Xanthine oxidase inhibitory properties of 1,2,3,4-tetrahydroisoquinoline derivatives.

Full text

48
www.medfak.ni.ac.rs/amm
Original article
UDC: 577.152.1:547.833
doi:10.5633/amm.2021.0106
XANTHINE OXIDASE INHIBITORY PROPERTIES OF
1,2,3,4-TETRAHYDROISOQUINOLINE DERIVATIVES
Mihajlo Gajić1, Budimir S. Ilić2, Bojan P. Bondžić3, Zdravko Džambaski3, Ana Filipović3,
Gordana Kocić4, Andrija Šmelcerović2
Xanthine oxidase (XO) is a versatile metalloflavoprotein enzyme that is best known for
its rate-limiting role in the purine degradation pathway. Therapeutic inhibition of XO is based on
its role in a variety of diseases that is attributed either to the hyperproduction of uric acid, or
the hyperproduction of reactive oxygen species. Herein, we report the assessment of XO
inhibitory properties of 24 1,2,3,4-tetrahydroisoquinoline derivatives, among which compound
16 exhibited IC50 value of 135.72 ± 2.71 µM. The interaction of compound 16 with XO enzyme
was simulated using the Site Finder module, molecular docking and molecular dynamics.
Molecular modeling suggests that interactions with Met 1038, Gln 1040, Thr 1077, Gln 1194
and Val 1259 are an important factor for inhibitor affinity toward the XO enzyme. Our proposed
binding model might be beneficial for the discovery of new active 1,2,3,4-tetrahydroisoquino-
line-based inhibitors of XO enzyme.
Acta Medica Medianae 2021;60(1):48-55.
Key words: xanthine oxidase inhibition, 1,2,3,4-tetrahydroisoquinolines, molecular
docking, molecular dynamic simulation
1University of Niš, Faculty of Medicine, Department of
Pharmacy, Niš, Serbia
2University of Niš, Faculty of Medicine, Department of
Chemistry, Niš, Serbia
3University of Belgrade, Institute of Chemistry, Technology and
Metallurgy, Belgrade, Serbia
4University of Niš, Faculty of Medicine, Department of
Biochemistry, Niš, Serbia
Contact: Andrija Šmelcerović
48 Dr Zoran Djindjić Blvd., 18000 Niš, Serbia
E-mail: andrija.smelcerovic@medfak.ni.ac.rs
a.smelcerovic@yahoo.com
Introduction
Xanthine oxidoreductase (XOR) is a versatile
molybdopterin-containing flavoprotein enzyme that
exists in two interconvertible forms, xanthine oxi-
dase (XO) and xanthine dehydrogenase (XDH) (1).
The enzyme is best known for its rate-limiting role in
the purine degradation pathway, where it catalyzes
oxidative hydroxylation of hypoxanthine and xan-
thine, subsequently producing uric acid (2, 3). Dehy-
drogenase form is predominant in healthy tissues,
while conversion to XO occurs in pathological con-
ditions, after XDH proteolysis or oxidation of some of
its sulfhydryl residues (1). Superoxide anion radical
and hydrogen peroxide that are generated as the
byproducts of enzyme activity are responsible for
oxidative stress which usually accompanies elevated
XO activity. The role of XO in diseases is attributed
either to the hyperproduction of uric acid, or the
hyperproduction of reactive oxygen species. There-
fore, pharmacological inhibition of xanthine oxidase
is proven to be invaluable for the treatment of
hyperuricemia and gout in the first place, but might
also be beneficial for plethora of conditions, such as
cholecystitis, hemorrhagic shock, ischemia-reperfu-
sion injuries, hypercholesterolemia and carcinogene-
sis (4).
Given the broad therapeutic potential of XO
inhibitors, the aim of the current study was to
assess a group of 24 1,2,3,4-tetrahydroisoquinoline
derivatives for potential inhibitory properties against
XO and to perform molecular docking and molecular
dynamics simulation on active compounds, in order
to elucidate key structural features responsible for
XO inhibitory activity.
Materials and methods
Compounds
The synthesis of 24 1,2,3,4-tetrahydroiso-
quinoline derivatives was preformed according to the
description in our previous study (5).

Xanthine oxidase inhibitory properties of 1,2,3,4-tetrahydroisoquinoline... Mihajlo Gajić et al.
