Synthesis and Structure–Activity relationship of 1-(5-isoquinolinesulfonyl)piperazine analogues as inhibitors of Mycobacterium tuberculosis IMPDH

Article abstract of DOI:10.1016/j.ejmech.2019.04.027

Tuberculosis (TB) is a major infectious disease associated increasingly with drug resistance. Thus, new anti-tubercular agents with novel mechanisms of action are urgently required for the treatment of drug-resistant TB. In prior work, we identified compound 1 (cyclohexyl(4-(isoquinolin-5-ylsulfonyl)piperazin-1-yl)methanone) and showed that its anti-tubercular activity is attributable to inhibition of inosine-5′-monophosphate dehydrogenase (IMPDH) in Mycobacterium tuberculosis. In the present study, we explored the structure–activity relationship around compound 1 by synthesizing and evaluating the inhibitory activity of analogues against M. tuberculosis IMPDH in biochemical and whole-cell assays. X-ray crystallography was performed to elucidate the mode of binding of selected analogues to IMPDH. We establish the importance of the cyclohexyl, piperazine and isoquinoline rings for activity, and report the identification of an analogue with IMPDH-selective activity against a mutant of M. tuberculosis that is highly resistant to compound 1. We also show that the nitrogen in urea analogues is required for anti-tubercular activity and identify benzylurea derivatives as promising inhibitors that warrant further investigation.

Full text of DOI:10.1016/j.ejmech.2019.04.027

Accepted Manuscript
Synthesis and Structure–Activity relationship of 1-(5-isoquinolinesulfonyl)piperazine
analogues as inhibitors of Mycobacterium tuberculosis IMPDH
Vinayak Singh, Angela Pacitto, Stefano Donini, Davide M. Ferraris, Sándor Boros,
Eszter Illyés, Bálint Szokol, Menico Rizzi, Tom L. Blundell, David B. Ascher, Janos
Pato, Valerie Mizrahi
PII:
S0223-5234(19)30333-2
DOI:
https://doi.org/10.1016/j.ejmech.2019.04.027
EJMECH 11262
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 28 December 2018
Revised Date: 11 April 2019
Accepted Date: 11 April 2019
Please cite this article as: V. Singh, A. Pacitto, S. Donini, D.M. Ferraris, Sá. Boros, E. Illyés, Bá. Szokol,
M. Rizzi, T.L. Blundell, D.B. Ascher, J. Pato, V. Mizrahi, Synthesis and Structure–Activity relationship
of 1-(5-isoquinolinesulfonyl)piperazine analogues as inhibitors of Mycobacterium tuberculosis IMPDH,
European Journal of Medicinal Chemistry (2019), doi: https://doi.org/10.1016/j.ejmech.2019.04.027.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
48
1
21
4
7
MtbIMPDH IC50 (µM):
M. tuberculosis MIC90 (µM):
Compound 1-resistant mutant MIC90 (µM):
On-target selectivity in M. tuberculosis:
0.1
2
250
Yes
0.3
15.6
3.9
0.035
3.1
>100
Yes
0.27
3.9
15.6
Yes
Yes

ACCEPTED MANUSCRIPT
1
2
3
Synthesis and Structure–Activity Relationship of
1-(5-isoquinolinesulfonyl)piperazine Analogues as Inhibitors of
Mycobacterium tuberculosis IMPDH
4
5
1
,2,3*
4
5
5
7
Vinayak Singh
, Angela Pacitto , Stefano Donini , Davide M. Ferraris , Sándor Boros ,
7
7
5
4
4,6
6
7
Eszter Illyés , Bálint Szokol , Menico Rizzi , Tom L. Blundell , David B. Ascher , Janos
7
2*
Pato , and Valerie Mizrahi
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
1
H3D Drug Discovery and Development Centre, Department of Drug Discovery and
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
Development & Institute of Infectious Disease and Molecular Medicine, University of Cape
Town, Rondebosch 7701, Cape Town, South Africa
2
MRC/NHLS/UCT Molecular Mycobacteriology Research Unit, DST/NRF Centre of
Excellence for Biomedical TB Research & Wellcome Centre for Infectious Diseases
Research in Africa, Institute of Infectious Disease and Molecular Medicine & Department of
Pathology, University of Cape Town, Anzio Road, Observatory 7925, South Africa
3
South African Medical Research Council Drug Discovery and Development Research Unit,
Department of Chemistry and Institute of Infectious Disease and Molecular Medicine,
University of Cape Town, Rondebosch 7701, Cape Town, South Africa
4
Department of Biochemistry, University of Cambridge, 80 Tennis Court Road, Cambridge
CB2 1GA, United Kingdom
5
Dipartimento di Scienze del Farmaco, University of Piemonte Orientale, Via Bovio 6, 28100
Novara, Italy
6
Department of Biochemistry and Molecular Biology, University of Melbourne, Bio21
Institute, 30 Flemington Road, Parkville, 3052, Australia
7
Vichem Chemie Research Ltd, Rákóczi utca 5, Veszprém, 8200, Hungary
2
2
2
2
6
7
8
9
*Corresponding authors: vinayak.singh@uct.ac.za; valerie.mizrahi@uct.ac.za
3
3
0
1
1

