Synthesis, biochemical evaluation, and molecular modeling studies of aryl and arylalkyl di-n-butyl phosphates, effective butyrylcholinesterase inhibitors

Article abstract of DOI:10.1016/j.bmc.2017.04.002

A series of dialkyl aryl phosphates and dialkyl arylalkyl phosphates were synthesized. Their inhibitory activities were evaluated against acetylcholinesterase (AChE) and butyrylcholinesterase (BChE). The di-n-butyl phosphate series consistently displayed selective inhibition of BChE over AChE. The most potent inhibitors of butyrylcholinesterase were di-n-butyl-3,5-dimethylphenyl phosphate (4b) [K_I?=?1.0?±?0.4?μM] and di-n-butyl 2-naphthyl phosphate (5b) [K_I?=?1.9?±?0.4?μM]. Molecular modeling was used to uncover three subsites within the active site gorge that accommodate the three substituents attached to the phosphate group. Phosphates 4b and 5b were found to bind to these three subsites in analogous fashion with the aromatic groups in both analogs being accommodated by the “lower region,” while the lone pairs on the P[dbnd]O oxygen atoms were oriented towards the oxyanion hole. In contrast, di-n-butyl-3,4-dimethylphenyl phosphate (4a) [K_I?=?9?±?1?μM], an isomer of 4b, was found to orient its aromatic group in the “upper left region” subsite as placement of this group in the “lower region” resulted in significant steric hindrance by a ridge-like region in this subsite. Future studies will be designed to exploit these features in an effort to develop inhibitors of higher inhibitory strength against butyrylcholinesterase.

Full text of DOI:10.1016/j.bmc.2017.04.002

Accepted Manuscript
Synthesis, Biochemical Evaluation, and Molecular Modeling Studies of Aryl
and Arylalkyl Di-n-Butyl Phosphates, Effective Butyrylcholinesterase Inhibi-
tors
Kensaku Nakayama, Jason P. Schwans, Eric J. Sorin, Trina Tran, Jeannette
Gonzalez, Elvis Arteaga, Sean McCoy, Walter Alvarado
PII:
S0968-0896(17)30014-7
DOI:
http://dx.doi.org/10.1016/j.bmc.2017.04.002
BMC 13671
Reference:
Bioorganic & Medicinal Chemistry
To appear in:
Received Date:
Revised Date:
Accepted Date:
4 January 2017
29 March 2017
3 April 2017
Please cite this article as: Nakayama, K., Schwans, J.P., Sorin, E.J., Tran, T., Gonzalez, J., Arteaga, E., McCoy, S.,
Alvarado, W., Synthesis, Biochemical Evaluation, and Molecular Modeling Studies of Aryl and Arylalkyl Di-n-
Butyl Phosphates, Effective Butyrylcholinesterase Inhibitors, Bioorganic & Medicinal Chemistry (2017), doi: http://
dx.doi.org/10.1016/j.bmc.2017.04.002
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Synthesis, Biochemical Evaluation, and Molecular Modeling Studies
of Aryl and Arylalkyl Di-n-Butyl Phosphates, Effective
Butyrylcholinesterase Inhibitors
Kensaku Nakayamaa∗, Jason P. Schwansa∗, Eric J. Sorina∗, Trina Trana, Jeannette
Gonzaleza, Elvis Arteagaa, Sean McCoya and Walter Alvaradob
aDepartment of Chemistry and Biochemistry, California State University, Long Beach, 1250
Bellflower Blvd., Long Beach, CA 90840, USA
bDepartment of Physics, California State University, Long Beach, 1250 Bellflower Blvd., Long
Beach, CA 90840, USA
Corresponding authors.
Tel.: +1-562-985-4850; fax: +1-562-985-8557; e-mail: Kensaku.nakayama@csulb.edu
Tel.: +1-562-985-7778; fax: +1-562-985-8557; e-mail: Jason.schwans@csulb.edu
Tel.: +1-562-985-4850; fax: +1-562-985-7537; e-mail: Eric.sorin@csulb.edu
Keywords:
Cholinesterase
Alzheimer’s disease
Organophosphates
Enzyme Inhibition
Computational docking
TOC Graphic

Abstract
A series of dialkyl aryl phosphates and dialkyl arylalkyl phosphates were synthesized. Their
inhibitory activities were evaluated against acetylcholinesterase (AChE) and
butyrylcholinesterase (BChE). The di-n-butyl phosphate series consistently displayed selective
inhibition of BChE over AChE. The most potent inhibitors of butyrylcholinesterase were di-n-
butyl-3,5-dimethylphenyl phosphate (4b) [KI = 1.0 + 0.4 µM] and di-n-butyl 2-naphthyl
phosphate (5b) [KI = 1.9 + 0.4 µM]. Molecular modeling was used to uncover three subsites
within the active site gorge that accommodate the three substituents attached to the phosphate
group. Phosphates 4b and 5b were found to bind to these three subsites in analogous fashion
with the aromatic groups in both analogs being accommodated by the “lower region,” while the
lone pairs on the P=O oxygen atoms were oriented towards the oxyanion hole. In contrast, di-n-
butyl-3,4-dimethylphenyl phosphate (4a) [KI = 9 + 1 µM], an isomer of 4b, was found to orient
its aromatic group in the “upper left region” subsite as placement of this group in the “lower
region” resulted in significant steric hindrance by a ridge-like region in this subsite. Future
studies will be designed to exploit these features in an effort to develop inhibitors of higher
inhibitory strength against butyrylcholinesterase.

