Identification and Repurposing of Trisubstituted Harmine Derivatives as Novel Inhibitors of Mycobacterium tuberculosis Phosphoserine Phosphatase

Article abstract of DOI:10.3390/molecules25020415

Mycobacterium tuberculosis is still the deadliest bacterial pathogen worldwide and the increasing number of multidrug-resistant tuberculosis cases further complicates this global health issue. M. tuberculosis phosphoserine phosphatase SerB2 is a promising target for drug design. Besides being a key essential metabolic enzyme of the pathogen’s serine pathway, it appears to be involved in immune evasion mechanisms. In this work, a malachite green-based phosphatase assay has been used to screen 122 compounds from an internal chemolibrary. Trisubstituted harmine derivatives were found among the best hits that inhibited SerB2 activity. Synthesis of an original compound helped to discuss a brief structure activity relationship evaluation. Kinetics experiments showed that the most potent derivatives inhibit the phosphatase in a parabolic competitive fashion with apparent inhibition constants (K_i) values in the micromolar range. Their interaction modes with the enzyme were investigated through induced fit docking experiments, leading to results consistent with the experimental data. Cellular assays showed that the selected compounds also inhibited M. tuberculosis growth in vitro. Those promising results may provide a basis for the development of new antimycobacterial agents targeting SerB2.

Full text of DOI:10.3390/molecules25020415

molecules
Article
Identification and Repurposing of Trisubstituted
Harmine Derivatives as Novel Inhibitors of
Mycobacterium tuberculosis
Phosphoserine Phosphatase
Elise Pierson 1,*,† , Marie Haufroid 1,*,† , Tannu Priya Gosain 2, Pankaj Chopra 2
Ramandeep Singh 2 and Johan Wouters 1,*
,
1
Laboratoire de Chimie Biologique Structurale (CBS), Namur Medicine and Drug Innovation Center
(NAMEDIC), Namur Research Institute for Life Sciences (NARILIS), University of Namur (UNamur),
B-5000 Namur, Belgium
Tuberculosis Research Laboratory, Translational Health Science and Technology Institute,
Faridabad 121001, Haryana, India
2
*
Correspondence: elise.pierson@unamur.be (E.P.); marie.haufroid@unamur.be (M.H.);
johan.wouters@unamur.be (J.W.)
†
These authors contributed equally to this work.
ꢀꢁꢂꢀꢃꢄꢅꢆꢇ
ꢀꢁꢂꢃꢄꢅꢆ
Received: 17 December 2019; Accepted: 15 January 2020; Published: 19 January 2020
Abstract: Mycobacterium tuberculosis is still the deadliest bacterial pathogen worldwide and the
increasing number of multidrug-resistant tuberculosis cases further complicates this global health
issue. M. tuberculosis phosphoserine phosphatase SerB2 is a promising target for drug design. Besides
being a key essential metabolic enzyme of the pathogen’s serine pathway, it appears to be involved in
immune evasion mechanisms. In this work, a malachite green-based phosphatase assay has been
used to screen 122 compounds from an internal chemolibrary. Trisubstituted harmine derivatives
were found among the best hits that inhibited SerB2 activity. Synthesis of an original compound
helped to discuss a brief structure activity relationship evaluation. Kinetics experiments showed
that the most potent derivatives inhibit the phosphatase in a parabolic competitive fashion with
apparent inhibition constants (Ki) values in the micromolar range. Their interaction modes with the
enzyme were investigated through induced fit docking experiments, leading to results consistent
with the experimental data. Cellular assays showed that the selected compounds also inhibited
M. tuberculosis growth in vitro. Those promising results may provide a basis for the development of
new antimycobacterial agents targeting SerB2.
Keywords:
M. tuberculosis; phosphoserine phosphatase; SerB2; 2,7,9-trisubstituted
harmine derivatives
1. Introduction
Tuberculosis (TB), a disease caused by Mycobacterium tuberculosis (Mtb), is still a matter of concern
for the World Health Organization (WHO). Each year, an estimated 10 million people fall ill with TB.
Although this number has been stable in recent years and the number of casualties is slowly decreasing,
drug-resistant cases are on the rise [1–3]. To achieve its final goal, a TB-free world by 2035, the WHO
invites researchers to find new diagnostics tools, drug-targets, scaffolds and vaccines [4].
To this end, many enzymes from Mtb metabolic pathways, such as secretory tyrosine phosphatases
(MptpA, MptpB) or the inosine 50-monophosphate dehydrogenase (GuaB2), have been identified
and characterized [5–8]. Many amino acid biosynthetic pathways, including those of arginine and
Molecules 2020, 25, 415; doi:10.3390/molecules25020415
www.mdpi.com/journal/molecules

Molecules 2020, 25, 415
2 of 17
methionine, have also been proposed as promising therapeutic targets given their essential role in
bacterial growth [ 11]. One recent focus is on the serine biosynthetic pathway which leads to the
9–
production of the non-essential amino acid L-serine [12]. The latter is the source for the production of
proteins, the one-carbon metabolism, other amino acids like glycine and tryptophan and phospholipids
turnover [13–16]. Three different enzymes catalyze the three steps involved in the transformation
of 3-phosphoglycerate into L-serine. The first enzyme is a phosphoglycerate dehydrogenase (SerA,
EC 1.1.1.95) which oxidizes 3-phosphoglycerate into 3-phosphohydroxypyruvate [17]. The latter is
transformed in O-phospho-L-serine by the phosphoserine aminotransferase (SerC, EC 2.6.1.52) [18].
Finally, a phosphoserine phosphatase (SerB2, EC 3.1.3.3) catalyzes the last irreversible step,
the dephosphorylation of O-phospho-L-serine into L-serine [19].
Shree et al. [20] showed that Mtb phosphoserine phosphatase SerB2 is secreted within the cytosol
of infected macrophages and initiates a dramatic cascade of dephosphorylation. Those events lead
to the down-regulation of the immune mediator interleukin 8, which could result in the lack of
immune response from the infected host. On the other hand, dephosphorylation of cofilin induces
cytoskeletal rearrangements in the macrophage, helping invasion and intracellular survival of the
pathogen. SerB2 has further been validated as a drug target and a few inhibitors are already described
in the literature [20–22].
In this context, this work aims to repurpose molecules from our in-house chemical library and
evaluate them for their ability to inhibit SerB2’s activity [23]. The strategy to repurpose already existing
drugs and assess their efficiency on Mtb can save time in the drug discovery process and has already
led to successful outcomes [24
the treatment regimen of multidrug-resistant TB and was shown to be active against SerB2 [20
–
27]. Clofazimine, part of the regimen therapy against leprosy is now in
28 30].
,
–
Anti-inflammatory drugs, like indomethacin or other non-steroidal anti-inflammatory drugs, can
also be co-administrated with anti-TB chemotherapy to induce a synergistic effect on the infection
site [31–34].
Our chemical library is enriched with a great range of chemical scaffolds synthesized as part of
drug discovery and medicinal chemistry programs from our group [23]. A basic set of a hundred
molecules was selected and screened on purified SerB2. Compounds inhibiting the phosphatase activity
of the enzyme were considered hits. Interestingly, what emerged out of the screening process was a set
of molecules derived from harmine, a natural
Such compounds are already known to have many biological activities such as antitumor, antifungal
and antibacterial properties among others [35 37]. Their specific activity against mycobacterial strains
β-carboline alkaloid obtained from Peganum harmala.
–
and especially Mtb was assessed in a few papers [38–40]. Herein we report the inhibition of SerB2 by
trisubstituted harmine derivatives that were originally developed during an anticancer drug discovery
program [41–46]. Novel activity for those derivatives will be shown in this article.
2. Results and Discussion
2.1. Screening and Identification of Inhibitors
SerB2 is known to catalyze the hydrolysis O-phospho-L-serine into L-serine. During this reaction,
the inorganic phosphate (Pi) that is released into the medium can be photometrically determined
using a malachite green-based reagent to evaluate the activity of the enzyme [47,48]. Subsequent
enzymatic assays were performed in the presence of 122 different therapeutically intended compounds
made available by the NAMEDIC research center to identify new anti-TB agents targeting SerB2
(Figure 1A). The preliminary screening was performed at 100
µM in a pH 7.4 reaction media containing
200
µ
M O-phospho-L-serine, 5 mM Mg2+ and 5% dimethyl sulfoxide (DMSO). In total, 19 compounds
inhibited the phosphatase activity of SerB2 by at least 70% and were considered to be primary hits.
Among them were a phenyl thiazolamine derivative, two tryptophane derivatives, two coumarin
derivatives and 14 harmine derivatives. The primary hits were further tested against SerB2 at 10 µM