49
Evaluation of xanthine oxidase inhibition
Compounds were studied for inhibitory pro-
perties against bovine milk xanthine oxidase. Spec-
trophotometric measurement of uric acid formation
at 293 nm was used for in vitro evaluation of en-
zyme inhibition, as described in our previous studies
(6, 7). Initially, all compounds were assayed at a
concentration of 150 µM, while those inhibiting more
than 50% of enzyme activity were subsequently
tested in a broader concentration range to allow for
IC50 determination. Allopurinol was used as a posi-
tive control. All experiments were performed in tri-
plicate and averaged.
Ligand preparation
Examined inhibitor was generated using the
builder panel in the Molecular Operating Environ-
ment (MOE) 2019.0101 software (8). Using the MOE
LigX module, partial atomic charges were ascribed
and possible ionization states were generated at a
pH of 7.0. The MMFF94x force field was used for
optimization and the resulting structure was used for
modeling studies. Conformational search was carried
out by MOE LowModelMD method which performs
molecular dynamic perturbations along with low
frequency vibrational modes with energy window of
7 kcal/mol, and conformational limits of 1000.
Receptor preparation
The X-ray crystallographic structure of XO
enzyme (PDB code: 1N5X), retrieved from the Pro-
tein Data Bank, was prepared using the Structure
Preparation process in MOE. After the correction, hy-
drogens were added and partial charges (Gasteiger
methodology) were calculated. Energy minimization
(AMBER14:EHT, RMS gradient: 0.100) was perfor-
med.
Binding site selection
The Site Finder module of the MOE was used
to identify possible ligand-binding sites within the
optimized structure of XO enzyme. Hydrophobic or
hydrophilic alpha spheres served as probes denoting
zones of tight atom packing. These alpha spheres
were utilized to define and rank potential ligand-
binding sites according to their propensity for ligand
binding (PLB) score, which was based on the amino
acid composition of the pocket (9).
Docking protocol
The molecular docking study was performed
using the MOE to understand the ligand/protein
interactions in detail. The default Triangle Matcher
placement method was used for the induced fit
docking. GBVI/WSA dG scoring function which esti-
mates the free energy of binding of the ligand from
a given pose was used to rank the final poses. The
ligand/protein complex with lowest relative binding
free energy (ΔG) score was selected for further
study.
Molecular dynamics simulation
The molecular dynamics simulation of com-
pound 16 on XO enzyme, was carried out using the
Desmond Molecular Dynamics System (Desmond)
2018.4 software (10). The structure of the added
water was based on the simple point charge (SPC)
solvent model. The system was neutralized with Na+
ions to balance the net charge of the whole simula-
tion box to neutral. The final system contained ap-
proximately 123,000 atoms. The system was passed
through a 6-step relaxation protocol before molecu-
lar dynamics simulations. The relaxed system was
simulated for 10 ns, using a normal pressure tempe-
rature (NPT) ensemble with a Nosé–Hoover thermo-
stat at 300 K and Martyna–Tobias–Klein barostat at
1.01325 bar pressure. Atomic coordinate data and
system energies were recorded every 1 ps. The root
mean square deviation (RMSD) and root mean
square fluctuation (RMSF) of the inhibitor/XO en-
zyme complex were analyzed with respect to the
simulation time.
Results and discussion
Previously synthetized 1,2,3,4-tetrahydroiso-
quinoline derivatives (5) were assayed for potential
in vitro XO inhibitory properties. Among 24 tested
compounds (Table 1), IC50 value below 150 µM was
observed only in the case of 16 (135.72 ± 2.71 µM).
Analysis of structural features of active compound
16 and its closest inactive analog 17 indicate to 7,8-
dimethoxy substitution on 1,2,3,4-tetrahydroiso-
quinoline core as a probable cause for the lack of XO
inhibitory properties of 17. Identified active com-
pound 2-(4-fluorophenyl)-1,2,3,4-tetrahydroisoquino-
line-1-carbonitrile (16) was subjected to docking
studies and molecular dynamics simulation with the
goal to provide insight into the key structural featu-
res required for its XO inhibitory activity.
By combining a novel and highly effective
algorithm for rapid binding-site evaluation with
easy-to-use property visualization tools, Site Finder
provides researchers with efficient means to identify
and characterize binding sites (9). The results from
the Site Finder analysis highlighted that catalytic
residues like Gln 767, Glu 802, Arg 880, Phe 914,
Phe 1009 and Glu 1261 (11, 12) constituted the top-
ranked binding pocket of XO enzyme (Table 2,
Figure 1).