1. Introduction
Cholinesterase inhibitors are currently the major drug type in use to manage the progression
of Alzheimer’s disease (AD), the most common type of adult-onset dementia.1 AD-associated
cognitive impairment is correlated with acetylcholine (ACh) level reduction in the brain. This
cognitive loss takes place in conjunction with changes in cholinesterase activity where
acetylcholinesterase (AChE) activity decreases and butyrylcholinesterase (BChE) activity
increases.2 BChE activity in the AD brain is known to be elevated 40 - 65% above normal while
AChE activity decreases to about 65% of the normal level.2
Since both AChE and BChE hydrolyze ACh, the use of cholinesterase inhibitors as treatment
for AD is based on the hypothesis that the inhibition of these enzymes will increase the
concentration of acetylcholine in the brain.3 Thus, current treatment for AD involves the
administration of reversible dual cholinesterase inhibitors, such as galantamine and tacrine, to
suppress the activity of both AChE and BChE.4 Recently, increased efforts are being made to
develop cholinesterase inhibitors that target BChE in order to treat AD at more advanced
stages.5, 6
For example, cymserine analogs that selectively inhibited BChE were shown to raise ACh
levels, improve cognition and lower β-amyloid peptide in rodent.7 In addition, N1-
phenethylnorcymserine, a BChE-selective inhibitor, suppressed cognitive dysfunction in mice
challenged by amyloid-β peptide.8 These results suggest that abnormal BChE activity increases
the severity of cognitive dysfunction associated with AD. Therefore, it is possible that suitable
BChE-inhibitors may have a significant role to play in the treatment of cognitive loss associated
with AD.

The inhibition of the cholinesterases by organophosphorus insecticides and nerve toxins is
widely appreciated.9-11 Investigations into the selective inhibition of BChE over AChE by
organophosphorus compounds is also receiving increased attention recently. For example, the
exploration of organophosphates as AD therapeutics that are BChE-selective have been recently
reported by Richardson and co-workers.12 Meanwhile, Sultatos and co-workers have reported the
higher reactivity of the organophosphate chlorpyrifos towards BChE over AChE.13 More recently,
Kaboudin and Emadi found that phosphorothioates were moderately more selective inhibitors of
BChE over AChE.14 A recent study by Vinsova and co-workers has established that salicyanilide
diethyl phosphates are excellent inhibitors of both AChE and BChE, with a preference for the
latter enzyme, surpassing in some cases the inhibitory activity of the currently approved AD
medications rivastigmine and galantamine.15
Some time ago, we systematically studied the inhibitory properties of a small library of
dialkyl phenyl phosphates (DAPPs; Figure 1) against AChE and BChE,16 a family of compounds
that possess some structural similarities to metrifonate, a dual cholinesterase inhibitor that raised
acetylcholine levels and cognitive ability in AD patients during clinical trials.17 We expected that
the DAPPs should act as reversible inhibitors since they lack a good leaving group. We found
that phosphate 1a displayed the weakest inhibition against BChE among the series, while it was
shown to be the most efficient inhibitor of AChE. Meanwhile, compound 1d was the best
inhibitor of BChE, while displaying no inhibition against AChE. Encouraged by these findings,
we decided to explore the effects on inhibition against the cholinesterases, with particular focus
on BChE, when systematic changes were introduced into the aromatic portion of the overall
scaffold. The results of these studies are described herein.

Figure 1. Library of dialkyl phenyl phosphates employed as inhibitors.
2. Results and Discussion
2.1 Chemistry
We studied a library of phosphates comprised of aryl di-n-butyl phosphates (3a - 3c, 4a - c,
5a, 5b), benzyl di-n-butyl phosphate (6a), di-n-butyl methylphenylmethyl phosphates (6a - 6d),
and di-n-butyl phenylethyl phosphate (7) (Figure 2). As mentioned earlier, the library of
compounds examined in our previous work16 did not probe the effects of changes in the phenyl
portion of the inhibitor scaffold on enzyme inhibition. In this study, one main goal was to
explore how the methyl group location on the aromatic ring affects inhibition. We also changed
the size of the phenyl ring by substituting it with a naphthyl group. In addition, we incorporated a
methylene group as well as an ethyl chain between the phenyl ring and the phosphate oxygen
atom to study the effects of these changes on enzyme inhibition.
In addition to the library of dibutyl phosphates listed in Figure 2, we also prepared two aryl
diethyl phosphates 8 and 9 (Figure 3) which allowed us to study the effects of shortened alkyl
chains on inhibition. These phosphates were prepared in acceptable yields and purity (>95% by
GC/MS) by allowing either diethyl chlorophosphate 10 or di-n-butyl chlorophosphate 11 to react
with an excess of the appropriate alcohol, R’OH, in the presence of pyridine (Scheme 1).
Typical reaction times were approximately 12 hours, except for the case when 1-naphthol was

employed as the nucleophile, which required three days. All compounds were smoothly purified
via column chromatography to afford colorless or near colorless oils that gave satisfactory 1H
NMR, 13C NMR and mass spectral data.
Scheme 1.a Synthesis of phosphates 3 – 9.
a(i) Reagents: R'OH, pyridine, CH2Cl2, 0 °C – rt
Figure 2. Library of aryl/alkyl di-n-butyl phosphates studied in this work.
Figure 3: Library of aryl diethyl phosphates studied in this work
.