Molecules 2020, 25, 415
3 of 17
in three independent reactions. As depicted in Figure 1B, compounds 88
,
91 and 95 (Figure 2) are of
particular interest since they showed inhibition percentages at least twice as high as the rest of the hits.
A
B
100
90
80
70
60
50
40
30
20
10
0
100.0
80.0
60.0
40.0
20.0
0.0
94.5
82.6
70.4
36.3
16.5
25.1
29.4
15.6
15.7
15.5
11.1
2.7
1.4
0.0 0.0
51 56 66 84 104 27 39 40 86 87 88 89 90 91 92 93 94 95 105
PTA Trp Cou Harmine
0.0 0.0
0.0 0.0
0
10 20 30 40 50 60 70 80 90 100 110 120
Compounds
Figure 1.
at 100 µM in presence of 0.2 mM O-phospho-L-serine. The data depicted are averaged from two
independent experiments. ( ) Evaluation of the 19 primary hits for their ability to inhibit SerB2’s
phosphatase activity at 10 µM in presence of 0.2 mM O-phospho-L-serine. The data depicted are
(A) Evaluation of 122 compounds for their ability to inhibit SerB2’s phosphatase activity
B
averaged from three independent experiments
Cou: coumarin.
±
S.E. PTA: phenyl thiazolamine. Trp: tryptophane.
Figure 2. Structures of harmine, harmol and the three 2,7,9-trisubstituted harmine derivatives
highlighted in this study.
A closer look at the structures of the 14 primary harmine-derived hits (Table 1) revealed that
they share a common 7,9- or 2,7,9-substituted 7-oxy-1-methyl-β-carboline nucleus. The substitution
of this scaffold appears to be essential for the inhibition of SerB2, as emphasized by the inactivity of
harmine and harmol (48 and 49, Figure 2) in the screening at 100 µM. Table 1 shows the respective
R7/9/2 substituents and the inhibition percentages at 10 µM of compounds 88
, 91 and 95 and other
structurally related hits. With the exception of compound 93, all the primary hits are 2,7,9-trisubstituted,
which suggests an important role for the positive charge carried by the N2 atom. It can also be
noted that the substituents are generally bulky aliphatic or aromatic groups. In order to rationalize
their effect, we chose to describe them by their Hansch hydrophobic parameter
using Molinspiration online tool. Knowing that highly hydrophobic substituents possess a large
value, it can be observed that the inhibition efficiency increases with the hydrophobicity of the
R2 substituent in each of the three A, B and C sets (Table 1). The same trend also applies to the
size of the substituents. On the contrary, derivatives with small polar R2 groups (86
87 89) are
mostly inactive. Compound 89 however still shows moderate activity but it could be due to the R7
π [49], calculated
π
,
,

Molecules 2020, 25, 415
4 of 17
and R9 methyl-cyclohexyl substituents. The latter seem to improve the inhibition efficiency when
comparing compounds of the C set (89 90 91) with their respective counterparts of the A set (86 92,
,
,
,
94). Once again, the effect could be related to the higher hydrophobicity of the cyclohexyl ring but also
its greater flexibility.
Table 1. Structures of the common 2,7,9-substituted 7-oxy-1-methyl-
β-carboline core and inhibition
percentages at 10 µM S.E.) of SerB2’s phosphatase activity for the 7,9-disubstituted and ten
(
±
2,7,9-trisubstituted harmine derivatives that were primary hits in the screening. The Hansch
hydrophobic parameters π of each substituent were calculated using Molinspiration and are given in
the third and fifth column for R7/9 and R2 respectively. The compounds have been classified into three
sets A, B and C according to the nature of their R7 and R9 substituents.
Set Compound R7/R9
R2
π(R2) Inhibition at 10 µM (%)
39
−0.02
−2.25
0.03
0 ± 10.2
0 ± 5.6
86
휋(R7)/휋(R9)
2.13/1.67
A
92
94
95
87
0 ± 7.6
0.19
15.6 ± 4.3
82.6 ± 9.9
0 ± 12.4
1.17
−2.25
B
휋(R7)/휋(R9)
2.19/1.72
88
93
89
0.03
2.80
70.4 ± 9.1
0 ± 17.4
-
−2.25
25.1 ± 8.2
90
91
0.03
0.19
16.5 ± 13.3
94.5 ± 1.4
휋(R7)/휋(R9)
2.81/2.36
C
2.2. Design, Synthesis and Evaluation of a New Derivative
Compounds 95 and 91 showed the highest inhibition percentages of the test and both of
them possess an aromatic R2 substituent. While the difference between the effect of aromatic
and aliphatic R7/R9 substituents was highlighted in the previous paragraph, we wanted to verify
whether the aromatic feature on position 2 was required to achieve a high inhibition of SerB2. In this

Molecules 2020, 25, 415
5 of 17
view, we designed compound 124, choosing a N2-ethylcyclohexyl group (π = 1.12) to replace the
N2-ethylbenzyl group of compound 91. The two moieties have a relatively similar size, with the
ethylcyclohexyl lacking the ability to form aromatic-aromatic interactions and being more flexible.
Compound 124 was synthesized according to a published procedure for the synthesis of mono-,
di- and trisubstituted harmine derivatives [42,43]. The strategy is depicted in Scheme 1 and starts
with the demethylation of harmine to harmol in acidic conditions at reflux. Compound 123 was
obtained by a simultaneous N9/O7- alkylation of harmol in presence of 1-bromomethyl cyclohexane
and potassium hydroxide in N,N-dimethylformamide (DMF). A final N2-alkylation of compound 123
in tetrahydrofuran (THF) using 10 equivalents of 1-bromo-2-cyclohexylethane yielded compound 124
After purification by flash chromatography and characterization, the effect of compound 124
was tested at 10 µM on SerB2 using the aforementioned assay. From three independent reactions,
.
we obtained an averaged inhibition percentage of 93.5
±
1.8%. As the result for compound 91 was
94.5 ± 1.4%, it can be concluded that the aromaticity of the R2 group does not influence the extent of
inhibition and that its direct environment upon binding is probably not made of aromatic residues.
4
5
6
3
1
7
2
8
9
[31%]
[Quant.]
[50%]
Scheme 1. Synthesis of target molecule 124. (a) HBr, acetic acid, reflux, 72 h; (b) Bromomethyl
cyclohexane (2 eq), KOH (5 eq), N,N-dimethylformamide (DMF), room ◦temperature, 48 h;
(c) 1-bromo-2-cyclohexylethane (10 eq), tetrahydrofuran (THF), microwave, 155 C, 6 h.
2.3. Kinetics Assay and Determination of the Inhibition Mode
The inhibition modes of SerB2 by the three best harmine derivatives hits (88, 91, 95) and the
newly synthesized derivative 124 were investigated through enzyme kinetics experiments. Initial
velocities were measured with O-phospho-L-serine concentrations ranging from 0 to 2 mM at
fixed inhibitor concentrations. The resulting double-reciprocal Lineweaver-Burk plots are shown
in Figure 3A,C,E,G and were used to judge the binding mechanism of the selected compounds to
SerB2. For all four compounds, the double-reciprocal plots yielded straight lines intersecting on the
ordinate. These patterns suggest a competitive inhibition mechanism, with the apparent Michaelis
constant (Km) increasing with inhibitor concentration and the maximum velocity (Vmax) remaining
unchanged. As inhibitors can be further classified as linear, hyperbolic or parabolic, slopes of the
double-reciprocal plots (slopeLB) were plotted versus inhibitor concentration to gain further insight into
the inhibition mechanism (Figure 3B,D,F,H). The curved-up profiles obtained for all four compounds
reveal a parabolic inhibition of SerB2. Such behavior suggests a complex non-linear competitive
inhibition type where more than one molecule of inhibitor can bind to the enzyme [50,51].

