Benzoquinones as inhibitors of botulinum neurotoxin serotype A

Article abstract of DOI:10.1016/j.bmc.2014.06.004

Although botulinum neurotoxin serotype A (BoNT/A) is known for its use in cosmetics, it causes a potentially fatal illness, botulism, and can be used as a bioterror weapon. Many compounds have been developed that inhibit the BoNTA zinc-metalloprotease lig

Full text of DOI:10.1016/j.bmc.2014.06.004

Bioorganic & Medicinal Chemistry 22 (2014) 3971–3981
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Benzoquinones as inhibitors of botulinum neurotoxin serotype A
Paul T. Bremer a, Mark S. Hixon b, Kim D. Janda a,
⇑
a Departments of Chemistry and Immunology, The Scripps Research Institute, 10550 North Torrey Pines Road, La Jolla, CA 92037, USA
b Discovery Biology, Takeda San Diego, Inc., 10410 Science Center Drive, San Diego, CA 92121, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Although botulinum neurotoxin serotype A (BoNT/A) is known for its use in cosmetics, it causes a poten-
tially fatal illness, botulism, and can be used as a bioterror weapon. Many compounds have been devel-
oped that inhibit the BoNTA zinc-metalloprotease light chain (LC), however, none of these inhibitors have
advanced to clinical trials. In this study, a fragment-based approach was implemented to develop novel
covalent inhibitors of BoNT/A LC. First, electrophilic fragments were screened against BoNT/A LC, and
benzoquinone (BQ) derivatives were found to be active. In kinetic studies, BQ compounds acted as irre-
versible inhibitors that presumably covalently modify cysteine 165 of BoNT/A LC. Although most BQ
derivatives were highly reactive toward glutathione in vitro, a few compounds such as natural product
naphthazarin displayed low thiol reactivity and good BoNT/A inhibition. In order to increase the potency
of the BQ fragment, computational docking studies were employed to elucidate a scaffold that could bind
to sites adjacent to Cys165 while positioning a BQ fragment at Cys165 for covalent modification; 2-
amino-N-arylacetamides met these criteria and when linked to BQ displayed at least a 20-fold increase
Received 8 April 2014
Revised 31 May 2014
Accepted 2 June 2014
Available online 16 June 2014
Keywords:
Benzoquinone
Irreversible inhibitor
Covalent inhibitor
Naphthoquinone
Juglone
Naphthazarin
Botulinum neurotoxin
Fragment-based design
Rational design
Computational docking
Cysteine targeting
in activity to low
lM IC50 values. Unlike BQ alone, the linked-BQ compounds demonstrated only weak
irreversible inhibition and therefore acted mainly as non-covalent inhibitors. Further kinetic studies
revealed a mutual exclusivity of BQ covalent inactivation and competitive inhibitor binding to sites adja-
cent to Cys165, refuting the viability of the current strategy for developing more potent irreversible
BoNT/A inhibitors. The highlights of this study include the discovery of BQ compounds as irreversible
BoNT/A inhibitors and the rational design of low
moiety for activity.
lM IC50 competitive inhibitors that depend on the BQ
Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction
weaponization.1,4 The toxin is naturally produced during sporula-
tion by Clostridium botulinum, an anaerobic, Gram-positive bacte-
rium. If grown in sufficient quantities C. botulinum can be
disseminated into food supplies or adsorbed onto fine particles
for aerosolization.4 An actual BoNT/A bioterror attack on a human
population would result in widespread acute flaccid paralysis and
bulbar palsies (resulting in difficulty speaking, swallowing and
chewing).1 Although no bioterror attacks involving BoNT/A have
been successfully executed, many countries such as Iran, Iraq,
North Korea and Syria have developed and/or stockpiled weapons
containing botulinum toxin.1
In contrast to bioterrorism, the most common human exposure
to botulinum toxin takes the form of a foodborne illness known as
botulism. Treatment for botulism consists of FDA-approved
antibody-derived antitoxins, however, antitoxins must be adminis-
tered immediately after exposure to the toxin to achieve efficacy.5
Moreover, these antitoxins cannot neutralize toxins that have been
endocytosed into neurons. The BoNT/A mechanism of action
involves endocytosis of the 150 kDa holotoxin via the 100 kDa
Botulinum toxin is the most toxic known substance and has an
estimated intravenous LD50 of 1–2 ng/kg in humans.1 Eight differ-
ent serotypes of botulinum toxin exist, each with their own poten-
cies and modes of action, however, all serotypes are neurotoxic by
means of blocking acetylcholine release at the neuromuscular
junction causing muscle paralysis. The most potent botulinum
neurotoxin serotype, serotype A (BoNT/A), is widely recognized
as the commercial product BOTOXÒ, used cosmetically to reduce
facial wrinkles. When administered in low doses, BoNT/A is a vital
therapeutic used to treat a variety of conditions characterized by
uncontrollable muscle spasms such as blepharospasm (spasmotic
eye closure) and dysphonia (vocal fold spasms).2,3 On the other
hand, BoNT/A is considered a significant bioterror threat due to
its high potency and relative ease of mass production and
⇑
Corresponding author. Tel.: +1 (858) 784 2516.
E-mail address: kdjanda@scripps.edu (K.D. Janda).
http://dx.doi.org/10.1016/j.bmc.2014.06.004
0968-0896/Ó 2014 Elsevier Ltd. All rights reserved.

3972
P. T. Bremer et al. / Bioorg. Med. Chem. 22 (2014) 3971–3981
heavy chain into neurons.6 Subsequently, the 50 kDa zinc-metallo-
protease light chain (LC) of BoNT/A cleaves the 25 kDa SNAP-25,
one of three SNARE complex proteins responsible for fusing acetyl-
choline-containing vesicles to synaptic plasma membranes.7 For
the past 10 years, a significant effort has been put forth to develop
peptide and small molecule inhibitors of the BoNT/A LC.8–11 With
the exception of chicoric acid as an exosite inhibitor, most BoNT/
A LC inhibitors bind to the active site and typically contain a zinc
chelating moiety such as hydroxamic acids, however, two reports
exist of covalent BoNT/A inhibitors.12,13 Unfortunately, no known
compounds possess noteworthy in vivo efficacy in ameliorating
BoNT/A-induced toxicity; therefore, discovery of novel BoNT/A LC
inhibitors continues to be an important research endeavor.
BoNT/A at [I] = 100 mM. We reasoned that the inhibitory activity of
N-ethylmaleimide may be attributed to its cyclic structure, prompt-
ing us to investigate other cyclic electrophiles. 1,4-Benzoquinones
(BQs) and 1,4-naphthoquinones (NQs) were selected for screening
due to their known cysteine reactivity. Results from the SNAPtide
assay revealed that BQs and NQs can be potent inhibitors of
BoNT/A (Table 1) and that their inhibitory activity was time
dependent suggesting an irreversible mode of inhibition.
The potency of BQs and irreversible inhibitors in general
depends on two factors: affinity for the target (KI) and rate of cova-
lent bond formation with the target residue (kinact) (Fig. 1). To
account for the inhibitory mechanism of covalent inhibitors, we
expressed the inhibitory potential of each compound in terms of
kinact/KI (Table 1). Additionally, we elucidated the inhibitory mech-
anism of 5-OH-NQ (7) and BQ (3) by testing these compounds at a
wide range of concentrations and preincubation times in the SNAP-
tide assay. Results from the multi-dose experiment indicate that
concentration of these compounds is directly proportional to kobs
of BoNT/A inhibition while saturating kinetics was not observed
The active site of BoNT/A contains a cysteine residue (165) that
has recently been shown to be essential for catalytic activity. In
mutagenesis studies, swapping Cys165 for a serine drastically
reduced catalytic activity 50-fold. Furthermore, incubation of
BoNT/A with a thiol reactive compound (3-aminopropyl)methane-
thiosulfonate (MTSPA) irreversibly inhibited catalytic activity
(Ki = 7.7
l
M).14 In light of this data, we sought to uncover novel
(although kobs for 5-OHNQ appears to tail off slightly at 200 lM)
covalent inhibitors of BoNT/A which have the advantage of
persistently inactivating the toxin long after initial exposure to
the inhibitor. Irreversible inhibition is especially desirable for
BoNT/A because the toxin has a very long half-life (ꢀ10 days)
causing symptoms of intoxication for 4–6 months.15 From screen-
ing electrophilic fragments, we have found that 1,4-benzoquinone
(BQ) derivatives are potent irreversible inhibitors of BoNT/A. We
attempted to enhance the activity of the BQs via fragment-based
design to increase the effective molarity of the electrophilic
warhead relative to Cys165.
(Fig. 2). Furthermore, when the compounds were preincubated
with BoNT/A and diluted 50 fold into substrate, remaining BoNT/
A activity decreased exponentially as a function of preincubation
time with inhibitor (Fig. 3). The resulting kinact/KI values in both
assays were almost identical, thus confirming the irreversible inhi-
bition mechanism of BQ and NQ.