ACCEPTED MANUSCRIPT
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
Abstract
Tuberculosis (TB) is a major infectious disease associated increasingly with drug resistance.
Thus, new anti-tubercular agents with novel mechanisms of action are urgently required for
the treatment of drug-resistant TB. In prior work, we identified compound 1 (cyclohexyl(4-
(isoquinolin-5-ylsulfonyl)piperazin-1-yl)methanone) and showed that its anti-tubercular
activity is attributable to inhibition of inosine-5′-monophosphate dehydrogenase (IMPDH) in
Mycobacterium tuberculosis. In the present study, we explored the structure–activity
relationship around compound 1 by synthesizing and evaluating the inhibitory activity of
analogues against M. tuberculosis IMPDH in biochemical and whole-cell assays. X-ray
crystallography was performed to elucidate the mode of binding of selected analogues to
IMPDH. We establish the importance of the cyclohexyl, piperazine and isoquinoline rings for
activity, and report the identification of an analogue with IMPDH-selective activity against a
mutant of M. tuberculosis that is highly resistant to compound 1. We also show that the
nitrogen in urea analogues is required for anti-tubercular activity and identify benzylurea
derivatives as promising inhibitors that warrant further investigation.
Keywords: Tuberculosis, GuaB2, IMPDH, SAR
2

ACCEPTED MANUSCRIPT
5
5
5
6
6
6
6
6
6
6
6
6
6
7
7
7
7
7
7
7
7
7
7
8
8
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
1. Introduction
Tuberculosis (TB), an infectious disease caused by Mycobacterium tuberculosis (M.
tuberculosis), is the leading infectious killer globally, claiming 1.7 million lives in 2017 [1].
The severity of the global health threat presented by TB has been heightened by the
emergence and spread of multi drug-resistant (MDR) and extensively drug-resistant (XDR)
strains of M. tuberculosis that are resistant to first- and second-line drugs. Of the 10 million
incident cases of TB recorded globally in 2017, approximately 458,000 were MDR, defined
as resistant to isoniazid (INH) and rifampicin (RIF), with or without resistance to other first-
line anti-tubercular drugs.[1] Of these, 8.5 % had XDR-TB, which is resistant to INH and
RIF (i.e., MDR-TB) in addition to any fluoroquinolone and at least one of the injectable
second-line drugs, kanamycin, amikacin, or capreomycin [1]. Against this background, the
“End TB” strategy, which aims to reduce TB deaths by 90% and TB incidence by 80% by
2035, was launched recently by the WHO (http://www.who.int/tb/strategy/en/). Among other
interventions, new drug combinations that can shorten the duration of TB treatment; are
effective against MDR and XDR-TB, and can be co-administered with drugs for the
treatment of the major co-morbidities, HIV/AIDS and diabetes, are required to achieve this
aspirational goal. While considerable progress has been made on establishing a TB drug
pipeline, the high attrition rate of compounds in clinical development reinforces the need to
fuel the pipeline with high-quality leads that act on novel targets [2].
In prior work conducted under the auspices of the More Medicines for Tuberculosis
(MM4TB) consortium [3], we identified compound 1 (cyclohexyl(4-(isoquinoline-5-
ylsulfonyl)piperazin-1-yl)methanone) as a promising new anti-tubercular agent with limited
mammalian cell toxicity [4]. Compound 1 belongs to the 1-(5-isoquinolinesulfonyl)
homopiperazine (Fasudil) scaffold, which is a known inhibitor of Rho-associated protein
3

ACCEPTED MANUSCRIPT
8
8
8
8
8
8
8
8
9
9
9
9
9
9
9
9
9
9
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
kinase (ROCK), but was found to inhibit inosine-5′-monophosphate dehydrogenase (IMPDH)
+
in M. tuberculosis. IMPDH is an essential enzyme that catalyzes the NAD -dependent
conversion of inosine 5′-monophosphate (IMP) to xanthosine 5′-monophosphate (XMP) in
the de novo purine biosynthesis pathway [5]. Compound 1 is one of the several small-
molecule inhibitors of M. tuberculosis IMPDH discovered in recent years [6-15]. IMPDH
was genetically validated as a vulnerable and bactericidal drug target, and compound 1 was
shown to be active against M. tuberculosis in macrophages. Moreover, in a parallel study, the
value of IMPDH as a new TB drug target was called into question by the lack of in vivo
efficacy of an IMPDH inhibitor from an indazole sulfonamide series [8]. This conclusion was
attributed to subversion of IMPDH essentiality in vivo via the purine salvage pathway, which
enables GMP production in M. tuberculosis by assimilation of exogenous guanine, a
metabolite found to be present at millimolar concentrations in normal and diseased tissue
from humans and rabbits [8]. However, this notion was questioned by Hedstrom [16] who
offered alternative explanations for the discrepant conclusions regarding the value – or
otherwise – of IMPDH as a new TB drug target. Importantly, Hedstrom and colleagues
recently reported a benzoxazole-based M. tuberculosis IMPDH inhibitor which retains
activity in the presence of exogenous guanine [14], suggesting that IMPDH may indeed be a
vulnerable target in M. tuberculosis [5].
1
1
1
1
1
1
1
00
01
02
03
04
05
06
In the present study, we set out to explore the structure-activity relationships (SAR)
around compound 1 with the aim of identifying more potent analogues with improved
activity against a compound 1-resistant mutant of M. tuberculosis to enable further
pharmacologic interrogation of this target. The compounds were tested for inhibitory activity
against the IMPDH enzyme and for whole-cell activity against replicating wild-type M.
tuberculosis strain H37Rv. Target selectivity in M. tuberculosis was assessed by testing a
subset of compounds for activity against a conditional knockdown mutant in which the level
4