2.2 Enzyme Inhibition Studies
2.2.1. Evaluating the effect of adding substituents on the phenyl group of alkyl phenyl
phosphates
As described above, Law et al. previously measured inhibition constants for a series of alkyl
phenyl phosphates for BChE and AChE (Figure 1).16 Increasing the length of the alkyl groups
from one to five carbons led to more potent inhibitors. Although comparison of the reported KI
values suggests the dipentyl phenyl phosphate was the most potent inhibitor relative to the
shorter chain analogs (KI = 6 µM (1e) versus KI = 45-80 µM (1c, 1d), we recently observed that
limited solubility affects the determination of the inhibition behavior of the dipentyl phenyl
analog (unpublished results). The di-n-butyl phenyl analog, however, was soluble in the range of
concentrations used to measure inhibition constants. We therefore investigated the effect of
adding phenyl substituents in the di-n-butyl phenyl scaffold.
We first investigated the inhibitory behavior of introducing a methyl group on the phenyl
ring at the 2-, 3-, or 4-position. The solubility of each analog in 2% methanol was evaluated
using absorbance spectra. The absorbance increased linearly up to 0.5 mM suggesting the
analogs are soluble at or below this concentration. To compare to prior results, we first measured
the KI value for dibutyl phenyl phosphate (1d) and obtained a KI value of 100 µM, which is in
agreement with previous studies. Butyrylcholinesterase from equine serum (E.C. Number
3.1.1.8) was used in all BChE inhibition studies. We then used the same approach to determine
the KI values for the phosphate analogs bearing methyl substituents on the phenyl ring (Table 1
and Figure 4). Addition of dibutyl 2-methylphenyl phosphate (3a) led to a decrease in enzyme
activity, but residual activity remained at the highest concentrations that could be tested. This
could be due to the 2-methyl phenyl analog being a partial competitive inhibitor, or due to the

inability to achieve concentrations high enough to abolish activity. Nevertheless, we can obtain a
lower limit of 50 µM for the dibutyl 2-methylphenyl phosphate KI value. In contrast, addition of
dibutyl 3- and 4-methylphenyl phosphate reduced activity to levels indistinguishable from
background suggesting that these analogs act as competitive inhibitors. The KI values for dibutyl
3- and 4-methylphenyl phosphates (3b and 3c) were 15 and 8 µM, respectively, indicating they
are more potent inhibitors compared to the parental dibutyl phenyl phosphate (1d). To test if the
analogs are selective inhibitors of BChE compared to other cholinesterases, we determined if the
analogs inhibit AChE and/or chymotrypsin (Table 2). Previous studies have shown inhibition of
each enzyme by organophosphates typically bearing smaller substituents than the compounds
investigated herein. No significant inhibition of AChE was observed using inhibitor
concentrations of 500 µM indicating the analogs are selective for BChE.
Figure 4. Effects of methyl substituents on the KI value. Values and errors are averages and SDs
from three or more independent measurements and are from Table 1.

Table 1. Potency of aryl/alkyl di-n-butyl
phosphates
No
R Group
KI (µM)
3a
50 16
3b
3c
4a
4b
4c
5a
5b
6a
6b
6c
6d
7
16
8
2
1
1
9
1.0 0.4
11
22
1
2
1.9 0.4
2
1
72
24
14
53
2
8
6
1
Prior studies reported organophosphate inhibitors of acetylcholinesterase are irreversible
inhibitors in which the inhibitor covalently modifies the nucleophilic serine in the active site.18, 19
Alternatively, the compounds could interact with the active site through noncovalent interactions
and bind reversibly. To test if compounds 3a - 3c act as irreversible or reversible inhibitors, we
conducted substrate rescue experiments. Briefly, if the inhibitors covalently modify the active
site (irreversible inhibition), then the addition of excess substrate is not expected to displace the

compound and reverse inhibition. If the compounds are reversible inhibitors, then addition of
excess substrate is expected to displace the inhibitor and rescue enzyme activity. As shown in
Supplementary Content Table 1 addition of substrate rescued activity for compounds 3a - 3c,
suggesting the compounds act as reversible inhibitors. As a control, excess substrate did not
rescue activity after incubation with the irreversible inhibitor dibutyl 2-chlorophenyl phosphate
(10 in Supplementary Content).
Table 2. Relative activity of AChE, BChE, and chymotrypsin
BChEa
Activity ratio*
AChEb
Activity ratio*
Chymotrypsinc
Activity ratio*
Analog
3a
0.27 0.03
0.15 0.02
0.02 0.01
0.04 0.02
0.01 0.01
0.09 0.01
0.21 0.01
0.01 0.01
0.01 0.01
0.04 0.01
0.12 0.01
0.04 0.02
0.11 0.01
0.99 0.01
0.99 0.07
0.89 0.01
1.04 0.04
1.00 0.08
1.09 0.02
1.07 0.01
1.05 0.07
1.00 0.01
0.98 0.07
0.86 0.01
0.87 0.05
1.02 0.03
0.92 0.05
0.89 0.04
0.96 0.01
0.94 0.02
0.95 0.04
0.84 0.10
0.93 0.04
0.76 0.05
0.97 0.02
0.94 0.08
0.94 0.06
0.92 0.08
0.89 0.02
3b
3c
4a
4b
4c
5a
5b
6a
6b
6c
6d
7
* Activity ratio = 500 µM Inhibitor/No Inhibitor
aBChE from equine serum (E.C. 3.1.1.8)
bAChE from Electrophorus electricus (E.C. 3.1.1.7)
cChymotrypsin from bovine pancreas (E.C. 3.4.21.1)
The KI values for all methylphenyl analogs indicate that the methyl group on the phenyl ring
leads to a better inhibitor compared to the phenyl group alone. The position of the methyl group
is important as substitution at the 3- and 4-position led to 6-12 fold decrease in KI, while
substitution at the 2-position led to only a 2-fold effect. If a more potent inhibitor results from the
addition of a methyl group that introduces favorable hydrophobic interactions with nonpolar

