Structural comparison of Mtb-DHFR and h-DHFR for design, synthesis and evaluation of selective non-pteridine analogues as antitubercular agents

Article abstract of DOI:10.1016/j.bioorg.2018.04.022

Tuberculosis is an infectious disease that affects millions of population every year. Mtb-DHFR is a validated target that is vital for nucleic acids biosynthesis and therefore DNA formation and cell replication. This paper report identification and synthesis of novel compounds for selective inhibition of Mtb-DHFR and unleash the selective structural features necessary to inhibit the same. Virtual screening of databases was carried out to identify novel compounds on the basis of difference between the binding pockets of the two proteins. Consensus docking was performed to improve upon the results and best ten hits were selected. Hit 1 was subjected to analogues design and the analogues were docked against Mtb-DHFR. From the docking results 11 compounds were selected for synthesis and biological assay against H₃₇Rv. Most potent compound (IND-07) was tested for selectivity using enzymatic assay against Mtb-DHFR and h-DHFR. The compounds were found to have good inhibitory activity (25–200 μM) against H₃₇Rv and in enzyme assay against Mtb-DHFR and h-DHFR the compound was found selective towards Mtb-DHFR with selectivity index of 6.53. This work helped to identify indole moiety as novel scaffold for development of novel selective Mtb-DHFR inhibitors as antimycobacterial agents.

Full text of DOI:10.1016/j.bioorg.2018.04.022

Accepted Manuscript
Structural Comparison of Mtb-DHFR and h-DHFR for Design, Synthesis and
Evaluation of Selective Non-Pteridine Analogues as Antitubercular Agents
Kalicharan Sharma, Omprakash Tanwar, Shweta Sharma, Shakir Ali, M.M.
Alam, M.S. Zaman, Mymoona Akhter
PII:
S0045-2068(18)30141-X
https://doi.org/10.1016/j.bioorg.2018.04.022
YBIOO 2345
DOI:
Reference:
Bioorganic Chemistry
To appear in:
Received Date:
Revised Date:
Accepted Date:
14 February 2018
18 April 2018
24 April 2018
Please cite this article as: K. Sharma, O. Tanwar, S. Sharma, S. Ali, M.M. Alam, M.S. Zaman, M. Akhter, Structural
Comparison of Mtb-DHFR and h-DHFR for Design, Synthesis and Evaluation of Selective Non-Pteridine Analogues
as Antitubercular Agents, Bioorganic Chemistry (2018), doi: https://doi.org/10.1016/j.bioorg.2018.04.022
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.

Structural Comparison of Mtb-DHFR and h-DHFR for Design, Synthesis and Evaluation
of Selective Non-Pteridine Analogues as Antitubercular Agents
Kalicharan Sharma1, Omprakash Tanwar1, Shweta Sharma1, Shakir Ali2, M. M. Alam1, M. S.
Zaman1, Mymoona Akhter*1, 3
1Department of Pharmaceutical Chemistry, SPER, Jamia Hamdard, New Delhi, 110062, India
2
Department of Biochemistry, School of Chemical and Life Sciences, Jamia Hamdard, New
Delhi-110062
3 Bioinformatics Infrastructure Facility Lab, Jamia Hamdard, New Delhi-110062
✉Dr.Mymoona Akhter
Address: Drug Design and Medicinal Chemistry Lab, Dept. of Pharmaceutical Chemistry,
School of Pharmaceutical Education & Research, Jamia Hamdard, New Delhi, India, 110062.
Ph.: +91 9891369482
Mail ID: mymoonaakhter@gmail.com
Abstract: Tuberculosis is an infectious disease that affects millions of population every
year. Mtb-DHFR is a validated target that is vital for nucleic acids biosynthesis and
therefore DNA formation and cell replication. This paper report identification and
synthesis of novel compounds for selective inhibition of Mtb-DHFR and unleash the
selective structural features necessary to inhibit the same. Virtual screening of databases
was carried out to identify novel compounds on the basis of difference between the

binding pockets of the two proteins. Consensus docking was performed to improve upon
the results and best ten hits were selected. Hit 1 was subjected to analogues design and
the analogues were docked against Mtb-DHFR. From the docking results 11 compounds
were selected for synthesis and biological assay against H37Rv. Most potent compound
(
IND-07) was tested for selectivity using enzymatic assay against Mtb-DHFR and h-
DHFR. The compounds were found to have good inhibitory activity (25-200µM) against
H37Rv and in enzyme assay against Mtb-DHFR and h-DHFR the compound was found
selective towards Mtb-DHFR with selectivity index of 6.53. This work helped to identify
indole moiety as novel scaffold for development of novel selective Mtb-DHFR inhibitors
as antimycobacterial agents.
Keywords: Mtb-DHFR inhibitors, antimycobacterial, Indole moiety, 1DF7, 1OHJ,
antitubercular, consensus docking

1. Introduction
DHFR is an important enzyme in both mammals and microorganism including Mycobacterium
tuberculosis (Mtb). It catalyzes folic acid conversion to dihydro and tetrahydro folic acid which
is a crucial step in folate pathway1. Inhibition of folate pathway leads to interruption of
thymidine supply resulting cell death. Folate metabolism plays an essential role in the
biosynthesis of purines and pyrimidines and therefore in the nucleic acid biosynthesis for all the
living organisms and is thus directly or indirectly involved in the processes of cell reproduction2.
Tuberculosis (TB) is a top infectious disease killer worldwide. In 2016 global statistics showed
that 10.4 million suffered with TB and 1.7 million died from the disease. TB is a leading killer of
HIV-positive people: in 2016, 1 in 3 HIV deaths was due to TB. Globally in 2016, an estimated
3
490,000 people developed multidrug-resistant TB (MDR-TB) . New forms of MDR-TB and
extensively drug resistant TB (XDR-TB), pose new challenge for medicinal chemist. To address
new threat, there is an urgent need to develop new anti-TB drugs on novel targets having
minimum cross resistance with existing drugs. Thus, new anti-tuberculosis agents which could
act via unique mechanism is need of the hour. DHFR being crucial for survival of microorganism
has been explored extensively as drug target for example methotrexate4, trimethoprim5 and Br-
WR992106 an analogue of the antimalarial agent R992107 (Figure 1) have been found effective
clinically against tuberculosis. However, all clinically effective DHFR inhibitors are non-
selective, and inhibit both human and pathogenic DHFR more or less to a similar extent.
Figure 1: Structure of some reported DHFR inhibitors.
Mycobacterium (Mtb) and human (h) DHFRs show 26% structure similarity, but their
active sites are not identical8. Dissimilarities in the enzymatic constitution of the
microorganisms and mammals, forms the basis for design of selective inhibitors devoid of
toxicity to human cells9. Therefore a comparative computational study will help design
selective Mtb-DHFR inhibitors. The structural based virtual screening (SBVS) has gained
enormous attention and success in pharmaceutical industry over the past few decades for
discovering novel drugs10. The aim of this study was to develop a novel series of selective
Mtb-DHFR inhibitors as antitubercular agents using SBVS and synthetic approach
(Figure 2).

