Content uploaded by Tara Chand Yadav
Author content
All content in this area was uploaded by Tara Chand Yadav on Mar 19, 2019
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: 0739-1102 (Print) 1538-0254 (Online) Journal homepage: https://www.tandfonline.com/loi/tbsd20
Exploration of interaction mechanism of tyrosol as
a potent anti-inflammatory agent
Tara Chand Yadav, Naresh Kumar, Utkarsh Raj, Nidhi Goel, Pritish Kumar
Vardawaj, Ramasare Prasad & Vikas Pruthi
To cite this article: Tara Chand Yadav, Naresh Kumar, Utkarsh Raj, Nidhi Goel, Pritish Kumar
Vardawaj, Ramasare Prasad & Vikas Pruthi (2019): Exploration of interaction mechanism of
tyrosol as a potent anti-inflammatory agent, Journal of Biomolecular Structure and Dynamics, DOI:
10.1080/07391102.2019.1575283
To link to this article: https://doi.org/10.1080/07391102.2019.1575283
Published online: 19 Mar 2019.
Submit your article to this journal
View Crossmark data
Exploration of interaction mechanism of tyrosol as a potent anti-
inflammatory agent
Tara Chand Yadav
a
, Naresh Kumar
b
, Utkarsh Raj
c
, Nidhi Goel
d
, Pritish Kumar Vardawaj
c
, Ramasare
Prasad
a
and Vikas Pruthi
a
a
Department of Biotechnology, Indian Institute of Technology Roorkee, Roorkee, India;
b
Discipline of Biosciences and Biomedical
Engineering, Indian Institute of Technology Indore, Indore, India;
c
Department of Bioinformatics, Indian Institute of Information Technology
Allahabad, Allahabad, India;
d
Department of Chemistry, Institute of Science, Banaras Hindu University, Varanasi, India
Communicated by Ramaswamy H. Sarma
ABSTRACT
Drug discovery for a vigorous and feasible lead candidate is achallenging scientific mission as it requires expert-
ise, experience, and huge investment. Natural products and their derivatives having structural diversity are
renowned source of therapeutic agents since many years. Tyrosol (a natural phenylethanoid) has been extracted
from olive oil, and its structure was confirmed by elemental analysis, FT-IR, FT-NMR, and single crystal X-ray crys-
tallography. The conformational analysis for tyrosol geometry was performed by Gaussian 09 in terms of density
functional theory. Validation of bond lengths and bond angles obtained experimentally as well as theoretically
were performed with the help of curve fitting analysis, and values of correlation coefficient (R) obtained as 0.988
and 0.984, respectively. The charge transfer within the tyrosol molecule was confirmed by analysis of
HOMO!LUMO molecular orbitals. In molecular docking with COX-2 (PDB ID: 5F1A), tyrosol was found to pos-
sess satisfactory binding affinity as compared to other NSAIDs (Aspirin, Ibuprofen, and Naproxen) and a COX-2
selective drug (Celecoxib). ADMET prediction, drug-likeness and bioactivity score altogether confirm the lead/
drug like potential of tyrosol. Further investigation of simulation quality plot, RMSD and RMSF plots, ligands
behavior plot as well as post simulation analysis manifest the consistency of 5F1A-tyrosol complex throughout
the 20 ns molecular simulation process that signifies its compactness and stability within the receptor pocket.
Abbreviations: ADMET: Absorption, Distribution, Metabolism, Excretion and Toxicity; Å: Angstrom;
COX-2: Cyclooxygenase-2; DFT: Density Functional Theory; DMF: Dimethylformamide; FMO: Frontier
Molecular Orbital; FT-IR: Fourier-transform Infrared Spectroscopy; FT-NMR: Nuclear Magnetic Resonance
Spectroscopy; HOMO: Highest Occupied Molecular Orbital; LUMO: Lowest Unoccupied Molecular
Orbital; MD: Molecular Dynamics; NS: Nanosecond; NSAIDs: Non-steroidal anti-inflammatory drugs;
OPE: Osiris Property Explorer; RMSD: Root-Mean-Square Deviation; RMSF: Root Sean Square Fluctuation
ARTICLE HISTORY
Received 14 November 2018
Accepted 22 January 2019
KEYWORDS
Tyrosol; ab initio; COX-2;
ADMET; molecular docking;
molecular dynam-
ics simulation
CONTACT Vikas Pruthi vikasfbs@iitr.ac.in;vikasfbs@gmail.com Department of Biotechnology, Indian Institute of Technology Roorkee, Roorkee,
Uttarakhand 247667, India
Tara Chand Yadav and Naresh Kumar contributed equally as first author.
Supplemental data for this article can be accessed online at https://doi.org/10.1080/07391102.2019.1575283.
ß2019 Informa UK Limited, trading as Taylor & Francis Group
JOURNAL OF BIOMOLECULAR STRUCTURE AND DYNAMICS
https://doi.org/10.1080/07391102.2019.1575283
1. Introduction
Natural products are admirable source in drug discovery due
to their structural diversity as well as wide range of bio-
logical activities (Beutler, 2009; Harvey, 2008; Koehn & Carter,
2005; Kumar et al., 2016; Shen, 2015). In past, there are large
numbers of articles have been published on the extraction of
natural bioactive molecules from plants, including polyphe-
nols and their beneficial effects on human being (Ferguson,
Zhu, & Harris, 2005; Fuggetta, Cottarelli, Lanzilli, Tricarico, &
Bernini, 2012; Kumar & Pruthi, 2014; Kumar, Pruthi, & Goel,
2015; Manach, Williamson, Morand, Scalbert, & R
em
esy, 2005;
Montedoro, Servili, Baldioli, & Miniati, 1992; Ou, Li, & Gao,
1999; Toshihiro et al., 2000). In the Mediterranean area, the
daily consumption of wine and virgin olive oil is related to
the protection against several types of cancer, and the bene-
fits are mainly contributed by the phenolic compounds such
as tyrosol (Kontou, Psaltopoulou, Panagiotakos, Dimopoulos,
& Linos, 2011). Tyrosol, 2-(4-hydroxyphenyl)-ethanol, a deriva-
tive of phenylethyl alcohol, is one of the key bioactive phen-
olic compound, abundantly found in olive oil, white wine,
and plant extracts (Thibault, Arseneault, Longpre, &
Ramassamy, 2011). It possesses scavenging properties on
peroxynitrite and superoxide anion (Bertelli et al., 2002;
Puerta, Mart
ınez Dom
ınguez, Ru
ız-Gut
ıerrez, Flavill, & Hoult,
2001). Tyrosol also exhibits anti-genotoxic activity, averts
apoptosis in keratinocyte, reduces the expression of cell
adhesion molecule resulting in prevention of endothelial
dysfunction, inhibition of platelet-induced aggregation,
antioxidant, anticancer, anti-depressant, anti-stress, cardio-
protective, anti-osteoporosis, anti-inflammatory, anti-hyper-
glycemic, and neural protective effects (Ahn et al., 2008;
Bertelli et al., 2002; Bu et al., 2007; Carluccio et al., 2003;
Chandramohan, Pari, Rathinam, & Sheikh, 2015; Dinnella,
Minichino, D’Andrea, & Monteleone, 2007; Giovannini et al.,
1999; Loru et al., 2009; Panossian et al., 2008; Panossian,
Wikman, Kaur, & Asea, 2009; Petroni et al., 1995; Puerta et al.,
2001; Salucci et al, 2015; Thibault et al., 2011). Keeping the
broad spectrum of biological functions of tyrosol in mind,
there is a necessity to gain more insight toward its 3D-geom-
etry, structural characteristics, and bioactive mechanism.
Cyclooxygenase (COX), known as prostaglandin-endoperox-
ide synthase (PTGS), is a rate determining enzyme in the syn-
thesis of eicosanoids. It has two isoforms; COX-1 and COX-2,
which differs mainly in their expression pattern (Funk, Funk,
Kennedy, Pong, & Fitzgerald, 1991; Hempel, Monick, &
Hunninghake, 1994). COX-1 is constitutively expressed,
mainly obtained from cell lines of mammalian tissues, and
often regarded as a housekeeping enzyme. It is primarily
involved in cell-to-cell signaling, tissue homeostasis, cytopro-
tection, kidney and platelet function, (Kelley et al., 1997) and
also helps to maintain normal mucus lining of the stomach
(Dubois et al., 1998), whereas COX-2 is responsible for the
formation of key biological mediators such as prostanoids.
Earlier experimental studies confirm the role of COX-2
enzyme in numerous inflammatory processes and different
carcinomas (Mestre et al., 1999). Fascinatingly, many
antioxidants with chemo-protective effects are reported to
suppress the expression of COX-2 by disturbing the signaling
pathway that regulates the COX-2 gene (Chinery et al., 1998).
Therefore, COX-2 gene is also used as a biomarker against
cancer to assess the chemo-protective effects of natural phy-
tochemicals, and the discovery of novel COX-2 inhibitor is
considered to be a ray of hope for the prevention of tumori-
genesis and inflammation (Cheki et al., 2018; Dhanjal et al.,
2015; Kumar & Pruthi, 2015; Kumar et al., 2018; Nile, Ko, Kim,
& Keum, 2016; Pang, Hurst, & Argyle, 2016).
