Content uploaded by Rutwick Surya
Author content
All content in this area was uploaded by Rutwick Surya on May 25, 2022
Content may be subject to copyright.
Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=tbsd20
Journal of Biomolecular Structure and Dynamics
ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/tbsd20
In-silico validation of novel therapeutic activities
of withaferin a using molecular docking and
dynamics studies
Rutwick Surya Ulhas & Alok Malaviya
To cite this article: Rutwick Surya Ulhas & Alok Malaviya (2022): In-silico validation of novel
therapeutic activities of withaferin a using molecular docking and dynamics studies, Journal of
Biomolecular Structure and Dynamics, DOI: 10.1080/07391102.2022.2078410
To link to this article: https://doi.org/10.1080/07391102.2022.2078410
View supplementary material
Published online: 24 May 2022.
Submit your article to this journal
View related articles
View Crossmark data
In-silico validation of novel therapeutic activities of withaferin a using molecular
docking and dynamics studies
Rutwick Surya Ulhas
a,b
and Alok Malaviya
a
a
Applied and Industrial Biotechnology Laboratory, Department of Life Sciences, CHRIST (Deemed-to-Be University), Bangalore, Karnataka,
India;
b
Faculty of life sciences, University of Jena, (Friedrich-Schiller-Universit€
at Jena), Jena, Germany
Communicated by Ramaswamy H. Sarma
ABSTRACT
Withaferin A is a bioactive molecule of W. somnifera. We access its efficacy against various target pro-
teins associated with Cancer, Type-II Diabetes and hypercholesterolemia using molecular docking.
Although it’s efficacy against some of these targets have been reported earlier, we validate each
mechanism in order to report the most appropriate mechanism of action. We explain the anti-cancer
activity of Withaferin A by inhibition of Mortalin (mtHsp70) and Nrf2 protein with binding energies
8.85 kcal/mol and 12.59 kcal/mol respectively. Similarly, the anti-diabetic activity could be explained
by inhibition of alpha and beta-glucosidase with binding energies 6.44 and 4.43 kcal/mol respect-
ively and the cholesterol reduction could be explained by its ability to inhibition of NPC1 and SRB1
with binding energies 5.73 and 7.16 kcal/mol respectively. The molecular dynamics of the apopro-
tein and the protein-ligand complex simulated for the best targets of each activity namely Nrf2 protein
for anti-cancer, a-glucosidase for anti-diabetic and SR-B1 for anti-hypercholesterolemia activity indi-
cated the formation of stable complexes due to low RMSD deviations, low RMSF fluctuations and low
RG values after the docking simulation. Finally, an ADME þT (Adsorption, distribution, metabolism,
excretion and toxicity) prediction on Withaferin A showed that it obeyed all the Lipinsky’s rules and
qualified the drug-like criteria. All these results validate that Withaferin A possess potential anti-cancer,
anti-diabetic and cholesterol reducing properties. This is the first report that indicates the possibility of
Withaferin A binding and inhibiting SR-B1 as a mechanism of its anti-hypercholesterolemia activity.
HIGHLIGHTS
This paper validates various mechanisms proposed for the anti-cancer, anti-diabetic and anti-hyper-
cholesterolemic properties of Withaferin A obtained from Withania somnifera L. Dunal
The 3D structure of Withaferin A was docked to the suitable target proteins using Molecular dock-
ing server (AutoDock algorithm) and the binding energies were assessed. Molecular dynamics simu-
lations of the protein ligand complex were also studied and found to indicated stable binding
Additionally, the ligand was accessed to be drug-like in terms of its ADME þT predictions
Inhibition of SR-B1 and NPC-1 as potential anti-hypercholesterolemic mechanism of Withaferin A
was reported for the first time in this paper
ARTICLE HISTORY
Received 15 October 2021
Accepted 11 May 2022
KEYWORDS
Anti-cancer; anti-diabetic;
cholesterol reduction;
docking; Withaferin A;
W. somnifera
1. Introdution
In the recent decade, Phyto therapeutics have gained much
attention. Over 60% of the world population have
undergone phytotherapy at least once in their lifetimes
(Maurya, 2010; Saleem et al., 2020). Plants exhibit a vast
molecular diversity showing immense potential to produce
CONTACT Alok Kumar Malaviya alokkumar.malaviya@christuniversity.in Applied and Industrial Biotechnology Laboratory, Department of Life Sciences,
CHRIST (Deemed-to-Be University), Hosur road, Bangalore, Karnataka, India
Supplemental data for this article can be accessed online at http://dx.doi.org/10.1080/07391102.2022.2078410.
ß2022 Informa UK Limited, trading as Taylor & Francis Group
JOURNAL OF BIOMOLECULAR STRUCTURE AND DYNAMICS
https://doi.org/10.1080/07391102.2022.2078410
compounds of medicinal importance. Most of the medicinal
compounds are either alkaloids or steroids of various forms
and obtained from plant organs such as roots, leaves and
stems (Rai et al., 2016). Due to the vast variety of available
compounds, it has become difficult to perform validatory
chemical assays to assess their medicinal properties.
Furthermore, it becomes more difficult to understand and
elucidate their mechanism of action. In this paper we per-
form molecular docking to understand the interaction of one
such compound, namely Withaferin A with some of the
reported target proteins known to play a critical role in med-
ical conditions such as type II diabetes, high blood choles-
terol and cancer.
The plant Withania somnifera L. Dunal (derived from Latin;
‘Sleep inducing’indicates its sedative properties) of the
nightshade family is commonly known by the name
Ashwagandha (derived from Sanskrit; on account of its char-
acteristic ‘horse-odor’) or Indian ginseng (the name indicates
its similarity with the Chinese plant Panax ginseng in terms
of usage) (Bhattacharya & Muruganandam, 2003; Rai et al.,
2016; Surya et al., 2021). It has been reported to possess vari-
ous therapeutic activities such as anti-cancer, anti-diabetes,
anti-malaria, anti-ulcerogenic, anti-Parkinson’s, anti-inflamma-
tory, anti-microbial, anti-fatigue, anti-stress, revenant, free
radical scavenger, immunomodulatory, analgesic, hepatopro-
tective, neurogenerative, cardio-protective, anti-oxidant, anti-
coagulant, anti-depressant, chondroprotective and a detoxify-
ing agent among many others (Maurya, 2010; Saleem et al.,
2020; Sharma et al., 2008). The major class of active com-
pounds derived from this plant are C-28 steroidal lactones
called Withanolides characterized by delta lactones formed
due to C-22 and C-26 oxidations on its ergostane skeleton
(Matsuda et al., 2001).
Withaferin A (Figure 1) is one of the most widely studied
compounds of this class with various therapeutic properties.
However, the mechanisms of action for most of these prop-
erties are not fully understood. Recent studies have primarily
worked on animal models and cell lines to understand these
effects; however any clue on a definite molecular mechanism
is still missing (Sari et al., 2020).
Cancer is a major cause of mortality worldwide with over
10 million deaths in 2020 alone (Jemal et al., 1999). It is sci-
entifically classified as a group of heterogeneous hyper pro-
liferative disorders which causes tumorigenesis due to
altered cellular signaling pathways leading to uncontrolled
cell division and inhibition of apoptosis (Sever & Brugge,
2015). The occurrence of cancer is triggered by hereditary
factors and internal or external causes (Hanahan & Weinberg,
2011). Cancer is characterized by distinct hallmarks such as
transformation, proliferation, inhibition of apoptosis, replica-
tive immortality, invasion and metastasis, angiogenesis and
inflammations that promote tumor(Aggarwal et al., 2006;
Malaviya et al., 2021). The cancers that are accounted for
maximum mortality are colorectal, stomach, breast, lung and
prostate cancer (Rai et al., 2016). Cancer treatment and thera-
peutics are often targeted towards interference and disrup-
tion of cell division and establish homeostasis of its
hallmarks (Malaviya et al., 2021; Szarc Vel Szic et al., 2014).
