Development of Potent Inhibitors of Botulinum Neurotoxin Type B

Article abstract of DOI:10.1021/jm0300680

Botulinum neurotoxins are the most potent toxins known to date. They are zinc-metalloproteases able to cleave selectively an essential component of neurotransmitter exocytosis, causing the syndrome of botulism characterized by a flaccid paralysis. There is a great interest in designing antagonists of the action of these toxins. One way is to inhibit their catalytic activity. In this study, we report the design of such inhibitors directed toward BoNT/B. A study of the S₁ subsite specificity, using several β-amino thiols, has shown that this subsite prefers a p-carboxybenzyl moiety. The specificity of the S₁′ and S₂′ subsites was studied using two libraries of pseudotripeptides containing the S₁ synthon derived from the best β-amino thiol tested. Finally, a selection of various non natural amino acids for the recognition of the "prime" domain led to the most potent inhibitor of BoNT/B described to date with a K_i value of 20 nM.

Full text of DOI:10.1021/jm0300680

4648
J . Med. Chem. 2003, 46, 4648-4656
Develop m en t of P oten t In h ibitor s of Botu lin u m Neu r otoxin Typ e B
Christine Anne,† Serge Turcaud,† J ean Quancard, Franck Teffo, Herve´ Meudal,
Marie-Claude Fournie´-Zaluski, and Bernard P. Roques*
De´partement de Pharmacochimie Mole´culaire et Structurale, INSERM U266 - CNRS FRE2463,
UFR des Sciences Pharmaceutiques et Biologiques 4, Avenue de l’Observatoire - 75270 Paris Cedex 06, France
Received February 11, 2003
Botulinum neurotoxins are the most potent toxins known to date. They are zinc-metallopro-
teases able to cleave selectively an essential component of neurotransmitter exocytosis, causing
the syndrome of botulism characterized by a flaccid paralysis. There is a great interest in
designing antagonists of the action of these toxins. One way is to inhibit their catalytic activity.
In this study, we report the design of such inhibitors directed toward BoNT/B. A study of the
S1 subsite specificity, using several â-amino thiols, has shown that this subsite prefers a
p-carboxybenzyl moiety. The specificity of the S1′ and S2′ subsites was studied using two libraries
of pseudotripeptides containing the S1 synthon derived from the best â-amino thiol tested.
Finally, a selection of various non natural amino acids for the recognition of the “prime” domain
led to the most potent inhibitor of BoNT/B described to date with a Ki value of 20 nM.
In tr od u ction
drawbacks have been frequently reported following the
use of antitoxin.12
The seven serotypes of botulinum neurotoxins (BoNTs)
designated A to G are 150 kDa-dichain proteins com-
posed of a 100 kDa-heavy chain (HC) linked by a disul-
fide bridge to a 50 kDa-light chain (LC) and are pro-
duced by anaerobic bacteria, mainly Clostridium botu-
linum.1 They are the most lethal substances known to
mankind.2 The toxins induce a flaccid paralysis, which
can lead to death by respiratory arrest,3 through inhibi-
tion of acetylcholine release at the neuromuscular junc-
tion.4 The HC is responsible for cell-surface binding and
internalization of the toxin whereas the LC posseses a
zinc-metallopeptidase activity4 and was shown to cleave
one of the three proteins involved in exocytosis.5,6 These
toxins belong to the same family (clostridial neurotoxins)
as tetanus toxin, which has a quite similar mode of
action.5,7 All these enzymes contain the consensus
sequence HExxH found in thermolysin8 and share a
closely related spatial arrangement of the active site.9,10
The BoNTs cause food-borne, wound, and infant
botulism, which is a rare but severe disease. Moreover
they could be used as biological warfare agents because
of their easiness of production and their very high
toxicity.3,11 On the other hand, for more than 10 years,
there has been a medical use of botulinum neurotoxin,
essentially BoNT/A and BoNT/B, for the treatment of
various diseases such as hemifacial spasms, cervical
dystonia and axilary hyperhydrosis.12 The well-demon-
strated efficacy of the neurotoxins toward these diseases
has triggered a dramatic increase in their use to
diminish facial lines and to reduce myofacial pain.
However, severe drawbacks have been reported follow-
ing local injection of botulinum toxin and even fatal
issues.13 An antitoxin exists, but its use during the first
hours of infection results only in an improvement of the
successful intensive care treatment and of a decrease
in the time of hospital stay. Moreover, immunological
Similarly to tetanus toxin, BoNT/B cleaves synapto-
brevin, an integral membrane protein of synaptic vesicles,
which was shown to be essential for the formation of a
ternary complex with two other proteins of the plasma
membrane (SNAP-25 and syntaxin). This fusion process
triggers the synaptic release of neurotransmitters con-
tained in the small vesicles.14 BoNT/B cleaves the
sequence of human synaptobrevin (Syb) between Glu
76 and Phe 77 which are located on the cytosolic part
of Syb.5 We have previously developed an efficient high-
throughput assay for measuring the proteolytic activity
of BoNT/B15 and to determine the inhibitory potency of
a great number of molecules derived from classical or
combinatorial chemistry. Using this method, inhibitors
of BoNT/B were rationally designed through a classical
structure-activity relationship study. As tetanus toxin
and BoNT/B cleave the same substrate at the same
peptide bond, the strategy used to design BoNT/B
inhibitors was initially based on that developed to obtain
tetanus toxin inhibitors.16
In a first step, an exploration of the S1 subsite of
BoNT/B toxin by various â-amino thiols was performed.
This subsite, which specifically recognizes the Glu 76
residue of its natural substrate, synaptobrevin,10 is
assumed to accommodate preferentially polar side chains,
such as carboxylate, carboxamide, sulfonate, etc. In a
second step, the optimized S1 synthon was incorporated
in structures able to fit the S1-S′2 subsites of BoNT/B.
After determination by combinatorial chemistry of the
best natural residues interacting with these subsites,
non natural amino acids were used to improve the
inhibitory potency. The best inhibitor of BoNT/B de-
scribed to date (Ki ) 20 nM) was obtained by this way.
Resu lts
1. Ch em istr y. 1.1. Th e â-Am in o Th iols. Among the
various â-amino thiols used in this work, only three of
them, compounds 4, 8, and 9, have not been previously
* Corresponding author.
′ Tel. (33)-1-43.25.50.45. Fax. (33)-1-
43.26.69.18. roques@pharmacie.univ-paris5.fr.
† Contributed equally to the work.
10.1021/jm0300680 CCC: $25.00 © 2003 American Chemical Society
Published on Web 09/19/2003

Botulinum B Toxin Inhibition
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 22 4649
Sch em e 1. Synthesis of Inhibitors 4, 8, and 9a
a
(a) NaBH4 LiCl; (b) DEAD, PPh3, CH3COSH, THF; (c) NaOH, MeOH, THF; (d) TFA, anisole, ∆; (e) 6 N HCl, ∆.
described.17-21 They were prepared following reported
methods22 from their corresponding noncommercially
available R-amino acids. The t-Boc-amino esters 1 and
5 were prepared using a solid-liquid-phase transfer
catalytic method with the (benzylideneamino)acetic acid
ethyl ester and either the 4-(chloromethyl)-(tert-butyl)-
benzamide, or the 4-(bromomethyl)-phenyl(tert-butyl)-
sulfonamide (prepared from the commercially available
4-(chloromethyl)benzoyl chloride and 4-(chlorosulfonyl)-
benzoic acid, respectively), followed by acidic hydrolysis
of the benzylidene moiety and protection of the free
amine by a Boc group (data not shown).
The â-amino thiols 4, 8, and 9 were prepared by the
following synthetic sequence: (i) reduction of the esters
of 1 and 5 to afford the corresponding alcohols 2 and 6;
(ii) Mitsunobu reaction yielding the thioacetates 3 and
7; (iii) a final deprotection step leading to the expected
â-amino thiols 4, 8, and 9 (Scheme 1). These three
compounds were obtained as racemic mixtures.
