Synthesis, biochemical evaluation, and molecular modeling of organophosphate-coumarin hybrids as potent and selective butyrylcholinesterase inhibitors

Article abstract of DOI:10.1016/j.bmcl.2020.127213

A small library of new organophosphorylated warfarins and 3-benzylcoumarins were synthesized and evaluated for in vitro cholinesterase inhibition by Ellman's method. Most of the compounds were found to be selective for butyrylcholinesterase (BChE) over acetylcholinesterase (AChE), with IC₅₀ values ranging from 0.363 μM to 53.0 μM determined after 15 s of enzyme exposure. Comparison of the most potent compound, 3b with its constitutional isomer 2b revealed the high importance of phosphate positioning. Reversed selectivity and a 100-fold reduction in anti-BChE activity was observed when the organophosphate was attached to the benzyl instead of the coumarin. Docking calculations suggest that 3b binds initially as a transition state mimic with near-optimal phosphate orientation relative to S198 and occupation of the oxyanion hole prior to phosphorylation. These results might inspire the design of a new type of non-neuropathic and irreversible coumarin-based inhibitor against BChE.

Full text of DOI:10.1016/j.bmcl.2020.127213

Bioorganic&MedicinalChemistryLetters30(2020)127213
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis, biochemical evaluation, and molecular modeling of
organophosphate-coumarin hybrids as potent and selective
butyrylcholinesterase inhibitors
Lee J. Macklin, Jason P. Schwans⁎
Department of Chemistry and Biochemistry, California State University, Long Beach, 1250 Bellflower Boulevard, Long Beach, CA 90840-9507, USA
A R T I C L E I N F O
A B S T R A C T
Keywords:
A small library of new organophosphorylated warfarins and 3-benzylcoumarins were synthesized and evaluated
for in vitro cholinesterase inhibition by Ellman’s method. Most of the compounds were found to be selective for
butyrylcholinesterase (BChE) over acetylcholinesterase (AChE), with IC50 values ranging from 0.363 μM to
53.0 μM determined after 15 s of enzyme exposure. Comparison of the most potent compound, 3b with its
constitutional isomer 2b revealed the high importance of phosphate positioning. Reversed selectivity and a 100-
fold reduction in anti-BChE activity was observed when the organophosphate was attached to the benzyl instead
of the coumarin. Docking calculations suggest that 3b binds initially as a transition state mimic with near-
optimal phosphate orientation relative to S198 and occupation of the oxyanion hole prior to phosphorylation.
These results might inspire the design of a new type of non-neuropathic and irreversible coumarin-based in-
hibitor against BChE.
Butyrylcholinesterase
Coumarin
SwissDock
Alzheimer's Disease
Warfarin
Alzheimer's Disease (AD) is a complex multi-factorial brain disorder
which is characterized by formation of deleterious Aβ42 plaques, and
concomitant degeneration of proximal neurons and synapses. The dis-
ease currently affects over 45 million people world-wide and continues
to have a major public health impact as the aging population increases.1
Symptoms include dementia, memory loss, cognitive impairment, and
debilitating psychiatric manifestations such as depression, anxiety, and
psychosis. To date, the most commonly prescribed treatments for mild
to moderate stage AD have targeted a family of enzymes known as
cholinesterases (ChEs), such examples include the reversible inhibitors
donepezil, galantamine, and rivastigmine.2,3 ChEs in the brain function
to terminate excitatory synaptic activity by rapid degradation of the
neurotransmitter acetylcholine. These drugs follow the classical choli-
nergic hypothesis which suggests that increased synaptic concentration
of acetylcholine, as a result of acetylcholinesterase (AChE) inhibition,
provides some relief of dementia and AD-related cognitive decline.4–6
Due to persistent dosing requirements of reversible inhibitors, and the
significant expression of AChE in peripheral tissues, such compounds
tend to have severe systemic side-effects associated with disruption of
the peripheral nervous system (PNS) and gastrointestinal toxicity.