The therapeutic mechanisms include p53 signaling, GM-CSF
signaling, apoptosis and death receptor signaling as well as
G2-M phase cell cycle arrest by DNA damage (Surya et al.,
2021; Widodo et al., 2008). These are carried out by three
major approaches - (i) surgical removal of tumor, (ii) chemo-
therapy and (iii) radiation therapy. However, these conven-
tional treatment plans have several drawbacks, while surgical
and radiation therapies are invasive and cause unintended
tissue damage, conventional chemotherapeutic drugs gener-
ate highly toxic and reactive metabolites in the liver
(Kashyap et al., 2021; Sharma et al., 2008). Targeted treat-
ment plans on the other hand are effective for benign
tumors but not for metastatic tumors. Therefore, there is a
high demand for novel Phyto-therapeutics and integrated
medicine approaches to treat cancer. Various plants are
established sources of anti-cancer drugs. Particularly, vin-
blastine, vincristine, and podophyllotoxins are obtained from
Catharanthus roseus, etoposide and teniposide are extracted
from Podophyllum sps, while paclitaxel is isolated from Taxux
brevifolia. Many other medicinal plants have traditionally
been reported to possess anti-cancer activity with not
enough evidence, studies or practical use in the field of
medicine. In this paper, we attempt to validate the mechan-
ism of action for the anti-cancer activity of W. somnifera.
Diabetes is another highly prevalent, heterogeneous,
severe, chronic and multifactorial disease diagnosed in about
9.3% of the world population in 2019 (National Diabetes
Report, 2020). The occurrence of diabetes is exponentially
increasing in the present times, from 108 million cases in
1980 to over 422 million cases 2014. The occurrence of Type
II Diabetes is also alarmingly high among children and ado-
lescents. The previously understood risk factors for Type II
diabetes are categorized mainly into genetic, behavioral and
lifestyle factors. However, these are now challenged by epi-
demiological and molecular studies which have brought sev-
eral epigenetic mechanisms and diseases into attention that
are either acquired due to diabetes itself as a risk factor or
promote diabetic occurrence(Zimmet et al., 2014). Even can-
cer has also been reported to be promoted by obesity,
hyperglycemia and oxidative stress which are common hall-
marks of type II diabetes (Vigneri et al., 2009). The relation-
ship between the two diseases results in a positive feedback
loop where the cancer therapeutics cause or up regulate dia-
betes. This relationship exists because insulin is a growth
Figure 1. Molecular structure of Withaferin A.
2 R. SURYA ULHAS AND A. MALAVIYA
factor with metabolic and mitogenic influence on malignant
cells at both receptor and post receptor level(Ferguson et al.,
2013). The reverse relationship of cancer promoting diabetes
is primarily due to the common substrates shared by IGF-1
and Insulin receptor which makes most of the chemothera-
peutic anti-cancer drugs diabetogenic. It is postulated that
molecules preventing this positive loop could effectively be
used for simultaneously addressing both these conditions
High cholesterol level is a direct or indirect cause of over
4.5% annual global mortality with about one third of all
ischemic heart disease attributed to high cholesterol levels
(Carroll & Lacher, 2010; Jeong et al., 2018). More importantly,
malignant transformation disrupts the cholesterol feedback
inhibition mechanism as well in cancer patients. This results
in increased concentrations of cholesterol and its precursors
in cancer cells. In fact, it has been hypothesized that choles-
terol reduction is an approach to prevent the growth of
tumor cells, and therefore effective adjuvant to cancer ther-
apy and prevention of carcinogenesis (Ferguson et al., 2013).
Obesity is now a worldwide epidemic that strongly promotes
and gets promoted by type II diabetes. High cholesterol con-
centration causes insulin resistance in cells which leads to
dyslipidemia and hyperglycemia that together causes type II
diabetes. As stated previously, hyperinsulinemia and dyslipi-
demia contribute to carcinogenesis and therefore, here we
are trying to find out a single Phyto-therapeutic candidate
which has the anti-cancer, anti-diabetic and hypo-cholestero-
lemic properties. While W. somnifera is has been reported to
reduce cholesterol levels, no validated molecular mechanisms
are elucidated for this property. Here, we have attempted to
answer this dire need by molecular docking and molecular
dynamic simulations.
In this work we have attempted to elucidate and validate
the mechanism of action for reported anti-cancer, anti-dia-
betic and hypocholesterolemic properties of Withaferin A
using a molecular docking approach (Rutwick Surya &
Praveen, 2021, Halder et al., 2022). Additionally, we have
tried to co-relate its indirect anti-cancer property through
the related anti-diabetic and anti-hypocholestromic proper-
ties, so that it could be taken forward as potential herbal
candidate for treatment of mentioned conditions.
2. Materials and methods
2.1. Obtaining ligand spatial data
The spatial co-ordinates of the constituent atoms of
Withaferin A were obtained as its molecular structure from
PubChem database (https://pubchem.ncbi.nlm.nih.gov/). The
structure was obtained as a spatial data file with
.SDF extension.
2.2. Conversion of ligand data to PDB format
The online tool by National cancer institute for structure file
conversion was used to convert the spatial data file .SDF for-
mat of the ligands to Protein data bank .PDB format (https://
cactus.nci.nih.gov/translate/). The 3 D structure of the
molecule was requested in kekule format (however, the lig-
and is non aromatic and would yield the same co-ordinates
for either aromatic or kekule PDB output).
2.3. Obtaining protein structure
a-glucosidase and bglucosidase were selected as potential
target proteins for the anti-diabetic activity, Mortalin (mtHsp
9) and Nrf2 protein for the anti-cancer activity and
Scavenger receptor BI (SR-BI) and NPC1 like intracellular chol-
esterol transporter 1 (NPC1L1) for anti- Hypercholesterolemic
activity. The 3-dimensional structural coordinates of these
proteins were obtained from the protein data bank in .pdb
format. Crystal structure of human lysosomal acid-a-glucosi-
dase (PDB ID-5NN3) (Roig-Zamboni et al., 2017) and Crystal
structure of human cytosolic b-glucosidase (PDB ID-2JFE)
(Tribolo et al., 2007) were used to test the anti-diabetic activ-
ity of Withaferin A, Similarly, Substrate binding domain of
the human Heat Shock 70 kDa protein 9 (Mortalin) (PDB ID-
3N8E), whose structure is not yet published in the literature,
and Crystal structure of the Kelch-Neh2 complex (PDB ID-
2LFU) (Lo et al., 2006) were used to test the anticancer activ-
ity, and C-terminal transmembrane domain of scavenger
receptor BI (SR-BI)(PDB ID-5KTF) (Chadwick et al., 2017) and
NPC1-like intracellular cholesterol transporter 1 (NPC1L1)
(PDB ID- 6V3F) (Huang et al., 2020) were used to test the
cholesterol reducing property of Withaferin A.
2.3. Preparation of target proteins and ligands
The target proteins were uploaded in the protein library and
all the mentioned ligands were uploaded in the ligand
library of the molecular docking server. In the initial cleaning
steps, pH was set to 7 and other parameters were left to
their default values. Upon successful cleaning, docking was
initiated for the ligand with the 6 target proteins.
Preparation of the protein molecule involved addition of
essential polar hydrogen atoms as detected by the AutoDock
tool and uniting the solvation parameters atom type charges
by Gasteiger (Morris et al., 1998).
2.4. Molecular docking
The molecular docking server was used to calculate the bind-
ing energies and other docking parameters. This server calls
for the AutoDock workflow (Bikadi & Hazai, 2009). Energy
minimization of Withaferin A was done using the MMFF94
force field (Halgren, 1996) and Gasteiger partial charges were
assigned within the docking server. Elimination of the non-
polar hydrogen bonds was carried out followed by defining
of the rotatable bonds as auto-detected by the tool. The par-
tial charges of the atoms of the ligand were added to the
PDBQT file.
The Auto grid program generated affinity (grid) maps of
20 20 20 Å grid points with a 0.375 Å spacing (Morris
et al., 1998). The AutoDock parameter set- and distance-
dependent dielectric functions were used to calculate the
Van der Waals and the electrostatic terms.
JOURNAL OF BIOMOLECULAR STRUCTURE AND DYNAMICS 3
Lamarckian genetic algorithm (LGA) and the Solis & Wets
local search method were used to generate docking results
(Solis & Wets, 1981). Initial position, torsions and orientation
of the ligand molecules were set randomly. Ten different
runs configured to terminate after a maximum of 250,000
number of energy evaluations were used to derive the
results of the docking experiment. The population size was
configured to 150. During the search, torsion and quaternion
steps of 5 and a translational step of 0.2 Å were applied
respectively.
The commands used to run the docking (mentioned in
parenthesis) are summarized as 0.2 Å torsion step (tstep),
5.0rigid body orientation step (qstep), 5.0dihedral angles
step (dstep), 2.0 Å root mean square deviation tolerance
(rmstol), 150 individuals in the population with each individ-
ual formed by a coupling of the genotype and its associated
phenotype (ga_pop_size), 250000 and 540000 limits for the
docking program as the number of energy evaluations and
number of generations counted by the AutoDock as the
docking run progresses such that the run terminates when
either of the limits is reached (ga_num_evals and ga_num_-
generations) and 10 runs executed in total as the command
involves the new hybrid Lamarckian genetic algorithm search
engine (ga_run).
