Structure activity relationship exploration of 5-hydroxy-2-(3-phenylpropyl)chromones as a unique 5-HT2B receptor antagonist scaffold

Article abstract of DOI:10.1016/j.bmcl.2020.127511

Antagonists for the serotonin receptor 2B (5-HT_2B) have clinical applications towards migraine, anxiety, irritable bowl syndrome, and MDMA abuse; however, few selective 5-HT_2B antagonists have been identified. Previous studies from these labs identified a natural product, 5-hydroxy-2-(2-phenylethyl)chromone (5-HPEC, 2) as the first non-nitrogenous ligand for the 5-HT_2B receptor. Studies on 5-HPEC optimization led to the identification of 5-hydroxy-2-(3-phenylpropyl)chromone (5-HPPC, 3), which showed a tenfold improvement in binding affinity over 2 at 5-HT_2B. This study aimed to further improve receptor pharmacology of this unique scaffold. Guided by molecular modeling studies modifications at the C-3′ and C-4′ positions of 3 were made to probe their effects on ligand binding affinity and efficacy. Among the derivatives synthesized 5-hydroxy-2-(3-(3-cyanophenyl)propyl)chromone (5-HCPC, 3d) showed the most promise with a multifold improvement in binding affinity (pK_i = 7.1 ± 0.07) over 3 with retained antagonism.

Full text of DOI:10.1016/j.bmcl.2020.127511

Bioorganic&MedicinalChemistryLetters30(2020)127511
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Structure activity relationship exploration of 5-hydroxy-2-(3-phenylpropyl)
chromones as a unique 5-HT2B receptor antagonist scaffold
Minsoo Kima, Myles Trussa, Piyusha P. Pagareb, Martha A. Essandohc, Yan Zhangb,
⁎
Dwight A. Williamsa,
a Department of Chemistry, Kalamazoo College, Kalamazoo, MI 49006, USA
b Department of Medicinal Chemistry, Virginia Commonwealth University, Richmond, VA 23298, USA
c Department of Chemistry, Oklahoma State University, Stillwater, OK 74078, USA
A R T I C L E I N F O
A B S T R A C T
Keywords:
Antagonists for the serotonin receptor 2B (5-HT2B) have clinical applications towards migraine, anxiety, irritable
bowl syndrome, and MDMA abuse; however, few selective 5-HT2B antagonists have been identified. Previous
studies from these labs identified a natural product, 5-hydroxy-2-(2-phenylethyl)chromone (5-HPEC, 2) as the
first non-nitrogenous ligand for the 5-HT2B receptor. Studies on 5-HPEC optimization led to the identification of
5-hydroxy-2-(3-phenylpropyl)chromone (5-HPPC, 3), which showed a tenfold improvement in binding affinity
over 2 at 5-HT2B. This study aimed to further improve receptor pharmacology of this unique scaffold. Guided by
molecular modeling studies modifications at the C-3′ and C-4′ positions of 3 were made to probe their effects on
ligand binding affinity and efficacy. Among the derivatives synthesized 5-hydroxy-2-(3-(3-cyanophenyl)propyl)
chromone (5-HCPC, 3d) showed the most promise with a multifold improvement in binding affinity
5-HT2B
Chromone
Antagonist
5-HPEC
5-HPPC
G-protein coupled receptors
(pKi = 7.1
0.07) over 3 with retained antagonism.
The neurotransmitter serotonin or 5-hydroxytryptamine (5-HT (1),
Fig. 1) plays a role in a vast array of biological functions. Studies on 1
have confirmed its clinical relevance in most physiological systems,
including gastrointestinal, cardiovascular and nervous systems.1 The
ability of a single molecule like serotonin to participate in mediating
such a diverse array of physiological functions can be attributed to its
complex receptor pharmacology. Signaling by 1 is mediated by at least
15 different serotonin receptor subtypes which are classified into 7
receptor subfamilies (5-HT1–7). All of them are G protein-coupled re-
ceptors with the exception being the ligand gated ion channel 5-HT3.2,3
The 5-HT2 receptor family presently has three identified subtypes 5-
HT2A, 5-HT2B, and 5-HT2C, and has been shown to regulate sleep, ap-
petite, cardiovascular function, muscle contractions, and the halluci-
nogenic activity of phenylalkylamines.4 Accordingly, ligands targeting
this receptor family carry significant clinical importance. For example,
antagonists for the 5-HT2A and 5-HT2C have shown potential as anti-
psychotics and antidepressants, respectively.5,6 Antagonist for 5-HT2B
have been explored as treatments for migraines, heart diseases, and
substance abuse.7,8 One of the challenges in realizing the clinical po-
tential of the 5-HT2 receptor family is a lack of selective ligands for one
subtype over the others.8,9 Thus, developing ligands targeting specific
subtypes within the 5-HT2 receptor family may yield useful scaffolds
upon which to design novel chemotherapeutics.
Previous work from our labs identified the natural product, 5-hy-
droxy-2-(2-phenylethyl)chromone (5-HPEC (2), Fig. 1), as an inter-
esting scaffold for the development of novel 5-HT2B selective antago-
nists.10 In an effort to understand the structural features affecting the
receptor pharmacology of this scaffold at 5-HT2B a small library of
analogues were designed and synthesized. Homologation of the C-2
alkyl chain to afford 5-hydroxyl-2-(3-phenylpropyl)chromone (5-HPPC,
3), improved ligand binding affinity at 5-HT2B ten-fold (pKi = 5.6, (2)
to 6.6, (3)) while also retaining moderate antagonism
(pIC50 = 5.05
0.05).10,11 The work presented here discloses our
continued efforts to improve 5-HT2B receptor pharmacology for this
class of molecules.
To guide further molecular design, in silico modeling studies of 3
within the putative binding pocket of the 5-HT2B receptor (PDB ID
4IB4) were undertaken to identify potential molecular interactions that
could be exploited to further improve ligand binding affinity. A genetic
12
algorithm GOLD 5.6 was used to explore docking poses of 3 at 5-HT2B
.
The generated poses were scored using CHEMPLP and GOLD scores.
Clustering of the docking poses for 3 showed one high scoring cluster
family and the highest ranked pose is shown in Fig. 2. Within the
identified binding pocket, the chromone core of 3 occupies a
⁎ Corresponding author at: 1200 Academy Street, Kalamazoo, MI 49006, USA.
E-mail address: dwight.williams@kzoo.edu (D.A. Williams).
https://doi.org/10.1016/j.bmcl.2020.127511
Received 28 April 2020; Received in revised form 17 August 2020; Accepted 18 August 2020
Availableonline24August2020
0960-894X/©2020ElsevierLtd.Allrightsreserved.