Molecules 2020, 25, 415
6 of 17
B
A
0.030
0.025
0.020
0.015
0.010
1.5
[88]
0 µM
2.5 µM
5.0 µM
7.5 µM
10.0 µM
15.0 µM
1.0
0.5
0.0
-0.5
-20
0.0
5.0
10.0
15.0
0
20
40
60
60
60
60
1/[PS] (mM-1)
[88] (µM)
C
D
2.0
0.04
0.03
0.02
0.01
[91]
1.5
1.0
0.5
0.0
0 µM
1.5 µM
2.0 µM
2.5 µM
3.5 µM
-0.5
-20
0
20
40
0.0
1.0
2.0
3.0
4.0
1/[PS] (mM-1)
[91] (µM)
E
F
2.5
0.05
0.04
0.03
0.02
0.01
[95]
2.0
1.5
1.0
0.5
0.0
0 µM
2.5 µM
5.0 µM
7.5 µM
10.0 µM
12.5 µM
-0.5
-20
0
20
40
0.0
5.0
10.0
15.0
1/[PS] (mM-1)
[95] (µM)
H
G
2.0
0.04
0.03
0.02
0.01
[124]
0 µM
1.5
1.0
0.5
0.0
1.0 µM
1.5 µM
2.0 µM
2.2 µM
-0.5
-20
0.0 0.5 1.0 1.5 2.0 2.5
0
20
40
1/[PS] (mM-1)
[124] (µM)
Figure 3.
(A,C,E,G) Double reciprocal plots of SerB2 inhibition by compounds 88, 91, 95 and 124
at various fixed micromolar concentrations in presence of 0 to 2 mM O-phospho-L-serine (PS).
The data depicted are averaged from two independent experiments
double-reciprocal plots A, C, E and G versus inhibitor concentration
line: fitted Hill-type function (Equation (3)).
±
±
S.E.; (
B,D,F,H) Slope replots of
S.E. on slope calculations. Solid

Molecules 2020, 25, 415
7 of 17
In an effort to quantify the inhibitory effect, the inhibition constants (
K
i) of the four derivatives
were evaluated by fitting Equation (3) (see Section 4) to the data of the slope replots. This equation is
based on the Hill-type inhibition model introduced by Cao et al. [52] and used to describe parabolic
inhibition [53,54]. While the Ki parameters given by the model are rather apparent values since they
include the intrinsic constants for the binding of each molecule, they are useful for comparing the
relative potencies of the inhibitors.
Good fits were obtained for each compound and Ki values in the micromolar range were
determined with low confidence intervals (Table 2). Results show that the most potent compound
is 124, with an affinity for SerB2 almost twice as high as that of compound 91 and almost four times
as high as that of compound 95. Derivative 88 is the least efficient, which underlines once again
the importance of the hydrophobicity and bulkiness of the substituents. Hill coefficients (n) were
also calculated for the four systems and the values around 2 that were obtained suggest positive
cooperativity in the binding of the inhibitor. That is, binding of a first molecule to SerB2 would
facilitate the interaction with a second molecule by changing the affinity of the free binding site or
creating it.
Table 2. Inhibition kinetics parameters, estimated binding free energies (∆Gbind) and cellular activities
of SerB2 inhibitors 88, 91, 95 and 124. Inhibition constants (Ki) and Hill coefficient (n) values are
reported with 95% confidence intervals. Goodness of fit of the Hill-type model used (Equation (3))
are reported as R2 values in the fourth column. ∆Gbind were calculated following induced fit docking
experiments by the MM/GBSA method. Minimum inhibitory concentrations (MIC99) were determined
against M. tuberculosis strain H37Rv in vitro.
Compound
Ki (µM)
n
R2
∆Gbind (kcal/mol) MIC99(µM)
88
91
95
18.3 ± 2.3
1.45 ± 0.38 0.985
−78.18
−108.72
−99.53
6.3
1.5
3.6
0.8
3.27 ± 0.12 2.39 ± 0.45 0.996
6.66 ± 0.39 1.42 ± 0.16 0.998
1.75 ± 0.11 1.78 ± 0.53 0.993
124
−111.80
2.4. Induced Fit Docking and Binding Free Energy Calculation
Considering the competitive nature of the inhibition kinetics mechanisms studied above, it is
likely that inhibitors interact directly with the active site of the enzyme, disrupting substrate binding.
Like other phosphatases, SerB2 possesses an open and a closed conformation in order to let substrates
and products of the reaction in and out [55]. This kind of structure possesses highly dynamic properties
already shown in similar enzymes [56–58]. The SerB2 model generated by homology modeling based
on the crystal structure of Mycobacterium avium SerB (Protein Data Bank (PDB) entry 3P96) is in the
closed conformation. Rigid docking of ligands as big as harmine derivatives was thus challenging.
The Induced Fit Docking protocol was therefore chosen to take the mobility of the closing loop into
account. All four inhibitors were successfully docked within the active site of SerB2 (Figure S1) with
Glide XP scores ranging from
compound 124, is shown in Figure 4.
Analysis of those structures shows that all three substituents on the
−
7.9 to
−
13.3 kcal/mol. The docked structure of the best inhibitor,
β
-carboline core fill different
pockets within the active site. Those pockets mainly contain hydrophobic residues like leucines,
valine and isoleucines, explaining the selectivity for lipophilic substituents. Also, the substitution on
position 2 is essential for activity as discussed in Section 2.1 since it introduces a positive charge on the
N2 atom. The latter seems to form a salt bridge with the catalytic aspartate and a glutamate residue.
There are also two phenylalanine residues within the active site that can form
π-stacking and T-shape
interactions with the β-carboline core or N-substituent of other derivatives (Figure S2).

Molecules 2020, 25, 415
8 of 17
Binding affinities of the four ligands were assessed using MM/GBSA calculations (Table 2).
∆Gbind values show that 124 and 91 are the best inhibitors of SerB2 (with a slight preference for
The
124). Compound 95 has less affinity for the enzyme since its value is 10 kcal/mol higher. Finally,
compound 88 is predicted to be the less potent inhibitor of SerB2. Those calculations are in agreement
with the experimental kinetics results.
Figure 4. Structure of docked inhibitor 124 (green) within the active site of SerB2.
2.5. MIC99 Determination Assays and Time-Dependent Bactericidal/bacteriostatic Activity
In order to tackle the issue of drug-resistance, there is urgent need to identify drugs that target
novel metabolic pathways and also possess bactericidal killing activity [59]. Compounds 88
124 were first evaluated for their ability to inhibit mycobacterial growth in vitro by determining their
minimal inhibitory concentration (MIC99) against M. tuberculosis H37Rv. Values in the range 0.8–6.3
, 91, 95 and
µM
were obtained (Table 2). The results correlate well with the inhibition constants, which means that
bacterial growth inhibition may be related to SerB2 inhibition. Next, time-dependent in vitro killing
experiments were performed by exposing early-log phase cultures to 10
×
MIC99 of compounds 88,
91 95 124 and isoniazid (INH). The activity of the hits was compared to untreated mycobacterial
,
,
cultures as well as to INH exposed cultures (Figure 5). We observed that the selected compounds
inhibited the growth of M. tuberculosis in a bactericidal manner. After 6 days of treatment, the range of
killing observed in the case of the various harmine treated samples was comparable to those seen in
the case of INH treated samples. Compounds 88, 91, 95 and 124 also appeared to be equally effective
as no statistically significant difference was observed in their respective range of killing. As expected,
no killing was observed in DMSO treated samples. These observations suggest that targeting of SerB2
by harmine derivatives results in bactericidal killing in vitro.

