identification of inhibitors targeting mycobacterium ... · 1 identification of inhibitors...
TRANSCRIPT
1
Identification of inhibitors targeting Mycobacterium tuberculosis cell
wall biosynthesis via dynamic combinatorial chemistry
Jian Fu,a Huixiao Fu,a Marc Dieu,b Iman Halloum,c Laurent Kremer,c Yufen Xia,d Weidong Pan,d and Stéphane P. Vincenta*
a Département de Chimie, Laboratoire de Chimie Bio-Organique, University of Namur (FUNDP), rue de
Bruxelles 61, Namur B-5000, Belgium
E-mail: [email protected]
b MaSUN, Mass Spectrometry Facility, University of Namur, rue de Bruxelles 61, 5000 Namur, Belgium
c Laboratoire de Dynamique des Interactions Membranaires Normales et Pathologiques, CNRS UMR 5235,
Université de Montpellier, France
d The Key Laboratory of Chemistry for Natural Products of Guizhou Province and Chinese Academy of
Sciences, 3491 Baijin Road, Guiyang 550014, China
Electronic Supplementary Material (ESI) for Chemical Communications.This journal is © The Royal Society of Chemistry 2017
2
Table of Contents
Materials S3
Synthetic chemistry, NMR spectra of synthetic products S4-S12
HPLC purity evaluation S13-S14
Overexpression of the three UGMs S15-S16
Stability study S17-S18
DCC results (HPLC) S19-S22
LC-MS analysis S23-S28
Combinatorial in situ FP screening S29-S31
Biochemical evaluations S32-S33
Kinetic assay of H3+A1 S34
Computational study S35-S38
In vitro anti-tubercular activity S39
References S39
3
Materials and methods:
Reagents and chemicals were purchased from Sigma-Aldrich or Acros at ACS grade and used without
purification. Yields refer to chromatographically and spectroscopically homogeneous materials. All
reactions were monitored by thin-layer chromatography (TLC) carried out on Merck aluminum roll silica
gel 60-F254 using UV light and a phosphomolybdic acid solution as revelator. Merck silica gel (60, particle
size 40-63 µm) was employed for flash column chromatography. NMR spectra were recorded on a JEOL
ECX 400 or 500 with solvent peaks as reference. All compounds were characterized by 1H and 13C NMR as
well as by 1H-1H and 1H-13C correlation experiments when necessary. The following abbreviations are used
to describe the multiplicities: s= singlet, d= doublet, t= triplet, q= quartet, m= multiplet, br= broad. The
numbering of the protons and carbons is illustrated in the Scheme below. Aromatic, benzyl and methyl
(carbons and protons) are respectively labeled with “Arom”, “CH2Bn”, quaternary carbons are indicated
with a “q” superscript. Chemical shifts (δ) are reported in ppm and referenced indirectly to residual solvent
signals. Melting points are uncorrected and recorded using BÜCHI Melting Point Apparatus B-545.
Multimode Detector DTX 880 Beckmann Coulter, 384 well black microtiter plates (Corning), Amicon® Ultra
Centrifugal Filter (MWCO 10 kDa), VIVASPIN 500 (MWCO 10 kDa), Waters 600 E, C18 Atlantis T3 column,
5 μM 4.6 x 250 mm, common laboratory apparatus in biology laboratory.
4
Chemical synthesis: general procedures
Method 1 (based on a literature procedure):1 To a solution of hydrazide H3 (1 mmol) in absolute ethanol
(5 mL) containing one drop of 37% hydrochloric acid was added 1 mmol of aldehyde. The mixture was
stirred at room temperature for 2 hours. The completion of the reaction was confirmed by TLC. Next, the
solvent was partially concentrated at reduced pressure, and the resulting mixture was poured into an ice/water mixture. After neutralization with 10% aqueous sodium bicarbonate solution, the precipitate
formed was filtered out and dried under vacuum to give the desired compound as brown solid.
Method 2 (based on a literature procedure):2 to a solution of hydrazide H4, H5 ( prepared as previously
described)3 and H6 (1 mmol) in absolute ethanol (5 mL) was added 1 mmol of aldehyde. The mixture was
refluxed at 80°C for 6 hours, and the completion of the reaction was confirmed by TLC. Next, the solvent
was partially concentrated at reduced pressure, the resulting precipitate was filtered out and dried under
vacuum to give the desired acylhydrazones.
5
Specific synthetic procedures
H3+A1: To a solution of Girard’s reagent T (139.1 mg, 1 equiv., 0.83 mmol) in absolute ethanol (5mL)
containing one drop of 37% HCl was added 3,4-Dihydroxybenzaldehyde (120.2 mg, 1 equiv., 0.83 mmol).
The reaction mixture was stirred for 2 hours at room temperature, TLC analysis showed complete
conversion of the starting material. The resulting product was collected after evaporating the solvent, and
the desired product was crystallized from ethanol as a hygroscopic brown powder consisting of two
isomers (E/Z = 1:2.4). Mp = 217-218 °C. 1H NMR (500 MHz, DMSO-d6): δ = 11.87 (s, 1H, H-9), 9.71 (s, 1H,
phenyl-OH), 9.33 (s, 1H, phenyl-OH), 8.15 (s, 1H, E, H-7), 7.92 (s, 1H, Z, H-7), 7.19 – 6.82 (m, 3H, H-1, 2 ,5),
4.76 (s, 2H, Z, H-11), 4.33 (s, 2H, E, H-11), 3.28 (S, 9H, 3CH3). 13C NMR (125 MHz, DMSO-d6): δ = 165.0,
159.4, 149.3, 148.5, 148.3, 146.1, 145.8, 145.8, 125.0, 124.9, 120.9, 120.4, 115.8, 113.3, 113.0, 63.4, 62.26,
53.5, 53.2. HRMS (ESI+) calcd. for C12H18N3O3 [H]+: 252.1343; found: 252.1343.
