Design, synthesis, biological evaluation and molecular dynamics studies of 4-thiazolinone derivatives as protein tyrosine phosphatase 1B (PTP1B) inhibitors

Article abstract of DOI:10.1080/07391102.2019.1664333

Protein tyrosine phosphatase 1B (PTP1B) is a key negative regulator of insulin signaling pathway, and more and more studies have shown that it is a potential target for the treatment of type 2 diabetes mellitus (T2DM). In this study, 17 new 4-thiazolinone derivatives were designed and synthesized as novel PTP1B inhibitors, and ADMET prediction confirmed that these compounds were to be drug-like. In vitro enzyme activity experiments were performed on these compounds, and it was found that a plurality of compounds had good inhibitory activity and high selectivity against PTP1B protein. Among them, compound 7p exhibited the best inhibitory activity with an IC₅₀ of 0.92 μM. The binding mode of compound 7p and PTP1B protein was explored, revealing the reason for its high efficiency. In addition, molecular dynamics simulations for the PTP1B^WT and PTP1B^comp#7p systems revealed the effects of compound 7p on PTP1B protein at the molecular level. In summary, the study reported for the first time that 4-thiazolinone derivatives as a novel PTP1B inhibitor had good inhibitory activity and selectivity for the treatment of T2DM, providing more options for the development of PTP1B inhibitors. AbbreviationsBBB blood-brain barrierCDC25B cell division cycle 25 homolog BCYP2D6 Cytochrome P450 2D6 bindingDCCM dynamic cross-correlation mapDS Discovery StudioH bond hydrogen bondHIA human intestinal absorptionLAR leukocyte antigen-related phosphataseMD molecular dynamicsMEG-2 maternal-effect germ-cell defective 2MM-PBSA molecular mechanics Poisson Boltzmann surface area)PCA principal component analysisPDB Protein Data BankpNPP p–nitrophenyl phosphatePPB plasma protein bindingPTP1B protein tyrosine phosphotase 1BRMSD root mean square deviationRMSF root mean square fluctuationSHP-1 src homologous phosphatase-1SHP-2 src homologous phosphatase-2SPC single-point chargeTCPTP T cell protein tyrosine phosphataseT2DM Type 2 diabetes mellitusVDW van der Waals Communicated by Ramaswamy H. Sarma.

Full text of DOI:10.1080/07391102.2019.1664333

Journal of Biomolecular Structure and Dynamics
ISSN: 0739-1102 (Print) 1538-0254 (Online) Journal homepage: https://www.tandfonline.com/loi/tbsd20
Design, synthesis, biological evaluation and
molecular dynamics studies of 4-thiazolinone
derivatives as Protein tyrosine phosphatase 1B
(PTP1B) inhibitors
Wen-Shan Liu, Rui-Rui Wang, Hai Yue, Zhi-Hui Zheng, Xin-Hua Lu, Shu-Qing
Wang, Wei-Li Dong & Run-Ling Wang
To cite this article: Wen-Shan Liu, Rui-Rui Wang, Hai Yue, Zhi-Hui Zheng, Xin-Hua Lu,
Shu-Qing Wang, Wei-Li Dong & Run-Ling Wang (2019): Design, synthesis, biological
evaluation and molecular dynamics studies of 4-thiazolinone derivatives as Protein tyrosine
phosphatase 1B (PTP1B) inhibitors, Journal of Biomolecular Structure and Dynamics, DOI:
10.1080/07391102.2019.1664333
To link to this article: https://doi.org/10.1080/07391102.2019.1664333
View supplementary material
Accepted author version posted online: 06
Sep 2019.
Submit your article to this journal
View related articles
View Crossmark data
Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=tbsd20

Design, synthesis, biological evaluation and molecular dynamics studies
of 4-thiazolinone derivatives as Protein tyrosine phosphatase 1B (PTP1B)
inhibitors
a, 1
a,1
b
c
c
a
Wen-Shan Liu , Rui-Rui Wang , Hai Yue , Zhi-Hui Zheng , Xin-Hua Lu , Shu-Qing Wang ,
a, *
a, *
Wei-Li Dong , Run-Ling Wang
a
Tianjin Key Laboratory on Technologies Enabling Development of Clinical Therapeutics and
Diagnostics (Theranostics), School of Pharmacy, Tianjin Medical University, Tianjin 300070,
China.
b
Inner Mongolia Institute for Drug Control, Huhhot, Inner Mongolia 010000, China.
c
New Drug Research & Development Center of North China Pharmaceutical Group Corporation,
National Microbial Medicine Engineering & Research Center, Hebei Industry Microbial
Metabolic Engineering & Technology Research Center, Key Laboratory for New Drug Screening
Technology of Shijiazhuang City, Shijiazhuang 050015, Hebei, China.
1
These authors contributed equally to this work.
*
Correspondence to: Wei-Li Dong Email: dongweili@tmu.edu.cn
Run-Ling Wang Email: wangrunling@tmu.edu.cn
Funding information
Contract grant sponsor: National Natural Science Foundation of China; Contract grant
number: 81773569.
Contract grant sponsor: National Natural Science Foundation of China; Contract grant
number: 21772146.
Contract grant sponsor: Natural Science Foundation of Tianjin; Contract grant number:
18JCYBJC89800.

Abstract
Protein tyrosine phosphatase 1B (PTP1B) is a key negative regulator of insulin signaling pathway,
and more and more studies have shown that it is a potential target for the treatment of type 2
diabetes mellitus (T2DM). In this study, 17 new 4-thiazolinone derivatives were designed and
synthesized as novel PTP1B inhibitors, and ADMET prediction confirmed that these compounds
were to be drug-like. In vitro enzyme activity experiments were performed on these compounds,
and it was found that a plurality of compounds had good inhibitory activity and high selectivity
against PTP1B protein. Among them, compound 7p exhibited the best inhibitory activity with an
IC of 0.92 μM. The binding mode of compound 7p and PTP1B protein was explored, revealing
5
0
WT
the reason for its high efficiency. In addition, molecular dynamics simulations for the PTP1B
comp#7p
and PTP1B
systems revealed the effects of compound 7p on PTP1B protein at the molecular
level. In summary, the study reported for the first time that 4-thiazolinone derivatives as a novel
PTP1B inhibitor had good inhibitory activity and selectivity for the treatment of T2DM, providing
more options for the development of PTP1B inhibitors.
Key words: T2DM, PTP1B inhibitor, activity, molecular docking, molecular dynamics
simulation
Abbreviations
BBB (blood-brain barrier), CDC25B (cell division cycle 25 homolog B), CYP2D6 (Cytochrome
P450 2D6 binding), DCCM (dynamic cross-correlation map), DS (Discovery Studio), H bond
(hydrogen bond), HIA (human intestinal absorption), LAR (leukocyte antigen-related
phosphatase), MD (molecular dynamics), MEG- 2 (maternal-effect germ-cell defective 2),
MM-PBSA (molecular mechanics Poisson Boltzmann surface area), PCA (principal component
analysis), PDB (Protein Data Bank), pNPP (p–nitrophenyl phosphate), PPB (plasma protein
binding), PTP1B (protein tyrosine phosphotase 1B), RMSD (root mean square deviation), RMSF
(
root mean square fluctuation), SHP-1 (src homologous phosphatase-1), SHP-2 (src homologous
phosphatase-2), SPC (single-point charge), TCPTP (T cell protein tyrosine phosphatase), T2DM
Type 2 diabetes mellitus), VDW (van der Waals).
(
1
. Introduction
Type 2 diabetes mellitus (T2DM), one of the most common chronic diseases associated with
numerous complications, is characterized by insulin resistance(Sun, Zhuang, et al., 2017; Tadic &
Cuspidi, 2015). In 2015, more than 415 million people worldwide have diabetes, of which T2DM
affects 90-95% of the total population with diabetes, and its prevalence is estimated to exceed 10%
of the global adult population by 2040(Fernandez-Silva, Alonso-Gonzalez, Gonzalez-Perez,
Gestal-Otero, & Diaz-Gravalos, 2019; Sun, Zhang, et al., 2017). Therefore, T2DM has become a
serious crisis threatening human health, and it is urgent to find effective strategies to treat this
disease.
Usually, the abnormality of insulin signal transduction is closely related to the pathogenesis of
T2DM(X. Chen, Gan, Feng, Liu, & Zhang, 2019). Protein tyrosine phosphatase 1B (PTP1B) is a
member of the non-transmembrane phosphotyrosine phosphatase family, and studies have
demonstrated that PTP1B is a key negative regulator in the insulin signaling pathway. PTP1B is