In contrast to the popular use of reversible inhibitors, notable work
has recently highlighted the importance of irreversible inhibition of
brain AChE in monkeys using methanesulfonyl fluoride.7 The work of
Moss et al. provided a robust illustration of how an irreversible ChE
inhibitor allows maximization of the therapeutic potential of the cho-
linergic hypothesis. As revealed in the study, an irreversible inhibitor
will be naturally selective for the central nervous system (CNS) due to
the slow AChE turnover in this tissue (ca. 12 days) compared to in-
testinal AChE turnover (ca. 1 day). With dose-stacking and accumula-
tion of brain AChE inhibition over time, this allows for abolishment of
the more persistent dosing requirements associated with reversible in-
hibitors that often lead to PNS disruption. The most effective use of
reversible inhibitors is limited by lack of prolonged brain AChE in-
hibition and higher doses are often constrained by inherent and un-
bearable gastrointestinal disruptions.7
An important isoform of AChE known as butyrylcholinesterase
(BChE) has been shown to be significantly upregulated in AD brains,
while AChE levels remain unchanged or are decreased.8 Synthetic and
biochemical work by Nakayama et al. with organophosphorous (OP)
inhibitors has highlighted the importance of BChE-selective inhibitors
that do not alter AChE activity.9 Also work by Greig et al. showed that
selective inhibition of BChE in transgenic mouse models overexpressing
amyloid precursor protein and Aβ peptide exhibited a significant re-
duction (47–54%) of synthesis and secretion of neurotoxic Aβ and in-
creased cognitive performance in elderly rats.10 The BChE inhibitor-
dependent reduction in Aβ was also verified in a neuronal human cell
⁎ Corresponding author.
E-mail address: jason.schwans@csulb.edu (J.P. Schwans).
https://doi.org/10.1016/j.bmcl.2020.127213
Received 26 February 2020; Received in revised form 20 April 2020; Accepted 23 April 2020
Availableonline25April2020
0960-894X/©2020ElsevierLtd.Allrightsreserved.

L.J. Macklin and J.P. Schwans
Bioorganic&MedicinalChemistryLetters30(2020)127213
line with no effect on cell viability. These results are indicative of the
importance and high clinical potential of BChE-selective inhibitors
given their ability to mediate critical neuropathological markers of AD.
Coumarin derivatives are
a widely appreciated class of bio-
flavonoids which have proven beneficial in medicine most notably as
potent anti-coagulants (warfarin), anti-viral, anti-inflammatory, anti-
bacterial, and anti-cancer agents, among many other pharmacologically
relevant targets.11–14 A variety of coumarins have been implicated as in
vitro inhibitors of several key proteins related to AD progression, in-
cluding ChEs. In addition, coumarins have been known to demonstrate
highly diverse activities toward various avenues of AD, such as anti-
Aβ42 aggregation, BACE-1 inhibitors, COX/LOX antagonists, dopamine
and serotonin mediators, CB2 receptor antagonists, GABA receptor
agonists, NMDA receptor antagonists, and monoamine oxidase (MAO)
inhibitors.15–24
Scheme 2. Synthesis of organophosphorylated 3-benzylcoumarins (2b, 2c, 3b).
Reagents and conditions: a) NaH, THF, RT, 15 min; b) excess dialkyl-
chlorophosphate, where Z = O, S, and R2 = ethoxy, 0 °C to RT, 7 days. Mono-
phosphorylated products were obtained from position “x” or position “y” from
the corresponding phenolic scaffold.
Several docking studies25–27 have shown the coumarin core (2H-
chromen-2-one) is attracted to the peripheral anionic site (PAS) of both
ChE isoforms, this affinity serves to mimic the natural substrate en-
snaring mechanism which the enzyme uses for canonical catalytic ef-
ficiency. Therefore, the coumarin template is an excellent scaffold for
hybridization with other pharmacophores due to its tendency to in-
teract with the PAS. For this reason, medicinal chemists investigating
new anti-AD agents have often linked a coumarin derivative to either a
more potent ChE inhibitor (ChEi) like tacrine, or a pharmacophore for a
completely different anti-AD target (e.g., β-secretase) to produce novel
multi-target directed ligands which have shown promise for more ef-
ficacious therapeutics.28,6 In addition, the PAS of AChE was proposed to
be a binding site for Aβ where it promotes aberrant misfolding. New
dual-binding site inhibitors of AChE have been shown to reduce in vitro
aggregation rates using ChE-induced Aβ aggregation assays.28
In the work presented herein, we describe the synthesis, ChE eva-
luation, and molecular modeling of OP-warfarins (Scheme 1) and OP-
benzylcoumarins (Scheme 2) and their phenolic scaffolds. Additionally,
we substituted the OP group with an organosulfonyl (OS) moiety to
generate the respective OS-coumarin hybrids (Scheme 3).