6
H4+A1: To a solution of L-Tyrosine (162.03 mg, 1 equiv., 0.83 mmol) in absolute ethanol (5mL) was added
3,4-dihydroxybenzaldehyde (120.64 mg, 1 equiv., 0.83 mmol). The reaction mixture was stirred for 6 hours
under reflux. Next, the solvent was partially concentrated at reduced pressure, and the precipitate formed
was filtered out and washed with cold ethanol, and by Et2O, dried under vacuum to give the desired
product as a light yellow powder consisting of two isomers (E/Z = 1.7:1). Mp = 185-186 °C. 1H NMR (500
MHz, DMSO-d6): δ = 8.00 (s, 1H, E, H-7), 7.78 (s, 1H, Z, H-7), 7.16 – 6.63 (m, 7H, H-1, 4, 5, 14, 15, 17, 18),
4.23 (dd, J = 5.3 Hz, J = 7.6 Hz, 1H, Z, H-11), 3.63 (q, 3J =6.4 Hz, 1H, E, H-11), 2.85 – 2.79 (m, 1H, H-12a),
2.59 – 2.54 (m, 1H, H-12b). 13C NMR (125 MHz, DMSO-d6): δ 175.0 (q, C=O), 170.6, 155.7, 155.6, 147.9,
147.8, 147.3, 145.7, 143.6, 130.3, 130.2, 128.8, 128.5, 125.7, 120.4, 120.0, 115.5, 115.0, 114.9, 112.6,
112.4, 56.0, 52.6. HRMS (ESI+) calcd. for C16H19N3O4 [H]+: 316.1292; found: 316.1294.
7
8
H5: To a solution of Methyl 2-naphthoate (372.4 mg, 1 equiv., 2 mmol) in absolute ethanol (2 mL), was
added hydrazine hydrate (20 mL, 10 equiv., 20 mmol) dropwise. The reaction mixture was stirred for 4
hours under reflux. Completion of the reaction was confirmed by TLC. The reaction mixture was cooled
down to RT, the precipitate was filtered and washed with cold ethanol to afford H5 (315 mg, 85%) as a
yellowish amorphous powder. Mp = 151-152 °C. 1H NMR (500 MHz, DMSO-d6): δ 9.94 (s, 1H, NH); 8.41 (s,
1H, Nap-1-H); 8.00-7.90 (m, 4H, Nap-3-H, -4H, -5H, -8H); 7.59-7.56 (m, 2H, Nap-6-H, Nap-7-H); 4.57 (s, 2H,
NH2). 13C NMR (125 MHz, DMSO-d6): δ165.9 (q, C=O), 134.1 (q, Ar), 132.2 (q, Ar), 130.7 (q, Ar), 128.9 (CH,
Ar), 127.9 (CH, Ar), 127.6 (CH, Ar), 127.5 (CH, Ar), 127.3 (CH, Ar), 126.7 (CH, Ar), 123.9 (CH, Ar). HRMS (ESI+)
calcd. for C11H11N2O [H]+: 187.0866; found: 187.0865.
9
H5+A1: To a solution of 2-naphthohydrazide (140.4 mg, 1 equiv., 0.75 mmol) in absolute ethanol (5 mL)
was added 3,4-dihydroxybenzaldehyde (103.8 mg, 1 equiv., 0.75 mmol). The reaction mixture was stirred
for 6 hours under reflux. Completion of the reaction was confirmed by TLC. The reaction mixture was then
cooled down to RT, the precipitate was filtered and washed with cold ethanol to afford H5+A1 (251 mg,
82%) as a brown amorphous powder. Mp = 247-249 °C. 1H NMR (500 MHz, DMSO-d6) δ 11.78 (s, 1H, H-9),
9.41 (s, 1H, phenyl-OH), 9.30 (s, 1H, phenyl-OH), 8.51 (s, 1H, H-12), 8.31 (s, 1H, H-5), 8.05-7.98 (m, 4H, H-
14, 15, 16, 17), 7.67-7.62 (m, 2H, H-19, 20), 7.27 (s, 1H, H-7), 6.94 (d, 1H, J = 8.0 Hz, H-2), 6.80 (d, 1H, J =
8.0 Hz, H-1). 13C NMR (125 MHz, DMSO-d6) δ162.8 (q, C=O), 148.4 (q), 148.1 (CH), 145.8 (q), 134.3 (q),
132.1 (q), 130.9 (q), 128.9 (CH), 128.0 (CH), 127.8 (CH), 127.7 (CH), 126.9 (CH), 125.8 (q), 124.3 (CH), 120.6
(CH), 115.6 (CH), 112.7 (CH). HRMS (ES+) calcd. for C18H15N2O3 [M+H]+: 307.1083; found: 307.1077.
10
11
H6+A1: To a solution of 2-thiophenecarboxylic acid hydrazide (142 mg, 1 equiv., 1 mmol) in absolute
ethanol (5 mL) was added 3,4-dihydroxybenzaldehyde (138 mg, 1 equiv., 1 mmol). The reaction mixture
was stirred for 6 hours under reflu. Completion of the reaction was confirmed by TLC. The reaction mixture
was then cooled down to RT, the precipitate was filtered and washed with cold ethanol to afford H6+A1
as a pale solid consisting of two isomers (E/Z = 1:1.5, 235 mg, 90%). Mp = 231-233 °C. 1H NMR (500 MHz,
DMSO-d6) δ 11.66 (s, 1H, H-9), 9.47 (s, 1H, phenyl-OH), 9.32 (s, 1H, phenyl-OH), 8.24 (s, 1H, E, H-7), 8.04–
7.84 (m, 2H), 7.92 (s, 1H, Z, H-7), 7.30-6.70 (m, 4H). 13C NMR (125 MHz, DMSO-d6) δ161.0 (q, C=O), 157.5,
148.1, 148.0, 145.8, 144.7, 134.8, 134.5, 133.3, 131.6, 128.6, 128.2, 126.7, 125.6, 120.8, 120.6, 115.7,
115.6, 113.2, 112.6. HRMS (ESI+) calcd. for C18H14N2O3 [H]+: 263.0485; found: 263.0484.