1.2. Syn th esis of th e Syn th on 16. The chosen
method was an Oppolzer synthesis23 using the 2-(ben-
zydrylideneamino)-1-(10,10-dimethyl-3,3-dioxo-3λ6-thia-
4-aza-tricyclo[5.2.1.01.5]dec-4-yl)ethanone previously
described.24 Its stereospecific alkylation with 4-(bro-
momethyl)benzoic acid tert-butyl ester in the presence
of butyllithium as base afforded 10 which was then
deprotected by acidic hydrolysis in 11 and reprotected
by a Boc function to give 12 (Scheme 2). The source of
chirality, camphorsultam, is then cleaved in aprotic
medium by a phase-transfer catalysis method,24 leading
to the pure S enantiomer 13. The next steps were (i)
the Arndt-Eistert homologation of 13 leading to the
â-amino ester 14 with retention of configuration; (ii) the
stereospecific sulfenylation of 14 with 1-(4-methoxyben-
zyl)disulfanyl-2,4-dinitrobenzene;25 and (iii) the hydroly-
sis of the ester with bis(tributyltin) oxide which avoids
racemization. In these conditions, the synthon 16 is
obtained as pure 2S,3S diastereoisomer (Scheme 2).
1.3. Com bin a tor ia l Syn th esis of Tw o Libr a r ies
of P seu d otr ip ep tid es. Two pseudotripeptide libraries,
one to study the S1′ subsite of BoNT/B LC and the other
one to investigate its S2′ subsite, were generated by the
Divide/Couple/Recombine method using natural R-ami-
no acids.16 The peptides were obtained by solid-phase
synthesis on a chlorotrityl resin. For the first library
(Figure 1A), an equimolar mixture of 18 natural amino
acids as P2′ residues was introduced on the resin and
coupled successively with 18 amino acids (all natural
amino acids except proline and cysteine) corresponding
to the P1′ residue and finally coupled with the protected
synthon 16. The second library (Figure 1B) was achieved
by a parallel synthesis of 10 pseudotripeptides contain-
ing the same P1 synthon, a Trp residue as P1′ group,
and one of the 10 selected amino acids for the S2′ subsite
recognition.
All the coupling steps were performed using (benzo-
triazol-1-yloxy)tris(dimethylamino)phosphonium hexaflu-
orophosphate (Bop) and minimum amounts of diisopro-
pylethylamine to preserve the stereochemistry. The final
deprotection necessitated anhydrous hydrogen fluoride
owing to the use of 4-methoxybenzyl group as thiol
protecting moiety.
2. In h ibitor y P oten cies of th e Va r iou s In h ibi-
tor s. The various â-amino thiols were tested in vitro
on BoNT/B metallopeptidase activity, as described
previously.15 From the results reported in Table 1, it
appears that (i) the decreasing order of BoNT/B LC
recognition for the various side chains in P1 position is
phenyl > cyclohexyl > alkyl; (ii) the inhibitory potency
is increased when the benzyl (or cyclohexyl) moiety
bears a substituent in position 4; and (iii) this substitu-
ent is preferentially a carboxylate group. Although the
study was not exhaustive, it can be noticed that the best
â-amino thiol of this series, (compound 24, p-carbox-
yphenylalanine thiol) exhibits a rather good affinity (Ki
) 11 µM) for BoNT/B.
From the data reported in Figure 1A, it can be
observed that aromatic residues were preferred in the
P1′ position. The tryptophan led to a slightly better
inhibition than a phenylalanine or a tyrosine residue
(30-35% inhibition at 1 µM), while the less active
mixtures contained charged (Glu, Lys, Arg, etc.) or small
(Ser, Gly) amino acids. In the second series of inhibitors
(Figure 1B), a Trp residue was introduced in P1′ posi-
tion, and among the 10 residues tested as P2′ compo-
nents, aromatic amino acids and Asn are preferred with
35-40% of inhibitory potency at 1 µM. These results
clearly indicate a large preference of the S1′-S2′ domain
of the peptidase for aromatic residues. Consequently,
to tentatively improve the inhibitory potency of these

4650 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 22
Sch em e 2. Synthesis of Synthon 26a
Anne et al.
a
(a) nBuLi 2.5 M/hexane, THF, HMPA; (b) citric acid 10%, THF; (c) NEt3, (Boc)2O, DMF; (d) LiOH, LiBr, nBu4NBr, acetonitrile; (e)
iBuOCOCl, NMM, THF; (f) diazomethane followed by silver benzoate, NEt3, methanol; (g) nBuLi 1.6 M/hexane, HMDS, THF then HMPA,
1-(4-methoxybenzyldisulfanyl)-2,4-dinitrobenzene; (h) bis(tributyltin) oxide, acetonitrile.
pseudotripeptides, various compounds were synthesized
in large quantities by liquid-phase method, with a
benzylamide at the C-terminus.16
in BoNT/B LC. To our knowledge, 35 is the best
inhibitor of BoNT/B reported so far with a Ki value of
20 nM.
3. Molecu la r Mod elin g. All the compounds de-
scribed in this work have been synthesized before the
publication of the crystal structure of BoNT/B26 and a
substrate-sequence of synaptobrevin bound to the
BoNT/B LC.10 The results of these two studies were used
to investigate the possible mode of interaction of 35 with
BoNT/B LC. Figure 2 shows a view of the enzyme-
inhibitor complex. The biphenyl group seems to fit the
hydrophobic and deep pocket corresponding to the S′1
subsite of BoNT/B LC and interacts with several
hydrophobic residues of the protein. The carboxylate
group, of the P1 residue, faces a highly hydrophilic area
and was found at distances compatible with hydrogen
bonding or ionic interactions with the amino acids Lys
242 and Arg 183 of the metallopeptidase. The P1 benzyl
moiety of 35 participates to the correct orientation of
the p-carboxylate group. The benzothienyl heterocycle
seems to be stacked on the aromatic residue Tyr 53. This
could explain its favorable influence on BoNT/B LC
inhibition. Regarding the amino group, no evidence for
specific interaction with the enzyme was observed. The
result of this docking experiment are consistent with
those of the structure-activity studies.
Compound 26 issued from the library, with the
dipeptide Trp-Phe-benzylamide for S1′-S2′ domain rec-
ognition, inhibits BoNT/B with a Ki value of 3.5 × 10-7
M. The replacement of Trp by either the 1-naphthylm-
ethyl (compound 27), the NMe-Trp (28), or the biphenyl
methyl moiety (29) induces a slight increase in the
inhibitory potency of the related inhibitors (Table 2).