widely expressed in insulin-sensitive peripheral tissues and negatively regulates insulin signal
transduction by dephosphorylating IRS on tyrosine residues(Byon, Kusari, & Kusari, 1998; Cheng
et al., 2002; Goldstein, Bittner-Kowalczyk, White, & Harbeck, 2000; Zabolotny et al., 2002).
Studies have shown that PTP1B-knockout mice show increased insulin sensitivity, improved
glucose tolerance and resistance to diet-induced obesity(Elchebly et al., 1999; Klaman et al.,
2000). Therefore, PTP1B is a potential new drug target for the treatment of T2DM.
In recent years, inhibitors designed against PTP1B had emerged in an endless stream. The
ertiprotafib developed by American Home Products was the first drug candidate to enter Phase II
clinical trials(Wang et al., 2015). ISIS-113715, an antisense inhibitor of the PTP1B gene, was the
second to reach phase II clinical trials(G. Liu, 2004). Unfortunately, these inhibitors were later
proven to fail. These failed cases tell us that the development of clinically relevant PTP1B
inhibitors still faces significant challenges and is often plagued by its poor selectivity and
unfavorable pharmacokinetics(Eldehna et al., 2019). Therefore, it is increasingly important to
develop PTP1B inhibitors with high selectivity and good pharmacokinetic characteristics.
Virtual screening has proven to be a very effective method for discovering new drugs with the
desired properties and structural diversity(Neves et al., 2018). In this study, the scaffold
ZINC99459 was found to have a good docking score and fitting value for the PTP1B protein by
virtual screening method based on docking. Subsequently, based on the scaffold ZINC99459, the
De novo design of the compounds was carried out to design 17 novel 4-thiazolinone derivatives.
These compounds had confirmed to be drug-like by ADMET prediction. The designed compounds
were subsequently synthesized. In vitro enzyme activity experiments were performed on PTP1B
and other phosphatases, and compound 7p as the most potent selective PTP1B inhibitor was
identified. Furthermore, the binding mode of the compound 7p and PTP1B protein was
investigated. The conformation difference of PTP1B protein after binding with compound 7p was
further explored at the molecular level by molecular dynamics (MD) simulation. Specifically, 300
comp#7P
ns molecular dynamics simulations were performed on the protein-ligand complex (PTP1B
)
WT
and protein system (PTP1B ), respectively. A series of post-dynamics analyses were then
explored the conformational differences between the two systems, including root mean square
deviation (RMSD), root mean square fluctuation (RMSF), principal component analysis (PCA),
and dynamic cross-correlation mapping (DCCM). In conclusion, the 4-thiazolinone derivatives
were investigated as PTP1B selective inhibitors, and it was hoped to provide more clues for the
development of more efficient PTP1B inhibitors.
2
. Materials and methods
2
.1. Compound design
2.1.1. Virtual screening based on molecular docking
Virtual screening was a very effective method for developing new drugs with the required
characteristics and structural diversity(Jiang et al., 2019; Neves et al., 2018). In this study, virtual
screening based on molecular docking was used for drug design by Discovery Studio (DS) 3.5
software. Firstly, the crystal structure of PTP1B protein (PDB ID: 2VEY) with the catalytic
domain was downloaded from the protein data bank (PDB) and used as the receptor protein(Douty
et al., 2008). The ligands for docking were retrieved from the ZINC database. Then, "Prepare

Ligands and Prepare Protein" modules were implemented for protein and ligand preparation. The
binding pocket was then defined by the "Define and Edit Binding Site" tool, which produced a
sphere around the receptor protein's own ligand. Finally, the retrieved ligands were docked into the
receptor protein. Ligand with higher CDOKER score and fitted value was selected for further De
nova design.
2.1.2. De Novo Design
Ludi was a widely used re-designed algorithm for developing new potential candidate
compounds(Bohm, 1992). In our study, the De novo design method(Schneider & Fechner, 2005)
was used to design new inhibitors by means of the Ludi algorithm. Firstly, the "De novo library
generation protocol" was used to generate the fragment library for the De novo design. Secondly,
the "De novo receptor protocol" in Discovery Studio v3.5 was used to define the binding pocket of
the receptor. Based on the scaffold structure, new derivatives were developed in the binding
pocket of the receptor. Finally, through the Ludi algorithm, the appropriate fragments were fitted
to the scaffold obtained by virtual screening, producing a set of new derivatives with higher
scores.
2
.2. In silico ADMET prediction
The ADMET (absorption, distribution, metabolism, excretion, and toxicity) studies were
performed using DS v3.5 software to assess the pharmacokinetic properties of the compounds.
Some important ADMET descriptors include human intestinal absorption (HIA)(Miyake et al.,
2
017), blood-brain barrier (BBB)(Egan, Walters, & Murcko, 2002), plasma protein binding
PPB)(Wesson & Eisenberg, 1992), aqueous solubility(Cheng & Merz, 2003), Cytochrome P450
D6 binding (CYP2D6)(Susnow & Dixon, 2003) and toxicity.
(
2
2
.3. Chemistry
Reagents and solvents were obtained from commercial suppliers and could be used directly
without further purification. Analytical thin layer chromatography (TLC) was performed on the
TLC plate (silica gel 60 F254 and aluminum foil). All compounds were detected by using UV
light (wavelength: 254 nm or 365 nm). The Isolation and purification of compound was carried
1
13
out by flash column chromatography on silica gel 60 (300-400 mesh). H (400 MHz) and
C
NMR (100 MHz) spectra were recorded in CDCl or DMSO-d by the spectrometer at room
3
6
temperature. Chemical shifts were reported in parts per million (ppm) and coupling constants (J)
were expressed in Hz. The characteristics of the signal were s (singlet), d (doublet), dd (doublet of
doublets), t (triplet) and m (multiplet). Mass spectra were measured on an Agilent 1100 series.
2
.4. In vitro phosphatase activity assay
Human recombinant PTP1B, SHP2, SHP1, LAR, TCPTP, MEG-2 or CDC25B was expressed in E
coli and purifed by Ni-NTA affinity chromotagraphy in our laboratory. The enzyme activity was
measured using p–nitrophenyl phosphate (pNPP) as substrate in a 96-well plate. Briefly, purified
recombinant PTP1B or other Phosphatases (0.05 μg) in 50 l buffer containing 50mM citrate (pH
6.0), 0.1M NaCl, 1mM Ethylenediaminetetraacetic acid (EDTA), and 1mM dithiothreitol (DTT)
and test compounds were added to each well of a 96-well plate. After preincubation for 15 minutes
at room temperature, 50 μl of reaction buffer containing 2mM pNPP was added and incubated at