Figure 2: Identification of potential hit through virtual screening protocol and modification of
the same at three different sites to obtain the best compound IND-07
2. Experimental
2.1 In silico Work:
2.1.1 NCI and Drug Bank database: The National Cancer Institute compounds library (NCI
release 265,242 compounds, published in May 2012 and drug bank database, 8709 compounds)
were used for virtual screening. The 3D conformations of all the molecules were generated using
the maestro 9.6 (Glide version) virtual screening tools.
2.1.2 Software tools:Protein structure preparation was done by using Prepwizard (Maestro 9.6)
and ligands databases were prepared by using Ligprep tool (Maestro 9.6).
2.1.3 Protein Selection and preparation: Two proteinsMtb-DHFR(PDB ID: 1DF7) and h-DHFR
(PDBID: 1OHJ) were selected for VS workflow and docking protocols for identification of
selective Mtb-DHFR inhibitors. The selection of PDB’s was done on the basis of species and
resolution. The target proteins were then imported from RCSB Protein Data Bank saved in folder
and was prepared. The structure was preprocessed by adding hydrogens and check for missing
residues etc. The selection of water molecules was set at default and protein was optimized
followed by energy minimization at default setting. The grid was prepared and docking was
performed. The docking was carried on 1DF7 without glycerol moiety.
2.1.4 Structural comparison of Mtb-DHFR to h-DHFRand their binding pocket:
Till date thirteen crystal structures of DHFR have been reported, out thirteen, ten structures are
of Mtb-DHFR and three are of h-DHFR with NADP and different inhibitors. The resolution of
the structures deposited in PDB range from 1.7 to 2.0 Å, thereby making them excellent for
insilico studies. The bound inhibitors include methotrexate, trimethoprim and Br-WR99210 11-14
.
Mtb-DHFR and h-DHFR are quite similar in structure; Mtb-DHFR contains 159 amino acid
residues compared to 187 of the human protein with active site located at the C-terminal side of
the sheet. The structural comparison of two complexes revealed 26% amino acid sequence
similarity and also similarity between overall protein folds. However the environment of the
inhibitor binding site of the two enzymes has interesting differences like a glycerol molecule can

be seen in pocket of the Mtb-DHFR: MTX complex, while the same is absent in h-DHFR-COP
complex as the site is filled with hydrophobic side-chains. These differences between the
pathogen enzyme and host enzyme provide opportunities for designing new selective inhibitors
of Mtb-DHFR12.
2.1.5 Study of key difference in binding regions of Mtb-DHFRand h-DHFR:
In the ternary complex of MTX with Mtb-DHFR, the side of the aminopterin ring is accessible to
the solvent which is not in case of h-DHFR (Figure 3). Also in Mtb-DHFR a glycerol molecule
is found in a depression near MTX, Trp22, Asp27 and Gln28 whereas in case ofh-DHFRs
(complexed folate or COP) the site is well packed with three hydrophobic residue side chains,
Leu22, Pro26 and Phe31 which correspond to Leu20, Arg23 and Gln28 in Mtb-DHFR(Figure 3a
and 3b). Therefore if a compound is designed which could replace the glycerol from the pocket
and interact with the site in addition to other important interactions may result in development of
specific Mtb-DHFR inhibitor. On this assumption the present study was undertaken to design
novel selective inhibitors for Mtb-DHFR.12
Figure 3: Comparison of inhibitor binding site in Mtb-DHFR and h-DHFR at catalytic domain.
In Mtb DHFR inhibitor binding site a glycerol molecule bound close to the inhibitor MTX is
visible. (Li R. et al12)
2.2 Virtual Screening against Mtb-DHFR and h-DHFR:
Virtual screening is one of the most widely used approaches for identifying novel hits and for
removing non complementary compounds, which can lower the selectivity of compound for
targets. Since virtual screening workflow is based on docking of ligands on protein active site,
validation of docking protocol become important15. The validation of docking procedure was
carried out by extracting the co-crystallized ligands MTX and COP from the respective protein
complexes obtained from protein data bank and re-docked on the active site of Mtb-DHFR and
h-DHFR respectively.
Compounds were collected from NCI and drug bank database and merged to create a single
database (2, 73,951 compounds). This database was subjected to ADME and Lipinski rule
screening followed by virtual screening workflow against PDB 1DF7. In virtual screening

workflow compounds were subjected to HTVS, SP and XP docking processes. In each one of
these steps, by default top 5% of the compounds were selected for the next step. On the basis of
score top 100 hits were selected from the XP screening and visually analysis and then subjected
to screening against h-DHFR (PDBID: 1OHJ) for selectivity. Fifty compounds which showed
lowest docking score against h-DHFR and high score against Mtb-DHFR were selected for
further study. Virtual screening work flow is depicted in Figure 4.
Figure 4. Virtual Screening Workflow
2.3 Consensus Docking:
compounds selective towards Mtb-DHFR were also subjected to consensus docking using
GLIDE module of Schrödinger 9.4 and GOLD suite 5.2.216-17
.
2.3.1 Docking by using Glide module of Maestro 9.4: Docking of the selected fifty compounds
was carried out on Glide module of Maestro 9.4 as per the Glide protocol given in Schrödinger.
All the default parameters were used. For ligand preparation the pH was 7.0±2.0, force field was
OPLS3 and ionization was done using Epik. For protein preparation the pH was 7.0±2.0, force
field was OPLS3, ionization was done using Epikand the water molecules within 5 Å were kept
and rest were deleted.
2.3.2 Docking by using GOLD Suite 5.2.2: GOLD suite 5.2.2 was used for consensus docking of
final 50 compounds selected using the wizard with default parameters population size (100);
selection- pressure (1.1); number of operations (10,000); number of islands (1); niche size (2);
and operator weights for migrate (0), mutate (100), and crossover (100) were applied. The active
site with a 10 Å radius sphere was defined by selecting an active site residue of protein. Default
Genetic Algorithm settings were used for all calculations and a set of 10 solutions were saved for
each ligand.
Table 1: Structure of final hits selected after virtual screening of NCI and drug bank followed by
consensus docking using Gold suite
On the basis of results of consensus docking ten (10) hits were selected out of which hit 1was
selected for synthesis.

2.4 Chemistry:
All chemicals were purchased from local commercial suppliers and used as such or otherwise
mentioned. Melting points were determined by open capillary method by digital melting point
1
13
apparatus by ‘VEEGO’ and are uncorrected. H NMR and C NMR spectra were recorded on a
BrukerAvance II 400 spectrometer, using DMSO-d6 as solvent and TMS as an internal standard.
Mass spectral analysis was carried out using Applied BiosystemQtrap 3200 MS/MS system in
ESI mode. CHN analysis was done by Elementar, Germany. Infrared (IR) spectra were recorded
on a Shimadzu FT-IR 8400S infrared spectrophotometer using the ATR accessory. Reactions
were monitored by TLC using pre-coated silica gel aluminium plates (Kiesel gel 60, 254, E.
Merck, Germany); zones were detected visually under ultraviolet irradiation or exposing to I2
vapour. Analytical reversed-phase high performance liquid chromatography (HPLC) was
performed on a Shimadzu CLASS-VP system equipped with an auto sampler and a photodiode
array detector. Purity of the final compounds was determined using chromatographic systems;
column, Princeton SPHER- 100, C18 (particle size = 5 mm, pore size = 10 nm, dimensions =
50mmx 4.6 mm); mobile phase A, methanol; mobile phase B, acetonitrile. Using a flow rate of
1.0 mL/min, gradient elution was performed from 10% B to 90% B over 6 min. In every case, 10
mL of a 20 µl solution was injected.

Scheme1: Synthetic route for generation of starting maaterialindole Scaffold
Scheme 2: Synthetic route of indoleester derivatives

Scheme 3: Synthesis route of indoleketo derivatives
2.4.1 Ethyl 1-benzyl-5-hydroxy-2-methyl-1H-indole-3-carboxylate (JM-01):
A mixture of benzyl amine (10.70g, 100 mmol), ethyl acetoacetate (12.98g, 100 mmol) and p-
⁰
TsOH (0.86g, 5 mmol) in MeOH in a 100 mL conical flask were refluxed for 2h at 45 C. The
reaction was followed by TLC and after completion of the reaction (~2h) the reaction mixture
was evaporated and quenched with cold water 200 mL and extracted with dichloromethane
(3x500 mL). The organic layer was washed successively with saturated sodium bicarbonate,
brine and finally with water, dried over anhydrous Na2SO4 and then concentrated. The
intermediate (JMA) obtained was directly used for next step without purification.
To a solution of 1, 4-benzoquinone (17.75 g, 0.1643 mol) in CH2Cl2 (300mL) was added ZnCl2
(14.68 g, 0.1095 mol). The resultant mixture was heated to boiling and a solution of JMA(30.0 g,
0.1369 mol) in CH2Cl2 (100 mL) was added drop wise with stirring in 10 minutes. The mixture
was stirred under boiling for additional 40 minutes19. After completion of the reaction, the
⁰
reaction mixture was kept at 0-5 C for 4-5 hours. The crystals precipitated were filtered off and
washed with CH2Cl2 (3 x 1 mL) and acetone (2 x 1 mL) to afford 30.5 g (82%) of JM-01, m.p.

