DMF anhydrous, -25 °C, 96 h, ii.
HCl 3M, (b) ethanediol (2.2 eq), Me3SiCl (4.4 eq), MeCN anhydrous, µW 80 °C, 25 min, (c) PTSA (2 eq),
trimethylorthoformate (5 eq), MeOH anhydrous, µW 80 °C, 25 min.
2.2.4. Modifications of the 4-chlorophenyl group
In a last step, we investigated the replacement of the 4-chlorophenyl group. Compounds 21−32 were
obtained by refluxing IDABO (compound 13) with substituted phenethylamines under basic
conditions to give the corresponding N-alkylated nortropinone derivatives (Scheme 7).[20]
a
21 - 32
13
Scheme 7. Synthesis of N-alkylated nortropinones 21-32.
Reagents and conditions: (a) Ar-(CH2)2-NH2 (1 eq), K2CO3 (2.1 eq), EtOH/H2O (1/1), reflux, 4 h.

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
2.3 Biological results
In this study, we explored the SAR by modifying the tropinone core (compounds 7, 9, 11, 12), the
ethylene linker (14, 15, 17), the ketone function (compounds 18-20) or the 4-chlorophenyl group
(compounds 21-32). All compounds were tested by TSA to evaluate their affinity for the
mycobacterial transcriptional regulator EthR2. To avoid possible problems of solubility during the
biological evaluation of analogues, compounds were tested at a lower concentration (20 µM) than
previously, using the protocol described for the evaluation of the SMARt-420 series.[14] For most
compounds displaying an affinity for EthR2 (ΔTm > 0.1 °C) we measured their potency to inhibit
EthR2/DNA interaction using the EthR2-based reporter gene assay hosted in mammalian cells.
2.3.1. Modifications of the tropinone core
First, we investigated the modification of the tropinone core by removing the ethylene bridge, by
extending it by homologation or by moving it in α-positions of the carbonyl group (Table 2). The
replacement of the tropinone core (compound 1) by a piperidone ring (compound 7) resulted in a
decrease in affinity for the protein EthR2 (ΔTm = 0.3 °C). On the contrary, the affinity and the
potency are preserved when the pseudopelletierine core (compound 9) is introduced. These results
suggest that an ethylene or propylene bridge connecting the two carbon atoms in β position of the
carbonyl function is necessary to improve the affinity. Moreover, the position of this bridge is
important, since a loss of affinity is observed when the bridge is introduced at α positions of the
carbonyl function regardless of its size (compound 11, ΔTm = 0.1 °C and compound 12, ΔTm < 0 °C).
ΔTm (°C)
b
Cpd
1
R1
pIC50
SAR conditionsa
0.6
6.0
7
9
0.3
0.6
ND
5.8
11
12
0.1
ND
ND
-0.1
Table 2. Biological activities of compounds 1, 7, 9, 11 and 12.
a compound tested at 20 µM; bpIC50 = -log(IC50); ND: not determined
2.3.2. Modifications of the ethylene linker
In a second step, the ethylene linker was modified (Table 3). These modifications had a deleterious
effect on affinity. Indeed, the reduction of the chain length (compound 14, ΔTm = 0.0 °C) was not

ACCEPTED MANUSCRIPT
tolerated and the introduction of a methyl group in α position of the nitrogen atom (compound 15)
led to a compound with a very low affinity for the protein (ΔTm = 0.1 °C) suggesting that steric
hindrance is not tolerated in this position. The introduction of an amide function (compound 17) was
also detrimental to affinity (ΔTm < 0 ° C). This result shows the importance of the basicity of the
amine function and can be rationalized by looking at the co-crystallographic structure between 1 and
EthR2 (figure 2) in which the nitrogen atom establishes a H-bond with Asp168 side chain.
ΔTm (°C)
Cpd
14
R2
SAR conditionsa
0.0
15
17
0.1
-0.1
Table 3. Biological activities of compounds 14, 15, 17.
a compound tested at 20 µM
2.3.3. Modifications of the ketone function
The ketone function was replaced by a difluoroethylene function as isoster (compound 18) [26]. This
led to a compound with the same affinity for the protein (ΔTm = 0.6 °C) but surprisingly inactive in
the cellular assay. The replacement of the ketone by acetal groups (compounds 19 and 20) was not
tolerated (Table 4). This result is consistent with the binding mode of compound 1 to EthR2 (Figure 2)
where the ketone is facing Glu70. Therefore, its replacement by acetal functions may impair the
binding due to steric hindrance.
ΔTm (°C)
b
Cpd
1
R1, R2
pIC50
SAR conditionsa
O
0.6
6.0
F
F
18
19
0.6
0.1
< 4.8
< 4.8
O
O

ACCEPTED MANUSCRIPT
O
20
0.2
< 4.8
O
Table 4. Biological activities of compounds 18-20.
a compound tested at 20 µM; bpIC50 = -log(IC50); ND: not determined
2.3.4. Modifications of the 4-chlorophenyl group
To further explore the SAR, the 4-chlorophenyl group was replaced by aromatic rings substituted in
different positions with electron donating or withdrawing groups (Table 5).
ΔTm (°C)
b
Cpd
1
R3
pIC50
SAR conditionsa
0.6
6.0
21
22
0.1
1.0
5.0
6.0
23
0.2
5.7
24
25
26
27
28
0.0
0.6
0.3
0.4
1.0
ND
5.8
ND
5.6
6.1
29
30
31
-0.2
1.8
ND
6.7
ND
-0.1

ACCEPTED MANUSCRIPT
32
0.3
ND
Table 5. Biological activities of compound 1 and compounds 21-32.
a compound tested at 20 µM; bpIC50 = -log(IC50); ND: not determined
Replacing the 4-chlorophenyl group with an unsubstituted phenyl group (compound 21) led to a
much less active compound (ΔTm = 0.1 °C, pIC50 = 5.0). The introduction of a second chlorine atom at
position 3 on the phenyl ring (compound 22) led to a compound slightly more active than the
reference compound 1 (ΔTm = 1.0 °C, pIC50 = 6.0). In contrast, the introduction of a second chlorine
atom at position 2 (compound 23) decreased the biological activities (ΔTm = 0.2 °C, pIC50 = 5.7). This
result suggests that position 2 of the phenyl ring may be sensitive to steric hindrance. The
replacement of the 4-chlorophenyl group by a pyridine (compound 24) is not tolerated (ΔTm = 0.0
°C). Substitution of the phenyl ring with an electron-donating group such as methoxy (compound 26)
reduced affinity (ΔTm = 0.3 °C), whereas introduction of a methyl group (compound 25) led to a
compound as active as reference compound 1 (ΔTm = 0.6 °C, pIC50 = 5.8). The replacement of the
chlorine atom by a smaller fluorine atom (compound 27) led to a less active compound (ΔTm = 0.4 °C,
pIC50 = 5.6) while the introduction of a larger bromine atom (compound 28) increased the activity
(ΔTm = 1.0 °C, pIC50 = 6.1) showing that a bulky atom is tolerated in this position. The introduction of
a trifluoromethyl group in position 4 of the phenyl ring (compound 30) led to a compound having a
better affinity for EthR2 as well as a better potency than the compound 1 (ΔTm = 1.8 °C, pIC50 = 6.7).
This result suggests that steric hindrance or hydrophobicity at the para position of the phenyl ring are
important criteria for activity. Substitution of the phenyl ring in para position with a nitro group
(compound 29) strongly affected affinity (ΔTm < 0 °C), which suggests that an electron withdrawing
group on the aromatic ring is not the only parameter necessary for the affinity of the compounds.
Substitutions of the phenyl ring in ortho and meta positions with a trifluoromethyl group
(compounds 32 and 31) strongly reduced affinity (ΔTm = 0.3 °C and ΔTm < 0 °C respectively),
confirming the steric hindrance issue aforementioned with compound 23.
For compounds evaluated in both biological assays, we plotted the potency (pIC50) as a function of
the affinity (ΔTm). Interestingly, we observed a strong correlation (R² = 0.87) between pIC50 and ΔTm
(Figure 3) ascertaining that the potency of the compounds to disrupt the interaction between EthR2
and DNA is linked to their binding to EthR2.

ACCEPTED MANUSCRIPT
7
6
5
cpd 30
R2 = 0.87
cpd 28
cpd 22
cpd 1
cpd 9
cpd 25
cpd 23
cpd 27
cpd 21
0.0
0.5
1.0
Tm (°C)
1.5
2.0
Figure 2. pIC50 versus ΔTm
In order to measure the capacity of the compounds to boost ten times the activity of ETH, nine
compounds (1, 9, 21, 22, 24, 25, 27, 28, 30) were tested on M. tuberculosis with subactive doses of
ETH (MIC/10 = 0.1 µg/mL). Unfortunately, none of the compounds was able to boost ETH
bioactivation at concentrations up to 30 µM through the EthR2/EthA2 pathway. This may be due to a
lack of bacterial membrane permeability.
2.4. Co-crystallization
Compounds 9 and 30 were successfully co-crystallized with the repressor EthR2 (Refinement
statistics listed in Table 6). The two molecules have similar binding mode that is also similar to the
one of compound 1. Both molecules are found at full occupancy only in one of the two monomers of
EthR2, and their tertiary amines are H-bonded to Asp168 side chain. Similarly to compound 1, the
aromatic ring is in contact with Ser134, Ile164, Leu167, Met177, Trp100 and Thr138 and the bicyclic
moiety is in contact with Ala130, Ile113, Phe126, Trp100, Met74 and Val73.