Molecules 2020, 25, 415
9 of 17
9
8
7
6
5
4
3
2
1
0
**
***
**
**
***
4
O
2
95
88
91
UT
S
1
INH
DM
Compounds
Figure 5. Time-kill kinetics of compounds 88
H37Rv. Early logarithmic phase cultures of M. tuberculosis were exposed to 10
concentrations (MIC99) of the compounds and bacterial enumeration (expressed in colony-forming
units (CFU)) was performed by plating 10-fold serial dilutions on MB7H11 plates at day 6 post-exposure.
,
91
,
95
,
124 and isoniazid (INH) against M. tuberculosis
minimal inhibitory
×
The data shown is mean
±
S.E. of data obtained from three experiments performed in duplicates.
Statistical significance with respect to DMSO treated samples is reported as ** (p
*** (p ≤ 0.001). UT: untreated.
≤
0.01) and
3. Conclusions
In this work, a first screening of 122 compounds from our internal chemolibrary led to the
identification of a series of harmine derivatives targeting SerB2, an essential metabolic enzyme and
suspected virulence factor of M. tuberculosis. Comparison of the best harmine-derived hits revealed that
they share a common 7,9- or 2,7,9-substituted 7-oxy-1-methyl-β-carboline scaffold. A brief structure
activity relationship evaluation led to the conclusion that this kind of substitution is essential for
the inhibition of SerB2, as highlighted by the inactivity of harmine and harmol. Primary hits have
been confirmed at 10
further study.
µM concentration and three 2,7,9-trisubstituted compounds were retained for
In those three selected molecules (88, 91, 95), substituents are bulky aliphatic or aromatic groups.
In order to verify whether the R2 substituent had to be aromatic for the molecule to achieve a high
inhibition of SerB2, an original compound (124) has been designed, synthesized and tested. Compound
124, bearing a N2-ethylcyclohexyl group that replaces the N2-ethylbenzyl group of compound 91
,
showed similar activity in our enzymatic assay with inhibition at 10
µ
M of 93.5
±
1.8 and 94.5
±
1.4%
for 124 and 91, respectively. Therefore, it could be concluded that the aromaticity of the R2 substituent
does not influence the inhibition of SerB2 and that its direct environment upon binding of the inhibitor
is probably not made of aromatic residues.
Inhibition kinetics experiments were performed to further study the effect of 124 and the selected
compounds. The shapes of the double-reciprocal plots and slope replots showed that all four
compounds inhibit SerB2 in a parabolic competitive fashion with apparent Ki values in the micromolar
range. These experiments revealed that the most effective inhibitor was the newly synthesized 124 and
confirmed the importance of hydrophobic and flexible substituents over aromatics. Hill coefficients
were also calculated for the four molecules and indicate positive binding cooperativity. Along with the
parabolic behavior, this interesting finding suggests that the interaction of a first inhibitor molecule
with SerB2 facilitates the binding of a second one.

Molecules 2020, 25, 415
10 of 17
The interactions between the enzyme and the most potent derivatives were investigated by
induced fit docking experiments and binding free energy calculations. Analysis of the simulated
complexes suggests that all three substituents on the β-carboline core fill different pockets within the
active site. Those pockets are mostly made of hydrophobic residues like leucines, valine and isoleucines,
explaining the selectivity for hydrophobic substituents. Also, the third substitution appears essential
for activity since it introduces a positive charge on the nitrogen atom of the pyridinium moiety.
The latter could form a salt bridge with the catalytic aspartate and a glutamate residue. The binding
free energy values calculated from the docking poses turned out to be consistent with the experimental
data. Crystallographic studies are now underway to confirm those results.
Finally, the selected
MIC99 values ranging from 0.8 to 6.3
the main target of the tested compounds. Time-dependent killing experiments performed at 10
of compounds 88 91 95 and 124 showed that they inhibit bacterial growth in a bactericidal manner.
β
-carbolines are also active against M. tuberculosis strain H37Rv in vitro with
M. The correlation with the i values suggests that SerB2 could be
MIC99
µ
K
×
,
,
Altogether, these promising results may provide a basis for the development of new antimycobacterial
agents targeting SerB2.
4. Materials and Methods
4.1. Expression and Purification of SerB2
Escherichia coli strains BL21 (DE3, pLysS) and BL21 (DE3) were respectively transformed with
pET28b-serb2 or pAVA0421-serb2. Both plasmids carry serb2 gene under control of the T7 RNA
polymerase promoter and encode for a N-terminal hexahistidine tag. The pET28b-serb2 plasmid
contains a gene for kanamycin resistance and its expression product was used in the screening of
inhibitors. The pAVA0421-serb2 was used for the kinetic studies and encodes for ampicillin resistance
along with a human rhinovirus 3C (HRV 3C) protease site allowing cleavage of the hexahistidine tag.
The cells were grown at 37 ◦C in LB liquid medium supplemented with the appropriate antibiotics
until an optical density at 600 nm of 0.6–0.8. Protein expression was induced at 20 ◦C by the addition
of 0.2 mM isopropyl β-D-1-thiogalactopyranoside (IPTG). The cells were grown for 18 h and harvested
by centrifugation, resuspended in lysis buffer (50 mM Tris-HCl pH 7.4, 200 mM NaCl, 1 mg.mL−1
lysozyme, supplemented with cOmplete EDTA-free protease-inhibitor cocktail from Roche) and
disrupted by sonication over ice. After centrifugation and filtration, the lysate was loaded onto
a HisTrap FF crude IMAC column (GE Healthcare) and washed with buffer A (50 mM Tris-HCl
pH 8.0, 200 mM NaCl, 20 mM imidazole). The protein was eluted with a linear gradient of buffer
B (50 mM Tris-HCl pH 8.0, 200 mM NaCl, 500 mM imidazole). For the pAVA0421-serb2 construct,
the hexahistidine tag was removed by incubation with HRV3C protease overnight and separated from
the cleaved protein by a second IMAC purification. The fractions containing SerB2 were dialyzed
◦
overnight at 4 C against a storage buffer (50 mM Tris-HCl pH 7.4, 100 mM NaCl), aliquoted and
flash-freezed for storage at
Blue and concentration was determined spectrophotometrically at 280 nm using 11,710 M−1cm−1 as
the extinction coefficient.
−
80 ◦C. Purity was verified by 12% SDS-PAGE stained with Coomassie
4.2. Enzymatic Assay, Screening of Inhibitors and Inhibition Kinetics Studies
Enzyme activity was assayed by Pi determination using a malachite green-based phosphatase
assay adapted from the protocol of Arora et al. [22]. SerB2 is incubated at 37 ◦C in a 180 µL volume
containing 25 mM Tris-HCl pH 7.4, 5 mM MgCl2, 1 mM DTT, 5% DMSO and the selected inhibitor (or no
inhibitor for 100% activity control). The reaction is initiated by adding 20 µL 2 mM O-phospho-L-serine.
After incubation for 10 min at 37 ◦C, the reaction is stopped by mixing 150 µL of the reaction volume
with 50 µL dye composed of 1.7% ammonium heptamolybdate and 0.22% malachite green in 2 M
HCl [47
,48]. The absorbance of the solution is measured at 660 nm and the activity (released
Pi) is calculated from a calibration curve constructed using dilutions of a phosphate standard