⁰
212-214 C,1H NMR (DMSO- d6): 1.33 (3H, t, CH3,J= 6.8), 2.62 (3H,s, CH3), 4.245 (2H, q, CH2,
J= 6.8), 5.41 (2H, s, CH2), 6.15 (1H, dd, ArCH, J= 8.4), 6.981 (1H, d, ArCH, J= 6.8), 7.23 (4H,
m, ArCH), 7.39 (1H, s, ArCH), 8.93(1H, s, OH); 13C NMR (100 MHz, DMSO-d6) δ 166.11,
153.712, 145.559, 136.308, 131.895, 128.958(2C), 127.696(3C), 125.916, 112.587, 110.312,
105.519, 104.480, 46.666, 14.644, 12.032; ESI-MS (m/z) : 310.20 (M+1)+.
2.4.2 1-(1-benzyl-5-hydroxy-2-methyl-1H-indol-3-yl)ethanone (JM-02):
Synthesis of JM-02 was carried out by following the same procedure as in JM-01 except that
acetyl acetone(100 mmol) was used in place of ethyl acetoacetate (100 mmol) to afford 28 g
⁰
1
(79%) of JM-02, mp, 206-208 C, H NMR (DMSO-d6): 2.48 (3H, s,CH3), 2.63 (3H, s, CH3),
5.43 (2H, s, CH2), 6.39 (1H, dd,ArCH, J=8.8), 7.23-7.35 (5H, m, ArCH), 7.48 (1H, d, ArCH,
J=8.8), 7.56 (1H, s, ArCH), 8.88(1H, s, OH); 13C NMR (100 MHz, DMSO-d6) δ 194.443,
153.867, 145.231, 136.044, 132.096, 129.018, 127.696(2C), 125.896, 114.722, 112.147,
110.576, 105.899, 46.666, 31.590, 12.948;ESI-MS (m/z) : 280.30 (M+1)+.
2.4.3 2-((3-acetyl-1-benzyl-2-methyl-1H-indol-5-yl) oxy) acetic acid (IND-06)
⁰
⁰
To a solution of IND-11 (1eq) in THF: H2O (1:1) was added LiOH (3eq) at 0 C to 80 C for 1h.
After completion of reaction, reaction mixture was acidified with HCl and extracted with ethyl
acetate followed by drying of organic layer over sodium sulphate. The evaporation of solvent
under vacuum gave white powdered compound.
White solid, Yield= 72%, HPLC purity 100%, Rf = 0.26, mp, 172-174 °C; 1H NMR (400 MHz,
DMSO-d6), δ 2.526 (3H, s, CH3), 2.642 (3H, s, CH3), 4.627 (2H, s, CH2), 5.477 (2H, s, CH2),
6.805 (2H, dd, Ar-CH, J=8.8), 6.999 (2H, dd, Ar-CH, J= 6.8), 7.251 (3H, m, Ar-CH), 7.403
13
(2H, d, Ar-CH, J= 8.8), 7.508(1H, s, Ar-CH); C NMR (100 MHz, DMSO-d6) δ 193, 171.18,
154.47, 145.37, 137.61, 131.84, 129.24, 127.20(3C), 114.27, 111.71, 105.05, 65.98, 46.26,
31.80, 13.24; IR (cm-1): 3743.13 (OH), 1699.59 (C=O), 1528.98 (C-O, Bend); elemental analysis
(%) calculated/found: C(71.20/71.04); H(5.68/5.55); N(4.15/4.03); O(18.97/18.69); ESI-MS
(m/z): 338.17(M+1)+.
2.4.4 4-((3-acetyl-1-benzyl-2-methyl-1H-indol-5-yl) oxy) butanoic acid (IND-07)

⁰
⁰
To a solution of IND-14 (1eq) in THF: H2O (1:1) mixture was added LiOH (3eq) at 0 C to 80 C
for 1h. After completion of reaction, mixture was acidified with HCl, extracted with ethyl acetate
and organic layer was dried over sodium sulphate. After removing the solvent under vacuum
white powdered compound was obtained which was characterized.
1
White solid, Yield=69%, HPLC purity 100%, Rf = 0.35, mp, 140-142 °C; H NMR (400 MHz,
DMSO-d6), δ 1.938(2H, q, CH2, J=13.2), 2.338 (2H, t, CH2, J=14.4), 2.540 (3H, s, CH3), 2.646
(3H, s, CH3), 3.994 (2H, q, CH2, J= 11.6), 5.486 (2H, s, CH2), 6.795 (2H, d, Ar-CH, J=8.8),
6.996 (2H, d, Ar-CH, J= 6.8), 7.253 (3H, m, Ar-CH), 7.398 (1H, d, Ar-CH, J=8.8), 7.535 (1H, s,
Ar-CH); 13C NMR (100 MHz, DMSO-d6) δ 193.67, 174.72, 155.10, 145.24, 137.65, 131.65,
129.22, 127.20 (3C), 114.32, 111.84, 104.91, 67.49, 46.25, 30.75, 24.96; IR (cm-1): 3558.99
(OH), 1696.95 (C=O), 1410.71 (C-O, Bend); elemental analysis (%) calculated/found:
C(72.31/72.04); H(6.34/6.19); N(3.83/3.73); O(17.51/17.49); ESI-MS (m/z) :366.12(M+1).
2.4.5 1-benzyl-5-(3-hydrazinyl-3-oxopropoxy)-2-methyl-1H-indole-3-carbohydrazide (IND-08)
⁰
⁰
To a solution of IND-08a (1eq) in ethanol was added hydrazine hydrate (5eq) at 0 C to 80 C for
16h. After the completion of reaction, mixture was poured onto ice, extracted with ethyl acetate
and organic layer was dried over sodium sulphate. After removing the solvent under vacuum
white powdered compound was obtained which was characterized.
1
⁰
White solid, Yield=82%, HPLC purity 98.62%, Rf = 0.25, mp, 152-154 C; H NMR (400 MHz,
DMSO-d6), δ 1.356 (3H, t, NH, J=6.8), 2.648 (3H, s, CH3), 4.282 (4H, m, CH2,NH), 4.478 (2H,
s, CH2), 6.856 (1H, d, Ar-CH, J=8.8), 6.974 (2H, d, Ar-CH, J=7.6 ), 7.249 (3H, m, Ar-CH),
7.394 (1H, d, Ar-CH, J= 8.8), 7.517(1H, s, Ar-CH); 13C NMR (100 MHz, DMSO-d6) δ 167.42,
165.43, 154.17, 145.98, 137.63, 131.88, 129.23, 127.33(3C), 114.92, 105.07, 103.62, 67.31,
59.46, 46.40, 14.90, 12.30; IR (cm-1): 3741.06 (NH), 1678.59 (C=O), 1526.45 (C-N, Bend),
1177.90 (C-O, Bend); elemental analysis (%) calculated/found: C(62.98/62.84); H(6.08/6.01);
N(18.36/18.03); O(12.58/12.49); ESI-MS (m/z) : 380.31 (M-1)-.
2.4.6 2-((1-benzyl-3-(ethoxycarbonyl)-2-methyl-1H-indol-5-yl) oxy) acetic acid (IND-09)
⁰
⁰
To a solution of IND-12 (1eq) in THF: H2O (1:1) was added LiOH (3eq) at 0 C to 80 C for 1h.
After completion of reaction, reaction mixture was acidified with HCl and extracted with ethyl