Scheme 3. Synthesis of organosulfonylated warfarin and 3-benzylcoumarin.
Reagents and conditions: a) excess ethanesulfonylchloride, acetone, RT, 7 days;
b) same as conditions in Scheme 2 but with ethanesulfonylchloride.
corresponding sodium phenolate, which was subsequently reacted with
excess dialkylchlorophosphate to produce compounds 2b, 2c, and 3b.
These same procedures were carried out in an identical manner but
instead with ethanesulfonylchloride to afford compounds 1d and 2d
(Scheme 3). The compounds were purified by flash-chromatography
using hexanes/ethyl acetate as the eluent and generated the final pro-
ducts in 35–60% yields. Sample homogeneity for the final compounds
was confirmed via GC–MS and 1H, 13C, 31P NMR spectra (see
Appendix).
All compounds were characterized by 1H NMR, 13C NMR, and gas
chromatography mass spectrometry (GC–MS). To identify the in-
hibitors, the pure compounds were screened against BChE and AChE by
the method of Ellman29 with only minor modifications. Relative po-
tencies were characterized by determination of IC50 values after 15 s of
enzyme exposure. Assay conditions with no pre-incubation time were
used so that IC50 values would better approximate initial binding and to
avoid shifting IC50 toward the enzyme concentration. Molecular mod-
eling analyses were performed using SwissDock software30 to provide
structure–activity relationships and top-ranked docking poses of the
initial reversible binding events.
The library of synthesized inhibitors (including the phenolic pre-
cursors and naphthol control analogs) are shown in Fig. 1. The effects of
the compounds on ChE activity were evaluated using Ellman’s assay as
previously described but modified for smaller reactions on a 120 μL
scale. Stock inhibitor solutions were made in a methanol vehicle and
further diluted for a general screen against equine BChE and electric eel
AChE. The results of this screen are shown in Fig. 2, and relative initial
activities are compared to a control reaction with vehicle only. The
activities were determined in triplicate after 15 s enzyme exposure and
reported as a mean value. All the compounds were selective for BChE
over AChE, except for compound 2b which was found to be selective for
As shown in Scheme 1, commercial sodium warfarin was reacted
with excess dialkylchlorophosphate at room temperature, the sluggish
progress likely due to the sterically hindered phenoxide on position 4.
Near complete conversion was observed after stirring for 7 days and
produced compounds 1a-1c. The 3-benzylcoumarin phenolic precursors
(Scheme 2) were synthesized by low-yielding Umpolung-Domino re-
actions similar to as reported by Toräng et al.,31 but with slight mod-
ifications (described in Appendix). The phenolic hydroxyl group either
on position 5 (Rx) or para to the 3-benzyl group (Ry) was deprotonated
at room temperature using sodium hydride in THF to give the
AChE by 3-fold (IC50
=
10.3
1 μM) compared to BChE
(IC50 = 30.2 4 μM). To help explain this apparent preference,
docking calculations were performed using the human isoform crystal
structures32 and correlated with van der Waals (VdW) volume calcu-
lations. Molecular dynamics calculations by Saxena et al.33 showed a
considerable volume difference (~200 Å3) for the catalytic gorge of
BChE (502 Å3) versus AChE (302 Å3) and mutagenesis experiments il-
lustrated the general rigidity of these catalytic centers, in addition to a
lack of conformational perturbation upon ligand binding. One possible
reason for BChE-selectivity is the smaller active site cavity of AChE
which tends to exclude certain larger ligands greater than about 302 Å3.
This size-dependent selectivity has been observed with several ChE
inhibitors in various reports9,33,34 and is the likely reason for the overall
BChE-selective nature of the compounds presented in this letter (Fig. 2).
The inhibitory activities of compounds are summarized in Table 1.
Scheme 1. Synthesis of organophosphorylated warfarin (1a-1c). Reagents and
conditions: a) excess dialkylchlorophosphate, where Z = O, S and R1 = ethoxy
or butoxy; acetone, RT, 7 days.
2

L.J. Macklin and J.P. Schwans
Bioorganic&MedicinalChemistryLetters30(2020)127213
Fig. 1. Library of organophosphorylated and organosulfonylated racemic warfarin, 3-benzylcoumarins, and 1-naphthol derivatives.