ACCEPTED MANUSCRIPT
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
of IMPDH can be progressively depleted by transcriptional silencing of the IMPDH-encoding
gene, guaB2 [5]. Cross-resistance was assessed by testing for activity against a mutant of M.
tuberculosis, SRMV2.6 [5], which shows high-level resistance to compound 1 as a result of a
Tyr487Cys substitution. As the the crystal structure of Mycobacterium thermoresistible (Mth)
IMPDH (GuaB2) was elucidated [5] during the course of this study, the modes of binding of
seven compounds to Mth IMPDH were also analysed by X-ray crystallography and docking
analyses performed for three others.
2. Results and Discussion
2.1 Design strategy for target compounds
To understand the SAR, a molecular library of 49 compounds was built around the compound
1 structure (Figure 1; Schemes 1 and 2; Tables S1 and S2). We made changes on the Fasudil-
like moiety; on the cyclohexyl ring, utilizing both alicyclic and aromatic substituents; and by
increasing the distance between the carbonyl group and the cyclohexyl ring to produce
phenylacetic acid, phenylurea, and benzylurea derivatives.
1
1
23
24
Figure 1. Design of target compound series derived from compound 1.
5

ACCEPTED MANUSCRIPT
1
1
1
1
25
26
27
28
2.2. Synthetic chemistry
The synthetic routes for the preparation of 1-(5-isoquinolinesulfonyl)piperazine derivatives
(2-49) derivatives are summarized in Schemes 1 and 2 and Table S1.
1
29
1
1
1
1
1
30
31
32
33
34
Scheme 1. The synthetic pathway of derivatives of compound 1. Reagents and conditions: (i),
EtOAc, triethylamine, ambient temp; (ii), 2N HCl, 3h reflux; (iii), TFA/DCM=1:1, 3h,
ambient temp; (iv), H /Pd-C, EtOH, overnight, ambient temp; (v), Py, ambient temp; (vi),
2
EtOAc, aq.NaHCO , overnight, ambient temp; (vii), EtOAc/DMF, carbodiimide (DCC or
3
°
EDCI) overnight, ambient temp; (viii), Py, 4h, 100 C.
1
1
35
36
6

ACCEPTED MANUSCRIPT
1
1
37
38
Scheme 2. Synthesis of derivatives of compound 1 (also see Table S1).
1
1
1
1
1
1
1
1
1
1
1
1
1
39
40
41
42
43
44
45
46
47
48
49
50
51
2.3. SAR profiling using biological assays
2.3.1. Set I: Modifications on the Fasudil-like moiety
The finding by Magnet et al. [4] that the Fasudil analog, 5-(piperazin-1-
ylsulfonyl)isoquinoline, has no MIC against M. tuberculosis, pointed to a critical role for the
cyclohexyl group in the activity of compound 1 – a notion substantiated subsequently by the
co-crystal structure of compound 1 bound to Mth IMPDH [5]. Therefore, in the first set of
compounds, we retained the cyclohexyl and modified the Fasudil-like moiety to produce
compounds 2-8 (Table 1). Replacement of the rigid piperazine ring with the flexible
ethylenediamine spacer resulted in the loss of whole-cell activity for compound 2 while
retaining activity against the enzyme. Modification of the piperazine ring by inclusion of a
methyl group at position-3 (compound 3) resulted in profound loss of both biochemical and
whole-cell activities (MIC : 100 µM, IC : >100 µM). Increasing the distance between the
90
50
7

ACCEPTED MANUSCRIPT
1
1
1
1
1
1
1
1
1
1
1
1
52
53
54
55
56
57
58
59
60
61
62
63
sulfonamide and carboxamide groups (compound 4 and 5) also resulted in the loss of both
whole-cell activity and activity against the enzyme. All other modifications of the piperazine
ring similarly ablated whole-cell activity while significantly reducing activity against the
enzyme. The introduction of a methyl group at position-3 of the isoquinoline ring (compound
6) ablated both whole-cell and enzyme activities (MIC : 100 µM, IC : >100 µM) (Table 1).
9
0
50
The isoquinoline substituent was then replaced by a naphthalene to produce the 1-naphthyl
derivative, 7, or 2-naphthyl derivative, 8. While these compounds retained modest activity
against the enzyme, both modifications completely ablated whole-cell activity (Table 1),
suggesting an essential role for the nitrogen atom in the isoquinoline moiety for compound
uptake.
Table 1. SAR profiling of modifications on the Fasudil-like moiety
1
1
64
65
CY: Cyclic moiety
BC: Bicyclic moiety
Cmpd Structure
IMPDH
MIC90
(µM)
against
WT
MIC (µM) Hyper-sensitivity to
90
No.
X=N
R =H
IC (µM)
against
guaB2 cKD
5
0
1
SRMV2.6
if not defined
otherwise
CY
N
1
0.1 ± 0.01
2
250
~32-fold
N
BC
8