acetate followed by drying of organic layer over sodium sulphate. The evaporation of solvent
under vacuum gave white powdered compound.
1
⁰
White solid, Yield=65%, HPLC purity 98.28%, Rf = 0.32, mp, 160-162 C; H NMR (400 MHz,
DMSO-d6), δ 1.338 (3H, t, CH3, J=7.2), 2.638 (3H, s, CH3), 4.248 (2H, q, Ar-CH, J= 6.8), 4.566
(2H, s, CH2), 5.452 (2H, s, CH2), 6.781 (1H, d, Ar-CH, J= 8.8), 6.972 (2H, d, Ar-CH, J= 7.2),
7.254 (3H, m, Ar-CH), 7.365 (1H, d, Ar-CH, J= 8.8), 7.448 (1H, s, Ar-CH); 13C NMR (100
MHz, DMSO-d6) δ 171.33, 165.65, 154.33, 145.87, 137.68, 131.68, 129.22, 127.20, 111.89,
104.59, 66.08, 59.42, 46.83, 14.82, 12.28;IR (cm-1): 3649.86 (OH), 1689.59 (C=O), 1184.76
(OH, Bend); elemental analysis (%) calculated/found: C(68.65/70.62); H(5.76/5.45);
N(3.81/3.78); O(21.77/21.49); ESI-MS (m/z) : 366.19 (M-1)-.
2.4.7 2-((3-acetyl-1-benzyl-2-methyl-1H-indol-5-yl) oxy) acetohydrazide (IND-10)
⁰
To a solution of IND-11 (1eq) in ethanol was added hydrazine hydrate (1.2eq) at 80 C for 12h.
After the completion of reaction, mixture was poured onto ice, extracted with ethyl acetate and
organic layer was dried over sodium sulphate. After removing the solvent under vacuum white
powdered compound was obtained which was characterized.
1
⁰
White solid, Yield=77%, HPLC purity 100%, Rf = 0.22, mp, 148-150 C; H NMR (400 MHz,
DMSO-d6), δ 2.548 (3H, s, CH3), 2.651 (3H, s, CH3), 4.331 (2H, s, NH2), 4.454 (2H, s, CH2),
5.493 (2H, s, CH2), 6.871 (1H, dd, Ar-CH, J=8.8 ), 6.974 (2H, m, Ar-CH), 7.255 (3H, m, Ar-
CH), 7.425(1H, d, Ar-CH, J=8.8), 7.572 (1H, s, Ar-CH), 9.348(1H, s, NH); 13C NMR (100 MHz,
DMSO-d6) δ,193.65, 167.59(2C), 154.37, 145.48, 137.67, 131.97, 129.24(2c), 127.12(3c),
114.27, 111.80(2C), 105.48, 67.45, 46.25, 31.89, 13.23; IR (cm-1): 3632.40 (OH), 1674.57
(C=O), 1614.55(C=O), 1164.81 (OH, Bend); elemental analysis (%) calculated/found:
C(71.20/71.09); H(5.68/5.57); N(4.15/3.99); O(18.97/18.79); ESI-MS (m/z) : 352.22 (M+1)+.
2.4.8 Ethyl 2-((3-acetyl-1-benzyl-2-methyl-1H-indol-5-yl) oxy) acetate (IND-11)
To a mixture of JM-02 (1eq) and ethylbromoacetate (1.2eq) in 50 ml dry acetone was treated
with anhydrous K2CO3 (3eq) and allowed to reflux for 6 h. The reaction mixture was then filtered
and washed thoroughly with acetone. The solvent was evaporated, the product thus separated
was purified by recrystallization from ethanol to afford the white crystalline compound.

1
⁰
White solid, Yield=78%,HPLC purity 97.35%, Rf = 0.56, mp, 240-242 C; H NMR (400 MHz,
DMSO-d6), δ 1.203 (3H, t, CH3, J=7.2), 2.534 (3H, s, CH3), 2.650(3H, s, CH3), 4.160(2H, q,
CH2, J=14.4), 4.802 (2H, s, CH2), 5.492 (2H, s, CH2), 6.827 (1H, dd, Ar-CH, J= 8.8), 7.000 (2H,
d, Ar-CH, J=7.2 ), 7.255 (3H, m, Ar-CH), 7.429(1H, d, Ar-CH, J=8.8), 7.512(1H, d, Ar-CH,
J=2.4); 13C NMR(100 MHz, DMSO-d6) δ, 193.61, 169.55, 154.18, 145.50, 137.59, 132.01,
129.24, 127.12(3c), 114.31, 111.75(2C), 105.24, 65.94, 61.02, 46.28, 31.79, 14.57, 13.24; IR
(cm-1): 3103.28 (Ar-Str-CH), 1621.07 (C=O), 1449.49(C-O); elemental analysis (%)
calculated/found: C(72.31/67.66); H(6.34/6.22); N(3.83/3.77); O(17.51/17.45); ESI-MS (m/z)
:366.17(M+1)+.
2.4.9 Ethyl 1-benzyl-5-(2-ethoxy-2-oxoethoxy)-2-methyl-1H-indole-3-carboxylate (IND-12)
To a mixture of JM-01 (1eq) and ethylbromoacetate (1.2eq) in 50 ml dry acetone was treated
with anhydrous K2CO3 (3eq) and allowed to reflux for 6 h. The reaction mixture was then filtered
and washed thoroughly with acetone. The solvent was evaporated, the product thus separated
was purified by recrystallization from ethanol to afford the white crystalline compound.
1
⁰
White solid, Yield=85%, HPLC purity 92.43%, Rf = 0.36, mp, 120-122 C; H NMR (400 MHz,
DMSO-d6), δ 1.185 (3H, t, CH3, J=7.2), 1.343 (3H, t, CH3, J=7.2), 2.649 (3H, s, CH3), 4.156
(2H, Q, CH2, J=14.4), 4.262 (2H, Q, CH2, J=14.4), 4.757 (2H, S, CH2, J=6.4), 5.469 (2H, S,
CH2), 6.811 (1H, q, Ar-CH, J=8.8), 6.978 (2H, d, Ar-CH, J=7.2), 7.250 (3H, m, Ar-CH), 7.414
(2H, m, Ar-CH, J=8.8); 13C NMR(100 MHz, DMSO-d6) δ, 169.61, 165.40, 153.97, 146.11,
137.61, 131.90, 129.23, 127.12(3c), 112.02(2C), 104.72, 65.77, 61.04, 59.43, 46.41, 31.79,
14.65, 12.25; IR (cm-1): 3103.28 (Ar-Str-CH), 1752.53 (C=O), 1682.58(C=O), 1527.93(C-O);
elemental analysis (%) calculated/found: C(69.86/69.77); H(6.37/6.19); N(3.54/3.51);
O(20.23/19.99); ESI-MS (m/z) :396.22 (M+1)+.
2.4.10 1-(1-benzyl-5-(benzyloxy)-2-methyl-1H-indol-3-yl) ethanone (IND-13)
To a mixture of JM-02 (1eq) and benzylbromide (1.2eq) in 50 ml dry acetone was treated with
anhydrous K2CO3 (3eq) and allowed to reflux for 6 h. The reaction mixture was then filtered and
washed thoroughly with acetone. The solvent was evaporated, the product thus separated was
purified by recrystallization from ethanol to afford the white crystalline compound.

White solid, Yield=86%, HPLC purity 99.93%, Rf = 0.58, mp, 180-182 °C; 1H NMR (400 MHz,
DMSO-d6), δ 2.532 (3H, s, CH3), 2.464 (3H, s, CH3), 5.122 (2H, s, CH2), 5.478 (2H, s, CH2),
6.888 (1H, d, Ar-CH, J=8.8), 7.000(2H, s, Ar-CH), 7.329 (9H, m, Ar-CH), 7.638 (1H, s, Ar-
CH);13C NMR(100 MHz, DMSO-d6) δ 193.64, 154.92, 145.31, 138.01, 137.62, 131.75,
128.76(5C), 114.31, 112.19, 105.42, 70.20, 46.26, 31.86, 13.24; IR (cm-1): 3030.28 (Ar-Str-CH),
1617.83 (C=O), 1485.53 (C-O); elemental analysis (%) calculated/found: C(81.27/81.06);
H(6.27/6.21); N(3.79/3.59); O(8.66/8.58); ESI-MS (m/z): 370.21 (M+1)+.
2.4.11 Ethyl 4-((3-acetyl-1-benzyl-2-methyl-1H-indol-5-yl) oxy) butanoate (IND-14)
To a mixture of JM-02 (1eq) and ethylbromobutyrate (1.2eq) in 50 ml dry acetone was treated
with anhydrous K2CO3 (3eq) and allowed to reflux for 6 h. The reaction mixture was then filtered
and washed thoroughly with acetone. The solvent was evaporated, the product thus separated
was purified by recrystallization from ethanol to afford the white crystalline compound.
1
⁰
White solid, Yield=69%, HPLC purity 100%, Rf = 0.68, mp, 232-234 C; H NMR (400 MHz,
DMSO-d6), δ 1.159 (3H, t, CH3, J=7.2), 1.975 (2H, t, CH2, J=6.8), 2.463 (2H, t, CH2, J=7.6),
2.538 (3H, s, CH3), 2.646 (3H, s, CH3), 4.027 (4H, m, CH2), 5.481 (2H, s, CH2), 6.784 (1H, d,
Ar-CH, J=8.8), 7.001 (2H, d, Ar-CH, J=7.2), 7.252 (3H, m, Ar-CH), 7.396 (1H, d, Ar-CH,
J=8.8), 7.537 (1H, s, Ar-CH); 13C NMR(100 MHz, DMSO-d6) δ 193.64, 173.11, 155.06, 145.24,
137.64, 131.61, 129.21, 127.31(3C), 114.33, 104.94, 67.38, 60.33, 46.25, 31.84, 30.74, 24.93,
14.58, 13.26;IR (cm-1): 2903.25 (Ar-Str-CH), 1703.76 (C=O), 1526.28 (C-O); elemental analysis
(%) calculated/found: C(73.26/73.21); H(6.92/6.83); N(3.56/3.51); O(16.26/15.99); ESI-MS
(m/z) :416.25 (M+Na)+.
2.4.12 Ethyl 3-((3-acetyl-1-benzyl-2-methyl-1H-indol-5-yl) oxy) propanoate (IND-15)
To a mixture of JM-02 (1eq) and ethylbromopropionate (1.2eq) in 50 ml dry acetone was treated
with anhydrous K2CO3 (3eq) and allowed to reflux for 6 h. The reaction mixture was then filtered
and washed thoroughly with acetone. The solvent was evaporated, the product thus separated
was purified by recrystallization from ethanol to afford the light brown compound.
1
⁰
Light brown solid, Yield=55%, Rf = 0.62, mp, 215-217 C; H NMR (400 MHz, DMSO-d6), δ
2.487 (3H, s, CH3), 2.629 (3H, s, CH3), 5.436 (2H, s, CH2), 6.641 (1H, dd, Ar-CH, J=8.8), 6.702