The effect of the stereochemistry of residues in P′1 and
P′2 positions on the metallopeptidase activity of BoNT/B
was also studied (compounds 29-31), and it was shown
that the S,S isomer yielded a slightly better BoNT/B
inhibitor. The rather weak stereochemical dependence
of BoNT/B LC for optimal inhibition is similar to that
observed with TeTx.16
In the second part of the study, the S′2 recognition of
BoNT/B LC was optimized. With this aim, three differ-
ent amino acids were used: an aromatic amino acid with
a hydroxyl group, tyrosine (compound 33), an amino acid
with a nonaromatic but hydrophobic side-chain, isoleu-
cine (compound 34). As compared to 29 which contains
a phenylalanine in P2′, 33 and 34 are slightly less
efficient consistent with the results of the library (Figure
1B). Then, a non natural aromatic amino acid with a
bulky side-chain corresponding to benzothienylalanine
was introduced in P2′ position, leading to 35. As shown
in Table 2, an improvement in inhibitory potency was
obtained, suggesting the presence of a large S′2 subsite
The inhibitory potency of compound 35 was tested
27
with BoNT/A LC
and with various physiological
enzymes such as angiotensin-converting enzyme (ACE),28
endothelin-converting enzyme (ECE),29 and aminopep-
tidases A and N (APA or APN).30 In all cases, the

Botulinum B Toxin Inhibition
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 22 4651
Ta ble 1. Inhibitory Potencies of Various â-Amino Thiols on
BoNT/B
compounds
R
Ki (BoNT/B)a
17
18
19
20
21
22
23
24
4
(CH2)3CH3
(CH2)2CO2H
CH2-cyclohexyl
425 µM
250 µM
160 µM
>1 mM
21 µM
100 µM
250 µM
11 µM
60 µM
65 µM
22 µM
97 µM
CH2-(3-COOH)-cyclohexyl
CH2-(4-COOH)-cyclohexyl
CH2-Ph
CH2-(3-COOH)-Ph
CH2-(4-COOH)-Ph
CH2-(4-CONH2)-Ph
CH2-(4-SO2NH2)-Ph
CH2-(4-SO3H)-Ph
8
9
25
CH2-(4-CH2COOH)-Ph
a
The Ki values are the mean ( SEM of three independent
experiments performed in triplicate.
subsites of the active site of BoNT/B. Indeed this
neurotoxin is easily produced and purified and is very
stable, accounting mainly for its possible use as a
weapon.11 Moreover BoNT/B is now currently used for
the treatment of various neuromuscular diseases12 and
for the reduction of wrinkles.31 In addition, BoNT/B
remains efficient in patients who have developed anti-
bodies against BoNT/A.32 Nevertheless, the use of
BoNT/B is sometimes associated with more or less
severe side effects.13
At this time, there is no potent inhibitor of BoNT/B.
Zinc chelating agents are effective blockers but are
devoid of selectivity and have important adverse effects
and thus cannot be used in therapy.33-35 Small peptides
derived from the substrate of BoNT/B are neither
substrates nor inhibitors.5,34 Classical inhibitors of zinc
metalloproteases (ACE, NEP, etc.) were shown to be
inactive.5,36-38 Other inhibitors have also been tested:
7-N-phenylcarbamoylamino-4-chloro-3-propyloxyisocou-
marin (elastase inhibitor) with an inhibitory potency of
27.6 µM,35,36 phosphonates and phenylmethylsulfonyl
fluoride (serine protease inhibitor) which were not really
effective,35,39 BABIM (bis(5-amidino-2-benzimidazolyl)-
methane) or keto-BABIM with inhibitory potencies of
1.6 and 0.8 µM,40 and buforin-1 with an inhibitory
potency of 1 µM were the first synthetic substances with
a relatively good affinity for BoNT/B.41 However, their
potency was too low for a possible clinical use. Conse-
quently, there was a great interest in designing highly
potent and selective inhibitors of BoNT/B.
F igu r e 1. Histograms representing the percentage of inhibi-
tion obtained with mixtures of pseudotripeptides derived from
the â-amino thiol 16 on botulinum B neurotoxin light chain
endoprotease activity. The general formula of these mixtures
of compounds are given above the histograms. (A) Each
mixture differs by an amino acid putatively interacting in S1′
which is represented in the abscissa as a three-letter code. A
stoechiometric amount of each amino acid is present in P2′.
These mixtures were tested at 1 µM on BoNT/B LC. (B) In
each case, a tryptophan is present in P1′ position. Each mixture
differs by the residue present in S2′ which corresponds to an
amino acid indicated as a three-letter code in the abscissa.
These mixtures were tested at 1 µM on BoNT/B LC.
inhibitory potency (IC50) was found at least 2 orders of
magnitude higher as for BoNT/B LC (data not shown).
Thus, 35 appears as a selective BoNT/B inhibitor.
With this aim, pseudo-tripeptides were synthesized
since such molecules have been shown to be quite
effective for the inhibition of TeTx,16 a protease very
similar to BoNT/B, cleaving the same substrate (syn-
aptobrevin) at the same position. This was achieved step
by step by characterizing the best residues for S1, S′1,
and S′2 recognition.
In the case of BoNT/B, the benzyl side chain in P1
position requires a p-carboxylate group for optimal
interaction whereas a m-sulfonamide group seems to be
preferred in tetanus toxin.18 The importance of the
carboxylate could be related to interactions with con-
stituting amino acids present in the S1 subsite as
suggested by molecular modeling. The inhibitors’ poten-
cies are modulated by the position of the charged
Discu ssion
Inhibitors of botulinum neurotoxins could be devoted
to the protection against infection by the toxins or to
alleviate adverse effects when the toxins are used in
therapy. The first role would be preventive in the case
where a contamination is suspected, too late to allow
the use of serum but soon enough so that the inhibitors
could prevent the intracellular effect of the toxin. The
second one would be to block the evolution of the disease
by protecting the intracellular target proteins which are
not already cleaved by the toxin. Thus the duration and
the severity of the disease could be reduced.
In this work, our main objective was to investigate
the nature of the best residues interacting with the

4652 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 22
Ta ble 2. Inhibitory Potency of Molecules 26-35 on BoNT/B LC
Anne et al.
absolute
configuration
absolute
configuration
CR(P2′)
Ki (BoNT/B),a
no.
P1′
indolyl
CR(P1′)
P2′
M
26
27
28
29
30
31
32
33
34
35
S
S
S
S
R
S
S
S
S
S
benzyl
benzyl
benzyl
benzyl
benzyl
benzyl
imidazolylmethyl
4-OH-benzyl
1-methylpropyl
benzothienylmethyl
S
S
S
S
S
R
S
S
S
(3.5 ( 0.7) × 10-7
(2.2 ( 0.4) × 10-7
(1.7 ( 0.3) × 10-7
(1.6 ( 0.2) × 10-7
(5.4 ( 0.3) × 10-7
(5.8 ( 0.4) × 10-7
(5.0 ( 0.9) × 10-7
(8.1 ( 0.9) × 10-7
(5.4 ( 0.4) × 10-7
(2.0 ( 0.2) × 10-8
1-naphthyl
N-Me-indolyl
biphenyl
biphenyl
biphenyl
biphenyl
biphenyl
biphenyl
biphenyl
a
The Ki values are the mean ( SEM of three independent experiments performed in triplicate.
explained by the deep and hydrophobic S′1 subsite as
evidenced by crystallographic studies10,26 and molecular
modeling (this study). It seems that this subsite is so
deep that side chains longer than biphenyl could pos-
sibly improve the enzyme recognition.
Cleavage of synaptobrevin analogues by BoNT/B has
suggested that interaction with the S′2 subsite would
be less essential as that with the S′1.42 This was not
the case with inhibitors, illustrating the absence of
optimization of metallopeptidase recognition by sub-
strate as already observed with another zinc peptidase,
NEP.9 Thus, the best inhibitor of BoNT/B LC was
obtained in our series with a large bicyclic heteroaro-
matic residue (compound 35) while a glutamate was
found at this P′2 level in the substrate.5 This suggests
that the S′2 interactions with synaptobrevin are not
optimal, the selectivity being related to the recognition
of the long cleft of BoNT/B LC by the side chains of the
constituent amino acids of the substrate.10 This could
explain why specifically changing the glutamate does
not really modify the kinetics of cleavage. In the case
of small inhibitors such as those reported here, there is
only a small number of possible interactions, and their
optimization is essential to improve the inhibitory
potency (Table 2). This is clearly demonstrated with the
benzothienyl group. This molecule is the best inhibitor
ever described for BoNT/B with an inhibitory potency
of 20 nM (patent Fr. 01.04.895).44 This is probably due
to the various favorable interactions (ionic, hydrophobic,
aromatic ring stackings) evidenced in the enzyme active
site by molecular modeling. Nevertheless, the precise
mechanism of BoNT/B LC inhibition remains to be
investigated in detail since several modes of enzyme-
recognition have been identified in crystal structures
of the complex between BoNT/B and the less potent
inhibitor BABIM.45
F igu r e 2. Molecular modeling of the best inhibitor 35.