2.2. SAR
A series of known BQ analogues was synthesized and tested
to thoroughly probe the structure–activity relationship of this
BQs are highly relevant to biological systems and are well
known for their therapeutic properties. Many BQs are produced
naturally by certain plants for example thymoquinone (23) is
found in black cumin (Nigella sativa) and juglone (7) and naphthaz-
arin (13) are found in certain species of walnut trees of the genus
Juglans.16,17 BQs, namely quinone anti-cancer drugs, can elicit cyto-
toxic effects via reduction by various enzymes forming reactive
oxygen species and quinone methides, both of which can damage
(or alkylate) biomolecules, for example, DNA.18,19 In contrast, many
quinone-containing molecules such as endogenously-synthesized
ubiquinone (coenzyme Q10) act as anti-oxidants.20 Upon bioreduc-
tion, ubiquinone and related compounds protect against lipid
peroxidation, DNA oxidation and protein degradation.21 Despite
potential toxicity associated with BQ compounds, medicinal
chemistry campaigns to develop irreversible inhibitors of VEGFR-
2 as anti-cancer drugs have employed BQ moieties to covalently
modify specific cysteine residues.22,23 In our study, we used a
similar strategy to target Cys165 in BoNT/A light chain.
Table 1
Activity of BQ and NQ derivatives as covalent inactivators of BoNT/A LC
O
O
8
7
6
2
3
6
5
2
3
1
4
1
4
5
O
O
1,4-Benzoquinone (BQ)
1,4-Naphthoquinone (NQ)
kinact/KI (Mꢁ1 sꢁ1
)
Compound #
Name
1
2
3
4
5
6
7
8
2,5-DiCl-BQ
2-Cl-BQ
BQ
2-(4-I-Ph)-BQ
2-Ph-BQ
2-OMe-3-Tol-BQ
5-OH-NQ
84
51
17
10
9.7
9.5
5.3
4.7
4.1
5-OCyclopentoyl-NQ
2-OMe-BQ
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
5-OAc-NQ
2-Estrone-BQ
2-Me-BQ
5,8-diOH-NQ
5-OBn-NQ
NQ
2-(2-COOH-Et)-BQ
2-Me-NQ
5-OMe-NQ
6-OH-NQ
2-Tol-NQ
2,5-diOMe-3-Tol-BQ
2-(COOH-Me)-BQ
2-iPr-5-Me-BQ
2,6-diOMe-BQ
2,6-diMe-BQ
2-OMe-5-Tol-BQ
4.0
3.9
3.5
2.2
2.0
2.0
1.2
1.1
0.99
0.94
0.91
0.82
0.56
0.38
NA
NA
NA
2. Results and discussion
2.1. Discovery of benzoquinones as irreversible BoNT/A
inhibitors
In an effort to discover new irreversible inhibitors of BoNT/A, we
screened a series of small molecular weight electrophiles mostly
containing an alpha, beta unsaturated carbonyl motif. We chose a
commonly used assay for BoNT/A inhibitor screening which
involves the SNAPtide™ FRET substrate.24 The 13 amino acid
SNAPtide substrate mimics the region of the native SNAP-25
substrate that binds to the BoNT/A active site. BoNT/A LC readily
cleaves SNAPtide producing fluorescence by releasing the donor
and acceptor chromophores at the SNAPtide termini. In our
SNAPtide assay screen, the only electrophile that possessed any
inhibitory activity was N-ethylmaleimide which weakly inactivated
Compounds were tested at 50
NA = not active.
lM in the SNAPtide assay over a 1.5 h period.

P. T. Bremer et al. / Bioorg. Med. Chem. 22 (2014) 3971–3981
3973
2.3. Thiol reactivity
KI
kinact
E
I
E I
E I
We measured thiol reactivity independent of BoNT/A affinity by
incubating our compounds with glutathione and measuring free
thiol concentrations spectrophotometrically at various time points
with Ellman’s reagent (5,50-dithiobis-(2-nitrobenzoic acid), DTNB).
Thiol reactivity is important in the context of covalent inhibitors
because electrophilic chemotypes can react readily with endoge-
nous thiols, for example, glutathione, creating off-target effects
while reducing drug efficacy.26,27 Overall, BQ thiol reactivity was
very high compared to linear alpha–beta unsaturated carbonyl
compounds, for example, acrylamides and only a few compounds
were unreactive enough to calculate a second order rate constant
(kGSH, Table 2). Even these compounds (13, 18, 19, 21) possessed
kGSH values of 30 times greater than N,N-dimethylacrylamide
(kGSH = 0.0011 minꢁ1), a Michael acceptor moiety similar to that
found in ibrutinib which is an FDA-approved drug for treatment
of mantle cell lymphoma and chronic lymphocytic leukemia.28–30
BQ compounds exhibiting the greatest inhibitory activity in the
SNAPtide assay typically demonstrated high thiol reactivity. As
expected, these compounds contained electron withdrawing
groups (see compound 1) which raise LUMO energies of the BQs,
causing BQs to more readily accept electrons from thiol HOMOs.22
However, many compounds did not exhibit a positive correlation
between LUMO energy and BoNT/A activity. For example N-ethyl-
maleimide showed one of the highest thiol reactivities but was one
of the weakest inhibitors tested. On the other hand, natural prod-
uct naphthazarin (13, 5,8-dihydroxynaphthoquinone) exhibited
one of the lowest thiol reactivity profiles while retaining good
inhibitory activity. Clearly, binding affinity in addition to thiol
reactivity, plays an important role in governing BoNT/A inhibition
of these irreversible inhibitors. Given our results, thiol reactivity
and binding affinity could in theory be further tuned to yield even
more potent covalent inhibitor fragments.
Figure 1. Mechanism of action of covalent inhibitors. Inhibitor (I) binds to the
enzyme (E) with a certain affinity (KI) to form the enzyme–substrate complex (EꢂI).
Then, the inhibitor irreversibly forms a covalent bond with the enzyme according to
a certain rate constant (kinact) to form the inactivated enzyme–substrate complex
(E–I).
Figure 2. Dose-dependent inactivation of BoNT/A LC by BQ derivatives. Inhibitors
were incubated with BoNT/A LC in the presence of SNAPtide substrate and
fluorescence was measured over a 1.5 h period. Values for kobs were calculated for
each inhibitor concentration.
2.4. Strategies to improve benzoquinone potency by linking
benzoquinone to other pharmacophores
Figure 3. Time-dependent inactivation of BoNT/A LC by BQ compounds. Inhibitors
were preincubated with BoNT/A LC for various time periods and diluted 50 fold into
SNAPtide substrate. Remaining activity was determined for each time point as a
ratio of initial velocities.
We sought to explore the possibility of combining BQ with
other pharmacophores to enhance inhibitor potency and selectiv-
ity. Since the BoNT/A active site zinc is directly adjacent to cysteine
165, we hypothesized that a dual zinc chelator/cysteine trap inhib-
itor could be highly potent. In theory, linking BQ to a fragment with
high affinity for the adjacent active site could increase the effective
chemotype. Manipulation of both the steric and electronic charac-
ter of the BQ scaffold had a large impact on inhibitory activity.
Generally, electronics had the greatest impact on inhibitory activ-
ity since electron withdrawing groups increased activity while
electron donating groups decreased it. The most significant substi-
tuent effect was observed with the addition of a chlorine atom at
the 2 and/or 5 position of the BQ ring. The dichlorinated BQ (1)
molarity of BQ relative to Cys165, thus greatly increasing kinact
.
Moreover, the crystal structure of MTSPA covalently bound to
BoNT/A reveals a potential interaction between the zinc and the
MTSPA amine.14 To evaluate our linked fragment inhibitors, we
employed a more robust assay involving a 66-mer peptide sub-
strate that contains the 66 amino acids most essential for BoNT/A
LC binding of SNAP-25. By means of LCMS, the assay quantifies
the amount of 9-mer cleaved by BoNT/A LC relative to an isotopi-
cally labeled 9-mer internal standard.31 Also, BQ (3) inactivation
in the 66-mer assay was comparable to the SNAPtide assay.
We pursued a rational, fragment-based design of a BQ-linked
inhibitor to effectively position the electrophile close to Cys165.