M. Kim, et al.
Bioorganic&MedicinalChemistryLetters30(2020)127511
Fig. 1. Molecular structures of 5-HT (1), 5-
HPEC (2), and 5-HPPC (3).
condensation between acetophenone 6 and esters 5a-l followed by an
acid catalyzed cyclization afforded the anticipated 5-HPPC (3) deriva-
tives 3a-l with substituents at C-3′ and C-4′ as outlined in Scheme 2.
Characterization of receptor pharmacology for the newly synthe-
sized derivatives was done in a three-stage process.17 In the initial
stage, each compound was screened for its ability to inhibit radioligand
binding at a fixed concentration of 10 μM at each of the 5-HT2 receptor
subtypes. At this concentration significant inhibition was
deemed ≥ 50%. Results from this screening are summarized in Table 1.
Most compounds maintained inhibitory activity at 5-HT2B comparable
to 3. In general it was observed that the C-3′ derivatives had greater
inhibitory activities than the C-4′ derivatives in this assay. From this
initial data it was also observed that the more hydrophobic groups
substituents such as those in 3a and 3g decreased selectivity between 5-
HT2B/2C while more polar substituents such as those in 3b and 3h
provided better selectivity towards 5-HT2B. The C-3′ methoxy analogue
3k showed greater inhibitory activity (75.3%) at the 5-HT2B than 3
though selectivity seemed to be lost over the 5-HT2C (87.9%). The C-4′
methoxy analogue 3f showed a similar loss in selectivity indicating
larger hydrophobic electron donating groups though potentially toler-
ated may reduce selectivity. The halogen containing derivatives 3c, 3e,
3i and 3j with substitutions at either C-3′ or C-4′ positions showed
decreased inhibitory activity and loss of selectivity when compared to 3
suggesting that highly electronegative groups may be unfavorable while
also supporting previous observations that increased hydrophobicity
decreases selectivity. The most promising analogue 5-hydroxy-2-(3-(3-
cyanophenyl)propyl)chromone (5-HCPC, 3d) showed significantly
greater inhibitory activity (91.1%) than the parent compound 3 while
still maintaining an apparent selectivity for the 5-HT2B over 5-HT2A/2C
(7.2%/15.1%). It is interesting that the cyano group in 3d is electron
withdrawing in character like that of the halogens however it is less
lipophilic; thus, suggesting that polar hydrophilic groups on C-3′ may
be favored for improved binding affinity.
Fig. 2. Plausible ligand binding mode of 3 in the 5-HT2B receptor. 5-HT2B re-
ceptor is shown in grey cartoon while the amino acid residues involved in in-
teractions are shown in grey stick representations. 3 is shown in stick re-
presentations (cyan). Hydrogen bonding interactions are shown in yellow
dashes.
hydrophobic pocket formed by Val136, Phe217, Phe340, Phe341, and
Val336. Hydrogen bonding interactions between the C-5 hydroxyl
group and the C-4 carbonyl with Thr140 and Ser139, respectively, may
help anchor the scaffold within the binding pocket. The pendant phenyl
ring of the C-2 alkyl chain seems oriented towards a largely hydro-
phobic pocket composed of Val366, Trp131, and Leu132. Given such an
observation, it was hypothesized that higher hydrophobicity would
further improve the binding affinity for this scaffold. Thus, it was de-
termined that initial structural modification on 3 would be focused on
the pendant phenyl ring with the substituents being chosen selected
from the Craig plot.13 In accordance with this decision a series of
analogues with substituents selected from each quadrant of the Craig
plot at either C-3′ or C-4′ of the phenyl ring were implemented.
Chemical syntheses of the newly designed 5-HPPC (3) derivatives
was accomplished using an established route from our labs.14 The
needed butanoate esters 5a-l were prepared via Heck reaction between
the appropriate benzyl chloride and ethyl acrylate followed by hydro-
genation (Scheme 1).15,16 With the needed esters in hand, Claisen
Following the initial screening, compounds showing significant in-
hibitory activity were moved forward and their binding affinities at the
given receptor determined. Binding affinity results are shown in
Table 2. Interestingly, the enhanced inhibitory activity of 3k did not
translate into a significant increase in affinity at the 5-HT2B but showed
comparable affinity to 3 (pKi = 6.0). The loss of selectivity inferred
from the initial screening was also observed in the secondary assay as
3k had a pKi = 6.1 at the 5-HT2C. The most encouraging data from this
Scheme 1. Heck reaction followed by hydrogenation of benzylchlorides yielded esters. Reagents and conditions: (i) ethyl acrylate, Pd(OAc)2, Bu3N, reflux 15 h,; (ii)
H2, 50 psi, Pd/C, EtOH, rt 2 h. (Yields 23–85%).
2