(2H, dd, Ar-CH, J=7.2), 7.265 (4H, m, Ar-CH), 7.428 (1H, d, Ar-CH, J=2.4); 13C NMR(100
MHz, DMSO-d6) δ 193.45, 153.61, 144.93, 137.72, 130.87, 129.20, 127.60(3C), 113.95, 111.95,
106.03, 46.18, 31.77, 13.23; IR (cm-1): 2903.25 (Ar-Str-CH), 1703.76 (C=O), 1526.28 (C-O);
elemental analysis (%) calculated/found: C(72.80/72.71); H(6.64/6.63); N(3.69/3.59);
O(16.87/16.81); ESI-MS(m/z): 402.20 (M+Na)+.
2.4.12 Ethyl 1-benzyl-5-(2-hydrazinyl-2-oxoethoxy)-2-methyl-1H-indole-3-carboxylate (IND-16)
⁰
To a solution of IND-12 (1eq) in ethanol was added hydrazine hydrate (1eq) at 80 C for 6h.
After the completion of reaction, mixture was poured onto ice, extracted with ethyl acetate and
organic layer was dried over sodium sulphate. After removing the solvent under vacuum white
powdered compound was obtained which was characterized.
1
⁰
White solid, Yield=77%, HPLC purity 100%, Rf = 0.55, mp, 198-200 C; H NMR (400 MHz,
DMSO-d6), δ 1.358 (3H, t, CH3, J=7.2), 2.649 (3H, s, CH3), 4.285 (4H, m, CH2, NH), 4.481 (2H,
S, CH2), 5.468 (2H, s, CH2), 6.854 (1H, dd, Ar-CH, J=8.8), 6.974 (2H, d, Ar-CH, J=7.2), 7.259
(3H, m, Ar-CH), 7.395 (1H, d, Ar-CH, J=8.8) 7.517 (1H, d, Ar-CH, J=2.4); 13C NMR(100 MHz,
DMSO-d6) δ 167.42, 165.43 154.18, 145.98, 137.63, 131.89, 129.23, 127.23(3C), 111.83(2C),
105.10, 103.63, 67.31, 59.46, 46.40, 14.90, 12.29; IR (cm-1): 3283.58 (NH2), 2963.25 (Ar-Str-
CH), 1682.34 (C=O), 1532.95 (C-O); elemental analysis (%) calculated/found: C(72.80/72.71);
H(6.64/6.63); N(3.69/3.59); O(16.87/16.81); ESI-MS (m/z): 404.23 (M+Na)+.
2.5 Biological Assay:
2.5.1 MIC against H37Rv assay:
The biological assay was performed by TAACF center, NIAD, USA. The MIC of compound was
determined by measuring bacterial growth after 5 days in the presence of test compounds. Ten–
point two-fold serial dilutions of compounds were prepared in DMSO and diluted with 7H9-Tw-
OADC medium in 96-well plates with a final DMSO concentration of 2%. The highest
concentration of compound was 200 µM where compounds were soluble in DMSO at 10 mM.
For compounds with limited solubility, the highest concentration was 50x less than the stock
concentration e.g. 100 µM for 5 mM DMSO stock, 20 µM for 1 mM DMSO stock. For potent
compounds, assays were repeated at lower starting concentrations. Each plate included assay

controls for background (medium/DMSO only, no bacterial cells), zero growth (100 µM
rifampicin) and maximum growth (DMSO only), as well as a rifampicin dose response
curve.OD590 and fluorescence (Ex 560/Em 590) were measured using BioTek™ Synergy 4 plate
reader for M. tuberculosis inoculated plates which successively incubated for 5 days. Growth
was calculated separately for OD590 and RFU. To calculate the MIC, the 10-point dose response
curve was plotted as % growth and fitted to the Gompertz model using Graph Pad Prism 5.
(Figure 5A). In addition, dose response curves were generated using the Levenberg-Marquardt
algorithm and the concentrations that resulted in 50% and 90% inhibition of growth were
determined (IC50 and IC90 respectively) (Figure 5B)20.
Figure 5: Dose response curve used to calculate MIC, IC50, and IC90. Data points obtained from
a dose response growth inhibition assay are curve fitted using (A) the Gompertz model to
calculate MIC and (B) the Levenberg-Marquardt algorithm to calculate IC50 and IC90.
2.5.2 In vitro Mtb-DHFR and h-DHFR enzyme assay:
DHFR enzyme was expressed and purified as described in literature (30). Enzyme assays were
⁰
performed in 100 mM HEPES, 50 mMKCl, pH 7.0 at 25 C. The decrease in absorbance at 340
nm representing the oxidation of NADPH was monitored with a spectrophotometer. Solution of
test compound (IND-07) was added to a 1 mL cuvette at various concentrations, with the
individual DHFR s (20 nM), and 40 μM of NADPH, and the reaction was initiated by the
addition of 40 μM of dihydrofolate. For IC50 determinations, 20 nM of DHFR s was incubated
with 40 μM of cofactor NADPH and seven serially diluted concentrations of the inhibitors for 1
minute. The reaction was initiated by the addition of 40 μM dihydrofolate21-22
.
2.6 Molecular dynamic simulations:
The synthesized compound (IND-07) which showed highest activity was selected for MD
simulation. MD simulations of IND-07 at the binding site of Mtb-DHFR (PDBID: 1DF7) were
performed using the Desmond package incorporated in the Maestro. The system was built by
applying OPLS-AA force field in an explicit solvent with the single point charges (SPC) water
model (OPLS-AA/SPC). The initial coordinates for the MD calculations were taken from the
docking experiments. The SPC water molecules were then added and system was neutralized by
adding Na+ counter-ion to balance the net charges of the system. After the construction of the

solvent environment, the complex system was composed of approximately 25241 atoms. Before
equilibration and long production MD simulations, the systems were minimized and pre-
equilibrated using the default relaxation routine implemented in Desmond. The MD simulations
were run for 10 ns and during the MD simulations, the equations of motion were integrated with
a 2 fs time step in the NVT ensemble. The SHAKE algorithm was applied to all hydrogen atoms;
the van der Waals (VDW) cutoff was set to 9Ǻ. The temperature was maintained at 300 K,
employing the Nosé–Hoover thermostat method with a relaxation time of 1ps. The trajectory
recording interval was kept for every 10 ps during entire MD runs23.
2.7 AMDET and MM/GBSA Study:
Eleven compound synthesized were subjected to ADMET analysis by QikProp (Qikprop,
Version 3.5). It provides ranges for comparing the properties of a particular molecule with those
of 95% of known drugs. The descriptors calculated were partition coefficient, human oral
absorption, CNS activity and gut-blood barrier permeability.
The free binding energy studies of complexes of final 11 compounds with 1DF7, was carried out
by running MM/GBSA energy calculations implemented in Prime module of the Schrödinger
molecular modeling package. Complex structures from last 10 ns (i.e., 1 trajectory frame in each
2ns) of the trajectory frames were selected and binding free energies of corresponding structures
were calculated using VSGB 2.0 solvation model and OPLS3 force field to predict binding free
energies24.
3. Results and discussion:
3.1 Comparative study of binding pocket of Mtb-DHFR and h-DHFR:
This intention of this study was to identify the selective structural features required to inhibit
Mtb-DHFR and to develop novel compounds for selective inhibition of the same, through
computational studies and wet lab techniques. Since DHFR is present in both humans as well as
bacteria, selectivity becomes an important criterion for design and development of the ligands25.
For developing selective inhibitors for Mtb-DHFR, VS protocol was run with two target proteins
Mtb-DHFR (PDB Id: 1DF7) and h-DHFR (PDB Id: 1OHJ). A database of 2,73,951 compounds