Herein, extraction and structural elucidation of tyrosol
from olive oil followed by ab initio calculations, molecular
docking, and molecular dynamics simulation studies with
COX-2 have been carried out to investigate its role as well as
mechanism in anti-inflammatory reactions and comparison
with clinically approved anti-inflammatory drugs (aspirin, ibu-
profen, naproxen, and celecoxib). This is the first report of
structure determination and comparison of experimental and
theoretical data on tyrosol to the best of our knowledge.
2. Experimental
2.1. Chemicals and extraction
The sample used for the extraction of tyrosol was oil of olive.
The reference compounds (tyrosol) and chromatographic sol-
vents like methyl alcohol, ethyl acetate, acetonitrile, heptanes,
syringic acid, gallic acid, DMF etc. others were analytical grade
and purchased from TCI Chemicals and Himedia, India. Tyrosol
was extracted in modifications of earlier reported methods
(Adhami et al., 2015; Brenes, Garc
ıa, Garc
ıa, & Garrido, 2000;
Romero & Brenes, 2012). Briefly, 2.0 g olive oil sample was
treated with 2.0 mL (3 times) of DMF followed by washing with
hexane. N
2
gas was fizzed into extract for the removal of
remaining hexane. Extract was filtered through 0.22 lm pore
size filter and introduced into the chromatograph. Separation
of tyrosol was attained by using an elution gradient technique
(initial composition of 90% water and phosphoric acid at pH
3.0) and 10% MeOH. Concentration of MeOH was raised to
30% over 10 min and it was continued for 20 min.
Subsequently, MeOH was lifted up to 40, 50, 60, 70, and 100%
in 5 min internals at 1 mL/min flow rate and 35 C temperature.
2.2. Instrumentation
Crystallized tyrosol was dried completely preceding to elem-
ental analysis, and checked by Elementar Vario EL III analyzer.
The FT-IR/NMR spectra were recorded on Tensor 27, BRUKER
FT-IR spectrometer, and Avance 500 Bruker Biospin Intl
500 MHz, respectively. Melting point recorded on a JSGW
apparatus and was uncorrected. X-ray data was collected by
Bruker Kappa Apex four circle-CCD diffractometer having
graphite monochromated Mo Karadiation (k¼0.71070 Å).
SAINT v7.34 and SADABS programs were used for Lorentz-
Polarization and empirical absorption corrections, respect-
ively (SAINT, 2004; Sheldrick, 1996). Crystal structure of tyro-
sol was solved by using direct methods, while structure
2 T. C. YADAV ET AL.
solution, refinement and data output were accomplished
with the SHELXTL program (Sheldrick, 1990,2000). Hydrogen
atoms were located in geometrically calculated positions
using riding model, while non-hydrogen atoms were enlight-
ened anisotropically. Images were produced in the crystal lat-
tice with DIAMOND software (Brandenburg, 1999;
Klaus, 1999).
2.3. DFT and statistical studies
Geometry of tyrosol, aspirin, naproxen, ibuprofen, and cele-
coxib were optimized at 6-311G (d,p) basis set with a hybrid
function B3LYP by Gaussian 09 (Becke, 1993; Frisch, 2009).
Conformational analysis was performed to find most stable
conformation of tyrosol (Hockridge et al., 1999). Inputs for
geometry optimization were given in the form of z-matrix
constructed by GaussView 5.0 (Goel & Kumar, 2018).
Structural parameters for optimized geometry like total
energy, dipole moment, bond lengths, bond angles, and
HOMO–LUMO energy have been computed at the same
basis set (Koopmans, 1934; Tsuneda & Hirao, 2014; Zhang &
Musgrave, 2007). The values of correlation coefficient (R)
were calculated for statistical validation between crystal
structure and DFT optimizations data for bond lengths and
angles (Kumar et al., 2015).
2.4. Calculation of ADME and toxicity
In the course of development of a new pharmacologically
active molecule, in vitro assessment at an early stage using
ADMET (Absorption, Distribution, Metabolism, Excretion, and
Toxicity) and in vivo evaluation of pharmacodynamics param-
eters offers a basis for identification of new molecular enti-
ties which may serves as probable lead drug entrant with
suitable efficacy, metabolism, and pharmacokinetics. The
ADME/TOX descriptors of a chemical entity allow the
researchers to comprehend the necessary safety and efficacy
information for regulatory approval. QikProp v4.4,
(Schr€
odinger, LLC, New York, 2015) was used to compute the
ADME descriptors like molecular weight (MW), total polar
surface area, dipole moment, solvent accessible surface area,
hydrophobic and hydrophilic component of SASA, MDCK
(Madin-Darby Canine Kidney), Caco2-cell (heterogeneous
human epithelial colorectal adenocarcinoma cell lines), pre-
dicted polarizability, predicted hexadecane/gas partition
coefficient, predicted octanol/gas partition coefficient, pre-
dicted water/gas partition coefficient, predicted octanol/
water partition coefficient, predicted aqueous solubility, con-
formation-independent predicted aqueous solubility, etc. for
tyrosol, aspirin, naproxen, ibuprofen, and celecoxib. A dis-
tinctive stress has been laid on the prediction of IC
50
value
for blockage of HERG K
þ
channels for all the screened dual
inhibitors to check their potency. Toxicity prediction was car-
ried out by using Osiris Property Explorer (OPE; a web-based
tool) of five docked compounds (http://www.organic-chemis-
try.org/prog/peo/).
2.5. Molecular docking
Molecular docking of tyrosol, aspirin, naproxen, ibuprofen,
and celecoxib with aspirin-acetylated human COX-2 (PDB ID:
5F1A) receptor has been carried out using GLIDE v6.7
(Schrodinger, LLC, New York, 2015). Extra precision (XP)
method was applied for docking purpose of all five com-
pounds into the defined binding cavity of receptor devoid of
constraint and positioned on the basis of docking score.
Glide internally generates different conformations of com-
pounds, which are then passed-on through a series of filters.
In XP docking, compounds exhibiting good scores for a
range of interactions achieved by extensive sampling with
advanced scoring and avoiding penalties are selected to
eliminate the false positives. Finally, docking results are
assessed on the basis of scoring function given by Glide
score or Glide G-score, which can be represented as:
GScore ¼0:065 VanderWaalsenergy þ0:130
Coulombenergy þLipo þHbond þMetal þBuryP
þRotB þSite
where, hydrophobic interactions are indicated as Lipo, metal
binding by Metal, buried polar group penalty was repre-
sented as BuryP, penalty for freezing rotatable bonds was
shown by RotB and polar interactions existing in the active
site represented by Site. A comparative analysis of tyrosol
with well-known clinically marketed drugs (aspirin, naproxen,
ibuprofen, and celecoxib) was carried out on the basis of
docking score and involved interacting residues of
active site.
2.5.1. Ligand preparation
Optimized 3D geometries of currently prescribed drugs for
clinical purpose along with tyrosol were converted into .pdb
file format prior to molecular docking (Kumar & Garg, 2014).
Ligand processing followed by expansion of protonation and
tautomeric states (7.0 þ2.0 pH units) was done by LigPrep
along with Epik module (Shaw, Yibing, John, Michael, & Ron,
2005). For each ligand, a total of five stereoisomers with min-
imum energy conformations were produced, and the lowest
energy conformer with accurate chirality was selected for fur-
ther study.
2.5.2. Preparation of target receptor and active site
identification
The 3D X-ray crystal structure of aspirin-acetylated human
COX-2 at a resolution of 2.38 Å was obtained from the RCSB-
PDB (Protein Data Bank; http://www.rcsb.org/) with PDB ID:
5F1A (Lucido, Orlando, Vecchio, & Malkowski, 2016). The pro-
tein structure of salicylate bound human COX-2 complex was
further processed and analyzed to assess conformational sta-
bility. Steric clashes, hetero-atoms and non-essential water
molecules were scraped off, as well as hydrogen bond and
proper bond orders were assigned to crystal structure with
the help of Maestro Protein Preparation Wizard Workflow
program suite v10.2, Schr€
odinger, LLC, New York, NY, 2015,
and the default parameters were used (Sastry, Adzhigirey,
JOURNAL OF BIOMOLECULAR STRUCTURE AND DYNAMICS 3
Day, Annabhimoju, & Sherman, 2013). SiteMap, v3.4,
Schr€
odinger, LLC, New York, NY, 2015 was utilized to ascer-
tain the active site of 5F1A protein structure (Halgren, 2009),
residues lying within the range of 5 Å in vicinity of Ser530
were defined as active site residues and the same has also
been substantiated from the literature as well around which
the docking receptor grid was generated. The preparation
and processing of receptor provides all the necessary envir-
onmental settings and validates the protein structure in-silico
which brings the structure to mimic the in vitro surrounding
condition. The ingrained default constraints were utilized for
cutoff, neutralization, scaling, and dimension of the binding
pocket. The electrostatic grid box was generated around the
target receptor via Glide v6.6, Schr€
odinger, LLC, New York,
NY, 2015 module (Halgren et al., 2004). The dimensions of
the receptor grid were kept as: Inner box: X¼10, Y¼10,
Z¼10, Outer box: X¼30, Y¼30, Z¼30 and Grid center
X¼42.73463, Y¼26.3990, Z¼240.0050.