Schematic representation of the putative interactions with
BoNT/B.
carboxyl group and the length of the side chain bearing
this group (compare compounds 23, 24, 25). This seems
to be in relation to an optimal interaction with amino
acids present in the S1 subsite.
In a second step, the interaction with the S′1 subsite
was studied using a library of inhibitors obtained by
combinatorial chemistry with natural amino acids. This
showed that the best residue is a hydrophobic aromatic
residue, which is consistent with the cleavage site of
synaptobrevin, the substrate of BoNT/B, where the P′1
amino acid is a phenylalanine. The nature of this
residue was shown to be essential for a good interaction
with BoNT/B LC38,42 and for optimal recognition of
almost all zinc metallopeptidases.43
An optimization of the S′1 interaction was achieved
using non natural amino acids, leading to bulky aro-
matic residues as preferred ligands. This could be
Because the inhibitory potency is already in the high
nanomolar range, we could expect a rapid optimization
of this type of compound. This promising approach could

Botulinum B Toxin Inhibition
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 22 4653
[2-(4-ter t-Bu tylsu lfa m oylp h en yl)-1-h yd r oxym eth yleth -
yl]ca r ba m ic a cid ter t-bu tyl ester (6): 0.64 g (95%). TLC
(dichloromethane:methanol, 90:10): Rf ) 0.83. 1H NMR (DMSO-
d6) δ 1.05 (s, 9H), 1.25 (s, 9H), 2.45 (m, 4H), 3.55 (m, 1H), 4.75
(t, 1H), 6.6 (d, 1H), 7.3 (d, 2H), 7.35 (s, 1H), 7.65 (d, 2H).
Gen er a l P r oced u r e for th e Syn th esis of Com p ou n d s
3 a n d 7. To a solution of triphenylphosphine (2 equiv) in
anhydrous tetrahydrofuran (2 mL/mmol), stirred at 0°C, was
added diisopropylazodicarboxylate (2 equiv), and the mixture
was stirred at this temperature during 30 min. A yellow solid
was formed at this step. Then a mixture composed of thiolace-
tic acid (2 equiv) and the alcohol 2 or 6 (1 equiv) in anhydrous
tetrahydrofuran (2 mL/mmol of alcohol) was added at 0 °C.
The reaction was stirred at room temperature during 72 h.
After evaporation in vacuo, the residue was dissolved in ethyl
acetate and the organic layer washed once with water, a
solution of hydrogenocarbonate, water, and brine. After evapo-
ration in vacuo, the residue was purified by column chroma-
tography on silica gel with 30% of ethyl acetate in cyclohexane
to give the compounds as a colorless oil.
allow the design of molecules for in vivo use. Thus,
docking experiments have shown that the amino group
seems to have no role in enzyme recognition. Therefore,
removal of this hydrophilic group could lead to com-
pounds with similar inhibitory potencies in vitro with
probably a better in vivo bioavailability.
Exp er im en ta l P r oced u r es
Molecu la r Mod elin g. Molecular mechanics calculations
were performed using InsightII package (Accelrys Inc.) and
AMBER force field. A distance-dependent dielectric screening
constant of 4r was used. The protein was built using as a
starting point the PDB crystal structure of BoNT/B (PDB entry
1EPN).26 The inhibitor was first docked into the enzyme by
performing a sequence of six molecular dynamics runs. Each
run was done at 300 K during 10000 steps of 1 fs. A distance
constraint was imposed between the S atom of the thiol group
and the metal ion of the enzyme. This distance constraint was
sequentially reduced from 12 Å with the inhibitor outside the
protein active site for the first dynamics run to 2.3 Å for the
sixth one. For these runs, the protein backbone was held frozen
but its side chains were relaxed and the inhibitor was
completely relaxed. A conformational search of the best
inhibitor position was then performed using the AFFINITY
module of InsightII with a distance between the metal ion and
the S atom of the thiol group constrained to 2.3 Å, assuming
that the thiol acts as zinc chelating group. Among 50, 17
structures fulfilling these criteria and low energy were found
by the program. One of these structures was chosen for its
very low nonbinding energy between the enzyme and the
inhibitor in comparison to the others. This structure also
seemed acceptable by geometrical criteria, considering experi-
mental values of Ki for different inhibitors. Finally, 50 steps
of dynamics were performed on this structure. Each was done
at 300 K during 2000 steps of 1 fs. For these final runs, no
distance constraint was imposed between the enzyme and the
inhibitor. The resulting structure was submitted to steepest-
descent and conjugate-gradient minimization and stored.
Analysis of these structures showed a large family of low
energy overlapping structures. The minimum-energy structure
was used as a model for this study.
4-[1-(Acetylsu lfa n yl)-2-(Boc-a m in o)p r op yl]-(N-ter t-bu -
tyl)ben za m id e (3): 1.01 g (69%). TLC (cyclohexane:ethyl
1
acetate, 70:30): Rf ) 0.31. H NMR (DMSO-d6) δ 1.25 (s, 9H),
1.3 (s, 9H), 2.25 (s, 3H), 2.8 (dd, 2H), 3.0 (dd, 2H), 3.7 (m, 1H),
6.8 (d, 2H), 7.2 (d, 2H), 7.6 (s, 1H), 7.7 (d, 2H).
4-[1-(Acetylsu lfa n yl)-2-(Boc-a m in o)p r op yl]p h en yl-(N-
ter t-bu tyl)su lfon a m id e (7): 0.5 g (68%). TLC (cyclohexane:
ethyl acetate, 70:30): Rf ) 0.26. 1H NMR (DMSO-d6) δ 1.05 (s,
9H), 1.25 (s, 9H), 2.25 (s, 3H), 2.75 (dd, 2H), 2.75 et 2.95 (dd,
2H), 3.65 (m, 1H), 6.85 (d, 1H), 7.25 (d, 2H), 7.35 (s, 1H), 7.65
(d, 2H).
4-(2-Am in o-3-su lfa n ylp r op yl)ben za m id e (4). To a solu-
tion of 3 (1.0 g, 2.46 mmol) in a mixture of methanol (5 mL)
and tetrahydrofuran (5 mL) at 0°C under argon was added a
1 M sodium hydroxide solution (2.6 mL, 2.6 mmol). The
mixture was stirred at room temperature during 90 min. The
organic solvent was evaporated in vacuo, and after acidification
with a 1 N HCl solution, the solid was filtered and dried to
yield 0.805 g (90%), which was used without purification.
A solution of this compound (795 mg, 2.18 mmol) in
trifluoroacetic acid (6.7 mL, 84 mmol) and anisole (3.6 mL)
was heated at 60 °C during one week. After drying, the product
was precipitated with diethyl ether and purified by HPLC (C18
Kromasil, 100 Å, 5 µm). After freeze-drying, compound 4 was
1
Ch em istr y. H NMR spectra were measured on a Bruker
AC 270 MHz spectrometer using tetramethylsilane as internal
standard. Electrospray mass spectra (MS-ES) were recorded
on a Esquier-Bru¨ker spectrometer. Flash column chromatog-
raphy was performed using 40-63 µm silica gel. Reaction
progress was determined by either TLC analysis or monitored
using analytical reverse-phase HPLC (Shimadzu, LC10 AD-
vp with a Class-VP5.03 software) using a Kromasil C18 column
(100 Å, 5 µm, 250 × 4.6 mm) (Touzart-Matignon, France), with
a mobile phase consisting of water containing 0.05% TFA (A)
and acetonitrile/water 9/1 containing 0.038% TFA (B). Re-
agents were obtained from commercial sources and are used
without further purification.
