Antimalarial activity of tetrahydro-β-carbolines targeting the ATP binding pocket of the Plasmodium falciparum heat shock 90 protein

Article abstract of DOI:10.1016/j.bmcl.2020.127502

A series of tetrahydro-β-carboline derivatives of a lead compound known to target the heat shock 90 protein of Plasmodium falciparum were synthesized and assayed for both potency against the parasite and toxicity against a human cell line. Using a rationa

Full text of DOI:10.1016/j.bmcl.2020.127502

Bioorganic&MedicinalChemistryLetters30(2020)127502
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Antimalarial activity of tetrahydro-β-carbolines targeting the ATP binding
pocket of the Plasmodium falciparum heat shock 90 protein
⁎
Scott Eagona, , Jared T. Hammillb, Jordan Bacha, Nikalet Eversonc, Tyler A. Sisleyd,e,
Michael J. Wallsf, Sierra Durhamg, Dylan R. Pillaih, Mofolusho O. Faladeb, Amy L. Riceb,
Joshua J. Kimballc, Horacio Lazaroc, Celine DiBernardoa, R. Kiplin Guyb
a Department of Chemistry and Biochemistry, California Polytechnic State University, San Luis Obispo, CA 93407, USA
b Department of Pharmaceutical Sciences, University of Kentucky, Lexington, KY 40508, USA
c Promega Biosciences, 277 Granada Drive, San Luis Obispo, CA 93401, USA
d Department of Chemistry and Chemical Biology, Harvard University, Cambridge, MA 02138, USA
e Department of Microbiology, Harvard Medical School, Boston, MA 02115, USA
f Department of Public Health, Tauro University, Vallejo, CA 94592, USA
g Department of Food Science and Technology, University of California Davis, Davis, CA 95616, USA
h Department of Pathology and Laboratory Medicine, University of Calgary, Calgary, Alberta, Canada
A R T I C L E I N F O
A B S T R A C T
Keywords:
A series of tetrahydro-β-carboline derivatives of a lead compound known to target the heat shock 90 protein of
Plasmodium falciparum were synthesized and assayed for both potency against the parasite and toxicity against
a human cell line. Using a rationalized structure based design strategy, a new lead compound with a potency two
orders of magnitude greater than the original lead compound was found. Additional modeling of this new lead
compound suggests multiple avenues to further increase potency against this target, potentially paving the path
for a therapeutic with a mode of action different than any current clinical treatment.
Malaria
Plasmodium falciparum
Tetrahydro-β-carbolines
Structure activity relationship
Structure based design
Malaria is caused by protozoan parasites of the Plasmodium genus,
with P. falciparum and P. vivax responsible for vast majority of cases. In
2018, an estimated 228 million cases and 405,000 deaths were caused
by Plasmodium, with most deaths occurring in young children and
pregnant women.1 P. falciparum is by far the most deadly, accounting
for approximately 90% of all deaths worldwide.2 Despite a world-wide
effort to combat this deadly disease, a number of clinical challenges
have prevented further progress in the battle to eliminate the disease. In
particular, the absence of a clinically useful vaccine,3 the complexity of
the Plasmodium lifecycle,4 emergence of strains resistant to artemisinin-
based combination therapies (ACTs)5 and a small number of validated
small molecule targets6 underscore the need for treatments with new
modes of action.
90 protein (PfHsp90) transcription rate to more than triple.8 The ATP
binding pocket of PfHsp90 is highly conserved among a number of
clinical isolates, underscoring the critical cellular function of this cha-
perone protein.9 A few small molecule inhibitors of PfHsp90, most
notably geldanamycin, are known to be potent anti-malarial agents.
However, unacceptable toxicity profiles preclude the use of these
compounds in the clinic.10 Still, the critical role of PfHsp90 in parasite
viability and the fact that no current clinical treatment possesses a si-
milar mechanism of action both make this an attractive small-molecule
target.
Our group previously reported that the β-carboline harmine binds
selectively to the PfHsp90 ATP binding domain relative to human
Hsp90.11 We developed a microwave-mediated methodology to readily
prepare a library of tetrahydro-β-carboline and β-carboline derivatives
of the Harmine scaffold12 and tested a number of compounds to assay
their antimalarial activity. Our most potent compound from this initial
collection was 1, a 5-chloro-tetrahydro-β-carboline derivative of har-
mine that displayed modest activity and low toxicity against human
cells (Fig. 1).13
Heat shock proteins (Hsp) are a major class of chaperone proteins
found in all life forms ranging from prokaryotes to humans with heat
shock protein 90 (Hsp90) being one of the most common of all such
proteins.7 Since the malaria parasite must be transmitted from a poi-
kilothermic mosquito to a homothermic human, heat shock proteins are
vital for the parasite to survive the change in temperature. In fact, a few
degrees of temperature change can cause the P. falciparum heat shock
To design a more potent second generation of inhibitors, we first
⁎ Corresponding author.
E-mail address: seagon@calpoly.edu (S. Eagon).
https://doi.org/10.1016/j.bmcl.2020.127502
Received 14 July 2020; Received in revised form 12 August 2020; Accepted 16 August 2020
Availableonline18August2020
0960-894X/©2020ElsevierLtd.Allrightsreserved.