Molecules 2020, 25, 415
11 of 17
solution. Tests are run in triplicate and include blanks run without O-phospho-L-serine. Inhibition
kinetics measurements are performed following the same procedure with varying O-phospho-L-serine
concentrations and fixed inhibitor concentration. For the analysis of kinetic data, initial velocities are
reported against substrate concentration as the amounts of released Pi after 10 min reaction. In the
screening and inhibition assays, inhibition percentages (inh%) are calculated according to Equation (1)
where nPi and nPi
are the amounts of released Pi for the inhibited condition and the 100% activity
100%
control respectively.
100 × nPi
inh% = 100 −
.
(1)
n
Pi100%
4.3. Evaluation of Kinetic Parameters
Analysis of kinetic data and curve fitting by linear and non-linear regression was performed
using GraphPad Prism 5 (GraphPad Software, La Jolla California USA). Values of apparent Km and
Vmax parameters were determined by fitting the Michaelis-Menten equation to the experimental
velocity curves (v vs. [S]). Values of Ki and n parameters were determined by fitting the slope function
(Equation (3)) of the Hill-type inhibition model of Cao et al. [52] (Equation (2)) to the slopesLB vs.
[I] replots.
Vmax × [S]
ꢀ
ꢁ
v =
(2)
(3)
[I]n
Km 1 +
+ [S]
Kin
ꢂ
ꢃ
Km
=
[I]n
Kin
slopeLB
1 +
.
Vmax
4.4. Evaluation of Hydrophobic Parameters
Hansch hydrophobic parameters
π
were calculated using the Molinspiration online Property
Calculation Service (www.molinspiration.com).
4.5. Induced Fit Docking
The structure of SerB2 was build from the known structure of Mycobacterium avium phosphoserine
phosphatase SerB (84% identity, PDB entry 3P96) using Maestro Homology Modelling tool. The model
was further minimized with the Protein Preparation Wizard [60]. The protonation of acid and basic
residues was adjusted using Epik at a fixed pH of 7.4 [61]. The global structure was then refined with
the OPLS3e forcefield. Inhibitors were prepared using the LigPrep tool of Maestro with the OPLS3e
forcefield [62]. The Induced Fit Docking protocol [63,64] from Schrödinger’s 2019 suite was employed
to dock tri-substituted harmine derivatives within SerB2’s active site. The receptor grid center was
fixed on the magnesium ion and the cubic grid was set to dock ligands with length equal to 25 Å.
Side chains were trimmed based on their B-factors during the initial docking phase and Glide XP
precision was used for the final redocking step [65]. The affinity of ligands was then calculated using
the Prime MM-GBSA [66–68] application from Maestro. This physics-based method is usually used to
rescore ligands using the force field energies calculated for bound and unbound ligand in an implicit
solvent model. The difference of energy between the enzyme-ligand complex and the unbound state
of enzyme and ligand gives the
OPLS3e forcefield.
∆
Gbind in kcal/mol. The VSGB solvation model was used with the
4.6. MIC99 Determination Assays
◦
M. tuberculosis H37Rv were cultured at 37 C with continuous agitation at 200 rpm in middlebrook
7H9 medium supplemented with 0.2% glycerol, 1x Albumin-Dextrose-Saline (ADS), 0.05% Tween-80.
For MIC99 assays, bacterial cultures were seeded at an OD600nm of 0.0002 in 7H9 medium in the presence
of varying concentrations of drugs in 96-well round-bottom plates (Nunc, USA). The last column wells
of each plate were used for negative control (medium plus bacterial cells only). For positive control,

Molecules 2020, 25, 415
12 of 17
antitubercular drug, INH, was used in each assay. Plates were incubated at 37 ◦C for 14 days. MIC99
was determined as the lowest concentration of drug at which no visible growth is observed.
4.7. Time-Dependent Bactericidal/Bacteriostatic Activity of Harmine Derivatives
Compounds 124, 95, 88 and 91 were selected for kill kinetics experiments against M. tuberculosis
H37Rv. For killing experiments, early-log phase cultures were exposed to these drugs and isoniazid
(positive control) at 10x MIC99 concentration for 6 days. At designated ◦time points, samples were
collected, diluted 10 folds and plated on middlebrook 7H11 medium at 37 C for 3–4 weeks. To obtain
time-kill curves, log10 CFU values were plotted against time (in days) using GraphPad Prism 5.
Statistical analyses were performed using GraphPad Prism 5. Data were checked for homoscedasticity
using a F-test before analysis. Data from each treated sample was compared with DMSO control data
by performing an unpaired Student T-test (two-tailed). P values lower than 0.05 were considered
statistically significant.
4.8. Synthesis
4.8.1. General
All commercial starting materials were used without further purification. Technical solvents
were distilled before use. Anhydrous THF was obtained by distillation under argon with
sodium/benzophenone. Thin-layer chromatography (TLC) was performed on silica gel 60 F254
TLC plates (Merck) and revelated at 254 nm or by staining with p-anisaldehyde. Melting points were
measured in open capillaries using a Büchi melting point B-450 apparatus. Mass spectra (MS) were
recorded using electron spray ionization in positive mode on a Bruker maXis Impact spectrometer.
Analytes were dissolved in dichloromethane (DCM) and injected at 3
liquid chromatography (HPLC) analyses were performed on an Agilent 1100 series HPLC using UV
detection at 254 nm. The method used consisted of the injection of 10 L of a 100 g/mL solution
onto an Agilent Zorbax Eclipse XDB C8 4.6 mm 150 mm, 5 m separation using a 30:70 solution of
µL/min. High performance
µ
µ
×
µ
0.01 M sodium butane sulfonate in water/ 0.01 M sodium butane sulfonate in methanol as the eluent
(flow: 1.0 mL/min). H and 13C NMR spectra were recorded at 400 and 100 MHz respectively on a Jeol
1
JNM ECX 400 spectrometer. Chemical shifts were reported in parts per million (ppm,
residual peak of the solvent (
1H CDCl3 7.26 ppm, 13C CDCl3 77.2 ppm, 1H DMSO-d6 2.50 ppm,
13C DMSO-d6 39.5 ppm). Coupling constants are given in Hz. Proton coupling patterns are described
as singlet (s), doublet (d), triplet (t), quartet (q), multiplet (m).
δ) relative to the
δ
δ
δ
δ
4.8.2. Synthesis of 1-methyl-9H-pyrido[3,4-b]indol-7-ol hydrochloride dihydrate (Harmol)
Harmine (2.00 g, 9.42 mmol, 1 equiv) was dissolved in glacial acetic acid (40 mL) followed
by addition of 47% aqueous hydrobromic acid (40 mL). The solution was refluxed for 72 h.
After completion of reaction was confirmed by TLC, the mixture was diluted in water and concentrated
under reduced pressure. Dilution and evaporation were repeated twice to afford a beige solid. Yield:
◦
1
99%, Rf: 0.37 (DCM/ethanol(EtOH): 85/15), mp 248 C, MS [M+H]+ : 199.1, H NMR (400 MHz,
DMSO-d6)
δ
(ppm) : 2.96 (s, 3H, CH3), 6.90 (dd, 1H, 3J = 8.70, 4J = 2.10, H-6), 7.04 (d, 1H, J = 2.10, H-8),
8.25 (d, 1H, J = 8.50, H-5), 8.35 (d, 1H, J = 6,18 H-3), 8.39 (d, 1H, J = 6,18 , H-4), 10.46 (s, 1H, O-H),
12.52 (s, 1H, N-H), 13C NMR (100 MHz, DMSO-d6)
128.67, 132.43, 133.55, 136.53, 145.72, 161.35.
δ (ppm) : 15.83, 96.49, 112.79, 112.88, 113.64, 124.78,
4.8.3. 7-(cyclohexylmethoxy)-9(cyclohexylmethyl)-1-methyl-β-carboline (123)
Compound 123 was synthesized according to the general procedure described in Reference [42
]
from harmol hydrobromide dihydrate (1.50 g, 4.76 mmol, 1 equiv) in the presence of potassium
hydroxide (1.34 g, 23.9 mmol, 5 equiv) and bromomethylcyclohexane (1.33 mL, 9.53 mmol, 2 equiv)
in DMF (30 mL). A white solid was obtained after purification. Yield: 50%, Rf: 0.75 (DCM/EtOH:

Molecules 2020, 25, 415
13 of 17
1
85/15), H NMR (400 MHz, CDCl3
)
δ
(ppm) : 1.10-1.96 (m, 22H, cyclohexyles); 3.04 (s, 3H, CH3);
3.88 (d, 2H, J = 6.18 , O-CH2); 4.28 (d, 2H, J = 7,10, N-CH2); 6.85 (d, 1H, J = 1,83, H-8); 6.87 (dd, 1H,
4J = 2.06 , 3J = 8.47, H-6) ; 7.76 (d, 1H, J = 5.27 , H-4); 7.95 (d, 1H, J = 8.47 , H-5); 8.27 (d, 1H, J = 5.50,
H-3), 13C NMR (100 MHz, CDCl3
) δ (ppm) : 23.71; 26.00; 26.39; 26.66; 30.15; 31.19; 37.96; 40.36; 50.93;
74.09; 95.00; 109.20; 112.20; 114.88; 122.25; 129.63; 135.67; 138.11; 140.68; 143.96; 160.49 , mp 148 ◦C,
MS [M+H]+: 391.3.
4.8.4. 7-(cyclohexylmethoxy)-9-(cyclohexylmethyl)-1-methyl-2-(2-cyclohexylethyl)-β-carboline
-2-ium bromide (124)
Compound 124 was synthesized according to the general procedure described in Reference [42
]
from compound 123 (200 mg, 0.51 mmol, 1 equiv) and 1-bromo-2-cyclohexylethane (0.8 mL, 5.11 mmol,
10 equiv) in THF (7 mL). The purified product was a beige solid. Yield: 31%, Rf: 0.23 (DCM/EtOH:
90/10), 1H NMR (400 MHz, CDCl3 (ppm) : 0.99-1.93 (m, 35H, N+-CH2
) δ -CH2-Cy, 2 cyclohexyles),
3.27 (s, 3H, CH3), 3.88 (d, 2H, J = 5.95, N-CH2), 4.37 (d, 2H, J = 7.10, O-CH2), 4.86 (t, 2H, J = 8.24,
N+-CH2), 6.84 (d, 1H, J = 1.60, H-8), 6.95 (dd, 1H, 4J = 1.83, 3J = 8.93, H-6), 8.04 (d, 1H, J = 8.70, H-5),
8.23 (d, 1H, J = 6.64, H-4), 8.79 (d, 1H, J = 6,64, H-3), 13C NMR (100 MHz, CDCl3
) δ (ppm): 16.79;
25.76; 25.88; 26.06; 26.30; 26.50; 29.77; 29.97; 31.06; 33.20; 35.44; 37.77; 39.06; 40.57; 51.96; 56.68; 74.30;
94.65; 112.87; 113.34; 114.77; 124.26; 133.93; 135.29; 135.29; 135.74; 136.97; 148.31; 163.66, mp 237 ◦C,
MS [M+H]+ : 501.4 , Purity: 100% by HPLC.
Supplementary Materials: Supplementary materials are available online at http://www.mdpi.com/1420-3049/
25/2/415/s1.
Author Contributions: Conceived and designed the experiment: M.H. and E.P. Performed enzymatic assays
and synthesis: E.P. Performed TB microbiology experiments: T.P.G. and P.C. Supervised TB experiments: R.S.
Designed and supervised the project: J.W. Manuscript redaction: E.P., M.H., J.W., R.S. All authors have read and
agreed to the published version of the manuscript.
Funding: EP acknowledges the Fonds de la Recherche Scientifique (F.R.S.-FNRS, Belgium) for her Research Fellow
grant. RS acknowledges the funding received from THSTI and Department of Biotechnology, Govt. of India (Grant
ID; BT/PR29075/BRB/10/1699/2018). TG and PC acknowledge Department of Biotechnology and Department
of Science and Technology (Grant ID: EMR/2016/0002405) for their respective fellowships.
Acknowledgments: The pAVA0421-serb2 plasmid was generously provided by the Seattle Structural Genomics
Center for Infectious Disease (www.SSGCID.org) which is supported Federal Contract No. HHSN272201700059C
from the National Institute of Allergy and Infectious Diseases, National Institutes of Health, Department of Health
and Human Services.
Conflicts of Interest: The authors declare no conflict of interest.
Abbreviations
The following abbreviations are used in this manuscript:
CFU
DCM
DMF
Colony Forming Unit
Dichloromethane
N,N-Dimethylformamide
HRV 3C Human rhinovirus 3C
IMAC
INH
Ki
MIC
Mtb
PDB
Pi
Immobilized metal affinity chromatography
isoniazid
inhibition constant
minimal inhibitory concentration
Mycobacterium tuberculosis
Protein Data Bank
inorganic phosphate
O-phospho-L-serine
Tuberculosis
PS
TB
THF
Tetrahydrofuran

Molecules 2020, 25, 415
14 of 17
References
1.
2.
3.
4.
World Health Organization. Global Tuberculosis Report 2018. Available online: https://www.who.int/
(accessed on 24 September 2019).
The Global Alliance for TB Drug Development. Available online: https://www.tballiance.org/ (accessed on
2 October 2019).
Lytras, T.; Kalkouni, O. The global tuberculosis epidemic: Turning political will into concrete action.
J. Thorac. Dis. 2018, 10, S3149–S3152. [CrossRef]
Floyd, K.; Glaziou, P.; Zumla, A.; Raviglione, M. The global tuberculosis epidemic and progress in care,
prevention, and research: An overview in year 3 of the End TB era. Lancet Respir. Med. 2018, 6, 299–314.
[CrossRef]
5.
Cole, S.T.; Brosch, R.; Parkhill, J.; Garnier, T.; Churcher, C.; Harris, D.; Gordon, S.; Eiglmeier, K.; Gas, S.;
Barry, C., III; et al. Deciphering the biology of Mycobacterium tuberculosis from the complete genome
sequence. Nature 1998, 393, 537–544. [CrossRef]
6.
7.
Sassetti, C.M.; Boyd, D.H.; Rubin, E.J. Genes required for mycobacterial growth defined by high density
mutagenesis. Mol. Microbiol. 2003, 48, 77–84. [CrossRef]
Singh, V.; Donini, S.; Pacitto, A.; Sala, C.; Hartkoorn, R.C.; Dhar, N.; Keri, G.; Ascher, D.B.; Mondesert, G.;
Vocat, A.; et al. The inosine monophosphate dehydrogenase, GuaB2, is a vulnerable new bactericidal drug
target for tuberculosis. ACS Infect. Dis. 2016, 3, 5–17. [CrossRef] [PubMed]
Sahu, N.U.; Singh, V.; Ferraris, D.M.; Rizzi, M.; Kharkar, P.S. Hit discovery of Mycobacterium tuberculosis
inosine 5-monophosphate dehydrogenase, GuaB2, inhibitors. Bioorg. Med. Chem. Lett. 2018, 28, 1714–1718.
[CrossRef]
Burke, C.; Abrahams, K.A.; Richardson, E.J.; Loman, N.J.; Alemparte, C.; Lelievre, J.; Besra, G.S. Development
of a whole-cell high-throughput phenotypic screen to identify inhibitors of mycobacterial amino acid
biosynthesis. FASEB BioAdv. 2019, 1, 246–254. [CrossRef]
8.
9.
10. Berney, M.; Berney-Meyer, L.; Wong, K.W.; Chen, B.; Chen, M.; Kim, J.; Wang, J.; Harris, D.; Parkhill, J.; Chan,
J.; et al. Essential roles of methionine and S-adenosylmethionine in the autarkic lifestyle of Mycobacterium
tuberculosis. Proc. Natl. Acad. Sci. USA 2015, 112, 10008–10013. [CrossRef] [PubMed]
11. Mizrahi, V.; Warner, D.F. Death of Mycobacterium tuberculosis by L-arginine starvation. Proc. Natl. Acad.
Sci. USA 2018, 115, 9658–9660. [CrossRef] [PubMed]
12. Haufroid, M.; Wouters, J. Targeting the Serine Pathway: A Promising Approach against Tuberculosis?
Pharmaceuticals 2019, 12, 66. [CrossRef]
13. Lowry, M.; Hall, D.E.; Hall, M.S.; Brosnan, J.T. Renal metabolism of amino acids in vivo: Studies on serine
and glycine fluxes. Am. J. Physiol.-Ren. Physiol. 1987, 252, F304–F309. [CrossRef] [PubMed]
14. Snell, K. The duality of pathways for serine biosynthesis is a fallacy. Trends Biochem. Sci. 1986, 11, 241–243.
[CrossRef]
15. Ros, R.; Muñoz-Bertomeu, J.; Krueger, S. Serine in plants: Biosynthesis, metabolism, and functions.
Trends Plant Sci. 2014, 19, 564–569. [CrossRef] [PubMed]
16. Bai, G.; Schaak, D.D.; Smith, E.A.; McDonough, K.A. Dysregulation of serine biosynthesis contributes to the
growth defect of a Mycobacterium tuberculosis crp mutant. Mol. Microbiol. 2011, 82, 180–198. [CrossRef]
[PubMed]
17. Dey, S.; Hu, Z.; Xu, X.L.; Sacchettini, J.C.; Grant, G.A. D-3-Phosphoglycerate dehydrogenase from
Mycobacterium tuberculosis is a link between the Escherichia coli and mammalian enzymes. J. Biol.
Chem. 2005, 280, 14884–14891. [CrossRef] [PubMed]
18. Coulibaly, F.; Lassalle, E.; Baker, H.M.; Baker, E.N. Structure of phosphoserine aminotransferase from
Mycobacterium tuberculosis. Acta Crystallogr. Sect. D Biol. Crystallogr. 2012, 68, 553–563. [CrossRef]
19. Yadav, G.P.; Shree, S.; Maurya, R.; Rai, N.; Singh, D.K.; Srivastava, K.K.; Ramachandran, R. Characterization
of M. tuberculosis SerB2, an essential HAD-family phosphatase, reveals novel properties. PLoS ONE 2014
9, e115409. [CrossRef]
,
20. Shree, S.; Singh, A.K.; Saxena, R.; Kumar, H.; Agarwal, A.; Sharma, V.K.; Srivastava, K.; Srivastava, K.K.;
Sanyal, S.; Ramachandran, R. The M. tuberculosis HAD phosphatase (Rv3042c) interacts with host proteins
and is inhibited by Clofazimine. Cell. Mol. Life Sci. 2016, 73, 3401–3417. [CrossRef]