12
13
HPLC assessment of compound purity
All tested compounds with a purity of > 95% (HPLC analysis) were used for biological study. We provided
the spectra of HPLC assays as below.
Column: C18 Atlantis T3 column, 5 μm 4.6 x 250 mm;
Mobile phase: MeCN / H2O (45 : 55, V/V);
Wavelength: 254 nm;
Rate: 1 mL/min;
Temperature: 25°C.
H3+A1
98.83%
H4+A1
96.57%
14
H6+A1
99.64%
H5+A1
97.45%
15
Biochemical part:
1) UGM preparation
Expression and purification of UGM from Mycobacterium tuberculosis
A vector construct (pET-29b) containing the gene encoding for UGM from Mycobacterium tuberculosis was
provided by Prof. Laura L. Kiessling. This construct was transformed into BL21(DE3) E.coli cells.
Transformed cells were grown in Terrific Broth and 50 μg/mL kanamycin at 37°C, culture overnight without
induction. Cells were harvested by centrifugation at 6000 rpm for 30 min at 4°C and the pellet was
resuspended in the lysis buffer (20 mM sodium phosphate, 25 mM imidazole, 500 mM NaCl, pH 7.4). The
disruption of the cells was achieved by lyzozyme, Triton X-100, and sonication. Lysed cells were centrifuged
at 16 000 rpm for 50 min at 4°C. The protein was purified by hexahistidine-Ni2+-nitrilotriacetic acid affinity
chromatography. After loading of the soluble fraction, the column was washed with a 50 mM phosphate
buffer containing 300 mM NaCl and 20 mM imidazole (pH 8). The elution of UGM was made by a linear
gradient (0-50%) to 50 mM sodium phosphate buffer (pH 8) containing 300 mM NaCl and 250 mM
imidazole. Fractions containing UGM were pooled and dialyzed overnight against 20mM sodium
phosphate buffer (pH 7) at 4°C. The purity was estimated by SDS-PAGE and the concentration was
measured by absorbance on a spectrophotometer (DTX 880 Multimode Detector) at 450 nm.
UGMs from E. coli and K. pneumoniae were expressed and purified following similar protocols.4
SDS-page
Yield: 65 mg in 1L culture
Standard 250kD 150kD 100kD
25kD
75kD
50kD
37kD
20kD
EcUGM
37kD
16
Yield: 101 mg in 1L culture
Yield: 11 mg in 1L culture
KpUGM
17
2) Stability study
In order to study the UGM stability in aniline-catalyzed conditions, the inhibition of UGM was performed
following the procedure described by Liu et al. and by by our group.5 All assays were performed at room
temperature using a phosphate buffer (NaH2PO4 50 mM, pH 7.0), and fresh solutions of sodium dithionite
to provide reductive conditions. The activity of the enzyme (in the presence and in the absence of an
inhibitor) is evaluated by measuring the conversion of UDP-Galf into UDP-Galp. The enzyme (12 nM
EcUGM) in phosphate buffer was first reduced with sodium dithionite (final concentration 12.5 mM) and
incubated at room temperature in the presence and in the absence of an inhibitor. The substrate UDP-
Galf (final concentration 25 μM) was added and allowed the reaction to proceed at specific times. The
reaction was stopped by quenching the samples with liquid N2. The conversion of UDP-Galf into UDP-Galp
was monitored by HPLC (Waters 600 E with a C18 Atlantis T3 column, 5 μM 4.6 x 250 mm, elution with 50
mM Triethylamine Acetic Acid pH 6.8, 0.5% CH3CN; UV detection at 262 nm and flow rate 1mL/min).
The extent of conversion was evaluated by the integration of the two peaks of the sugar nucleotides UDP-
Galf and UDP-Galp (Eq. (1))
% conversion = (Area UDP-Galp peak)/ [(Area UDP-Galp peak) + (Area UDP-Galf peak)]
By calculating the turnover of inhibited reactions compared to reactions without inhibitor
(Eq. (2))
% inhibition = [(% conversion - % conversion (inhibitor))/ % conversion] x 100
The results are as follows.
T im e (s e c o n d s )
Co
nv
ers
ion
(%
)
3 0 6 0 9 0 1 2 0 1 5 0 1 8 0
0
2 0
4 0
6 0
8 0
0 h
2 h
4 h
8 h
1 0 m M a n ilin e 8 h
Figure S1. The UGM-catalyzed conversion of UDP-Galf into UDP-Galp as a function of incubation times and presence of aniline.
The fresh solution of UGM was kept at room temperature, small samples were removed at different points in time and their UGM
activities were measured, following the aforementioned procedure. Also 10 mM aniline contained 5% DMSO was prepared,
monitoroing activity of UGM under aniline-catalyzed DCL conditions.
18
En
zy
me
ac
tiv
ity
(%
)
0h
2h
4h
8h
10m
M a
nilin
e 8
h
0
2 0
4 0
6 0
8 0
1 0 0
1 2 0
0 h
2 h
4 h
8 h
1 0 m M a n ilin e 8 h
T im e (h )
Figure S2. Relative enzyme activity of UGM. The activity slightly decreased during 8h, as well as in the presence of 10 mM
aniline. Each point was determined in triplicate experiments.
Figure S2 indicated that the UGM catalytic activity slightly decreased after keeping the EcUGM at room
temperature for 8 hours compared to fresh one. The residual activity of UGM incubated with 10 mM aniline for 8
hours, was still 80% of the original activity. Based on this small loss of activity, we considered that UGM is stable
long enough and in presence of aniline, two important criteria to set-up DCC conditions.
19
DCC study
AU
0.00
0.04
0.08
Minutes
0.00 6.00 12.00 18.00 24.00 30.00
Figure S3 Dynamic combinatorial library control assay DCC time-dependency study (in the presence and absence of aniline).