From the co-crystal structure of a previously reported peptide
inhibitor of BoNT/A,9 we gleaned that an amino acetamide moiety
could chelate zinc, while accommodating both a fragment to bind
into a nearby hydrophobic pocket (S10 pocket) and fragment to
covalently modify Cys165. Indeed, we found that a series of
2-amino-N-(aryl)acetamides could weakly inhibit BoNT/A with
when incubated with BoNT/A at 50 lM completely abrogated cat-
alytic activity after 20 min, and therefore stands as one of the most
potent covalent BoNT/A inhibitors ever reported. However, the
major liability of this inhibitor is that the chlorination appears to
almost exclusively influence electronics over sterics; the electro-
negative atom heightens the electrophilicity and thiol reactivity
(Table 2) without increasing binding affinity. In terms of steric
characteristics that influence BQ inhibitory activity, the 2,3-substi-
tution appears to be most favorable compared to 2,5; 2,6; or 2,3,5/
6. The best demonstration of this is comparing 2,3 (6) and 2,5-
methoxytolylBQ (26) in which case the former is much more
potent (Table 1). The favorability of the 2,3-substitution prompted
us to test a series of juglones (5-hydroxynaphthoquinones) which
were functionalized at the 5-OH, a strategy previously used to
develop anti-cancer and anti-inflammatory compounds.25 None
of the tested derivatives displayed better activity relative to the
parent compound (Table 1).
the most potent being aryl = 4-phenoxyphenyl (27), IC50=120 lM
(Fig. 6). We performed docking studies using previous methodol-
ogy10,47 to find the best amino acetamide that when linked to
BQ, would properly position it near Cys165. Results from the

3974
P. T. Bremer et al. / Bioorg. Med. Chem. 22 (2014) 3971–3981
Table 2
Thiol reactivity of select compounds
Compound
#
%GSH remaining at
10 s
%GSH remaining at
30 min
kGSH
(minꢁ1
)
21
13
18
19
14
10
15
8
24
7
12
6
89
85
80
71
57
51
47
46
45
44
35
25
24
22
21
0
49 (35)
39 (26)
27 (14)
32 (22)
39
26
20
25
0
32
5
19
6
0.032
0.025
0.053
0.036
3
25
5
0
8
0
1
Figure 5. Docking studies for optimizing lipophilic tail. Docking studies with
various lipophilic tails indicate that phenoxyphenyl, benzyloxyphenyl and biphenyl
groups gave high fitness scores while a methylene linker in between the tail and
aminoacetamide provides an ideal angle to project a linked BQ toward Cys165.
Blue = enzyme, green = compound without CH2 linker, pink = compound with CH2
linker.
1 h GSH% values are indicated with (). 1 mM of each compound was incubated with
1 mM glutathione (GSH) and the amount of free GSH remaining was determined by
DTNB. Only 4 compounds were unreactive enough to calculate a second order rate
constant (kGSH) of Michael adduct formation.
docking studies indicated that the amino acetamide with aryl =
3-benzyloxyphenyl (28) connected to BQ via a methylene linker,
perfectly positioned all three pharmacophores (compound 30,
Fig. 4). The benzyloxyphenyl moiety is buried within the S10 pocket
(back left), the amino acetamide is interacting with zinc (green
sphere) while the BQ is ideally positioned to covalently modify
Cys165 (yellow sphere = S). In light of this discovery and with fur-
ther docking studies, we designed two other analogues (31, 32).
We found that phenoxyphenyl and biphenyl with a methylene
linker to the aminoacetamide gave favorable fitness scores while
providing a perfect angle to position a linked BQ fragment near
Cys165 (Fig. 5). Heterocycles (pyridine and triazole) were substi-
tuted in place of the secondary amine in order to constrain the
rotatable bonds and to decrease the basicity of the amine (a poten-
tial liability in the presence of BQ).
comparing 28 to 31 and 64-fold in comparing 29 to 32) (Fig. 6).
However, the BQ linked inhibitors showed only weak time-depen-
dent inactivation of BoNT/A, suggesting that the compounds acted
primarily as competitively inhibitors. In fact, when 31 was prein-
cubated for 30 min at 50 lM with BoNT/A and diluted into sub-
strate, enzyme activity was only slightly reduced (kinact
/
KI = 1.7 Mꢁ1 sꢁ1). In comparing 31 to other BQs (Table 1) it lies
within the weakest 25% of all BQs tested based on covalent inacti-
vation, despite giving a low IC50 value in the competitive 66-mer
assay.
We further probed the SAR of the linked BQ inhibitors by testing
various fragments and 2,5-dimethoxyphenyl precursors of inhibi-
tors 30–32. Although the 2,5-dimethoxyphenyl moiety in place of
BQ was well tolerated in 33 versus 30, it led to a decrease in activ-
ity for 34 versus 31 and 38 versus 32 (Fig. 7). Evaluation of frag-
ments of 31 (35–37), elucidated that the carbonyl, pyridine and
benzoquinone were the most important functional groups for the
In comparing the activities of the linked BQ inhibitors versus
the analogous amino acetamides, the addition of the BQ increased
potency to low lM IC50 (26-fold in comparing 27 to 30, 21-fold in
Figure 4. Compound 30 docked into the active site of BoNT/A LC. Lowest energy conformations were generated using OMEGA and then docked into 4ELC and 2IMB BoNT/A LC
crystal structures using FRED. Compound 30 gave one of the highest fitness scores while perfectly aligning all pharmacophores within the enzyme. The lipophilic
benzyloxyphenyl tail is buried within the S10 pocket (back left), the aminoacetamide is interacting with the zinc (green sphere) while the benzoquinone is perfectly
positioned for covalent interaction with cys 165 (S = yellow sphere).

P. T. Bremer et al. / Bioorg. Med. Chem. 22 (2014) 3971–3981
3975
O
O
O
O
O
O
H
N
NH2
N
H
N
H
30
27
IC50=18 µM
IC50=417 µM
O
O
O
H
N
O
N
O
N
H
NH2
N
H
O
31
28
IC50=2.3 µM
IC50=120 µM
O
O
O
NH2
N
N
N
H
N
H
O
N
29
32
IC50=160 µM
IC50=2.5 µM
Figure 6. Activity and structures of aminoacetamide inhibitors and their BQ-linked analogues. As predicted by docking studies, linkage of BQ to aminoacetamide inhibitors
greatly enhances inhibitory activity in the 66-mer assay. However, BQ covalent inactivation by these compounds did not significantly contribute to overall BoNT/A inhibition.
activity of 31. Furthermore, high IC50 values from fragments 35 and
39 suggest that the lipophilic aryl groups do not possess good
group efficiency (GE). This result corroborates the low activity of
aminoacetamides found in Figure 6. However, a 7-fold increase
in activity was observed by adding the phenoxyphenyl fragment
(compare 36 to 31). Overall SAR studies show that every fragment
of linked inhibitors 30–32 contributes to activity, suggesting that
the inhibitors are assuming the Figure 4 binding mode we had
intended based on docking studies. Although significant covalent
modification of Cys165 was not observed, this result can be
explained by the fact that the docking software does not select
for possible inhibitor–protein covalent interactions; the software
identified a BQ binding site near Cys165 (Fig. 4) in both Cys165
accessible (4ELC) and inaccessible (2IMB) crystal structures, how-
ever, that is no guarantee the BQ will covalently interact with
Cys165.
In light of the failure of our linked benzoquinones to act as
effective BoNT/A covalently inactivators, we simplified our strategy
to include small molecular weight dual zinc chelator/cysteine trap
compounds. We tested both pyridyl acrylamides and carboxyl
O
O
O
H
N
N
H
O
33
30
(precursor to
)
IC50=25 µM
O
O
O
H
N
N
N
H
O
34
31
(precursor to
IC50= 200 µM
)
O
O
H
N
O
O
O
H
N
N
N
N
N
H
O
O
36
31
)
O
(fragment of
37
31
)
(fragment of
35
31
(fragment of
)
IC50=14 µM
IC50=97 µM
IC50=570 µM
O
O
O
N
N
H
NH
N
N
H
O
N
N
N
39
32
)
(fragment of
IC50=210 µM
38
32
)
(precursor to
IC50=40 µM
Figure 7. Activity and structures of various precursors and fragments of compounds 30–32. Results from testing compounds in the 66-mer assay reveal that the BQ moiety is
the most important moiety for activity; removal of lipophilic tail is less critical. Replacement of BQ with precursor dimethoxyphenyl groups led to a decrease in activity.