S. Eagon, et al.
Bioorganic&MedicinalChemistryLetters30(2020)127502
most deeply embedded end of the ADP molecule, making polar contacts
with residues Thr171 and Asp79. Below the adenine rings, isoleucine
and phenylalanine residues form an aliphatic floor, although ADP itself
does not appear to interact with these hydrophobic residues. The more
polar phosphate end of ADP makes several polar contacts with Arg98
and Asn37, and extends away from the depths of the pocket, pre-
sumably exposed to solvent. Given that both adenine and harmine share
a similar fused 5- and 6-membered heterocyclic system, we hypothe-
sized that 1 was binding in similar fashion, presumably making polar
contact with the same Asp79 and Thr171 residues (Fig. 2). While this
binding model is highly speculative, it suggests two potential paths to
increase potency. One path would be substituting larger aromatic/ali-
phatic groups at the methyl position of 1 to take advantage of the ali-
phatic floor formed by Ile77 and Phe124. A second approach could be
the introduction of new groups at the 5-chloro position to take ad-
vantage of the polar residues that interact with the phosphate group of
the native ligand. Since it is synthetically easier to vary the methyl
position of 1, we decided to concentrate on synthesizing a library of
compounds with various phenyl rings in place of the methyl group of 1.
Fig. 1. Structure, cytotoxicity and antimalarial activity of initial lead com-
pound 1.
attempted to rationalize a pharmacophore model of 1. Fortunately, a
solved protein structure for the N-terminus region of PfHsp90 bound to
ADP is available (PDB ID: 3 k60, Fig. 2). The adenine heterocycle is the
Fig. 2. Crystal structure of ADP bound to PfHsp90. (A) Three dimensional view of the top of the binding pocket surface of PfHsp90 with ADP. Red protein surfaces are
oxygen atoms, blue are nitrogen atoms, and grey are carbon atoms. ADP is shown with a yellow backbone. (B) Binding pocket shown with the protein as a ribbon. (C)
Two dimensional representation of ADP bound to PfHsp90. ADP is shown in blue, polar contacts are indicated by dotted red lines, and aliphatic regions indicated
with pink. (D) Putative binding model of 1.
2