Gen er a l P r oced u r e for th e Syn th esis of Com p ou n d s
2 a n d 6. A mixture composed of the appropriate carboxylic
acid in dimethoxyethane (1 mL/mmol) was stirred at -15 °C
under argon. N-Methylmorpholine (1 equiv) and isobutylchlo-
roformate (1 equiv) were then added successively and the
reaction was stirred during 2 min. at -15 °C. After filtration,
a solution of sodium borohydride (1.5 equiv) in water (0.7 mL/
mmol) was added, and the reaction was immediately hydro-
lyzed with water (26 mL/mmol). The product was extracted
two times with ethyl acetate, and the combined organic layers
were washed with HCl (1 N) and brine. After evaporation in
vacuo, the title compounds were obtained as colorless oils
which were used without further purification.
1
obtained as a white powder: 0.13 g, (30%). H NMR (DMSO-
d6 + TFA) δ 2.6 (m, 2H), 2.9 (d, 2H), 3.5 (m, 1H), 7.2 (d, 2H),
7.65 (d, 2H), 7.9 (m, 3H). MS (ESI) (M + H)+ m/z ) 210.9.
4-(2-Am in o-3-su lfa n ylp r op yl)ben zen esu lfon a m id e (8).
To a solution of a compound 7 (0.25 g, 0.56 mmol) in a mixture
of methanol and tetrahydrofuran (1/1, 2 mL) at 0 °C under
argon was added NaOH (0.6 mL, 0.59 mmol). The mixture was
stirred at room temperature during 90 min. After acidification
with HCl (1 N), the product was filtered and dried to yield a
white powder, 0.157 g (70%).
A solution of this crude product in trifluoroacetic acid (1 mL)
and anisole (0.5 mL) was stirred overnight at room tempera-
ture. After drying, the product was precipitated with diethyl
ether. Compound 8 was obtained as a white powder: 0.08 g
(94%). 1H NMR (DMSO-d6 + TFA) δ 2.6-3.1 (m, 4H), 3.65 (m,
1H), 7.4 (d, 2H), 7.75 (d, 2H), 8.05 (m, 3H). MS (ESI) (M -
Cl)+ m/z ) 247.
4-(2-Am in o-3-su lfa n ylp r op yl)ben zen esu lfon ic Acid (9).
A solution of 7 (0.25 g, 0.56 mmol) in HCl (6 N) (2 mL) was
heated at 120 °C overnight. After drying, the product was
purified on cation-exchange resin DOWEX 50W×8. After
freeze-drying, the product was obtained as yellow powder: 0.06
1
g (40%). H NMR (DMSO-d6 + TFA) δ 2.75 (m, 2H), 2.85 (m,
2H), 3.65 (m, 1H), 7.4 (d, 2H), 7.75 (d, 2H), 8.05 (m, 3H). MS
(ESI) (M - Cl)+ m/z ) 247.0
[1-(4-ter t-Bu tylca r ba m oylben zyl)-2-h yd r oxyeth yl]ca r -
ba m ic a cid ter t-bu tyl ester (2): 1.25 g (71.5%). TLC
(dichloromethane:methanol, 90:10): Rf ) 0.43. 1H NMR (DMSO-
d6) δ 1.25 (s, 9H), 1.3 (s, 9H), 2.55 (dd, 2H), 2.8 (dd, 2H), 3.55
(m, 1H), 4.7 (t, 1H), 6.6 (d, 1H), 7.2 (d, 2H), 7.6 (s, 1H), 7.7 (d,
2H).
4-[2-(Ben zh yd r ylid en ea m in o)-3-(10,10-d im eth yl-3,3-d i-
oxo-3λ6-th ia -4-a za -tr icyclo[5.2.1.01.5]d ec-4-yl)-3-oxop r op y-
l]ben zoic Acid ter t-Bu tyl Ester (10). A solution of 2-(ben-
zhydrylideneamino)-1-(10,10-dimethyl-3,3-dioxo-3λ6-thia-4-
aza-tricyclo[5.2.1.01.5]dec-4-yl)ethanone (17 g, 38.9 mmol) in

4654 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 22
Anne et al.
200 mL of anhydrous tetrahydrofuran was stirred at -78 °C
under argon. nBuLi (2.5 M in hexane, 16.8 mL, 42.5 mmol)
was added dropwise, and the mixture was stirred 15 min at
this temperature. A solution of 4-bromomethylbenzoic acid tert-
butyl ester (12.7 g, 46.68 mmol) in 30 mL of hexamethylphos-
phoramide and 100 mL of anhydrous tetrahydrofuran was
added dropwise at -78 °C. The mixture was allowed to warm
to room temperature under stirring during 3 h. The reaction
was stopped with a mixture of tetrahydrofuran and acetic acid
(4/1, 15 mL) at 0 °C. After removal of the solvent, the residue
was dissolved in 300 mL of ether. The organic layer was
washed once with 10% NH4Cl solution, water, and brine. After
drying with Na2SO4 the solvent was removed. The oily residue
was triturated in a mixture of cyclohexane/ ethyl acetate 95/5
purified by silica gel column chromatography (cyclohexane:
ethyl acetate, 6:4) to give the product as a yellow solid: 2.36
g (67%). mp ) 71 °C. [R]22D ) -82.8 (c ) 1.03 in CHCl3). HPLC
(A:B, 20:80): tR ) 5.50 min. 1H NMR (DMSO-d6) δ 1.25 (s, 9H),
1.50 (s, 9H), 2.35 (d, 2H), 2.71 (d, 2H), 3.45 (s, 3H), 3.90 (m,
1H), 6.80 (d, 1H), 7.25 (d, 2H), 7.75 (d, 2H).
4-[2-ter t-Bu t oxyca r b on yla m in o-3-(4-m et h oxyb en zyl-
su lfa n yl)-3-m eth oxyca r bon ylp r op yl]ben zoic Acid ter t-
Bu t yl E st er (15). To a solution of freshly distilled hexa-
methyldisilazane (3.46 mL, 16.4 mmol) in 40 mL of dry
tetrahydrofuran at 0 °C under argon was added nBuLi (1.6 M
in hexane, 10.24 mL, 16.4 mmol). The mixture was stirred 15
min at 0 °C and cooled to -70 °C. A solution of 14 (2.3 g, 5.85
mmol) in 23 mL of dry tetrahydrofuran was added dropwise.
After 2 h under argon at -70 °C, 1-(4-methoxybenzyldisulfa-
nyl)-2,4-dinitrobenzene (2.89 g, 8.19 mmol) and hexameth-
ylphosphoramide (2 mL, 8.19 mmol) were added, and the
mixture was stirred 90 min at -70 °C. The reaction was
quenched with 10% NH4Cl, diluted with 150 mL of ethyl
acetate. The organic layer washed with water and brine, dried
over Na2SO4, and concentrated. The crude mixture was puri-
fied by silica gel column chromatography (cyclohexane:dichlo-
romethane: ethyl acetate, 7:1.5:1.5) to give the product as pale
yellow solid: 1.55 g (47%). mp ) 79 °C. HPLC (A:B, 20:80): tR
to give a white solid: 17.6 g (72%). mp ) 174 °C. [R]22
)
D
-605.1 (c ) 0.787 in MeOH). 1H NMR (DMSO-d6) δ 0.45 (s,
3H), 0.75 (s, 3H), 1.45 (s, 9H), 1.50-1.85 (m, 6H), 2.35 (m, 2H),
2.86 (m, 1H), 3.40-3.75 (m, 3H), 4.80 (t, 1H), 6.85-7.0 (m,
4H), 7.30-7.50 (m, 8H), 7.50 (d, 2H).