Molecules 2020, 25, 415
15 of 17
21. Anishetty, S.; Pulimi, M.; Pennathur, G. Potential drug targets in Mycobacterium tuberculosis through
metabolic pathway analysis. Comput. Biol. Chem. 2005, 29, 368–378. [CrossRef]
22. Arora, G.; Tiwari, P.; Mandal, R.S.; Gupta, A.; Sharma, D.; Saha, S.; Singh, R. High throughput screen
identifies small molecule inhibitors specific for Mycobacterium tuberculosis phosphoserine phosphatase.
J. Biol. Chem. 2014, 289, 25149–25165. [CrossRef]
23. Marx, S.; Van Gysel, M.; Breuer, A.; Dal Maso, T.; Michiels, C.; Wouters, J.; Le Calvé, B. Potentialization of
anticancer agents by identification of new chemosensitizers active under hypoxia. Biochem. Pharmacol. 2019
162, 224–236. [CrossRef] [PubMed]
,
24. Parida, S.; Axelsson-Robertson, R.; Rao, M.; Singh, N.; Master, I.; Lutckii, A.; Keshavjee, S.; Andersson, J.;
Zumla, A.; Maeurer, M. Totally drug-resistant tuberculosis and adjunct therapies. J. Intern. Med. 2015
277, 388–405. [CrossRef] [PubMed]
,
25. Maitra, A.; Bates, S.; Kolvekar, T.; Devarajan, P.V.; Guzman, J.D.; Bhakta, S. Repurposing—A ray of hope in
tackling extensively drug resistance in tuberculosis. Int. J. Infect. Dis. 2015, 32, 50–55. [CrossRef] [PubMed]
26. Maitra, A.; Bates, S.; Shaik, M.; Evangelopoulos, D.; Abubakar, I.; McHugh, T.D.; Lipman, M.; Bhakta, S.
Repurposing drugs for treatment of tuberculosis: A role for non-steroidal anti-inflammatory drugs.
Br. Med. Bull. 2016, 118, 138. [CrossRef] [PubMed]
27. Harausz, E.P.; Garcia-Prats, A.J.; Seddon, J.A.; Schaaf, H.S.; Hesseling, A.C.; Achar, J.; Bernheimer, J.;
Cruz, A.T.; D’Ambrosio, L.; Detjen, A.; et al. New and repurposed drugs for pediatric multidrug-resistant
tuberculosis. Practice-based recommendations. Am. J. Respir. Crit. Care Med. 2017, 195, 1300–1310. [CrossRef]
28. Gopal, M.; Padayatchi, N.; Metcalfe, J.; O’Donnell, M. Systematic review of clofazimine for the treatment of
drug-resistant tuberculosis. Int. J. Tuberc. Lung Dis. 2013, 17, 1001–1007. [CrossRef]
29. Falzon, D.; Schünemann, H.J.; Harausz, E.; González-Angulo, L.; Lienhardt, C.; Jaramillo, E.; Weyer, K.
World Health Organization treatment guidelines for drug-resistant tuberculosis, 2016 update. Eur. Respir. J.
2017, 49, 1602308. [CrossRef]
30. Silva, D.R.; Dalcolmo, M.; Tiberi, S.; Arbex, M.A.; Munoz-Torrico, M.; Duarte, R.; D’Ambrosio, L.; Visca, D.;
Rendon, A.; Gaga, M.; et al. New and repurposed drugs to treat multidrug-and extensively drug-resistant
tuberculosis. J. Bras. Pneumol. 2018, 44, 153–160. [CrossRef]
31. Tonby, K.; Wergeland, I.; Lieske, N.V.; Kvale, D.; Tasken, K.; Dyrhol-Riise, A.M. The COX-inhibitor
indomethacin reduces Th1 effector and T regulatory cells in vitro in Mycobacterium tuberculosis infection.
BMC Infect. Dis. 2016, 16, 599. [CrossRef]
32. Kroesen, V.M.; Gröschel, M.I.; Martinson, N.; Zumla, A.; Maeurer, M.; van der Werf, T.S.; Vilaplana, C.
Non-steroidal anti-inflammatory drugs as host-directed therapy for tuberculosis: A systematic review.
Front. Immunol. 2017, 8, 772. [CrossRef]
33. Gold, B.; Pingle, M.; Brickner, S.J.; Shah, N.; Roberts, J.; Rundell, M.; Bracken, W.C.; Warrier, T.; Somersan, S.;
Venugopal, A.; et al. Nonsteroidal anti-inflammatory drug sensitizes Mycobacterium tuberculosis to
endogenous and exogenous antimicrobials. Proc. Natl. Acad. Sci. USA 2012, 109, 16004–16011. [CrossRef]
[PubMed]
34. Ivanyi, J.; Zumla, A. Nonsteroidal Antiinflammatory Drugs for Adjunctive Tuberculosis Treatment.
J. Infect. Dis. 2013, 208, 185–188. [CrossRef] [PubMed]
35. Cao, R.; Chen, Q.; Hou, X.; Chen, H.; Guan, H.; Ma, Y.; Peng, W.; Xu, A. Synthesis, acute toxicities, and
antitumor effects of novel 9-substituted
[CrossRef] [PubMed]
β-carboline derivatives. Bioorg. Med. Chem. 2004, 12, 4613–4623.
36. Patel, K.; Gadewar, M.; Tripathi, R.; Prasad, S.; Patel, D.K. A review on medicinal importance,
pharmacological activity and bioanalytical aspects of beta-carboline alkaloid “Harmine”. Asian Pac. J.
Trop. Biomed. 2012, 2, 660–664. [CrossRef]
37. Song, H.; Liu, Y.; Liu, Y.; Wang, L.; Wang, Q. Synthesis and antiviral and fungicidal activity evaluation of
β
-carboline, dihydro-
2014, 62, 1010–1018. [CrossRef] [PubMed]
38. Cao, R.; Peng, W.; Wang, Z.; Xu, A. -Carboline alkaloids: Biochemical and pharmacological functions.
β-carboline, tetrahydro-β-carboline alkaloids, and their derivatives. J. Agric. Food Chem.
β
Curr. Med. Chem. 2007, 14, 479–500. [CrossRef]
39. Begum, S.; Ali, S.N.; Siddiqui, B.S. Antitubercular Alkaloid. U.S. Patent 8420660 B2, 16 April 2013.
40. Davoodi, H.; Ghaemi, E.; Mazandarani, M.; Shakeri, F.; Javid, S.; Klishadi, M. Anti-mycobacterial and
anti-inflammatory activity of Peganum harmala. Chem. Pharm. Res. 2015, 7, 1611–1616.