The Ki values were obtained from the .dlg file generated
by Autodock. These values are calculated from the function
Ki ¼e
(DG/(RT)
. However, these are theoretical Ki values that
only serve an indicative purpose and needn’t always coincide
with the experimental values. Ki values are more specific and
appropriate quantification of the enzyme-ligand complex
and were thus chosen over Kd value calculation.
2.5. Molecular dynamics simulations
Molecular dynamics simulations were performed on the tar-
gets showing best binding energies for each of activities
namely the Nrf2 transcription factor for Anti-cancer, a-gluco-
sidase for Anti-diabetic and SR-B1 for Anti-hypercholesterole-
mic respectively using Gromacs-2019.4 (Abraham et al.,
2015). The selected ligand topology was downloaded from
the PRODRG sever to obtain the force field parameters
(Gromos54a7.ff with spc216) to simulate the molecular
dynamics of ligand (Sch€
uttelkopf & Van Aalten, 2004). The
docking poses obtained from the previous step was used to
simulate the molecular dynamics of the ligand-protein com-
plex and the apo-protein. The energy of the system in vac-
uum was minimized using a steepest descent algorithm for
1500 steps with stable RMSD values as convergence criteria.
All bonds were constrained, and electrostatic interactions
were treated as Particle Mesh Ewald for long-range electro-
statics with a particle cut-off value of 4. In a cubic periodic
box of 0.5 nm, the complex structures were solvated using a
simple point charge (SPC) water model. The temperature was
set to 310 K and pressure to 1.01 bar, the values were fixed
with the Verlet algorithms. The complex system was main-
tained at a salt concentration of 0.15 M by adding sufficient
numbers of Na þand Cl- counter ions (35000 water mole-
cules with 40 Na þand 24 Cl- for a-glucosidase; 32000 water
molecules with 37 Na þand 24 Cl- for SR-B1; 29000 water
molecules with 32 Na þand 19 Cl- for Nrf2 respectively). After
the NPT equilibration phase for a time set value of 100 ps
(100 ps each for the 5 position restrains of 1000, 100, 10, 1
and 0), a final simulation run of 100 ns was conducted in the
ensemble (Gangadharappa et al., 2020). The GROMACS simu-
lation package was used to analyze the Root mean square
deviation (RMSD), Root mean square fluctuation (RMSF),
Radius of gyration (RG), Secondary structure element analysis
(SSE), Ligand-RMSD and Hydrogen bonds (Prasanth
et al., 2021).
In order to estimate the binding free energy variation (DG
binding) of an inhibitor with protein over simulation time,
Molecular Mechanics Poisson-Boltzmann Surface Area (MM-
PBSA) was used. The binding free energy was estimated
using the GROMACS utility g_mmpbsa. Our results were
obtained by computing DG in the last 20 ns within 1000
frames (Kumari et al., 2014).
2.6. ADME and druglikeness prediction
pkCSM pharmacokinetics tool (http://biosig.unimelb.edu.au/
pkcsm/) was used to predict the adsorption, distribution,
metabolism, exertion and toxicity (ADME þT) characteristics
of Withaferin A. The ligand file format was .SDF through
which the SMILES of Withaferin A were extracted and used
as the input for the ADME þT prediction.
The molecule was screened for adherence to Lipinski’s
rule, ensuring no more than 5 hydrogen bond donors and
10 hydrogen bond acceptors, and provided the molecular
mass doesn’t exceed 500 Da and octanol–water partition
coefficient (log P) is less than 5. These parameters were used
to determine the druglikeness of Withaferin A.
3. Results and discussion
The ligand molecule namely Withaferin A has been reported as
the potential ant-cancer, anti-diabetic and cholesterol-reducing
active compound of the plant Withania somnifera in the literature
(Saleem et al., 2020;Suryaetal.,2021). The potential therapeutic
activity of Withaferin A as a cholesterol reducing-, anti-diabetic-
and anti-cancer agent was assessed by docking Withaferin A to
various receptors and enzymes such as NPC1 and SRB1 for evalu-
ating cholesterol reducing potential, alpha and b-glucosidase for
anti-diabetic potential, and Nrf2 and Mortalin for anticancer activ-
ity assessment. Furthermore, the low binding energy (high mag-
nitude negative values) of these docking simulations in kcal/mol
is a strong indicator that these bindings were spontaneous.
Therefore, docking feasibility was quantified in terms of the bind-
ing energy. It may be implied that the docking of the ligand to
the target protein results in inhibition of the target protein and
this inhibitory potential was quantified in terms of the estimated
inhibitory constant value which is the concentration of the ligand
(withaferin A) needed to cause 50% inhibition of the target pro-
tein. Other parameters such as the individual energy contribu-
tions (Van der Waals þhydrogen bond þdesolvation energy
(vdW þHbond þdesolv) and the electrostatic forces of attraction
respectively) were individually obtained. The interatomic contact
4 R. SURYA ULHAS AND A. MALAVIYA
frequency, was calculated based on a knowledge-based scoring
function that assumes a more favorable interaction would occur
at a higher frequency (Brooijmans & Kuntz, 2003;Guedesetal.,
2014). All the docking was visualized to highlight the interacting
amino acids (with their respective binding energies for specific
atoms of Withaferin A). The H bond plot was obtained to identify
the location of docking in terms of the secondary structure of
the target protein. The lowest binding energies of all the docking
along with their inferred comments is mentioned in Table 1.The
energy contributions of the best hits in all the dockings per-
formed are described in Table 2.
3.1. Visualization of all the docking simulations
The protein level interactions namely docking visualization
and geometric representation of all the target proteins are
depicted in Figure 2 and their respective amino-acid level
interaction namely H-bond plots and 2 D interaction plots
have been presented in Figure 3 respectively. The geometric
representation of the protein-ligand interaction at the struc-
tural level involves the interacting peptide represented as a
cartoon and the ligand molecule as a ball and stick structure.
Moreover, the structure of Withaferin A is represented in ball
and stick form with all the carbon-amino acid interaction dis-
tinctly numbered and labeled. The visualization is obtained
at the most optimal viewing angle with maximum visibility
of the interaction. The amino-acid level visualization of the
interactions is studied where the 2 D interaction plots repre-
sent the individual amino acid-carbon interaction with
respect to the Withaferin A molecule placed centrally. The H-
bond plot is a scatter plot with the binding locations repre-
sented by red dots.
3.2. Docking simulations pertaining to anti-
cancer property
The anti-cancer activity was assessed by docking
Withaferin A with the target proteins Nrf2 and Mortalin.
Mortalin is a protein belonging to the Hsp70 family with
numerous reported mechanisms for carcinogenesis such as
p53 inactivation, activation of EMT signaling and
Apoptosis dysregulation. Studies involving upregulation of
Mortalin have reported overexpression of cancer cell stem-
ness markers including ABCG2, OCT-4, CD9, ALDH1, MRP1
and CD133. Various clinical studies have also reported that
Mortalin is unresponsive to a large number of chemother-
apeutic drugs and its inhibition led to sensitization of can-
cer cells to the drugs previously reported unresponsive.
Therefore inhibition of Mortalin has been proposed as a
potential mechanism of cancer therapy, which would tar-
get cancer stem cells and consequently prevent further
malignancy (Deocaris et al., 2012;Yangetal.,2013;Yun
et al., 2017). Nrf2 on the other hand is a transcription fac-
tor involved in antioxidant response to oxidative stress. It
activates the antioxidant response element (ARE) which in
turn induces the phase II detoxication genes. In normal
cells, it is bound to the Kelch domain of the Keap1 protein
and thus results in a low basal expression of the antioxi-
dant genes. However, in various cancer cells the hallmark
MAPKinasesignalingcausesreleaseofNrf2fromKeap1
due to its phosphorylation. Nrf2 then translocated to the
Table 1. Summary of the docking results and inferences.
Target therapeutic Target protein Binding energy (kcal/mol) Comments
Cholesterol reduction NPC1 5.72 Withaferin A can bind to and potentially inhibit NPC1 which disrupts the
cholesterol uptake by the enterocytes and lead to its excretion. However,
the binding energy is relatively less favorable than that of SRB1
SRB1 7.16 Withaferin A can bind to and hence inhibit SRB1 more efficiently than NPC1.