2.6. Molecular dynamics simulations
Structural and functional veracity of proteins owes to its
molecular function which is ultimately influenced by its con-
formational stability and integrity, can be studied by using
atomic level perturbation in simulated real time environment
via MD (Molecular dynamics) simulation. In our present
investigation, MD simulation has been carried out to assess
the conformational stability of protein-ligand complex by
using Desmond tool (Bowers et al., 2006, Maestro-Desmond
Interoperability Tools, version 4.1, Schr€
odinger, New York,
2015). MD simulations analysis for tyrosol in complex with
aspirin-acetylated human COX-2 receptor was evaluated on
the basis of binding affinity, number of H-bonds, ADME and
toxicity parameters. Stability and conformational behavior
was assessed by analyzing the simulation quality plots, pro-
tein–ligand RMSD, protein–ligand contact graph and root
mean square fluctuation (RMSF) plot throughout the
20 ns simulation.
2.6.1. System building
The receptor–ligand complex system was processed and side
chains of different residues are refined, followed by the strain
minimization using protein preparation interface of Maestro
(Schr€
odinger, LLC, New York, NY, 2015) prior to MD simula-
tions. In the course of complex processing, all the missing
atoms of 5F1A-tyrosol system were added and processed
complex was then introduced to the system builder module
of Desmond, encompassing ions and solvation tabs.
Undesirable water molecules were scraped off from the crys-
tal structure of complex and TIP3P water model system was
chosen for the solvation of 5F1A-tyrosol complex by means
of solvation panel (Jorgensen, Chandrasekhar, Madura,
Impey, & Klein, 1983). Optimized potentials for liquid simula-
tions (OPLS_2005) force-field was utilized for amino acid
interaction (Shivakumar et al., 2010), encompassing SPC
(Simple point charge) solvent archetype for docked complex
and orthorhombic water box of dimension (10 10 10 Å)
was created to confirm the complete solvent coverage of
entire surface 5F1A-tyrosol system creating a box volume of
613,915 Å. Net charge on the system was neutralized and
balanced by adding two Na
þ
ions. To provide the real time
environment for simulation purpose, 0.15 M NaCl was also
added to stabilize the system (Bowers et al., 2006).
2.6.2. Molecular dynamics
System building of 5F1A-tyrosol complex was followed by
molecular dynamics simulation of the complex. Periodic
boundary conditions were applied and all the bad contacts
of residues were removed by energy minimization with the
help of hybrid method steepest decent and the limited-
memory Broyden Fletcher Goldfarb Shanno (LBFGS)
algorithms (Guo et al., 2010). Prior to equilibration and final
MD production run, the system was minimized by applying
1 kcal/mol/Å of convergence threshold followed by 2000 iter-
ations and pre-equilibration was carried out with the help of
ingrained relaxation protocol employed in Desmond. The
complete system was exposed to 300 K and 1 bar pressure
for 20 ns at NPT ensemble of MD simulation with recording
interval maintained at 10 ps for total energy (kcal mol). The
structural variations along with dynamic behavior were
examined by determining the stability of protein via calcula-
tion of root mean square deviation (RMSD) –defines the sta-
bility of protein, potential energy and RMSF determining the
thermal movement of an individual residue and their fluctu-
ation, its scoring guarantees the overall alignment of the sys-
tem in surrounding and observance of large structural
changes and stable nature of the receptor–ligand complex
can be analyzed (Raj, Kumar, Gupta, & Varadwaj, 2015).
3. Results and discussions
3.1. Elemental analysis, FT-IR, and FT-NMR studies
Purified tyrosol was further processed for its structural char-
acterization. Melting point of purified tyrosol was recorded
as 91–93 C. Elemental analysis calculated for tyrosol
(C
8
H
10
O
2
, 138.16) was C: 69.54% and H: 7.30%, while experi-
mentally found C: 68.21% and H: 7.23% for the same.
According to the FT-IR spectrum (KBr, Figure S1), tyrosol
exhibits the transmittance peaks at 3426 cm
1
, which is due
to asymmetric stretching vibrations of phenolic as well as
alcoholic O–H groups. In addition, stretching band at
1262 cm
1
, this reflects the presence of C–O. The weak
absorption of 2930 cm
1
has been designated to m
as
(C–H)
stretching vibration, C ¼C stretching vibrations occurs at
1632 and 1563 cm
1
. The bands in the region of
950–550 cm
1
, represent strong C–H bending transmittance.
1
H-NMR (500 MHz, DMSO-d
6
,ppm, Figure S2) d: 2.49-2.61 (dd,
2H, –CH
2
,J¼6 Hz), 3.49–3.54 (dd, 2H, –CH
2
,J¼2.5 Hz),
4.56–4.59 (dd, 1H, Alcohol, J¼1.5 Hz), 6.64–6.66 (d, 2H,
Aromatic, J¼1 Hz), 6.97–6.99 (d, 2H, Aromatic, J¼6.1 Hz),
9.168 (s, 1H, –OH exch). The presence of exchangeable pro-
tons (phenolic and carboxylic) in tyrosol molecule has been
confirmed by adding a few drops of D
2
O in NMR sample
4 T. C. YADAV ET AL.
tube.
13
C-NMR (125 MHz, DMSO-d
6
,ppm, Figure S3) d: 38.69,
63.06, 115.39, 130.14 and 155.88.
3.2. Crystal structure
Tyrosol crystallizes in monoclinic system with space group P
21 (Figure 1(a)). Refinement parameters, crystallographic
data, bond lengths, bond angles, isotropic and anisotropic
displacement parameters are provided in Table 1 and Tables
S1–S3. The single crystal X-ray structure analysis reveals that
one tyrosol molecule is coordinated to four other tyrosol
molecules through O–HO non-covalent interactions
(O1–H1O2, 1.889(3) Å; O2–H2AO1, 1.992(3) Å), and associ-
ated to create a two dimensional ladder like perspective
view (Figure 1(c)).
3.3. Geometry optimization, FMOs and
statistical analysis
Conformational analysis for tyrosol (Figure S4) has been car-
ried out, and the lowest energy structure (Figure S4e) was
processed for further theoretical calculation and comparison
with X-ray crystal structure (Hockridge et al., 1999). It showed
the positive harmonic vibrational frequencies, indicated glo-
bal minimum on the potential energy surface. It also showed
a resemblance with the structure obtained by single crystal
diffraction (Figure 1(a)). Computation of frontier molecular
orbitals (FMOs), which plays a major role in the prediction of
electronic and optical properties of a molecule, has been
achieved by HOMO (highest occupied molecular orbital) and
Table 1. Crystallographic data collection and refinement parameters for iso-
lated tyrosol.
Empirical formula C
8
H
10
O
2
Color White
Formula weight (g mol
1
) 138.16
Crystal system Monoclinic
Space group P21
a(Å) 5.3179(3)
b(Å) 8.3579(4)
c(Å) 8.1495(4)
a() 90.0
b() 94.705(4)
c() 90.0
V(Å
3
) 361.00(3)
Z2
D
Calc
(g/cm
3
) 1.271
m(Mo K
a
)mm
1
) 0.090
Crystal size (mm) 0.32 0.26 0.19
h
max
() 2.508–28.306
Reflections measured 1727
Observed reflections 1369
Data/restraints/parameters 1727/1/93
Final Rindices[I>2(I)}]
R
1a
0.0406
wR
2b
0.0978
CCDC number 1584517
a
R
1
¼PkF
o
jjF
c
k=PjF
o
j,
b
wR
2
¼{P[w(F
o2
F
c2
)
2
]/Pw(F
o2
)
2
.
Figure 1. Ball and stick models of tyrosol’s geometry (a) solid state single crystal, (b) DFT optimized, and (c) 3D packing showing various O–HO non-covalent
interactions. Color code: C, gray/black; H, orange/white; O, red.
JOURNAL OF BIOMOLECULAR STRUCTURE AND DYNAMICS 5
LUMO (lowest unoccupied molecular orbital) analysis. HOMO
is known as the ability of a molecule to donate an electron,
while LUMO corresponds to the ability to obtain an electron.
The energy gap between HOMO and LUMO is used to calcu-
late the optical and electronic properties viz. chemical hard-
ness, chemical softness, kinetic stability, and optical
polarizability of a molecule (Goel & Kumar, 2017; Pearson,
1988; Vasilescu & Adrian-Scotto, 2010). The values of these
parameters are provided in Table 2. As shown in Figure 2(a),
the pictorial representations of isodensity plots energy transi-
tion for tyrosol clearly confirm that in ground state (HOMO),
the density is distributed throughout the molecule mainly at
benzene ring, while in case of LUMO the density is con-
densed around the upper and lower four carbon atoms of
benzene ring. The phenol group also lost the density from
HOMO!LUMO indicating its role in biological properties of
tyrosol. Most stable geometry of tyrosol (Figure 1(b)) con-
tains 37 occupied and 73 unoccupied virtual molecular orbi-
tals, respectively out of 110 alpha molecular orbitals.
Comparisons between simulated and experimentally
obtained bond lengths and bond angles data have been
summarized in Table S1. These data were statistically ana-
lyzed for its mathematical validation, and the values of cor-
relation coefficient (R) for bond lengths and bond angles
were found to be 0.988 and 0.984, respectively, in curve fit-
ting analysis (Figure 2(b), (c)). Commenting on these findings,
we inferred that simulated values for structural parameters
lie statistically closed to the experimentally obtained results
and outcome in these cases to be worthy in respect to their
statistical significance.