groups in the active site, then the position of these hydrophobic interactions relative to the
remainder of the phosphate inhibitor may be important. That is, each of the three analogs is
expected to introduce the same increase in hydrophobicity compared to the unsubstituted case.
However, substitution at the 3- or 4-position leads to a more potent inhibitor, relative to
substitution at the 2-position, suggesting that unfavorable interactions are present when binding
3a that is not found in 3b or 3c.
Building on the observation that methyl substitution at the 3- or 4-position of the phenyl ring
leads to a better inhibitor compared to the phenyl group alone, we next tested di- and trimethyl
analogs with methyl groups at the 3,4-, 3,5-, or 3,4,5-positions. If hydrophobic interactions with
the individual methyl groups at the 3- and 4-position lead to a better inhibitor, then the
introduction of methyl groups at multiple positions may further increase these interactions. The
KI values for 3,4-dimethyl phenyl phosphate (4a), 3,5-dimethyl phenyl phosphate (4b), and
3,4,5-trimethylphenyl phosphate (4c) are presented in Table 1. Results indicating the analogs are
selective for BChE relative to AChE are presented in the Table 2. The KI value for 3,5-
dimethylphenyl phosphate, 4b, was 95-fold lower than dibutyl phenyl phosphate 1d and 15-fold
lower than the analog bearing a single methyl group at the 3-position (3b) (Table 1). These
results suggest that additional favorable hydrophobic interactions are made with an analog
bearing methyl groups at the two meta positions. Meanwhile, where addition of a methyl group
at the 4-position led to a lower KI value relative to the unsubstituted phenyl ring, addition of a 4-
methyl group in an analog bearing methyl groups at the 3, 5-postions actually increased the KI
value by 10-fold. This suggests that simply increasing the steric bulk does not necessarily lead to
an inhibitor with a lower KI value. The context in which the steric bulk is added affects whether
additional methyl groups lead to a better inhibitor.

2.2.2. Evaluating dialkyl benzyl analogs as BChE inhibitors
We next examined benzyl dialkyl phosphate analogs to test if introducing conformational
flexibility of the aromatic group relative to the phosphate leads to better inhibitors. Inspection of
BChE X-ray structures with ligands bearing aromatic groups bound show the aromatic groups
situated in different conformations. While previous modeling studies have suggested
conformations in which the dialkyl phenyl phosphates are bound, experimental data supporting
these binding conformations are not available. The greater conformational flexibility of the
benzyl versus phenyl group might allow the substituted aromatic group to adopt a conformation
leading to more favorable hydrophobic interactions and thus increase binding.
We first measured the KI value for benzyl dibutyl phosphate (6a) and compared the results to
the dibutyl phenyl phosphate (1d) (Table 1 and Figure 5). The KI value of 2.3 µM was ~50-fold
lower than the value for the corresponding phenyl analog. This suggests that the additional
conformational flexibility afforded by the benzyl linker allows the phenyl group to adopt a
conformation that leads to more favorable binding interactions or may permit the aromatic group
to adopt a conformation that reduces unfavorable interactions.
If a single methylene group between the phosphate and aromatic ring results in a ~50-fold
lower KI value, then additional methylene groups that introduce even greater conformational
flexibility may lead to the aromatic group being positioned to make more favorable binding
interactions. To test this possibility, we determined the KI value for phenethyl phosphate 7
(Figure 5). The 2-carbon linker analog displayed a KI value of 53 µM, within 2-fold of dibutyl
phenyl phosphate, and a 28-fold higher KI value than that for 6a. Inhibition of 6a-6d was rescued
by excess substrate suggesting the compounds are reversible inhibitors (Supplementary Content
Table 1).

Figure 5. Effects of methyl substitutions on KI values in the benzyl scaffold. Values and errors
are averages and SDs from three or more independent measurements and are from Table 1.
As described above, methyl substitutions on the aromatic ring lead to lower KI values. To test
if methyl substitutions on the aromatic ring lead to better inhibitors in the benzyl dibutyl
scaffold, we measured the KI values for dibutyl phosphates bearing a 2-, 3-, or 4-methyl benzyl
group (6b – 6d) (Figure 5). In contrast to the results for the methylphenyl series, all methyl
group substitutions on the benzyl scaffold led to poorer inhibitors relative to 6a. This result
suggests that simply introducing steric bulk to increase van der Waals interactions may not lead
to better inhibitors.
2.2.3. Effect of expanding the aromatic moiety for dialkyl aryl phosphate inhibitors.
A number of cholinesterase inhibitors, such as tacrine, rivastigmine and galantamine, bear
aromatic groups. The X-ray crystallographic structures of BChE show the active site lined with
aromatic and hydrophobic groups and inhibitors containing aromatic groups can make favorable

binding interactions via van der Waals interactions and/or pi-pi stacking. To test if expanding the
aromatic moiety in a dialkyl phosphate leads to a more potent inhibitor, we examined the effect
of introducing a 1- and 2-naphthyl group in the dibutyl phosphate scaffold. The KI values of 22.2
and 1.9 µM for dibutyl 1-naphthyl phosphate (5a) and dibutyl 2-naphthyl phosphate (5b),
respectively, were lower than the dibutyl phenyl analog suggesting that the larger aromatic
moiety leads to more potent inhibitors (Table 1 and Figure 6). Enzyme activity was rescued in
the presence of excess substrate suggesting the compounds are reversible inhibitors
(Supplementary Content Table 1). The ~10-fold difference in KI values for the naphthyl analogs
suggests that the position of the naphthyl group relative to the phosphate affects the inhibition
properties. It is possible that this is related to the earlier observation that methyl group
substitution at the 3- and 4-position of the phenyl group leads to more effective inhibitors
compared to substitution of the 2-position. In the 2-naphthyl analog, the benzene ring is
connected to another aromatic ring at the 3- and 4-position, whereas in the 1-naphthyl analog the
benzene ring is connected to another ring at the 2- and 3-positon. For the phenyl analogs
described above methyl substitution at the 3- and 4-position lead to inhibitors with significantly
lower KI values than the phenyl analog substituted at the 2-position. This suggests that alkyl or
aromatic substitutions at the 2-position prevents the alkyl or aromatic group from making
favorable interactions, or substitution at the position closest to the phosphate leads to unfavorable
interactions leading to weaker binding.