ACCEPTED MANUSCRIPT
CY
NH
0
.5 ± 0.02
2
3
50
>250
100
ND*
ND
HN
BC
CY
N
CH3
>100
100
N
BC
CY
HN
4
5
6
7
9.7 ± 0.3
>100
>100
>100
>100
>100
>100
>100
>100
ND
ND
ND*
ND
N
BC
CY
N
14.7 ± 0.3
HN
BC
1
R =CH
3
CY
N
>100
N
BC
X=C
CY
N
3.0 ± 0.1
N
BC
9

ACCEPTED MANUSCRIPT
CY
N
O
N
S
O
8
1.6 ± 0.1
>100
>100
ND
1
1
1
1
1
1
1
1
1
1
1
1
66
67
68
69
70
71
72
73
74
75
76
77
*ND, not determined.
2.3.2. Set II: Modifications on compound 2
Based on the data obtained from the compounds in Set I, the ethylenediamine compound 2
was selected for further development. We explored the effect of ring size and found a
correlation with the inhibitory activity against the enzyme, with the IC value of 9 > 10 > 2
5
0
(Table 2). The cyclobutyl (10) and t-butyl (11) analogues showed similar activity profiles
presumably owing to their comparable size and polarity. However, all of the ethylenediamine
analogues had weak activity against M. tuberculosis (MIC : 50-100 µM) underscoring the
9
0
importance of the piperazine ring for whole-cell activity.
Table 2. SAR profiling of compound 2
CY
NH
HN
BC
1
78
Cmpd No
CY
IMPDH
MIC90
(µM)
against WT
MIC (µM)
against
SRMV2.6
Hyper-
sensitivity to
guaB2 cKD
90
IC (µM)
5
0
2
0.5 ± 0.02 50
>250
ND*
1
0

ACCEPTED MANUSCRIPT
9
1
6.9 ± 0.2
2.7 ± 0.1
100
100
100
100
ND
ND
0
1
1
2.8 ± 0.2
100
100
ND
1
1
1
1
1
1
1
1
1
1
1
79
80
81
82
83
84
85
86
87
88
89
*ND, not determined.
2.3.3. Set III: Modifications on the cyclohexyl ring
As reported previously, the cyclohexyl group of compound 1 was orientated to form strong
pi interactions with the Y487’ (Y471’ in the Mth structure) side chain from the adjacent
protomer in the IMPDH tetramer [5]. To explore the SAR around the cyclohexyl ring,
compounds 12-21 were synthesized and analysed in terms of IMPDH inhibition and whole-
cell activity (Table 3).
IIIa: Alicyclic derivatives
Compound 12, in which a methyl group was added at position-1 of the cyclohexyl ring of 1
maintained inhibitory activity against IMPDH (IC : 2.4 µM), but anti-tubercular activity was
5
0
1
1
1
1
90
91
92
93
ablated (MIC : 250 µM) suggesting that this modification affected compound permeation
90
and/or metabolism in M. tuberculosis. In contrast, the addition of a methyl group at position-
2 of the cyclohexyl ring moderately improved the inhibitory activity against IMPDH while
retaining modest target-selective, whole-cell activity (compound 13, MIC : 31.2 µM; IC :
9
0
50
1
1
1
94
95
96
0.084 µM). Replacement of the cyclohexyl by a cycloheptyl ring (compound 14) reduced
both enzyme and whole-cell activities by approximately one order of magnitude (MIC : 31.2
9
0
µM, IC : 0.83 µM) while maintaining target selectivity. The co-crystal structure of 14 with
5
0
1
1