37°C for 30 min. The enzyme activity was measured by detecting the absorbance at 405 nm for
the amount of produced p-nitrophenol.
2
.5. Protein and ligand preparation
The crystal structure of the PTP1B protein was obtained from the Protein Data Bank (PDB) with
entries 2VEY. In order to optimize the initial structure, DS v3.5 software package was used to
repair the missing residues of PTP1B protein by the protocol of "Fill loops". The repaired PTP1B
protein was then prepared by the Prepare Protein module, including removing water, assigning
bond order, adding the hydrogen atoms, treat metals and disulfides. In addition, the structure of
compound 7p was obtained by ChemDraw 8.0 and converted into MOL2 formation by Chem 3D.
Subsequently, compound 7p was prepared by the "prepare ligand" module of DS v3.5, including
keeping ionization, desalt and generating tautomer.
2
.6. Molecular docking
Molecular docking study was performed by using the DS 3.5 software. Firstly, the preparation of
proteins and the energy of the ligands was minimized by the CHARMm force field(Zhai et al.,
2
019). Subsequently, the receptor binding pocket was defined by the coordinate space of the
extracted co-crystallized ligand(Jin, Ma, Li, Li, & Wang, 2018; Zhang, Luan, Chou, & Johnson,
002). The binding pocket of a protein was generally defined by a residue having at least one
2
heavy atom, and the distance between this heavy atom and the heavy atom of the ligand was 5
Å(Ma, Wang, Xu, Wang, & Chou, 2012). Finally, the docking module in the DS v3.5 program was
performed to dock the ligands into the protein. The conformation with the highest docking score
was used to analyze the interaction between the protein and the ligand.
2
.7. Molecular Dynamics Simulation
WT
comp#7p
MD simulations for PTP1B and PTP1B
systems were performed using the GROMACS
4.5.5 software package with CHAMM27 force field(Pol-Fachin, Fernandes, & Verli, 2009). First,
a file called "topology" was generated that contained simulation information about the nonbonding
parameters (atom types and charges) and the bonding parameters (bonds, angles and dihedrals).
Then, the system was simulated by incorporating space-filling dodecahedron boxes and filled with
explicit single-point charge (SPC) water molecules (the number of water molecules was 3400).
The distance from the surface of the protein to the edge of the box was limited to at least 1 nm and
the system is neutralized by adding 6 sodium ions. Subsequently, the steepest descent method was
used to ensure that there were no steric clash and inappropriate geometry in the system, and the
energy of the system was minimized to 1000 kJ· mol-1 / nm(Tang et al., 2018). At the same time,
the system got a better balance by the 100 ps position restrained equilibration dynamics simulation,
namely NVT and NPT canonical ensemble. Where, N was part number, P was system pressure, V
was volume, and T was temperature. During the NVT simulation, the system was heated from 0 to
300 K with the thermostat and stabilized at 300 K. During the NPT simulation, the system
pressure was maintained at 1 bar by using a constant pressure device. In addition, all
hydrogen-bonded atoms in the system were suitable for the LINCS algorithm(Hess, 2008). Finally,
300 ns MD simulation was performed.
2
.8. Principal Component Analysis (PCA)

PCA was a statistical technique that has been widely used to understand the conformational
differences of protein system(Balmith & Soliman, 2017). This method was based on building a
covariance matrix of complex sets of variables and was used to reduce the dimensionality of the
data to gain valuable information from the protein system throughout the molecular dynamics
simulation(Maisuradze, Liwo, & Scheraga, 2009; Yesudhas et al., 2016). The ensemble formula
used to calculate the covariance matrix C for each pair of Cα atoms i and j was based on the
ij
following equation(Zhou, Zhang, Chen, Zhao, & Zhong, 2016):
C =〈( x -〈x 〉) (x -〈x 〉)〉(i, j = 1, 2, 3,..., N)
ij
i
i
j
j
Here, x and x were the ith and jth Cartesian coordinates of Cα atoms, respectively.〈x 〉and〈x 〉
i
j
i
j
were the time average over all the configurations gained in MD simulation, and N was the number
of carbon atoms. The eigenvectors of the matrix were taken as the principal component (PC),
which indicated the projections of trajectory on the principal mode. And, the corresponding
eigenvalues represented the direction of PC. Usually, the first few principal components were used
to describe the overall motion of the system(Fakhar et al., 2017). The PCA scatter plot was
generated by Bio3d software(Grant, Rodrigues, ElSawy, McCammon, & Caves, 2006).
2
.9. Dynamic Cross-Correlation Map (DCCM) Analysis
To detect the dynamic correlation among different protein domains, DCCM analysis was
WT
performed to analyze the correlation matrix across all Cα atoms of PTP1B
system and
comp#7p
PTP1B
system(Xu, Kong, Zhu, Sun, & Chang, 2016). Cross-correlation coefficient C for
ij
the pair of each Cα atoms i and j was calculated the following equation(Ndagi, Mhlongo, &
Soliman, 2017):
2
i
2
j
1/2
C = 〈Δr • Δr 〉/ 〔〈 Δr 〉〈 Δr 〉〕
ij
i
j
Here, Δr (Δr ) represented the displacement vector of the ith (jth) atom of the systems, and i and j
i
j
represented the i-th and j-th atom, and〈.......〉indicated ensemble average. The value of C was
ij
from -1 to 1. Positive values indicated positively correlated movement (the same direction), and
negative values indicated anti-correlated movement (the opposite direction)(Li et al., 2014). The
higher the absolute value of C , the greater the correlation (or inverse correlation) between the two
ij
residues(Olde Scheper, Meredith, Mansvelder, van Pelt, & van Ooyen, 2017).
3
. Results and discussion
3
.1. Compound design
The scaffold ZINC99459 (Figure 1) was found by virtual screening with a higher docking score
and fitted value. The searched fragments were fitted to the scaffold ZINC99459 by the De nova
design method, and a series of new compounds 7a-q were obtained based on the docking scores
and fitted value.