Scheme S1 Flow chart of the pre-equilibrated and adaptive DCC
Strategy 1: pre-equilibrated DCC in the presence of aniline. Four hydrazides H1-4 (4 × 5 µL, 100 mM,
DMSO), aldehyde A1 (1 µL, 100 mM, DMSO) and aniline (10 µL, 1 M, DMSO) were added to a mixture of
20
DMSO (19 µL) and ammonium acetate buffer (960 µL, 50 mM, pH 6.2). The DCL was allowed to stay
overnight (14 h) at room temperature, subsequently divided into two aliquots. Were then added 500 µL
EcUGM (60 µM) and 500 µL buffer.
The DCL was allowed to stand at room temperature, with occasional shaking for 6 h. The pH of the sample
was raised to 8 by the addition of NaOH, and the protein was removed by ultrafiltration using a 10,000
MWCO filter (Vivaspin). HPLC analysis was performed and the traces were compared with the experiment
performed without enzyme.
AU
0.00
0.02
0.04
Minutes
5.00 10.00 15.00 20.00 25.00
Figure S4 HPLC chromatography of DCC (pre-equilibrated DCC)
Strategy 2 Adaptive DCC condition: Four hydrazides H1-4 (4 × 5 µL, 100mM, DMSO), aldehyde A1 (1 µL, 100 mM, DMSO) and aniline (10 µL, 1 M, DMSO) were added to a mixture of DMSO (19 µL) and ammonium
acetate buffer (960 µL, 50 mM, pH 6.2). Divided into 2 aliquots, then added 500 µL EcUGM (60 µM) and
500 µL buffer. The DCL was allowed to stand at room temperature, with occasional shaking for 6 h. The
pH of the sample was raised to 8 by the addition of NaOH, and the protein was removed by ultrafiltration using a 10,000 MWCO filter (Vivaspin). HPLC analysis was performed and the traces were compared with
the experiment performed without enzyme.
AU
0.000
0.002
0.004
Minutes
12.00 15.00 18.00 21.00 24.00 27.00
Figure S5 Area of peaks in the presence and absence of EcUGM (1st, adaptive-DCC).
21
AU
0.000
0.003
0.006
Minutes
12.00 15.00 18.00 21.00 24.00 27.00
Figure S6 Area of peaks in the presence and absence of EcUGM (2nd, adaptive-DCC)
HPLC condition: C18 Atlantis T3 column, 5 μm 4.6 x 250 mm; flow rate, 1 mL min-1; wavelength, 254 nm,
gradient, MeCN / H2O (0.01% TFA) from 5% to 15% over 10 min, then to 25% over 10 min, and eventually
to 40% over 20 min)
Enlarge the scale
Eleven hydrazides/hydrazines H2-12 (11 × 5 µL, 100mM, DMSO), aldehyde A1 (5 µL, 100 mM, DMSO) and
aniline (10 µL, 1 M, DMSO) were added to an ammonium acetate buffer (930 µL, 50 mM, pH 6.2). Divided
into 2 aliquots, then added 500 µL EcUGM (60 µM) and 500µL buffer. The DCL was allowed to stand at
room temperature, with occasional shaking for 6 h. The pH of the sample was raised to 8 by the addition
of NaOH, and the protein was removed by ultrafiltration using a 10,000 MWCO filter (Vivaspin). HPLC
analysis was performed and the traces were compared with the experiment performed without enzyme.
22
AU
0.00
0.20
0.40
Minutes
0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00 90.00
Figure S7 The reaction of aldehyde with a library of 11 hydrazides/hydrazines (top). HPLC chromatogram (bottom). Blue, in the
presence of UGM; Black, in the absence of UGM. H3+A1 and H5+A1 significantly increased (red asterisk). HPLC condition: C18
Atlantis T3 column, 5 μm 4.6 x 250 mm; flow rate, 1 mL min-1; wavelength, 254 nm, gradient, MeCN / H2O (0.01% TFA) from 5%
to 20% over 40 min, then to 50% over another 40 min.
23
LC-MS analysis
LC-MS analyses were realized with a maXis Impact UHR-TOF (Bruker, Germany) coupled with an UltiMate
3000 nano-UPLC system (Thermo, USA). The mixture was separated by reverse-phase liquid
chromatography using a 1mm × 150 mm reverse phase Thermo column (Acclaim PepMap 100 C18, Thermo,
USA) in an Ultimate 3000 liquid chromatography system. The flow rate was 30 µL/min. Mobile phase A
was water containing 0.1% formic acid (FA). Mobile phase B was 20% water−80% acetonitrile containing
0.1% FA. The mobile phase was increased linearly from 5 to 20% B in 40 min and from 20 to 50% B in 40
min, then washed with 95% B for 5 min, and equilibrated with 5% B for 15 min, for a total of 100 min.
Electrospray ionization source; positive mode; Scan between 90 m/z and 600 m/z.
The components were clearly identified by LC-MS, except H3+A1, H8+A1 and H9+A1 because MS
acquisition started from 8min to avoid contamination and impurities (compounds H3+A1 and H8+A1 are
polar molecules). However, the retention of each compound was confirmed by injecting the synthetic pure
compounds.
24
25
26
27
28
29
Combinatorial in situ FP screening
5-methoxyanthranilic acid as aniline surrogate
In our first attempts of in situ screening, aniline was used for promoting (acyl)hydrazones formation.
We started with five hydrazides H1-5 and seven aldehydes A1-7. In each well, to a solution of hydrazides
(7.5 µL, 1.0 equiv., 4 mM in 2 % DMSO in buffer: NaH2PO4, pH= 6.2, 50mM) and aldehydes (7.5 µL, 1.0
equiv., 4 mM in 2 % DMSO in buffer) was added 5 µL aniline (60 mM in 2 % DMSO in buffer, 10.0 equiv.).