3976
P. T. Bremer et al. / Bioorg. Med. Chem. 22 (2014) 3971–3981
H
H
N
N
N
N
O
O
40
41
IC50=100 µM
IC50=48 µM
O
O
O
HO
HO
Derivatization
R
HO
O
O
R = H, 2,5-diMeO, 4-Cl, 4-Ph, 3,4-OCH2O
No Activity
O
22
16
Not Active
O
IC50=9.3 µM
0.57 LE
Figure 8. Activity and structures of dual zinc chelating/cysteine trap inhibitors. 66-mer assay results indicate that 22 is a high LE inhibitor of BoNT/A LC, and that pyridyl
acrylamides are weak inhibitors. However, BoNT/A inhibition by these compounds is not significantly impacted by covalent inactivation. Conversion of the benzoquinone to
analogous phenylacetic acid compounds ablated BoNT/A LC inhibitory activity, further demonstrating the necessity of the BQ moiety for activity.
benzoquinones (Fig. 8), and given the ꢀ7 Å distance between Zn
and Cys165, these compounds would be the ideal length to access
both. Pyridyl acrylamides 40 and 41 weakly inhibited BoNT/A. 2-
(2-Carboxyethyl)-1,4-BQ (22) was fairly active against BoNT/A
and possessed good ligand efficiency (LE) while 2-(3-carboxypro-
pyl)-1,4-BQ (16) was inactive. However, in the 66-mer preincuba-
tion assay, 22 inhibited BoNT/A no more than related compound 12
(2-Me-BQ) with a weak kinact/KI of 2.5 Mꢁ1 sꢁ1 while 41 displayed
no irreversible inhibition. Overall, our simplified dual-action inhib-
itors still suffered from the same problem as 31 where their cova-
lent modification only slightly contributed to their inhibitory
activity. Despite this, 22 is a high LE fragment that could easily
be derivatized for development of a more potent competitive
inhibitor. In fact, we tested phenylacetic acid derivatives as ana-
logues of 22 and we found that they were completely inactive at
50 lM (Fig. 8). The lack of activity in the phenylacetic acid com-
pounds also suggests that the 1,4-benzoquinone moiety is critical
for achieving potency.
2.5. Kinetic studies
We investigated the discrepancy between inhibitory activity
and covalent modification by observing BQ inactivation via the
SNAPtide assay in the presence of other inhibitors. Competitive
Figure 9. Dose-dependent competition of zinc-chelating, active site inhibitors with BQ. (A) Impact of AdH on initial rates of BoNT/A LC. (B) Impact of AdH on kobs of BQ
inactivation. (C) Impact of 28 on initial rates of BoNT/A LC. (D) Impact of 28 on kobs of BQ inactivation. Active site inhibitors AdH and 28 compete with BQ covalent
inactivation.

P. T. Bremer et al. / Bioorg. Med. Chem. 22 (2014) 3971–3981
3977
inhibitor adamantanemethylhydroxamate (AdH) has
a Ki of
4. Experimental
460 nM, and crystallographic data shows AdH interacts with the
S10 pocket while chelating the zinc of BoNT/A.10 Since no overlap
exists between the binding site of this compound and Cys165, we
hypothesized that AdH and BQ could inhibit BoNT/A coopera-
tively or in a non-mutually exclusive manner. Kinetic assays
reveal that AdH actually competes with BQ-mediated inactivation
of BoNT/A (Fig. 9A and B). Furthermore, 28 which presumably has
the same binding mode as AdH, also competed with BQ-mediated
BoNT/A inhibition (Fig. 9C and D). A possible explanation for this
phenomenon is that AdH binding induces an enzyme conforma-
tion in which Cys165 is locked in a solvent inaccessible position.
Cys165 is also solvent inaccessible in the uninhibited BoNT/A
conformation,32 although a degree of enzyme flexibility must
exist that allows electrophiles like BQ and MTSPA to access
Cys165; BoNT/A is known to be a fairly flexible enzyme.33 Lastly,
we tested BQ in the presence of chicoric acid (CA), a known exo-
site inhibitor, and BQ inactivation was not affected by CA (data
not shown). In contrast to AdH, CA induces a catalytically inactive
conformation that must not render Cys165 inaccessible to BQ.
Kinetics studies support our hypothesis that linked inhibitors
30–32 are acting as a competitive inhibitors via the binding mode
predicted in Figure 4; binding of the lipophilic tail renders Cys165
inaccessible for covalent modification but does not disrupt the
favorable non-covalent binding of BQ adjacent to the active site
zinc. In light of these results, development of ‘enhancer’ com-
pounds that could shift Cys165 to a solvent accessible position
would be highly desirable in the context of BoNT/A irreversible
inhibitors.
4.1. Synthesis
4.1.1. General methods
NMR spectra were recorded on a Bruker 600 MHz spectrometer.
All chemical shifts are reported in ppm using the CDCl3 solvent
peak as a reference. All starting materials and reagents were
purchased from commercial sources and used without further
purification. All reactions were run under N2 gas and with dry,
distilled solvents unless otherwise noted. LCMS as well as TLC
visualized by UV light and/or ninhydrin staining were routinely
used to monitor reactions. Following aqueous workups, the organic
layer was always dried using MgSO4 and then filtered. Compounds
1–3, 7, 9, 12, 13, 15, 17, 23–25 were obtained from commercial
sources while compounds 4–6, 11, 20, 21, 26 were obtained from
the Baran lab.34 Compounds 10,25 14,35 16, 18,36 19,37 40/41,38
were synthesized as reported previously.
4.1.2. General procedure for amide couplings
To a 0.2 M solution of amine (0.1–10 mmol, 1 equiv) and car-
boxylic acid (1 equiv) in DCM was added Cl-HOBt (1.05 equiv)
and Et3N (1.3 equiv) followed by EDC-HCl (1.2 equiv). The mixture
was stirred at rt for 12 h. The crude mixture was diluted with DCM
and washed once with 1 M HCl, once with satd NaHCO3 and once
with brine. The DCM solution was evaporated to afford the product
with >90% purity. The product was recrystallized from Et2O/hexane
or purified by silica gel chromatography if necessary. Yields were
typically >75%.
3. Conclusion
4.1.3. 5,8-Dioxo-5,8-dihydronaphthalen-1-yl
cyclopentanecarboxylate (8)
We have discovered that 1,4-benzoquinones and naphthoqui-
nones are irreversible inhibitors of BoNT/A. Important factors that
enhance BQ activity are substitution at the 2 and 3 positions (such
as in NQ) and substituents that are electron withdrawing, for
example, chloro. Respectively, these factors allow steric accessibil-
ity of cysteine to the electrophilic portion of BQ (in contrast to 2,6
and 2,5 substitution) and enhance the thiol reactivity by increasing
LUMO energies. An increase in BoNT/A activity of BQs typically led
to a concurrent increase in non-specific thiol reactivity although
natural product NQ, naphthazarin, stood out as possessing low
thiol reactivity while retaining good BoNT/A activity. The require-
ment for high thiol reactivity to achieve potency highlights the
difficulty in developing drug-like irreversible inhibitors of BoNT/
A with low thiol reactivities on the level of ibrutinib. The buried
position of Cys165 is likely to blame for the need for ‘hot’
electrophiles in order to covalently modify Cys165. An alternative
strategy for covalent inhibitor design would be screening
for ‘enhancer’ compounds that would shift the enzyme to a
Cys165-exposed conformation for modification by mildly-reactive
electrophiles.
Despite the failure of our attempts to create more potent irre-
versible inhibitors via a fragment-based approach, the process
has led to the discovery of low micromolar IC50 competitive inhib-
itors (30–32); docking and SAR studies suggest that linking 1,4-
benzoquinone to zinc and S10 site pharmacophores was successful
for inhibitor design. Lastly, kinetic studies indicate that binding of
zinc-chelating, active site inhibitors and BQ covalent inhibition are
mutually exclusive, refuting the strategy of targeting the zinc and
S10 site for irreversible inhibitor development. However, the fact
that our rationally-designed compounds achieved potency stands
as a demonstration of how computational docking can identify
active compounds without the need for the synthesis and screen-
ing of large small-molecule libraries.
Similar to previously reported procedure,25 cyclopentoyl chlo-
ride (3 equiv, 29
lL) was added to a solution of 7 (0.08 mmol,
14 mg) and DMAP (0.2 equiv, 2 mg) in 200
lL pyridine/200
lL
DCM. After stirring for 45 min, the reaction was diluted with 1N
HCl and extracted with DCM. Purification by pTLC with 30% EtOAc
in hexane afforded 8 as a yellow solid (12 mg, 55%).
1H NMR (600 MHz, CDCl3) d 8.04 (dd, J = 7.7, 1.3 Hz, 1H), 7.75 (t,
J = 7.9 Hz, 1H), 7.37 (dd, J = 8.1, 1.3 Hz, 1H), 6.93 (d, J = 10.3 Hz, 1H),
6.84 (d, J = 10.3 Hz, 1H), 3.22–3.09 (m, 1H), 2.19–2.05 (m, 4H),
1.90–1.76 (m, 2H), 1.76–1.63 (m, 2H).
13C NMR (151 MHz, CDCl3) d 184.44, 183.78, 174.96, 149.94,
140.11, 137.37, 134.83, 133.68, 130.00, 125.00, 123.74, 44.12,
30.00, 26.01.
ESI-TOF-MS (m/z): [M+H]+ calcd 271.0965, obsd 271.0965.