Inhibition of SRB1 disrupts influx of HDL and lead to its excretion
Anti-diabetic aGlucosidase 6.44 Withaferin A docks with a-glucosidase spontaneously and hence potentially
inhibits it. This would lead to disruption of intestinal glucose metabolism
and improve postprandial hyperglycemia. However, the binding energy for
this docking is relatively less favorable than that of bGlucosidase.
bGlucosidase 4.43 Withaferin A docks with a-glucosidase more efficiently than b-glucosidase.
The mechanism of glucose reduction is similar to that of a-Glucosidase
Anti-cancer Mortalin (mtHsp70) 8.85 Withaferin A docks spontaneously with Mortalin at very low doses and
inhibits it. This will potentially downregulate the cancer cell stemness
markers, prevent further malignancy and enhance the responsiveness of
other chemotherapeutic drugs
Nrf2 12.59 Withaferin A shows the most effective binding energy while docking to Nrf2.
Inhibition of Nrf2 leads to induction of apoptosis by downregulation of the
cytoprotective genes which are overexpressed in cancer cells
Table 2. Best binding energies of all the molecular dockings conducted.
Target protein
Est. Free Energy
of Binding
Est. Inhibition
Constant Ki
vdW þHbond þdesolv
Energy
Electrostatic
Energy
Total
Intermolecular
Energy Frequency
Interact.
Surface
Mortalin (mtHsp70) 8.85 kcal/mol 328.30 nM 9.26 kcal/mol þ0.12 kcal/mol 9.14 kcal/mol 50% 800.37
Nrf2 12.59 kcal/mol –12.46 kcal/mol 0.07 kcal/mol 12.53 kcal/mol 20% 951.475
SRB1 7.16 kcal/mol –7.30 kcal/mol þ0.00 kcal/mol 7.29 kcal/mol 30% 669.837
NPC1 5.72 kcal/mol –5.45 kcal/mol 0.19 kcal/mol 5.64 kcal/mol 30% 615.946
a-glucosidase 8.07 kcal/mol 1.21 mM3.65 kcal/mol 0.16 kcal/mol 3.81 kcal/mol 50% 541.832
b-glucosidase 6.92 kcal/mol 8.41 mM3.99 kcal/mol 0.02 kcal/mol 4.00 kcal/mol 50% 434.4
JOURNAL OF BIOMOLECULAR STRUCTURE AND DYNAMICS 5
nucleus and expresses several anti-oxidant and cytoprotec-
tive genes that bypasses the cancer cell from apoptosis
due to mitochondrial dysfunction induced oxidative stress
(Angelica & Fong, 2008;Nguyenetal.,2009;Vaishnavi
et al., 2012). Therefore, inhibition of Mortalin as well as
Nrf2 mediated by Withaferin A could be a potential non-
specific anti-cancer mechanism.
The lowest binding energy for Withaferin A bound to Nrf2
was 12.59 kcal/mol with an inhibition constant of just
587.19 pM and that for Mortalin was found to be 8.85 kcal/
mol with inhibition constant of only 328.30 nM. The dosages
to inhibit 50% of these targets were in the range of nano
and even picomolar with favorable and spontaneous ability
to bind. In fact, of all the targets considered for docking, the
Figure 2. (From left to right) Protein structural level visualization of Withaferin A docked to Nrf2, Mortalin (mtHsp70), SRB1, NPC1, a-glucosidase and b-glucosidase
respectively. The docking poses at the top panel represents the interacting domains of the target protein colored according to the secondary structures with the
yellow ribbons representing beta sheets and pink ribbons representing alpha helices. The interacting ligand is represented as cylindrical molecular conformations.
The bottom panel illustrates the entire protein as a cylinder with distinct colors for each domain, the ligand binding is represented as balls with distinct colors for
each atom.
Figure 3. (From left to right) Amino acid level visualization of Withaferin A docked to Nrf2, Mortalin (mtHsp70), SRB1, NPC1, a-glucosidase and b-glucosidase
respectively. The top panel shows a 2-dimentional map of the ligand molecule interacting with specific amino acids of the protein. The amino acids are represented
as labelled green arcs pointing towards interacting atoms of the ligand. The legend within the figure indicates that the lines connecting the atoms within the lig-
and are colored distinctly with purple indicating a covalent bond, red indicating a non-covalent bond and green dotted line indicating a hydrogen bond (with the
hydrogen bond accepting atom colored red while all other atoms colored black). The bottom panel shows the hydrogen bond plots for all the interactions. In these
dot plots, the distances between every possible amino acid residue pair are represented on a 2-dimensional plane, and those amino acids that interact with the lig-
and are represented as red squares while the others remain as black squares. The position and clustering of these squares also indicates their interaction with the
neighbouring amino acids and thus the secondary structure of the protein domain that contains the interacting residue.
Table 3. Binding energies as predicted by MM/PBSA estimation during the MD simulations.
Target Protein Van der Waals energy Electrostatic energy Polar solvation energy Binding energy
Nrf2 222.293 þ/9.916 kJ/mol 17.072 þ/7.362 kJ/mol 117.060 þ/15.178 kJ/mol 141.591 þ/17.521 kJ/mol
SR-B1 107.459 þ/6.851 kJ/mol 6.418 þ/5.390 kJ/mol 39.865 þ/9.181 kJ/mol 85.755 þ/8.732 kJ/mol
a-Glucosidase 116.242 þ/17.511 kJ/mol 7.075 þ/11.569 kJ/mol 54.429 þ/16.763 kJ/mol 81.418 þ/15.594 kJ/mol
6 R. SURYA ULHAS AND A. MALAVIYA
best score was seen for Nrf2 and Mortalin. This indicates the
strong potential of Withaferin A as a potential therapeutic
agent against cancer stem cells, mediated by inhibition of
Mortalin and Nrf2.Tables 1 and 2of the supplementary
material contains detailed energies (of those of the electro-
static and non-electrostatic interactions) of all the possible
docking of Withaferin A with these two targets.
The total intermolecular energy of the docking between
Withaferin A and Nrf2 was found to be 12.53 kcal/mol with
only 0.07 kcal/mol being electrostatic energy and the
remaining 12.46 kcal/mol accounted for vdW þHbond þ
devolv Energy. The amino acids directly interacting with
Withaferin A along with their interaction energies were
ILE559 (-1.7415), VAL420 (-0.8171), VAL606 (-0.6635), VAL467
(-0.4902), CYS368 (-0.5841), THR560 (-0.1606), CYS513
(-0.1842). On the other hand, total intermolecular energy in
case of Mortalin was 9.14 kcal/mol distributed as
9.26 kcal/mol of vdW þHbond þdesolv energy and a posi-
tive electrostatic energy of þ0.12 kcal/mol. The amino acids
directly interacting with Withaferin A along with their
respective interaction energies in kcal/mol were GLU448
(-1.1391) LEU450 (-2.7415) GLN479 (-0.7709) THR449 (1.9489)
PHE472 (-1.7215) GLU483 (-0.6323) ALA475 (-0.832). These
interactions are mentioned along with the interacting atoms
of Withaferin A in Supplementary Tables 3–6.
3.3. Docking simulations pertaining to anti-
diabetic property
The anti-diabetic activity was assessed by docking Withaferin
A with the target proteins alpha-Glucosidase and beta-
Glucosidase. Glucosidases are enzymes that cleave glycosidic
linkage as a part of carbohydrate digestion. They show
selectivity based on number of monosaccharides, stereo-
chemical configuration (of the hydroxyl group) and the
cleavage site position. Alpha- and beta-Glucosidase act spe-
cifically on the terminal glucose (Andrade et al., 2015;
Hakamata et al., 2009).Inhibition of these enzymes has shown
a downregulation of intestinal glucose absorption and
reduced postprandial blood glucose level which is a poten-
tial mechanism of therapeutic for type II diabetes(Derosa &
Maffioli, 2012; Ghani, 2015; Panwar et al., 2014). Glucosidase
inhibitors were approved as anti-diabetic drugs in 1990s.
Withaferin A displayed the ability to spontaneously bind and
inhibit these enzymes and therefore, we are proposing here
that it could be used as a potential therapeutic agent for the
treatment of Type II Diabetes.
The lowest binding energy for Withaferin A bound to
alpha-Glucosidase was found to be 6.44 kcal/mol with an
inhibition constant of 19 mM and that for beta-Glucosidase
was found to be 4.43 kcal/mol with an inhibition constant
of 567.07 mM which indicates that it strongly inhibits alpha-
Glucosidase at much smaller dosage than beta-Glucosidase.
Supplementary Tables 7 and 8 gives a detailed account on
the binding energies and decomposed energies of all the
possible docking of Withaferin A for these 2 target proteins.