The reaction mixture was left at room temperature for 4 h, with occasional shaking. Subsequently, the
fluorescent probe (5µL, 90nM) and EcUGM (5µL, 3 µM) was added in each well. The microtiter plate was
directly read by using a microplate reader. The inhibition percentage was given by data were fitted to
equation.
(Eq. (3))
Inhibition % = [100-(FP signal for the inhibitor / FP signal without inhibitor)]*100
Figure S8. Preliminary in situ screening against EcUGM
We observed that the combinations made up of aldehydes A1 or A6 exhibited higher inhibition than other
aldehydes as depicted in Fig. S8. In contrast, and very strangely, no specific hydrazide seem to play a pivotal role
in the inhibition process, which could lead us to the conclusion that the inhibition was almost exclusively due to
the aldehyde subunit. We thus hypothesized that the intermediate imine generated between aldehyde An and
0
20
40
60
80
100
A 1 A 2 A 3 A 4 A 5 A 6 A 7
Inh
ibit
ion
%
Inhibiton of in situ screening against EcUGM
30
aniline could produce some UGM inhibition (or binding), which means that the imine, or its hemiaminal
intermediate (Scheme S2), could behave as UGM inhibitor. We could first show that, under these specific (dynamic)
conditions, the presence of aniline was required to observe some UGM inhibition (no inhibition observed in
absence of aniline). We then chemically synthesized the individual imine products that had shown some apparently
“strong” inhibition. Once individually assayed against UGM, the “best” imines only displayed inhibition at high
concentrations (1 mM or 500 μM), whereas no inhibition was observed at low concentration. This is the reason
why we coined this phenomenon “false positive” result. In the context of the in situ screening, the aniline catalyst
clearly participated to the inhibition process, but without leading to the identification of a potent UGM inhibitor.
Scheme S2 Imine formation
Therefore, we concluded that aniline could interfere with this in situ screening and we tried to find some alternative
catalysts. A survey of the literature lead us to try 5-methoxyanthranilic acid. In the aniline-catalyzed
oxime/hydrazone formations, Kool et al. 6demonstrated that the intermediate-imine was found to be moderately
stable, with a half-life for hydrolysis of 20 min. This relatively slow equilibration, may likely explain why we
observed false positive results.
Enlarge the scale of building blocks
The in situ screening protocol described above was followed again but with minor modification: i) the real
therapeutic target MtUGM was used for screening; ii) we performed the reaction in an eppendorf by
occasionally shaking instead of directly in the 384 well, which allowed the building blocks to react
completely. Thirteen hydrazides/hydrazines H1-13 and seven aldehydes A1-7 were chosen as building
blocks. All reactions are conducted in Eppendorf tube. In each tube, to a solution of hydrazides/hydrazines
(20 µL, 1.0 equiv., 2.25 mM in 2.25% DMSO in buffer) and aldehydes (20 µL, 1.0 equiv., 2.25 mM in 2.25%
DMSO in buffer) was added 20 µL 5-methoxyanthranilic acid (22.5 mM in 2.25% DMSO in buffer, 10.0
equiv.). The reaction mixture was left at room temperature for 4 h, with occasional shaking. Subsequently,
20 µL reaction mixtures were transferred into a 384-well plate (in triplicate), the fluorescent probe (5µL
90nM) and MtUGM (5µL 3 µM) was added in each well. In addition, UDP (500 µM) was used as the positive
control. The microtiter plate was directly read using FP plate reader. The inhibition percentage was given
by data were fitted to equation.
(Eq. (3))
Inhibition % = [100-(FP signal for the inhibitor / FP signal without inhibitor)]*100
31
Figure S9. Structure of building blocks. Inhibition percentage against MtUGM at 0.5 mM for each combination by FP approach.
A1
A3
A5
A7
20
30
40
50
60
70
80
90
100
H1 H2 H3 H4 H5 H6 H7 H8 H9 H10 H11 H12 H13
Inh
ibitio
n %
A1
A2
A3
A4
A5
A6
A7
UDP
73%
81% 88%
32
UGM binding assays With the enzymes and compounds in hand, we were keen to carry out a biological evaluation. The assay
described by Kiessling et al. was strictly followed, including the synthesis of the fluorescent probe (UDP-
fluorescein).7
To determine the binding affinity of fluorescent probe towards MtUGM, serial dilutions of dialyzed UGM
(final concentration: 1x10-5 to 12 μM) were incubated with 15 nM of the fluorescent probe in 50 mM
sodium phosphate buffer, pH 7.0 at room temperature. Final volumes were 30 μl in 384 well black
microtiter plates and the measurements were realized in triplicate. Fluorescence polarization was
analyzed using DTX880 Multimode Detector Beckman-Coulter device (λexcitation = 485 nm,λemission = 535 nm).
Data were fitted to equation (Eq. (4)) with Prism 5 GraphPad Software.
y = FPmin + (FPmax – FPmin)*1/ (1+10^ (log Kd – x)*slope
with y = fluorescence polarization
FPmin = minimal fluorescence polarization signal
FPmax = maximum fluorescence polarization signal
Kd = dissociation constant
x = log [UGM]
-6 -4 -2 0 2
1 0 0
2 0 0
3 0 0
4 0 0
mil
lip
ola
riz
ati
on
un
its
A ff in ity o f K p U G M fo r f lu o re s c e n t p ro b e
K i= 0 .5 4 μ M
lo g [U G M ]
-6 -4 -2 0 2
1 0 0
2 0 0
3 0 0
4 0 0
lo g [U G M ]
mil
lip
ola
riz
ati
on
un
its
A ff in ity o f M tU G M fo r f lu o re s c e n t p ro b e
K i= 0 .3 9 μ M
Figure S10. Ki determination for the fluorescent probe with KpUGM, MtUGM respectively.
