Rational drug design and synthesis of new α-Santonin derivatives as potential COX-2 inhibitors

Article abstract of DOI:10.1016/j.bmcl.2018.02.036

Sesquiterpene compounds are widely known for their numerous pharmacological activities. Herein the focus of the authors was on α-Santonin, a sesquiterpene lactone from the Artemisia genus: the aim was to determine whether α-Santonin could be considered in the treatment of inflammation and pain. To this purpose, a small series of derivatives was designed and screened in silico against the enzyme COX-2 along with the parent compound. Drug-likeness parameters were also assessed. The compounds were eventually synthesized, and few were tested to determine their efficacy in the inhibition of COX-2 activity and expression. Overall, compound A2 was the only one with a detectable inhibitory potential of COX-2 activity whilst two of its ether derivatives demonstrated improved ability in the inhibition of COX-2 expression.

Full text of DOI:10.1016/j.bmcl.2018.02.036

Bioorganic & Medicinal Chemistry Letters xxx (2018) xxx–xxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Rational drug design and synthesis of new
potential COX-2 inhibitors
a-Santonin derivatives as
Adriana Coricello a, Asma El-Magboub b, Marian Luna c, Angela Ferrario c, Ian S. Haworth b,
Charles J. Gomer c, Francesca Aiello a, , James D. Adams b
⇑
a Dipartimento di Farmacia e Scienze della Salute e della Nutrizione, Edificio Polifunzionale, Università della Calabria, Arcavacata di Rende, 87036 Rende (CS), Italy
b Department of Pharmacology and Pharmaceutical Sciences, School of Pharmacy, University of Southern California, 1985 Zonal Avenue, Los Angeles 90089, CA, USA
c Children’s Hospital Los Angeles, University of Southern California, 4650 W Sunset Boulevard, Los Angeles 90027, CA, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Sesquiterpene compounds are widely known for their numerous pharmacological activities. Herein the
Received 20 November 2017
Revised 10 February 2018
Accepted 17 February 2018
Available online xxxx
focus of the authors was on
a-Santonin, a sesquiterpene lactone from the Artemisia genus: the aim
was to determine whether -Santonin could be considered in the treatment of inflammation and pain.
a
To this purpose, a small series of derivatives was designed and screened in silico against the enzyme
COX-2 along with the parent compound. Drug-likeness parameters were also assessed. The compounds
were eventually synthesized, and few were tested to determine their efficacy in the inhibition of COX-
2 activity and expression. Overall, compound A2 was the only one with a detectable inhibitory potential
of COX-2 activity whilst two of its ether derivatives demonstrated improved ability in the inhibition of
COX-2 expression.
Keywords:
a-Santonin
COX-2_inhibition
COX-2_expression
Inflammation
Ó 2018 Elsevier Ltd. All rights reserved.
Drug_design
Nowadays great attention is focused on natural remedies that
could help in the fight against several human diseases.1–4 The
treatment of inflammation and pain has always been of interest
for the scientific community. Nonsteroidal Anti-Inflammatory
Drugs (NSAIDs) are first-line drugs in the treatment of pain and
inflammation. Their biological targets are the enzymes Cyclooxy-
genase 1 and 2 (COX-1 and COX-2).5 NSAIDs inhibit the catalytic
activity of these enzymes and consistently decrease the production
of prostaglandins which are important mediators in the pathogen-
esis of inflammation.6 The significant side effects associated with
NSAIDs consumption spurred the researchers to investigate a
new class of molecules which were COX-2 selective inhibitors. In
past years, many molecules featuring heterocyclic rings and polar
groups have been designed and synthesized: overall, they
expressed high selectivity and low IC50 values in the inhibition
analgesic effects but it has shown promising results although the
mechanisms of action have not yet been proposed or analyzed fur-
ther.11,12 Many other properties are otherwise attributed to this
compound. Artemisia absinthium and Artemisia annua were tradi-
tionally known for their anthelminthic properties probably due
to their content of a
-Santonin.13 The compound itself was proven
to have anthelmintic activity against several strains of helminths
but due to its toxicity at a low dosage (300 mg in adults) it is
no longer used as
helminthiasis.14–16
reported the development of three derivatives of
exhibited anti-oxidant effect.17 Some data showed how more mod-
ifications of the structure of -Santonin could confer the ability to
be active against abnormal cell growth suggesting a potential role
in the treatment of cancer.18 Overall,
-Santonin was discretely
a
medicament in the treatment of
1997 research led by Kulyyasov et al.,
-Santonin that
A
a
a
a
of COX-2.7 Sesquiterpene lactones, such as Leucodin and
a-San-
investigated but when it comes to the fields of pain and inflamma-
tonin, which can be detected in many species of the Artemisia
genus, have been recently investigated and proved important effi-
cacy in the inhibition of COX-2.8–10 Sesquiterpene lactones’ multi-
ringed nucleus may be considered as an interesting scaffold for the
design of COX-2 selective inhibitors. a-Santonin, in particular, has
been briefly examined for anti-inflammatory, anti-pyretic and
tion still limited data are available.