Compound 3b exhibited a 100-fold increase in potency for BChE
compared to its constitutional isomer 2b, indicating that serine-inter-
acting moieties on position 5 of the coumarin may be important for
developing more effective irreversible BChE-selective inhibitors. When
the OP group is attached directly to the coumarin core as opposed to the
benzyl moiety, the more efficient leaving group may lead to a more
rapid phosphorylation rate (kmax). However, the maximum rate of
phosphorylation for 3b could not be measured due to inherent temporal
limitations of the assay. The linear phase of time-dependent inhibition
passes almost immediately after enzyme exposure at nM concentrations
near the enzyme. This was also the case for OP-warfarin 1c which in-
hibited too rapidly to measure kmax
.
Using Chimera software, the calculated VdW volume of 2b (321 Å3)
was larger than the theoretical active-site gorge volume of AChE by
~20 Å3. By the exclusion hypothesis, one might expect 2b to exhibit
poor inhibitory activity toward AChE. Kinetic experiments for 2b
showed substantial AChE inhibition (IC50 = 10.3
1 μM) that lacks
any time-dependence, compared to BChE which inhibits progressively
Fig. 2. Bar graphs illustrating general selectivity of OP-coumarin hybrids for BChE over AChE. A) BChE inhibition by OP-coumarins at a general screening con-
centration of 34 μM, B) AChE screen under identical conditions. Compounds were screened by the method of Ellman29 using 2% MeOH (v/v) as a vehicle. For 1c and
3b, BChE activity above background was not observed, and substantial AChE activity was observed at the same concentration.
3

L.J. Macklin and J.P. Schwans
Bioorganic&MedicinalChemistryLetters30(2020)127213
Table 1
In vitro inhibitory activities of warfarin-OPs, 3-
benzylcoumarin-OPs (and phenolic pre-
cursors), OP-naphthyl and OS-naphthyl ana-
logs, against horse BChE after 15 s exposure.
Entry
IC50 (µM)
1a
1ba
1c
1d
2a
2b
2c
2d
3a
3b
4
11.2
> 100
3
1.52
0.2
ND
51.2
30.2
53.0
4
4
4
ND
> 700
0.363
0.1
3
32.0
54.0
71.5
5a
5b
5
5
IC50 is the concentration of inhibitor which
results in 50% of maximal activity. Mean va-
lues are reported with standard deviation
values.
ND = not determined; a = determined in 2%
DMSO.
Fig. 4. Top-ranked docking pose of 2b into hAChE (PDB 4PQE). A molecular
volume calculation indicates that 2b is too large (321 Å3, cyan) to fit into the
rigid active site of hAChE (302 Å3, white box) but inhibits from a location distal
from the active center. Even though 2b is predicted to bind outside of the AChE
active site, it was experimentally found to be 3-fold more potent than for BChE.
The inability to enter the active site and modify S203 is supported by a lack of
observed time-dependent inhibition. The H405 residue is shown in magenta,
with VdW contacts shown as yellow lines.
Fig. 3. Top-ranked docking pose of 2b into hBChE (PDB 1P0I). 2b (gray) can fit
into the larger active site cavity and interact with several key catalytic residues
and subsites (labeled with color) prior to phosphorylation of S198, giving rise to
time-dependent inhibition. White lines here indicate the favorable VdW con-
tacts.
over time. These results suggest that 2b binds initially to the active site
of BChE prior to phosphorylation of S198, whereas 2b inhibits AChE by
some other mechanism that does not involve active-site entry or
covalent modification. To test this hypothesis, a comparison of the
docking poses of 2b inside AChE and BChE supported this implication.
2b can easily be accommodated into the larger active site of BChE
where it binds initially to the acyl pocket and oxyanion hole, with the
OP group interacting with the anionic site and H438 residue, prior to
phosphorylation (Fig. 3). On the other hand, none of the top-ranked
docking poses of 2b inside AChE modeled entry into the active site
(Fig. 4). Compound 2b was found to dock distally from the active
center, but proximal to the gorge entry channel where it is modeled to
interact primarily with H405 through hydrogen bonding and VdW
contacts, while the benzyl group was found to interact with the back-
bone carbonyl of N233 through VdW interactions. It is through these
interactions that 2b may exert an interference upon the active site
Fig. 5. Compound 2b exhibits time-dependent inhibition of BChE but not
AChE. A lack of time-dependent inhibition was observed for 2b in AChE, while
at the same concentration and time points showed progressive inhibition of
BChE indicative of irreversible covalent modification. The slight inhibition in-
crease of 2b in AChE over the first two minutes is likely due to the rate of the
initial binding event and time needed for complete saturation of the enzyme
population.