Fluorescence polarization binding assays were performed with EcUGM, KpUGM and MtUGM. Serial
dilutions of the inhibitor (final concentrations from 0 μM to 1 mM) and 15 nM of the fluorescent probe
were mixed in 50 mM phosphate buffer pH 7.0 at room temperature. UGM (final concentration of EcUGM
and KpUGM are 500 nM, MtUGM is 580 nM) was added to start the experiment. Final volumes were 30 μl
in 384 well black microtiter plates and the measurements were realized in triplicate. Fluorescence
polarization was analyzed using DTX880 Multimode Detector Beckman-Coulter device (λexcitation = 485
nm,λemission = 535 nm).
Determination of Kd :
Data were fitted to equation (Eq. (5)) with Prism 5 GraphPad Software.
y = FPmin + (FPmax – FPmin)*1/(1+10^(x-log Kd)
log Kd = log (10^ (logKi * (1+Cf/ Kd f)))
with y = fluorescence polarization,
FPmin = minimal fluorescence polarization signal
FPmax = maximum fluorescence polarization signal
Kd = dissociation constant
x = log [UGM]
Ki = inhibition constant
Cf = concentration of the fluorescent probe
33
C o m p e tit io n H 3 + A 1 / f lu o re s c e n t p ro b e
-6 -4 -2 0 2 4
1 0 0
2 0 0
3 0 0
lo g [H 3 + A 1 ]
mil
lip
ola
riz
ati
on
un
its
E c U G M
K d = 4 .6 μ M
C o m p e tit io n H 3 + A 1 / f lu o re s c e n t p ro b e
-6 -4 -2 0 2 4
5 0
1 0 0
1 5 0
2 0 0
lo g [H 3 + A 1 ]
mil
lip
ola
riz
ati
on
un
its
K p U G M
K d = 1 3 μ M
C o m p e tit io n H 3 + A 1 / f lu o re s c e n t p ro b e
-6 -4 -2 0 2 4
1 0 0
2 0 0
3 0 0
lo g [H 3 + A 1 ]
mil
lip
ola
riz
ati
on
un
its
M tU G M
K d = 2 .9 μ M
Figure S11. Kd determination for compound H3+A1 against EcUGM, KpUGM and MtUGM respectively.
C o m p e tit io n H 4 + A 1 / f lu o re s c e n t p ro b e
-6 -4 -2 0 2 4
1 0 0
2 0 0
3 0 0
lo g [H 4 + A 1 ]
mil
lip
ola
riz
ati
on
un
its
E c U G M
K d > 1 m M
C o m p e tit io n H 4 + A 1 / f lu o re s c e n t p ro b e
-6 -4 -2 0 2 4
5 0
1 0 0
1 5 0
lo g [H 4 + A 1 ]
mil
lip
ola
riz
ati
on
un
its
K p U G M
K d > 1 m M
C o m p e tit io n H 4 + A 1 / f lu o re s c e n t p ro b e
-6 -4 -2 0 2 4
1 0 0
2 0 0
3 0 0
lo g [H 4 + A 1 ]
mil
lip
ola
riz
ati
on
un
its
M tU G M
K d > 1 m M
Figure S12. Kd determination for compound H4+A1 against EcUGM, KpUGM and MtUGM respectively.
C o m p e tit io n H 5 + A 1 / f lu o re s c e n t p ro b e
-6 -4 -2 0 2 4
1 0 0
2 0 0
3 0 0
lo g [H 5 + A 1 ]
mil
lip
ola
riz
ati
on
un
its
E c U G M
K d = 3 7 7 μ M
C o m p e tit io n H 5 + A 1 / f lu o re s c e n t p ro b e
-6 -4 -2 0 2 4
2 0
4 0
6 0
8 0
1 0 0
lo g [H 5 + A 1 ]
mil
lip
ola
riz
ati
on
un
its
K p U G M
K d = 6 3 1 μ M
C o m p e tit io n H 5 + A 1 / f lu o re s c e n t p ro b e
-6 -4 -2 0 2 4
1 0 0
2 0 0
3 0 0
lo g [H 5 + A 1 ]
mil
lip
ola
riz
ati
on
un
its
M tU G M
K d = 5 4 0 μ M
Figure S13. Kd determination for compound H5+A1 against EcUGM, KpUGM and MtUGM respectively.
C o m p e tit io n H 6 + A 1 / f lu o re s c e n t p ro b e
-6 -4 -2 0 2 4
1 0 0
2 0 0
3 0 0
lo g [H 6 + A 1 ]
mil
lip
ola
riz
ati
on
un
its
E c U G M
K d > 1 m M
C o m p e tit io n H 6 + A 1 / f lu o re s c e n t p ro b e
-6 -4 -2 0 2 4
5 0
1 0 0
1 5 0
lo g [H 6 + A 1 ]
mil
lip
ola
riz
ati
on
un
its
K p U G M
K d > 1 m M
C o m p e tit io n H 6 + A 1 / f lu o re s c e n t p ro b e
-6 -4 -2 0 2 4
1 0 0
2 0 0
3 0 0
lo g [H 6 + A 1 ]
mil
lip
ola
riz
ati
on
un
its
M tU G M
K d > 1 m M
Figure S14. Kd determination for compound H6+A1 against EcUGM, KpUGM and MtUGM respectively.
34
Kinetic assay of H3+A1
In this assay, MtUGM activity was performed by HPLC according to the following procedure, adapted from
previous work.4, 8 For reactions in the presence of the inhibitor H3+A1, the enzyme was reduced and
incubated with the inhibitor before allowing it to react with the substrate. A final MtUGM concentration
of 60 nM in sodium phosphate buffer (50 mm, pH 7.0) was reduced by the addition of sodium dithionite
(12.5 mM) and in presence of H3+A1 (300 µM, 150 µM, 100µM, 50 µM, 25 µM, 0 µM, each of them
contains 3% DMSO) at room temperature for 5 minutes. The substrate, UDP-Galf (concentration ranging
from 200 µM to 10 µM), was then added, and the reaction was allowed to proceed and quenched in liquid
N2 at 60, 120, 180 seconds.