The purpose of this study was to investigate whether an
involvement of
a-Santonin and some of its derivatives could be
considered in the treatment of inflammation via the inhibition of
COX-2 activity and expression. The development of new remedies
capable of decreasing inflammation by more than one mode of
action is still an unbeaten track. It piqued the interest of the
authors how many sesquiterpene lactones were reported to be
very active in lowering COX-2 levels in treated cells: they seem
⇑
Corresponding author.
E-mail address: francesca.aiello@unical.it (F. Aiello).
https://doi.org/10.1016/j.bmcl.2018.02.036
0960-894X/Ó 2018 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Coricello A., et al. Bioorg. Med. Chem. Lett. (2018), https://doi.org/10.1016/j.bmcl.2018.02.036

2
A. Coricello et al. / Bioorganic & Medicinal Chemistry Letters xxx (2018) xxx–xxx
Table 1
Drug-likeness parameters.
to interfere with NF-jB and MAPK pathways, which regulate the
expression of several mediators of the inflammation process,
including COX-2.19,20 After a careful assessment, the authors
decided to evaluate in the first place whether the study compounds
were in any way liable to interfere with such pathways, and hence,
if they could also be appraised for the direct inhibition of COX-2.
Compound
-Santonin
HBD(ꢀ5)
HBA(ꢀ10)
MW(<500)
LogP(<5)
a
0
0
1
0
0
0
0
0
0
3
4
3
3
3
5
4
4
5
246.3
1.81
3.33
3.35
5.35
6.27
4.05
4.54
4.69
5.09
A1
A2
A3
A4
R0
R1
R3
R5
288.34
246.30
336.42
404.42
414.51
364.43
382.42
298.88
The first approach was a comparative analysis between
tonin, its isomer -Desmotroposantonin and the synthetic inter-
mediate A1. In fact, -Desmotroposantonin (A2) was obtained
through two synthetic approaches (Scheme 1).21
-Santonin was
converted into -Desmotroposantonin acetate (A1) using acetic
anhydride and concentrated sulphuric acid. The reaction is known
as the Thiele reaction.22
-Desmotroposantonin acetate was hence
a-San-
a
a
a
a
a
de-acetylated to A2 using a mixture of methanol and ammonium
Table 2
Docking results.
hydroxide.21 Reactions’ times and purification methods were prop-
erly adjusted to maximize the yields.
obtained with an 89% yield and as a highly pure white solid as con-
firmed by mass spectrometry.
a-Desmotroposantonin was
Compound
Celecoxib
Binding Affinity (kcal/mol)
À12.5
À10.3
À9.4
À9.4
À9.4
À7.6
À7.3
À9.2
À9.1
À8.9
a-Santonin
The compounds were tested on RIF-1 cells cultures previously
treated with Photofrin-Photodynamic Therapy (PH-PDT).23 PH-
PDT is very effective in the treatment of solid tumors. The method
induces oxidative stress in treated cells, hence stimulating the pro-
duction of several cellular mediators including COX-2. Two param-
eters were evaluated to establish whether Santonin, A2 and A1
interfered with COX-2 production and functioning: COX-2 expres-
sion and PGE2 production. Results are shown in Figs. 4 and 5.