4-[2-Am in o)-3-(10,10-d im eth yl-3,3-d ioxo-3λ6-th ia -4-a za -
tr icyclo[ 5.2.1.01.5]dec-4-yl)-3-oxopr opyl]ben zoic Acid ter t-
Bu tyl Ester (11). To a solution of 10 (16.1 g, 25.8 mmol) in
270 mL of tetrahydrofuran was added 272 mL of a 10% citric
acid solution. The mixture was stirred at room temperature
during 2 h. After removal of the solvent, the aqueous layer
was washed twice with ether and the pH adjusted to 7-8 with
saturated aqueous NaHCO3. The product was extracted with
dichloromethane. The organic layer washed with brine, dried
over Na2SO4, and concentrated to give the product as a
1
) 10.05 min. H NMR (DMSO-d6) δ 1.15 (s, 9H), 1.50 (s, 9H),
2.35 (d, 2H), 3.0 (m, 1H), 3.60 (s, 3H), 3.70 (s, 3H), 3.8-4.0
(m, 3H), 6.80 (m, 3H), 7.25 (m, 4H), 7.70 (m, 2H).
4-[2-ter t-Bu t oxyca r b on yla m in o-3-ca r b oxy-3-(4-m et h -
oxyben zylsu lfa n yl)p r op yl]ben zoic Acid ter t-Bu tyl Ester
(16). To a solution of bis(tributyltin) oxide (8.2 mL, 16.1 mmol)
in 32 mL of acetonitrile was added a solution of 15 (1.5 g, 2.75
mmol) in 5.5 mL of acetonitrile. The mixture was heated under
reflux. After 5 days, the mixture was cooled to 0 °C and 20
mL of aqueous 2 N HCl was added. After stirring 3 h at room
temperature, the reaction was diluted in 100 mL of ethyl
acetate, washed with water and brine, dried over Na2SO4, and
concentrated to give a pale yellow oil. Silica gel chromatogra-
phy (cyclohexane:ethyl acetate:acetic acid, 8:2:0.5) gave the
product as a yellow solid: 1.15 g (79%). mp ) 76 °C. HPLC
(A:B, 20:80): tR ) 6.1 min. [R]22D ) -489.2 (c ) 0.732 in MeOH).
1H NMR (DMSO-d6) δ 1.2 (s, 9H), 1.45 (s, 9H), 2.35 (d, 2H),
3.0 (m, 1H), 3.75 (s, 3H), 3.8-4.0 (m, 3H), 6.80 (m, 3H), 7.25
(m, 4H), 7.70 (m, 2H).
1
colorless oil: 9.3 g (78%). H NMR (DMSO-d6) δ 0.75 (s, 3H),
0.85 (s, 3H), 1.50 (s, 9H), 1.60-2.10 (m, 6H), 2.35 (m, 2H), 2.70
(m, 1H), 2.95 (m, 1H), 3.50-3.80 (m, 3H), 4.0 (t, 1H), 7.30 (d,
2H), 7.75 (d, 2H).
4-(2-ter t-Bu t oxyca r b on yla m in o-2-ca r b oxyet h yl)b en -
zoic Acid ter t-Bu tyl Ester (13). To a solution of 11 (8.7 g,
18.7 mmol) and triethylamine (5.26 mL, 37.42 mmol) in 90
mL of dimethylformamide at 0 °C was added di-tert-butyl
dicarbonate (4.5 g, 20.58 mmol). The mixture was stirred
during 48 h at room temperature. After removal of the solvent,
the crude mixture was diluted with 250 mL of ethyl acetate,
washed with 10% KHSO4, water, and brine, dried over Na2-
SO4, and concentrated. The resulting crude oil 12 (8.5 g, 75%)
was used without purification (TLC (cyclohexane:ethyl acetate:
dichloromethane, 7:1.5:1.5): Rf ) 0.21).
Gen er a l P r oced u r e for th e Syn th esis of Dep r otected
Dip ep tid es. Syn th esis of Com bin a tor ia l Libr a r ies of
Th iol-Con ta in in g In h ibitor s. A pseudotripeptide library,
corresponding to 18 mixtures of thiol-containing peptides of
formula (synthon 26)-AA′1x-AA′2x, where AA′1x and AA′2x
represent the 18 different natural amino acids (Cys and Pro
omitted), was obtained by solid-phase peptide synthesis on
2-chlorotrityl chloride resin, by using a previously described
protocol.16 The Divide/Couple/Recombine method was used to
synthesize 18 sublibraries. A first amino acid mixture is
prepared by combining the 18 different amino acids linked to
the 2-chlorotrityl resin. This homogenized mixture is then split
into 18 identical samples to which a known amino acid is
coupled. This gives rise to 18 different mixtures, each of them
containing synthon 26 in each sample, allowing the synthesis
of 18 mixtures containing 18 couples of diastereomers. The
18 fractions of peptidyl resin were then cleaved from the resin
by using a dichloromethane/trifluoroethanol mixture (8/2). The
deprotection of the TFA-labile protecting groups was achieved
by using trifluoroacetic acid, and the final deprotection step
used to hydrolyze the benzylcarbamate group and the 4-meth-
oxybenzyl group was performed with anhydrous hydrogen
fluoride. The freeze-dried sublibraries were soluble in water
at 1 mM. Electrospray mass spectra and quantitative amino
acid analysis showed in each case a correct mass distribution
and a good stoechiometry in the S2′ position, respectively (data
not shown).
A solution of 12 (8.4 g, 15.03 mmol), LiOH (1.44 g, 60.12
mmol), LiBr (6.51 g, 75.15 mmol), and nBu4NBr (1.93 g, 6.01
mmol) in 150 mL of acetonitrile was stirred overnight at room
temperature. After removal of the solvent, 150 mL of water
was added and the aqueous layer washed twice with ether.
The pH was adjusted at 2-3 with 10% KHSO4, and the product
was extracted with dichloromethane. The organic layer washed
with brine, dried over Na2SO4, and concentrated. The resulting
crude mixture was purified by silica gel column chromatog-
raphy (cyclohexane:ethyl acetate:acetic acid, 7.5:2.5:0.25) to
give the product as a colorless oil: 6.2 g (71%). [R]22 ) 100.4
D
(c ) 1.068 in MeOH). TLC: (cyclohexane:ethyl acetate:acetic
acid, 5:5:0.5) Rf ) 0.75. 1H NMR (DMSO-d6) δ 1.25 (s, 9H),
1.50 (s, 9H), 2.80 (m, 1H), 3.0 (m, 1H), 4.05 (t, 1H), 7.00 (d,
1H), 7.30 (d, 2H), 7.75 (d, 2H).
4-(2-ter t-Bu t oxyca r b on yla m in o-3-m et h oxyca r b on yl-
p r op yl)ben zoic Acid ter t-Bu tyl Ester (14). Isobutyl chlo-
roformate (0.99 mL, 9 mmol) was added to a solution of 13
(3.29 g, 9 mmol) and N-methylmorpholine (0.99 mL, 9 mmol)
in 27 mL of tetrahydrofuran at -20 °C. The mixture was
stirred 15 min and filtered. At 0 °C a solution of diazomethane
(80 mmol) in 100 mL of ether was added to the filtrate. After
2 h at room temperature and removal of the solvent, 170 mL
of methanol, silver benzoate (144 mg, 0.72 mmol), and tri-
ethylamine (3.79 mL, 27 mmol) were successively added. The
mixture was stirred 15 min at 0 °C and 30 min at room
temperature. The dark suspension was filtered on Celite 545.
After removal of the methanol, the crude mixture was diluted
with ether (250 mL), washed with water and brine, dried over
Na2SO4, and concentrated. The resulting crude yellow oil was
A second library corresponding to the parallel synthesis of
pseudotripeptides of formula (synthon 26)-Trp-AA2 was achieved
on 2-chlorotrityl chloride resin with the same protocol de-
scribed above.