ACCEPTED MANUSCRIPT
Figure 4: (A) X-ray structure representations of the ligand-binding pocket of EthR2 filled with compound 9 (PDB ID 6HS1).
(B) X-ray structure representations of the ligand-binding pocket of EthR2 filled with compound 30 (PDB ID 6HS2). Surface
of the ligand binding domain is highlighted. Hydrogen bonds with Asn179 and Asn176 are represented with dotted lines.
Colors legend: magenta (compound) or dark green (EthR2) = carbon, dark blue = nitrogen, red = oxygen, yellow = sulfur,
light green = chlorine, light blue = fluorine. Images were generated with Pymol.
3. Conclusion
Through a fragment-based approach, we first identified five new chemical series as ligands of the
mycobacterial transcriptional regulator EthR2. By combining structure-based design from the protein
−ligand complexes obtained by co-crystallization, with in vitro assays (target-based and functional),
we were able to study structure-activity relationships and delineate the key molecular features that
are essential for the molecular recognition of the analogues in the tropinone series by EthR2.
Compounds 9, 22 and 28 were equipotent as the hit, and the trifluoromethyl derivative (compound
30) was more potent at disrupting the interaction between the transcriptional repressor and its
targeted promoter (pIC50 = 6.7). Interestingly, this compound displayed a high ligand efficiency index
(LE = 0.44 = (1.37*pIC50)/(HAC); HAC: Heavy Atoms Count). Optimization of the physico-chemical
properties in this series will be pursued in order to find compounds capable of boosting ethionamide
activity in M. tuberculosis infected macrophages.
Acknowledgements
We thank Hélène Cristini, Guillaume Blanc and Birgit Schellhorn for technical assistance. This work
was supported by l’Agence Nationale de la Recherche (ANR), France (Tea-4-Two, ANR-14-CE14-0027-
01), PRIM (NewBio4Tb), Institut National de la Santé et de la Recherche Médicale, Université de Lille,
Institut Pasteur de Lille, Centre National de la Recherche Scientifique, Région Hauts-de-France
(convention no.12000080), and Société d’Accélération du Transfert de Technologie Nord. The NMR
facilities were funded by the Région Nord-Pas de Calais (France), the Ministère de la Jeunesse, de
l'Education Nationale et de la Recherche (MJENR), the Fonds Européens de Développement Régional
(FEDER) and Lille University. René Wintjens is a Research Associate at the Belgian National Fund for
Scientific Research (FRS-FNRS). We are indebted to the ID30A-1 (MASSIF-1) ESRF (proposals MX1866

ACCEPTED MANUSCRIPT
and MX1960) and to the SOLEIL-PROXIMA-I/II (BAG proposal 20171555) synchrotron facilities for
beam-time allocations.
4. Experimental section
4.1. Chemistry
Solvents for synthesis, analysis and purification were purchased as analytical grade from commercial
suppliers and used directly without further purification. Chemical reagents were purchased from
Acros, Aldrich, Fluka, Merck, Maybridge, Fluorochem, TCI or Alfa Aesar as reagent grade and used
without further purification.
HPLC-MS analysis was performed on two different HPLC-MS systems:
LC-MS Waters Alliance Micromass ZQ 2000 system was equipped with a Waters 2747 sample
manager, a Waters 2695 separations module, a Waters 2996 photodiode array detector (200-800
nm) and a Waters Micromass ZQ2000 detector (scan 100-800).
LC-MS Waters 3100 Mass Detectors system was equipped with a Waters 2767 sample manager, a
Waters 515 HPLC pump, a Waters Systems Fluidics Organizer, a Waters 2545 Binary Gradient
Module, a Waters 2487 Dual λ Absorbance (215 nm and 254 nm) detector and a Waters 3100 Mass
detectors.
XBridge C18 column (3.5 µm particle size, dimensions 50 mm x 4.6 mm) was used for HPLC analysis.
The injection volume was 20 µL. A mixture of water and acetonitrile was used as mobile phase in
gradient-elution. The pH of the mobile phase was adjusted with HCOOH and NH4OH to form a buffer
solution at pH 3.8 or pH 9.2. The analysis time was 5 minutes (at a flow rate at 2 mL/min), 10 minutes
(at a flow rate at 1 mL/min) or 30 minutes (at a flow rate at 1 mL/min). Purity (%) was determined by
reversed phase HPLC, using UV detection (215 nm), and all isolated compounds showed purity
greater than 95%.
HRMS analysis was performed on a LCT Premier XE Micromass, using a C18 X-Bridge 3.5 μm particle
size column, dimensions 50 mm * 4.6 mm. A gradient starting from 98% H2O 5 mM Ammonium
Formate pH=3.8 and reaching 100% CH3CN 5 mM Ammonium Formate pH=3.8 within 3 min at a flow
rate of 1 mL/min was used.
NMR spectra were recorded on a Bruker DRX-300 spectrometer. The results were calibrated to
signals from the solvent as an internal reference [e.g. 5.32 (residual CD2Cl2) and 53.84 (CD2Cl2) ppm,
2.50 (residual DMSO-d6) and 39.52 (DMSO-d6) ppm for ¹H and ¹³C NMR spectra, respectively].
Chemical shifts (δ) are in parts per million (ppm) downfield from tetramethylsilane (TMS). The
1
assignments were made using one-dimensional (1D) H and 13C spectra and two-dimensional (2D)
HSQC-DEPT, COSY and HMBC spectra. NMR coupling constants (J) are reported in Hertz (Hz), and
splitting patterns are indicated as follows: s for singlet, brs for broad singlet, d for doublet, t for
triplet, a for quartet, quin for quintet, sex for sextet, dd for doublet of doublet, ddd for doublet of
doublet of doublet, dt for doublet of triplet, qd for quartet of doublet, m for multiplet, δ for chemical
shift, J for coupling constant.
Flash chromatography was performed using silica gel cartridges. The reactions were monitored by
thin layer chromatography on silica gel pre-coated F254 Merck plates and the resulting eluted plated
were viewed under ultraviolet light at 254 nm and/or stained using a solution potassium
permanganate.
General procedure A for the synthesis of the tertiary ammonium iodide salt.

ACCEPTED MANUSCRIPT
The corresponding cyclic ketone (1 eq.) was dissolved in acetone (C = 1 M) under stirring. Methyl
iodide (1.1 eq.) was added dropwise. Reaction mixture was stirred at RT for 1 hour. Precipitate was
filtered and washed with acetone and with cyclohexane/ethyl acetate solution (1/1). The resulting
solid was dried under reduced pressure to give the desired compound.
1,1-Dimethylpiperidin-1-ium-4-one iodide (6).
The title compound was synthesized according the general procedure A using the N-methyl-4-
1
piperidone as starting material. Yield: 80% as a white solid. H-NMR (DMSO-d6): δ (ppm) 3.74 (t, J =
7.1 Hz, 2H), 3.36 – 3.33 (m, 1H), 3.32 (s, 3H), 3.26 (s, 3H), 2.72 – 2.68 (m, 2H), 1.89 – 1.84 (m, 3H).
9,9-Dimethyl-9-azoniabicyclo[3.3.1]nonan-3-one iodide (8)
The title compound was synthesized according the general procedure A using the pseudopelletierine
1
as starting material. Yield: 54% as a white solid. H-NMR (DMSO-d6): δ (ppm) 3.93 - 3.89 (m, 2H),
3.36 (s, 3H), 3.32 (s, 3H), 3.26 - 3.24 (m, 1H), 2.60 - 2.54 (m, 2H), 2.42 - 2.28 (m, 2H), 1.80 - 1.74 (m,
2H), 1.64 - 1.56 (m, 1H), 1.39 - 1.22 (m, 2H).
8,8-Dimethyl-8-azoniabicyclo[3.2.1]octan-3-one iodide (13)
The title compound was synthesized according the general procedure A using the tropinone as
starting material. Yield: 92% as a white solid. 1H-NMR (DMSO-d6): δ (ppm) 4.17 (m, 2H), 3.39 (s, 3H),
3.19 (s, 3H), 2.97 (m, 2H), 2.65 (m, 2H), 2.50 (m, 2H), 1.95 (m, 2H).
General procedure B for the alkylation of the tertiary ammonium iodide salt.
To a solution of the appropriate amine (1 eq.) in EtOH (C = 0.5 M) was added a solution of K2CO3 (2.1
eq.) in water (C = 3 M). The resulting solution was refluxed and a solution of the appropriate tertiary
ammonium iodide (1.1 eq.) in water (C = 0.8 M) was added dropwise. The reaction mixture was
refluxed for 4 hours. The reaction was monitored by TLC. After completion, the reaction was cooled
down to RT and volatiles were removed under vacuum. The residue was extracted with DCM (3x),
washed with brine (1x), dried over MgSO4 and reduced in vacuo. The residue was purified by flash
chromatography to give the desired product.
8-[2-(4-Chlorophenyl)ethyl]-8-azabicyclo[3.2.1]octan-3-one (1)
The title compound was synthesized according the general procedure B using the 8,8-dimethyl-8-
azoniabicyclo[3.2.1]octan-3-one iodide (13) and 2-(4-chlorophenyl)ethylamine as starting materials.
1
Yield: 76%. H-NMR (CD2Cl2): δ (ppm) 7.30 (d, J = 8.7 Hz, 2H), 7.23 (d, J = 8.5 Hz, 2H), 3.58 - 3.53 (m,
2H), 2.88 - 2.79 (m, 4H), 2.63 (dd, J = 16.0 Hz, J = 4.5 Hz, 2H), 2.15 (d, J = 15.8 Hz, 2H), 2.05 - 2.00 (m,
2H), 1.59 (d, J = 7.8 Hz, 2H). 13C-NMR (CD2Cl2): δ (ppm) 209.4, 139.2, 131.6, 130.2, 128.3, 58.9, 52.1,
47.5, 35.2, 27.8. HRMS (TOF, ES+) m/z [M+H]+ calculated for C15H19NOCl 264.1155, found 264.1159.
1-[2-(4-Chlorophenyl)ethyl]piperidin-4-one (7)
The title compound was synthesized according the general procedure B using the 1,1-
dimethylpiperidin-1-ium-4-one iodide (6) and 2-(4-chlorophenyl)ethylamine as starting materials.
1
Flash chromatography cyHex to cyHex/EtOAc 7/3. Yield: 57%. H-NMR (CD2Cl2): δ (ppm) 7.29 (d, J =
8.8 Hz, 2H), 7.21 (d, J = 8.6 Hz, 2H), 2.85 - 2.68 (m, 8H), 2.43 (t, J = 6.3 Hz, 4H). 13C-NMR (CD2Cl2): δ
(ppm) 208.5, 139.1, 131.6, 130.1, 128.3, 58.7, 53.1, 41.2, 33.2. HRMS (TOF, ES+) m/z [M+H]+
calculated for C13H17NOCl 238.0999, found 238.0995.