The intermolecular contacts between exam-
ined inhibitor and XO enzyme were analyzed using
the ligand interaction diagram of MOE suite (Table 3,
Figure 2). It illustrates the existence of hydrogen
bond and pi-H interactions. Additionally, the bond
distances, bond energy and binding free energy be-
tween the inhibitor and receptor atoms were also
examined (Table 3). The molecular docking high-
lighted the importance of Met 1038, Gln 1040, Thr
1077, Gln 1194 and Val 1259 in the formation of
inhibitor/XO enzyme complex (Table 3, Figure 2).
Observed interactions between 16 and non-catalytic

Acta Medica Medianae 2021, Vol.60(1) Xanthine oxidase inhibitory properties of 1,2,3,4-tetrahydroisoquinoline...
50
residues (Table 3, Figure 2), are similar to molecular
interplay involving non-purine XO inhibitors (13-16).
Some of recently synthesized pyrazole derivatives
were found to inhibit XO enzyme via Met 1038, Gln
1040 and Gln 1194 residues (13). Furthermore,
inhibitory effect of verbascoside on XO activity, can
be attributed to the formation of hydrogen bond
with Gln 1194 residue (14). Moreover, molecular
simulation revealed that newly synthesized hesperi-
din derivatives interacted with XO residues Met
1038, Gln 1040, Thr 1077, Gln 1194 and Val 1259
(15). Additionally, molecular docking revealed that
1,4-dicaffeoylquinic acid interacted with XO enzyme
via Gln 1040 and Thr 1077 residues (16).
Table 1. In vitro XO inhibitory activity of 1,2,3,4-tetrahydroisoquinolines
Compd.
R1
R2
R3
R4
IC50 values (µM) ± SD
1
H
CH2COCH3
H
H
> 150
2
H
CH2COCH3
OCH3
OCH3
> 150
3
Ph
CH2COCH3
H
H
> 150
4
Ph
CH2COCH3
OCH3
OCH3
> 150
5
Ph
CH2COCH2CH3
H
H
> 150
6
Ph
CH2COCH2CH3
OCH3
OCH3
> 150
7
Ph
CH2COPh
H
H
> 150
8
Ph
CH2COPh
OCH3
OCH3
> 150
9
Ph
2-oxocyclohexane
H
H
> 150
10
o-Me-Ph
CH2COCH3
H
H
> 150
11
p-F-Ph
CH2COCH3
H
H
> 150
12
p-F-Ph
CH2COCH3
OCH3
OCH3
> 150
13
p-F-Ph
CH2COCH2CH3
H
H
> 150
14
p-F-Ph
CH2COCH2CH3
OCH3
OCH3
> 150
15
p-F-Ph
CH2COPh
H
H
> 150
16
p-F-Ph
CN
H
H
135.72 ± 2.71
17
p-F-Ph
CN
OCH3
OCH3
> 150
18
p-F-Ph
2-oxocyclohexane
H
H
> 150
19
p-F-Ph
2-oxocyclohexane
OCH3
OCH3
> 150
20
p-Me-Ph
CH2COCH3
H
H
> 150
21
p-Me-Ph
CH2COPh
H
H
> 150
22
p-Me-Ph
CH2COPh
OCH3
OCH3
> 150
23
p-Me-Ph
CH2COCH2CH3
OCH3
OCH3
> 150
24
p-Me-Ph
CN
OCH3
OCH3
> 150

Xanthine oxidase inhibitory properties of 1,2,3,4-tetrahydroisoquinoline... Mihajlo Gajić et al.