3.4. ADMET analysis
Most of the lead molecules often turn out to be unsuccessful
before clinical solicitation due to their poor pharmacokinetic
properties (Darvas et al., 2002). Therefore, it is extremely sig-
nificant to design and develop such kind of lead molecules
which can offer better gut absorption, safely reach the target
site and should not transform into toxic metabolite before it
reaches to the desired site of action followed by its safe
elimination prior to its accumulation inside the body result-
ing in harmful side effects. Evaluation of pharmacokinetic
factors in initial stages of designing and development of
novel lead candidate using computational methods is exten-
sively prevalent nowadays (Gleeson, Hersey, & Hannongbua,
2011; Hodgson, 2001; Kumar, Goel, Yadav, & Pruthi, 2017;
Lipinski, Lombardo, Dominy, & Feeney, 1997; Lombardo,
Gifford, & Shalaeva, 2003; Navia & Chaturvedi, 1996; Yadav
et al., 2018). ADMET calculation using QikProp program of
mentioned drugs along with tyrosol has been given in Table
3. The pharmacokinetic screening of tyrosol molecule was
found in the satisfactory parameters of Lipinski’s rule of five.
Octanol–water partition coefficient (QP logP(o/w)), conform-
ation-independent aqueous solubility (CIQP log S), and the
predicted aqueous solubility (QP log S), reported as critical
factors in determining absorption and distribution properties
of drug inside the body was found in the suitable range,
whereas cell permeability (QPPCaco), a vital feature influenc-
ing the access (to biological membranes) along with metab-
olism of drugs and the percentage human oral absorptions
values were observed in acceptable range. General rule of
thumb says, greater the bioactivity score higher will be the
possibility of the molecule to be active. Compounds possess-
ing bioactivity score >0.00 manifest substantial biological
actions, while score ranging from -0.50 to 0.00 were of mod-
erately active compounds, and score found below -0.50 were
assumed to be inactive (Verma, 2012). The predicted bioac-
tivities viz. as GPCR (G-protein coupled receptor), LI (ligand
gated ion channel modulator), KI (kinase inhibitor), NRL
(nuclear receptor ligand), PI (protease inhibitor), and EI
(enzyme inhibitor) computed by using MOLINSPIRATION
server and have been shown in Table 3.
3.5. Interaction study of the protein–ligand complex
All the reference compounds along with tyrosol are listed in
Table 4 with their docking score, Glide g-score, and Glide e-
model. These ligands are ranked based on their docking
score. The 2D receptor–ligand interactions of the mentioned
compounds are shown below in Figure 3, as protein–ligand
interaction diagram. Tyrosol shows docking score of
5.528 kcal/mole and forms hydrogen bond with backbone
of Met522 amino acid residue, polar interaction with Ser530,
hydrophobic interaction with Val349, Leu352, Ala527, Tyr348,
Tyr385, Phe381, Leu384, Gly526, Trp387, Phe518, and Val523
amino acid residues. Celecoxib has docking score of
7.145 kcal/mole and forms H-bond with backbone of
Phe518 and Leu352 amino acid residue, polar interaction
with Ser530, Ser353, Gln192 and His90, p–pinteraction with
positively charged residue Arg120, positive charge interaction
with Arg513 and hydrophobic interaction with Val349,
Val523, Val116, Leu359, Leu384, Leu531, Ala527, Ala516,
Tyr348, Tyr355, Tyr385, Phe381, Gly526, Trp387, and Met522.
Aspirin, the top scoring compound exhibits docking score
7.499 kcal/mol and form two hydrogen bonds with side
chains of Ser530 and Tyr385 amino acid residues, polar inter-
action with positively charged Ser530 residue, hydrophobic inter-
actions with residues Val349, Tyr348, Phe205, Leu384, Phe381,
Met522, Leu352, Ala527, Trp387, Phe518, Val523 and Gly526.
Ibuprofen shows the value of docking score as 6.636 kcal/mole
and it forms two H-bonds; one with side chain of Tyr355, while
other with side chain of Arg120. Polar interactions were also
noticed with Ser530 as well as Ser353, and hydrophobic interac-
tions with Leu359, Ala527, Val349, Leu384, Trp381, Gly526,
Table 2. Molecular properties calculated for tyrosol at DFT/B3LYP/6-311G
basis set.
Parameters (unit) Computed values
Total energy (eV) 12556.2849
Dipole moment (Debye) 2.4798
E
HOMO
(eV) 6.0851
E
LUMO
(eV) 0.3505
E
HOMO
–E
LUMO
5.7343
Chemical potential (eV) 3.2175
Chemical hardness (eV) 3.2175
Chemical softness (eV)
1
0.1554
Electronegativity (eV) 2.7898
Electrophilic index (eV) 1.2095
6 T. C. YADAV ET AL.
Tyr385, Met522, Phe518, Leu352, Val523, Leu531, and Val116
amino acid residues. Naproxen shows 7.352 kcal/mol docking
score and form H-bond with side chain of Ser530, polar inter-
action with Ser353, hydrophobic interactions with Tyr355,
Leu359, Leu531 Val349, Val523, Ala527, Trp583, Tyr385, Trp548,
Phe381, Met522, Leu384, Phe518, Leu352, Val116, Val525, Gly526
amino acid residues and positive charge interaction with residue
Arg120. We have also calculated the change in the torsion angles
for tyrosol in free state as well as complex with COX-2 (Table S4)
and noticed almost same values, indicating its structure stability
during the interaction process.
3.6. Dynamics simulations study
A variety of fundamental biological features including con-
formational behavior of proteins (Balasco, Barone, &
Vitagliano, 2015), unraveling of key structural insights as well
as designing and development of novel molecules (Arfeen,
Patel, Khan, & Bharatam, 2015) can be studied by using
molecular dynamic simulations via atomic-level perturbation.
A molecular dynamics time step consists of a computation-
ally intensive force calculation for each atom of a chemical
system, following less intensive integration step which aids
in movement and positioning of atoms according to classical
Newtonian laws of motion. Herein, we are displaying in-
depth analysis of results observed during MD simulation to
evaluate the stability of tyrosol with 5F1A receptor (Aspirin-
Acetylated Human COX-2, Docking score = 5.528, GLIDE
g-score = 5.528 and GLIDE e-model = 6.940) complex
obtained from docking. Desmond is an extremely efficient
high performance MD simulation package available to carry
out simulations for bio-molecular systems employing parallel
algorithms and numerical techniques to accomplish accuracy.
3.6.1. Simulation quality plot analysis
During the course of 20 ns simulation process, system behav-
ior of protein–ligand complex was mainly observed via tem-
perature (T), pressure (P), volume (V), potential energy (E_p),
and total energy (E) plots as shown in Figure 5. Volume of
system ranges from 590,000 to 575,000 nm
3
and average
volume of the system was maintained 576239.398 nm
3
throughout the simulation. P(1000 to þ300 bar) and T
(300 K) of the system were also nearly constant during the
simulation (Figure 4). The behavior of potential energy (E_p)
is consistent (190,500 to 192,500 kcal/mol), while total
energy (E) increases suddenly from -170,000 to -155,000 kcal/
mol and then gradually becomes constant as the time of the
simulation increases till it reaches -150,000 kcal/mol. Plot ana-
lysis infers that the protein ligand complex shows consistent
Figure 2. FMOs diagram (a) correlation graph between calculated and experimental bond lengths, (b) and bond angles, (c) for tyrosol.
JOURNAL OF BIOMOLECULAR STRUCTURE AND DYNAMICS 7
behavior into the system, and can carry out the inhibition of
human COX-2 by tyrosol.
3.6.2. RMSD and RMSF analysis
Functional veracity of protein–ligand complex is governed by
their folding and movement of backbone within the system.
RMSD analysis of receptor offers insight of the structural
deviations ensuing in Ca, backbone, and side chain residues
of the target receptor throughout the 20 ns simulation
(Shivakumar et al., 2010). Calculation of protein_C-alpha,
backbone, side-chain, and ligand_RMSD was carried out dur-
ing the progression of simulation trajectory of 20 ns in order
to compute the mean variation in shift of a group of atoms
for a specific frame with respect to a reference frame. The
MD analysis of 5F1A (COX-2) receptor with ligand (Aspirin,
Ibuprofen, Naproxen, Celecoxib, and Tyrosol) complexes dur-
ing simulation process was monitored via RMSD profiles as
calculated from the atomic fluctuations from MD trajectories.