The total intermolecular energy of the docking between
Withaferin A and a-glucosidase was found to be 6.40 kcal/
mol with the electrostatic energy being 0.13 kcal/mol and
the rest 6.27 kcal/mol were found to be
vdW þHbond þdesolv energy. The amino acids directly
interacting with Withaferin A along with their respective
interaction energies in kcal/mol were SER940
(-0.0414),LYS903 (-0.5033),GLN900 (-0.3607), ASN925 (-0.2382),
SER924 (0.2189), GLN902 (0.6033). Similarly, the total inter-
molecular energy in case of beta-Glucosidase was 4.13 kcal/
mol distributed as 0.07 kcal/mol of electrostatic energy and
4.07 kcal/mol of vdW þHbond þdesolv energy. Only a sin-
gle amino acid namely ASN46 interacted directly with
Withaferin A showing interaction energy of 0.8015 kcal/mol.
It is interesting to note that while most of the interaction
energy in alpha-Glucosidase was physical forces such as Van
der Waals and Hydrogen bonding, beta-Glucosidase majorly
interacted with it in terms of electrostatic forces. These inter-
actions are mentioned along with the interacting atoms of
Withaferin A in Supplementary Tables 9–12.
3.4. Docking simulations pertaining to cholesterol
reduction property
The cholesterol reducing activity was assessed by docking
Withaferin A with the target proteins Scavenger receptor B1
(SR-B1) and Niemann-Pick-C 1 like intracellular cholesterol
transporter 1 (NPC1L1). SR-B1 is a membrane bound HDL
receptor that is involved in selective influx of HDL derived
cholesteryl esters into the cells and efflux of peripheral tissue
and macrophage based cholesterol back to the liver (Shen
et al., 2018). As a result, it is proposed that the inhibition of
this receptor could be of potential therapeutic importance.
NPC1L1 is a cholesterol transporter on the plasma membrane
that assists in uptake and absorption of cholesterol from the
small intestine enterocytes. Inhibition of this uptake essen-
tially prevents cholesterol absorption and retains it in the
lumen of the intestine for excretion(Davis & Veltri, 2007).
Withaferin A showed a favorable binding to NPC1L1 and SR-
B1 which would potentially inhibit these targets and as a
result, could be used to reduce the body cholesterol. Based
on this observation, here we are proposing inhibition of
NPC1L1 and SR-B1 by Withaferin A as a potential therapeutic
mechanism to address the problem of Hypercholesterolemia.
To the best of our knowledge, this is the first report on
this aspect.
The lowest binding energy for Withaferin A bound to SR-
B1 was found to be 7.16 kcal/mol with an inhibition con-
stant of 5.67 mM and that for NPC1L1 was found to be
5.72 kcal/mol with an inhibition constant of 63.94 mM.
Therefore, Withaferin A inhibits SR-B1 much more strongly at
much smaller dosage than NPC1L1. Supplementary Tables 13
and 14 gives a detailed account on the binding energies and
decomposed energies of all the possible docking of
Withaferin A to the 2 targets proteins.
The total intermolecular energy of the docking between
Withaferin A and SR-B1 was found to be 7.29 kcal/mol and
the interaction was entirely electrostatic with no
vdW þHbond þdesolv energy. Only the amino acid TYR443
was found to directly interact with Withaferin A with energy
JOURNAL OF BIOMOLECULAR STRUCTURE AND DYNAMICS 7
of interaction 1.650 kcal/mol. In contrast to SR-B1, the total
intermolecular energy of interaction of 5.64 kcal/mol for
NPC1L1 was distributed as 0.19 kcal/mol of electrostatic
energy and the remaining 5.45 kcal/mol of
vdW þHbond þdesolv energy. Even the amino acids directly
interacting with the ligand were much more diverse and
showed interactions majorly through polar amino acids
among others. The decomposed interaction energies of the
amino acids were found to be ARG498 (-1.2607),ASN497
(-1.0911),LEU482 (-0.5009), ARG452 (-0.9196),ASP545 (-0.469),
THR499 (-0.5247),GLU449 (-0.4472) and ASN496 (-0.0826). The
decomposed energies along with interacting atoms of
Withaferin A are represented in Supplementary Tables 15–18.
3.5. Molecular dynamic simulations
Since the molecular docking results only depict the rigid
poses of the ligand-protein complexes, they fail to explain
the flexibility of the residues and secondary structural ele-
ments that explain the stability of the docked complexes.
MD simulations show the structural disturbances of a target
protein in its biological environment and as a result, the
movement and conformational shifts of the protein as a con-
sequence of the ligand docking can be visualized (Bhardwaj
& Purohit, 2020; Purohit, 2014). Thus, MD simulations were
performed on the apoproteins of Nrf2 (for the anti-cancer
activity), SR-B1 (for the anti-hypercholesterolemic activity)
and a-glucosidase (for the anti-diabetic activity) and their
respective docked complexes with Withaferin A respectively
for 100 Nanoseconds (ns). The simulation curves were
obtained for the calculations of RMSD, RMSF, RG, H-Bonds,
Ligand RMSD, SASA, SSE and MMPSA calculations
respectively.
The RMSD curves depict the stability and balance of the
protein after the docking with the ligand. The RMSD was
selectively performed on the backbone Carbon atoms to
study the protein activity during the MD simulation lasting
100 ns. The deviations in the structure of the protein before
(apo-protein) and after (docked complex) binding to the lig-
and were measured as a function of time. The curves were
plotted for the apoprotein and the complex to access the
deviation before and after the docking. All the RMSD values
obtained during the MD Simulations are attached in
Supplementary spreadsheet 1. The RMSF curves depict the
fluctuation of specific protein residues upon its docking with
the ligand. Higher fluctuations of the residues that partici-
pate in protein-ligand interaction signify a greater destabil-
ization due to the ligand docking and is thus undesirable.
The RMSF curves were plotted for the Apo-protein and the
protein-ligand complex against all the amino acid residues of
the protein. Upon docking, the protein may fold or unfold
leading to a variation in its RG value. RG value indicates the
solidity or the compactness of the protein structure as a con-
sequence of docking. Thus, a smaller fluctuation in the RG
value can be inferred as a more rigid target, and it is ideal if
a fluctuation in the RG is consistent between the apo protein
and the protein-ligand complex. The RG plots were accessed
for the apo-proteins and the protein ligand complexes
through the entirety of the simulation lasting 100 ns.
Another parameter to determine the compactness of the
protein is the SASA which would indicate the modulation of
inhibiting ligands on the protein. In our study, the SASA val-
ues changed negligibly in all 3 of the target proteins
throughout the 100 ns simulation. The changes to the sec-
ondary structure of the protein were also directly observed
in terms of the number of residues influencing the propen-
sity of the secondary structure throughout the 100 ns.
However, the propensity of the secondary structure was also
found to be consistent with those of the respective apo-pro-
tein for all the 3 complexes. Although the potential H-bonds
were predicted in the docking analysis, they were re-calcu-
lated during the 100 ns simulation. The H-bonds influence
the interaction between the ligand and the protein. The
number of H-bonds remained consistent with the predictions
made during the docking study.
Finally, the python script for MmPbSaStat from the pack-
age by g_mmpbsa was used to estimate the average free
binding energy of the complexes during the last 20 ns of the
MD simulation. The binding energy of Withaferin A with
Nrf2, a-glucosidase and SR-BI was found to be 141.591 þ/-
17.521 kJ/mol, 85.755 þ/- 8.732 kJ/mol and 81.418 þ/-
15.594 kJ/mol respectively with the polar solvation energy
being 117.060 þ/- 15.178 kJ/mol, 39.865 þ/- 9.181 kJ/mol
and 54.429 þ/- 16.763 kJ/mol respectively; electrostatic
energy being 17.072 þ/- 7.362 kJ/mol, 6.418 þ/- 5.390 kJ/
mol and 7.075 þ/- 11.569 kJ/mol respectively and Van der
Waals energy being 222.293 þ/- 9.916 kJ/mol, 107.459
þ/- 6.851 kJ/mol and 116.242 þ/- 17.511 kJ/mol respect-
ively. These values are summarized in Table 3.
3.5.1. MD simulations for NRF2
The RMSD curve for the Nrf2 apo-protein and that docked
with the ligand showed an average RMSD value of 0.20 nm
and 0.21 nm with a standard deviation of 0.02 nm and
0.03 nm respectively. This average was calculated during the
simulation from 20 ns to 100 ns accounting to the variation
in the RMSD values up to 20 ns and the curve reaching a sta-
ble plateau from 20 ns to 100 ns. The RMSD plot for the
NRF2 protein and that of the ligand has been depicted in
Figure 4a and b. The Figure 4c depicts the RG variation curve
of Nrf apo-protein and complex throughout the simulation.