M. Kim, et al.
Bioorganic&MedicinalChemistryLetters30(2020)127511
Scheme 2. Claisen condensation and cyclization of 6 and 5a-l yielded derivatives 3a-l. Reagents and conditions: (i) NaH, THF, reflux 1 h; (ii) 5a-l, THF, reflux 4 h;
(iii) HCl, MeOH, reflux 45 min. (Yields 2–67%).
Table 1
Percent inhibition data of compounds 3a-l obtained from primary radioligand binding assays at 10 µM (n = 4).
Compound
% Inhibition
Compound
% Inhibition
5-HT2B
5-HT2B
5-HT2A
5-HT2C
5-HT2A
5-HT2C
2
62.6
57.7
65.8
61.7
91.1
52.7
69.3
14.2
14.4
8.7
4.8
59.4
43.5
44.8
15.5
44.4
67.4
3
70.4
50
5.4
7.9
3a
3g
3h
3i
3j
3k
3l
17.2
19.5
9.3
51.2
36
3b
50
3c
13.9
7.2
50.4
49.2
75.3
35.3
35.8
33.9
87.9
32.6
3d (5-HCPC)
14.2
−1.6
−5
3e
3f
8.4
9.8
Table 2
pKi values of compounds showing ≥ 50% inhibition at the 5-HT2B in
the primary assays.
Compound
pKi at 5-HT2B (pKi at 5-HT2C)
2
5.6
3
6.6
3a
5.7 (5.5)
6.0
3b
3c
5.7
3d (5-HCPC)
7.1
3e
3f
5.6
5.9 (5.7)
5.8 (5.5)
5.6
3g
3h
3i
5.6
3k
6.0 (6.1)
assay was the significant improvement in inhibitory activity observed in
Fig. 3. Plausible ligand binding mode of 3d in the 5-HT2B receptor. 5-HT2B
receptor is shown in grey cartoon while the amino acid residues involved in
interactions are shown in grey stick representations. 3d is shown in stick re-
presentations (pink) Hydrogen bonding interactions are shown in yellow da-
shes.
the primary assay for 3d correlated to a 4-fold improvement in binding
affinity (pKi = 7.1
0.07) over 3.
With 5-HCPC showing improved affinity at 5-HT2B, its functional
activity at 5-HT2B was then determined. In the calcium mobilization
agonist assay, 3d did not display any significant agonist activity. The
pEC50 for 3d was determined to be 6.3
0.11 compared to the en-
0.05, which is a 40-fold
Val136, Phe217, Phe340, Phe341, and Val336 with the C-5 hydroxyl
group showing potential hydrogen bonding interactions with Thr140
and the C-4 carbonyl showing potential hydrogen bonding interactions
with Ser139 (Fig. 3). The C-2 alkyl chain of 3d appeared to be directed
towards the cytoplasmic surface of the receptor. The phenyl ring sub-
stituents of 3d seemed positioned within the hydrophobic pocket
formed between extracellular loop 2 and helix 3 consisting of Trp131,
Leu132, Val208, and Leu209 residues. It seemed that the C-3′ cyano
group in 3d may form hydrogen bonding interaction with the hydroxy
group of Tyr370 on TM7. This additional hydrogen bonding interaction
could be perceived as a favorable interaction and helped provide an
explanation for the increased affinity of 3d over 3.
dogenous ligand 5-HT with a pEC50 of 10.15
difference in the EC50 values. Additionally, the maximal response for 3d
was only 12% of the maximal response at a concentration of 10 μM
compared to the maximal response seen by the endogenous agonist 1 at
the significantly lower concentration of 3.0 nM. These data taken to-
gether suggest that 3d may not act as an agonist at 5-HT2B. In the an-
tagonist assay an EC50 dose of 5-HT (1.6 nM) was challenged with
various concentrations of 3d. It was determined that the pIC50 for 3d is
5.2
0.31 which was a moderate improvement over the original lead
0.05).
(pIC50 = 5.05
To help rationalize the improved binding affinity of 3d at 5-HT2B
,
molecular docking studies were conducted using the same approach as
for 3.12 3d showed very similar binding modes to that of 3. Its chro-
mone ring occupied the same hydrophobic pocket formed by Val136,
In summary, this study aimed to improve the in vitro pharmacology
of the 5-HPPC (3) scaffold with a focus on the 5-HT2B receptor. A series
3