All receptor frames were first superimposed on the reference
frames of 5F1A receptor’s C-alpha atoms and side chain
atoms followed by RMSD calculation on the basis of selected
atoms as depicted in Figure 5. The RMSD of C-alpha atoms
with Aspirin (mean - 1.920; median - 2.030), Ibuprofen (mean
- 1.583; median - 1.611), Naproxen (mean - 2.211; median -
2.226), Celecoxib (mean - 2.207; median - 2.107), and Tyrosol
(mean - 1.678; median - 1.878) were computed. The RMSD of
backbone atoms with Aspirin (mean - 1.948; median - 2.050),
Ibuprofen (mean - 1.619; median - 1.647), Naproxen (mean -
2.238; median - 2.252), Celecoxib (mean - 2.072; median -
2.159), and Tyrosol (mean -1.878; median - 1.902) were
calculated. The RMSD of side chain atoms with Aspirin (mean
- 2.889; median - 3.040), Ibuprofen (mean - 2.536; median -
2.601), Naproxen (mean - 3.145; median - 3.180), Celecoxib
(mean - 3.014; median - 3.102), and Tyrosol (mean - 2.758;
median - 2.798) were also calculated. The ligand RMSD of
Aspirin (mean - 0.293; median - 0.295), Ibuprofen (mean -
0.383; median - 0.388), Naproxen (mean - 0.160; median -
0.154), Celecoxib (mean - 0.755; median - 0.796), and Tyrosol
(mean - 0.141; median - 0.138) showed a significant observa-
tions. Fluctuation in RMSD of the ligand signifies the stability
of ligand with respect to the receptor between its active site
cleft. The RMSD plot of the 5F1A-tyrosol complex depicted in
Figure 5 manifests early conformational oscillations occurring
within the receptor complex system till 15 ns, which later on
became stabilized and converged afterwards in production
phase. RMSD backbone and RMSD_Caplot indicate steady
behavior, and also shows deviations initially from 1 to 10 Å
and further converged in the succeeding 10 ns signifying
Table 3. ADMET profiling calculated for tyrosol, aspirin, naproxen, ibuprofen, and celecoxib.
ADME descriptors Aspirin Ibuprofen Naproxen Tyrosol Celecoxib Range
MW 190.239 214.347 242.358 144.213 381.372 500
SASA 400.502 488.846 489.334 370.974 621.969 300–1000
FOSA 287.161 399.128 405.817 265.809 89.007 0–750
FISA 113.341 89.718 83.517 105.165 151.466 7–330
Volume 667.639 840.079 857.861 589.253 1079.981 500–2000
Donor HB 3 2 2 2 2 0–6
Acceptor HB 6.8 3.4 5.1 3.4 5.5 2–20
Glob 0.922405 0.880792 0.892286 0.916268 0.818478 0.75–0.95
QP polrz 16.712 23.613 24.967 14.86 38.233 13–70
QPlogPC16 6.159 7.174 7.374 5.001 10.758 4–18
QPlogPoct 12.584 11.784 13.134 9.063 19.79 8–35
QPlogPw 11.222 6.46 8.315 6.723 11.941 4–45
QPlogPo/w 0.04 2.336 1.884 0.635 3.346 2–6.5
CIQPlogS 0.831 2.157 2.124 0.935 5.728 6.5–0.5
QPlogKHSA 0.796 0.051 0.198 0.558 0.37 1.5–1.5
QPlogHERG 2.788 3.461 3.275 2.887 5.693 10–(5)
QPPCaco 833.853 1396.712 1599.228 496.823 362.702 25–500
QPlogBB 0.686 0.569 0.437 0.513 0.758 3.0–1.2
QPPMDCK 406.497 709.894 821.775 493.011 799.513 <25 poor
>500 great
QPlogKp 3.032 2.597 2.579 3.074 3.215 8.0–0.1
% Human Oral Absorption 78.991 96.912 95.321 84.333 92.349 >80 high
<25 low
Tumorigenic Yes No No No No NA
Mutagenic Yes Yes Yes No No NA
Irritant No No No No No NA
Reproductive effect Yes Yes Yes No No NA
Drug-likeness 1.63 0.08 0.36 4.13 8.10 NA
Drug score 0.48 0.31 0.26 0.32 0.36 NA
GPCR ligand 0.76 0.17 0.11 0.80 0.06 NA
Ion channel modulator 0.32 0.01 0.06 0.13 0.27 NA
Kinase inhibitor 1.06 0.72 0.38 0.87 0.01 NA
Nuclear receptor ligand 0.44 0.05 0.14 0.65 0.28 NA
Protease inhibitor 0.82 0.21 0.26 0.97 0.06 NA
Enzyme inhibitor 0.28 0.12 0.15 0.21 0.17 NA
Table 4. Docking and GLIDE score of tested ligands with COX-2 (PDB
ID: 5F1A).
Ligand name Docking score GLIDE g-score GLIDE e-model
Tyrosol 5.528 5.528 6.940
Asprin 7.499 7.499 0.863
Naproxen 7.352 7.352 32.263
Celecoxib 7.145 7.146 47.630
Ibuprofen 6.636 6.366 14.466
8 T. C. YADAV ET AL.
that the binding affinity and ligands position were main-
tained within receptor pocket (Figure 5). Similarly, big fluctu-
ation in RMSD of ligand plot implies away movement of
ligand from the receptor pocket during simulation, but
herein, the ligand illustrates a consistent and steady perform-
ance throughout entire 20 ns simulation process (range
Figure 3. Molecular docking interactions of (a) Aspirin (b) Ibuprofen (c) Naproxen (d) Celecoxib, and (e) Tyrosol with COX-2 (PDB ID 5F1A).
JOURNAL OF BIOMOLECULAR STRUCTURE AND DYNAMICS 9
0.00–0.274 Å; mean 0.141; median 0.138), signifying the
5F1A-Tyrosol complex stability within the system. While
higher variation in RMSD plot of side chain has been
observed initially from 1 to 15 Å and after 15 ns got con-
verged became stable showing consistent behavior steadily
till the termination of simulation process.
RMSF parameter determines the structural and conform-
ational variations occurring locally along the receptor. The
RMSF is used to analyze the deviation of residues of the
protein with respect to the reference during the entire simu-
lation. As shown in Figure 6, The RMSF calculation of 5F1A
(COX-2) with Aspirin, Ibuprofen, Naproxen, Celecoxib com-
plexes depicts higher fluctuation in the side chain amino
acid residues as compared to Tyrosol. Consequently, the fluc-
tuation of core binding site amino acid residue of receptor
with tyrosol belonging to active site cavity viz., Ser530,
Tyr385, Leu359, Ala527, Val349, Leu384, Trp381, Gly526,
Tyr385, Met522, Phe518, Leu352, Val523, Leu531, Val116 and
Figure 4. Simulation quality analysis plot of volume, temperature, pressure, potential energy, and total energy during the 20 ns simulations.
Figure 5. RMSD plots of ligand, C-alpha, backbone, and side chain for 20 ns.
10 T. C. YADAV ET AL.
Arg120 are intacted residues, and showed the consistency
throughout the 20 ns simulation. The stability trend for the
receptor-ligand complexes were found in decreasing order;
Tyrosol >Aspirin >Celecoxib >Ibuprofen >Naproxen.
Active site amino acid residues manifest comparatively low
side chain and backbone fluctuations as compared to other
residues. The small range RMSFs and RMSDs reflects insignifi-
cant structural rearrangements in the docking complex dur-
ing 20 ns simulation.
3.6.3. Energy measurement and variation analysis
The observed average total energy for 5F1A (COX-2) complex
with compounds aspirin, ibuprofen, naproxen, celecoxib and
tyrosol were -153,463 kcal/mol, -153,578 kcal/mol,
-153,444 kcal/mol, -151,008 kcal/mol, and 152,670 kcal/mol
along with a standard deviation of 1472.459 kcal/mol,
1437.294 kcal/mol, 1481.000 kcal/mol, 1254.944 kcal/mol, and
1462.00 kcal/mol, respectively, during the whole simulation
process. Correspondingly, the recorded average potential
energy (a vital factor in simulation analysis) was
-192,252 kcal/mol for aspirin, -192,319 kcal/mol for ibuprofen,
-192,296 kcal/mol for naproxen, -188,428 kcal/mol for cele-
coxib and 192,276 kcal/mol for tyrosol with a standard
deviation of 242.469 kcal/mol, 222.039 kcal/mol, 246.311 kcal/
mol, 232.739 kcal/mol and 258.799 kcal/mol, respectively, dur-
ing the whole simulation process. From the graphs, it is evi-
dent that the system 5F1A-Tyrosol was more stable as
compared to other systems and remains stable during the
20 ns simulation process as there were not any significant
variations happening in the conformation of the protein–li-
gand complex as shown in Figure 7. Potential energy E_p
(kcal/mol) plot of the ligand-complex with respect to
time (ps).
3.6.4. Interaction fraction plot of 5F1A-tyrosol complex
during simulation
5F1A receptor interactions with the ligand tyrosol were
observed during the course of 20 ns simulation process and
a stacked bar chart is also plotted for the same that displays
the interaction fraction of the amino acid residues compris-
ing active site of 5F1A receptor as shown in Figure 8. Amino
acid residues Leu352, Leu384, Tyr385, Trp387, and Ser530
hydrophobic groove of COX channel of 5F1A receptor are
involved in the regulation of the activity of the protein. Out
of them, Leu384, Tyr385, Trp387, and Ser530 amino acid resi-
dues were found to be more promising residues for the cata-
lytic activity of this receptor as these have interaction
fraction values of 0.9, 0.8, 0.95 and 0.8, respectively, and they
also were involved in the formation of H-bonds with ligand
after 20 ns simulation process. Leu352, Phe518, and Val523
amino acid residues were also observed in the creation of
hydrophobic interaction and salt bridge, respectively, with
the ligand after 20 ns simulation process.