Botulinum B Toxin Inhibition
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 22 4655
To a solution of Boc-amino acid (5 mmol) in dichloromethane
(3 mL/mmol) was added successively benzylamine (6 mmol),
(benzotriazol-1-yloxy)tris(dimethylamino)phosphonium hexaflu-
orophosphate (Bop) (6 mmol), and diisopropylethylamine (17.5
mmol). The reaction was stirred 3 h at room temperature. After
evaporation to dryness, the oily residue was diluted with ethyl
acetate and washed with aqueous 1 M KHSO4, 1 M NaHCO3,
water, and brine. The organic phase was concentrated to give
the product as an oil which was used in the next step without
purification.
To a solution of the Boc-benzylamide amino acid (2.5 mmol)
in dichloromethane (5 mL/mmol) at 0 °C was added trifluoro-
acetic acid (TFA) (50 mmol). The mixture was stirred 2 h at 0
°C and 1 h at room temperature. After evaporation to dryness,
the product was obtained as a white solid after trituration of
the oily residue with ether/hexane (1/1). The product was used
in the next step without purification.
(m, 21H), 7.75 (d, 2H), 7.95 (m, 3H), 8.45 (d, 1H), 8.65 (m, 2H).
MS (ESI) (M + H)+ m/z ) 715.6.
4-{(2S)-2-Am in o-3-[1-(1-ben zylca r ba m oyl-(2R)-2-p h en -
yl-eth ylcar bam oyl)-(2S)-2-biph en yl-4-yl-eth ylcar bam oyl]-
3(S)-su lfa n ylp r op yl)}ben zoic a cid (31): 32 mg (32%). HPLC
(gradient 40-100% B in 20 min): tR ) 13.2 min. 1H NMR
(DMSO-d6 + TFA) δ 2.7-3.2 (m, 6H), 3.55 (m, 1H), 3.65 (m,
1H), 4.15-4.25 (2dd, 2H),4.65 (m, 2H), 7.05-7.5 (m, 21H), 7.8
(d, 2H), 7.95 (m, 3H), 8.45 (d, 1H), 8.65 (m, 2H). MS (ESI) (M
+ H)+ m/z ) 715.5.
4-(2S)-2-Am in o-3-{1-[1-ben zylca r ba m oyl-(2S)-2-(3H-im -
id a zol-4-yl)-et h ylca r b a m oyl]-(2S)-2-b ip h en yl-4-ylet h yl-
ca r ba m oyl}-3(S)-su lfa n ylp r op yl)ben zoic a cid (32): 21 mg
(25%). HPLC (35% B): tR ) 6.2 min. 1H NMR (DMSO-d6
+
TFA) δ 2.6-3.15 (m, 6H), 3.35-3.45 (dd, 1H), 3.7 (m, 1H), 4.1
(2dd, 2H), 4.65 (m, 2H), 6.7 (d, 2H), 7.05-7.55 (m, 18H), 7.8
(d, 2H), 8.0 (d, 1H), 8.45 (t, 1H), 8.65 (d, 1H), 8.55 (m, 1H), 8.8
(s, 1H). MS (ESI) (M + H)+ m/z ) 705.3.
The deprotected dipeptide was obtained after coupling with
Boc-amino acid using Bop and deprotected with TFA. The
dipeptides are used without purification in the next step.
4-{(2S)-2-Am in o-3-[1-(1-ben zylca r ba m oyl-(2S)-2-(4-h y-
d r oxyp h en yl)eth ylca r ba m oyl)-(2S)-2-bip h en yl-4-yleth yl-
ca r ba m oyl]-3(S)-su lfa n ylp r op yl)}ben zoic a cid (33): 13 mg
(14%). HPLC (gradient 40-100% B in 20 min): tR ) 11.1 min.
1H NMR (DMSO-d6 + TFA) δ 2.6-3.2 (m, 6H), 3.65 (m, 2H),
4.2 (d, 2H), 4.55-4.65 (m, 2H), 6.55 (d, 2H), 6.9-7.45 (m, 18H),
7.8 (d, 2H), 7.9 (m, 3H), 8.55 (m, 3H). MS (ESI) (M + H)+ m/z
) 731.1.
4-{(2S)-2-Am in o-3-[1-(1-ben zylca r ba m oyl-(2S)-2-m eth -
ylbu tylca r ba m oyl)-(2S)-2-bip h en yl-4-yleth ylca r ba m oyl]-
3(S)-su lfa n ylp r op yl)}ben zoic a cid (34): 28 mg (26%). HPLC
(gradient 40-100% B in 20 min): tR ) 11.0 min. 1H NMR
(DMSO-d6 + TFA) δ 0.75 (m, 6H), 1.1-1.45 (m, 2H), 1.75 (m,
1H), 2.75-3.05 (m, 6H), 3.65 (m, 2H), 4.2 (m, 3H), 4.7 (m, 1H),
7.1-7.55 (m, 16H), 7.8 (d, 2H), 7.95 (m, 3H), 8.3 (d, 1H), 8.55
(t, 1H), 8.6 (d, 1H). MS (ESI) (M + H)+ m/z ) 681.1.
4-{(2S)-2-Am in o-3-[1-(2S)-2-b en zo[b]t h iop h en -3-yl-1-
ben zylca r b a m oylet h ylca r b a m oyl)-(2S)-2-b ip h en yl-4-yl-
eth ylca r ba m oyl]-3(S)-su lfa n ylp r op yl)}ben zoic a cid (35):
Gen er a l P r oced u r e for th e Syn th esis of P seu d otr ip -
ep tid es 26-35. To a solution of (2S,3S)-3-amino-3-(3-tert-
butylsulfamoylphenyl)-2-(4-methoxybenzylsulfanyl)propionic
acid23 or synthon 16 (0.24 mmol) in dimethylformamide (6 mL/
mmol) were added the selected dipeptide (0.29 mmol), Bop
(0.29 mmol), and diisopropylethylamine (0.84 mmol). The
mixture was stirred 3 h at room temperature. After removal
of the solvent, the crude solid was washed with aqueous 10%
KHSO4 solution and water, dried, and used without purifica-
tion.
The residue (0.12 mmol) was stirred at 0 °C for 1 h with 10
mL of anhydrous hydrogen fluoride (HF), 0.6 mL of anisole,
and 0.6 mL of dimethyl sulfide. After evaporation of HF, the
residue was taken up with TFA and precipitated with a cold
mixture of ether:n-hexane (1/1). After centrifugation, the
precipitate was taken up with water and freeze-dried. The
product was finally purified by semipreparative HPLC on
Kromasil C18 column (100 Å, 5 µm, 250 × 20 mm). Analytical
HPLC and RMN analysis confirmed the presence of one single
diastereoisomer.
27 mg (30%). HPLC (gradient 55-100% B in 15 min): tR
)
1
8.7 min. H NMR (DMSO-d6 + TFA) δ 2.6-2.33 (m, 6H), 3.6
(m, 2H), 4.2 (d, 2H), 4.65-4.75 (m, 2H), 6.9-7.45 (m, 19H),
7.7-7.85 (m, 4H), 7.9 (m, 3H), 8.6 (m, 3H). MS (ESI) (M +
H)+ m/z ) 769.6.
4-{(2S)-2-Am in o-3-[1-(1-b en zylca r b a m oyl-(2S)-2-p h e-
n ylet h ylca r ba m oyl)-(2S)-2-(1H -in d ol-3-yl)et h ylca r ba m -
oyl]-3(S)-su lfa n ylp r op yl)}ben zoic a cid (26): 22 mg (27%).
Ack n ow led gm en t. The authors gratefully acknowl-
edge C. Dupuis for her assistance in drafting this
manuscript and Dr. A. Blommaert for helpful comments
and stylistic revision.