4

L.J. Macklin and J.P. Schwans
Bioorganic&MedicinalChemistryLetters30(2020)127213
architecture. This model is supported by a lack of observed progressive
inhibition for AChE (Fig. 5).
(kobs = 142
38 mM−1 s−1) by approximately 20-fold (Fig. S2).
Although the severity of the thion effect appears less pronounced, a
similar trend was observed when comparing the 3-benzylcoumarin
oxon 2b with thion 2c, which caused a ~2-fold decrease in potency
with the sulfur substitution (Table 1), indicative of the overall reduced
chemical reactivity of O-arylphosphorothioates with BChE.
The 3-benzylcoumarin phenols 3a, 2a, and diol 4 were also eval-
uated to test the non-covalent influence of the coumarin template and
whether the hydroxyl group’s extent and position are important for
inhibition. Compound 3a (the precursor to lead compound 3b) did not
inhibit BChE (IC50
>
700 μM), suggesting that electron donating
Previous reports36–38 have identified a crucial off-target enzyme
(neuropathy target esterase, NTE), which is of great importance when
discussing ChE-based OP inhibitors as a possible AD treatment. Many
OP compounds undergo a time-dependent aging process with esterases;
after initial covalent modification, the enzyme-inhibitor (EI) adduct
may further react to form an anionic species which cannot be rescued
by any nucleophilic oxime antidote. To date, the biological functions of
NTE are unclear, however it is thought to be implicated in early neu-
ronal growth and development. When NTE is inhibited and aged via
pesticidal OP compounds, detrimental neuropathic symptoms (e.g.,
axonal degeneration) may develop in a matter of days to weeks (it is
still unclear why the aged EI adduct is generally required for OP-in-
duced delayed neuropathy). For this reason, development of irrever-
sible BChE-selective inhibitors which are non-neuropathic is of great
interest. Organosulfonylated inhibitors are non-ageable in esterases and
are not known to undergo subsequent modification of the sulfonylated
EI adduct.
groups on position 5 may lead to poorer inhibitors, and the OP group
likely makes significant energetic contributions to initial binding.
Docking of deprotonated phenol 3a in BChE shows it clinging to the
exterior surface of the protein and does not enter or get near the cat-
alytic gorge (Fig. S4). On the other hand, only a 1.5-fold decrease in
potency was observed for phenolic precursor 2a (IC50 = 51.2 μM)
compared to the OP 2b, indicating that the coumarin template plays a
substantial role for the initial reversible binding event. When com-
paring these results to the diol 4 (IC50 = 32.0 μM), a 1.5-fold increase
in potency was observed. Using ChemAxon software, the predicted pKa
of the hydroxyl group associated with 3a and 2a is 7.63 and 9.5, re-
spectively. At pH 7.5, the phenol attached directly to the coumarin
nucleus is deprotonated ~43% of the time, whereas when linked to the
benzyl group is protonated 99% of the time. For compound 4, this gives
rise to two possible inhibitory modes (Z = −1 and Z = 0 states) which
both simultaneously contribute to inhibition at different sites (Fig. S5).
Based on docking, we speculate the poor BChE inhibition by 3a is due
to unfavorable electrostatic repulsion associated with the deprotonated
state and E197 in the active site (Fig. S4).
We synthesized the OS-warfarin, OS-3-benzylcoumarin, and OS-
naphthol analogs (1d, 2d, 5b). Unfortunately, due to high hydro-
phobicity, 1d and 2d could not be dissolved in vehicle and could not be
diluted into aqueous buffer solutions for evaluation. Compound 5b,
The OP-warfarin oxons 1a and 1c differ only in their alkyl chain
length (diethyl and dibutyl) and both exhibit rapid time-dependent
inhibition of BChE while lacking any AChE inhibition (data not shown).
Warfarin was selected as a convenient 3-benzylcoumarin control scaf-
fold which is commercially available as a sodium salt and is easily
conjugated to various electrophiles to afford novel warfarin derivatives.