ACCEPTED MANUSCRIPT
1
1
1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
97
98
99
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
Mth IMPDH confirmed the binding of this compound to the NAD-binding pocket of IMPDH
in a similar manner to compound 1. Pharmacophore analysis of the binding (Figure 4; Table
5) highlighted the pivotal role of interactions with 14 made at both hotspot regions identified
(Figures 2 and 3).
IIIb: Aromatic derivatives
Substitution of the cyclohexyl by a phenyl ring (compound 15) resulted in loss of both
biochemical and whole-cell activities (IC : 8.1 µM, MIC : >100 µM). However, other
5
0
90
derivatives containing substituted phenyl rings were more active than compound 15 (Table
3). The 3-cyano substitution (compound 16) had no effect on the IC but improved whole-
5
0
cell activity (MIC : 25 µM) whereas the 4-cyano substitution (compound 17) had a
9
0
deleterious effect on both biochemical and whole-cell activity. However, substitution of the
phenyl ring in compound 16 by a 3-nitrophenyl in compound 18 increased the inhibitory
activity against both the enzyme and M. tuberculosis (IC : 1.1 µM, MIC : 12.5 µM).
5
0
90
Electron density of the soaked complex with Mth IMPDH revealed that compound 18
exhibits extensive pi interactions with both IMP and A285 (A269 in the Mth structure) in the
first hotspot region. The tail of 18 extended up into the second hotspot region where it made
strong pi interactions with the Y487′ (Y471′ in the Mth structure) side chain from the
adjacent molecule in the tetramer (Figure 2, 3, and 4; Table 5). The addition of a second
nitro group at the meta position in the phenyl ring (compound 19) had no further effect on
either IC or MIC (3,5-dinitrophenyl, IC : 1.4 µM, MIC : 12.5 µM). Interestingly, the
5
0
90
50
90
3,4,5-trimethoxyphenyl derivative, compound 20, displayed reasonably good whole-cell
activity (MIC : 15.6 µM) despite having a poor IC (12 µM) suggesting that the
9
0
50
modifications are important for compound uptake.
Based on the inhibitory characteristics of compounds 18/19 and 2, we synthesized
compound 21, the ring-opened derivative of compound 19. Compound 21 (IC : 0.3 µM,
5
0
1
2

ACCEPTED MANUSCRIPT
2
2
2
2
2
2
2
2
2
2
22
23
24
25
26
27
28
29
30
31
MIC : 15.6 µM) was highly active against the enzyme, and was 4-fold more active against
9
0
the compound 1-resistant mutant strain, SRMV2.6, than wild-type M. tuberculosis (MIC :
9
0
3.9 µM vs. 15.6 µM). This compound also showed a 4-fold increase in potency against the
guaB2 cKD mutant upon guaB2 silencing, thus confirming its target selectivity for IMPDH
in M. tuberculosis and ability to bind at the active site of a mutant form of the enzyme in
which the Tyr487 residue is substituted by Cys. Docking analysis supported the notion that
compound 21 may exhibit a distinct mode of binding at the active site being independent of
interaction with Y487’ (Figure 5).
Table 3. SAR profiling of modifications on the cyclohexyl ring.
CY
N
N
BC
2
32
Cmpd
No
CY
IMPDH MIC90
MIC90
(µM)
against
SRMV2.6
Hyper-
sensitivity
to guaB2
cKD
IC50
(µM)
against
WT
(µM)
0
0
.1 ±
.01
1
2
250
~32-fold
ND*
2
0
.4 ±
.01
12
250
>250
1
3

ACCEPTED MANUSCRIPT
0
0
.084 ±
.0
1
3
31.2
31.2
>250
16-32-fold
0
0
.83 ±
.04
1
1
4
5
>250
>100
16-fold
ND
8.1 ± 0.1 >100
16
8.6 ± 0.5 25
50
ND
1
0
6.5 ±
.5
1
1
7
8
100
100
50
ND
1.1 ± 0.1 12.5
8-fold
1
0
.4 ±
.01
19
12.5
100
ND
1
4

ACCEPTED MANUSCRIPT
1
0
2.0 ±
.5
20
15.6
>125
ND
0
0
.3 ±
.02
21
15.6
3.9
4-fold
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
*ND, not determined.
2.3.4. Set IV: Increasing the distance between the carbonyl group and the cyclohexyl ring or
its analogues
In this set of compounds, the effect of (i) increasing the spacing between carbonyl group and
the cyclohexyl ring or its analogues and (ii) change of one nitrogen of piperazine to carbon
on the activity profile was investigated (Table 4).
IVa: Phenylacetic acid derivatives: methylene insertion
Insertion of a methylene between the carbonyl group and the cyclohexyl ring transformed the
compounds into phenylacetic acid derivatives (compounds 22-30; Table 4). Most of the
compounds in this series showed reasonable activity against IMPDH, with 22
(cyclohexylacetic acid) and 29 (4-fluorophenylacetic acid) being the most potent (IC : 0.23
5
0
µM and 0.68 µM, respectively). Although compound 22 and 29 were reasonably active
against M. tuberculosis and 22 retained target selectivity in whole cells, as evidenced by the
hypersensivity of the guaB2 cKD mutant to this compound, the MIC values of the other
9
0
phenylacetic acid analogues were unfavorable. Electron density of the soaked complexes of
1
5