Figure 1. Design strategy of the target compounds.
3
.2. In silico ADMET prediction
ADMET prediction was considered a key step for drug candidates(Thiyagarajan, Lin, Chang, &
Weng, 2016). The ADMET properties of all design compounds 7a-q were predicted and compared
by using DS v3.5. For the aqueous solubility level: 0 (extremely low); 1 (very low, but possible), 2
(low), 3 (good); for the PPB level: 0 meant that the binding level was below 90%, 1 meant that the
binding level was above 90%, and 2 meant that the binding level was above 95%; for the BBB
level: 0 (very good), 1 (good), 2 (moderate), 3 (poor), 4 (undefined); for the CYP2D6 Probability
level: Less than 0.5 meant that it was unlikely to inhibit CYP2D6 (Non-CYP2D6 inhibitors), and
above 0.5 meant that it was likely to inhibit CYP2D6 enzyme (CYP2D6 Inhibitors); for the
hepatotoxic prediction, true represented non-hepatotoxic and for HIA level: 0 (good), 1 (moderate),
2
(poor), 3 (very poor). The ADMET prediction results of these compounds were listed in Tables
S1 and S2. According to the predicted results, these 17 compounds were confirmed to be drug-like
and had low side effects.
3
.3. Chemistry
A mixture of benzyl chloride (1) (30 mmol), hydroxybenzaldehyde derivatives 2a-b (40 mmol),
and anhydrous potassium carbonate (K CO ) (90 mmol) in N,N-Dimethylformamide (DMF) (80
2
3
mL). The completion of the reaction was detected by TLC analysis. The mixture was extracted
with ethyl acetate (EtOAc) (3 × 60 mL) and washed with 5% brine solution (3 × 60 mL). The
combined organic layers were dried with anhydrous sodium sulphate (NaSO ). The obtained
4
residue was purified by silica gel column chromatography (PE : EA = 4 : 1) to afford Intermediate
2a-i.
The p-aminobenzoic acid derivatives 4a-b (50 mmol) was added to absolute ethanol (EtOH) (100
mL) to form a white suspension. Under ice bath conditions, thionyl chloride (175 mmol) was
slowly added dropwise to the suspension. After the dropwise addition was completed, the
suspension was stirred at room temperature for 30 min, and then heated to reflux for 4 h. After
completion of the reaction, the volatile solvent was concentrated under reduced pressure to obtain
white solid. The solid was dissolved in EtOAc (150 mL) to obtain a white suspension. Under ice
bath, 25% Sodium hydroxide (NaOH) aqueous solution (60 mL) was slowly added and stirred for
30 min. The white suspension gradually became clear solution with pH of from 7 to 8. After
standing, the aqueous layer was removed, and the organic layer was washed with saturated salt
solution (30 × 60 mL), and dried over anhydrous NaSO . The solvent was removed under reduced
4
pressure to obtain intermediates 5a-b.
The ethyl p-aminobenzoate derivatives 5a-b (50 mmol) were dissolved in tetrahydrofuran (THF)

(300 mL) and benzaldehyde (100 mmol) and dicyclohexylcarbodiimide (DCC) (60 mmol).
Mercapto acetic acid (100 mmol) was slowly added dropwise to the mixture under ice bath. After
the dropwise addition was completed, the reaction was stirred at room temperature for 3 h. After
the reaction was completed, the white solid in the mixture was removed by filtration. The volatile
solvent was then removed under reduced pressure to obtain white solid. EtOAc (350 mL) was
added to white solid to dissolve, and then the organic layer was washed successively with 5%
citric acid aqueous solution (3 × 70 mL), distilled water (3 × 70 mL), 5% sodium bicarbonate
(
NaHCO ) aqueous sodium (3 × 70 mL) and saturated aqueous sodium chloride solution (3 × 70
3
mL). Finally, the white solid was recrystallized from absolute EtOH to obtain intermediates 6a-b.
Sodium metal (4 mmol) was cut into small sodium particles and added to absolute EtOH (20 mL)
to form an alcohol solution of sodium ethoxide (EtONa). Intermediates 6a-b (4 mmol) and
benzaldehyde derivatives (4 mmol) were added to absolute EtOH (30 mL) and stirred to dissolve
completely. Then, alcohol solution of EtONa was slowly added dropwise, and the mixture was
heated under reflux for 30 min. After completion of the reaction, the solvent was evaporated under
reduced pressure, and the residue obtained was was purified by silica gel chromatography on silica
gel (PE : EA = 5 : 1) to obtain the final products 7a-q.
Scheme 1. Synthetic scheme for the synthesis of compounds 7a-q. Reagents and conditions: (a)
o
o
o
K CO , DMF, 55 C, 3 h; (b) SOCl , EtOH, 0 C ~ 80 C, 4.5h; (c) Mercapto acetic acid, DCC, THF,
2
o
3
2
o
0
C ~ rt, 4 h; (d) Sodium, EtOH, 80 C, 0.5 h.

3
.4. Biological evaluation
Table 1 listed the inhibitory activities of 17 4-thiazolinone derivatives against PTP1B and other
phosphatases. It could be seen from the table that multiple compounds exhibited good inhibitory
activity against PTP1B protein with IC50 values ranging from 0.92 to 9.64 μM. Among them,
compound 7p showed the most potent inhibitory activity with an IC50 value of 0.92 μM.
Subsequently, study of the structure-activity relationship (SAR) was carried out.
When n = 0, different R substituents had large effects on the inhibitory activity. Specifically, when
there was no substitution on the substituent R, the compound 7a exhibited an inhibitory activity
with an IC of 3.23 μM. When the substituent R was a 2-position substitution, the activity of the
50
2
-position Cl substitution (7c, IC = 4.02 μM) was significantly higher than that of the 2-position
50
F substitution (7b, IC50 = 9.64 μM). When the substituent R was a 3-position substitution, the
activity of the 3-position CH substitution (7e, IC = 2.09 μM) was significantly higher than that
3
50
of the other 3-position substituents (7d, 7f and 7g). When the substituent R was a 4-position
substitution, the activity of the 4-position CH substitution (7g, IC = 1.39 μM) was significantly
3
50
higher than that of the other 4-position substituents (7h and 7i). Moreover, the activity of the
-position CH substitution was higher than that of the 3-position CH substitution (7g vs 7e),
4
3
3
indicating that the 4-position CH substitution was a dominant substituent. When n = 1, its SAR
3
was similar to that of n = 0. It was found that the activity of the 4-position CH substitution (7p,
3
IC50 = 0.92 μM) was significantly higher than that of the other substituents. Furthermore, by
comparing the values of n, it was found that the introduction of CH (n = 1) in the structure could
2
significantly increase the inhibitory activity (7p vs 7g) with a 1.5-fold increase in activity.
As shown in Table 1, we also explored the selectivity of these 17 compounds for other
phosphatases, including homogeneous T cell protein tyrosine phosphatase (TCPTP), src
homologous phosphatase-2 (SHP-2), src homologous phosphatase-1 (SHP-1), cell division cycle
2
5 homolog B (CDC25B), leukocyte antigen-related phosphatase (LAR) and maternal-effect
germ-cell defective 2 (MEG- 2). We had focused on the selectivity of compounds 7e, 7g, 7m and
p with good inhibitory activity against other phosphatases. It could be seen from the table that
7
these four compounds showed more than 15-fold greater selectivity for PTP1B than for TCPTP,
more than 12-fold selectivity for PTP1B over SHP-1, more than 18-fold selectivity for PTP1B
over CDC25B, and more than 10-fold selectivity for PTP1B over MEG-2. In addition, these four
compounds had almost no activity on SHP-2 and LAR, both with IC50 values over 100 μM. In
short, after our efforts, we had developed multiple compounds that had good inhibitory activity
and high selectivity for PTP1B protein.