S. Eagon, et al.
Bioorganic&MedicinalChemistryLetters30(2020)127502
the (R) and (S) forms of 7, with both programs predicting nearly
identical poses for each enantiomer (Fig. 3). We were pleased to see
that the computed binding pose of (R)-7 fit as predicted, with the meta-
tolyl ring residing in the aliphatic pocket and likely polar contacts be-
tween the piperidine nitrogen and the nearby Asp79 and Thr171 re-
sidues. The harmine core of (R)-7 also resides in roughly the same
position as the adenine ring of ADP. The (S)-7 pose, however, was
markedly different, with the harmine core rotated away from the
pocket such that the indole nitrogen appears to interact with Asp79
while the piperidine nitrogen continues to maintain contact Thr171.
Moreover, the chlorine atom appears to be orientated toward Arg98,
suggesting a possible ion–dipole interaction. In addition to the different
predicted poses, both programs also predicted stronger binding for the
(R) form, with Vina favoring (R)-7 by 0.70 kcal/mol, and iDock fa-
voring (R)-7 by 0.72 kcal/mol. While these results may suggest that the
(R) form is more potent, we do not believe that our simulations alone
are sufficient to make this conclusion without additional in vitro work
utilizing enantiopure compounds.
Scheme 1. Preparation of ring substituted derivatives of 1.
We also investigated N-alkylation at the indole nitrogen with the goal of
increasing the non-polar interactions.
To prepare analogs of 1 in which the methyl group was replaced by
a variety of ring systems we utilized a previously reported synthesis of
tetrahydro-β-carbolines via a microwave mediated Pictet-Spengler re-
action11 (Scheme 1).
To synthesize derivatives of 1 that were alkylated at the indole ni-
trogen, a protection-alkylation-deprotection strategy was used, as out-
lined in Scheme 2. While we were able to successfully isolate a me-
thylated and ethylated derivative, attempts to alkylate using this
methodology with larger alkyl groups lead to only trace product which
could not be isolated in sufficient amounts to screen. While additional
synthetic strategies could potentially yield more derivatives, we
decided to screen the two compounds we had on hand first before ex-
panding upon other N-alkylated products by alternative routes.
We then tested the proliferation inhibitory activity of our com-
pounds against the chloroquine-sensitive P. falciparum 3D7 strain
(MRA-102) in O-positive erythrocytes (Table 1). We also tested these
compounds for inhibition of proliferation of a human fibroblast cell line
(BJ).
Our simulations provide a rational SAR model that explains the
antimalarial activity of our compounds. Replacing the methyl group of
1 with a phenyl ring allows for additional hydrophobic interactions,
giving most phenyl-substituted derivatives improved activity. The re-
latively small size of the hydrophobic floor formed by Ile77 and Leu93
explains why substituting a larger benzyl or napthyl group is unfavor-
able, as shown by the lower activity displayed by 12–14. The meta-
methyl group in lead compound 7 appears to fit tightly in the aliphatic
floor of the binding pocket in close proximity to Ile77. This close
proximity and aliphatic environment provide a reasonable explanation
for why other phenyl substitutions alter the binding orientation due to
steric and/or electronic repulsion, effectively weakening the putative
hydrogen bonding interactions with Thr171 and Asp79 which is the
likely driver of binding affinity. Alkylation of the indole nitrogen also
causes significant loss of activity either through unfavorable steric in-
teractions or the loss of the hydrogen bonding interaction predicted
between the (S) form and Asp79. Our docking models also suggest that
the chlorine atom, which points away from the binding pocket, could
serve as a useful functional handle for the addition of substituents
which could form more favorable interactions with Arg98.
The majority of phenyl-substituted derivatives displayed improved
potency, typically around one order of magnitude better than 1, with 7
presenting the best potency (352 nM) and an improvement of two or-
ders of magnitude compared to 1. Larger substituents including nap-
thyl, benzyl, and cyclohexyl derivates showed little improvement, and
indole N-alkylated compounds were completely inactive. Most com-
pounds displayed little to no effect on the proliferation of the human BJ
cell line.
To predict the binding mode of our new analogs, we docked our lead
compound to the ATP pocket of PfHsp90 with the Vina14 and iDock15
programs. Default parameters were used for both docking programs as
longer search times have demonstrated little improvement against
known benchmarks.16 Since our compounds are chiral, we docked both
In conclusion, we were able to rationally design and synthesize a
library of tetrahydro-β-carboline derivatives of our original lead com-
pound 1 by appending various aromatic substitutions in order to make
Scheme 2. Preparation of N-alkylated derivatives of 1.
3