ACCEPTED MANUSCRIPT
2
2
2
2
2
2
2
2
2
2
2
2
2
2
49
50
51
52
53
54
55
56
57
58
59
60
61
62
22, 27, and 29 with Mth IMPDH revealed that these molecules bound to the NAD binding
pocket and made strong pi interactions with IMP and A285, in addition to a network of polar
and hydrogen bonds to surrounding residues including T343, G4225 and E458 (Figure 3 and
4). Compounds 27 and 29 were also able to make interactions with adjacent protomer through
the side chain of P61′ and Y487′.
IVb: Phenylurea derivatives: NH group insertion
A series of analogues containing mono or di substitutions of the phenyl ring were synthesized
(compounds 31-45, Table 4). These compounds were similar to the phenylacetic acid
derivatives, but with a nitrogen in place of carbon. Some of the compounds showed good
inhibitory activity against the enzyme but limited whole-cell activity (Table 4). When the
phenyl ring was substituted at position 3, the resulting compounds generally displayed
reduced whole-cell activity while maintaining IC values comparable to that of compound 1
5
0
(Table 4). Similarly, when the phenyl ring was substituted at other positions, the resulting
compounds 33 (4-chloro phenylurea, MIC : 250 µM, IC : 15 µM); 39 (2-nitro, 4-methyl
9
0
50
2
63
phenylurea; MIC : 25 µM, IC : 0.6 µM); and 40 (2-chloro, 6-methyl phenylurea, MIC :
90 50 90
2
2
2
2
2
2
2
2
2
2
64
65
66
67
68
69
70
71
72
73
250 µM, IC : 3.4 µM) were markedly less active against M. tuberculosis. Despite displaying
50
limited whole-cell activity, the electron density of the soaked complex suggested that
compound 37 was able to bind to Mth IMPDH in a mode similar to 1 (Figure 4; Table 5). An
additional chlorine atom at position-4 resulted in substantial loss of both biochemical and
anti-mycobacterial activities (compounds 34 and 33). Thus, while substituting a carboxamide
by a urea moiety improved the inhibitory activities of the 3-substituted phenyl derivatives
against the enzyme, this did not translate into improved whole-cell activity, again pointing to
an impact on compound permeation and/or metabolism in M. tuberculosis.
Next, based on the inhibitory characteristics of compound 20, the
methylenedioxyphenyl ring was selected for further evaluation (compound 41-45, Table 4).
1
6

ACCEPTED MANUSCRIPT
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
In contrast to the other phenylureas, the 3,4 methylenedioxyphenyl urea derivative, 41,
showed promising whole-cell activity (MIC : 6.3 µM), cross-resistance to 1, and a 4-fold
9
0
increase in potency against the guaB2 cKD mutant coupled with potent activity against the
enzyme (IC : 0.08 µM). While compound 42, the thiophenylurea analogue of 41, retained
5
0
comparable activity against the enzyme (IC : 1.1 µM), the replacement of oxygen by sulfur
5
0
resulted in an 8-fold reduction in whole-cell activity. Replacement of the piperazine ring in
41 by the seven-membered homopiperazine ring in 44 resulted in an 8-fold reduction in
biochemical activity and a 2-fold reduction in whole-cell activity. Interestingly, the
phenyulrea compound 45, in which the piperazine ring in 41 was replaced by 3-
methylpiperazine, had a negligible effect on both biochemical and whole-cell activity. This is
in contrast to the dramatic loss of activity observed when the same piperazine ring
substitution was made in compound 1 to produce compound 3, as described above.
Moreover, both compounds 44 and 45 showed increased potency against the guaB2 cKD
mutant confirming that these compounds retained target selectivity for IMPDH in whole M.
tuberculosis cells. Structural analysis confirmed tight binding of compound 45 to Mth
IMPDH, similar to compound 1 (Figures 2, 3, and 4; Table 5).
IVc : Benzylurea derivatives: NHCH group insertion
2
Finally, to further increase the spacing between the carbonyl group and the cyclohexyl ring,
we synthesized a series of benzylurea derivatives (46-49, Table 4). Compound 46 (IC : 0.35
5
0
µM, MIC : 3.1 µM), exhibited biochemical and whole-cell activities comparable to 1 and
9
0
also retained target selectivity in M. tuberculosis. The 3-methoxybenzylurea derivative,
compound 47, was the most potent IMPDH inhibitor analysed in this study (IC : 0.035 µM),
5
0
demonstrating ~3-fold greater potency than 1, while maintaining whole-cell activity (MIC :
9
0
3.1 µM) as well as target selectivity in M. tuberculosis. However, compound 47 was not
active against SMRV2.6, a finding consistent with the docking analysis, which predicted a
1
7

ACCEPTED MANUSCRIPT
2
3
3
3
3
3
3
3
3
3
3
3
3
99
00
01
02
03
04
05
06
07
08
09
10
11
strong interaction between 47 and the Y487’ residue in the adjacent promoter (Figure 5).
Compound 48 (2-methyl benzylurea) also exhibited good anti-mycobacterial (MIC : 3.9
9
0
µM) and biochemical (IC : 0.27 µM) activity, while maintaining target selectivity in M.
50
tuberculosis. This compound was also reasonably active against the resistant mutant,
SRMV2.6 (MIC : 15.6 µM) even though docking analysis suggested an interaction with the
9
0
Y487’ residue (Figure 5). We also evaluated the benzylurethane, compound 49, which differs
from the benzylurea compound 46 by only one atom. While retaining activity against the
enzyme (IC : 0.8 µM), this substitution abolished whole-cell activity (MIC : 100 µM)
5
0
90
(Table 5), underscoring the importance of the nitrogen atom in the urea moiety for compound
engagement of the IMPDH target in M. tuberculosis.
Table 4. SAR profiling of compounds with increased distance of carbonyl group to the
cyclohexyl ring or its analogs.
CY
N
N
BC
3
12
Cmpd
No
CY
IMPDH
MIC90
(µM)
MIC90
(µM) sensitivity
Hyper-
IC (µM)
5
0
against against to guaB2
WT
SRMV
cKD
2
.6
1
0.1 ± 0.01
2
250
~32-fold
1
8