Figure 6. Effects of naphthyl substitutions on KI. Values and errors are averages and SDs from
three or more independent measurements and are from Table 1.
2.2.4. Using dialkyl aryl phosphates to probe the connection of groups within inhibitors and
inhibition behavior.
The idiosyncratic composition of enzyme active sites introduces many challenges in the
design of effective inhibitors. While a series of molecules with varying substituents at a position
can be screened to identify candidate inhibitors, these changes can affect the contribution to
binding for other groups in the inhibitor. Phosphates 3 – 7 offered a convenient system in which
to explore the connection between different groups on the inhibitor and the inhibition behavior,
as changes can be readily introduced on the phenyl ring and the alkyl chains can be varied to
probe the connection between the different groups and the inhibition properties.
To further probe this connection we investigated the effect of perturbing the dialkyl chains
using the two inhibitors that showed the lowest KI values in the results described above: 3,5-
dimethyl phenyl and 2-naphthyl analogs (4b and 5b). As previously noted, Law et al. reported
that increasing the length of the dialkyl groups lead to inhibitors with lower KI values and

inhibitors more selective for BChE relative to AChE. To evaluate the connection between the
length of the alkyl chain and the identity of the aromatic group, we compared the inhibition
properties of aryl diethyl phosphates 8 and 9 and aryl di-n-butyl phosphates 4b and 5b, which
bear either 3,5-dimethylphenyl or 2-naphthyl groups.
In the diethyl phosphate scaffold, introduction of the 3,5-dimethyl groups led to a 3-fold
decrease in the KI value (Table 3 and Figure 7), and is significantly less than the 94-fold lower KI
value determined for introducing the 3,5-dimethylgroups in the dibutyl scaffold. Similarly,
expanding the aryl system from phenyl to 2-naphthyl led to smaller lowering in KI value in the
diethyl relative to the dibutyl analog, 11-fold compared to 50-fold, respectively. Our results
suggest that contribution of the aromatic moiety to inhibition is affected by the identity of the
alkyl chains as well. The presence of the longer dialkyl chains could affect the position of the
ligand bound in the active site allowing more extensive van der Waals interactions in the dibutyl
analogs leading to lower KI values.
Table 3. Potency of aryl di-n-ethyl phosphates
No
R Group
KI (µM)
a
-
1590 40
8
9
583 61
147 10
a Synthesis of di-n-ethyl phenyl phosphate reported
in Law et al.

Figure 7. Testing the effect of aromatic substitutions in diethyl and dibutyl phosphate scaffolds.
Effects of the 3,5-dimethyl (A) and 2-naphthyl (B) substitutions on the KI value in the diethyl
and dibutyl phosphate scaffolds. Values and errors are averages and SDs from three or more
independent measurements and are from Tables 1 and 3.

3. Molecular modeling studies
In discussing our molecular modeling results, we focus on the following three subgroups of
inhibitors examined in this study: the di-n-butyl methylphenyl series; the di-n-butyl naphthyl
series; and the di-n-butyl dimethylphenyl and trimethylphenyl series (Figure 8).
Figure 8: Inhibitors discussed in this subsection.
Our massive docking trials for each of the inhibitors in these three subgroups have led to two
primary observations regarding possible binding modes for these species. First and foremost, the
aromatic-group bearing substituent, ArO-, may be located in one of two possible regions within
the active site pocket, the first of which will be defined as the “lower region” of the pocket, near
residues Y128 and Q67, and bounded by residues T120 and choline-binding W82. This first
orientation is that assumed by the control species, dibutyl phenyl phosphate (1d). The alternative
ArO-group position, near W430 and Y440, will be defined as the “upper left region” of the
pocket. Figures 9a, 9b, and 9c provide identical views of bound inhibitors of the methylphenyl
series from the perspective of one looking directly into the binding pocket from the external
opening of the active site gorge and illustrate these two regions clearly. The preference of a
given ArO-group for one or both of these regions is dependent on the steric factors and the
chemistry of these regions and of the respective inhibitor under consideration.
As a result of ArO-group positioning, the P=O oxygen can take on multiple orientations with
respect to the binding pocket. In one possible orientation, this oxygen points “out of pocket”

Figure 9. Binding poses of the (a) o-methylphenyl (3a), (b) m-methylphenyl (3b), and (c) p-
methylphenyl dibutyl phosphate (3c) inhibitors in the active site of BChE from the perspective of
one viewing from the external entrance of the binding site gorge. The pocket surface is color-
coded to distinguish neutral (white), hydrophobic (green), anionic/hydrogen bond acceptor (red),
and cationic/hydrogen bond donor (blue) regions. Secondary structure of the surrounding region
is shown as white ribbons to convey the relative orientation of the protein.