Injections of the samples were realized by HPLC (Waters 600 E with a C18 Atlantis T3 column, 5 µM 4.6 x
250 mm, elution with a 50 mM triethylammonium acetate buffer, pH 6.8, 0.5% CH3CN; detection at 262
nm; flow rate 1mL/min. The area under the substrate and product peaks was recorded, and the conversion
percentage of substrate was calculated by using the following formula :
[Eq. (1)]: %conversion = (Area UDP-Galp peak)/ [(Area UDP-Galp peak) + (Area UDP-Galf peak)] x 100).
The quantity of UDP-Galp was calculated: [S] corresponding × conversion%. Then the data fitting in Excel
to obtain the slope as rate-1 (s/µM)
The plots of initial rates versus substrate concentrations were caculated by using Origin8 (Microcal
Software, Northhampton, MA) and GraphPad Prism software (GraphPad Software, San Diego, CA)
Figure S15. Lineweaver-Burk plot, showing 1/rate (y-axis) versus 1/ [UDP-Galf] (x-axis), in presence of the inhibitor at various
concentrations (0 to 300 µM). Assayed on MtUGM in reduced conditions. Each point on the Lineweaver-Burk plot corresponds
to reactions performed at three distinct times (60, 120 and 180 seconds).
35
Computational study
Molecular docking was performed in an attempt to determine the interaction network between the
synthesized compounds and the protein. To model the binding mode of these compounds, we performed
a docking study with the UGM in complex with substrate UDP-Galp (PDB ID: 4RPH, reduced form), obtained
from the PDB database.9 Molecular modeling studies were carried with Discovery Studio 4.0 and AutoDock
Vina software.10 Examination of the resulting model indicated that H3+A1 (E isomer) adopted the same
binding mode in the active site.
It’s notable that the docked H3+A1 occupied the 3 main binding subsites, from the FAD pocket to the
uridine pocket, the orientation approximately meshes with substrate UDP-Galp, although the size of
H3+A1 is smaller than UDP-Galp (Figure S18). Indeed, the linkage of acylhydrazone would stack in the
pyrophosphate pocket whereas the trimethylamine would point toward the FAD cofactor next to the
carbohydrate binding site. The predicted poses also highlighted potential interactions of the inhibitor with
UGM key catalytic residues such as Arg 292, Arg 180, Tyr 328, Tyr 366, Tyr 161 and the FAD cofactor (Figure
S16 and S17).
Also, it is interesting to note that H3+A1 has two isomers, then the Z isomer was also modeled, with
essentially the same binding mode was observed, as well as the key residues (Arg 292, Arg 180, Tyr 328,
Tyr 366, Tyr 161 and the FAD cofactor) of UGM interact with H3+A1 (Figure S19). In addition, the
overlapping of E and Z isomer indicates that they perhaps synergistic bind to UGM to enhance binding
affinity (Figure S20).
Figure S16. Selected docking conformations of compound H3+A1 (E isomer, magenta). The best pose was depicted in the UGM
binding site, as well FAD cofactor (gray), the residue interaction d < 4Å
36
Figure S17.Schematic interaction map of H3+A1 and MtUGM (d < 4Å).
Figure S18. Overlay with compound H3+A1 (E isomer, magenta) and substrate UDP-Galp (cyan)
37
Figure S19. Selected docking conformations of compounds H3+A1 (Z isomer, green). The best pose was depicted in the UGM
binding site, as well FAD cofactor (gray), the residue interaction d < 4Å
Figure S20. (A) Overlay with E (magenta) and Z (green) of H3+A1. (B) Superimposition of the crystal structure substrate UDP-
Galp (cyan) and modeled structures H3+A1: E (magenta) and Z (green)
A potential allosteric site has recently been study by Pinto and Sanders,8 using an opened form11 (PDB
code: 1V0J) as a receptor. To further validate the binding mode of H3+A1, we performed a modeling study
using opened form as well. Likewise, the same superimposable orientation was observed, both isomers
were located at UGM catalytic site (Figure 21). Thus the results described here, revealed that ligand H3+A1
adopted either closed form or opened form.
A B
38
Figure S21. Overlay with E (magenta) and Z (green) of H3+A1 in opened form.
In summary, the docking study allowed to propose a binding mode for most potent inhibitor and provided
insights for further inhibitor design. The docking of compound H3+A1 into the binding pocket of UGM
showed a similar manner to a natural substrate UDP-Galp, and the key residues interaction with H3+A1
was predicted. Furthermore the best compound H3+A1 was docked to the active site of UGM with good
energy scores supporting the inhibitory activity.
As experimentally demonstrated in the manuscript, the most potent UGM inhibitor H3+A1 is a competitive
inhibitor of MtUGM (Ki = 35.3 uM). The resulting model clearly indicated that the inhibitor H3+A1 binds to the
substrate catalytic pocket of MtUGM. As suggested by the reviewer, we carried out H3+A1 (E isomer) docked with
another two UGMs: EcUGM (PDB: 1I8T, Naismith, Nat. Struct. Mol. Biol., 2001, 8, 858-863) and KpUGM (PDB: 3INT,
Forest, Biochemistry, 2009, 48, 9171-9173) using the same software and protocol as previously described.
Figure S22. Selected docking conformations of compound H3+A1 (E isomer, green) with EcUGM (PDB: 1I8T). The best pose was
depicted in the UGM binding site, as well the FAD cofactor (gray).
39
Figure S23. Selected docking conformations of compound H3+A1 (E isomer, green) with KpUGM (PDB: 3INT). The best pose was
depicted in the UGM binding site, as well the FAD cofactor (gray).
Docking results indicated that H3+A1 adopted the same global binding mode in the active sites of both EcUGM and
KpUGM, though with slightly different orientations.
Then, the molecular modeling was performed with compounds 3 and 4 and the three UGMs. The results indicate
that compounds 3 and 4 both locate in the UGM active site next to cofactor-FAD (Figures S24, S25 and S26).
Presumably, this coincidence is due to the fact that many residues involved in the FAD binding site are highly
conserved within all known bacterial UGMs.