Molecules 2020, 25, 415
16 of 17
41. Reniers, J.; Robert, S.; Frederick, R.; Masereel, B.; Vincent, S.; Wouters, J. Synthesis and evaluation of
-carboline derivatives as potential monoamine oxidase inhibitors. Bioorg. Med. Chem. 2011, 19, 134–144.
β
[CrossRef]
42. Frédédrick, R.; Bruyere, C.; Vancraeynest, C.; Reniers, J.; Meinguet, C.; Pochet, L.; Backlund, A.; Masereel, B.;
Kiss, R.; Wouters, J. Novel trisubstituted harmine derivatives with original in vitro anticancer activity.
J. Med. Chem. 2012, 55, 6489–6501. [CrossRef]
43. Meinguet, C.; Bruyère, C.; Frédérick, R.; Mathieu, V.; Vancraeynest, C.; Pochet, L.; Laloy, J.; Mortier, J.;
Wolber, G.; Kiss, R.; et al. 3D-QSAR, design, synthesis and characterization of trisubstituted harmine
derivatives with in vitro antiproliferative properties. Eur. J. Med. Chem. 2015, 94, 45–55. [CrossRef]
44. Meinguet, C.; Masereel, B.; Wouters, J. Preparation and characterization of a new harmine-based
antiproliferative compound in complex with cyclodextrin: Increasing solubility while maintaining biological
activity. Eur. J. Pharmaceut. Sci. 2015, 77, 135–140. [CrossRef] [PubMed]
45. Carvalho, A.; Chu, J.; Meinguet, C.; Kiss, R.; Vandenbussche, G.; Masereel, B.; Wouters, J.; Kornienko, A.;
Pelletier, J.; Mathieu, V. A harmine-derived beta-carboline displays anti-cancer effects in vitro by targeting
protein synthesis. Eur. J. Pharmacol. 2017, 805, 25–35. [CrossRef] [PubMed]
46. Marx, S.; Bodart, L.; Tumanov, N.; Wouters, J. Design and Synthesis of a New Soluble Natural β-Carboline
Derivative for Preclinical Study by Intravenous Injection. Int. J. Mol. Sci. 2019, 20, 1491. [CrossRef] [PubMed]
47. Itaya, K.; Ui, M. A new micromethod for the colorimetric determination of inorganic phosphate.
Clin. Chim. Acta 1966, 14, 361–366. [CrossRef]
48. Baykov, A.; Evtushenko, O.; Avaeva, S. A malachite green procedure for orthophosphate determination and
its use in alkaline phosphatase-based enzyme immunoassay. Anal. Biochem. 1988, 171, 266–270. [CrossRef]
49. Fujita, T.; Iwasa, J.; Hansch, C. A new substituent constant, π, derived from partition coefficients. J. Am.
Chem. Soc. 1964, 86, 5175–5180. [CrossRef]
50. Leskovac, V.
Hyperbolic and parabolic inhibition. In Comprehensive Enzyme Kinetics; Kluwer:
New York, NY, USA, 2003; pp. 95–110.
51. Segel, I. Simple inhibition systems. In Enzyme Kinetics: Behaviour and Analysis of Rapid Equilibrium and
Steady-State Enzyme Systems; John Wiley & Sons: Hoboken, NJ, USA, 1993; pp. 100–160.
52. Cao, R.; Zeidan, A.A.; Rådström, P.; van Niel, E.W. Inhibition kinetics of catabolic dehydrogenases by
elevated moieties of ATP and ADP—Implication for a new regulation mechanism in Lactococcus lactis.
FEBS J. 2010, 277, 1843–1852. [CrossRef]
53. Pala, D.; Rivara, S.; Mor, M.; Milazzo, F.M.; Roscilli, G.; Pavoni, E.; Giannini, G. Kinetic analysis and molecular
modeling of the inhibition mechanism of roneparstat (SST0001) on human heparanase. Glycobiology 2016
26, 640–654. [CrossRef]
,
54. Hammond, E.; Handley, P.; Dredge, K.; Bytheway, I. Mechanisms of heparanase inhibition by the heparan
sulfate mimetic PG545 and three structural analogues. FEBS Open Bio 2013, 3, 346–351. [CrossRef]
55. Abendroth, J.; Gardberg, A.S.; Robinson, J.I.; Christensen, J.S.; Staker, B.L.; Myler, P.J.; Stewart, L.J.;
Edwards, T.E. SAD phasing using iodide ions in a high-throughput structural genomics environment.
J. Struct. Funct. Genomics 2011, 12, 83–95. [CrossRef]
56. Haufroid, M.; Mirgaux, M.; Leherte, L.; Wouters, J. Crystal structures and snapshots along the reaction
pathway of human phosphoserine phosphatase. Acta Crystallogr. Sect. D Struct. Biol. 2019, 75, 592–604.
[CrossRef] [PubMed]
57. Kim, H.Y.; Heo, Y.S.; Kim, J.H.; Park, M.H.; Moon, J.; Kim, E.; Kwon, D.; Yoon, J.; Shin, D.; Jeong, E.J.; et al.
Molecular basis for the local conformational rearrangement of human phosphoserine phosphatase.
J. Biol. Chem. 2002, 277, 46651–46658. [CrossRef] [PubMed]
58. Peeraer, Y.; Rabijns, A.; Verboven, C.; Collet, J.F.; Van Schaftingen, E.; De Ranter, C. High-resolution structure
of human phosphoserine phosphatase in open conformation. Acta Crystallogr. Sect. D Biol. Crystallogr. 2003
59, 971–977. [CrossRef] [PubMed]
,
59. Dhiman, R.; Singh, R. Recent advances for identification of new scaffolds and drug targets for Mycobacterium
tuberculosis. IUBMB Life 2018, 70, 905–916. [CrossRef]
60. Sastry, G.M.; Adzhigirey, M.; Day, T.; Annabhimoju, R.; Sherman, W. Protein and ligand preparation:
Parameters, protocols, and influence on virtual screening enrichments. J. Comput.-Aided Mol. Des. 2013
27, 221–234. [CrossRef] [PubMed]
,

Molecules 2020, 25, 415
17 of 17
61. Greenwood, J.R.; Calkins, D.; Sullivan, A.P.; Shelley, J.C. Towards the comprehensive, rapid, and accurate
prediction of the favorable tautomeric states of drug-like molecules in aqueous solution. J. Comput.-Aided
Mol. Des. 2010, 24, 591–604. [CrossRef]
62. Harder, E.; Damm, W.; Maple, J.; Wu, C.; Reboul, M.; Xiang, J.Y.; Wang, L.; Lupyan, D.; Dahlgren, M.K.;
Knight, J.L.; et al. OPLS3: A force field providing broad coverage of drug-like small molecules and proteins.
J. Chem. Theory Comput. 2015, 12, 281–296. [CrossRef]
63. Sherman, W.; Day, T.; Jacobson, M.P.; Friesner, R.A.; Farid, R. Novel procedure for modeling ligand/receptor
induced fit effects. J. Med. Chem. 2006, 49, 534–553. [CrossRef]
64. Sherman, W.; Beard, H.S.; Farid, R. Use of an induced fit receptor structure in virtual screening. Chem. Biol.
Drug Des. 2006, 67, 83–84. [CrossRef]
65. Friesner, R.A.; Murphy, R.B.; Repasky, M.P.; Frye, L.L.; Greenwood, J.R.; Halgren, T.A.; Sanschagrin, P.C.;
Mainz, D.T. Extra precision glide: Docking and scoring incorporating a model of hydrophobic enclosure for
protein- ligand complexes. J. Med. Chem. 2006, 49, 6177–6196. [CrossRef]
66. Greenidge, P.A.; Kramer, C.; Mozziconacci, J.C.; Wolf, R.M. MM/GBSA binding energy prediction on the
PDBbind data set: Successes, failures, and directions for further improvement. J. Chem. Inf. Model. 2012
53, 201–209. [CrossRef] [PubMed]
,
67. Lyne, P.D.; Lamb, M.L.; Saeh, J.C. Accurate prediction of the relative potencies of members of a series of
kinase inhibitors using molecular docking and MM-GBSA scoring. J. Med. Chem. 2006, 49, 4805–4808.
[CrossRef] [PubMed]
68. Sirin, S.; Kumar, R.; Martinez, C.; Karmilowicz, M.J.; Ghosh, P.; Abramov, Y.A.; Martin, V.; Sherman, W.
A computational approach to enzyme design: Predicting
ω-aminotransferase catalytic activity using docking
and MM-GBSA scoring. J. Chem. Inf. Model. 2014, 54, 2334–2346. [CrossRef] [PubMed]
c
ꢀ 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).