ACCEPTED MANUSCRIPT
9-[2-(4-Chlorophenyl)ethyl]-9-azabicyclo[3.3.1]nonan-3-one (9)
The title compound was synthesized according the general procedure B using the 9,9-dimethyl-9-
azoniabicyclo[3.3.1]nonan-3-one iodide (8) and 2-(4-chlorophenyl)ethylamine as starting materials.
1
Flash chromatography cyHex to cyHex/EtOAc 8/2. Yield: 69%. H-NMR (CD2Cl2): δ (ppm) 7.30 (d, J =
8.5 Hz, 2H), 7.22 (d, J = 8.5 Hz, 2H), 3.40 - 3.35 (m, 2H), 2.95 - 2.90 (m, 2H), 2.80 - 2.75 (m, 2H), 2.66
(dd, J = 16.4 Hz, J = 6.4 Hz, 2H), 2.20 (d, J = 16.6 Hz, 2H), 1.94 - 1.82 (m, 2H), 1.57 - 1.41 (m, 4H). 13C-
NMR (CD2Cl2): δ (ppm) 210.7, 139.4, 131.5, 130.3, 128.2, 54.5, 54.2, 42.6, 34.7, 29.7, 16.6. HRMS
(TOF, ES+) m/z [M+H]+ calculated for C16H21NOCl 278.1312, found 278.1319.
2-(4-Chlorophenyl)-N,N-bis(methoxymethyl)ethanamine (10). Anhydrous K2CO3 (1 eq.) was added to
a mixture of anhydrous MeOH (C = 2 M) and paraformaldehyde (3.25 eq.). Then, 2-(4-
chlorophenyl)ethylamine (1 eq.) was added over a period of 30 minutes and the resulting mixture
was stirred for 48 hours at RT. The crude reaction was filtered to remove all solids and the filtrate
was concentrated under vacuum. The residue was suspended in anhydrous DCM and filtered. The
filtrate was evaporated under reduced pressure to afford 2-(4-chlorophenyl)-N,N-
1
bis(methoxymethyl)ethanamine (10). Yield: 90%. H-NMR (CD2Cl2): δ (ppm) 7.27 (d, J = 8.6 Hz, 2H),
7.19 (d, J = 8.6 Hz, 2H), 4.24 (s, 4H), 3.20 (s, 6H), 3.11 – 3.06 (m, 2H), 2.83 – 2.78 (m, 2H).
Procedure for the synthesis of compounds 11 and 12.
To a mixture of the appropriate cyclic ketone (cyclopentanone or cyclohexanone, 1 eq.) and 2-(4-
chlorophenyl)-N,N-bis(methoxymethyl)ethanamine (10, 2 eq.) in anhydrous MeCN (C = 1 M) was
added trichloromethylsilane (2 eq.). The mixture was stirred at RT for 3 to 8 hours (monitoring done
by TLC, 2,4-DNPH staining), quenched with an aqueous saturated NaHCO3 and extracted with EtOAc
(3x). The combined organic layers were washed with brine (1x), dried over MgSO4 and concentrated
under vacuum. The residue was purified by flash chromatography to give the desired final product
(11 and 12).
3-[2-(4-Chlorophenyl)ethyl]-3-azabicyclo[3.2.1]octan-8-one (11)
Flash chromatography cyHex to cyHex/DCM/EtOAc 8/1/1. Yield: 40%. 1H-NMR (CD2Cl2): δ (ppm) 7.29
(d, J = 8.4 Hz, 2H), 7.22 (d, J = 8.8 Hz, 2H), 3.06 - 3.01 (m, 2H), 2.81 - 2.75 (m, 2H), 2.72 - 2.67 (m, 2H),
2.56 (d, J = 10.4 Hz, 2H), 2.15 - 2.10 (m, 2H), 1.92 - 1.86 (m, 2H), 1.84 - 1.78 (m, 2H). 13C-NMR
(CD2Cl2): δ (ppm) 220.0, 139.7, 132.0, 130.1, 128.2, 61.8, 56.7, 45.4, 33.1, 22.6. HRMS (TOF, ES+) m/z
[M+H2O+H]+ calculated for C15H21NO2Cl 282.1261, found 282.1252.
3-[2-(4-Chlorophenyl)ethyl]-3-azabicyclo[3.2.1]octan-8-one (12)
Flash chromatography petroleum ether to petroleum ether/(DCM-EtOH 3/1) 9/1. Yield: 25%. 1H-NMR
(CD2Cl2): δ (ppm) 7.30 (d, J = 8.6 Hz, 2H), 7.21 (d, J = 8.6 Hz, 2H), 3.23 (dd, J = 12.6 Hz, J = 1.6 Hz, 2H),
2.81 (t, J = 7.7 Hz, 2H), 2.65 - 2.51 (m, 5H), 2.33 - 2.29 (m, 2H), 2.14 - 2.06 (m, 2H), 2.04 - 1.91 (m, 2H),
1.41 - 1.33 (m, 1H). 13C-NMR (CD2Cl2): δ (ppm) 217.9, 139.6, 131.9, 130.1, 128.2, 60.3, 58.4, 48.0,
34.8, 33.1, 20.9. HRMS (TOF, ES+) m/z [M+H]+ calculated for C16H21NOCl 278.1312, found 278.1310.
8-[(4-Chlorophenyl)methyl]-8-azabicyclo[3.2.1]octan-3-one (14)
The title compound was synthesized according the general procedure B previously described using
the 8,8-dimethyl-8-azoniabicyclo[3.2.1]octan-3-one iodide (13) and the 4-chlorobenzylamine as
starting materials.