3.6.5. Ligand property calculation during 20 ns simulation
To elucidate the stability of tyrosol in the 5F1A receptor, five
properties were examined during the course of 20 ns simula-
tion: (1) Ligand RMSD—it is root mean square deviation of a
ligand with respect to reference conformation (normally, first
frame is considered as reference at time t¼0); (2) rGyr
(radius of gyration) calculates the ligand extendedness, corre-
sponding to its principal moment of inertia; (3) MolSA
(molecular surface area)—computation of molecular surface
with a probe of radius 1.4 Å; (5) SASA (solvent accessible sur-
face area)—it is considered as surface area accessible to
water molecules; (6) PSA (polar surface area); defined as the
solvent accessible surface area of the molecule attributed
Figure 6. RMSF plot of amino acid residues C-alpha (red color) and backbone (blue color) of COX-2 protein with respect to fluctuation (Å) during 20 ns simulation.
JOURNAL OF BIOMOLECULAR STRUCTURE AND DYNAMICS 11
due to O
2
and N
2
atoms. From the plot, it is a clear evident
that RMSD of ligand remains constant during the whole
simulation process. Overall ligand RMSD was observed below
0.5 Å. In the initial stages of simulation, it shows slight fluctu-
ation from 0.00 to 0.274 Å with an average of 0.141 Å, but
later on maintains a constant RMSD throughout the simula-
tion process. This investigation manifests that the complex
acquires a stable conformation during the entire simulation
process. Throughout the 20 ns simulation, radius of gyration
of ligand ranges from 2.25 to 2.55 Å. MolSA, SASA, and from
the prior PSA plots exhibited the steady behavior of the lig-
and during complete process of simulation (Figure 9).
3.6.6. Post-simulation analysis of 5F1A-tyrosol complex
Following MD simulation, the post simulation analyses of the
receptor (5F1A) as well as ligand (Tyrosol) interactions were
carried out (Figure 10). Before MD simulation, only Meth522
residue was found to be interacting via H-bond formation,
polar interaction with Ser530 and hydrophobic interactions
with amino acid residues like Tyr348, Leu352, Tyr385, Trp387,
Phe518 and Val523, etc. Post-simulation analysis revealed
better conformational stability of ligand within active site of
5F1A receptor as the hydrophobic groove was occupied by
the hydroxy-ethyl moiety of ligand coordinated (H-bond)
predominantly by Leu384 and Trp387 via side chain and
backbone respectively, and a hydrophobic interaction with
Leu352. Phenol moiety of ligand interacts via side chain
H-bond with Tyr385 and Ser530. From the plot, it is clear
that the conformational integrity of the complex was greatly
improved after MD simulation as the ligand was kept tightly
intact within hydrophobic groove of COX channel of tar-
get receptor.
Figure 7. Potential energy (kcal/mol) plots of ligands-COX-2 complex with respect to time (ps).
Figure 8. Histogram for tyrosol binding with COX-2.
12 T. C. YADAV ET AL.
Figure 9. Variation in the ligand properties with respect to time during the course of 20 ns simulation.
Figure 10. Post simulation interaction diagram of 5F1A-tyrosol after 20 ns simulation.
JOURNAL OF BIOMOLECULAR STRUCTURE AND DYNAMICS 13
4. Conclusions
Stimulation of COX-2 enzyme results in production of prosta-
glandins in swollen as well as erythematous tissues, and
therefore, has been acknowledged as a significant target to
design and develop more selective anti-inflammatory agents.
The use of conventional prescription including NSAIDs and
selective inhibitor such as celecoxib, accounts for the treat-
ment of inflammation, have been reported to be accompany-
ing several unusual side effects, which chiefly includes
cardiovascular and GI (gastro-intestinal) toxicity. Even though,
the availability of structural information is amass. The charac-
teristics of the drug-enzyme interaction that are responsible
for determining the COX-2 selectivity of drugs are not always
clear in several condition. No uninterrupted ligand–protein
interactions have been observed with amino acid residues
that are exclusive to COX-2. Henceforth, despite of remark-
able advancement carried out in development of newer anti-
inflammatory drugs, designing safer and cost-effective drugs
to treat inflammatory processes is still a crucial challenge.
Since tyrosol is obtained from plant source possessing anti-
oxidant properties with proven potential as anti-cancerous as
well as cell rejuvenation which is under oxidative stress. It
can be used alone or in combination with other marketed
drugs as anti-inflammatory substance possessing minimum
side effects and contraindications compared with other state
of the art prescribed NSAIDs. The model developed here
comprises of geometry optimization at DFT/B3LYP/6-311G
(d,p) basis set where structural constraints like bond lengths
and angles have been calculated and found in accord with
the X-ray crystallography findings. Calculations for electronic
properties and HOMO!LUMO energies were also carried
out. Calculated ADMET parameters for tyrosol confirmed no
risk, and can be further studied to design and develop safe
and selective COX-2inhibitors. The HOMO!LUMO analysis
manifests maximum percentage of charge transfer, support
our study that it holds the ligand molecules strongly within
the receptor. The molecular docking also confirms the bind-
ing of tyrosol in active site of COX-2, followed by molecular
dynamics simulation, and post-simulation analysis unraveled
the intricate interactions as well as strong affinity due to the
increase in hydrophilic interactions by the formation of H-
bonds with key residues viz. Leu384, Tyr385, Trp387, Met522
and Ser530 accompanied with hydrophobic interactions with
Tyr348, Val349, Leu352, Phe381, Leu384, Phe518, Val523,
Gly526, and Ala527 to stabilize the receptor–ligand complex.
Undoubtedly, ample research remains to be carried out to
evaluate the binding mechanism as well as the designing
and development of specific COX-2 inhibitors. Our investiga-
tion signifies state-of-the-art way of performing rational drug
design since facts apprehended by static experiments such
as X-ray and in silico inference can now be aided by informa-
tion resulting from the transitory protein-ligand interac-
tions involved.
Acknowledgements
The authors are grateful to IIT Roorkee, IIT Indore, and IIIT Allahabad for
instrumentation facilities and research environment. TCY and NK are
thankful to MHRD and UGC (award letter no. 201718-PDFSS-2017-18-
UTT-17095) for SRF and Postdoctoral fellowship, respectively. Authors
are highly thankful to Dr. Nidhi Goel (Department of Chemistry, BHU
Varanasi) for solving the crystal structure of tyrosol. Tara Chand Yadav
and Naresh Kumar contributed equally in this research article.
Disclosure statement
No potential conflict of interest was reported by the authors.
ORCID
Tara Chand Yadav http://orcid.org/0000-0003-4941-1946
Naresh Kumar http://orcid.org/0000-0002-0551-3302
References
Adhami, H. R., Zehl, M., Dangl, C., Dorfmeister, D., Stadler, M., Urban, E.,
…Krenn, L. (2015). Preparative isolation of oleocanthal, tyrosol, and
hydroxytyrosol from olive oil by HPCCC. Food Chemistry,170,
154–159.
Ahn, E. Y., Jiang, Y., Zhang, Y., Son, E. M., You, S., Kang, S. W., …Kim,
D. K. (2008). Cytotoxicity of p-tyrosol and its derivatives may correlate
with the inhibition of DNA replication initiation. Oncology Reports,
19(2), 527–534.
Arfeen, M., Patel, R., Khan, T., & Bharatam, P. V. (2015). Molecular dynam-
ics simulation studies of GSK-3bATP competitive inhibitors:
Understanding the factors contributing to selectivity. Journal of
Biomolecular Structure & Dynamics,33(12), 2578–2593.
Balasco, N., Barone, D., & Vitagliano, L. (2015). Structural conversion of
the transformer protein RfaH: New insights derived from protein
structure prediction and molecular dynamics simulations. Journal of
Biomolecular Structure & Dynamics,33(10), 2173–2179.
Becke, A. D. (1993). Density functional thermochemistry. III. The role of
exact exchange. Journal of Chemical Physics.,98(7), 5648–5652. doi:
10.1063/1.464913
Bertelli, A. A., Migliori, M., Panichi, V., Longoni, B., Origlia, N., Ferretti, A.,
…Giovannini, L. (2002). Oxidative stress and inflammatory reaction
modulation by white wine. Annals of the New York Academy of
Sciences,957, 295–301.
Beutler, J. A. (2009). Natural products as a foundation for drug discovery.
Current Protocols in Pharmacology,46, 9.11.1–9.11.21.
Bowers, K. J., Chow, E., Xu, H., Dror, R. O., Eastwood, M. P., Gregersen, B.,
& Shaw, D. E. (2006). Scalable algorithms for molecular dynamics sim-
ulations on commodity clusters. In SC 2006 Conference, Proceedings of
the ACM/IEEE (p. 43). Tampa, FL: IEEE.
Brandenburg, K. (1999). DIAMOND: Visual Crystal Structure Information
System. Version 2.1c. Bonn, Germany: Crystal Impact GbR.
Brenes, M., Garc
ıa, A., Garc
ıa, P., & Garrido, A. (2000). Rapid and complete
extraction of phenols from olive oil and determination by means of a
coulometric electrode array system. Journal of Agricultural and Food
Chemistry,48(11), 5178–5183.