Though warfarin is most well-known as a potent anti-clotting agent
through inhibition of vitamin K epoxide reductase (VKOR), we re-
purposed this compound toward BChE inhibition by exploiting the
specific affinity of the coumarin nucleus for the PAS, ultimately al-
lowing for more rapid covalent modification of S198. At 34 μM, the
warfarin sodium salt exhibited only very minor inhibition of BChE
(~10%) which is likely due to transient PAS binding (Fig. 2A). To test
whether the addition of simple steric bulk to 1a would result in a more
effective inhibitor, compound 1c was synthesized and demonstrated
~7-fold increased potency, probably due to more favorable initial
binding and increased VdW interactions in the active cavity. This in-
ference is supported by docking calculations which show a higher
overall active site occupancy by 1c, making contacts with each key
subsite and the OP group positioned closely to S198. In contrast, 1a was
found to lack VdW contacts with the acyl pocket and oxyanion hole,
while positioned significantly further away from the catalytic triad,
resulting in a lower total occupancy (Fig. S6).
with the simpler naphthol group was soluble (IC50 = 71.5
5 μM)
and was synthesized as a control to optimize the representative orga-
nosulfonylation procedure. Current efforts are underway to synthesize a
more structurally diverse library of OS-warfarins and OS-3-benzylcou-
marin derivatives. We are also exploring the potential modification of
assay conditions which would be more amenable to evaluation of these
hydrophobic compounds. Some previous reports described the suc-
cessful encapsulation of active AChE into liposomes in which a defined
membrane environment mimics that of a living cell and allows for ki-
netic characterization of hydrophobic inhibitors via Ellman’s
method.39,40
Out of the compounds synthesized and evaluated in this report, lead
compound 3b was identified (IC50 = 0.363
0.1 μM). When the OP
group is placed directly on the coumarin core (position 5) as opposed to
the benzyl ring of 2b, a 100-fold increase in potency is observed. The
enhanced inhibitory activity of 3b can be attributed to a more optimal
initial binding event as well as a faster phosphorylation rate due to the
coumaric leaving group. The top-ranked docking pose of 3b nested in
the BChE active site is shown in Fig. 6a.
As can be seen in the docking pose, the planar coumarin core spans
the catalytic gorge while interacting with Y128 and W82 of the anionic
site (AS), allowing the benzyl group to π- π stack with Y332 face-on. In
this model, 3b binds initially with the phosphorous atom forming a
contact between the hydroxyl group of S198, positioned less than 4 Å
away. Importantly, the non-bridging oxygen of the phosphate forms a
hydrogen bond with the backbone amide of G116 in the oxyanion hole,
an interaction which brings 3b kinetically closer to the transition state
upon initial binding. Thus, 3b likely binds as a transition state mimic
prior to rapid phosphorylation of S198 (Fig. 6b).
When the non-bridging OP oxygen of warfarin 1a was replaced with
a sulfur, this reduced inhibitory strength by at least an order of mag-
nitude (1b, IC50 > 100 μM). The loss in potency for the thiophosphate
1b is likely due to a combination of reduced initial binding strength and
a diminished phosphorylation rate. Though the bulkier sulfur sub-
stitution may cause increased steric clashing and interfere with the
initial reversible binding event, work by Zhang et al.35 demonstrated
the substantially reduced chemical reactivity of O-arylpho-
sphorothioates with protein-tyrosine phosphatases (PTPases). The re-
ported enzymatic thio effect generally results in slower associative re-
actions which rely on the electron withdrawing power of the
phosphorous for transition state stabilization. With the non-bridging
sulfur substitution, the Lewis acid character of the OP group is de-
creased, which may disrupt the transition state complementarity and
hinder the association of S198 with the OP moiety. These factors likely
contribute to the observed slower covalent modification rate of BChE by
In summary, we have identified a new class of BChE-selective in-
hibitors that combines the structural features of two different areas of
study. Although many new reversible cholinesterase inhibitors are
continually being described in the literature, we are unaware of any
other reports describing irreversible BChE-selective coumaric in-
hibitors. Herein we have identified a new versatile coumarin scaffold as
an efficient active-site director of covalent modifiers toward the CAS of
BChE. The general affinity of the coumarin template for the PAS allows
for more precise active-site targeting of serine-interacting moieties,
such as OP or OS groups. With OP-warfarins 1a-1c, we found the simple
1b (kobs
=
8.4
0.4 mM−1 s−1
)
compared to 1a
5

L.J. Macklin and J.P. Schwans
Bioorganic&MedicinalChemistryLetters30(2020)127213
inhibition, indicative of the coumarin template attraction to the cata-
lytic gorge. We have also identified OP-coumarin 2b as an inhibitor of
both ChEs, but paradoxically has a 3-fold preference for AChE over
BChE despite its inability to enter the active site. Comparison of phenols
2a and 3a with their OP derivatives (2b and 3b) highlights the com-
plexity of designing selective ChE inhibitors.