A1
A2
A3
A4
R0
R1
R3
R5
a
-Santonin demonstrated a minor ability to inhibit COX-2
expression and only at higher concentrations (500 M); moreover,
l
Tools and the molecular structures were added with polar hydro-
gens and torsion bonds defined. The files were saved as .pdbqt files.
Once the Vina Wizard was started, a Vina search box was defined
to include the COX-2 active site. Docking calculations were hence
performed and the in-target bindings were analyzed. In order to
validate the method Celecoxib was run first. Results are shown in
Table 2.
Among the three compounds, A2 appeared as the most suitable
for this study: its docking analysis showed good binding affinity
values and its contribution to COX-2 inhibition in the bioassays
the decrease of PGE2 levels (Fig. 5) was likely a direct consequence
of the lowered number of expressed enzymes. Treatment with
compound A1 caused an increased COX-2 expression at lower
dosages (400
lM) and only slightly decreased protein levels at
800 M. Despite the increase of COX-2 levels in the cells, PGE2
l
levels were importantly reduced at 200
lM and 400
lM but they
also remained dramatically high at 800
l
M. Despite the fluctua-
tions of the activity profile the data implied that compound A1
could likely be a good inhibitor of COX-2 activity. Anyway, due
to its enhancing effect on COX-2 expression at 400 lM, the com-
pound was not investigated further. Compound A2 did not show
any contribution in the reduction of COX-2 expression at lower
dosages (250 lM) whereas a marked efficiency in inhibiting PGE2
seemed the most valuable. Since
a-Desmotroposantonin can also
easily be the target of functionalizing reactions because of its
hydroxyl group, the compound was selected as a scaffold for the
design of more derivatives (Fig. 1).
production was observed at this concentration. Furthermore,
higher amounts of the administered substance were also reported
to noticeably decrease COX-2 expression levels.
After the aforementioned biological approach, Lipinski’s rule of
five parameters were evaluated to determine drug-likeness for the
tested compound. All the compounds responded positively (See
Table 1).24
A docking analysis was hence performed using Autodock Vina
Wizard in PyRx platform.25,26 The input protein file was retrieved
from the Protein Data Bank as the 3LN1 crystal structure of Cele-
coxib bound to COX-2. The macromolecule was prepared using
AutoDock Tools (ADT);27 multiple chains were deleted, water
molecules and other small molecules were removed and polar
hydrogens were added. The ligands were prepared using Chem-
Draw Ultra 12.0. The ligands’ PDB files were loaded into AutoDock
RO
O
O
R
R
O
O
R1
R3
R5
A3
A4
R0
F
F3C
O
O
O
a
b
O
O
HO
O
O
O
O
O
O
O
Santonin
A2
A1
Cl
S
Scheme 1. Reagents and conditions: a. Ac2O, H2SO4 (dropped); 0 °C. b. NH4OH:
CH3OH (1:1); rt.
Fig. 1. Designed derivatives of a-Desmotroposantonin.
Please cite this article in press as: Coricello A., et al. Bioorg. Med. Chem. Lett. (2018), https://doi.org/10.1016/j.bmcl.2018.02.036

A. Coricello et al. / Bioorganic & Medicinal Chemistry Letters xxx (2018) xxx–xxx
3
Compounds A3, A4 and R0 were proposed as O-benzyl ethers of
-Desmotroposantonin. The para-substitution on the aromatic ring
a
was meant to test how different substituents could change the
affinity of the protein-ligand interactions. Similarly, R1, R3 and
R5 were designed as oxo-phenyl-ethyl derivatives. The oxo-phe-
nyl-ethyl moiety is even bulkier than the benzyl group and the car-
bonyl was considered likely to form H-bonds with polar residues.
Lipinski’s parameters were calculated and none of the com-
pounds were reported for more than one violation of the rule of
five (Table 1).24
A docking analysis was performed for all the designed com-
pounds using the reported method. Overall, all the screened
ligands showed good binding affinity values. Special attention
was directed to compounds A3 and R1 which gave the best scores
also suggesting that the para-substitution on the phenyl ring is not
likely to improve affinity.