Figure S24. Overlay with inhibitor 3 (green) and 4 (cyan) docked in EcUGM (PDB: 1I8T).
40
Figure S25. Overlay with inhibitor 3 (green) and 4 (cyan) docked in KpUGM (PDB: 3INT).
Figure S26. Overlay with inhibitor 3 (green) and 4 (cyan) docked in MtUGM (PDB: 4RPH).
Compounds H4+A1, H5+A1 and H6+A1 that could not be docked into the active site of MtUGM (4RPH). We found
that these compounds located on the surface of UGM, rather than in the active site (Figure S27).
41
Figure S27. Superimposition of the crystal structure substrate UDP-Galp (cyan) and modeled structures
H4+A1 (wheat), H5+A1 (yellow) and H6+A1 (pink).
42
In vitro anti-tubercular activity
Materials and Methods
Bacterial strains and growth conditions. M. tuberculosis mc26030 was grown at 37°C in Sauton’s medium
supplemented with 20 µg/mL of pantothenic acid.12
Drug susceptibility testing. The susceptibility of M. tuberculosis to the various compounds was determined
as reported previously.13 In brief, Middlebrook 7H10 solid medium containing oleic-albumin-dextrose-
catalase enrichment (OADC) and 20 µg/mL of pantothenic acid was supplemented with increasing
concentrations of the chemical analogues. Stock solutions at 10 mg/mL were diluted in DMSO. Serial 10-
fold dilutions of the actively growing culture were plated and incubated at 37°C for 2-3 weeks. The minimal
inhibitory concentration (MIC) was defined as the minimum concentration required to inhibiting 99% of
the growth.
Cytotoxicity assay
Cytotoxicity study was performed on A549 cells in the laboratory of Professor Carine MICHIELS by Maude
Fransolet (University of Namur, NARILIS). A cytotoxicity assay was used to assess viability of A549 cells in
the presence of H3+A1. An adequate density of A549 cells were added to wells of a sterile 24-well plate.
Cells were then treated with serially diluted H3+A1 or a vehicle control. After 24 hours, MTT solution was
added incubated for 2 hours at 37 °C in a 5% CO2 atmosphere. After discarding medium, 1mL of lysis buffer
was added and allowed to incubate for an additional 1 hour. The viability of the treated cells was analyzed
using plate reader, by measuring the OD at 570nm. Viability readings were normalized to DMSO control
wells.
[ In h ib ito r ] ( M )
% V
iab
le c
ell
s
0 1 0 0 2 0 0 3 0 0
0
2 0
4 0
6 0
8 0
1 0 0
1 2 0H 3 + A 1
Figure S28 Cytotoxicity study of H3+A1
Reference
1. A. E. Kummerle, M. Schmitt, S. V. Cardozo, C. Lugnier, P. Villa, A. B. Lopes, N. C. Romeiro, H. l. n. Justiniano,
M. A. Martins and C. A. Fraga, J. Med. Chem., 2012, 55, 7525-7545.
2. T. Sedaghat, A. Tarassoli, Z. Ansari-Asl and H. Motamedi, J. Coord. Chem., 2013, 66, 2549-2557.
3. K. M. Khan, M. Rasheed, Z. Ullah, S. Hayat, F. Kaukab, M. I. Choudhary and S. Perveen, Biorg. Med. Chem.,
2003, 11, 1381-1387.
4. I. N'Go, S. Golten, A. Arda, J. Canada, J. Jimenez-Barbero, B. Linclau and S. P. Vincent, Chem. Eur. J., 2014,
20, 106-112.
5. S. El Bkassiny, I. N'Go, C. M. Sevrain, A. Tikad and S. P. Vincent, Org. Lett., 2014, 16, 2462-2465.
6. P. Crisalli and E. T. Kool, J. Org. Chem., 2013, 78, 1184-1189.
43
7. (a) E. E. Carlson, J. F. May and L. L. Kiessling, Chem. Biol., 2006, 13, 825-837; (b) M. Soltero-Higgin, E. E.
Carlson, J. H. Phillips and L. L. Kiessling, J. Am. Chem. Soc., 2004, 126, 10532-10533; (c) C. J. Frederic, A.
Tikad, J. Fu, W. Pan, R. B. Zheng, A. Koizumi, X. Xue, T. L. Lowary and S. P. Vincent, Chem. Eur. J., 2016, 22,
15913-15920.
8. Y. Shi, C. Colombo, J. R. Kuttiyatveetil, N. Zalatar, K. E. van Straaten, S. Mohan, D. A. Sanders and B. M.
Pinto, ChemBioChem, 2016, 17, 2264-2273.
9. K. E. van Straaten, J. R. Kuttiyatveetil, C. M. Sevrain, S. A. Villaume, J. s. Jimenez-Barbero, B. Linclau, S. p.
P. Vincent and D. A. Sanders, J. Am. Chem. Soc., 2015, 137, 1230-1244.
10. O. Trott and A. J. Olson, J. Comput. Chem., 2010, 31, 455-461.
11. K. Beis, V. Srikannathasan, H. Liu, S. W. Fullerton, V. A. Bamford, D. A. Sanders, C. Whitfield, M. R. McNeil
and J. H. Naismith, J. Mol. Biol., 2005, 348, 971-982.
12. V. K. Sambandamurthy, S. C. Derrick, T. Hsu, B. Chen, M. H. Larsen, K. V. Jalapathy, M. Chen, J. Kim, S. A.
Porcelli, J. Chan, S. L. Morris and W. R. Jacobs, Jr., Vaccine, 2006, 24, 6309-6320.
13. L. Kremer, J. D. Douglas, A. R. Baulard, C. Morehouse, M. R. Guy, D. Alland, L. G. Dover, J. H. Lakey, W. R.
Jacobs, Jr., P. J. Brennan, D. E. Minnikin and G. S. Besra, J. Biol. Chem., 2000, 275, 16857-16864.