Table 1. Inhibitory activity of compound 7a-q on PTP1B and other phosphatases.
7a-p
Compd.
ZINC-9
n
-
R
PTP1B TCPTP SHP2
SHP1 CDC25
LAR MEG-2
IC50
IC50
(μM)
NA
IC50
(μM)
NA
IC50
(μM)
NA
B IC50
(μM)
NA
IC50
(μM)
NA
IC50
(μM)
NA
(μM)
-
50.24
9459
7
7
7
7
7
a
b
c
d
e
0
0
0
0
0
0
0
H
2-F
3.23
9.64
4.02
3.97
2.09
4.05
2.88
>100
>100
9.05
>100
>100
>100
>100
>100
>100
0.96
3.60
3.08
>100
>100
>100
>100
39.03
>100
>100
>100
>100
>100
>100
>100
>100
>100
4.16
4.34
2-Cl
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
3-Cl
>100
30.21
13.20
8.63
3-CH3
3-OCH3
3-OCH2
Ph
7
f
7g
7
h
0
0
0
1
1
1
1
1
1
1
4-F
7.51
4.89
1.39
3.22
8.05
2.40
3.03
6.26
0.92
5.61
>100
>100
>100
24.75
>100
>100
29.31
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
62.85
>100
>100
4.74
2.97
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
6.87
12.92
14.85
>100
6.03
7
7
i
j
4-Cl
4-CH3
H
16.69
>100
2.71
7k
7
l
2-F
7m
3-Cl
>100
>100
2.53
>100
>100
32.35
>100
>100
7
n
o
p
q
3-OCH3
4-F
7
7
7
4-CH3
4-OCH2
Ph
22.23
1.14
The “NA” indicated that the compound was not tested.
3
.5. Study of the binding mode
Molecular docking was used to simulate the binding mode between compound 7p and PTP1B
protein and to understand their interactions in depth(J. Liu, Wang, Ma, Wang, & Wang, 2011). The
PTP1B protein had two binding sites: a catalytically active site (CYS215-ARG221) and a second
aromatic binding site (ARG24, ARG254, MET258, GLY259, GLN262 and GLN266)(Y. Y. Chen
et al., 2008). As shown in Figure 2A, compound 7p with a higher inhibitory activity was well
docked to the active pocket of PTP1B with the help of DS 3.5 software. The top-ranked docking
score for compound 7p and PTP1B protein was 45.1248. As shown in Figure 2B, the interaction of
compound 7p and PTP1B was clearly shown. Among them, the compound 7p formed hydrogen

bond (H bond) interactions with residues ASP181, CYS215 and GLN262, formed van der Waals
VDW) interactions with a plurality of residues, formed a Sigma-Pi interaction with residue
(
TYR46, and formed a cation-Pi interaction with the residue LYS120. Therefore, the compound 7p
formed stable interactions with PTP1B, further revealing the reason why the compound had high
inhibitory activity against PTP1B protein.
Figure 2. The binding mode of compound 7p and PTP1B protein. (A) The docking pocket of
compound 7p and PTP1B. (B) 2D diagram of the interaction of compound 7p and PTP1B. Here,
the H bond interactions with residues are denoted by green dotted arrow pointing to the electron
donor; the H bond interactions with residues are denoted by blue dashed arrow pointing to the
electron donor; the Sigma-Pi and cation-Pi interactions are represented by the orange line; the pink
rectangles represent residues with charge or H bond interactions; the green rectangles represent
residues with VDW interactions.
3
.6. Molecular dynamics study
3.6.1 Investigation on the stabilities of two simulated systems
To investigate the stability and conformational differences of the PTP1B receptor induced by the
binding with compound 7p, we carried out molecular dynamics simulations of the PTP1B receptor
WT
comp#7p
structures in two systems (PTP1B
and PTP1B
). The conformational stability and
convergence of two systems during MD simulations was evaluated by calculating the root mean
square deviation (RMSD) of the backbone atoms(Wan, Hu, Tian, & Chang, 2013). Figure 3
reflected the trend of RMSD in the two systems relative to the crystal structure. System stability
WT
and convergence with maximum fluctuation of 0.314 nm between 0 and 300 ns in PTP1B
comp#7p
system were observed, whereas in the PTP1B
system, the highest fluctuation was 0.205 nm
at 10 ns, after which the value of RMSD fluctuated around 0.220 nm. However, after
approximately 10 ns, the RMSD trajectories converged and the fluctuation rested below 0.300 nm
for all the systems. In general, the RMSD plot from 300 ns MD trajectories showed no remarkable
difference for the three systems after the system convergence. However, the average RMSD of
WT
comp#7p
0
.2764 nm and 0.2413 nm was observed in PTP1B and PTP1B
systems, respectively,
comp#7p
which indicated that the PTP1B
system maintained the stability of the conformation to the
greatest extent during the dynamics simulation(Su et al., 2018). These results confirmed that both
systems reached equilibrium after 10 ns, so the 11-300 ns trajectories were intercepted for the
remaining post-dynamic analysis.

Figure 3. RMSD of all backbone atoms of two systems. The black line represents the RMSD of
WT
comp#7p
the PTP1B system. The red line means the RMSD of the PTP1B
system. The average
values of RMSD are represented as bar graph of the upper left corner.
3.6.2. Evaluation on the flexibility of side chain residues
Residues were component of protein that determined its conformational features. Interactions such
ligand-residue interaction in the active site might induce conformational changes in protein
structure and alter its function(J. Chen, Wang, Pang, Zhang, & Zhu, 2019). More specifically, the
conformational changes occur as a result of ligand-induced motion during ligand binding(Loeffler
&
Winn, 2013). Understanding ligand-induced conformational changes in the protein structure are
critical to structure-based rational drug design. RMSF was a measure of average atomic mobility
of backbone atoms (N, Cα, and C) during MD simulations(Ndagi, Mhlongo, & Soliman, 2018). To
understand and explore the structural dynamics that took place upon the ligand binding, RMSF of
the subject systems was calculated from the 300 ns MD trajectories and the plot is presented in
WT
Figure 4. It was reflected that the flexibility of most residues in the PTP1B system was similar
comp#7p
to that in the PTP1B
system. The greater the value of RMSF, the greater the flexibility of
residues. The differences in flexibility were observed in the catalytic active region (residues
CYS215-ARG221). Similarly, the region (residues ARG254-GLN266) also exhibited higher
WT
fluctuations in the PTP1B system. The RMSF values of the residues CYS215-ARG221 and
WT
comp#7p
ARG254-GLN266 in the PTP1B were 0.11 nm and 0.09 nm higher than that in PTP1B
respectively. Obviously, the compound 7p had reduced the flexibility in regions
CYS215-ARG221 and ARG254-GLN266), which might lead to the decrease in the catalytic
activity of the PTP1B protein.
,
(

WT
comp#7p
Figure 4. The RMSF of the side‐ chain atoms for PTP1B system and PTP1B
system. The
comp#7p
red line indicates the outcome for the PTP1B
system and the black line represents the
WT
outcome for the PTP1B system. Meanwhile, the region (residues CYS215-ARG221) and region
(residues ARG254-GLN266) are highlighted by the green frame and yellow frame, respectively.
3.6.3. The analysis for the conformational states of the two systems
In order to explore the conformational state of the two proteins during the MD simulation, PCA
was performed on all Cα atoms in the PC phase space. The conformational behavior of the
WT
comp#7p
PTP1B system and the PTP1B
system was obtained by projecting their trajectories into a
two-dimensional subspace spanning along first major PCs (PC1 and PC2)(W. S. Liu, Wang, Sun,
et al., 2019). From the Figure 5, the first two eigenvectors (PC1 and PC2) that occupied most of
the variance in the original distribution of the protein conformation space could be used to analyze
the conformational state of the two systems. PCA scatter plot showed two kinds of conformational
states in subspace. Blue dots represented unstable conformational states, red dots represented
stable conformational states, and white dots represent intermediate states between two
conformations. The protein system periodically switched between two conformational states (blue
and red)(W. S. Liu, Wang, Li, et al., 2019). From PCA results, it could be seen that the
conformational state of proteins had changed significantly between the two systems. Compared
WT
comp#7p
with PTP1B system (Figure 5A), the conformational state of PTP1B
system (Figure 5B)
was more concentrated, suggesting that the existence of ligands restricted the movement of
WT
proteins. From the 11-300 ns simulation trajectories, the top 20 PCs in the PTP1B system and
comp#7p
WT
PTP1B
system occupied 73% and 80.9% of the total variation, respectively. In PTP1B
system, the contributions of the first two PCs (PC1 and PC2) to the variance were 31.0% and
comp#7p
1
1.6%, respectively. In PTP1B
system, the contributions of the first two PCs to the variance
comp#7p
were 38.2% and 15.0%, respectively. It could be seen that PTP1B
system occupied smaller
WT
phase space and exhibited lower flexibility than PTP1B system.