The results were analyzed using BIOVIN Discovery Studio.28
Compound A3 binding is stabilized by hydrophobic interactions.
The phenyl ring protrudes into the COX-2 selectivity pocket setting
up Pi- interactions with residues VAL523 and ALA516; the lactone
moiety is oriented towards the channel gate and SER353 interacts
with the carbonyl group by a carbon hydrogen bond. Overall the
compound is stabilized by van der Waals and alkyl interactions
(Fig. 2).
Fig. 3. 2D plot of compound R1 interactions with the COX-2 active site.
The pose of R1 is totally different from A3. The selectivity
pocket remains basically uninvolved in the interactions: the com-
pound orients the lactone moiety toward the back of the channel
and the phenyl ring toward the entrance thus suggesting that the
addition of a carbon in the linker between the scaffold and the aro-
matic ring could make the moiety too bulky for the selectivity
pocket. The compound is mainly stabilized by hydrophobic inter-
actions and only one hydrogen bond interaction with SER530
was detected (Fig. 3).
i
HO
RO
O
O
O
O
A2
A3, A4, R0, R1, R3, R5
Scheme 2. Reagents and conditions: ì. R-CH2-Br, K2CO3, Acetone; reflux.
All the screened compounds were synthesized according to the
method described by Chintakindi et al. for the Williamson synthe-
sis (Scheme 2).21 The conditions involved potassium carbonate as a
base to activate the nucleophile and acetone as solvent. All the
reactions were heated to reflux and run under N2 atmosphere.
The reactions’ progression was frequently monitored, and reaction
times were properly adjusted. The final products were purified and
isolated using either crystallization or column chromatography to
maximize the yields.
than those needed for the parent compound: compound A3 was
already active at 62.5 M while compound A4 showed an impor-
tant effect only at 400 M which is still lower than 500 M at
which A2 was active. Compound A3 revealed an important efficacy
against PGE2 production but, as for -Santonin, it remains unclear
whether this effect was due to a direct inhibition of the enzyme or
to its lowered expression. 500 M and 750 M dosages of A3 were
less efficacious against both protein expression and PGE2 synthesis
than lower doses, thus suggesting a potential toxicity of the com-
pound at higher concentrations. Compound A4 contribution in
lowering PGE2 levels was less important than that of A3; also,
when the histograms are compared, it appears clear that its effect
was more likely due to interactions with COX-2 expression
l
l
l
a
l
l
Compounds A3 and A4 were tested as previously described.
They both showed a marked improvement in the ability to sup-
press COX-2 expression consistently and at lower concentrations
Fig. 4. Relative optical density (OD) of COX-2 to actin at different dosages of the
Fig. 2. 2D plot of compound A3 interactions with the COX-2 active site.
tested compounds. Histograms represent the means SEM of three experiments.
Please cite this article in press as: Coricello A., et al. Bioorg. Med. Chem. Lett. (2018), https://doi.org/10.1016/j.bmcl.2018.02.036

4
A. Coricello et al. / Bioorganic & Medicinal Chemistry Letters xxx (2018) xxx–xxx
Acknowledgements
The authors would like to thank Dr. Eric J. Lien from the Univer-
sity of Southern California – Los Angeles for the time spent in sup-
porting the team and giving advice about the work.
A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at https://doi.org/10.1016/j.bmcl.2018.02.036.
References
1. Aiello F, Badolato M, Pessina F, et al. ACS Chem Neurosci. 2016;7:737.
2. Aiello F, Armentano B, Polerà N, et al. J Agric Food Chem. 2017;65:7944.
3. Aiello F, Carullo G, Badolato M, Brizzi A. ChemMedChem. 2016;11:1686.
4. Carullo G, Cappello A, Frattaruolo L, Badolato M, Armentano B, Aiello F. Future
Med Chem. 2017;9:79.
Fig. 5. Prostaglandins production after treatments with different dosages of the
tested compounds. Histograms represent the means SEM of three experiments.