ACCEPTED MANUSCRIPT
Flash chromatography cyHex to cyHex/EtOAc 85/15. Yield: 48%. 1H-NMR (CD2Cl2): δ (ppm) 7.42 (d, J =
8.4 Hz, 2H), 7.35 (d, J = 8.5 Hz, 2H), 3.75 (s, 2H), 3.50 - 3.46 (m, 2H), 2.68 (dd, J = 15.7 Hz, J = 4.6 Hz,
2H), 2.22 - 2.10 (m, 4H), 1.64 - 1.61(m, 2H). 13C-NMR (CD2Cl2): δ (ppm) 209.4, 138.4, 132.4, 129.8,
128.3, 58.7, 54.4, 48.2, 27.7. HRMS (TOF, ES+) m/z [M+H]+ calculated for C14H17NOCl 250.0999,
found 250.1003.
8-[2-(4-Chlorophenyl)-1-methyl-ethyl]-8-azabicyclo[3.2.1]octan-3-one (15)
To a solution of 4-chlorophenylacetone (1 eq.) in DCE (C = 1.6 M) with activated molecular sieves 3 Å
was added a solution of 8-azabicyclo[3.2.1]octan-3-one hydrochloride (1.1 eq.) and triethylamine (1.1
eq.) in DCE (C = 0.8 M). Acetic acid (1 eq.) and sodium triacetoxyborohydride (1.4 eq.) were then
added and the resulting solution was stirred at RT for 4 hours. The solution was then basified with
aqueous 1N NaOH solution and extracted with DCM (3x). The combined organic layers were washed
with brine (1x), dried over MgSO4 and concentrated under reduced pressure. The residue was
purified by flash chromatography (cyHex to cyHex/EtOAc 4/6) to give 8-[2-(4-chlorophenyl)-1-methyl-
1
ethyl]-8-azabicyclo[3.2.1]octan-3-one (15). Yield: 68%. H-NMR (CD2Cl2): δ (ppm) 7.30 (d, J = 8.5 Hz,
2H), 7.18 (d, J = 8.5 Hz, 2H), 3.93 - 3.89 (m, 1H), 3.82 - 3.77 (m, 1H), 3.08 - 2.97 (m, 2H), 2.75 - 2. 52
(m, 3H), 2.20 - 2.09 (m, 2H), 2.03 - 1.97 (m, 2H), 1.65 - 1.59 (m, 3H), 1.04 (d, J = 6.5 Hz, 3H). 13C-NMR
(CD2Cl2): δ (ppm) 209.4, 138.3, 131.7, 131.0, 128.2, 56.2, 55.1, 52.4, 45.8, 45.6, 41.4, 28.9, 28.4, 17.9.
HRMS (TOF, ES+) m/z [M+H]+ calculated for C16H21NOCl 278.1312, found 278.1310.
2 steps procedure for the synthesis of 8-[2-(4-chlorophenyl)acetyl]-8-azabicyclo[3.2.1]octan-3-one
(17).
1-(Benzotriazol-1-yl)-2-(4-chlorophenyl)ethanone (16)
1H-benzotriazole (3 eq.) and SOCl2 (1 eq.) in DCM (C = 1 M) were added dropwise to a solution of 2-
(4-chlorophenyl)acetic acid (1 eq.) in DCM (C = 1 M). The reaction mixture was stirred for 24 hours at
RT. The precipitate was filtered off and the filtrate was concentrated under vacuum. The residue was
purified by flash chromatography (cyHex to cyHex/DCM 8/2) to give 1-(benzotriazol-1-yl)-2-(4-
1
chlorophenyl)ethanone (16). Yield : 74%. H-NMR (CD2Cl2): δ (ppm) 8.28 (dt, J = 8.2 Hz, J = 0.9 Hz,
1H), 8.16 (dt, J =8.3 Hz, J = 0.9 Hz, 1H), 7.70 (ddd, J = 8.2 Hz, J = 7.2 Hz, J = 1.1 Hz, 1H), 7.56 (ddd, J =
8.3 Hz, J = 7.2 Hz, J = 1.1 Hz, 1H), 7.43 (d, J = 8.5 Hz, 2H), 7.38 (d, J = 8.7 Hz, 2H), 4.74 (s, 2H).
8-[2-(4-Chlorophenyl)acetyl]-8-azabicyclo[3.2.1]octan-3-one (17)
Nortropinone hydrochloride (1.1 eq.) and DMAP (1.1 eq.) were stirred in chloroform (C = 0.3 M) for 5
minutes at RT. Then, 1-(benzotriazol-1-yl)-2-(4-chlorophenyl)ethanone (16, 1 eq.) was added and the
solution was heated under microwave irradiation for 15 minutes at 100 °C. The resulting solution was
washed with aqueous saturated Na2CO3 (3x), HClaq 1N (3x) and brine (1x). The organic phase was
dried over MgSO4 and concentrated to dryness to afford 8-[2-(4-chlorophenyl)acetyl]-8-
azabicyclo[3.2.1]octan-3-one (17). Yield: 81%. 1H-NMR (CD2Cl2): δ (ppm) 7.36 (d, J = 8.6 Hz, 2H), 7.28
(d, J = 8.6 Hz, 2H), 4.95 - 4.91 (m, 1H), 4.52 - 4.48 (m, 1H), 3.80 (d, J = 15.1 Hz, 1H), 3.70 (d, J = 15,1
Hz, 1H), 2.69 (dd, J = 16.0 Hz, J = 4.7 Hz, 1H), 2.40 – 2.32 (m, 3H), 2.17 - 2.0 (m, 2H), 1.82 - 1.67 (m,
2H). 13C-NMR (CD2Cl2): δ (ppm) 206.7, 166.7, 133.7, 132.7, 130.4, 128.7, 54.2, 51.1, 49.4, 48.7, 40.5,
29.9, 27.8. HRMS (TOF, ES+) m/z [M+H]+ calculated for C15H17NO2Cl 278.0948, found 278.0946.
8-[2-(4-Chlorophenyl)ethyl]-3-(difluoromethylene)-8-azabicyclo[3.2.1]octane (18)
Under argon atmosphere, potassium tert-butoxide (2.2 eq.) dissolved in dry DMF (C = 0.9 M) was
added to a solution of 8-[2-(4-chlorophenyl)ethyl]-8-azabicyclo[3.2.1]octan-3-one (1, 1.2 eq.) and 2-

ACCEPTED MANUSCRIPT
(difluoromethylsulfonyl)pyridine (1 eq.) in dry DMF (C = 0.25 M) at -50 °C. The reaction mixture was
allowed to warm up to -25 °C. The mixture was stirred for 4 days at -25 °C. Then, the reaction mixture
was quenched with aqueous saturated NH4Cl (C = 0.5 M), followed by HCl 3 M (C = 0.5 M). The
reaction mixture was allowed to warm up to RT. A solution of NaOH 3 M (C = 0.5 M) was added and
the product was extracted with DCM (3x). The combined organic layers were dried over MgSO4 and
concentrated in vacuo. The residue was purified by flash chromatography (cyHex to cyHex/EtOAc
9/1) to give 8-[2-(4-chlorophenyl)ethyl]-3-(difluoromethylene)-8-azabicyclo[3.2.1]octane (18). Yield:
1
30%. H-NMR (CD2Cl2): δ (ppm) 7.31 (d, J = 8.6 Hz, 2H), 7.24 (d, J = 8.4 Hz, 2H), 3.54 - 3.46 (m, 2H),
2.97 - 2.92 (m, 2H), 2.79 - 2.74 (m, 2H), 2.53 - 2.49 (m, 2H), 2.16 (dd, J = 15.2 Hz, J = 1.1 Hz, 2H), 2.04 -
1.99 (m, 2H), 1.61 (d, J = 8.0 Hz, 2H). 13C-NMR (CD2Cl2): δ (ppm) 153.3 (t, J = 288.3 Hz), 138.3, 131.8,
130.2, 128.4, 81.4 (t, J = 18.9 Hz), 59.2, 50.0, 33.8, 29.2, 26.3. HRMS (TOF, ES+) m/z [M+H]+
calculated for C16H19NF2Cl 298.1174, found 298.1172.
8'-[2-(4-Chlorophenyl)ethyl]spiro[1,3-dioxolane-2,3'-8-azabicyclo[3.2.1]octane] (19)
To a solution of 8-[2-(4-chlorophenyl)ethyl]-8-azabicyclo[3.2.1]octan-3-one (1, 1 eq.) in dry MeCN (C
= 0.15 M) under an argon stream was added ethanediol (2.2 eq.) followed by chlorotrimethylsilane
(4.4 eq.). The reaction mixture was heated under microwave irradiation for 25 minutes at 80 °C. Then
a saturated solution of NaHCO3 was added and the resulting kksolution was extracted with DCM (3x).
The combined organic layer were washed with brine (1x), dried over MgSO4 and concentrated to
dryness to give 8'-[2-(4-chlorophenyl)ethyl]spiro[1,3-dioxolane-2,3'-8-azabicyclo[3.2.1]octane] (19).
1
Yield: 83%. H-NMR (CD2Cl2): δ (ppm) 7.28 (d, J = 8.5 Hz, 2H), 7.21 (d, J = 8.5 Hz, 2H), 3.95 (t, J = 6.2
Hz, 2H), 3.79 (t, J = 6.5 Hz, 2H), 3.33 - 3.28 (m, 2H), 2.83 - 2.75 (m, 2H), 2.70 - 2.64 (m, 2H), 2.0 - 1.94
(m, 4H), 1.92 - 1.83 (m, 2H), 1.67 (d, J = 13.6 Hz, 2H). 13C-NMR (CD2Cl2): δ (ppm) 139.4, 131.5, 130.2,
128.2, 107.2, 64.2, 62.9, 58.5, 39.8, 34.9, 29.7, 25.7. HRMS (TOF, ES+) m/z [M+H]+ calculated for
C17H23NO2Cl 308.1417, found 308.1419.
8-[2-(4-Chlorophenyl)ethyl]-3,3-dimethoxy-8-azabicyclo[3.2.1]octane (20)
Under argon atmosphere, trimethylorthoformate (5 eq.) was added to a solution of 8-[2-(4-
chlorophenyl)ethyl]-8-azabicyclo[3.2.1]octan-3-one (1, 1 eq.) and PTSA (2 eq.) in dry MeOH (C = 0.5
M). The reaction mixture was heated under microwave irradiation for 25 minutes at 80 °C. Then a
saturated solution of NaHCO3 was added and the resulting solution was extracted with DCM (3x). The
combined organic layers were dried over MgSO4 and concentrated under vacuum. The residue was
purified by flash chromatography (cyHex to cyHex/[EtOAc/MeOH 3/1]) to yield 8-[2-(4-
1
chlorophenyl)ethyl]-3,3-dimethoxy-8-azabicyclo[3.2.1]octane (20). Yield: 60%. H-NMR (CD2Cl2): δ
(ppm) 7.29 (d, J = 8.6 Hz, 2H), 7.21 (d, J = 8.6 Hz, 2H), 3.27 - 3.21 (m, 2H), 3.16 (s, 3H), 3.11 (s, 3H),
2.78 - 2.73 (m, 2H), 2.66 - 2.60 (m, 2H), 1.92 - 1.86 (m, 2H), 1.81 - 1.73 (m, 6H). 13C-NMR (CD2Cl2): δ
(ppm) 139.7, 131.3, 130.2, 128.2, 98.9, 57.9, 52.7, 47.9, 46.8, 37.1, 35.2, 25.8. HRMS (TOF, ES+) m/z
[M+H]+ calculated for C17H25NO2Cl 310.1574, found 310.1569.
8-(2-Phenylethyl)-8-azabicyclo[3.2.1]octan-3-one (21)
The title compound was synthesized according the general procedure B previously described using
the 8,8-dimethyl-8-azoniabicyclo[3.2.1]octan-3-one iodide (13) and 2-phenylethylamine as starting
1
materials. Yield: 54%. H-NMR (CD2Cl2): δ (ppm) 7.34 - 7.19 (m, 5H), 3.6 1- 3.55 (m, 2H), 2.91 - 2.80
(m, 4H), 2.66 (dd, J = 16.0 Hz, J = 4.5 Hz, 2H), 2.19 - 2.13 (m, 2H), 2.06 - 2.01 (m, 2H), 1.63 - 1.55 (m,