Figure 5. Projections of trajectories onto the first two principal components (PC1 and PC2) in the
WT
comp#7p
phase space. (A) Projection for PTP1B
system. (B) Projection for PTP1B
system.
Different conformational states are represented by two different color points, here, red represents
the stable conformational state and blue represents the unstable conformational state.
3.6.4. The analysis for the correlated motions of the two systems
Dynamic cross-correlation map (DCCM) showed the overall landscape of the correlated motion
between residues during MD simulation(Ndagi et al., 2017). To investigate the difference in
WT
comp#7p
correlated motion between PTP1B system and PTP1B
system, the DCCM of all Cα atoms
during the entire simulations was analysis(Yan et al., 2019). The correlated motions of the Cα
atoms from the 11-300 ns simulated trajectories of two systems were displayed in Figure 6. The
different colors represented the different correlations. The strong negative correlated motions of
the residues ranged from 0 to -1 (from white to light blue to dark blue), and conversely, strong
positive correlated motions of the residues ranged from 0 to 1 (from white to light red to dark
red)(Machaba, Mhlongo, & Soliman, 2018). The darker the areas, the stronger the correlation
between the residues. By analyzing the DCCM, the correlated motions between the regions of the
PTP1B protein was significantly altered due to the binding of the compound 7p. It was found that

comp#7p
the overall correlated motion of residues in the PTP1B
system was significantly reduced
WT
compared to the PTP1B system. For better observation, the regions with significant differences
WT
in the correlated motion were highlighted by the black frames. In PTP1B system, the catalytic
active region (residues CYS215–AYS221) and regions (residues GLU8–TYR46 and PRO180–
comp#7p
PRO210) showed highly positive correlation. However, in PTP1B
system, the positive
correlation between the region (residues CYS215–ARG221) and regions (residues GLU8–TYR46
and PRO180–PRO210) was significantly attenuated. Therefore, the compound 7p resulted in a
significant decrease in the correlated motions of residues in the PTP1B protein, indicating that
comp#7p
WT
PTP1B
system had a higher stability in conformation than PTP1B system.
WT
Figure 6. A, the DCCM analyses of Cα atoms for PTP1B system. B, the DCCM analyses of Cα
comp#7p
atoms for PTP1B
system. Blue regions mean negative correlation while red regions mean
positive correlation, and the deeper the color, the stronger the correlation. The catalytic active
region is marked in black frames.
4
. Conclusion
PTP1B has attracted widespread attention as a target for the treatment of T2DM. In this study, 17
-thiazolinone derivatives were designed and synthesized as novel PTP1B inhibitors. ADMET
4
prediction confirmed that these compounds were potential to be drug-like. In vitro enzyme activity
experiment found multiple compounds with developmental value. Among them, compound 7p had
the best inhibitory activity with an IC50 of 0.92 μM. The binding mode of compound 7p and
PTP1B protein was explored by molecular docking. It was found that the stable interactions
between compound 7p and PTP1B protein was formed, revealing the underlying reason why
compound 7p had high inhibitory activity on PTP1B protein at the molecular level. Furthermore,
WT
comp#7p
3
00 ns molecular dynamics simulations were performed on both PTP1B and PTP1B
systems. A series of the post-dynamics analyses were performed on the simulated trajectories of
these two systems to investigate the effects of compound 7p on PTP1B protein. The stability of the
two systems was evaluated by RMSD, and it was found that both systems reached steady state
comp#7p
around 10 ns. The average RMSD value of the PTP1B
system was smaller than that of the
WT
PTP1B
system, indicating that the stability of the protein after binding to the ligand was
significantly enhanced. The flexibility of the two systems was evaluated by RMSF and it was

found that there were significant differences in the RMSF of the two systems at the catalytic active
region (residues CYS215-ARG221) and the region (residues ARG254-GLN266). It was found that
the binding of compound 7p and PTP1B protein caused significant decrease in the flexibility of
comp#7p
these two regions. Analyses of PCA and DCCM revealed that the PTP1B
system occupied a
smaller phase space and the correlated motions between the residues were significantly reduced,
indicating increase in protein rigidity after binding with the ligand. In summary, through this series
of studies, more clues are provided for further development of more efficient selective PTP1B
inhibitors.
CONFLICT OF INTEREST
The authors report no conflicts of interest in this work.
References
Balmith, M., & Soliman, M. E. (2017). Non-active site mutations disturb the loop dynamics,
dimerization, viral budding and egress of VP40 of the Ebola virus. Mol Biosyst, 13(3),
585-597. doi: 10.1039/c6mb00803h
Bohm, H. J. (1992). The computer program LUDI: a new method for the de novo design of
enzyme inhibitors. J Comput Aided Mol Des, 6(1), 61-78.
Byon, J. C., Kusari, A. B., & Kusari, J. (1998). Protein-tyrosine phosphatase-1B acts as a
negative regulator of insulin signal transduction. Mol Cell Biochem, 182(1-2), 101-108.
Chen, J., Wang, X., Pang, L., Zhang, J. Z. H., & Zhu, T. (2019). Effect of mutations on binding
of ligands to guanine riboswitch probed by free energy perturbation and molecular
dynamics
simulations.
Nucleic
Acids
Res,
47(13),
6618-6631.
doi:
10.1093/nar/gkz499
Chen, X., Gan, Q., Feng, C., Liu, X., & Zhang, Q. (2019). Investigation of selective binding of
inhibitors to PTP1B and TCPTP by accelerated molecular dynamics simulations. J
Biomol Struct Dyn, 37(14), 3697-3706. doi: 10.1080/07391102.2018.1526117