S. Eagon, et al.
Bioorganic&MedicinalChemistryLetters30(2020)127502
Table 1
Bioactivity, therapeutic index (T.I.) and cLogP values for all tested compounds.
Compound
R1
R2
EC50 3D7a (μM)
EC50 BJb (μM)
T.I.
cLogPc
1
2
3
H
H
H
Me
Ph
12.2
1.67
1.94
2.3d
1.7
–
–
2.05
3.45
4.00
> 24
> 24
14
12
1.09
4
5
H
H
2.53
13.3
0.75
23.6
23
0.744
9
2
4.00
4.00
0.495
0.35
6
H
2.08
2.03
> 24
12
3.93
7
8
9
H
H
H
0.352
17.9
4.35
0.487
0.088
1.22
> 24
> 24
15.7
68
1
3.93
3.93
3.06
4.89
4
10
11
H
H
7.53
4.45
1.0
> 24
> 24
3
5
3.06
3.06
0.423
12
H
7.06
2.34
> 24
3
4.44
13
14
H
H
16.6
10.9
2.03
1.15
17.2
3.38
1
2
4.44
3.72
> 24
(continued on next page)
4

S. Eagon, et al.
Bioorganic&MedicinalChemistryLetters30(2020)127502
Table 1 (continued)
Compound
R1
H
R2
EC50 3D7a (μM)
EC50 BJb (μM)
> 24
T.I.
2
cLogPc
3.69
15
12.1
5.25
16
17
Me
Et
Me
Me
> 18
> 18
> 24
> 24
1
1
2.29
2.62
a
EC50 values are reported as the mean
SD of five measurements. Assay endpoints were normalized from 0% (DMSO only) to 100% inhibition (16 µM
SD of six measurements. Assay endpoints were normalized from 0% (DMSO only) to 100% inhibition (58 µM
Mefloquine).
b
EC50 values are reported as the mean
Idarubicin).
c
d
Calculated log P of the free base.
Reported previously. See Ref. 12.
additional non-polar contacts. Our new front runner compound, 7,
displays a 35-fold increase in potency inhibiting proliferation of P.
falciparum and no detectable antiproliferative activity against a human
cell line. Docking simulations indicate that the most potent stereo-
isomer of 7 maintains the binding orientation of 1, taking advantage of
additional non-polar contacts in the ATP pocket of PfHsp90. While the
gains in potency are notable, several challenges still remain with this
scaffold. Future studies will include investigating other meta-substituted
groups that are less sensitive to metabolism, additional functionaliza-
tion utilizing the chloro group as a handle for coupling reactions,
resolution of the two enantiomers such that they can be assayed in-
dependently, or oxidizing the compound to the achiral β-carboline
system using a gentle oxidizing agent such as Ag2CO3 or IBX.17
Declaration of Competing Interest
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.
Fig. 3. Computed docking pose of lead compound 7 in the ATP pocket of PfHsp90. (A) Docking pose of (R)-7 as calculated by Vina (gold backbone) and iDock (teal
backbone). RMSD = 0.267 Å. (B) 2D representation for (R)-7, with polar interactions indicated by dotted red lines, and non-polar interactions shown as pink lines.
(C) Docking pose of (S)-7 as calculated by Vina (gold backbone) and iDock (teal backbone). RMSD = 0.096 Å. (D) 2D representation for (S)-7, with polar interactions
indicated by dotted red lines, and non-polar interactions shown as pink lines.
5