ACCEPTED MANUSCRIPT
2H). 13C-NMR (CD2Cl2): δ (ppm) 209.5, 140.6, 128.7, 128.3, 126.0, 58.8, 52.3, 47.4, 35.9, 27.9. HRMS
(TOF, ES+) m/z [M+H]+ calculated for C15H20NO 230.1545, found 230.1546.
8-[2-(3,4-Dichlorophenyl)ethyl]-8-azabicyclo[3.2.1]octan-3-one (22)
The title compound was synthesized according the general procedure B previously described using
the 8,8-dimethyl-8-azoniabicyclo[3.2.1]octan-3-one iodide (13) and 3,4-dichlorophenylethylamine as
starting materials. Yield: 60%. 1H-NMR (CD2Cl2): δ (ppm) 7.41 - 7.39 (d, J = 8.2 Hz, 1H), 7.41 (d, J =2.1
Hz, 1H), 7.17-7.13 (dd, J = 8.2 Hz, J =2.1 Hz, 1H), 3.53 (m, 2H), 2.82 (s, 4H), 2.65 - 2.59 (dd, J = 15.91
Hz, J = 4.30 Hz, 2H), 2.19 - 2.13 (m, 2H), 2.05 - 2.01 (m, 2H), 1.63 - 1.56 (m, 2H). 13C-NMR (CD2Cl2): δ
(ppm) 209.3, 141.2, 131.8, 130.7, 130.1, 129.7, 128.5, 58.9, 51.9, 47.6, 35.0, 27.8. HRMS (TOF, ES+)
m/z [M+H]+ calculated for C15H18NOCl2 298.0765, found 298.0768.
8-[2-(2,4-Dichlorophenyl)ethyl]-8-azabicyclo[3.2.1]octan-3-one (23)
The title compound was synthesized according the general procedure B previously described using
the 8,8-dimethyl-8-azoniabicyclo[3.2.1]octan-3-one iodide (13) and 2,4-dichlorophenylethylamine as
1
starting materials. Yield: 32%. H-NMR (CD2Cl2): δ (ppm) 7.41 (d, J = 2.0 Hz, 1H), 7.29 (d, J = 8.2 Hz,
1H), 7.23 (dd, J = 8.2 Hz, J = 2.0 Hz, 1H), 3.59 - 3.55 (m, 2H), 3.0 - 2.94 (m, 2H), 2.83 - 2.77 (m, 2H),
2.64 (dd, J = 16.0 Hz, J = 4.5 Hz, 2H), 2.16 (d, J = 15.9 Hz, 2H), 2.06 - 2.01 (m, 2H), 1.58 (d, J = 7.9 Hz,
2H). 13C-NMR (CD2Cl2): δ (ppm) 209.3, 136.8, 134.6, 132.3, 131.9, 129.0, 127.0, 59.0, 50.4, 47.6, 33.1,
27.8. HRMS (TOF, ES+) m/z [M+H]+ calculated for C15H18NOCl2 298.0765, found 298.0764.
8-[2-(4-Pyridyl)ethyl]-8-azabicyclo[3.2.1]octan-3-one (24)
The title compound was synthesized according the general procedure B previously described using
the 8,8-dimethyl-8-azoniabicyclo[3.2.1]octan-3-one iodide (13) and 4-(2-aminoethyl)pyridine as
1
starting materials. Flash chromatography cyHex to cyHex/(EtOAc-MeOH 3/1) 75/25. Yield: 64%. H-
NMR (CD2Cl2): δ (ppm) 8.49 (d, J = 4.3 Hz, 2H), 7.22 (d, J = 4.2 Hz, 2H), 3.57 - 3.52 (m, 2H), 2.86 (s, 4H),
2.66 - 2.59 (dd, J = 15.9 Hz, J = 4.4 Hz, 2H), 2.19 - 2.13 (m, 2H), 2.06 - 2.01 (m, 2H), 1.63 - 1.56 (m, 2H).
13C-NMR (CD2Cl2): δ (ppm) 209.3, 149.5, 133.3 (HMBC coupling), 124.2, 58.9, 51.2, 47.6, 35.1, 27.8.
HRMS (TOF, ES+) m/z [M+H]+ calculated for C14H19N2O 231.1497, found 231.1502.
8-[2-(p-Tolyl)ethyl]-8-azabicyclo[3.2.1]octan-3-one (25)
The title compound was synthesized according the general procedure B previously described using
the 8,8-dimethyl-8-azoniabicyclo[3.2.1]octan-3-one iodide (13) and 2-(4-methylphenyl)ethylamine as
1
starting materials. Flash chromatography cyHex to cyHex/EtOAc 8/2. Yield: 59%. H-NMR (CD2Cl2): δ
(ppm) 7.17 - 7.10 (m, 4H), 3.58 - 3.45 (m, 2H), 2.95 - 2.82 (m, 4H), 2.69 (dd, J = 16 Hz, J = 5.2 Hz, 2H),
2.34 (s, 3H), 2.18 - 2.12 (m, 2H), 2.06 - 2.00 (m, 2H), 1.62 - 1.55 (m, 2H). 13C-NMR (CD2Cl2): δ (ppm)
209.5, 137.4, 135.6, 128.9, 128.6, 58.8, 52.4, 47.3, 35.4, 27.9, 20.7. HRMS (TOF, ES+) m/z [M+H]+
calculated for C16H22NO 244.1701, found 244.1703.
8-[2-(4-Methoxyphenyl)ethyl]-8-azabicyclo[3.2.1]octan-3-one (26)
The title compound was synthesized according the general procedure B previously described using
the 8,8-dimethyl-8-azoniabicyclo[3.2.1]octan-3-one iodide (13) and 2-(4-methoxyphenyl)ethylamine
as starting materials. Flash chromatography cyHex to cyHex/EtOAc 8/2. Yield: 59%. 1H-NMR (CD2Cl2):
δ (ppm) 7.20 (d, J = 8.7 Hz, 2H), 6.87 (d, J = 8.7 Hz, 2H), 3.80 (s, 3H), 3.61 - 3.56 (m, 2H), 2.81 (s, 4H),
2.66 (dd, J = 15.8 Hz, J = 4.3 Hz, 2H), 2.17 (d, J = 15.6 Hz, 2H), 2.08 - 2.0 (m, 2H), 1.60 (d, J = 7.9 Hz,

ACCEPTED MANUSCRIPT
2H). 13C-NMR (CD2Cl2): δ (ppm) 209.6, 158.0, 132.5, 129.6, 113.6, 58.8, 55.1, 52.5, 47.4, 35.0, 27.9.
HRMS (TOF, ES+) m/z [M+H]+ calculated for C16H22NO2 260.1651, found 260.1649.
8-[2-(4-Fluorophenyl)ethyl]-8-azabicyclo[3.2.1]octan-3-one (27)
The title compound was synthesized according the general procedure B previously described using
the 8,8-dimethyl-8-azoniabicyclo[3.2.1]octan-3-one iodide (13) and 2-(4-fluorophenyl)ethylamine as
1
starting materials. Flash chromatography cyHex to cyHex/EtOAc 6/4. Yield: 51%. H-NMR (CD2Cl2): δ
(ppm) 7.25 (dd, J = 5.6 Hz, J = 8.7 Hz, 2H), 7.0 (t, J = 9 Hz, 2H), 3.57 - 3.54 (m, 2H), 2.88 - 2.77 (m, 4H),
2.63 (dd, J = 16.0 Hz, J = 4.6 Hz, 2H), 2.16 (d, J = 15.7 Hz, 2H), 2.06 - 2.00 (m, 2H), 1.63 - 1.55 (m, 2H).
13C-NMR (CD2Cl2): δ (ppm) 209.4, 161.4 (d, J = 242 Hz), 136.4 (d, J = 3 Hz), 130.2 (d, J = 7.8 Hz), 114.9
(d, J = 21 Hz), 58.9, 52.4, 47.5, 35.1, 27.8. HRMS (TOF, ES+) m/z [M+H]+ calculated for C15H19NOF
248.1451, found 248.1448.
8-[2-(4-Bromophenyl)ethyl]-8-azabicyclo[3.2.1]octan-3-one (28)
The title compound was synthesized according the general procedure B previously described using
the 8,8-dimethyl-8-azoniabicyclo[3.2.1]octan-3-one iodide (13) and 2-(4-bromophenyl)ethylamine as
1
starting materials. Flash chromatography cyHex to cyHex/EtOAc 8/2. Yield: 64%. H-NMR (CD2Cl2): δ
(ppm) 7.44 (d, J = 8.4 Hz, 2H), 7.18 (d, J = 8.4 Hz, 2H), 3.56 - 3.53 (m, 2H), 2.86 - 2.82 (m, 4H), 2.63
(dd, J = 15.9 Hz, J = 4.3 Hz, 2H), 2.19 - 2.12 (m, 2H), 2.05 - 2.01 (m, 2H), 1.62 - 1.55 (m, 2H). 13C-NMR
(CD2Cl2): δ (ppm) 208.5, 139.8, 131.7, 130.9, 119.7, 59.0, 52.2, 47.4, 35.3, 27.8. HRMS (TOF, ES+) m/z
[M+H]+ calculated for C15H19NOBr 308.0650, found 308.0656.
8-[2-(4-Nitrophenyl)ethyl]-8-azabicyclo[3.2.1]octan-3-one (29)
The title compound was synthesized according the general procedure B previously described using
the 8,8-dimethyl-8-azoniabicyclo[3.2.1]octan-3-one iodide (13) and 2-(4-nitrophenyl)ethylamine as
1
starting materials. Yield: 45%. H-NMR (CD2Cl2): δ (ppm) 8.17 (d, J = 8.7 Hz, 2H), 7.46 (d, J = 8.7 Hz,
2H), 3.57 - 3.52 (m, 2H), 3.0 - 2.95 (m, 2H), 2.90 - 2.84 (m, 2H), 2.62 (dd, J = 16.1 Hz, J = 4.4 Hz, 2H),
2.17 (d, J = 16.1 Hz, 2H), 2.06 - 2.00 (m, 2H), 1.58 (d, J = 8.7 Hz, 2H). 13C-NMR (CD2Cl2): δ (ppm) 209.2,
148.7, 129.7, 123.4, 121.1 (HMBC coupling), 59.0, 61.7, 47.6, 35.6, 27.8. HRMS (TOF, ES+) m/z
[M+H]+ calculated for C15H19N2O3 275.1396, found 275.1401.
8-[2-[4-(Trifluoromethyl)phenyl]ethyl]-8-azabicyclo[3.2.1]octan-3-one (30)
The title compound was synthesized according the general procedure B previously described using
the
8,8-dimethyl-8-azoniabicyclo[3.2.1]octan-3-one
iodide
(13)
and
2-(4-
trifluoromethylphenyl)ethylamine as starting materials. Yield: 26%. 1H-NMR (CD2Cl2): δ (ppm) 7.59 (d,
J = 8.0 Hz, 2H), 7.42 (d, J = 7.8 Hz, 2H), 3.58 - 3.54 (m, 2H), 2.97 - 2.91 (m, 2H), 2.88 - 2.83 (m, 2H),
2.64 (dd, J = 16.0 Hz, J = 4.6 Hz, 2H), 2.17 (d, J = 16.7 Hz, 2H), 2.07 - 2.02 (m, 2H), 1.64 - 1.56 (m, 2H).
13C-NMR (CD2Cl2): δ (ppm) 209.3, 145.0, 129.2, 128.1 (q, J = 31.7 Hz), 125.1 (q, J = 3.7 Hz), 124.4 (q, J
= 271.5 Hz), 58.9, 51.9, 47.5, 35.7, 27.8. HRMS (TOF, ES+) m/z [M+H]+ calculated for C16H19NOF3
298.1419, found 298.1422.
8-[2-[3-(Trifluoromethyl)phenyl]ethyl]-8-azabicyclo[3.2.1]octan-3-one (31)
The title compound was synthesized according the general procedure B previously described using
the
8,8-dimethyl-8-azoniabicyclo[3.2.1]octan-3-one
iodide
(13)
and
2-(3-
trifluoromethylphenyl)ethylamine as starting materials. Flash chromatography cyHex to cyHex/EtOAc