(2,65,242 molecule from NCI and 8709 from drug bank) was screened against Mtb-DHFR and
the results obtained were screened against h-DHFR for selectivity.
It is very important to understand the composition of active site of the two target proteins to
explore and use the difference therein for development of selective ligands. The comparative
analysis of the two target proteins Mtb-DHFR (PDB: 1DF7) and h-DHFR (PDB: 1OHJ) was
carried out and difference in the type of amino acid making catalytic triad and other important
amino acids involved in interaction with ligand was noted. The results (Table 2) obtained were
similar to the experimental results presented by Li R. et al12 but what could be inferred from in
silico studies was not possible from the experimental data. As per the X-ray crystallographic data
two interactions are important for Mtb-DHFR inhibition i.e. H-bond interaction with Asp27 and
Arg32 and for h-DHFR inhibition the two important interactions are Val115 andGln35. This was
also observed in cross dockingstudies of the two ligands MTX and COP with the two target
proteins. What caught our attention was that for inhibition of Mtb-DHFR, formation of H-bond
with Asp27 and Arg32 was crucial as it could be seen in docking results of both the ligands
(MTX and COP) and similarly the docking results of MTX and COP with 1OHJ revealed
formation of H-bond with Val115 and Gln35 is crucial for h-DHFR inhibition. This supports the
fact that MTX and COP are non-selective inhibitors as both of them bind to crucial amino acids
in the two proteins. Second observation from the study was π stacking of MTX and COP with
Phe31 and Phe34 in Mtb-DHFR and h-DHFR active site respectively. These results provide very
important insight for the development of selective inhibitors of Mtb-DHFR and that is ligands
must forms H-bond with Asp27 and Arg32 and not with Val115 andGln35 and must shows π-
stacking with Phe31 only.
Table 2: Important interaction of Methotrexate, COP and Hit1 identified with the active site of
Mtb-DHFR and h-DHFR
For the same purpose the binding pocket of the two proteins were critically studied for
size and type of small molecules that could bind in the active site. Studies have revealed
that in addition to the MTX in active site of Mtb-DHFR, a glycerol molecule is also
present whereas in h-DHFRpocket is filled with COP only (Figure 3a and 3b). The

glycerol binding site near DHFR binding site is relatively small in size and may be treated
as extension of the DHFR site. In h-DHFRs, the active site is well packed with three
hydrophobic residue side chains, Leu22, Pro26 and Phe31 and no extension could be
observed suggesting a compounds with side chain which could mimic the binding mode
of glycerol to protein, may bind to Mtb-DHFRselectively as also suggested by Wei H. et
al.26 Such a ligand will be sterically and chemically hindered from forming a complex
with h-DHFR. These hypothesis have been used as basis for present study and for
identifying novel selective hits a virtual screening protocol was run using two databases
viz. NCI and Drug Bank database (2,73,951 compounds) against the two selected targets.
3.2 Virtual Screening and MD Simulation study:
The virtual screening protocols use docking to knock out the hits which are not of interest
i.e. don’t show good interactions with the target. Therefore it is crucial to use correct
docking strategy and validate the procedure before proceeding to VS27. The docking study
on Mtb-DHFR was carried out without bound glycerol in the pocket. The docking
procedure was evaluated by re-docking the co- crystal ligand and reproducing the crystal
orientation of the co-crystal ligand. The methotrexate and COP were extracted, refined
and docked into their respective protein active sites using the similar parameters which
were set for virtual screening and RMSD were calculated between their docked poses and
co-crystal conformations. The RMSD value of 1.2Å for methotrexate and COP docked
pose with their corresponding crystal orientation was obtained indicating the correctness
of the docking procedure. Also the performance of the virtual screening is evaluated by its
ability to pick the active compounds for a particular target present in the database. To
check the same 20 actives were added into the database and docked compounds was
performed. The docking experiment resulted in 20 actives as top scoring ligands among
the top 100 compounds. Moreover, all the low energy conformers of the co-crystalized
ligands were on the top of the docking scores.
In the VS protocol the selected databases (2,73,951) were screened against Mtb-DHFR
(1DF7) using virtual screening protocol (HTVS, SP and XP) of GLIDE module
(Schrödinger). The first screening which was high throughput in nature resulted in 17877

compounds which were subjected to SP screening (1225 compounds) followed by XP
screening which resulted in 122 compounds. From these results top 100 hits (72 from NCI
and 28 from Drug Bank) were selected on the basis of docking score, binding
interactions, visual analysis and removal of duplicates and other molecules reported with
Mtb-DHFR activity. The hits selected were inspected manually to cover maximum
amount of chemical space and screened against h-DHFR (PDB: 1OHJ). To work towards
selectivity, 50 hits were shortlisted after analyzing the binding interaction and visual
inspection but this time the compounds with lowest docking score on 1OHJ and high
score with Mtb-DHFR were selected assuming that these will be selective towards Mtb-
DHFR. For reassuring the results obtained and selectivity consensus docking was
performed. Since different docking software’s use different algorithms, so compounds
which show results in consensus with each other might be potential hits. Therefore,
consensus docking was carried out by using GLIDE and GOLD 5.2.2. Top ten hits were
selected and analyzed visually and for synthetic feasibility. On the basis of synthetic
feasibility hit-1 was selected, synthesized, derivatized and evaluated for biological
activity. To improve upon the chemical space and interaction of the hits, 30 derivatives of
hit 1 were designed and docked on the Mtb-DHFR and h-DHFR. On the basis of docking
ratio and synthetic feasibility 11 compounds were finally synthesized. The structure of
synthesized compounds was confirmed by spectral analysis. All the synthesized
compounds were screened for in vitro antimycobacterial assay on H37RV. From the
literature, we could be observed that most of the compounds reported for inhibitors of
DHFR are either pteridine or pyrimidine based with an exception of one or two. The
advantage of our study was identification of completely new class of Mtb-DHFR
inhibitors which are indole based.
3.3 Classification of Mtb-DHFR catalytic domain:
Hit-1 exhibit four sites (Figure 6) of modifications which could be explored to design
novel selective Mtb-DHFR inhibitors. Here are presented some assumptions based on the
insilico studies of the targets for the selectivity against Mtb-DHFR inhibition.
Figure 6: Basic indole moiety and sites of modification