51
Table 2. Summary of the top five inhibitor-binding sites in XO
Site
Size
PLB
Hyd
Side
Residues
1
172
1.31
49
95
Gln 112, Cys 150, Gln 767, Gly 796, Gly 797, Phe 798, Gly
799, Glu 802, Leu 873, Ser 876, Arg 880, Ala 910, Phe 911,
Arg 912, Gly 913, Phe 914, Phe 1005, Ser 1008, Phe 1009,
Thr 1010, Val 1011, Leu 1014, Met 1038, Gly 1039, Gln 1040,
Gly 1041, Leu 1042, Lys 1045, Pro 1076, Thr 1077, Ala 1078,
Ala 1079, Ser 1080, Val 1081, Ser 1082, Thr 1083, Ile 1190,
Asp 1191, Gln 1194, Gly 1197, Lys 1257, Ala 1258, Val 1259,
Gly 1260, Glu 1261
2
98
0.87
35
63
Cys 662, Val 663, Ile 696, Thr 697, Ile 698, Glu 699, Tyr 735,
Gly 738, Gln 739, Asp 740, His 741, Asp 832, Met 833, Leu
834, Ile 835, Thr 836, Gly 837, Gly 838, Arg 839, Pro 841, Asn
866, Asn 904, Leu 905, Ser 906, Asn 908, Leu 1211, Tyr
1213, Ser 1214, Pro 1215, Gly 1217, Ser 1218, Leu 1219, Thr
1221, Arg 1222
3
101
0.68
20
60
His 614, Asp 651, Glu 652, Thr 653, Thr 661, Cys 662, Val
663, Gly 664, His 665, Ile 666, Pro 693, Ala 694, Ile 695, Ser
706, Tyr 708, Arg 804, Leu 807, Leu 834, Ile 835, Thr 836,
Gly 837, Gly 868, Asn 869, Ser 870, Arg 871, Lys 902, Thr
903, Asn 904, Ser 906, Ser 907
4
185
0.59
51
95
Gln 62, Ile 66, His 67, Phe 68, Ser 69, Ser 123, Thr 126, Leu
127, Asn 130, Gln 131, Glu 137, Glu 138, Asp 141, Ala 142,
Gln 144, Ala 304, Ser 306, Ser 307, Glu 309, Lys 310, Leu
313, Arg 328, Leu 331, Glu 332, Leu 334, Arg 335, Trp 336,
Lys 340, Gln 341, Lys 343, Ser 344, Leu 548, Gln 550, Lys
551
5
95
0.54
34
63
Leu 952, Asn 956, Gln 957, Arg 958, Leu 959, Glu 960, Gly
961, Ser 963, Pro 1136, Asn 1137, Leu 1138, Asn 1148, Phe
1150, His 1151, Tyr 1152, Phe 1153, Tyr 1155, Asp 1181, Gly
1183, Cys 1247, Pro 1248, Asn 1249, Lys 1250, Lys 1251, Lys
1257
Figure 1. The top ranked XO binding site, represented by a grey-red surface map

Acta Medica Medianae 2021, Vol.60(1) Xanthine oxidase inhibitory properties of 1,2,3,4-tetrahydroisoquinoline...
52
Table 3. Summary of the 16 interactions with protein residues in XO
Inhibitor
Inhibitor
atoms
Protein
residue
Inhibitor-Protein
Interactions
Distance
(Å)
E
(kcal/mol)
ΔG binding
(kcal/mol)
16
C10
Met 1038
H-donor
4.29
-0.3
-8.36
C12
Gln 1194
H-donor
3.06
-0.5
N19
Thr 1077
H-acceptor
3.47
2.2
6-ring
Gln 1040
pi-H
4.41
-0.2
6-ring
Val 1259
pi-H
4.64
-0.4
Figure 2. 3D/2D view of compound 16 (A, B) bound in the active site of XO.
The polar part of the active site is shown as a pink surface, hydrophobic part as a green surface,
while the solvent exposed part is shown as a red surface.
The study was further extended to assess the
stability of 16/XO enzyme complex through the
molecular dynamics simulation. The RMSD and RMSF
plots for XO enzyme and examined inhibitor showed
that docking complex was stable during entire simu-
lation period (Figures 3-4). The RMSD for Cα, side
chains and heavy atoms remained within the limit of
2 Å. The similar situation was noted for RMSF
values. The obtained results indicated small struc-
tural rearrangements, less conformational changes
and confirmed stability of 16/XO enzyme complex
(17).
Figure 3. RMSD (A) and RMSF (B) plot of XO, obtained during the course of 10 ns molecular dynamics simulation

Xanthine oxidase inhibitory properties of 1,2,3,4-tetrahydroisoquinoline... Mihajlo Gajić et al.