Clearly, the RG curve fluctuates between 1.75 nm and
1.85 nm but remain consistent throughout the simulation,
and these fluctuations in the complex remained consistent
with the apo-protein. The SASA simulations in Figure 4d also
show a negligible variation in the area throughout the simu-
lation for both the apo-protein and the complex. Figure 4e
shows the RMSF curve for the Nrf apo-protein and complex.
It can be seen that the curve shows significant flexibility in
all the residues of the protein and these flexibilities are con-
served before and after the docking. Finally, the secondary
structure propensity for Nrf2 is depicted in Figure 4f and g
shows the H-bonds in Nrf2 to reach as many as 4 in certain
instances of the simulation indicating a stable binding.
8 R. SURYA ULHAS AND A. MALAVIYA
3.5.2. MD simulations for SR-B1
The RMSD curve for the SR-B1 apo-protein and that docked
with the ligand, Withaferin A showed an average RMSD value
of 1.25 nm and 1.53 nm with a standard deviation of 0.16 nm
and 0.14 nm respectively. Although this is relatively higher
than the RMSD observed for Nrf2, it is important to consider
that the RMSD variations for the apo-protein and the com-
plex is still consistent since the RMSD of the apo-protein also
showed a considerably high magnitude. This average was
calculated during the simulation from 10 ns to 100 ns since
the curve in case of SR-B1 showed a greater variation up to
the initial 10 ns and then stabilized to a plateau from 10 ns
to 100 ns. The RMSD plot for the SR-B1 protein and that of
the ligand has been depicted in Figure 5a and b. The
Figure 5c depicts the RG variation curve of SR-B1 apo-protein
and complex throughout the simulation. The RG value varia-
tions are also significantly higher in SR-B1 as compared to
Nrf2, however the fluctuations in the complex remained con-
sistent with the apo-protein. The SASA simulations in Figure
5d once again showed a negligible variation in the area
throughout the simulation for both the apo-protein and the
complex for SR-B1. Figure 5e shows the RMSF curve for the
SR-B1 apo-protein and complex. It can be seen that the
curve shows a relatively lower flexibility compared to Nrf2
but remain significantly flexible in all the residues of the SR-
B1 protein these flexibilities are also conserved before and
Figure 4. MD simulation curves for Nrf2 protein; a. RMSD of the protein; b. RMSD of the ligand; c. RG; d. SASA; e. RMSF; f. Secondary structure analysis; g. H
bond analysis.
Figure 5. MD simulation curves for SR-B1; a. RMSD of the protein; b. RMSD of the ligand; c. RG; d. SASA; e. RMSF; f. Secondary structure analysis; g. H
bond analysis.
JOURNAL OF BIOMOLECULAR STRUCTURE AND DYNAMICS 9
after the docking. Finally, the secondary structure propensity
for SR-B1 is depicted in Figure 5f and g shows the H-bonds
in SR-B1 to reach as many as 3 in certain instances of the
simulation while remaining 2 in most cases, thus indicating a
stable binding.
3.5.3. MD simulations for a-glucosidase
Finally, the RMSD curve for the a-glucosidase apo-protein
and that docked with the ligand also showed an average
RMSD value of 0.24 nm and 0.25 nm with a standard devi-
ation of 0.03 for both the means respectively. This average
was also calculated during the simulation from 10 ns to
100 ns since the curve reached a stable plateau after the first
10 ns. The RMSD plot for the a-glucosidase protein and that
of the ligand has been depicted in Figure 6a and b. The
Figure 6c depicts the RG variation curve of a-glucosidase
apo-protein and complex throughout the simulation. Clearly,
the RG curve fluctuates between 2.85 nm and 2.90 nm but
remain consistent throughout the simulation, and these fluc-
tuations in the complex also remained consistent with the
apo-protein. The SASA simulations in Figure 6d once again
showed a negligible variation in the area throughout the
simulation for both the apo-protein and the complex. Figure
6e shows the RMSF curve for the a-glucosidase apo-protein
and complex. It can be seen that the curve shows significant
flexibility in all the residues of the protein and these flexibil-
ities are conserved before and after the docking. Finally, the
secondary structure propensity for a-glucosidase is depicted
in Figure 6f and g shows the H-bonds in a-glucosidase to
reach as many as 3 in certain instances of the simulation
while being between 1 and 2 for most part of the simulation
indicating a stable binding.
3.6. ADME, toxicity predictions and drug likeness
predictions
The ADMET studies on Withaferin A passed the preliminary
Lipinski’s rule. The molecular weight of Withaferin A was found
to be 470.606 Da with a Log P value of 3.3529. It displayed 6
Hydrogen bond acceptors and 2 Hydrogen bond donors
respectively. Additionally, 3 of its bonds were rotatable and
the molecule showed a polar surface area of 201.317 Å
2
.
Since the weight is below 500 Da with less than 5 H bond
donors and 10 H bond acceptors, and a log P value less than
10, Withaferin A was found to be a compatible druglike com-
pound. Additionally, 30 modules were run based on the cate-
gories Adsorption, Distribution, Metabolism, Excretion and
Toxicity. The output for these virtual module simulations is
shown in Supplementary Table 19.
4. Conclusion
The preliminary molecular docking of Withaferin A conducted
on the target proteins with their binding energies namely
Mortalin (-8.85 kcal/mol), Nrf2 (-12.59 kcal/mol), a-glucosidase
(-6.44 kcal/mol), b-glucosidase (-4.43 kcal/mol), NPC1
(-5.73 kcal/mol) and SRB1 (-7.16 kcal/mol) clearly suggests that
Withaferin A has potential anti-cancer, anti-diabetic and choles-
terol reducing properties where the mechanism of action
involves inhibition of these target proteins post binding.
Further, ADME þT studies show that Withaferin A is compound
which could be used as drug and obeys the Lipinski’s rule.
It is however important to understand that further experi-
mental studies are essential to clearly determine and validate
the proposed mechanism of action. This could include
molecular dynamics simulations of the other targets we pro-
posed. In addition, the cholesterol reducing property of W.
somnifera and hence Withaferin A is not validated by animal
trials and clinical studies. This is a potential area of future
Figure 6. MD simulation curves for a-Glucosidase; a. RMSD of the protein; b. RMSD of the ligand; c. RG; d. SASA; e. RMSF; f. Secondary structure analysis; g. H
bond analysis.
10 R. SURYA ULHAS AND A. MALAVIYA
examination. While animal studies for the anti-diabetic prop-
erties of W. somnifera are successfully reported and validates
our observation here in this work, very few studies have
used isolated Withaferin A as the drug while others continue
to use whole plant organ extracts which holds true for most
of the anti-cancer studies as well. It is important to under-
stand that isolated Withaferin A must be the test subject for
future work, since it has already been screened as a potential
bioactive compound and our work has even validated the
mechanism at a molecular level. Targeted use of Withaferin
A instead of the plant extracts also holds potential to opti-
mization, chemical manipulation, dosage standardization and
production at an industrial level.
Acknowledgements
We would like to thank Christ (Deemed to be University) for providing
support and encouragement for this research. We also thank Prof T.
Usha for her help and feedback during preparation of this MS.
Disclosure statement
No potential conflict of interest was reported by the authors.
Funding
This work was carried out as a part of RGS/F Project (GRD Number -
894) funded by Vision Group on Science and Technology (VGST),
Department of Electronics, IT, BT and S&T, Government of Karnataka.
ORCID
Alok Malaviya http://orcid.org/0000-0001-9808-3366
References
Abraham, M. J., Murtola, T., Schulz, R., P
all, S., Smith, J. C., Hess, B., &
Lindahl, E. (2015). GROMACS: High performance molecular simulations
through multi-level parallelism from laptops to supercomputers.
SoftwareX,1–2,19–25. https://doi.org/10.1016/j.softx.2015.06.001
Aggarwal, B. B., Ichikawa, H., Garodia, P., Weerasinghe, P., Sethi, G., Bhatt,
I. D., Pandey, M. K., Shishodia, S., & Nair, M. G. (2006). From traditional
Ayurvedic medicine to modern medicine: Identification of therapeutic
targets for suppression of inflammation and cancer. Expert Opinion on
Therapeutic Targets,10(1), 87–118. https://doi.org/10.1517/14728222.