4.1.4. 3-(3,6-Dioxocyclohexa-1,4-dien-1-yl)propanoic acid (16)
2,5-Dimethoxypropionic acid was oxidized via a previously
reported procedure employing oxone and 4-iodophenoxyacetic
acid to the benzoquinone 16 as an orange solid (12.9 mg, 55%) with
pTLC (70% EtOAc in hexane). Characterization agreed with a previ-
ous report of 16.39
4.1.5. 2-(3,6-Dioxocyclohexa-1,4-dien-1-yl)acetic acid (22)
2,5-Dimethoxyphenylacetic acid was oxidized via a previously
reported procedure40 employing oxone and 4-iodophenoxyacetic
acid to the benzoquinone 22 as an orange solid (16.5 mg, 75%)
without the need for a purification step.
1H NMR (600 MHz, MeOD) d 6.83 (d, J = 10.1 Hz, 1H), 6.80–6.78
(m, 1H), 6.76–6.75 (m, 1H), 3.47 (d, J = 1.2 Hz, 2H).
13C NMR (151 MHz, MeOD) d 188.90, 187.89, 173.10, 143.98,
137.68, 135.89, 35.55.
ESI-TOF-MS (m/z): [M+H]+ calcd 167.0339, obsd 167.0339.

3978
P. T. Bremer et al. / Bioorg. Med. Chem. 22 (2014) 3971–3981
4.1.6. 2-Amino-N-[3-(benzyloxy)phenyl]acetamide (27)
Compound 27 was prepared via the general amide coupling
procedure with 3-benzyloxyaniline and Boc-Gly-OH to afford the
Boc protected product as a white solid (3.2 g, 82%). Boc deprotec-
tion with 1:1 TFA/DCM over 30 min quantitatively produced 27
as a colorless oil.
13C NMR (151 MHz, CDCl3) d 159.32, 153.94, 151.82, 136.91,
129.96, 128.69, 128.11, 127.68, 118.30, 117.56, 116.55, 112.83,
112.67, 111.90, 111.83, 106.76, 70.04, 56.10, 55.88.
ESI-TOF-MS (m/z): [M+H]+ calcd 377.1496, obsd 377.1496.
4.1.10. 6-[(3,6-Dioxocyclohexa-1,4-dien-1-yl)amino]-N-(4-
phenoxybenzyl)picolinamide (31)
1H NMR (600 MHz, MeOD) d 7.43–7.40 (m, 2H), 7.38–7.34 (m,
3H), 7.31–7.28 (m, 1H), 7.22 (t, J = 8.2 Hz, 1H), 7.11–7.08 (m, 1H),
6.79–6.75 (m, 1H), 5.07 (s, 2H), 3.83 (s, 2H).
Compound 34 (12 mg, 0.026 mmol) was dissolved in 30
lL
MeOH, 30 L MeCN and 400 L H2O and cooled to 0 °C. PhI(OAc)2
l
l
13C NMR (151 MHz, MeOD) d 165.43, 160.67, 140.27, 138.53,
130.77, 129.49, 128.89, 128.53, 113.36, 111.98, 107.89, 70.98,
42.14.
(1.2 equiv, 10.5 mg) was added gradually and the mixture was stir-
red for 1.5 h, allowing to warm to rt. The reaction mixture was
diluted with satd NaHCO3 and extracted with EtOAc. Purification
by pTLC with 70% EtOAc in hexane afforded 31 as a red solid
(1.6 mg, 14%). (Adapted from a previously reported method)
(Fig. 11).42
ESI-TOF-MS (m/z): [M+H]+ calcd 257.1284, obsd 257.1284.
4.1.7. 2-Amino-N-(4-phenoxyphenyl)acetamide (28)
Compound 28 was prepared via the general amide coupling
procedure with 4-phenoxyaniline and Boc-Gly-OH to afford the
Boc protected product as a white solid (3.3 g, 89%) after recrystal-
lization. Boc deprotection with 1:1 TFA/DCM over 30 min quantita-
tively produced 28 as a white solid.
1H NMR (600 MHz, CDCl3) d 7.95 (dd, J = 7.6, 0.9 Hz, 1H), 7.90 (d,
J = 6.3 Hz, 1H), 7.86 (t, J = 8.1 Hz, 1H), 7.78 (br s, 1H), 7.39–7.36 (m,
2H), 7.33–7.30 (m, 2H), 7.19 (d, J = 2.4 Hz, 1H), 7.13 (dd, J = 8.1,
0.8 Hz, 1H), 7.09 (tt, J = 7.5, 1.1 Hz, 1H), 7.04–7.00 (m, 4H), 6.80
(d, J = 10.1 Hz, 1H), 6.75 (dd, J = 10.1, 2.4 Hz, 1H), 4.67 (d,
J = 6.0 Hz, 2H).
1H NMR (600 MHz, MeOD) d 7.58–7.55 (m, 2H), 7.35–7.30 (m,
2H), 7.08 (tt, J = 7.4, 1.1 Hz, 1H), 6.98–6.93 (m, 4H), 3.85 (s, 2H).
13C NMR (151 MHz, MeOD) d 165.32, 158.89, 155.16, 134.66,
130.88, 124.28, 122.66, 120.36, 119.49, 42.04.
13C NMR (151 MHz, CDCl3) d 187.17, 183.31, 163.80, 157.37,
156.80, 151.19, 148.99, 139.81, 139.76, 139.09, 132.98, 132.94,
129.86, 129.42, 123.35, 119.47, 118.96, 117.67, 116.54, 107.97,
43.28.
ESI-TOF-MS (m/z): [M+H]+ calcd 243.1128, obsd 243.1128.
ESI-TOF-MS (m/z): [M+H]+ calcd 426.1448, obsd 426.1442.
4.1.8. N-([1,10-Biphenyl]-4-ylmethyl)-2-aminoacetamide (29)
Compound 29 was prepared via the general amide coupling
procedure with 4-phenylbenzylamine and Boc-Gly-OH to afford
the Boc protected product as a white solid (2.6 g, 83%) after
recrystallization. Boc deprotection with 1:1 TFA/DCM over
30 min quantitatively produced 29 as a white solid.
4.1.11. N-([1,10-Biphenyl]-4-ylmethyl)-1-[(3,6-dioxocyclohexa-
1,4-dien-1-yl)methyl]-1H-1,2,4-triazole-3-carboxamide (32)
See preparation of 30 for details.
Compound 38 (7.2 mg, 0.017 mmol) was oxidized with 2 equiv
CAN over 8 h to produce 32 as a brown solid (0.6 mg, 9%) (Fig. 12).
Chemical instability precluded the acquisition of clean NMR
spectra.
1H NMR (600 MHz, MeOD) d 7.59–7.56 (m, 3H), 7.44–7.37 (m,
5H), 7.34–7.30 (m, 1H), 4.47 (s, 2H), 3.73 (s, 2H).
ESI-TOF-MS (m/z): [M+H]+ calcd 399.1452, obsd 399.1455.
13C NMR (151 MHz, MeOD) d 165.76, 140.57, 140.31, 137.06,
128.49, 127.84, 127.00, 126.77, 126.48, 42.57, 40.13.
4.1.12. N-[3-(Benzyloxy)phenyl]-2-[(2,5-dimethoxybenzyl)
amino]acetamide (33)
ESI-TOF-MS (m/z): [M+H]+ calcd 241.1335, obsd 241.1335.
3-Benzyloxyaniline (80 mg, 0.40 mmol) was dissolved in DCM
and cooled to 0 °C. Bromoacetyl bromide (2 equiv, 70
Et3N (5 equiv, 280 L) were added slowly and the mixture was stir-
red at 0 °C for 6 h. 2,5-Dimethoxybenzylamine (2.5 equiv, 151 L)
lL) and
4.1.9. N-[3-(Benzyloxy)phenyl]-2-{[(3,6-dioxocyclohexa-1,4-
dien-1-yl)methyl]amino}acetamide (30)
l
l
N-Boc-33 (14 mg, 0.027 mmol) was dissolved in 0.4 mL CHCl3,
0.8 mL MeCN and 0.2 mL H2O and cooled to 0 °C. CAN (2 equiv,
27 mg) was gradually added, and the mixture was stirred for
2.5 h.41 The mixture was diluted with water and extracted with
DCM. The crude product was purified by pTLC with 40% EtOAc in
hexane to afford the Boc protected product (4.7 mg, 36%). The
Boc group was removed by stirring with 1:1 TFA/DCM over
30 min to quantitatively produce 30 as a yellow oil (Fig. 10).
1H NMR (600 MHz, CDCl3) d 7.42–7.33 (m, 4H), 7.31 (d,
J = 6.8 Hz, 2H), 7.15 (t, J = 8.1 Hz, 1H), 7.05 (d, J = 7.6 Hz, 1H), 6.90
(dd, J = 9.0, 2.9 Hz, 1H), 6.84 (d, J = 9.2 Hz, 1H), 6.80 (d, J = 3.0 Hz,
1H), 6.76–6.66 (m, 1H), 4.99 (s, 2H), 3.87 (s, 2H), 3.72 (s, 2H).
was added at 0 °C and the mixture was allowed to warm over
16 h of stirring. Boc2O (6 equiv, 524 mg) was added along with 2
addition equiv of Et3N and DMAP (0.2 equiv, 10 mg) and the mix-
ture was stirred for 1 h at rt. The reaction mixture was diluted with
satd NaHCO3 and extracted with DCM. The crude product was puri-
fied by pTLC with 30% EtOAc in hexane to afford N-Boc-33 as a col-
orless oil (49 mg, 24%).