53
Figure 4. RMSD (A) and RMSF (B) plot of compound 16,
obtained during the course of 10 ns molecular dynamics simulation
Conclusion
A series of 24 1,2,3,4-tetrahydroisoquinoline
derivatives were screened for potential XO inhibitory
properties. Among them, only compound 16 (IC50 =
135.72 ± 2.71 µM) exhibited IC50 value below 150
µM, and was further subjected to molecular docking
and molecular dynamic simulation. Molecular model-
ing suggests that interactions with Met 1038, Gln
1040, Thr 1077, Gln 1194 and Val 1259 are an
important factor for inhibitor affinity toward the XO
enzyme. Observed interactions with non-catalytic
residues, might be beneficial for the discovery of
new active 1,2,3,4-tetrahydroisoquinoline-based in-
hibitors of XO enzyme.
Acknowledgements
The financial support of this work by the
Ministry of Education, Science and Technological
Development of the Republic of Serbia (Grant
numbers 451-03-68/2020-14/200113 and 451-03-
68/2020-14/200026) and the Faculty of Medicine of
the University of Niš (Internal project No. 40) are
gratefully acknowledged. The authors would like to
thank D. E. Shaw Research, for providing us the
Desmond software package free of cost for this
study.

Acta Medica Medianae 2021, Vol.60(1) Xanthine oxidase inhibitory properties of 1,2,3,4-tetrahydroisoquinoline...
54
References
1. Pacher PA, Nivorozhkin A, Szabó C. Therapeutic
effects of xanthine oxidase inhibitors: renaissance half
a century after the discovery of allopurinol. Pharmacol
Rev 2006;58(1):87-114. [CrossRef][PubMed]
2. Berry CE, Hare JM. Xanthine oxidoreductase and
cardiovascular disease: molecular mechanisms and
pathophysiological implications. J Physiol 2004;555
(3):589-606. [CrossRef][PubMed]
3. Ojha R, Singh J, Ojha A, Singh H, Sharma S, Nepali K.
An updated patent review: xanthine oxidase inhibitors
for the treatment of hyperuricemia and gout (2011-
2015). Expert Opin Ther Pat 2017;27(3):311-45.
[CrossRef][PubMed]
4. Šmelcerović A, Tomović K, Šmelcerović Ž, Petronijević
Ž, Kocić G, Tomašič T, et al. Xanthine oxidase inhi-
bitors beyond allopurinol and febuxostat; an overview
and selection of potential leads based on in silico
calculated physico-chemical properties, predicted
pharmacokinetics and toxicity. Eur J Med Chem
2017;135:491-516. [CrossRef][PubMed]
5. Džambaski Z, Bondžić BP. Dehydrogenative C(sp3)–H
bond functionalization of tetrahydroisoquinolines me-
diated by organic oxidants under mild conditions. Org
Biomol Chem 2019;17(26):6420-5.
[CrossRef][PubMed]
6. Šmelcerovic Ž, Veljković A, Kocić G, Yancheva D,
Petronijević Ž, Anderluh M, Šmelcerović A. Xanthine
oxidase inhibitory properties and anti-inflammatory
activity of 2-amino-5-alkylidene-thiazol-4-ones. Chem
Biol Interact 2015;229:73-81. [CrossRef][PubMed]
7. Tomović K, Ilić BS, Šmelcerović Ž, Miljković M,
Yancheva D, Kojić M, Mavrova AT, Kocić G,
Šmelcerović A. Benzimidazole-based dual dipeptidyl
peptidase-4 and xanthine oxidase inhibitors. Chem
Biol Interact 2020;315:108873. [CrossRef][PubMed]
8. MOE, Molecular Operating Environment (computer
program). Version 2019.0101. Montreal (Canada):
Chemical Computing Group ULC; 2019.
9. Soga S, Shirai H, Kobori M, Hirayama N. Use of amino
acid composition to predict ligand-binding sites. J
Chem Inf Model 2007;47(2):400-6.
[CrossRef][PubMed]
10. Desmond, Desmond Molecular Dynamics System
(computer program). Version 2018.4. New York (NY):
D.E. Shaw Research; 2018.
11. Okamoto K, Matsumoto K, Hille R, Eger BT, Pai EF,
Nishino T. The crystal structure of xanthine oxidore-
ductase during catalysis: implications for reaction
mechanism and enzyme inhibition. Proc Natl Acad Sci
U S A 2004;101(21):7931-6. [CrossRef][PubMed]
12. Pauff JM, Cao H, Hille R. Substrate orientation and
catalysis at the molybdenum site in xanthine oxidase
crystal structures in complex with xanthine and
lumazine. J Biol Chem 2009;284(13):8760-7.