ACCEPTED MANUSCRIPT
75/25. Yield: 25%. 1H-NMR (CD2Cl2): δ (ppm) 7.56 (s, 1H), 1.53 - 7.43 (m, 3H), 3.58 - 3.54 (m, 2H), 3.0 -
2.91 (m, 2H), 2.88 - 2.83 (m, 2H), 2.64 (dd, J = 16.1 Hz, J = 4.1 Hz, 2H), 2.16 (d, J = 16.1 Hz, 2H), 2.07 -
2.01 (m, 2H), 1.58 (d, J = 7.9 Hz, 2H). 13C-NMR (CD2Cl2): δ (ppm) 209.2, 141.6, 132.4, 130.2 (q, J = 31.6
Hz), 128.7, 125.5 (q, J = 3.7 Hz), 124.4 (q, J = 272 Hz), 122.8 (q, J = 3.8 Hz), 58.9, 52.0, 47.5, 35.6, 27.8.
HRMS (TOF, ES+) m/z [M+H]+ calculated for C16H19NOF3 298.1419, found 298.1420.
8-[2-[2-(Trifluoromethyl)phenyl]ethyl]-8-azabicyclo[3.2.1]octan-3-one (32)
The title compound was synthesized according the general procedure B previously described using
the
8,8-dimethyl-8-azoniabicyclo[3.2.1]octan-3-one
iodide
(13)
and
2-(2-
trifluoromethylphenyl)ethylamine as starting materials. Flash chromatography cyHex to
cyHex/(EtOAc-EtOH 3-1) 9/1. Yield: 25%. 1H-NMR (CD2Cl2): δ (ppm) 7.67 (d, J = 7.9 Hz, 1H), 7.54 (t, J =
7.9 Hz, 1H), 7.47 (d, J = 7.4 Hz, 1H), 7.36 (t, J = 7.4 Hz, 1H), 3.63 – 3.57 (m, 2H), 3.08 – 3.03 (m, 2H),
2.85 – 2.79 (m, 2H), 2.65 (dd, J = 16.1 Hz, J = 4.4 Hz, 2H), 2.18 (dd, J = 17.1 Hz, J = 1.8 Hz, 2H), 2.07 –
2.00 (m, 2H), 1.6 (d, J = 8.0 Hz, 2H). 13C-NMR (CD2Cl2): δ (ppm) 209.4, 138.9, 131.8, 131.7, 128.3 (q, J =
29.9 Hz), 126.3, 125.9 (q, J = 4.7 Hz), 124.6 (q, J = 274.0 Hz), 59.0, 52.4, 47.6, 32.7, 27.8. HRMS (TOF,
ES+) m/z [M+H]+ calculated for C16H19NOF3 298.1419, found 298.1417.
4.2. Crystallography
Protein production and purification were performed as previously reported [7]. Crystals in
orthorhombic space group P212121 were obtained in several conditions, with the best conditions
being hanging drop vapor diffusion technique at 20 °C and 200 mM ammonium chloride, 100 mM
MES pH 6.0, 20% (w/v) PEG6000, as crystallization buffer. In addition, EthR2 crystals in monoclinic
space group C121 were produced by changing imidazole concentration of the protein sample from
100 to 400 mM and by ranging the pH and the ammonium chloride concentration of crystallization
buffer in 6.0-7.0 and 200-400 mM, respectively. Crystals were then transferred to a new drop for
soaking experiments with small molecules at 1.0 mM concentration for 1-3 days at 20 °C. Molecules
9 and 30 were solved by co-crystallization, as attempts to soak EthR2 crystals with these molecules
failed. In that cases the following procedure was applied: a diluted protein sample (60 µM) was
incubated at 4 °C for overnight with 1 mM compound (protein:/compound molar ratio 6/100 and
DMSO final concentration of 5% (v/v)); the solution was then concentrated to ~10 mg/ml protein
concentration and screened for crystallization with the initial crystallographic conditions of
orthorhombic form.
The X-ray diffraction data were collected in-house or using synchrotron radiation facilities at ESRF-
ID30A-1 (Grenoble, France) and at SOLEIL-PX1 (Paris, France). Data indexations were automatically
performed with autoProc or XDSME, both based on the XDS program.[27] Data collection statistics
are provided in Table 6. Structures were solved by molecular replacement with MOLREP [28] using as
searching template the PDB ID 5N7O [15]. Structure refinements were done without non-
crystallographic symmetry restraints by iteratively cycling through REFMAC5 [29] and COOT [30].
Stereochemical description and restraints for small molecules were produced using AceDRG [31]
from the SMILES string and the initial ligand pose was set up using the FINDLIGAND tool of the CCP4
software suite. In each solved crystal structures, ligands were modeled and refined at full occupancy.
The fit between modeled ligand and electron density was assessed with EDSTATS [32], using four
local metrics, the real-space R-factor (RSR), the real-space correlation coefficient (RSCC), the real-
space Z-difference score (RSZD) and the real-space Z-observed score (RSZO). Final refinement