Chen, Y. Y., Chu, H. M., Pan, K. T., Teng, C. H., Wang, D. L., Wang, A. H., . . . Meng, T. C.
2008). Cysteine S-nitrosylation protects protein-tyrosine phosphatase 1B against
oxidation-induced permanent inactivation. J Biol Chem, 283(50), 35265-35272. doi:
0.1074/jbc.M805287200
(
1
Cheng, A., & Merz, K. M., Jr. (2003). Prediction of aqueous solubility of a diverse set of
compounds using quantitative structure-property relationships. J Med Chem, 46(17),
3572-3580. doi: 10.1021/jm020266b
Cheng, A., Uetani, N., Simoncic, P. D., Chaubey, V. P., Lee-Loy, A., McGlade, C. J., . . .
Tremblay, M. L. (2002). Attenuation of leptin action and regulation of obesity by
protein tyrosine phosphatase 1B. Dev Cell, 2(4), 497-503.
Douty, B., Wayland, B., Ala, P. J., Bower, M. J., Pruitt, J., Bostrom, L., . . . Yue, E. W. (2008).
Isothiazolidinone inhibitors of PTP1B containing imidazoles and imidazolines. Bioorg
Med Chem Lett, 18(1), 66-71. doi: 10.1016/j.bmcl.2007.11.012
Egan, W. J., Walters, W. P., & Murcko, M. A. (2002). Guiding molecules towards drug-likeness.
Curr Opin Drug Discov Devel, 5(4), 540-549.
Elchebly, M., Payette, P., Michaliszyn, E., Cromlish, W., Collins, S., Loy, A. L., . . . Kennedy, B.
P. (1999). Increased insulin sensitivity and obesity resistance in mice lacking the
protein tyrosine phosphatase-1B gene. Science, 283(5407), 1544-1548. doi:
10.1126/science.283.5407.1544
Eldehna, W. M., Abdelrahman, M. A., Nocentini, A., Bua, S., Al-Rashood, S. T., Hassan, G.
S., . . . Supuran, C. T. (2019). Synthesis, biological evaluation and in silico studies with
4-benzylidene-2-phenyl-5(4H)-imidazolone-based benzenesulfonamides as novel

selective carbonic anhydrase IX inhibitors endowed with anticancer activity. Bioorg
Chem, 90, 103102. doi: 10.1016/j.bioorg.2019.103102
Fakhar, Z., Govender, T., Maguire, G. E. M., Lamichhane, G., Walker, R. C., Kruger, H. G., &
Honarparvar, B. (2017). Differential flap dynamics in l,d-transpeptidase2 from
mycobacterium tuberculosis revealed by molecular dynamics. Mol Biosyst, 13(6),
1223-1234. doi: 10.1039/c7mb00110j
Fernandez-Silva, M. J., Alonso-Gonzalez, A., Gonzalez-Perez, E., Gestal-Otero, J. J., &
Diaz-Gravalos, G. J. (2019). [Health literacy in patients with type 2 diabetes: A
cross-sectional study using the HLS-EU-Q47 questionnaire]. Semergen, 45(1), 30-36.
doi: 10.1016/j.semerg.2018.08.003
Goldstein, B. J., Bittner-Kowalczyk, A., White, M. F., & Harbeck, M. (2000). Tyrosine
dephosphorylation and deactivation of insulin receptor substrate-1 by protein-tyrosine
phosphatase 1B. Possible facilitation by the formation of a ternary complex with the
Grb2 adaptor protein. J Biol Chem, 275(6), 4283-4289. doi: 10.1074/jbc.275.6.4283
Grant, B. J., Rodrigues, A. P., ElSawy, K. M., McCammon, J. A., & Caves, L. S. (2006). Bio3d:
an R package for the comparative analysis of protein structures. Bioinformatics,
22(21), 2695-2696. doi: 10.1093/bioinformatics/btl461
Hess, B. (2008). P-LINCS: A Parallel Linear Constraint Solver for Molecular Simulation. J
Chem Theory Comput, 4(1), 116-122. doi: 10.1021/ct700200b
Jiang, C. S., Ge, Y. X., Cheng, Z. Q., Wang, Y. Y., Tao, H. R., Zhu, K., & Zhang, H. (2019).
Discovery of New Selective Butyrylcholinesterase (BChE) Inhibitors with Anti-Abeta
Aggregation Activity: Structure-Based Virtual Screening, Hit Optimization and

Biological Evaluation. Molecules, 24(14). doi: 10.3390/molecules24142568
Jin, W. Y., Ma, Y., Li, W. Y., Li, H. L., & Wang, R. L. (2018). Scaffold-based novel SHP2
allosteric inhibitors design using Receptor-Ligand pharmacophore model, virtual
screening and molecular dynamics. Comput Biol Chem, 73, 179-188. doi:
10.1016/j.compbiolchem.2018.02.004
Klaman, L. D., Boss, O., Peroni, O. D., Kim, J. K., Martino, J. L., Zabolotny, J. M., . . . Kahn, B.
B. (2000). Increased energy expenditure, decreased adiposity, and tissue-specific
insulin sensitivity in protein-tyrosine phosphatase 1B-deficient mice. Mol Cell Biol,
20(15), 5479-5489. doi: 10.1128/mcb.20.15.5479-5489.2000
Li, X., Wang, X., Tian, Z., Zhao, H., Liang, D., Li, W., . . . Lu, S. (2014). Structural basis of
valmerins as dual inhibitors of GSK3beta/CDK5. J Mol Model, 20(9), 2407. doi:
10.1007/s00894-014-2407-1
Liu, G. (2004). Technology evaluation: ISIS-113715, Isis. Curr Opin Mol Ther, 6(3), 331-336.
Liu, J., Wang, F., Ma, Z., Wang, X., & Wang, Y. (2011). Structural determination of three
different series of compounds as Hsp90 inhibitors using 3D-QSAR modeling,
molecular docking and molecular dynamics methods. Int J Mol Sci, 12(2), 946-970. doi:
10.3390/ijms12020946
Liu, W. S., Wang, R. R., Li, W. Y., Rong, M., Liu, C. L., Ma, Y., & Wang, R. L. (2019).
Investigating the reason for loss-of-function of Src homology 2 domain-containing
protein tyrosine phosphatase 2 (SHP2) caused by Y279C mutation through molecular
dynamics
simulation.
J
Biomol
Struct
Dyn,
1-12.
doi:
10.1080/07391102.2019.1634641

Liu, W. S., Wang, R. R., Sun, Y. Z., Li, W. Y., Li, H. L., Liu, C. L., . . . Wang, R. L. (2019).
Exploring the effect of inhibitor AKB-9778 on VE-PTP by molecular docking and
molecular dynamics simulation. J Cell Biochem. doi: 10.1002/jcb.28963
Loeffler, H. H., & Winn, M. D. (2013). Ligand binding and dynamics of the monomeric
epidermal growth factor receptor ectodomain. Proteins, 81(11), 1931-1943. doi:
10.1002/prot.24339
Ma, Y., Wang, S. Q., Xu, W. R., Wang, R. L., & Chou, K. C. (2012). Design novel dual agonists
for treating type-2 diabetes by targeting peroxisome proliferator-activated receptors
with
core
hopping
approach.
PLoS
One,
7(6),
e38546.
doi:
10.1371/journal.pone.0038546
Machaba, K. E., Mhlongo, N. N., & Soliman, M. E. S. (2018). Induced Mutation Proves a
Potential Target for TB Therapy: A Molecular Dynamics Study on LprG. Cell Biochem
Biophys, 76(3), 345-356. doi: 10.1007/s12013-018-0852-7
Maisuradze, G. G., Liwo, A., & Scheraga, H. A. (2009). Principal component analysis for
protein folding dynamics. J Mol Biol, 385(1), 312-329. doi: 10.1016/j.jmb.2008.10.018
Miyake, M., Koga, T., Kondo, S., Yoda, N., Emoto, C., Mukai, T., & Toguchi, H. (2017).
Prediction of drug intestinal absorption in human using the Ussing chamber system: A
comparison of intestinal tissues from animals and humans. Eur J Pharm Sci, 96,
373-380. doi: 10.1016/j.ejps.2016.10.006
Ndagi, U., Mhlongo, N. N., & Soliman, M. E. (2017). The impact of Thr91 mutation on c-Src
resistance to UM-164: molecular dynamics study revealed a new opportunity for drug
design. Mol Biosyst, 13(6), 1157-1171. doi: 10.1039/c6mb00848h