S. Eagon, et al.
Bioorganic&MedicinalChemistryLetters30(2020)127502
Acknowledgements
Micro. 2018;16:156–170.
6. Gamo FJ. Antimalarial drug resistance: new treatment options for Plasmodium. Drug
Dis Today: Technol. 2014;11:81–88.
This work was supported by the Center for Applied Biotechnology at
California Polytechnic State University, the Bill and Linda Frost Fund
and Grand Challenges Canada.
7. Shrestha L, Bolaender A, Patel HJ, Taldone T. Heat Shock Protein (HSP) Drug
Discovery and Development: Targeting Heat Shock Proteins in Disease. Curr Top Med
Chem. 2016;16:2753–2764.
8. Su XZ, Wellems TE. Sequence, Transcript Characterization and Polymorphisms of a
Plasmodium falciparum Gene Belonging to the Heat-Shock Protein (Hsp) 90 Family.
Gene. 1994;151:225–230.
Appendix A. Supplementary data
9. Shahinas D, Liang M, Datti A, Pillai DR. A Repurposing Strategy Identifies Novel
Synergistic inhibitors of Plasmodium falciparum Heat Shock Protein 90. J Med Chem.
2010;53:3552–3557.
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.bmcl.2020.127502.
10. Taldone T, Gozman A, Maharaj R, Chiosis G. Targeting Hsp90: Small-Molecule
Inhibitors and their Clinical Development. Curr Opin Phamacol. 2008;8:370–374.
11. Shahinas D, Macmullin G, Benedict C, Crandall I, Pillai DR. Harmine is a Potent
Antimalarial Targeting Hsp90 and Synergizes with Chloroquine and Artemisinin.
Antimicrob Agents Chemother. 2012;56:4207–4213.
References:
12. Eagon S, Anderson MO. Microwave-Assisted Synthesis of Tetrahydro-β-carbolines
and β-Carbolines. Eur J Org Chem. 2014;8:1653–1665.
1. World Malaria Report 2019. World Health Organization; Geneva, Switzerland. 2019.
https://www.who.int/publications-detail/world-malaria-report-2019.
2. Spangenberg T, Burrows JN, Kowalczyk S, McDonald S, Wells TN, Willis P. The open
access malaria box: a drug discovery catalyst for neglected diseases. PLoS One.
2013;8:62901–62908.
13. Bayih AG, Folefoc A, Mohon AN, Eagon S, Anderson M, Pillai DR. In Vitro and in Vivo
Anti-malarial Activity of Novel Harmine-Analog Heat Shock Protein 90 Inhibitors: A
Possible Partner for Artemisinin. Malar J. 2016;15:579.
14. Trott O, Olson AJ. AutoDock Vina: Improving the Speed and Accuracy of Docking
with a New Scoring Function, Efficient Optimization, and Multithreading. J Comput
Chem. 2010;31:455–461.
3. Arama C, Troye-Blomberg M. The path of malaria vaccine development: challenges
and perspectives. J Intern Med. 2014;275:456–466.
4. (a) Phillips MA, Burrows JN, Manyandoo C, van Huijsduijnen RH, van Voorhis WC,
Wells TNC. Malaria Nat Rev Dis Prim. 2017;3:17050
15. Li H, Leung K, Wong M. iDock: A Multithreaded Virtual Screening Tool for Flexible
Ligand Docking. CIBCB. 2015;34:115–126.
(b) Durrows JN, Burlot E, Campo B, et al. Antimalarial Drug Discovery – the Path
towards Eradication. Parasitology. 2014;141:128–139
16. Masters L, Eagon S, Heying M. Evaulation of Consensus Scoring Methods for
AutoDock Vina, Smina and iDock. J Mol Graph Model. 2020;96:107532.
17. (a) Durham SD, Sierra B, Gomez MJ, et al. Synthesis of β-Carbolines via a Silver-
Mediated Oxidation of Tetrahydro-β-carbolines. Tetrahedron Lett.
2017;58:2747–2750
(c) Flannery EL, Chatterjee AK, Winzeler EA. Antimalarial Drug Discovery:
Approaches and Progress toward New Medicines. Nat Rev Microbiol.
2013;11:849–862
(d) Renslo AR. Antimalarial drug discovery: from quinine to the dream of eradica-
tion. ACS Med Chem Lett. 2013;4:1126–1128.
(b) Panarese JD, Waters SP. Tandem Iodine-Mediated Oxidations of Tetrahydro-β-
carbolines: Total Synthesis of Eudistomins Y1–Y7. Org Biomol Chem.
2013;11:3428–3431.
5. (a) Dondorp AM, Nosten F, Yi P, et al. Artemisinin resistance in plasmodium falci-
parum malaria. N Engl J Med. 2009;361:455–467
(b) Haldar K, Bhattacharjee S, Safekui I. Drug resistance in plasmodium. Nat Rev
6