M. Kim, et al.
Bioorganic&MedicinalChemistryLetters30(2020)127511
of derivatives were synthesized via an efficient route and their receptor
pharmacology examined. While several compounds showed compar-
able results to the original lead compound, one new compound 3d
showed the most promise with a 4-fold increase in binding affinity at
the 5-HT2B compared to 3 while retaining antagonist functional ac-
tivity. Ligand docking studies suggested this improvement in binding
affinity may be due to favorable hydrogen bonding interactions of the
Psychoactive Drug Screening Program, Contract # HHSN-271-2013-
00017-C (NIMH PDSP).
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.bmcl.2020.127511.
C-3′ cyano group with the hydroxy group of Tyr370 of the 5-HT2B
.
Further optimization of this new lead is underway and will be reported
in due course.
References
1. Berger, M.; Gray, J. A.; Roth, B. L. In Annual Review of Medicine; Annual Reviews:
Palo Alto, 2009; Vol. 60, p 355.
Declaration of Competing Interest
2. Roth BL. Ann Clin Psychiatry. 1994;6:67.
3. Kroeze WK, Kristiansen K, Roth BL. Curr Top Med Chem. 2002;2:507.
4. Roth BL, Willins DL, Kristiansen K, Kroeze WK. Pharmacol Ther. 1998;79:231.
5. Heisler LK, Zhou L, Bajwa P, Hsu J, Tecott LH. Genes Brain and Behavior. 2007;6:491.
6. Eison AS, Mullins UL. Behav Brain Res. 1996;73:177.
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
ence the work reported in this paper.
7. Brea J, Castro-Palomino J, Yeste S, et al. Curr Top Med Chem. 2010;10:493.
8. Poissonnet G, Parmentier JG, Boutin JA, Goldstein S. Mini-Rev Med Chem.
2004;4:325.
Acknowledgments
9. Shyu K-G. Circ Res. 2009;104:1.
10. Williams DA, Zaidi SA, Zhang Y. Bioorg Med Chem Lett. 2014;24:1489.
11. Williams DA, Zaidi SA, Zhang Y. J Nat Prod. 1859;2015:78.
12. Jones G, Willett P, Glen RC, Leach AR, Taylor R. J Mol Biol. 1997;267:727.
13. Craig PN. J Med Chem. 1971;14:680.
The authors would like to thank the Thomas J. Smith Student
Research Fellowship for Stipend support (MK), the Kalamazoo College
Department of Chemistry Start-up Funds (DAW), and The Course-Hero-
Woodrow Wilson Fellowship for Excellence in Teaching (DAW). This
work was also partially supported by NIH/NIDA Grants DA024022,
DA044855 and DA050311 (Y.Z.). Receptor binding profiles were gen-
erously provided by the National Institute of Mental Health's
14. Williams DA, Smith C, Zhang Y. Tetrahedron Lett. 2013;54:4292.
15. Heck RF, Nolley JP. J Org Chem. 1972;37:2320.
16. Ghosh AK, Nicponski DR. Org Lett. 2011;13:4328.
17. Jensen NH, Roth BL. Comb Chem High Throughput Screen. 2008;11:420.
4