Ndagi, U., Mhlongo, N. N., & Soliman, M. E. (2018). Emergence of a Promising Lead
Compound in the Treatment of Triple Negative Breast Cancer: An Insight into
Conformational Features and Ligand Binding Landscape of c-Src Protein with UM-164.
Appl Biochem Biotechnol, 185(3), 655-675. doi: 10.1007/s12010-017-2677-z
Neves, B. J., Braga, R. C., Melo-Filho, C. C., Moreira-Filho, J. T., Muratov, E. N., & Andrade, C.
H. (2018). QSAR-Based Virtual Screening: Advances and Applications in Drug
Discovery. Front Pharmacol, 9, 1275. doi: 10.3389/fphar.2018.01275
Olde Scheper, T. V., Meredith, R. M., Mansvelder, H. D., van Pelt, J., & van Ooyen, A. (2017).
Dynamic Hebbian Cross-Correlation Learning Resolves the Spike Timing Dependent
Plasticity
Conundrum.
Front
Comput
Neurosci,
11,
119.
doi:
10.3389/fncom.2017.00119
Pol-Fachin, L., Fernandes, C. L., & Verli, H. (2009). GROMOS96 43a1 performance on the
characterization of glycoprotein conformational ensembles through molecular
dynamics
simulations.
Carbohydr
Res,
344(4),
491-500.
doi:
10.1016/j.carres.2008.12.025
Schneider, G., & Fechner, U. (2005). Computer-based de novo design of drug-like molecules.
Nat Rev Drug Discov, 4(8), 649-663. doi: 10.1038/nrd1799
Su, J., Liu, X., Zhang, S., Yan, F., Zhang, Q., & Chen, J. (2018). A computational insight into
binding modes of inhibitors XD29, XD35, and XD28 to bromodomain-containing
protein 4 based on molecular dynamics simulations. J Biomol Struct Dyn, 36(5),
1212-1224. doi: 10.1080/07391102.2017.1317666
Sun, W., Zhang, B., Zheng, H., Zhuang, C., Li, X., Lu, X., . . . Xiu, Z. (2017). Trivaric acid, a

new inhibitor of PTP1b with potent beneficial effect on diabetes. Life Sci, 169, 52-64.
doi: 10.1016/j.lfs.2016.11.012
Sun, W., Zhuang, C., Li, X., Zhang, B., Lu, X., Zheng, Z., & Dong, Y. (2017). Varic acid
analogues from fungus as PTP1B inhibitors: Biological evaluation and
structure-activity relationships. Bioorg Med Chem Lett, 27(15), 3382-3385. doi:
10.1016/j.bmcl.2017.06.001
Susnow, R. G., & Dixon, S. L. (2003). Use of robust classification techniques for the prediction
of human cytochrome P450 2D6 inhibition. J Chem Inf Comput Sci, 43(4), 1308-1315.
doi: 10.1021/ci030283p
Tadic, M., & Cuspidi, C. (2015). Type 2 diabetes mellitus and atrial fibrillation: From
mechanisms to clinical practice. Arch Cardiovasc Dis, 108(4), 269-276. doi:
10.1016/j.acvd.2015.01.009
Tang, J., Qu, Z., Luo, J., He, L., Wang, P., Zhang, P., . . . Huang, Y. (2018). Molecular
Dynamics Simulations of the Oil-Detachment from the Hydroxylated Silica Surface:
Effects of Surfactants, Electrostatic Interactions, and Water Flows on the Water
Molecular Channel Formation.
0.1021/acs.jpcb.7b09716
J
Phys Chem B, 122(6), 1905-1918. doi:
1
Thiyagarajan, V., Lin, S. H., Chang, Y. C., & Weng, C. F. (2016). Identification of novel FAK
and S6K1 dual inhibitors from natural compounds via ADMET screening and
molecular
0.1016/j.biopha.2016.02.020
Wan, H., Hu, J. P., Tian, X. H., & Chang, S. (2013). Molecular dynamics simulations of wild
docking.
Biomed
Pharmacother,
80,
52-62.
doi:
1

type and mutants of human complement receptor 2 complexed with C3d. Phys Chem
Chem Phys, 15(4), 1241-1251. doi: 10.1039/c2cp41388d
Wang, M. Y., Jin, Y. Y., Wei, H. Y., Zhang, L. S., Sun, S. X., Chen, X. B., . . . Wang, R. L.
(
2015).
Synthesis,
biological
evaluation
and
3D-QSAR
studies
of
imidazolidine-2,4-dione derivatives as novel protein tyrosine phosphatase 1B
inhibitors. Eur J Med Chem, 103, 91-104. doi: 10.1016/j.ejmech.2015.08.037
Wesson, L., & Eisenberg, D. (1992). Atomic solvation parameters applied to molecular
dynamics of proteins in solution. Protein Sci, 1(2), 227-235. doi:
10.1002/pro.5560010204
Xu, L., Kong, R., Zhu, J., Sun, H., & Chang, S. (2016). Unraveling the conformational
determinants of LARP7 and 7SK small nuclear RNA by theoretical approaches. Mol
Biosyst, 12(8), 2613-2621. doi: 10.1039/c6mb00252h
Yan, F., Liu, X., Zhang, S., Su, J., Zhang, Q., & Chen, J. (2019). Electrostatic
interaction-mediated conformational changes of adipocyte fatty acid binding protein
probed by molecular dynamics simulation. J Biomol Struct Dyn, 37(14), 3583-3595.
doi: 10.1080/07391102.2018.1520648
Yesudhas, D., Anwar, M. A., Panneerselvam, S., Durai, P., Shah, M., & Choi, S. (2016).
Structural Mechanism behind Distinct Efficiency of Oct4/Sox2 Proteins in Differentially
Spaced
DNA
Complexes.
PLoS
One,
11(1),
e0147240.
doi:
10.1371/journal.pone.0147240
Zabolotny, J. M., Bence-Hanulec, K. K., Stricker-Krongrad, A., Haj, F., Wang, Y., Minokoshi,
Y., . . . Neel, B. G. (2002). PTP1B regulates leptin signal transduction in vivo. Dev Cell,

2
(4), 489-495.
Zhai, M., Liu, S., Gao, M., Wang, L., Sun, J., Du, J., . . . Zhang, W. (2019).
,5-Diaryl-1H-pyrazolo[3,4-b]pyridines as potent tubulin polymerization inhibitors:
3
Rational design, synthesis and biological evaluation. Eur J Med Chem, 168, 426-435.
doi: 10.1016/j.ejmech.2018.12.053
Zhang, J., Luan, C. H., Chou, K. C., & Johnson, G. V. (2002). Identification of the N-terminal
functional domains of Cdk5 by molecular truncation and computer modeling. Proteins,
48(3), 447-453. doi: 10.1002/prot.10173
Zhou, Y., Zhang, N., Chen, W., Zhao, L., & Zhong, R. (2016). Underlying mechanisms of cyclic
peptide inhibitors interrupting the interaction of CK2alpha/CK2beta: comparative
molecular dynamics simulation studies. Phys Chem Chem Phys, 18(13), 9202-9210.
doi: 10.1039/c5cp06276d