ACCEPTED MANUSCRIPT
0
0
.23 ±
.02
2
2
2
3
31.2
250
>250
>250
16-fold
ND*
3.3 ± 0.2
24
5.96 ± 0.2 125
>250
ND
2
2
5
6
4.0 ± 0.2
125
250
>250
>250
ND
ND
2.6 ± 0.2
27
1.98 ± 0.1 250
>250
ND
1
9

ACCEPTED MANUSCRIPT
1
0
1.48 ±
.3
28
250
>250
ND
0
0
.68 ±
.05
2
3
3
9
0
1
25.0
250
>100
>250
31.2
ND
ND
ND
5.3 ± 0.2
0
0
.15 ±
.03
15.6
0
0
.08 ±
.005
3
3
3
2
3
4
62.5
125
>250
50
ND
ND
ND
15.0 ± 0.5 250
0.16 ± 0.1 25
2
0

ACCEPTED MANUSCRIPT
3
3
5
6
5.6 ± 0.2
12.5
>100
ND
ND
0
0
.042 ±
.002
31.2
62.5
3
3
7
8
0.9 ± 0.06 125
>250
ND
ND
1.6 ± 0.2
100
100
3
4
9
0
0.6 ± 0.03 25
25
ND
ND
3.4 ± 0.1
250
>250
2
1

ACCEPTED MANUSCRIPT
0
0
.08 ±
.005
4
4
1
2
6.3
50
>100
>100
4-fold
ND
1.1 ± 0.1
O
O
O
43
>100
>100
>100
ND
NH
N
NH
BC
O
O
O
44
0.6 ± 0.02 12.5
>100
8-16-fold
NH
N
N
BC
O
O
O
CH3
0
0
.12 ±
.02
4
4
5
6
12.5
>100
>100
8-16-fold
NH
N
N
BC
0
0
.35 ±
.03
3.1
8-fold
2
2

ACCEPTED MANUSCRIPT
0
.035±
4
4
7
8
3.1
3.9
>100
15.6
8-fold
0
.002
0
0
.27 ±
.03
16-32-fold
49
0.8 ± 0.05 100
100
ND
3
3
13
14
*ND, not determined.
3
3
3
3
3
3
3
3
15
16
17
18
19
20
21
22
2.4. X-ray crystallography
All compounds were evaluated for structural analysis and co-crystal structures were obtained
for seven of these, namely compounds 14, 18, 22, 27, 35, 37 and 45. Computational analysis
of the Mth IMPDH (GuaB2) structure revealed two main hotspot regions, both of which were
located within the NAD binding pocket (Figure 2). One of these was located near the bound
IMP, where compounds could make apolar interactions to both the IMP and surrounding
residues. The other was located close to the entrance of the active site, where apolar
interactions, including to the neighbouring subunit, could be taken advantage of.
2
3

ACCEPTED MANUSCRIPT
3
3
23
24
Figure 2. Hotspot maps of IMP bound IMPDH. Calculated hydrogen donor regions are
3
25
coloured in blue, hydrogen acceptor regions in red and apolar regions in yellow.
3
3
3
3
3
3
3
3
3
26
27
28
29
30
31
32
33
34
Pharmacophore analysis of the binding of the IMPDH inhibitors (including 5J5R[5],
5K4X and 5K4Z [8]) (Figure 3) highlighted the pivotal role of interactions they made at both
hotspot regions identified. In the first hotspot region, all of the analogues made extensive pi
interactions with both IMP and A285 (A269 in the Mth structure). They could also make
weak hydrogen bonds to the side chain of T343 (T327 in the Mth structure) and the main
chain atoms of G334, V335, G336 and G425 (G318, V319, G320 and G409 in the Mth
structure). The tails of the enzyme inhibitors extended up into the second hotspot region,
where they all made strong pi interactions with the Y487′ (Y471′ in the Mth structure) side
chain from the adjacent molecule in the tetramer.
3
3
35
36
2
4

ACCEPTED MANUSCRIPT
3
3
37
38
Figure 3. Pharmacophore model of the NAD-binding pocket, showing the important
interactions made by our recently described inhibitors (including 5J5R[5], 5K4X and 5K4Z
[8]). The chemical characteristics of the small molecule binders are represented by mesh
spheres, with green representing hydrogen donors and positive ions, purple aromatic groups,
white and orange hydrogen donors and acceptors respectively, and blue and red positive and
negative ions respectively.
3
3
3
3
3
3
3
3
3
3
3
3
39
40
41
42
43
44
45
46
47
48
49
50
The compounds 14, 18, 22, 27, 35, 37 and 45 were also able to make additional
interactions to P61′ side chain from the adjacent protomer (Figure 4). In summary, it was
observed that the compounds establish strong interactions to the Mth IMPDH, suggesting that
factors such as compound permeation, metabolism and/or efflux could be the reason for their
moderate to weak anti-tubercular activities.
2
5