10.1.87
Andrade, E. F., Lobato, R. V., De Ara
ujo, T. V., Zanger^
onimo, M. G., De
Sousa, R. V., & Pereira, L. J. (2015). Efecto de los beta-glucanos en el
control de los niveles de glucosa en pacientes diab
eticos: revisi
on sis-
tem
atica. Nutricion Hospitalaria,31(1), 170–177. https://doi.org/10.
3305/nh.2015.31.1.7597
Angelica, M. D., & Fong, Y. (2008). Molecular mechanisms of Nrf2-medi-
ated antioxidant response wenge. Journal of Molecular Biology,141(4),
520–529. 10.1002/mc.20465.Molecular
Bhardwaj, V. K., & Purohit, R. (2020). A new insight into protein-protein
interactions and the effect of conformational alterations in PCNA.
International Journal of Biological Macromolecules,148, 999–1009.
https://doi.org/10.1016/j.ijbiomac.2020.01.212
Bhattacharya, S. K., & Muruganandam, A. V. (2003). Adaptogenic activity
of Withania somnifera: An experimental study using a rat model of
chronic stress. Pharmacology, Biochemistry, and Behavior,75(3),
547–555. https://doi.org/10.1016/S0091-3057(03)00110-2
Bikadi, Z., & Hazai, E. (2009). Application of the PM6 semi-empirical
method to modeling proteins enhances docking accuracy of
AutoDock. Journal of Cheminformatics,1(1), 15–16. https://doi.org/10.
1186/1758-2946-1-15
Brooijmans, N., & Kuntz, I. D. (2003). Molecular recognition and docking
algorithms. Annual Review of Biophysics and Biomolecular Structure,32,
335–373. https://doi.org/10.1146/annurev.biophys.32.110601.142532
Carroll, M. D., & Lacher, A. (2010). Total and high-density lipoprotein chol-
esterol in adults: National Health and Nutrition Examination Survey,
2009–2010 (Vol. 92, pp. 1–8). U.S. Department of Health and Human
services - Centers for Disease Control and Prevention National Center
for Health Statistics.
Chadwick, A. C., Jensen, D. R., Hanson, P. J., Lange, P. T., Proudfoot, S. C.,
Peterson, F. C., Volkman, B. F., & Sahoo, D. (2017). NMR structure of the
C-terminal transmembrane domain of the HDL receptor, SR-BI, and a
functionally relevant leucine zipper motif. Structure (London, England :
1993),25(3), 446–457. https://doi.org/10.1016/j.str.2017.01.001
Davis, H. R., & Veltri, E. P. (2007). Zetia: Inhibition of Niemann-Pick C1
like 1 (NPC1L1) to reduce intestinal cholesterol absorption and treat
hyperlipidemia. Journal of Atherosclerosis and Thrombosis,14(3),
99–108. https://doi.org/10.5551/jat.14.99
Deocaris, C. C., Lu, W.-J., C. Kaul, S., & Wadhwa, R. (2012). Druggability of
mortalin for cancer and neuro-degenerative disorders. Current
Pharmaceutical Design,19(3), 418–429. https://doi.org/10.2174/
13816128130308
Derosa, G., & Maffioli, P. (2012). a-Glucosidase inhibitors and their use in
clinical practice. Archives of Medical Science : AMS,8(5), 899–906.
https://doi.org/10.5114/aoms.2012.31621
Ferguson, R. D., Gallagher, E. J., Scheinman, E. J., Damouni, R., & LeRoith,
D. (2013). The epidemiology and molecular mechanisms linking obes-
ity, diabetes, and cancer. In Gerald Litwack (Ed.), Vitamins and hor-
mones (1st ed., Vol. 93). Elsevier Inc. https://doi.org/10.1016/B978-0-
12-416673-8.00010-1
Gangadharappa, B. S., Sharath, R., Revanasiddappa, P. D., Chandramohan, V.,
Balasubramaniam, M., & Vardhineni, T. P. (2020). Structural insights of met-
allo-beta-lactamase revealed an effective way of inhibition of enzyme by
natural inhibitors. Journal of Biomolecular Structure & Dynamics,38(13),
3757–3771. https://doi.org/10.1080/07391102.2019.1667265
Ghani, U. (2015). Re-exploring promising a-glucosidase inhibitors for
potential development into oral anti-diabetic drugs: Finding needle in
the haystack. European Journal of Medicinal Chemistry,103, 133–162.
https://doi.org/10.1016/j.ejmech.2015.08.043
Guedes, I. A., de Magalh~
aes, C. S., & Dardenne, L. E. (2014). Receptor-lig-
and molecular docking. Biophysical Reviews,6(1), 75–87. https://doi.
org/10.1007/s12551-013-0130-2
Hakamata, W., Kurihara, M., Okuda, H., Nishio, T., & Oku, T. (2009). Design
and Screening Strategies for alpha-glucosidase inhibitors based on
enzymological information . Current Topics in Medicinal Chemistry,
9(1), 3–12. https://doi.org/10.2174/156802609787354306
Halder, D., Das, S., Joseph, A., & Jeyaprakash, R. S. (2022). Molecular
docking and dynamics approach to in silico drug repurposing for
inflammatory bowels disease by targeting TNF alpha. Journal of
Biomolecular Structure and Dynamics,40,1–14. https://doi.org/10.
1080/07391102.2022.2050948
Halgren, T. A. (1996). Merck molecular force field. I. Basis, form, scope,
parameterization, and performance of MMFF94. Journal of
Computational Chemistry,17(5–6), 520–552. https://doi.org/10.1002/
(SICI)1096-987X(199604)17:5/6<490::AID-JCC1>3.0.CO;2-P
Hanahan, D., & Weinberg, R. A. (2011). Hallmarks of cancer: The next gener-
ation. Cell,144(5), 646–674. https://doi.org/10.1016/j.cell.2011.02.013
Huang, C. S., Yu, X., Fordstrom, P., Choi, K., Chung, B. C., Roh, S. H., Chiu,
W., Zhou, M., Min, X., & Wang, Z. (2020). Cryo-EM structures of
NPC1L1 reveal mechanisms of cholesterol transport and ezetimibe
inhibition. Science Advances,6(25), 1–12. https://doi.org/10.1126/
sciadv.abb1989
Jemal, A., Bray, F., & Ferlay, J. (1999). Global cancer statistics: 2011. CA: A
Cancer Journal for Clinicians,49(2), 33–64. https://doi.org/10.3322/
caac.20107
Jeong, S. M., Choi, S., Kim, K., Kim, S. M., Lee, G., Son, J. S., Yun, J. M., &
Park, S. M. (2018). Association of change in total cholesterol level with
JOURNAL OF BIOMOLECULAR STRUCTURE AND DYNAMICS 11
mortality: A population-based study. Plos One,13(4), e0196030.
https://doi.org/10.1371/journal.pone.0196030
Kashyap, D., Tuli, H. S., Yerer, M. B., Sharma, A., Sak, K., Srivastava, S.,
Pandey, A., Garg, V. K., Sethi, G., & Bishayee, A. (2021). Natural prod-
uct-based nanoformulations for cancer therapy: Opportunities and
challenges. Seminars in Cancer Biology,69,5–23. https://doi.org/10.
1016/j.semcancer.2019.08.014
Kumari, R., Kumar, R., & Lynn, A. (2014). g_mmpbsa-a GROMACS tool for
high-throughput MM-PBSA calculations . Journal of Chemical
Information and Modeling,54(7), 1951–1962. https://doi.org/10.1021/
ci500020m
Lo, S. C., Li, X., Henzl, M. T., Beamer, L. J., & Hannink, M. (2006). Structure
of the Keap1:Nrf2 interface provides mechanistic insight into Nrf2 sig-
naling. The EMBO Journal,25(15), 3605–3617. https://doi.org/10.1038/
sj.emboj.7601243
Malaviya, A., Paari, K. A., Malviya, S., Kondapalli, V. K., Ghosh,., & A., S. R.
A. (2021). Gut microbiota and cancer correlates. In Indu Pal Kaur
(Editor-in-Chief) and Parneet Kaur Deol (Ed.), Probiotic research in ther-
apeutics: Applications in cancers and immunological diseases (Vol. 1,
pp. 1–27). Singapore: Springer.
Matsuda, H., Murakami, T., Kishi, A., & Yoshikawa, M. (2001). Structures of
withanosides I, II, III, IV, V, VI, and VII, new withanolide glycosides,
from the roots of Indian Withania somnifera DUNAL. and inhibitory
activity for tachyphylaxis to clonidine in isolated guinea-pig ileum.
Bioorganic & Medicinal Chemistry,9(6), 1499–1507. https://doi.org/10.