Though the introduction of an OP group led to an overall more
active inhibitor, OP attachment to the benzyl moiety instead of the
coumarin resulted in reversed selectivity for AChE. The constitutional
isomer 3b was highly selective for BChE and exhibited 100-fold en-
hanced inhibitory activity, revealing the importance of relative phos-
phate positioning around the coumarin template, influencing both se-
lectivity and potency. The versatile lead scaffold 3b represents a proof
of principle for a new class of irreversible coumarin-based inhibitors
that are BChE-selective and are not limited to use of organophosphorus
covalent modifiers. Experimental work was validated by SwissDock
calculations which provided top-ranked poses of the initial binding
events. Due to possibility of OP-induced delayed neuropathy (OPIDNP)
associated with off-target NTE inhibition, the development of non-
neuropathic and irreversible BChE-selective OS analogs are of great
appeal. Continued work will focus on synthesis and evaluation of a
second-generation library of irreversible non-neuropathic OS deriva-
tives of compound 3b, by substituting the OP component entirely. In
concert to this, we plan to explore alkaloidal and hydroxylated benzyl
substitutions which may facilitate more favorable contacts needed for
stronger PAS binding.
Fig. 6a. Top-ranked docking pose of 3b in hBChE active site (PDB 1P0I).
Numerous π- π stacking interactions are found between the 3-benzylcoumarin
template with the PAS and AS subsites. The OP-moiety is found facing towards
S198 and exhibits numerous VdW contacts with the oxyanion hole (blue), in
addition the P]O moiety exhibited a hydrogen bond with the backbone amide
of G116. The gorge access point is denoted by the white arrow. To emphasize
occupancy, a solid surface was applied to the ligand with 80% transparency.
Color coding follows that of Fig. 3.
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.
Acknowledgements
We would like to thank California State University, Long Beach and
the College of Natural Sciences and Mathematics for providing partial
funding through the Graduate Research Fellowship. Many thanks to Dr.
Gregory Whitaker for his gracious donation to this work through a
scholarship to Lee J. Macklin. We would also like to thank Dr. Kensaku
Nakayama and Dr. Eric Sorin for their critical reading of the manuscript
and for providing valuable feedback.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.bmcl.2020.127213.
References
1. Crous-Bou M, Minguillón C, Gramunt N, Molinuevo JL. Alzheimers Res Ther.
2017;9:71.
2. Hayashi K, Hamano T, Asano R, et al. J Neurol Sci. 2017;381:329–330.
3. Bhattacharjee S, Patanwala AE, Lo-Ciganic WH, et al. Alzheimer's Dement (NY).
2019;5:294–302.
4. Derakhshankhah H, Sajadimajd S, Jafari S, et al. Nanomedicine. 2020:102149.
5. Kozin SA, Makarov AA. The convergence of Alzheimer’s Disease pathogenesis con-
cepts. Mol Biol (Mosk). 2019;53:896–903.
Fig. 6b. Diethylphosphate moiety of 3b found to interact with key catalytic
residues. From the docking pose shown in Fig. 6a, emphasis was given to the
binding of the OP-moiety. Specifically, a contact is found between the phos-
phorous of 3b and oxygen atom of the Ser198 hydroxyl group (d = 3.67 Å,
white line/orange arrow), in addition to an H-bond of the P]O group with the
backbone amide (–NH) of G116 of the oxyanion hole (d = 2.52 Å, yellow line).