toward the external opening of the gorge, fully solvent-exposed and lacking interactions with
residues within the gorge. Alternatively, this oxygen can point “into pocket” and take on one of
two notable orientations. When this oxygen takes on a downward facing orientation, strong
interactions with hydrogen bonding partners in the oxyanion hole (G116, G117, and A199) are
present. When this oxygen instead assumes an upward or rear-facing orientation, hydrogen
bonding with these residues is less significant, or lost entirely, and an unfavorable interaction
with the carboxyl group of residue E197 in the rear of the pocket occurs, making the former
orientation the preferred of the three orientations observed and the latter the least favorable.
As illustrated in Figures 9a - 9c, the di-n-butyl methylphenyl series 3a – 3c prefer ArO-group
placement in the upper left region of the pocket with the P=O oxygen pointing “out of pocket”
and thus leaving this moiety largely unimportant in the binding energetics of these species.
Observed docking scores reproduced the experimentally observed inhibition trend, with the p-
methylphenyl analog binding favored over the m-methylphenyl, and the m-methylphenyl analog
binding favored over o-methyl. Structural analysis shows that the o-methylphenyl substituent
(3a) takes on a rear-facing orientation (Figure 9a), located in a vacant region of the pocket
relatively void of potential interaction partners, and rotation of the o-methylphenyl group by
~180° would lead to considerable steric clash with the phosphate group. This lack of significant
participation by the methyl substituent during binding results in this species having relatively
weak inhibitory power, on the order of that of the control species, dibutyl phenyl phosphate (1d).
In comparison, the m-methyl substituent (3b) prefers a forward facing orientation while avoiding
the previously mentioned steric clash with the inhibitor phosphate group (Figure 9b). This
orientation allows m-methyl interaction with the sidechain of W430, thereby stabilizing the
binding of this species. Of the three possible substituent placements, this interaction is

maximized by the p-methyl species (3c), resulting in further enhancement of binding strength
and inhibitory power (Figure 9c).
The preferred binding modes of the 1-naphthyl (5a) and 2-naphthyl (5b) substituted species
are shown in Figures 10a and 10b, respectively, with lower panels presenting the same poses as
viewed after rotation of 80° about the vertical axis. The lower region of the pocket provides
ample volume for the placement of the naphthyl group substituents. In both cases, ArO-group
interaction with the choline-binding W82 residue is apparent and the P=O oxygen is oriented
“into pocket.” Yet, subtle but important differences are also observed, leading to the 2-naphthyl
species being a significantly more effective inhibitor, again in agreement with our experimental
observations. The lower panels of Figure 10 illustrate the contrast in positioning of the 1-
naphthyl (5a) and 2-naphthyl (5b) ArO-groups, with the former being positioned farther inside
the binding pocket and thus more sterically confined than the latter species. This positioning of
the 1-naphthyl species leaves the P=O oxygen positioned in significant contact with the E197
carboxyl group, while also limiting interactions with hydrogen bonding partners in the oxyanion
hole (G116, G117, and A199). In addition, the 1-naphthyl species has only limited interaction
with the hydroxyl group of T120 and cannot interact with residue Q67. In contrast, the location
of the 2-naphthyl substituent allows significant interactions with both the entire T120 sidechain
and the aminethe amine group of Q67, thereby providing numerous interactions by which this
inhibitor is stabilized.
Following our observations that larger ArO-groups can be sterically limited to the lower
region of the pocket, and that the location of a methyl substituent on the phenyl group can
enhance inhibitory activity to varying degrees, we assessed the experimentally observed trends
among di-n-butyl dimethylphenyl and trimethylphenyl phosphates 4a-4c. Of the three species

studied, a structural trend was quickly identified – though larger than their monomethyl analogs,
the 3,4-dimethylphenyl (4a) and 3,4,5-trimethylphenyl (4c) ArO-groups are unable to occupy the
lower region of the pocket, as do the naphthyl groups: the p-methyl substituent present in these
species conflicting with a structural “ridge” present at the bottom of the pocket between residues
Y128 and Q67. Figure 11a illustrates the resulting pose of the 3,4-dimethylphenyl species, with
the ArO-group located in the upper left region of the pocket. This species assumes the same
docking pose as that of the para monomethyl species (3c), with an added methyl group in the
meta position on the phenyl ring. Yet this added m-methyl group adds no significant interactions
with pocket residues, resulting in nearly identical inhibitory power as that of di-n-butyl p-
methylphenyl phosphate (3c).
In striking contrast, the 3,5-dimethylphenyl substituent of 4b, with a more broad ArO-group
profile, does not suffer the p-methylphenyl steric clash in the lower region of the pocket
described above, and this ArO-group is thus located in that lower region (Figure 11b). This
orientation allows the P=O oxygen to point “into pocket” in a downward facing direction,
thereby fostering interactions with hydrogen bonding partners in the oxyanion hole that are
lacking in the “out of pocket” oxygen orientation of the 3,4-dimethyl species (4a). Similar to the
observed docking poses of the 1-naphthyl (5a) and 2-naphthyl species (5b), placement of the
ArO-group in the lower region of the pocket (Figure 11c) allows significant interaction between
the 3,5-dimethyl groups and both W82 and T120. Additional interactions between the two m-
methyl groups and residues Q67 and Y128 provide added binding strength, thereby giving this
species inhibitory power on par with the highly-effective 2-naphthyl species (5b).

Figure 10. Binding poses of the (a) 1-napthyl and (b) 2-naphtyl dibutyl phosphate inhibitors in
the BChE active site, with color coding following that of Figure 9. Top panel images are oriented
identically to Figure 9 images, with bottom panels rotated 80° around the vertical axis.
Secondary structure of the surrounding region is omitted to best illustrate protein-inhibitor
interactions.

Figure 11. Binding poses of the (a) 3,4-dimethylphenyl and (b) 3,5-dimethylphenyl dibutyl
phosphate inhibitors in the BChE active site oriented identically to that in Figure 9. (c) Binding
pose of the 3,5-dimethylphenyl species shown in (b) rotated 80° around the vertical axis. Color
coding follows that of Figure 9 and secondary structure of the surrounding region is omitted to
best illustrate protein-inhibitor interactions.