1H NMR w/Boc (600 MHz, CDCl3) d 7.45–7.41 (m, 2H), 7.40–7.36
(m, 2H), 7.34–7.30 (m, 1H), 7.16 (t, J = 8.1 Hz, 1H), 6.88–6.73 (m,
3H), 6.70 (dd, J = 8.2, 2.2 Hz, 1H), 5.04 (s, 2H), 4.55 (s, 2H), 3.98
(s, 2H), 3.75 (s, 3H), 3.74 (s, 3H), 1.45 (s, 9H).
O
O
(b, c)
(a)
(d, e)
33
30
N-Boc-
O
Br
3 steps, 1 pot
N
NH2
H
Figure 10. Synthesis of 30. (a) Bromoacetyl bromide, Et3N, DCM, 0 °C; (b) 2,5-dimethoxybenzylamine, 0 °C to rt; (c) Boc2O, Et3N, cat. DMAP, DCM, 24% (3 steps); (d) CAN,
MeCN/H2O; (e) TFA, DCM, 36% (2 steps).

P. T. Bremer et al. / Bioorg. Med. Chem. 22 (2014) 3971–3981
3979
O
O
O
O
O
H
H
N
N
N
Br
R
N
N
R
N
(d)
(c)
(a, b)
N
H
O
O
42
31 R
,
=NHMePh-4-OPh
34 R
,
=NHMePh-4-OPh
36, R=MeOOC
37, R=H
43, R=MeOOC
44, R=H
Figure 11. Synthesis of 31 and related compounds. (a) H2, Pd/C, 88%; (b) 6-bromopicolinic acid, EDC, Cl-HOBt, Et3N, DCM 76%; (c) 2,5-dimethoxyaniline, BINAP, Pd(OAc)2,
Cs2CO3, toluene, 40–50%; (d) PhI(AcO)2, MeOH/H2O, 14–26%.
4.1.15. Methyl 6-[(3,6-dioxocyclohexa-1,4-dien-1-
yl)amino]picolinate (36)
O
(a)
(b)
(c)
N
HO
38
32
39
NH
Compound 43 (Methyl 6-[(2,5-dimethoxyphenyl)amino]picoli-
nate) was prepared via the same aryl amination reaction used to
prepare 31. Methyl-6-bromopyridine-2-carboxylate was reacted
with 2,5-dimethoxyaniline to afford 43 as a colorless oil (41 mg,
44%) which was subsequently used in the next reaction.
1H NMR (600 MHz, CDCl3) d 8.11 (d, J = 3.0 Hz, 1H), 7.67–7.55
(m, 2H), 7.03 (dd, J = 7.8, 1.4 Hz, 1H), 6.81 (d, J = 8.8 Hz, 1H), 6.51
(dd, J = 8.8, 3.0 Hz, 1H), 3.97 (s, 3H), 3.85 (s, 3H), 3.84 (s, 3H).
Compound 43 was oxidized with PhI(OAc)2 to afford 36 as a red
solid (2.0 mg, 22%). (See preparation of 31 for full conditions)
1H NMR (600 MHz, CDCl3) d 7.82–7.78 (m, 3H), 7.12 (dd, J = 7.0,
2.0 Hz, 1H), 6.79–6.74 (m, 2H), 4.02 (s, 3H).
N
Figure 12. Synthesis of 32. (a) 4-Phenylbenzylamine, EDC, Cl-HOBt, Et3N, DMF,
68%; (b) 2,5-dimethoxybenzyl bromide, K2CO3, DMF, 81%; (c) CAN, 9%.
13C NMR w/Boc (151 MHz, CDCl3) d 174.41, 168.14, 159.25,
151.70, 136.91, 129.61, 128.55, 127.94, 127.49, 126.40, 111.99,
110.98, 106.13, 81.32, 69.94, 55.72, 52.90, 28.31, 20.49.
ESI-TOF-MS (m/z) w/Boc: [M+H]+ calcd 507.2490, obsd
507.2493.
ESI-TOF-MS (m/z) no Boc: [M+H]+ calcd 407.1965, obsd
407.1966.
13C NMR (151 MHz, CDCl3) d 187.89, 183.62, 165.51, 152.52,
146.70, 139.78, 139.14, 138.97, 132.87, 119.67, 116.95, 109.54,
53.16.
4.1.13. 6-[(2,5-Dimethoxyphenyl)amino]-N-(4-phenoxybenzyl)
picolinamide (34)
ESI-TOF-MS (m/z): [M+H]+ calcd 259.0713, obsd 259.0714.
4-Phenoxybenzonitrile (200 mg, 1.00 mmol) was dissolved in
8 mL MeOH and 5% Pd/C (0.2 equiv, 400 mg) was added. The reac-
tion was stirred for 3 h under an H2 atmosphere and filtered
through celite. Purification by pTLC with 10% MeOH, 2% Et3N in
DCM afforded 4-phenoxybenzylamine as a white solid (179 mg,
88%). 42 (1.1 equiv, 22 mg) was dissolved in toluene with 2,5-
dimethoxyaniline (1 equiv, 8 mg), BINAP (0.3 equiv, 11 mg),
Pd(OAc)2 (0.15 equiv, 2 mg) and Cs2CO3 (2.5 equiv, 46 mg) and
the mixture was stirred for 15 h at 105 °C. The solvent was
removed and the crude product was filtered through silica, eluting
with EtOAc. Final purification by pTLC with 50% EtOAc in hexane
afforded 34 as colorless oil (12.4 mg, 48%).
4.1.16. 2-(Pyridin-2-ylamino)cyclohexa-2,5-diene-1,4-dione
(37)
See preparation of 31 for details.
Compound 44 was oxidized with PhI(OAc)2 to afford 37 as a red
solid (3.2 mg, 26%).
1H NMR (600 MHz, CDCl3) d 8.41–8.35 (m, 1H), 7.73 (d,
J = 2.1 Hz, 1H), 7.65 (ddd, J = 8.3, 7.4, 1.9 Hz, 1H), 6.97 (ddd,
J = 7.3, 5.0, 0.9 Hz, 1H), 6.95–6.90 (m, 1H), 6.78–6.70 (m, 2H).
13C NMR (151 MHz, CDCl3) d 188.02, 183.84, 152.84, 148.45,
140.09, 139.19, 137.99, 132.82, 118.34, 113.59, 108.50.
ESI-TOF-MS (m/z): [M+H]+ calcd 201.0659, obsd 201.0659.
1H NMR (600 MHz, CDCl3) d 8.18 (t, J = 6.1 Hz, 1H), 7.71 (dd,
J = 7.4, 1.0 Hz, 1H), 7.70–7.64 (m, 1H), 7.58 (d, J = 3.0 Hz, 1H),
7.38–7.29 (m, 4H), 7.13–7.07 (m, 1H), 7.03–6.95 (m, 6H), 6.82 (d,
J = 8.8 Hz, 1H), 6.48 (dd, J = 8.8, 3.0 Hz, 1H), 4.64 (d, J = 6.1 Hz,
2H), 3.86 (s, 3H), 3.64 (s, 3H).
4.1.17. N-([1,10-Biphenyl]-4-ylmethyl)-1-(2,5-
dimethoxybenzyl)-1H-1,2,4-triazole-3-carboxamide (38)
Compound 39 (11 mg, 0.040 mmol) was dissolved in 0.75 mL
DMF and K2CO3 (2 equiv, 11 mg) and 2,5-dimethoxybenzyl bro-
mide (11 mg, 1.2 equiv) were added (treatment of 2,5-dimethoxy-
benzyl alcohol with PBr3 afforded the corresponding benzyl
bromide).43 The reaction mixture was stirred at rt. for 2.5 h and fil-
tered. The crude product was purified by pTLC with 5% MeOH in
DCM to afford 38 as a colorless oil (13.8 mg, 81%).
13C NMR (151 MHz, CDCl3) d 164.67, 157.35, 156.70, 154.00,
153.88, 148.16, 143.13, 138.88, 133.33, 130.46, 129.88, 129.48,
123.38, 119.23, 118.95, 114.25, 113.48, 111.16, 105.76, 105.35,
56.37, 55.65, 43.03.
ESI-TOF-MS (m/z): [M+H]+ calcd 456.1918, obsd 456.1919.
1H NMR (600 MHz, CDCl3) d 8.05 (s, 1H), 7.58–7.53 (m, 4H),
7.48–7.38 (m, 4H), 7.37–7.32 (m, 1H), 6.89–6.78 (m, 3H), 5.34 (s,
2H), 4.69 (d, J = 6.0 Hz, 2H), 3.80 (s, 3H), 3.75 (s, 3H).