Bu, Y., Rho, S., Kim, J., Kim, M. Y., Lee, D. H., Kim, S. Y., …Kim, H. (2007).
Neuroprotective effect of tyrosol on transient focal cerebral ischemia
in rats. Neuroscience Letters,414(3), 218–221.
Carluccio, M. A., Siculella, L., Ancora, M. A., Massaro, M., Scoditti, E.,
Storelli, C., …De Caterina, R. (2003). Olive oil and red wine antioxi-
dant polyphenols inhibit endothelial activation: Antiatherogenic prop-
erties of Mediterranean diet phytochemicals. Arteriosclerosis,
Thrombosis, and Vascular Biology,23(4), 622–629.
Chandramohan, R., Pari, L., Rathinam, A., & Sheikh, B. A. (2015). Tyrosol, a
phenolic compound, ameliorates hyperglycemia by regulating key
enzymes of carbohydrate metabolism in streptozotocin induced dia-
betic rats. Chemico-Biological Interactions,229,44–54.
Cheki, M., Yahyapour, R., Farhood, B., Rezaeyan, A., Shabeeb, D., Amini,
P., …Najafi, M. (2018). COX-2 in radiotherapy: A potential target for
14 T. C. YADAV ET AL.
radioprotection and radiosensitization. Current Molecular
Pharmacology,11(3), 173–183.
Chinery, R., Beauchamp, R. D., Shyr, Y., Kirkland, S. C., Coffey, R. J., &
Morrow, J. D. (1998). Antioxidants reduce cyclooxygenase-2 expres-
sion, prostaglandin production, and proliferation in colorectal cancer
cells. Cancer Research,58(11), 2323–2327.
Darvas, F., Keseru, G., Papp, A., Dorm
an, G., Urge, L., & Krajcsi, P. (2002).
In silico and ex silico ADME approaches for drug discovery. Current
Topics in Medicinal Chemistry,2(12), 1287–1304.
Dhanjal, J. K., Sreenidhi, A. K., Bafna, K., Katiyar, S. P., Goyal, S., Grover,
A., & Sundar, D. (2015). Computational structure-based de novo
design of hypothetical inhibitors against the anti-inflammatory target
COX-2. PLoS One,10(8), e0134691.
Dinnella, C., Minichino, P., D’Andrea, A. M., & Monteleone, E. (2007).
Bioaccessibility and antioxidant activity stability of phenolic com-
pounds from extra-virgin olive oils during in vitro digestion. Journal
of Agricultural and Food Chemistry,55(21), 8423–8429. doi:10.1021/
jf072244þ
Dubois, R. N., Abramson, S. B., Crofford, L., Gupta, R. A., Simon, L. S., Van
De Putte, L. B., & Lipsky, P. E. (1998). Cyclooxygenase in biology and
disease. FASEB Journal: Official Publication of the Federation of
American Societies for Experimental Biology,12(12), 1063–1073.
Ferguson, L. R., Zhu, S-T., & Harris, P. J. (2005). Antioxidant and antigeno-
toxic effects of plant cell wall hydroxycinnamic acids in cultured HT-
29 cells. Molecular Nutrition & Food Research,49(6), 585–593.
Frisch, M. J. (2009). Gaussian 09, Revision A.1. Wallingford, CT.
Fuggetta, M. P., Cottarelli, A., Lanzilli, G., Tricarico, M., & Bernini, R.
(2012). In vitro antitumor activity of olive oil tyrosol and hydroxytyro-
sol and their methyl carbonate derivatives. Medicinal and Aromatic
Plant Science and Biotechnology,6,25–30.
Funk, C. D., Funk, L. B., Kennedy, M. E., Pong, A. S., & Fitzgerald, G. A.
(1991). Human platelet/erythroleukemia cell prostaglandin G/H syn-
thase: cDNA cloning, expression, and gene chromosomal assignment.
FASEB Journal: Official Publication of the Federation of American
Societies for Experimental Biology,5(9), 2304–2312.
Giovannini, C., Straface, E., Modesti, D., Coni, E., Cantafora, A., De
Vincenzi, M., …Masella, R. (1999). Tyrosol, the major olive oil biophe-
nol, protects against oxidized-LDL-induced injury in Caco-2 cells. The
Journal of Nutrition,129(7), 1269–1277.
Gleeson, M. P., Hersey, A., & Hannongbua, S. (2011). In-silico ADME mod-
els: A general assessment of their utility in drug discovery applica-
tions. Current Topics in Medicinal Chemistry,11(4), 358–381.
Goel, N., & Kumar, N. (2017). A stable nonanuclear Tb(III) cluster for
selective sensing of picric acid. Inorganica Chimica Acta,463,14–19.
doi:10.1016/j.ica.2017.04.010
Goel, N., & Kumar, N. (2018). A dual-functional luminescent Tb(iii) metal–-
organic framework for the selective sensing of acetone and TNP in
water. RSC Advances,8(20), 10746–10755. doi:10.1039/C7RA13494K
Guo, Z., Mohanty, U., Noehre, J., Sawyer, T. K., Sherman, W., & Krilov, G.
(2010). Probing the alpha-helical structural stability of stapled p53
peptides: Molecular dynamics simulations and analysis. Chemical
Biology & Drug Design,75(4), 348–359.
Halgren, T. A. (2009). Identifying and characterizing binding sites and
assessing druggability. Journal of Chemical Information and Modeling,
49(2), 377–389.
Halgren, T. A., Murphy, R. B., Friesner, R. A., Beard, H. S., Frye, L. L.,
Pollard, W. T., & Banks, J. L. (2004). Glide: A new approach for rapid,
accurate docking and scoring. 2. Enrichment factors in database
screening. Journal of Medicinal Chemistry,47(7), 1750–1759.
Harvey, A. L. (2008). Natural products in drug discovery. Drug Discovery
Today,13(19–20), 894–901.
Hempel, S. L., Monick, M. M., & Hunninghake, G. W. (1994).
Lipopolysaccharide induces prostaglandin H synthase-2 protein and
mRNA in human alveolar macrophages and blood monocytes. The
Journal of Clinical Investigation,93(1), 391–396.
Hockridge, M. R., Knight, S. M., Robertson, E. G., Simons, J. P., McCombie,
J., & Walker, M. (1999). Conformational landscapes in flexible organic
molecules: 4-Hydroxy phenyl ethanol (p-tyrosol) and its singly
hydrated complex. Physical Chemistry Chemical Physics,1(3), 407–413.
doi:10.1039/a808701f
Hodgson, J. (2001). ADMET-turning chemicals into drugs. Nature
Biotechnology,19(8), 722–726.
Jorgensen, W. L., Chandrasekhar, J., Madura, J. D., Impey, R. W., & Klein,
M. L. (1983). Comparison of simple potential functions for simulating
liquid water. Journal of Chemical Physics,79(2), 926–935. doi:10.1063/
1.445869
Kelley, D. J., Mestre, J. R., Subbaramaiah, K., Sacks, P. G., Schantz, S. P.,
Tanabe, T., …Dannenberg, A. J. (1997). Benzo[a]pyrene up-regulates
cyclooxygenase-2 gene expression in oral epithelial cells.
Carcinogenesis,18(4), 795–799. doi:10.1093/carcin/18.4.795
Klaus, B. (1999). DIAMOND version 1.2c. Bonn, Germany: University of
Bonn.
Koehn, F. E., & Carter, G. T. (2005). The evolving role of natural products
in drug discovery. Nature Reviews. Drug Discovery,4(3), 206–220.
Kontou, N., Psaltopoulou, T., Panagiotakos, D., Dimopoulos, M. A., &
Linos, A. (2011). The mediterranean diet in cancer prevention: A
review. Journal of Medicinal Food,14(10), 1065–1078.
Koopmans, T. (1934). About the allocation of wave functions and eigen-
values of the individual electrons one atom. Physica,1(1–6), 104–113.
Kumar, N., & Garg, A. (2014). Structural optimization and docking studies
of anatoxin-a: A potent neurotoxin. African Journal of Biotechnology,
13(30), 3092–3100. doi:10.5897/AJB2014.13671
Kumar, N., Goel, N., Yadav, T. C., & Pruthi, V. (2017). Quantum chemical,
ADMET and molecular docking studies of ferulic acid amide deriva-
tives with a novel anti-cancer drug target. Medicinal Chemistry
Research,26(8), 1822–1834. doi:10.1007/s00044-017-1893-y
Kumar, N., Gupta, S., Yadav, T. C., Pruthi, V., Varadwaj, P. K., & Goel, N.
(2018). Extrapolation of phenolic compounds as multi-target agents
against cancer and inflammation. Journal of Biomolecular Structure
and Dynamics, 1. doi:10.1080/07391102.2018.1481457
Kumar, N., Kumar, S., Abbat, S., Nikhil, K., Sondhi, S. M., Bharatam, P. V.,
…Pruthi, V. (2016). Ferulic acid amide derivatives as anticancer and
antioxidant agents: Synthesis, thermal, biological and computational
studies. Medicinal Chemistry Research,25(6), 1175–1192. doi:10.1007/
s00044-016-1562-6
Kumar, N., & Pruthi, V. (2014). Potential applications of ferulic acid from
natural sources. Biotechnology Reports (Amsterdam)),4,86–93.