Lastly, unlike the 3,5-dimethyl species (4b), the p-methyl substituent on the 3,4,5-
trimethylphenyl moiety (4c) causes the ArO-group to again be located in the upper left region of
the pocket (pose not shown), as in the case of di-n-butyl p-methylphenyl phosphate (3c). The
added m-methyl substituent results in a slight rotation of the ArO-group, similar to that observed
for the o-methyl species (3a) and positions the added m-methyl group in the vacant region
toward the rear of the pocket, as described above for 3a (Figure 9a). No significant interactions
to affect binding strength result from this added methyl substituent, leaving this species nearly
equal in inhibitory power as di-n-butyl 3,4-dimethylphenyl phosphate (4a) and di-n-butyl p-
methylphenyl phosphate (3c).
4. Conclusions
By studying the effects of introducing relatively simple changes in the structure of the
original phosphate 1d on BChE inhibition, we were able to develop several very potent, highly
selective BChE inhibitors with KI values down to 1/95th that of phosphate 1d. Our computational
studies uncovered two major binding modes for this class of inhibitors, where the substituent
bearing the aromatic group could occupy either the “upper left region” or the “lower region” of
the three subsites in the BChE active site gorge. Phosphates 4b and 5b, two of the most effective
inhibitors, both preferred to place their aromatic groups in the “lower region”, which also
allowed the P=O moiety to interact with hydrogen bond donor groups in the oxyanion hole.
This study demonstrates the utility of employing P(V) triesters as possible transition state or
tetrahedral intermediate mimics in developing effective inhibitors for the cholinesterases. While
clinically important cholinesterase inhibitors generally have KI values in the nanomolar range
and the current inhibitors show KI values in the micromolar range, the ease and flexibility by

which structural variations can be introduced into these triesters magnifies their utility as a
scaffold for new inhibitors. We are currently studying ways to exploit these findings to design
new inhibitors possessing even greater inhibitory properties.
5. Experimental Section
Chemistry
Diethyl chlorophosphate, di-n-butyl chlorophosphate and the alcohols were used as received
from commercial suppliers without further purification. Methylene chloride (DCM) and pyridine
were dried and stored over 3 Å molecular sieves. All glassware was dried and purged with dry
nitrogen before use. All reactions were performed under dry nitrogen. The rotary evaporator used
was Buchi R-3. Flash chromatographic purifications were performed by using silica gel (flash-
grade) from Dynamic Adsorbents Inc. and hexane and ethyl acetate as solvent system. TLCs
were ran on Whatman glass-backed silica gel plates and visualized with UV light and
phosphomolybdic acid (10%) in ethanol solution. NMR spectra were obtained on either a Bruker
Fourier 300 MHz or Bruker Avance III 400 MHz spectrometer with CDCl3 as the solvent and
CHCl3 residual peak as reference. Chemical shifts are given in parts per million (ppm), and all
coupling constants (J) are given in hertz (Hz). GC/MS analyses were performed on an Agilent
7820A/5975 GC/MS instrument. High resolution mass spectra (HRMS) were obtained by the
UCR Mass Spectrometry Facility on an Agilent 6210 LCTOF instrument operated in the
multimode source configuration.
General procedure for the preparation of triphosphates. To a dried two or three-necked 150
mL round bottom flask, 1.04 mL (1.14 g, 5.00 mmol) of di-n-butyl chlorophosphate was

dissolved in 12 mL CH2Cl2 with stirring. The solution was cooled to 0 oC. Then 2.0 equivalents
of alcohol and 2.0 equivalents of pyridine were added to the solution via syringe. The reaction
was stirred overnight at room temperature. The reaction mixture was diluted with 80 mL of
diethyl ether and washed three times with 30 mL of 10% HCl. The aqueous layers were
combined and washed with 40 mL of CH2Cl2. The combined organic layer was washed once
with 40 mL of saturated sodium bicarbonate, once with 40 mL of brine, dried over magnesium
sulfate, filtered, concentrated in vacuo. The products were purified by flash chromatography.
Di-n-butyl phenyl phosphate (1d). Following the general procedure described above and using
phenol, 1d was obtained as a yellow oil: 50% yield. 1H NMR (300 MHz, CDCl3) ꢀ 0.91 (t, J =
7.8 Hz, 6H, (OCH2CH2CH2CH3)2), 1.33-1.45 (m, 4H, (OCH2CH2CH2CH3)2), 1.62-1.71 (m, 4H,
(OCH2CH2CH2CH3)2), 4.06-4.19 (m, 4H, (OCH2CH2CH2CH3)2), 7.13-7.35 (m, 5H, Ph-H); 13C
NMR (75 MHz, CDCl3) ꢀ 13.52, 18.60, 32.24, 68.32, 119.93, 124.94, 129.67, 150.74. MS
(gc/ms) m/z (%): 286 (M+, 8.4), 175 (100), 94 (54.9).
Di-n-butyl o-methylphenyl phosphate (3a). Following the general procedure described above
and using o-methylphenol, 3a was obtained as a light yellow oil: 37% yield. 1H NMR (300 MHz,
CDCl3) ꢀ 0.91 (t, J = 7.4 Hz, 6H, (OCH2CH2CH2CH3)2), 1.33-1.45 (m, 4H,
(OCH2CH2CH2CH3)2), 1.62-1.71 (m, 4H, (OCH2CH2CH2CH3)2), 2.31 (s, 3H, Ph-CH3), 4.09-
4.19 (m, 4H, (OCH2CH2CH2CH3)2), 7.03-7.29 (m, 4H, Ar-H); 13C NMR (75 MHz, CDCl3) ꢀ
15.53, 16.34, 18.62, 32.19, 68.26, 119.70, 124.86, 126.96, 129.25, 131.26,149.22. MS (gc/ms)
m/z (%): 300 (M+, 45.9), 188 (100), 107 (100). HRMS: m/z calculated for C15H26O4P (MH+)
301.1563. Found 301.1521.