ACCEPTED MANUSCRIPT
statistics are given in Table 6 using MolProbity [33]. The refined coordinates and the structure factors
were deposited in the Protein Data Bank under the PDB IDs indicated in Table 6.
compound
PDB ID
1
2
3
4
5
9
30
6HRW
BDM14272
6HRX
6HRY
6HRZ
6HS0
6HS1
6HS2
Structure ID
BDM72201
BDM72719
BDM72170
BDM71847
BDM76060
BDM76150
Data collection statistics
Beamline
Rigaku
MicroMax-
007
SOLEIL/PX1
SOLEIL/PX1
ESRF/MASSIF
-1
ESRF/MASSIF
-1
ESRF/MASSIF
-1
ESRF/MASSIF
-1
Detector type
Wavelength (Å)
Autoprocessing
software
Saturn944+
1.541870
XDSME
Pilatus 6M
0.97857
XDSME
Pilatus 6M
0.97857
XDSME
Pilatus3 2M
0.96600
Pilatus3 2M
0.96600
Pilatus3 2M
0.96600
Pilatus3 2M
0.96600
autoProc
autoProc
autoProc
autoProc
Space group
a, b, c (Å)
P212121
P212121
P212121
C121
C121
P212121
P212121
59.81; 74.57; 59.54; 74.15; 59.52; 75.09; 109.91;
88.78 91.58 88.48
90.00; 90.00; 90.00; 90.00; 90.00; 90.00; 90.00; 95.67; 90.00; 96.75; 90.00; 90.00; 90.00; 90.00;
110.09;
59.46; 74.01; 59.48; 73.57;
40.72; 108.93 40.56; 108.74 91.65
91.10
α, β, γ (°)
90.00
Resolution range 30.00-2.45
(Å) (2.60-2.45)
90.00
90.00
90.00
90.00
90.00
90.00
50.00-1.87
(1.98-1.87)
34087 (5344)
50.00-1.84
(1.95-1.84)
34594 (5290)
54.69-1.57
(1.61-1.57)
66857 (4922)
107.99-1.45
(1.48-1.45)
80773 (4071)
57.58-1.69
(1.72-1.69)
45240 (2215)
49.80-1.87
(1.90-1.87)
33585 (1659)
No. of unique 15012 (2334)
reflections
Multiplicity
Completeness (%)
mean (I/σ(I))
Rmeas (%)
6.99 (6.57)
99.4 (97.5)
9.65 (2.04)
23.5 (120.3)
2.05
5.41 (5.36)
99.5 (98.2)
20.66 (2.92)
4.8 (61.4)
2.10
5.37 (5.34)
98.9 (95.2)
18.34 (2.35)
6.1 (66.3)
2.05
3.1 (2.9)
98.7 (99.0)
16.98 (2.37)
4.0 (54.8)
2.52
2.7 (2.4)
95.4 (98.3)
13.95 (2.23)
4.5 (57.0)
2.50
4.6 (4.8)
99.0 (98.8)
14.25 (2.17)
5.6 (82.7)
2.09
4.9 (4.6)
99.5 (99.5)
17.12 (1.95)
5.2 (86.9)
2.07
VM (A3 Da-1)
Solvent content (%) 40.17
41.41
40.10
51.17
50.88
41.27
40.58
Refinement statistics (MolProbity)
Number of refined 2914/18/9
2910/28/27
2961/28/54
3030/26/159
3037/28/240
2914/19/85
2914/21/41
atoms
protein/ligand/wat
er
Final Rwork/Rfree 22.14/26.98
(%)
22.49/26.08
0.0149
21.34/24.24
0.0151
21.26/23.03
0.0156
20.30/21.77
0.0149
22.38/25.50
0.0144
24.33/22.60
0.0141
RMSD
bond 0.0149
lengths (Å)
RMSD bond angles 2.01
1.79
1.80
1.89
1.87
1.76
1.76
(°)
Overall mean
factor (Å2)
B
35.1/57.1/23
.1
37.3/52.5/33
.9
32.6/36.6/34. 26.5/26.5/30. 21.1/25.2/28. 33.8/41.1/34. 39.3/44.7/38.
0
1
2
9
1
protein/ligand/wat
er
Rama
favored/outliers
(%)
99.47/0.26
5.26 (99th)
100.00/0.00
100.00/0.00
99.75/0.00
100.00/0.00
100.00/0.00
99.74/0.00
Clashscore
(percentile)
4.74 (97th)
1.20 (99th)
3.66 (98th)
1.17 (99th)
3.42 (98th)
1.13 (99th)
2.43 (99th)
1.03 (99th)
3.56 (98th)
1.15 (99th)
4.06 (98th)
1.19 (99th)
MolProbity score 1.47 (99th)
(percent)
Ligand RSR/RSCC
Ligand RSZD/RSZO
0.120/0.887
0.763/1.779
0.109/0.929
0.791/2.140
0.097/0.922
0.692/2.68
0.070/0.970
0.843/3.936
0.0835/0.943 0.102/0.923
1.114/3.861 0.493/2.262
0.074/0.954
0.603/3.072
Table 6 : crystallographic data

ACCEPTED MANUSCRIPT
4.3. Biology
4.3.1 Thermal Shift Assay on EthR2 (screening conditions)
60 nL of each fragment at 100 mM in DMSO was dispensed in a 384-well plate using the Echo®550
liquid handler (LabCyte). The recombinant M. tuberculosis protein EthR2 (Rv0078) was used at a final
concentration of 20 µM. The fluorescent dye NanoOrange® (Life technologies) was used to monitor
EthR2 unfolding. This dye is environmentally sensitive and leads to an increase in fluorescence
following exposure of hydrophobic regions during protein unfolding. The final sample concentration
in a total assay volume of 6.06 µL was 20 μM EthR2, 3x NanoOrange®, 1.6% DMSO, and 1 mM
fragment in the Thermal Shift Assay-buffer (100 mM Glycine, pH 7, 150 mM NaCl). The thermal shift
assay was conducted in a Lightcycler 480 (Roche). The system contained a heating/cooling device for
temperature control and a charge-coupled device (CCD) detector for real-time imaging of the
fluorescence changes in the wells of the microplate. The samples were heated from 37 to 85 °C with
a heating rate of 0.05 °C/s. The fluorescence intensity was measured at Ex/Em = 465/510 nm. The
LightCycler® exported .ixo files were processed using the protein melting analysis tool (Roche). The
output was a .csv file containing the inflection point (Tm).
4.3.2 Thermal Shift Assay on EthR2 (SAR conditions)
Compounds in DMSO were dispensed in a 96-well plate. The recombinant M. tuberculosis protein
EthR2 (Rv0078) was used at a final concentration of 10 μM. The fluorescent dye SYPRO® Orange
(SigmaAldrich) was used to monitor EthR2 unfolding. This dye is environmentally sensitive and leads
to an increase in fluorescence following exposure of hydrophobic regions during protein unfolding.
The thermal shift assay was conducted in a Lightcycler 480 (Roche). The system contained a
heating/cooling device for temperature control and a charge-coupled device (CCD) detector for real-
time imaging of the fluorescence changes in the wells of the microplate. The final sample
concentration was 10 μM EthR2, 3x SYPRO® Orange, 1% DMSO, and 20 μM ligand in the Thermal
Shift Assay-buffer (100 mM Glycine, pH 7, 150 mM NaCl). The samples were heated from 37 to 85 °C
with a heating rate of 0.04 °C/s. The fluorescence intensity was measured at Ex/Em = 465/510 nm.
The LightCycler® exported ixo files that were processed using the protein melting analysis tool
(Roche). The output was a csv file containing the inflection point (Tm).
4.3.3 Mammalian SEAP reporter assay
Vector design. A chimeric mammalian transcriptional regulator was constructed by fusing the coding
sequence of EthR2 (Rv0078) from Mycobacterium tuberculosis to the Herpes simplex derived
transactivator protein VP16. The resulting ORF was sequence optimized for expression in
human/mouse, synthesized (Genscript) and introduced into pSEAP2-control (Clontech) using
EcoRI/XbaI to generate the expression vector pCK289 (PSV40-EthR2-VP16-HA-pA). To enable EthR2-
VP16 based transcriptional control, we designed a synthetic DNA by fusing the 61-nucleotide
spanning intergenic region of ethA2 (Rv0077c) and ethR2 (Rv0078) upstream of a minimal variant of
the human
cytomegalovirus derived promotor (PhCMVmin). This synthetic promoter was cloned into pSEAP2-
basic (Clontech) by restriction (XhoI/EcoRI) and ligation to generate pCK287 (OEthR2-PhCMVmin-
SEAP-pA), which enables EthR2-VP16 based control of human placental secreted alkaline
phosphatase (SEAP) expression.
Cell culture. Baby hamster kidney cells (BHK-21, American Type Culture Collection CCL-10) were
cultured in Dulbecco’s modified Eagle’s medium (DMEM, Gibco, catalog no. 41966) supplemented

ACCEPTED MANUSCRIPT
with 10% (v/v) heat-inactivated fetal bovine serum (FBS, Gibco, catalog no. 10500) and 1% (v/v)
penicillin/streptomycin (Gibco, catalog no. 15140) in a humidified atmosphere with 5% CO2 at 37°C.
Prior to transfection, cells from a confluent grown culture dish were split 1:2 to a new petri dish and
grown for 16 hours. For transfection, 10μg of total plasmid DNA (5μg pCK287 and 5μg pCK289, ratio
1:1) were added to 1mL Opti-MEM medium (Gibco, catalog no. 11058021), mixed with 30μL
MegaTran 1.0 transfection reagent (Origene, catalog no. TT200003) and incubated for 15min at room
temperature. The DNA mix was transferred to the cells and incubated for 6h for plasmid uptake. The
cells were washed with PBS, detached from the plate with 0.05% trypsin-EDTA (Gibco, catalog no.
25300), counted and diluted to 500.000 cells/mL using fresh culture medium. Hundred microliter of
cell suspension was dispersed per well of a 96 well culture dish. Compound dilutions were prepared
by adding 1.2μL compounds (10mM stock solution in DMSO) to 600μL culture medium, 1:2 dilution
rows were prepared, and 100μL of compound dilutions were finally added to the cells. After
incubation for 48h the SEAP expression was quantified from cell culture supernatants using a p-
nitrophenyl phosphate based method.