3.3.1 Site-1 Acidic side chain:
The acidic group of this side chain is crucial for forming salt bridge between two
positively charged arginine residues (Arg32 and Arg60). This salt bridge interaction is
very important as can be seen with methotrexate also. The benzene ring along with the
side chain of hit -1 corresponds to the p-amino benzoic acid part of methotrexate.
Replacement of acidic group with sulphate, sulphone, sulphonamide and phosphate was
found to decrease glide score which may be due to their inability to form the proper salt
bridge. Position of this side chain is also very crucial for the better binding of hit-1 to the
receptor. Substitution at 4 and 5th position decrease binding energy as reflected by
decreased glide score while substitution at 6 and 7th position are favourable for binding.
Optimum binding was observed with substitution at 6th position because the side chain
length is optimum for forming the salt bridge with Arg32 and Arg60.
3.3.2 Site-2 Acid amide group modification:
This part corresponds to the N12 of methotrexate which faces the NADPH. At this site
methotrexate does not form any hydrogen bond whereas hit-1 forms H-bond with Ser49.
This part may help in attaining the selectivity by properly positioning the ligands in the
pocket. Replacing this acid amide group with acetic acid, propionic acid,
propionamideand diketoamide groups reduced the binding scores and also the selectivity
as it failed to form hydrogen bond with Ser49. Thus the optimum group at this position is
ketone group.
3.3.3 Site-3 Modification:
Site 3 is sterically hindered site as the adjacent sites bear large substitutions. Therefore
modification at this site is not favorable. Changing the methyl group with bigger
substituents e.g. ethyl, propyl, butyl or phenyl does not allow acid amide group to form
H-bond with Ser49 and thus reducing the interaction of the compounds with receptor.
3.3.4 Site-4 Modification:

One of the critical differences between Mtb-DHFR and h-DHFR structure is glycerol
binding site present in Mtb-DHFR active site. The docked pose of Mtx in Mtb-
DHFRrevealed that N7 and C8 atom of methotrexate are accessible to solvent and are
facing hydrophilic side chains. At this site an L type of narrow cavity formed by residues
including Phe31, Asp27, Gln28, Ile94, Ile5, Tyr100, Ala7, Trp6, His30, Gly95 NADP and
glycerol. Therefore the opportunity to design selective Mtb-DHFR inhibitors lies at this
site also i.e. to design the molecules which possess site chain characteristics of glycerol
and could bind to the glycerol binding pocket. So the molecules which could
accommodate in this cavity can be selective Mtb-DHFR inhibitors because this cavity is
different in h-DHFR enzyme. The best virtual hit (Hit 1) found to occupy the same pose
and interaction as methotrexate as well as the benzyl part is aligned towards the
negatively charged Asp27 which is catalytic amino acid and responsible for reductase
activity. This interaction is most important for Mtb-DHFR selectivity as glycerol also
forms H-bond with acidic group of Asp27 and is absolutely absent in h-DHFR. So the
functional groups which are forming hydrogen bond with this amino acid will definitely
have the Mtb-DHFRactivity.
Another interesting finding of the docking analysis is that the ligands which could from π-
stacking interaction with Phe31 in Mtb-DHFRcan be selective Mtb-DHFRinhibitors as it
is absent in h-DHFR(π-stacking with Phe34). Phe31 residue forms π-stacking interaction
with pteridine ring of methotrexate while in h-DHFRit forms π-stacking with Phe34 so it
active in both human and mycobacterium. But the Hit 1 ligand forms π-stacking with
Phe31 in Mtb-DHFR and not with Phe34. Thus this finding suggests that the structure
which forms π-stacking with Phe31 in Mtb-DHFR and not with Phe34 in h-DHFR will
provide the selective inhibition towards Mtb-DHFR.
Therefore modification at this site provides further opportunities for developing selective
Mtb-DHFR inhibitors. The modification at site 4 were done in order to optimization the
activity using hydrophobic and hydrophilic substitutions like aliphatic moieties, benzyl
ring with or without substitution etc. In silico studies were carried out to observe the
effect of these groups on the selectivity of these agents towards Mtb-DHFR. It was
observed that hydrophilic groups like hydroxyl and methoxy substitutions were more
selectivity towards Mtb-DHFR compared to hydrophobic substituents like halogen. This

may be due to the fact that in Mtb-DHFR the amino acids facing pteridine (Trp22, Asp27
and Gln28) are hydrophilic in nature while amino acids facing pteridine ring (Leu22,
Pro26 and Phe31) in h-DHFR are hydrophobic (Figure 3a and 3b). Hydrophobic groups
do not form hydrogen bond with Asp27 however they form π-stacking interaction with
Phe31. Most important interactions of indole ligand, methotrexate and COP are
summarized in Table 1
.
Figure 7: (A) Docking pose of Hit 1 against Mtb-DHFRin GOL catalytic domain, (B)
Ligplot of Hit 1 against Mtb-DHFR catalytic domain (C) Binding interaction of IND-
07against Mtb-DHFR, (D) Ligplot of IND-07against Mtb-DHFR, (E) Docking pose of
MTX against Mtb-DHFR, (F) Ligplot of MTX against Mtb-DHFR.
Table 3: Docking score of synthesized compounds against Mtb DHFR and h DHFR, their in
vitro antimycobacterial activity and enzymatic inhibition data
3.4 Chemistry:
As a part of our ongoing research, number of protocols were developed for synthesis of
indole moiety. Intermediate JM-01 and JM-02 were obtained through JMA and JMB
from benzylamine reaction with ethylacetoacte and acetyl acetone respectively.
From the two intermediates JM-01 and JM-02 rest of the compounds were synthesized
by condensation reactions with bromoethylacetate/ethyl-4-bromo butyric ester followed
by either hydrolysis with lithium hydroxide or treatment with hydrazine hydrate28.
The final compounds were purified by crystallization from ethanol and the structures
1
were confirmed by IR, H NMR, Mass and elemental analysis. 1H NMR of all the
compounds showed a singlet corresponding to CH2 proton between δ 5.4-5.6 and CH3
protons could be located around δ 2.5-2.7. The aromatic protons were observed as triplet,
doublet and multiples range δ 6.66 to 7.60. The IR spectra showed characteristic peaks at
3200 to 3400 (NH str.), 1670 to 1700 (C=O str.), 2925 to 3050 (Ar C-H str.), 1570 to
1590 (C=C Str.). The yield of synthesized compounds was between 60-85%.
3.5 Biological Evaluation:

3.5.1 MIC Antimycobacterial assay:
The MIC of the compounds was determined by measuring bacterial growth after 5 days in
the presence of test compounds against M. tuberculosis H37Rv colonies29. For potent
compounds, assays were repeated at lower concentrations. All the 11 compounds showed
significant inhibition against H37Rv cell lines between ranges (25 to 200µM). Out of 11
compounds IND-07 showed highest activity and was selected for in vitro testing against
Mtb-DHFR and h-DHFR.
Figure 8: Dose response curve for IND-07
3.5.2 In vitro Mtb-DHFR and h-DHFR enzyme assay:
The selectivity of the compound towards Mtb-DHFR over human DHFR was done by
performing enzyme assay using h-DHFR. The identified compound was found selective
towards Mtb-DHFR. Most potent derivative (IND-07) inhibited the two enzymes at an
IC50 of 150 µM and 980 µM respectively. The selectivity index was found to be 6.53
(Table 3).
3.6 MD Simulation Studies:
The stability of docked pose at the binding site of Mtb-DHFR (PDBID: 1DF7) was
evaluated by molecular dynamics simulation. The RMSD of protein backbone and heavy
atoms during the entire 10 ns MD simulation are given in Figure 9. The MD simulation
studies revealed that the compound (IND-07) is highly stable within the active pocket of
the enzyme. The studies displayed that the compound interacts profoundly with important
amino acids viz. Arg60, Arg32, Ser49 and Phe31. The comparison of the MD simulation
studies of compound (IND-07) and the reference ligand (MTX) displayed similar
interaction.
Figure 9: (A) Time dependence of the total energy and protein backbone, (B) RMSD
relative to the initial minimized complex of molecule during MD simulations.
The evaluation of MD trajectories reveals that the protein-ligand complex was stable after
initial 1 ns period. Throughout the simulation studies, the molecule IND-07 maintained