Di-n-butyl m-methylphenyl phosphate (3b). Following the general procedure described above
and using m-methylphenol, 3b was obtained as a light yellow oil: 39% yield. 1H NMR (300
MHz, CDCl3) ꢀ 0.84 (t, J = 7.4 Hz, 6H, (OCH2CH2CH2CH3)2), 1.26-1.38 (m, 4H,
(OCH2CH2CH2CH3)2), 1.55-1.64 (m, 4H, (OCH2CH2CH2CH3)2), 2.26 (s, 3H, Ph-CH3), 4.03-
4.10 (m, 4H, (OCH2CH2CH2CH3)2), 6.88-7.21 (m, 4H, Ar-H); 13C NMR (75 MHz, CDCl3) ꢀ
13.53, 18.61, 21.30, 32.16, 68.22, 116.84, 120.59, 125.67, 129.31, 139.86,150.67. MS (gc/ms)
m/z (%): 300 (M+, 43.7), 188 (100), 108 (49.2). HRMS: m/z calculated for C15H26O4P (MH+)
301.1563. Found 301.1531.
Di-n-butyl p-methylphenyl phosphate (3c). Following the general procedure described above
and using p-methylphenol, 3c was obtained as a colorless oil: 40% yield. 1H NMR (300 MHz,
CDCl3) ꢀ 0.92 (t, J = 7.4 Hz, 6H, (OCH2CH2CH2CH3)2), 1.33-1.46 (m, 4H,
(OCH2CH2CH2CH3)2), 1.62-1.71 (m, 4H, (OCH2CH2CH2CH3)2), 2.31 (s, 3H, Ph-CH3), 4.09-
4.18 (m, 4H, (OCH2CH2CH2CH3)2), 7.07-7.14 (m, 4H, Ar-H); 13C NMR (75 MHz, CDCl3) ꢀ
13.54, 18.62, 20.71, 32.17, 68.14, 119.71, 130.10, 134.48, 148.64. MS (gc/ms) m/z (%): 300
(M+, 38.6), 188 (100), 108 (76.8). HRMS: m/z calculated for C15H26O4P (MH+) 301.1563.
Found 301.1528.
Di-n-butyl 3, 4-dimethylphenyl phosphate (4a). Following the general procedure described
above and using 3, 4-dimethylphenol, 4a was obtained as a light yellow oil: 39% yield. 1H NMR
(300 MHz, CDCl3) ꢀ 0.92 (t, J = 7.4 Hz, 6H, (OCH2CH2CH2CH3)2), 1.34-1.46 (m, 4H,
(OCH2CH2CH2CH3)2), 1.62-1.71 (m, 4H, (OCH2CH2CH2CH3)2), 2.21 (s, 3H, Ph-CH3), 2.23 (s,

3H, Ph-CH3), 4.09-4.16 (m, 4H, (OCH2CH2CH2CH3)2), 6.92-7.07 (m, 3H, Ar-H) ; 13C NMR (75
MHz, CDCl3) ꢀ 14.12, 19.18, 20.45, 3.1.51, 68.63, 69.01, 117.49, 121.46, 130.96, 133.68,
138.65, 149.17. MS (GC/MS) m/z (%): 314 (M+, 71), 258 (71), 203 (100). HRMS: m/z
calculated for C16H28O4P (MH+) 315.1720. Found 315.1663.
Di-n-butyl 3, 5-dimethylphenyl phosphate (4b). Following the general procedure described
above and using 3, 5-dimethylphenol, 4b was obtained as a yellow oil: 43% yield. 1H NMR (400
MHz, CDCl3) ꢀ 0.89 (t, J = 7.4 Hz, 6H, (OCH2CH2CH2CH3)2), 1.33-1.42 (m, 4H,
(OCH2CH2CH2CH3)2), 1.61-1.68 (m, 4H, (OCH2CH2CH2CH3)2), 2.26 (s, 6H, Ph-(CH3)2), 4.08-
4.13 (m, 4H, (OCH2CH2CH2CH3)2), 6.76 (s, 1H, Ar-H), 6.81 (s, 2H, Ar-H); 13C NMR (100
MHz, CDCl3) ꢀ 11.66, 16.79, 19.35, 30.43, 66.31, 115.73, 124.72, 137.61, 148.88. MS (gc/ms)
m/z (%): 314 (M+, 8.4), 202 (100), 122 (41.4). HRMS: m/z calculated for C16H28O4P (MH+)
315.1720. Found 315.0501.
Di-n-butyl 3, 4, 5-trimethylphenyl phosphate (4c). Following the general procedure described
above and using 3, 4, 5-trimethylphenol, 4c was obtained as a light yellow oil: 28% yield. 1H-
NMR (300 MHz, CDCl3) ꢀ 0.92 (t, J = 7.4 Hz, 6H, (OCH2CH2CH2CH3)2), 1.34-1.47 (m, 4H,
(OCH2CH2CH2CH3)2), 1.62-1.72 (m, 4H, (OCH2CH2CH2CH3)2), 2.11 (s, 3H Ph-CH3), 2.25 (s,
6H, Ph-(CH3)2), 4.09-4.16 (m, 4H, (OCH2CH2CH2CH3)2), 6.86 (s, 2H, Ar-H); 13C NMR (100
MHz, CDCl3) ꢀ 13.72, 15.00, 18.75, 20.85, 32.34, 68.24, 118.81, 131.83, 137.96, 147.95. MS
(GC/MS) m/z (%): 328 (M+, 25), 216 (100), 91 (63). HRMS: m/z calculated for C17H30O4P
(MH+) 329.1876. Found 329.1846.