[CrossRef][PubMed]
13. Li P, Tian Y, Zhai H, Deng F, Xie M, Zhang X. Study on
the activity of non-purine xanthine oxidase inhibitor by
3D-QSAR modeling and molecular docking. J Mol
Struct 2013;1051:56-65. [CrossRef]
14. Wan Y, Zou B, Zeng H, Zhang L, Chen M, Fu G.
Inhibitory effect of verbascoside on xanthine oxidase
activity. Int J Biol Macromol 2016;93:609-14.
[CrossRef][PubMed]
15. Malik N, Dhiman P, Khatkar A. Mechanistic approach
towards interaction of newly synthesized hesperidin
derivatives against xanthine oxidase. Int J Biol
Macromol 2019;135:864-76. [CrossRef][PubMed]
16. Cao W, Fang Y, Wu T, Liang F, Cheng Y, Salah M, et
al. Insights from multispectral and molecular docking
investigation on the xanthine oxidase inhibition by
1,4-dicaffeoylquinic acid. J Mol Struct 2020:128475.
[CrossRef]
17. Liu K, Kokubo H. Exploring the stability of ligand
binding modes to proteins by molecular dynamics
simulations: a cross-docking study. J Chem Inf Model
2017;57(10):2514-22. [CrossRef][PubMed]

Xanthine oxidase inhibitory properties of 1,2,3,4-tetrahydroisoquinoline... Mihajlo Gajić et al.
55
Originalni rad
UDC: 577.152.1:547.833
doi:10.5633/amm.2021.0106
INHIBICIJA KSANTIN OKSIDAZE DERIVATIMA
1,2,3,4-TETRAHIDROIZOHINOLINA
Mihajlo Gajić1, Budimir S. Ilić2, Bojan P. Bondžić3, Zdravko Džambaski3, Ana Filipović3,
Gordana Kocić4, Andrija Šmelcerović2
1Univerzitet u Nišu, Medicinski fakultet, Katedra za farmaciju, Niš, Srbija
2Univerzitet u Nišu, Medicinski fakultet, Katedra za hemiju, Niš, Srbija
3Univerzitet u Beogradu, Institut za hemiju, tehnologiju i metalurgiju, Beograd, Srbija
4Univerzitet u Nišu, Medicinski fakultet, Katedra za biohemiju, Niš, Srbija
Kontakt: Andrija Šmelcerović
Bulevar dr Zorana Đinđića 48, 18000 Niš, Srbija
E-mail: andrija.smelcerovic@medfak.ni.ac.rs
a.smelcerovic@yahoo.com
Ksantin oksidaza (XO) je metaloflavoproteinski enzim, koji je najpoznatiji po svojoj
ulozi ograničavanja brzine razgradnje purinskih nukleotida. Terapijska inhibicija XO zasniva se
na njenoj ulozi u brojnim bolestima, koje su povezane bilo sa hiperprodukcijom mokraćne
kiseline ili hiperprodukcijom reaktivnih kiseoničnih vrsta. U ovom radu izvršeno je ispitivanje
sposobnosti inhibicije XO 24 derivata 1,2,3,4-tetrahidroizohinolina, od kojih je jedinjenje 16
pokazalo IC50 vrednost od 135,72 µM ± 2,71 µM. Interakcija jedinjenja 16 sa XO enzimom
simulirana je korišćenjem Site Finder modula molekularnog dokinga i molekularne dinamike.
Molekulsko modelovanje ukazuje na to da su interakcije sa Met 1038, Gln 1040, Thr 1077,
Gln 1194 i Val 1259 važan faktor postojanja afiniteta inhibitora prema XO enzimu. Naš
predloženi model vezivanja mogao bi biti od značaja za razvoj novih aktivnih inhibitora XO
zasnovanih na 1,2,3,4-tetrahidroizohinolinskom heterociklusu.
Acta Medica Medianae 2021;60(1):48-55.
Ključne reči: inhibicija ksantin oksidaze, 1,2,3,4-tetrahidroizohinolini, molekularni
doking, simulacija molekularne dinamike
This work is licensed under a Creative Commons Attribution 4.0 International (CC BY 4.0) Licence