4.1.14. N-(4-phenoxybenzyl)picolinamide (35)
Compound 35 was prepared via the general amide coupling
procedure in 1:1 DCM to DMF with 4-phenoxybenzylamine and
picolinic acid to afford the product as a white solid (18 mg, 84%).
1H NMR (600 MHz, CDCl3) d 8.53 (ddd, J = 4.7, 1.7, 0.9 Hz, 1H),
8.37 (br s, 1H), 8.24 (dt, J = 7.8, 1.1 Hz, 1H), 7.86 (td, J = 7.7,
1.7 Hz, 1H), 7.43 (ddd, J = 7.6, 4.7, 1.3 Hz, 1H), 7.37–7.29 (m, 4H),
7.09 (tt, J = 7.5, 1.1 Hz, 1H), 7.03–6.95 (m, 4H), 4.65 (d, J = 6.1 Hz,
2H).
13C NMR (151 MHz, CDCl3) d 159.19, 156.88, 153.75, 151.50,
144.11, 140.85, 140.67, 137.03, 128.89, 128.54, 127.55, 127.43,
127.19, 122.96, 116.92, 115.26, 111.79, 55.89, 55.88, 49.77, 43.15.
ESI-TOF-MS (m/z): [M+H]+ calcd 429.1921, obsd429.1923.
4.1.18. N-([1,10-Biphenyl]-4-ylmethyl)-1H-1,2,4-triazole-3-
carboxamide (39)
Compound 39 was prepared via the general amide coupling
procedure in DMF with 4-phenylbenzylamine and 1,2,4-triazole-
3-carboxylic acid to afford the product which precipitated out of
the reaction as a white solid (48 mg, 68%).
13C NMR (151 MHz, CDCl3) d 164.35, 157.31, 156.75, 149.93,
148.22, 137.53, 133.21, 129.86, 129.48, 126.37, 123.40, 122.50,
119.21, 118.96, 43.06.
ESI-TOF-MS (m/z): [M+H]+ calcd 305.1284, obsd 305.1287.

3980
P. T. Bremer et al. / Bioorg. Med. Chem. 22 (2014) 3971–3981
1H NMR (600 MHz, DMSO-d6) d 9.85 (s, 1H), 8.46–8.43 (m, 2H),
4.2.3. 66-mer assay31
8.43–8.40 (m, 2H), 8.28–8.24 (m, 2H), 8.21 (d, J = 8.2 Hz, 2H), 8.17–
8.14 (m, 1H), 5.28 (d, J = 6.4 Hz, 2H).
All 66-mer assays were run in 40 mM HEPES pH 7.4. BoNT/A LC
concentrations were 0.8 nM, 66-mer substrate (prepared in-house
13C NMR (151 MHz, DMSO-d6) d 140.01, 138.84, 138.69, 128.90,
127.91, 127.30, 126.57, 41.74.
via solid phase synthesis) concentrations were 5
lM. Inhibitors
were tested at 50 M and substrate cleavage was allowed to occur
l
ESI-TOF-MS (m/z): [M+H]+ calcd 279.1240, obsd 279.1240.
for 25 min at which point the reaction was quenched with 20% TFA
solution. A 13C labeled 9-mer cleavage product was added as an
internal standard (IS) and each sample was analyzed by LCMS to
quantify the amount of cleavage product relative to the IS. Initial
velocities were used to calculate IC50 values from Eq. (4) (a rear-
ranged form of Eq. (2)). In the preincubation experiments, inhibi-
4.1.19. 6-Bromo-N-(4-phenoxybenzyl)picolinamide (42)
Compound 42 was prepared via the general amide coupling
procedure with 4-phenoxybenzylamine and 6-bromopicolinic acid
to afford the product as a colorless oil (73 mg, 76%).
1H NMR (600 MHz, CDCl3) d 8.21 (dd, J = 7.6, 1.0 Hz, 1H), 8.18 (s,
1H), 7.74 (t, J = 7.8 Hz, 1H), 7.63 (dd, J = 7.9, 1.0 Hz, 1H), 7.37–7.32
(m, 4H), 7.12 (tt, J = 7.4, 1.1 Hz, 1H), 7.06–6.97 (m, 4H), 4.65 (d,
J = 6.1 Hz, 2H).
tors at 50 lM of inhibitor was incubated with 40 nM enzyme and
diluted 50-fold into substrate and initial rates of 66-mer cleavage
were measured. Redetermination of competitive inhibitor IC50 val-
ues with a different batch of 66-mer substrate yielded up to a ~5
fold increase in relative IC50 values, however this result does not
affect the SAR or conclusions herein.
13C NMR (151 MHz, CDCl3) d 162.85, 157.20, 156.84, 151.01,
140.68, 139.81, 132.83, 130.91, 129.86, 129.51, 123.43, 121.56,
119.14, 118.99, 43.10.
V
½Iꢃ V
0
ESI-TOF-MS (m/z): [M+H]+ calcd 383.0390, obsd 383.0391.
IC50
¼
ð4Þ
V
0
1 ꢁ V
4.1.20. N-(2,5-Dimethoxyphenyl)pyridin-2-amine (44)
Compound 44 was prepared via the same aryl amination reac-
tion used to prepare 31. 2-Bromopyridine was reacted with 2,5-
dimethoxyaniline to afford 44 as a colorless oil (20 mg, 40%).
1H NMR (600 MHz, CDCl3) d 8.27–8.23 (m, 1H), 7.87 (d,
J = 3.0 Hz, 1H), 7.53–7.47 (m, 1H), 7.02 (s, 1H), 6.86 (d, J = 8.4 Hz,
1H), 6.80 (d, J = 8.8 Hz, 1H), 6.76–6.72 (m, 1H), 6.45 (dd, J = 8.8,
3.0 Hz, 1H), 3.85 (s, 3H), 3.80 (s, 3H).
4.2.4. Regression analysis
All curve-fitting was performed in GraphPad PRISM version 6
using Eqs. 1–4 and standard linear and exponential regressions.
4.2.5. Computational studies
A previously used computational model and scoring function10
derived from the Autocorrelator program46 was used for docking
studies. OMEGA v2.4.6 was used to generate lowest energy con-
formers of query molecules and FRED was used to dock these con-
formers into 2IMB and 4ELC co-crystal structure of BoNT/A LC.47
13C NMR (151 MHz, CDCl3) d 155.42, 154.10, 148.25, 142.87,
137.49, 131.38, 115.21, 110.92, 110.13, 105.15, 104.72, 56.40,
55.83.
ESI-TOF-MS (m/z): [M+H]+ calcd 231.1128, obsd 231.1130.
Acknowledgments
4.2. Enzyme assays
The authors would like to thank the NIH for funding under
grants 5R01AI080671-05 and 1F31DA037709-01, Dr. Peter Silhar
for proposing juglone as an inhibitor, Professor Phil Baran and
his research group for providing the noted BQ compounds,
Dr. Matthew Lardy for his assistance with docking studies and Greg
McElhaney and Jing Yu for synthesis of aminoacetamides.
4.2.1. SNAPtide assay44
Recombinant 425aa BoNT/A LC was used throughout all assays
and was prepared as previously described.45 All SNAPtide assays
were run in 40 mM HEPES + 0.1% Triton X-100 at pH 7.4. BoNT/A
LC concentrations were 18.5 nM while SNAPtide (prod. no. 521, List
Labs) concentrations were
5 lM. Fluorescence was recorded
continuously for 1.5 h to calculate an accurate kobs using Eq. (1). In
the preincubation experiments, inhibitors at the specified concen-
trations were incubated with 925 nM enzyme and diluted 50-fold
into substrate and initial rates of SNAPtide cleavage were measured.
References and notes
1. Arnon, S. S.; Schechter, R.; Inglesby, T. V., et al. JAMA 2001, 285, 1059.
2. Schantz, E. J.; Johnson, E. A. Microbiol. Rev. 1992, 56, 80.
3. Jankovic, J. J. Neurol. Neurosurg. Psychiatry 2004, 75, 951.
Vcontrol
kobs
obst
ð1 ꢁ eꢁk Þ þ b
ð1Þ
4. D. Hanson, Botulinum Toxin:
A Bioterrorism Weapon, 2004. http://
RFU ¼
www.emsworld.com/article/10324792/botulinum-toxin-a-bioterrorism-
weapon (Accessed March 2014).
5. Willis, B.; Eubanks, L. M.; Dickerson, T. J.; Janda, K. D. ACIE 2008, 47, 8360.
6. Simpson, L. I. JPET 1989, 251, 1223.
7. Blasi, J.; Chapman, E. R.; Link, E.; Binz, T.; Yamasaki, S.; De Camilli, P.; Sudhof, T.
C.; Niemann, H.; Jahn, R. Nature 1993, 365, 160.