1016/S0968-0896(01)00024-4
Maurya, R. (2010). Chemistry and pharmacology of Withania coagulans :
An Ayurvedic remedy. The Journal of Pharmacy and Pharmacology,
62(2), 153–160. https://doi.org/10.1211/jpp.62.02.0001
Morris, G. M., Goodsell, D. S., Halliday, R. S., Huey, R., Hart, W. E., Belew,
R. K., & Olson, A. J. (1998). Automated docking using a Lamarckian
genetic algorithm and an empirical binding free energy function.
Journal of Computational Chemistry,19(14), 1639–1662. https://doi.
org/10.1002/(SICI)1096-987X(19981115)19:14%3C1639::AID-
JCC10%3E3.0.CO;2-B
Nguyen, T., Nioi, P., & Pickett, C. B. (2009). The Nrf2-antioxidant response
element signaling pathway and its activation by oxidative stress. The
Journal of Biological Chemistry,284(20), 13291–13295. https://doi.org/
10.1074/jbc.R900010200
Panwar, H., Calderwood, D., Grant, I. R., Grover, S., & Green, B. D. (2014).
Lactobacillus strains isolated from infant faeces possess potent inhibi-
tory activity against intestinal alpha- and beta-glucosidases suggest-
ing anti-diabetic potential. European Journal of Nutrition,53(7),
1465–1474. https://doi.org/10.1007/s00394-013-0649-9
Prasanth, D. S. N. B. K., Murahari, M., Chandramohan, V., Panda, S. P.,
Atmakuri, L. R., & Guntupalli, C. (2021). In silico identification of
potential inhibitors from Cinnamon against main protease and spike
glycoprotein of SARS CoV-2. Journal of Biomolecular Structure &
Dynamics,39(13), 4618–4632. https://doi.org/10.1080/07391102.2020.
1779129
Purohit, R. (2014). Role of ELA region in auto-activation of mutant KIT
receptor: A molecular dynamics simulation insight. Journal of
Biomolecular Structure & Dynamics,32(7), 1033–1046. https://doi.org/
10.1080/07391102.2013.803264
Rai, M., Jogee, P. S., Agarkar, G., & Santos, C. A. D. (2016). Anticancer
activities of Withania somnifera: Current research, formulations, and
future perspectives. Pharmaceutical Biology,54(2), 189–197. https://
doi.org/10.3109/13880209.2015.1027778
Roig-Zamboni, V., Cobucci-Ponzano, B., Iacono, R., Ferrara, M. C.,
Germany, S., Bourne, Y., Parenti, G., Moracci, M., & Sulzenbacher, G.
(2017). Structure of human lysosomal acid a-glucosidase-A guide for
the treatment of Pompe disease. Nature Communications,8(1), 1-8.
https://doi.org/10.1038/s41467-017-01263-3
Rutwick Surya, U., & Praveen, N. (2021). A molecular docking study of
SARS-CoV-2 main protease against phytochemicals of Boerhavia dif-
fusa Linn. for novel COVID-19 drug discovery. Virusdisease,32(1),
46–54. https://doi.org/10.1007/s13337-021-00683-6
Saleem, S., Muhammad, G., Hussain, M. A., Altaf, M., & Abbas Bukhari,
S. N. (2020). Withania somnifera L.: Insights into the phytochemical
profile, therapeutic potential, clinical trials, and future prospective.
Iranian Journal of Basic Medical Sciences,23(12), 1501–1526. https://
doi.org/10.22038/ijbms.2020.44254.10378[33489024
Sari, A. N., Bhargava, P., Dhanjal, J. K., Putri, J. F., Radhakrishnan, N.,
Shefrin, S., Ishida, Y., Terao, K., Sundar, D., Kaul, S. C., & Wadhwa, R.
(2020). Combination of withaferin-A and CAPE provides superior anti-
cancer potency: Bioinformatics and experimental evidence to their
molecular targets and mechanism of action. Cancers,12(5),
1160–1125. https://doi.org/10.3390/cancers12051160
Sch€
uttelkopf, A. W., & Van Aalten, D. M. F. (2004). PRODRG: A tool for
high-throughput crystallography of protein-ligand complexes. Acta
Crystallographica. Section D, Biological Crystallography,60(Pt 8),
1355–1363. https://doi.org/10.1107/S0907444904011679
Sever, R., & Brugge, J. S. (2015). Genetic and epigenetic mechanisms of
cancer progression. Cold Spring Harbor Perspectives in Medicine,5(4),
A006098–21. https://doi.org/10.1101/cshperspect.a006098
Sharma, A., Shanker, C., Tyagi, L. K., Singh, M., & Rao, C. V. (2008). Herbal
medicine for market potential in India: An overview. Academic Journal
of Plant,1(2), 26–36.
Shen, W. J., Azhar, S., & Kraemer, F. B. (2018). SR-B1: A unique multifunc-
tional receptor for cholesterol influx and efflux. Annual Review of
Physiology,80,95–116. https://doi.org/10.1146/annurev-physiol-
021317-121550
Solis, F. J., & Wets, R. J. B. (1981). Minimization by random search techni-
ques. Mathematics of Operations Research,6(1), 19–30. https://doi.org/
10.1287/moor.6.1.19
Surya, U. R., Naik, P. M., Vinayak, U., & Praveen, N. (2021). A critical
review of anticancer properties of Withania somnifera (L.) Dunal with
respect to the biochemical mechanisms of its phytochemical constitu-
ents. Plant Science Today,8(2), 236–249. https://doi.org/10.14719/pst.
2021.8.2.969
Szarc Vel Szic, K., Op De Beeck, K., Ratman, D., Wouters, A., Beck, I. M.,
Declerck, K., Heyninck, K., Fransen, E., Bracke, M., De Bosscher, K.,
Lardon, F., Van Camp, G., & Vanden Berghe, W. (2014).
Pharmacological levels of Withaferin A (Withania somnifera) trigger
clinically relevant anticancer effects specific to triple negative breast
cancer cells. PLoS ONE,9(2), e87850. https://doi.org/10.1371/journal.
pone.0087850
Tribolo, S., Berrin, J. G., Kroon, P. A., Czjzek, M., & Juge, N. (2007). The
Crystal Structure of Human Cytosolic beta-glucosidase unravels the
substrate aglycone specificity of a family 1 glycoside hydrolase .
Journal of Molecular Biology,370(5), 964–975. (https://doi.org/10.1016/
j.jmb.2007.05.034
US Department of Health and Human Services. (2020). National Diabetes
Statistics Report, 2020. National Diabetes Statistics Report, 2.
Vaishnavi, K., Saxena, N., Shah, N., Singh, R., Manjunath, K., Uthayakumar,
M., Kanaujia, S. P., Kaul, S. C., Sekar, K., & Wadhwa, R. (2012).
Differential activities of the two closely related withanolides,
Withaferin A and Withanone: Bioinformatics and experimental eviden-
ces. PLoS ONE,7(9), e44419. https://doi.org/10.1371/journal.pone.
0044419
Vigneri, P., Frasca, F., Sciacca, L., Pandini, G., & Vigneri, R. (2009).
Diabetes and cancer. Endocrine-Related Cancer,16(4), 1103–1123.
https://doi.org/10.1677/ERC-09-0087
Widodo, N., Takagi, Y., Shrestha, B. G., Ishii, T., Kaul, S. C., & Wadhwa, R.
(2008). Selective killing of cancer cells by leaf extract of
Ashwagandha: Components, activity and pathway analyses. Cancer
Letters,262(1), 37–47. https://doi.org/10.1016/j.canlet.2007.11.037
Yang, L., Li, H., Jiang, Y., Zuo, J., & Liu, W. (2013). Inhibition of mortalin
expression reverses cisplatin resistance and attenuates growth of
ovarian cancer cells. Cancer Letters,336(1), 213–221. https://doi.org/10.
1016/j.canlet.2013.05.004
Yun, C. O., Bhargava, P., Na, Y., Lee, J. S., Ryu, J., Kaul, S. C., & Wadhwa,
R. (2017). Relevance of mortalin to cancer cell stemness and cancer
therapy. Scientific Reports,7, 42016–42010. https://doi.org/10.1038/
srep42016
Zimmet, P. Z., Magliano, D. J., Herman, W. H., & Shaw, J. E. (2014).
Diabetes: A 21st century challenge. The Lancet. Diabetes &
Endocrinology,2(1), 56–64. https://doi.org/10.1016/S2213-8587(13)70
112-8
12 R. SURYA ULHAS AND A. MALAVIYA