Kumar, N., & Pruthi, V. (2015). Structural elucidation and molecular dock-
ing of ferulic acid from Parthenium hysterophorus possessing COX-2
inhibition activity. 3 Biotech,5(4), 541–551.
Kumar, N., Pruthi, V., & Goel, N. (2015). Structural, thermal and quantum
chemical studies of p-coumaric and caffeic acids. Journal of Molecular
Structure,1085, 242–248. doi:10.1016/j.molstruc.2014.12.064
Lipinski, C. A., Lombardo, F., Dominy, B. W., & Feeney, P. J. (1997).
Experimental and computational approaches to estimate solubility
and permeability in drug discovery and development settings.
Advanced Drug Delivery Reviews,23(1–3), 3–25. doi:10.1016/S0169-
409X(96)00423-1
Lombardo, F., Gifford, E., & Shalaeva, M. Y. (2003). In silico ADME predic-
tion: Data, models, facts and myths. Mini Reviews in Medicinal
Chemistry,3(8), 861–875.
Loru, D., Incani, A., Deiana, M., Corona, G., Atzeri, A., Melis, M. P., …
Dessi, M. A. (2009). Protective effect of hydroxytyrosol and tyrosol
against oxidative stress in kidney cells. Toxicology and Industrial
Health,25(4–5), 301–310.
Lucido, M. J., Orlando, B. J., Vecchio, A. J., & Malkowski, M. G. (2016).
Crystal structure of aspirin-acetylated human cyclooxygenase-2:
Insight into the formation of products with reversed stereochemistry.
Biochem,55(8), 1226–1238. doi:10.1021/acs.biochem.5b01378
Manach, C., Williamson, G., Morand, C., Scalbert, A., & R
em
esy, C. (2005).
Bioavailability and bioefficacy of polyphenols in humans. I. Review of
97 bioavailability studies. The American Journal of Clinical Nutrition,
81(1 Suppl), 230S–242S.
Mestre, J. R., Chan, G., Zhang, F., Yang, E. K., Sacks, P. G., Boyle, J. O., …
Dannenberg, A. J. (1999). Inhibition of cyclooxygenase-2 expression.
An approach to preventing head and neck cancer. Annals of the New
York Academy of Sciences,889,62–71.
Montedoro, G., Servili, M., Baldioli, M., & Miniati, E. (1992). Simple and
hydrolyzable phenolic compounds in virgin olive oil. 1. Their extrac-
tion, separation, and quantitative and semiquantitative evaluation by
JOURNAL OF BIOMOLECULAR STRUCTURE AND DYNAMICS 15
HPLC. Journal of Agricultural and Food Chemistry,40(9), 1571–1576.
doi:10.1021/jf00021a019
Navia, M. A., & Chaturvedi, P. R. (1996). Design principles for orally bio-
available drugs. Drug Discovery Today,1(5), 179–189. doi:10.1016/
1359-6446(96)10020-9
Nile, S. H., Ko, E. Y., Kim, D. H., & Keum, Y. S. (2016). Screening of ferulic
acid related compounds as inhibitors of xanthine oxidase and cyclo-
oxygenase-2 with anti-inflammatory activity. Brazilian J
Pharmacognosy,26(1), 50–55. doi:10.1016/j.bjp.2015.08.013
Ou, S., Li, Y., & Gao, K. (1999). A study on scavenging activity of wheat
bran dietary fiber for free radical. Acta Nutrimenta Sinica,21, 191–194.
Pang, L. Y., Hurst, E. A., & Argyle, D. J. (2016). Cyclooxygenase-2: A role
in cancer stem cell survival and repopulation of cancer cells during
therapy. Stem Cells International,2016,1–11. doi:10.1155/2016/
2048731
Panossian, A., Nikoyan, N., Ohanyan, N., Hovhannisyan, A., Abrahamyan,
H., Gabrielyan, E., & Wikman, G. (2008). Comparative study of
Rhodiola preparations on behavioral despair of rats. Phytomedicine:
International Journal of Phytotherapy and Phytopharmacology,15(1–2),
84–91.
Panossian, A., Wikman, G., Kaur, P., & Asea, A. (2009). Adaptogens exert a
stress-protective effect by modulation of expression of molecular chaper-
ones. Phytomedicine,16(6–7), 617–622. doi:10.1016/j.phymed.2008.12.003
Pearson, R. G. (1988). Absolute electronegativity and hardness:
Application to inorganic chemistry. Inorganic Chemistry,27(4),
734–740. doi:10.1021/ic00277a030
Petroni, A., Blasevich, M., Salami, M., Papini, N., Montedoro, G. F., & Galli,
C. (1995). Inhibition of platelet aggregation and eicosanoid produc-
tion by phenolic components of olive oil. Thrombosis Research,78(2),
151–160.
Puerta, R., Mart
ınez Dom
ınguez, M. E., Ru
ız-Gut
ıerrez, V., Flavill, J. A., &
Hoult, J. R. (2001). Effects of virgin olive oil phenolics on scavenging
of reactive nitrogen species and upon nitrergic neurotransmission.
Life Sciences,69(10), 1213–1222.
Raj, U., Kumar, H., Gupta, S., & Varadwaj, P. K. (2015). Novel DOT1L recep-
tor natural inhibitors involved in mixed lineage leukemia: a virtual
screening, molecular docking and dynamics simulation study. Asian
Pacific Journal of Cancer Prevention: APJCP,16(9), 3817–3825.
Romero, C., & Brenes, M. (2012). Analysis of total contents of hydroxytyr-
osol and tyrosol in olive oils. Journal of Agricultural and Food
Chemistry,60(36), 9017–9022.
SAINT version 7.34 (2004). Software for the integration of CCD detector
system. Madison, WI: Bruker Analytical X-ray Systems.
Salucci, S., Burattini, S., Battistelli, M., Buontempo, F., Canonico, B.,
Martelli, A. M., …Falcieri, E. (2015). Tyrosol prevents apoptosis in
irradiated keratinocytes. Journal of Dermatological Science,80(1),
61–68.
Sastry, G. M., Adzhigirey, M., Day, T., Annabhimoju, R., & Sherman, W.
(2013). Protein and ligand preparation: Parameters, protocols, and
influence on virtual screening enrichments. Journal of Computer-Aided
Molecular Design,27(3), 221–234.
Shaw, D. E., Yibing, S., John, L. K., Michael, P. E., & Ron, O. D. (2005).
Gaussian split Ewald: A fast Ewald mesh method for molecular simula-
tion. Journal of Chemical Physics,122,1–13.
Sheldrick, G. M. (1990). Phase annealing in SHELX-90: Direct methods for
larger structures. Acta Crystallographica Section A Foundations of
Crystallography,A46, 467–473. doi:10.1107/S0108767390000277
Sheldrick, G. M. (1996). SADABS: Program for empirical absorption correc-
tion of area detector data. Gottingen, Germany: University of
Gottingen.
Sheldrick, G. M. (2000). SHELXTL-NT, Version 6.12, Reference manual.
Germany: University of Gottingen.
Shen, B. (2015). A new golden age of natural products drug discovery.
Cell,163(6), 1297–1300.
Shivakumar, D., Williams, J., Wu, Y., Damm, W., Shelley, J., & Sherman, W.
(2010). Prediction of absolute solvation free energies using molecular
dynamics free energy perturbation and the OPLS force field. Journal
of Chemical Theory and Computation,6(5), 1509–1519. doi:10.1021/
ct900587b
Thibault, S. L. C., Arseneault, M., Longpre, F., & Ramassamy, C. (2011).
Tyrosol and hydroxytyrosol, two main components of olive oil, pro-
tect N2a cells against amyloid-b-induced toxicity. Involvement of the
NF-kB signaling. Current Alzheimer Research,8, 543–551. doi:10.2174/
156720511796391845
Toshihiro, A., Ken, Y., Miho, Y., Motohiko, U., Yumiko, K., Naoto, S., &
Koichi, A. (2000). Triterpene alcohol and sterol ferulates from rice
bran and their anti-inflammatory effects. Journal of Agricultural and
Food Chemistry,48, 2313–2319.
Tsuneda, T., & Hirao, K. (2014). Long-range correction for density func-
tional theory. Wiley Interdisciplinary Reviews: Computational Molecular
Science,4(4), 375–390. doi:10.1002/wcms.1178
Vasilescu, D., & Adrian-Scotto, M. (2010). From Democritus to
Schr€
odinger: A reflection on quantum molecular modeling. Structural
Chemistry,21(6), 1289–1314. doi:10.1007/s11224-010-9665-z
Verma, A. (2012). Lead finding from Phyllanthus debelis with hepatopro-
tective potentials. Asian Pacific Journal of Tropical Biomedicine,2,
1735–1737.
Yadav, T. C., Srivastava, A. K., Dey, A., Kumar, N., Raghuwanshi, N., &
Pruthi, V. (2018). Application of computational techniques to unravel
structure–function relationship and their role in therapeutic develop-
ment. Current Topics in Medicinal Chemistry,18(20), 1769–1791.
Zhang, G., & Musgrave, C. B. (2007). Comparison of DFT methods for
molecular orbital eigenvalue calculations. The Journal of Physical
Chemistry A,111(8), 1554–1561.
16 T. C. YADAV ET AL.