5. Smith WL, DeWitt DL, Garavito RM. Ann Rev Biochem. 2000;69:145.
6. Funk CD. Science. 2001;294:1871.
7. Carullo G, Galligano F, Aiello F. Med Chem Comm. 2017;8:492.
8. Khafagy SM, Gharbo SA, Sarg TM. Planta Med. 1971;20:90.
9. Adams JD. Pharmaceuticals (Basel). 2012;5:1045.
10. Adams JD, Haworth IS, Coricello A, et al. Front Nat Prod Chem. 2017;3:176.
11. Al-Harbi MM, Qureshi S, Ahmed MM, Raza M, Miana GA, Shah AH. Jpn J
Phrmacol. 1994;64:135.
12. Martin ML, Moran A, Carron R, Montero MJ, Roman LS. J Ethnopharmacol.
1988;23:285.
13. Ferreira JFS, Peaden P, Keiser J. Parasitol Res. 2011;109:1585.
14. Satou T, Koga M, Matsuhashi R, Koike K, Tada I, Nikaido T. Vet Parasitol.
2002;104:131.
modulators rather than with COX-2 itself (Figs. 4 and 5). Sesquiter-
pene lactones have been known for their potential in suppressing
COX-2 expression and it is not excluded that a similar effect might
be associated with
a
-Santonin derivatives as well.29,30 The com-
pounds were also tested for viability after 24 h on RIF-1 cells.
Among the five, A3 turned out to be the most toxic (Data shown
in Supplementary Data Contents 6) whereas the whole experiment
provided an overall good safety profile for the tested compounds.
15. Githiori JB, Hoglund J, Waller PJ, Baker RL. Vet Parasitol. 2005;118:215.
16. Khare CP. Indian Medicinal Plants: An Illustrated Dictionary. New York: Springer
Science & Business Media; 2008.
17. Kulyyasov AT, Seitembetov TS, Adekenov SM. Chem Nat Compd. 1997;33:185.
18. Kim SH, Song JH, Choi BG, Kim HJ, Kim TS. J Biol Chem. 2006;281:13117.
19. Tamura R, Morimoto K, Hirano S, et al. Biosci Biotechnol Biochem. 2012;76:2360.
20. Hwang D, Fischer NH, Jang BC, Tak H, Kim JK, Lee W. Biochem Biophys Res
Commun. 1996;226:810.
21. Chintakindi PK, Sangwan PL, Farooq S, et al. Eur J Med Chem. 2013;60:365.
22. Colombo MI, Testero SA, Pellegrinet SC, Bohn ML, Rùveda EA. Chem Educ.
2001;6:350.
23. Dougherty TJ, Gomer CJ, Henderson BW, et al. J Natl Cancer Ist. 1998;90:889.
24. Lipinski CA. Drug Discov Today. 2004;4:337.
25. Trott O, Olson AJ. J Comput Chem. 2010;31:455.
26. Dallakyan S, Olson AJ. Methods Mol Biol. 2015;1263:243.
27. Morris GM, Huey R, Lindstrom W, et al. J Comput Chem. 2009;16:2785.
28. Dassault Systèmes BIOVIA. Discovery Studio, R2. San Diego: Dassault Systèmes;
2017.
29. Schmidt BM, Belolipov IV, Kurmukov AS, Zarikov S, Raskin I. US Patent 2008/
0145465A1, 2008.
Conclusions
More analyses are still on-going over the other derivatives but,
overall, it seems likely, at least according to the theoretical predic-
tions, that the implementation of the scaffold with an unsubsti-
tuted aromatic ring could improve the binding affinity and hence
the inhibitory activity against COX-2, while the addition of bulkier
moieties would make it harder for the ligand to make good inter-
actions inside the selectivity pocket.
On the other hand, the biological profile shows that the parent
compound A2 remains the best candidate as a direct inhibitor of
the enzyme while the two bulkier derivatives A3 and A4 rather
appear to be involved in the interference with COX-2 expression
pathways.
30. Nam YJ, Lee DH, Lee MS, Lee CS. N-S Arch Pharmakol. 2015;388:921.
Please cite this article in press as: Coricello A., et al. Bioorg. Med. Chem. Lett. (2018), https://doi.org/10.1016/j.bmcl.2018.02.036