its interaction with active site residues of protein. Both RMSD values remained within the
range 1 Å for the system after reaching equilibrium, which further proved the
conformational stabilities of the protein-ligand complex.
3.7 ADMET and MM/GBSA Studies:
The ADME properties of the 11 compounds were analyzed using Qikprop. The result of
Qikprop analysis of most potent compound, IND-07 is presented Table 4. MM/GBSA
studies of IND-07 revealed that its binding free energy is less than that of MTX which is
-84.83 and -23.58 respectively indicating that the compound IND-07 may be stable in
the receptor pocket (Table 4).
Table 4: ADMET analysis and MM/GBSA binding score
4. Conclusion:
By using in silico approach, the glycerol binding site of Mtb-DHFR binding pocket was
exploited successfully to identify and develop novel inhibitors for same. Based on the virtual
screening hit 1 was obtained, and a series of molecules was designed, by using different
substituents at 3, 5 position of indole moiety of hit 1. Compound IND-07 was developed as novel
selective inhibitors of Mtb-DHFR. It contains a novel central core (1-(1-benzyl-5-hydroxy-2-
methyl-1H-indol-3-yl) ethanone), which will significantly expand the chemical space of novel
Mtb-DHFR inhibitors. This compound did show selectivity against Mtb-DHFR and can be used
as a lead compound for further optimization. Further, the present findings will help in designing
novel Mtb-DHFR inhibitors in future to improve the selectivity of compounds against Mtb-
DHFR.
Software used: (a) Schrödinger Suite 2017 Protein Preparation Wizard; LLC, New York, NY,
Glide 7.1 Schrödinger, LLC; New York, NY, 2017, LigPrep, version 4.6, Schrödinger, LLC,
New York, NY, 2017. (b) Molinspiron property calculation program.
Acknowledgements
The authors wish to acknowledge the Department of Biotechnology (DBT), Govt of India, for
providing Bioinformatics infrastructure facility (via letterno.AS/MP (RES.)/JH-5/2013). The

author KC Sharma thanks to C. Karthikeyan for his valuable support for this study. We
gratefully acknowledge the TAACF (Tuberculosis Antimicrobial Acquisition Coordinating
Facility, National Institute of Allergy and Infectious Diseases (Contract No.
HHSN272201100009I), USA for carrying out the biological activity. One of the authors (KC)
also thanks Dr. G S Deora, Monash Institute of Pharmaceutical Sciences, Monash University,
and Melbourne, Australia for his help in MD Simulation studies.
Conflict of interest:
The authors report no declarations of interest.
† Supplementary data: The supporting information includes spectral data details (NMR, IR,
HRMS, HPLC and Mass), MD simulation and molecular modelling study of the compounds.
This material is available free of charge.
References:
1. Argyrou, A.; Vetting, M. W.; Aladegbami, B.; Blanchard, J. S. Nature Structural &
Molecular Biology 2006, 13, 408.
2. Hagner, N.; Joerger, M. Cancer Management and Research 2010, 2, 293.
3. World
Health
Organisation,
Tuberculosis
fact
sheet,
Available
at;
http://www.who.int/mediacentre/factsheets/fs104/en, data accessed: December 2017.
4. Robson, C.; Meek, M. A.; Grunwaldt, J. D.; Lambert, P. A.; Queener, S. F.; Schmidt, D.;
Griffin, R. J. Journal of Medicinal Chemistry 1997, 40, 3040.
5. Russell, P. B.; Hitchings, G. H. Journal of the American Chemical Society 1951, 73, 3763.
6. Hekmat-Nejad, M.; Rathod, P. K. Experimental Parasitology 1997, 87, 222.
7. Yuthavong, Y.; Tarnchompoo, B.; Vilaivan, T.; Chitnumsub, P.; Kamchonwongpaisan, S.;
Charman, S. A.; McLennan, D. N.; White, K. L.; Vivas, L.; Bongard, E.; Thongphanchang,
C.; Taweechai, S.; Vanichtanankul, J.; Rattanajak, R.; Arwon, U.; Fantauzzi, P.;
Yuvaniyama, J.; Charman, W. N.; Matthews, D. Proceedings of the National Academy of
Sciences of the United States of America 2012, 109, 16823.
8. Kobayashi, M.; Kinjo, T.; Koseki, Y.; Bourne, C. R.; Barrow, W. W.; Aoki, S. Journal of
Chemical Information and Modeling 2014, 54, 1242.
9. Lele, A. C.; Raju, A.; Khambete, M. P.; Ray, M. K.; Rajan, M. G. R.; Arkile, M. A.; Jadhav,
N. J.; Sarkar, D.; Degani, M. S. ACS Medicinal Chemistry Letters 2015, 6, 1140.

10. Anderson, A. C. Chemistry & Biology 2003, 10, 787.
11. Cody, V.; Galitsky, N.; Luft, J. R.; Pangborn, W.; Rosowsky, A.; Blakley, R. L. Biochemistry
1997, 36, 13897.
12. Li, R.; Sirawaraporn, R.; Chitnumsub, P.; Sirawaraporn, W.; Wooden, J.; Athappilly, F.;
Turley, S.; Hol, W. G. J. Journal of Molecular Biology 2000, 295, 307.
13. Dias, Marcio Vinicius B.; Tyrakis, P.; Domingues, Romenia R.; Leme, Adriana Franco P.;
Blundell, Tom L. Structure 2014, 22, 94.
14. Hajian, B.; Scocchera, E.; Keshipeddy, S.; G-Dayanandan, N.; Shoen, C.; Krucinska, J.;
Reeve, S.; Cynamon, M.; Anderson, A. C.; Wright, D. L. PLOS One 2016, 11, e0161740.
15. Katsila, T.; Spyroulias, G. A.; Patrinos, G. P.; Matsoukas, M.-T. Computational and
Structural Biotechnology Journal 2016, 14, 177.
16. Jones, G.; Willett, P.; Glen, R. C.; Leach, A. R.; Taylor, R. Journal of Molecular Biology
1997, 267, 727.
17. Friesner, R. A.; Murphy, R. B.; Repasky, M. P.; Frye, L. L.; Greenwood, J. R.; Halgren, T.
A.; Sanschagrin, P. C.; Mainz, D. T. Journal of Medicinal Chemistry 2006, 49, 6177.
18. Schrödinger Release 2017-4: LigPrep, Schrödinger, LLC, New York, NY, 2017.
19. Pawlak, J. M.; Khau, V. V.; Hutchison, D. R.; Martinelli, M. J. The Journal of Organic
Chemistry 1996, 61, 9055.
20. Ollinger, J.; Bailey, M. A.; Moraski, G. C.; Casey, A.; Florio, S.; Alling, T.; Miller, M. J.;
Parish, T. PLOS One 2013, 8, e60531.
21. White, E. L.; Ross, L. J.; Cunningham, A.; Escuyer, V. FEMS Microbiology Letters 2004,
232, 101.
22. Nixon, M. R.; Saionz, K. W.; Koo, M. S.; Szymonifka, M. J.; Jung, H.; Roberts, J. P.;
Nandakumar, M.; Kumar, A.; Liao, R.; Rustad, T.; Sacchettini, J. C.; Rhee, K. Y.;
Freundlich, J. S.; Sherman, D. R. Chemical Biology 2014, 21, 819.
23. Schrödinger Release 2017-4: Desmond Molecular Dynamics System, D. E. Shaw Research,
New York, NY, 2017. Maestro-Desmond Interoperability Tools, Schrödinger, New York,
NY, 2017.
24. Schrödinger Release 2017-4: QikProp, Schrödinger, LLC, New York, NY, 2017.
25. Gond, D. S.; Meshram, R. J.; Jadhav, S. G.; Wadhwa, G.; Gacche, R. N. Drug Invention
Today 2013, 5, 182.

26. Hong, W.; Wang, Y.; Chang, Z.; Yang, Y.; Pu, J.; Sun, T.; Kaur, S.; Sacchettini, J. C.; Jung,
H.; Lin Wong, W.; Fah Yap, L.; Fong Ngeow, Y.; Paterson, I. C.; Wang, H. Scientific
Reports 2015, 5, 15328.
27. Lionta, E.; Spyrou, G.; Vassilatis, D. K.; Cournia, Z. Current Topics in Medicinal Chemistry
2014, 14, 1923.
28. Littell, R.; Morton, G. O.; Allen, G. R. Journal of the American Chemical Society 1970, 92,
3740.
29. Zelmer, A.; Carroll, P.; Andreu, N.; Hagens, K.; Mahlo, J.; Redinger, N.; Robertson, B. D.;
Wiles, S.; Ward, T. H.; Parish, T.; Ripoll, J.; Bancroft, G. J.; Schaible, U. E. The Journal of
Antimicrobial Chemotherapy 2012, 67, 1948.

HO
O
O
NH2
N
NH2
N
O
NH2
N
N
O
O
O
N
O
N
N
N
OH
N
N
H2N
N
H2N
H2N
O
Br-WR 99210 Br
Trimethoprim
Methotrexate
Figure 1: Structure of some reported DHFR inhibitors