6. Wang Y, Wang H, Chen HZ. Curr Neuropharmacol. 2016;14:364–375.
7. Moss DE, Perez RG, Kobayashi H. J Alzheimer's Dis. 2017;55:1285–1294.
8. Korábečný J, Nepovimová E, Cikánková T, et al. Neuroscience. 2018;370:191–206.
9. Nakayama K, Schwans JP, Sorin EJ, et al. Bioorg Med Chem. 2017;25:3171–3181.
10. Greig NH, Utsuki T, Ingram DK, et al. Proc Natl Acad Sci USA.
2005;102:17213–17218.
11. Shen G, Cui W, Zhang H, et al. Nat Struct Mol Biol. 2017;24:69–76.
12. Sandhu S, Bansal Y, Silakari O, Bansal G. Bioorg Med Chem. 2014;22:3806–3814.
13. Pereira TM, Franco DP, Vitorio F, Kummerle AE. Curr Top Med Chem.
2018;18:124–148.
addition of steric bulk in the form of elongated alkyl chains led to a 7-
fold increase in potency, while sulfur substitution on the non-bridging
position led to a > 10-fold reduction of inhibition. The non-phos-
phorylated coumarins 2a and 4 exhibited moderate non-covalent
14. Srikrishna D, Godugu C, Dubey PK. Mini Rev Med Chem. 2018;18:113–141.
15. Soto-Ortega DD, Murphy BP, Gonzalez-Velasquez FJ, et al. Bioorg Med Chem.
6

L.J. Macklin and J.P. Schwans
Bioorganic&MedicinalChemistryLetters30(2020)127213
2011;19:2596–2602.
28. Mishra P, Kumar A, Panda G. Bioorg Med Chem. 2019;27:895–930.
16. Marumoto S, Miyazawa M. Bioorg Med Chem. 2012;20:784–788.
17. Revankar HM, Bukari SN, Kumar GB, Qin HL. Bioorg Chem. 2017;71:146–159.
18. Abdelhafez OM, Amin KM, Ali HI, Maher TJ, Batran RZ. Neurochem Int.
2011;59:906–912.
29. Ellman GL, Courtney KD, Andres Jr V, Feather-Stone RM. Biochem Pharmacol.
1961;7:88–95.
30. Grosdidier A, Zoete V, Michielin O. Nucleic Acids Res. 2011;39:W270–277.
31. Toräng J, Vanderheiden S, Nieger M, Bräse S. Eur J Org Chem. 2007;2007:943–952.
32. Nicolet Y, Lockridge O, Masson P, Fontecilla-Camps JC, Nachon F. J Biol Chem.
2003;278:41141–41147.
19. Ostrowska K, Grzeszczuk D, Gluch-Lutwin M, et al. Bioorg Med Chem.
2018;26:527–535.
20. Rempel V, Volz N, Hinz S, et al. J Med Chem. 2012;55:7967–7977.
21. Riedel G, Davies SN. Cannabinoids. 2005;168:445–477.
22. Singhuber J, Baburin I, Ecker GF, Kopp B, Hering S. Eur J Pharmacol.
2011;668:57–64.
33. Saxena A, Redman AM, Jiang X, Lockridge O, Doctor BP. Biochemistry.
1997;36:14642–14651.
34. Gonzalez J, Ramirez J, Schwans JP. Amino Acids. 2016;48:2755–2763.
35. Zhang YL, Hollfelder F, Gordon SJ, et al. Biochemistry. 1999;38:12111–12123.
36. Glynn Paul. Biochem J. 1999;344(3):625–631.
23. Irvine MW, Costa BM, Volianskis A, et al. Neurochem Int. 2012;61:593–600.
24. Patil PO, Bari SB, Firke SD, Deshmukh PK, Donda ST, Patil DA. Bioorg Med Chem.
2013;21:2434–2450.
37. Richardson RJ, Hein ND, Wijeyesakere SJ, Fink JK, Makhaeva GF. Chem Biol Interact.
2013;203:238–244.
25. Schalk I, Ehret-Sabatier L, Le Feuvre Y, Bon S, Massoulie J, Goeldner M. Mol
Pharmacol. 1995;48:1063–1067.
38. Jokanović M. Toxicology. 2018;410:125–131.
39. Ochoa EL. J Eur Biochem. 1980;107:415–422.
26. Radić Z, Taylor P. Chem Biol Interact. 1999;119:111–117.
27. Fallarero A, Oinonen P, Gupta S, et al. Pharmacol Res. 2008;58:215–221.
40. Chaize B, Fournier D. Biosens Bioelectron. 2004;20:628–632.
7