ACCEPTED MANUSCRIPT
3
3
51
52
Figure 4. The X-ray crystal structures of compounds bound to Mth IMPDH. (A) 1, (B) 14,
(C) 18, (D) 22, (E) 27, (F) 35, (G) 37, and (H) 45. Interactions made by the inhibitors in the
X-ray crystal structures of the complexes with Mth IMPDH (gray; and adjacent protomer in
cyan) were calculated by Arpeggio [17], with ring and pi interactions shown in green, amide–
amide in blue, hydrogen bonds in red, and polar interactions in orange. The electron density
of the soaked complexes revealed that all the compounds bound into the NAD-binding pocket
of IMPDH in a similar manner to compound 1.
3
3
3
3
3
3
3
3
3
3
3
3
53
54
55
56
57
58
59
60
61
62
63
64
In the case of compounds 21, 47, and 48, our attempts to obtain co-crystals with Mth
IMPDH were unsuccessful despite extensive efforts. As a result, docking analyses were
performed instead for these compounds (Figure 5). These analyses revealed that binding of
compound 21 to the IMPDH is independent of interaction with Y487’.
2
6

ACCEPTED MANUSCRIPT
3
3
65
66
Figure 5. Docking analyses of compounds 21 (A), 47 (B), and 48 (C) bound to Mth IMPDH.
Interactions made by compounds with Mth IMPDH (gray; and adjacent protomer in cyan)
were calculated by Arpeggio [17], with ring and pi interactions shown in green, amide–amide
in blue, hydrogen bonds in red, and polar interactions in orange
3
3
3
3
67
68
69
70
2
7

ACCEPTED MANUSCRIPT
Table 5. Data collection and refinement statistics. Statistics for the highest-resolution shell are shown in parentheses.
compound 14
6D4Q
compound 18
6D4R
compound 22
6D4V
compound 27
6D4U
compound 35
6D4W
compound 37
6D4S
compound 45
6D4T
PDB Code
Beamline
I24
I24
I24
I24
I24
I24
I24
Wavelength (Å)
Resolution range (Å)
0.97
0.97
0.97
0.97
0.97
0.97
0.97
61.64 - 1.71
42.33 - 1.34
(1.39 - 1.34)
I 4
63.5 - 2.02
(2.09 - 2.02)
I 4
61.21 - 1.7
(1.76 - 1.7)
I 4
35.93 - 1.8
(1.86 - 1.8)
I 4
63.22 - 1.63
(1.69 - 1.63)
I 4
61.18 - 1.54
(1.60 - 1.54)
I 4
(
1.77 - 1.71)
Space group
Unit cell
I 4
89.6 89.6 85.0 88.7 88.7 84.7 89.8 89.8 84.9 88.7 88.7 84.6 88.8 88.8 84.6 89.4 89.4 85.3 88.7 88.7 84.5
9
0 90 90
90 90 90
90 90 90
90 90 90
71796 (7189)
36025 (3612)
2.0 (2.0)
90 90 90
56465 (5608)
29914 (2956)
1.9 (1.9)
90 90 90
83302 (8319)
41811 (4173)
2.0 (2.0)
90 90 90
96068 (9539)
48275 (4799)
2.0 (2.0)
Total reflections
Unique reflections
Multiplicity
Completeness (%)
Mean I/sigma(I)
Wilson B-factor
R-merge
72178 (7214) 145626(14577) 44053 (4407)
36236 (3621)
2.0 (2.0)
73343 (7332)
2.0 (2.0)
22169 (2223)
2.0 (2.0)
100.0 (100.0)
13.6 (1.8)
24.16
99.9 (99.9)
12.9 (1.4)
17.45
100.0 (100.0)
10.6 (2.3)
29.05
100.0 (99.9)
5.4 (1.7)
98.5 (99.1)
12.5 (2.6)
21.84
100.0 (99.9)
16.2 (1.3)
22.71
100.0 (100.0)
11.9 (2.0)
16.46
20.55
0.047 (0.530)
0.066
0.037 (0.656)
0.052
0.105 (0.512)
0.149
0.082 (0.536)
0.116
0.057 (0.429)
0.080
0.034 (0.684)
0.048
0.040 (0.414)
0.057
R-meas
CC1/2
0.998 (0.34)
0.16 (0.28)
0.998 (0.487)
0.17 (0.29)
0.979 (0.23)
0.17 (0.28)
0.987 (0.283)
0.17 (0.25)
0.978 (0.57)
0.17 (0.24)
0.999 (0.51)
0.17 (0.27)
0.997 (0.472)
0.17 (0.26)
R-work
2
8