ACCEPTED MANUSCRIPT
References
[1] Global tuberculosis report 2017. Geneva: World Health Organization., 2017.
[2] N. Willand, B. Dirie, X. Carette, P. Bifani, A. Singhal, M. Desroses, F. Leroux, E. Willery, V. Mathys,
R. Deprez-Poulain, G. Delcroix, F. Frenois, M. Aumercier, C. Locht, V. Villeret, B. Deprez, A.R. Baulard,
Synthetic EthR inhibitors boost antituberculous activity of ethionamide, Nature Med., 15 (2009) 537.
[3] M. Flipo, M. Desroses, N. Lecat-Guillet, B. Dirie, X. Carette, F. Leroux, C. Piveteau, F. Demirkaya, Z.
Lens, P. Rucktooa, V. Villeret, T. Christophe, H.K. Jeon, C. Locht, P. Brodin, B. Deprez, A.R. Baulard, N.
Willand, Ethionamide Boosters: Synthesis, Biological Activity, and Structure-Activity Relationships of
a Series of 1,2,4-Oxadiazole EthR Inhibitors, J. Med. Chem., 54 (2011) 2994-3010.
[4] M. Flipo, M. Desroses, N. Lecat-Guillet, B. Villemagne, N. Blondiaux, F. Leroux, C. Piveteau, V.
Mathys, M.P. Flament, J. Siepmann, V. Villeret, A. Wohlkönig, R. Wintjens, S.H. Soror, T. Christophe,
H.K. Jeon, C. Locht, P. Brodin, B. Déprez, A.R. Baulard, N. Willand, Ethionamide boosters. 2.
Combining bioisosteric replacement and structure-based drug design to solve pharmacokinetic issues
in a series of potent 1,2,4-oxadiazole EthR inhibitors, J. Med. Chem., 55 (2012) 68-83.
[5] J. Costa-Gouveia, E. Pancani, S. Jouny, A. Machelart, V. Delorme, G. Salzano, R. Iantomasi, C.
Piveteau, C.J. Queval, O.R. Song, M. Flipo, B. Deprez, J.P. Saint-André, J. Hureaux, L. Majlessi, N.
Willand, A. Baulard, P. Brodin, R. Gref, Combination therapy for tuberculosis treatment: pulmonary
administration of ethionamide and booster co-loaded nanoparticles, Sci Rep, 7 (2017) 5390.
[6] N.J. Tatum, J.W. Liebeschuetz, J.C. Cole, R. Frita, A. Herledan, A.R. Baulard, N. Willand, E. Pohl,
New active leads for tuberculosis booster drugs by structure-based drug discovery, Org Biomol Chem,
15 (2017) 10245-10255.
[7] V. Mendes, T.L. Blundell, Targeting tuberculosis using structure-guided fragment-based drug
design, Drug Discov Today, 22 (2017) 546-554.
[8] C. Marchetti, D.S.H. Chan, A.G. Coyne, C. Abell, Fragment-based approaches to TB drugs,
Parasitology, 145 (2018) 184-195.
[9] B. Villemagne, M. Flipo, N. Blondiaux, C. Crauste, S. Malaquin, F. Leroux, C. Piveteau, V. Villeret, P.
Brodin, B.O. Villoutreix, O. Sperandio, S.H. Soror, A. Wohlkönig, R. Wintjens, B. Deprez, A.R. Baulard,
N. Willand, Ligand efficiency driven design of new inhibitors of Mycobacterium tuberculosis
transcriptional repressor EthR using fragment growing, merging, and linking approaches, J Med
Chem, 57 (2014) 4876-4888.
[10] S. Surade, N. Ty, N. Hengrung, B. Lechartier, S.T. Cole, C. Abell, T.L. Blundell, A structure-guided
fragment-based approach for the discovery of allosteric inhibitors targeting the lipophilic binding site
of transcription factor EthR, Biochem J, 458 (2014) 387-394.
[11] P.O. Nikiforov, S. Surade, M. Blaszczyk, V. Delorme, P. Brodin, A.R. Baulard, T.L. Blundell, C. Abell,
A fragment merging approach towards the development of small molecule inhibitors of
Mycobacterium tuberculosis EthR for use as ethionamide boosters, Org. Biomol. Chem., 14 (2016)
2318-2326.
[12] P.O. Nikiforov, M. Blaszczyk, S. Surade, H.I. Boshoff, A. Sajid, V. Delorme, N. Deboosere, P.
Brodin, A.R. Baulard, C.E. Barry, T.L. Blundell, C. Abell, Fragment-Sized EthR Inhibitors Exhibit
Exceptionally Strong Ethionamide Boosting Effect in Whole-Cell Mycobacterium tuberculosis Assays,
ACS Chem Biol, 12 (2017) 1390-1396.
[13] D.S. Chan, V. Mendes, S.E. Thomas, B.N. McConnell, D. Matak-Vinković, A.G. Coyne, T.L. Blundell,
C. Abell, Fragment Screening against the EthR-DNA Interaction by Native Mass Spectrometry, Angew
Chem Int Ed Engl, 56 (2017) 7488-7491.
[14] N. Blondiaux, M. Moune, M. Desroses, R. Frita, M. Flipo, V. Mathys, K. Soetaert, M. Kiass, V.
Delorme, K. Djaout, V. Trebosc, C. Kemmer, R. Wintjens, A. Wohlkönig, R. Antoine, L. Huot, D. Hot, M.
Coscolla, J. Feldmann, S. Gagneux, C. Locht, P. Brodin, M. Gitzinger, B. Déprez, N. Willand, A.R.
Baulard, Reversion of antibiotic resistance in Mycobacterium tuberculosis by spiroisoxazoline SMARt-
420, Science, 355 (2017) 1206-1211.

ACCEPTED MANUSCRIPT
[15] A. Wohlkönig, H. Remaut, M. Moune, A. Tanina, F. Meyer, M. Desroses, J. Steyaert, N. Willand,
A.R. Baulard, R. Wintjens, Structural analysis of the interaction between spiroisoxazoline SMARt-420
and the Mycobacterium tuberculosis repressor EthR2, Biochem Biophys Res Commun, 487 (2017)
403-408.
[16] M.C. Lo, A. Aulabaugh, G. Jin, R. Cowling, J. Bard, M. Malamas, G. Ellestad, Evaluation of
fluorescence-based thermal shift assays for hit identification in drug discovery, Anal Biochem, 332
(2004) 153-159.
[17] C.H. Luan, S.H. Light, S.F. Dunne, W.F. Anderson, Ligand screening using fluorescence thermal
shift analysis (FTS), Methods Mol Biol, 1140 (2014) 263-289.
[18] M. Flipo, N. Willand, N. Lecat-Guillet, C. Hounsou, M. Desroses, F. Leroux, Z. Lens, V. Villeret, A.
Wohlkönig, R. Wintjens, T. Christophe, H. Kyoung Jeon, C. Locht, P. Brodin, A.R. Baulard, B. Déprez,
Discovery of novel N-phenylphenoxyacetamide derivatives as EthR inhibitors and ethionamide
boosters by combining high-throughput screening and synthesis, J. Med. Chem., 55 (2012) 6391-
6402.
[19] W. Weber, R. Schoenmakers, B. Keller, M. Gitzinger, T. Grau, M. Daoud-El Baba, P. Sander, M.
Fussenegger, A synthetic mammalian gene circuit reveals antituberculosis compounds, Proc Natl
Acad Sci U S A, 105 (2008) 9994-9998.
[20] N. Willand, B. Folléas, C. Boutillon, L. Verbraeken, J.-C. Gesquière, A. Tartar, B. Deprez, Efficient,
two-step synthesis of N-substituted nortropinone derivatives, Tetrahedron Lett., 48 (2007) 5007-
5011.
[21] A.P. Mityuk, A.V. Denisenko, O.P. Dacenko, O.O. Grygorenko, P.K. Mykhailiuk, D.M. Volochnyuk,
O.V. Shishkin, A.A. Tolmachev, An Approach to Azabicyclo[n.3.1]alkanes by Double Mannich
Reaction, Synthesis, 2010 (2010) 493-497.
[22] M.A. Brimble, C. Brocke, Efficient Synthesis of the Azabicyclo[3.3.1]nonane Ring System in the
Alkaloid Methyllycaconitine Using Bis(alkoxymethyl)alkylamines as Aminoalkylating Agents in a
Double Mannich Reaction, Eur. J. Org. Chem., 2005 (2005) 2385-2396.
[23] A.F. Abdel-Magid, S.J. Mehrman, A Review on the Use of Sodium Triacetoxyborohydride in the
Reductive Amination of Ketones and Aldehydes, Org. Process Res. Dev., 10 (2006) 971-1031.
[24] Y. Zhao, W. Huang, L. Zhu, J. Hu, Difluoromethyl 2-pyridyl sulfone: a new gem-difluoroolefination
reagent for aldehydes and ketones, Org. Lett., 12 (2010) 1444-1447.
[25] S.J. McGinty, A. Finch, R. Griffith, R.M. Graham, J.B. Bremner, Synthesis and biological evaluation
of bicyclic and tricyclic substituted nortropane derivatives: discovery of a novel selective alpha1D-
adrenergic receptor ligand, Bioorg. Med. Chem., 12 (2004) 5639-5650.
[26] N.A. Meanwell, Synopsis of some recent tactical application of bioisosteres in drug design, J.
Med. Chem., 54 (2011) 2529-2591.
[27] W. Kabsch, XDS, Acta Crystallogr D Biol Crystallogr, 66 (2010) 125-132.
[28] A. Vagin, A. Teplyakov, MOLREP: an Automated Program for Molecular Replacement, J. Appl.
Cryst., 30 (1997) 1022-1025.
[29] G.N. Murshudov, P. Skubák, A.A. Lebedev, N.S. Pannu, R.A. Steiner, R.A. Nicholls, M.D. Winn, F.
Long, A.A. Vagin, REFMAC5 for the refinement of macromolecular crystal structures, Acta Crystallogr
D Biol Crystallogr, 67 (2011) 355-367.
[30] P. Emsley, K. Cowtan, Coot: model-building tools for molecular graphics, Acta Crystallogr D Biol
Crystallogr, 60 (2004) 2126-2132.
[31] F. Long, R.A. Nicholls, P. Emsley, S. Graǽulis, A. Merkys, A. Vaitkus, G.N. Murshudov, AceDRG: a
stereochemical description generator for ligands, Acta Crystallogr D Struct Biol, 73 (2017) 112-122.
[32] I.J. Tickle, Statistical quality indicators for electron-density maps, Acta Crystallogr D Biol
Crystallogr, 68 (2012) 454-467.
[33] V.B. Chen, W.B. Arendall, J.J. Headd, D.A. Keedy, R.M. Immormino, G.J. Kapral, L.W. Murray, J.S.
Richardson, D.C. Richardson, MolProbity: all-atom structure validation for macromolecular
crystallography, Acta Crystallogr D Biol Crystallogr, 66 (2010) 12-21.