8. Boldt, G. E.; Kennedy, J. P.; Janda, K. D. Org. Lett. 2006, 8, 1729.
9. Kumaran, D.; Rawat, R.; Ludivico, M. L.; Ahmed, S. A.; Swaminathan, S. JBC
2008, 283, 18883.
10. Silhar, P.; Silvaggi, N. R.; Pellett, S.; Capkova, K.; Johnson, E. A.; Allen, K. N.;
Janda, K. D. BMC 2013, 21, 1344.
11. Silhar, P.; Capkova, K.; Salzameda, N. T.; Barbieri, J. T.; Hixon, M. S.; Janda, K. D.
JACS 2010, 132, 2868.
12. Li, B.; Cardinale, S. C.; Butler, M. M.; Pai, R.; Nuss, J. E.; Peet, N. P.; Bavari, S.;
Bowlin, T. L. BMC 2011, 19, 7338.
13. Capkova, K.; Hixon, M. S.; Pellett, S.; Barbieri, J. T.; Johnson, E. A.; Janda, K. D.
BMCL 2010, 20, 206.
IC50 values for Figure 9 were calculated using Eq. (2).
Vcontrol
V ¼
½Iꢃ
ð2Þ
1 þ IC
50
4.2.2. Glutathione reactivity assay
1 mM glutathione and 1 mM BQ compound were incubated in
PBS (Fisher Bioreagents) + 1 mM EDTA buffer at pH 7.4. At certain
time points, the reaction solution was diluted 1:10 into a 4 mg/
mL DTNB solution and the absorbance was measured at 412 nm.
Appropriate blanks were used to account for potential absorbance
by the BQ compounds. A standard curve was run alongside the
assay using [GSH] = 0, 0.25, 0.5, 0.75, 1, 1.25 mM. Eq. (3) was used
to calculate the second order rate constant (kGSH) of Michael adduct
formation.
14. Stura, E. A.; Le Roux, L.; Guitot, K.; Garcia, S.; Bregant, S.; Beau, F.; Vera, L.;
Collet, G.; Ptchelkine, D.; Bakirci, H.; Dive, V. JBC 2012, 287, 33607.
15. Foran, P. G.; Mohammed, N.; Lisk, G. O.; Nagwaney, S.; Lawrence, G. W.;
Johnson, E.; Smith, L.; Aoki, K. R.; Dolly, J. O. JBC 2003, 278, 1363.
16. Papageorgiou, V. P.; Assimopoulou, A. N.; Couladouros, E. A.; Hepworth, D.;
Nicolaou, K. C. ACIE 1999, 38, 270.
1
17. Dandawate, P. R.; Vyas, A. C.; Padhye, S. B.; Singh, M. W.; Baruah, J. B. Mini-Rev.
Med. Chem. 2010, 10, 436.
ꢁ 1 ¼ 2kGSH
t
ð3Þ
½GSHꢃ

P. T. Bremer et al. / Bioorg. Med. Chem. 22 (2014) 3971–3981
32. Breidenbach, M. A.; Brunger, A. T. Nature 2004, 432, 925.
3981
18. Kleyer, D. L.; Koch, T. H. JACS 1984, 106, 2380.
19. Ollinger, K.; Brunmark, A. JBC 1991, 266, 21496.
33. Silhar, P.; Lardy, M. A.; Hixon, M. S.; Shoemaker, C. B.; Barbieri, J. T.; Struss, A.
K.; Lively, J. M.; Javor, S.; Janda, K. D. ACS Med. Chem. Lett. 2013, 4, 283.
34. Fujiwara, Y.; Domingo, V.; Seiple, I. B.; Gianatassio, R.; Del Bel, M.; Baran, P. S.
JACS 2011, 133, 3292.
35. Brimble, M. A.; Brenstrum, T. J. J. Chem. Soc., Perkin. Trans. 1 2001, 1624.
36. Contant, P.; Haess, M.; Riegl, J.; Scalone, M.; Visnick, M. Synthesis 1999,
821.
20. Littarru, G. P.; Tiano, L. Mol. Biotechnol. 2007, 37, 31.
21. Stocker, R.; Bowry, V. W.; Frei, B. PNAS 1991, 88, 1646.
22. Wissner, A.; Floyd, M. B.; Johnson, B. D.; Fraser, H.; Ingalls, C.; Nittoli, T.;
Dushin, R. G.; Discafani, C.; Nilakantan, R.; Marini, J.; Ravi, M.; Cheung, K.; Tan,
X.; Musto, S.; Annable, T.; Siegel, M. M.; Loganzo, F. J. Med. Chem. 2005, 48,
7560.
23. Wissner, A.; Fraser, H. L.; Ingalls, C. L.; Dushin, R. G.; Floyd, M. B.; Cheung, K.;
Nittoli, T.; Ravi, M. R.; Tan, X. Z.; Loganzo, F. BMC 2007, 15, 3635.
24. Boldt, G. E.; Kennedy, J. P.; Hixon, M. S.; McAllister, L. A.; Barbieri, J. T.; Tzipori,
S.; Janda, K. D. J. Comb. Chem. 2006, 8, 513.
25. Maruo, S.; Kuriyama, I.; Kuramochi, K.; Tsubaki, K.; Yoshida, H.; Mizushina, Y.
BMC 2011, 19, 5803.
26. Wissner, A.; Overbeek, E.; Reich, M. F.; Floyd, M. B.; Johnson, B. D.; Mamuya, N.;
Rosfjord, E. C.; Discafani, C.; Davis, R.; Shi, X.; Rabindran, S. K.; Gruber, B. C.; Ye,
F.; Hallett, W. A.; Nilakantan, R.; Shen, R.; Wang, Y. F.; Greenberger, L. M.; Tsou,
H. R. J. Med. Chem. 2003, 46, 49.
37. Ameer, F.; Green, I. R.; Krohn, K.; Sitoza, M. Synth. Commun. 2007, 37, 3041.
38. Eriksson, J.; Aberg, O.; Langstrom, B. Eur. J. Org. Chem. 2007, 455.
39. Eipert, M.; Maichle-Mossmer, C.; Maier, M. E. J. Org. Chem. 2002, 67, 8692.
40. Yakura, T.; Omoto, M.; Yamauchi, Y.; Tian, Y.; Ozono, A. Tetrahedron 2010, 66,
5833.
41. Jacob, P., 3rd; Callery, P. S.; Shulgin, A. T.; Castagnoli, N., Jr. J. Org. Chem. 1976,
41, 3627.
42. Nawrat, C. C.; Lewis, W.; Moody, C. J. J. Org. Chem. 2011, 76, 7872.
43. Ma, Y. J.; Zhang, Z. B.; Ji, X. F.; Han, C. Y.; He, J. M.; Abliz, Z.; Chen, W. X.; Huang,
F. H. Eur. J. Org. Chem. 2011, 5331.
27. Gan, J. P.; Harper, T. W.; Hsueh, M. M.; Qu, Q. L.; Humphreys, W. G. Chem. Res.
Toxicol. 2005, 18, 896.
44. Boldt, G. E.; Kennedy, J. P.; Hixon, M. S.; McAllister, L. A.; Barbieri, J. T.; Tzipori,
S.; Janda, K. D. J. Comb. Chem. 2006, 8, 513.
28. Dasmahapatra, G.; Patel, H.; Dent, P.; Fisher, R. I.; Friedberg, J.; Grant, S. Br. J.
Haematol. 2013, 161, 43.
45. Eubanks, L. M.; Hixon, M. S.; Jin, W.; Hong, S.; Clancy, C. M.; Tepp, W. H.;
Baldwin, M. R.; Malizio, C. J.; Goodnough, M. C.; Barbieri, J. T.; Johnson, E. A.;
Boger, D. L.; Dickerson, T. J.; Janda, K. D. PNAS 2007, 104, 2602.
46. Lardy, M. A.; LeBrun, L.; Bullard, D.; Kissinger, C.; Gobbi, A. J. Chem. Inf. Model.
2012, 52, 1328.
47. The OEChem Toolkit, OMEGA, ROCS, FRED, and SZYBKI are distributed by
OpenEye Scientific Software, Santa Fe, NM: http://www.eyesopen.com
(Accessed May 2012).
29. Honigberg, L. A.; Smith, A. M.; Sirisawad, M.; Verner, E.; Loury, D.; Chang, B.; Li,
S.; Pan, Z.; Thamm, D. H.; Miller, R. A.; Buggy, J. J. PNAS 2010, 107, 13075.
30. Ponader, S.; Chen, S. S.; Buggy, J. J.; Balakrishnan, K.; Gandhi, V.; Wierda, W. G.;
Keating, M. J.; O’Brien, S.; Chiorazzi, N.; Burger, J. A. Blood 2012, 119, 1182.
31. Capkova, K.; Hixon, M. S.; McAllister, L. A.; Janda, K. D. Chem. Commun 2008,
3525.