1
HPLC (55% B): tR ) 5.2 min. H NMR (DMSO-d6 + TFA) δ
2.6-3.15 (m, 6H), 3.6 (m, 2H), 4.15-4.25 (m, 2H), 4.6 (m, 2H),
6.8-7.25 (m, 16H), 7.55 (d, 1H), 7.85 (d, 2H), 7.9 (m, 3H), 8.5
(m, 3H), 10.7 (s, 1H). MS (ESI) (M + H)+ m/z ) 678.4.
Refer en ces
4-{(2S)-2-Am in o-3-[1-(1-b en zylca r b a m oyl-(2S)-2-p h e-
n yleth ylcar bam oyl)-(2S)-2-n aph th alen -1-yl-eth ylcar bam -
oyl]-3(S)-su lfa n ylp r op yl)}ben zoic a cid (27): 15 mg (18%).
(1) Simpson, L. L. The origin, structure, and pharmacological
activity of botulinum toxin. Pharmacol. Rev. 1981, 33, 155-188.
(2) Middlebrook, J . L.; Brown, J . E. Immunodiagnosis and immu-
notherapy of tetanus and botulinum neurotoxins. Curr. Top.
Microbiol. Immunol. 1995, 195, 89-122.
(3) Shapiro, R. L.; Hatheway, C.; Swerdlow, D. L. Botulism in the
United States: a clinical and epidemiologic review. Ann. Intern.
Med. 1998, 129, 221-228.
(4) Montecucco, C.; Papini, E.; Schiavo, G. Bacterial protein toxins
penetrate cells via a four-step mechanism. FEBS Lett. 1994, 346,
92-98.
(5) Schiavo, G.; Benfenati, F.; Poulain, B.; Rossetto, O.; Polverino
de Laureto, P. DasGupta, B. R.; Montecucco, C. Tetanus and
botulinum-B neurotoxins block neurotransmitter release by
proteolytic cleavage of synaptobrevin. Nature 1992, 359, 832-
835.
(6) Humeau, Y.; Doussau, F.; Grant, N. J .; Poulain, B. How
botulinum and tetanus neurotoxins block neurotransmitter
release. Biochimie 2000, 82, 427-446.
(7) Schiavo, G.; Matteoli, M.; Montecucco, L. Neurotoxins affecting
neuroexocytosis. Physiol. Rev. 2000, 80, 717-766.
1
HPLC (60% B): tR ) 5.7 min. H NMR (DMSO-d6 + TFA) δ
2.7-3.7 (m, 8H), 4.1-4.8 (m, 4H), 4.6 (m, 2H), 6.8-8.8 (m,
27H). MS (ESI) (M + H)+ m/z ) 689.5.
4-{(2S)-2-Am in o-3-[1-(1-b en zylca r b a m oyl-(2S)-2-p h e-
n ylet h ylca r b a m oyl)-(2S)-2-(1-m et h yl-1H -in d ol-3-yl)et h -
ylca r ba m oyl]-3(S)-su lfa n ylp r op yl)}ben zoic a cid (28): 18
1
mg (22%). HPLC (50% B): tR ) 7.4 min. H NMR (DMSO-d6
+ TFA) δ 2.6-3.15 (m, 6H), 3.55 (m, 5H), 4.15-4.25 (m, 2H),
4.6 (m, 2H), 6.8-7.25 (m, 16H), 7.55 (d, 1H), 7.85 (d, 2H), 7.9
(m, 3H), 8.5 (m, 3H). MS (ESI) (M + H)+ m/z ) 692.6.
4-{(2S)-2-Am in o-3-[1-(1-b en zylca r b a m oyl-(2S)-2-p h e-
n yleth ylcar bam oyl)-(2S)-2-biph en yl-4-yl-eth ylcar bam oyl]-
3(S)-su lfa n ylp r op yl)}ben zoic a cid (29): 19 mg (22%). HPLC
1
(60% B): tR ) 5.5 min. H NMR (DMSO-d6 + TFA) δ 2.6-3.1
(m, 6H), 3.6 (m, 2H), 4.2 (d, 2H), 4.6 (m, 2H), 7.0-7.4 (m, 21H),
7.8 (d, 2H), 7.9 (m, 3H), 8.55 (m, 3H). MS (ESI) (M + H)+ m/z
) 715.9.
(8) Matthews, B. W. Structural basis of the action of thermolysin
and related zinc peptidases. Acc. Chem. Res. 1988, 21, 333-340.
(9) Roques, B. P.; Noble, F.; Dauge´, V.; Fournie´-Zaluski, M. C.;
Beaumont, A. Neutral endopeptidase 24.11. Structure, inhibi-
tion, and experimental and clinical pharmacology. Pharmacol.
Rev. 1993, 45 (1), 87-146.
4-{(2S)-2-Am in o-3-[1-(1-b en zylca r b a m oyl-(2S)-2-p h e-
n yleth ylcar bam oyl)-(2R)-2-biph en yl-4-yl-eth ylcar bam oyl]-
3(S)-su lfa n ylp r op yl)}ben zoic a cid (30): 26 mg (31%). HPLC
(gradient 40-100% B in 20 min): tR ) 12.6 min. 1H NMR
(DMSO-d6 + TFA) δ 2.85-3.45 (m, 6H), 3.55 (m, 1H), 3.7 (m,
1H), 4.25-4.35 (2dd, 2H),4.75 (m,1H), 4.85 (m, 1H), 7.1-7.5
(10) Hanson, M. A.; Stevens, R. C. Cocrystal structure of synapto-
brevin-II bound to botulinum neurotoxin type
B at 2.0 A
resolution. Nat. Struct. Biol. 2000, 7, 687-692.

4656 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 22
Anne et al.
(11) Arnon, S. S.; Schechter, R.; Inglesby, T. V.; Henderson, D. A.;
Bartlett, J . G.; Ascher, M. S.; Eitzen, E.; Fine, A. D.; Hauer, J .;
Layton, M.; Lillibridge, S.; Osterholm, M. T.; O’Toole, T.; Parker,
G.; Perl, T. M.; Russell, P. K.; Swerdlow, D. L.; Tonat, K.;
Working Group on Civilian Biodefense. Botulinum toxin as a
biological weapon: Medical and public health management.
J AMA 2001, 285, 1059-1070.
(28) Piquilloud, Y.; Reinharz, A. Roth, M. Studies of the angiotensin
converting enzyme with different substrates. Biochim. Biophys.
Acta 1970, 206, 136-142.
(29) Luciani, N.; de Rocquigny, H.; Turcaud, S.; Romieu, A.; Roques,
B. P. Highly sensitive and selective fluorescence assays for rapid
screening of endothelin-converting enzyme inhibitors. Biochem.
J . 2001, 356, 813-819.
(30) Goudreau, N.; Guis, C.; Soleilhac, J . M.; Roques, B. P. Dns-Gly-
(p-NO2)Phe-beta Ala, a specific fluorogenic substrate for neutral
endopeptidase 24.11. Anal. Biochem. 1994, 219, 87-95.
(31) Klein, A. W. Treatment of wrinkles with Botox. In Hyperhydrosis
and botulinum toxin in dermatology; Kreyden, O. P.; Bo¨ni, R.;
Burg, G., Eds.; Karger, Basel, 2002; pp 188-221.
(32) Lew, M. F.; Brashear, A.; Factor, S. The safety and efficacy of
botulinum toxin type B in the treatment of patients with cervical
dystonia: summary of three controlled clinical trials. Neurology
2000, 55, S29-S35.
(33) Adler, M.; Dinterman, R. E.; Wannemacher, R. W. Protection
by the heavy metal chelator N,N,N′,N′-tetrakis (2-pyridylmeth-
yl)ethylenediamine (TPEN) against the lethal action of botuli-
num neurotoxin A and B. Toxicon 1997, 35, 1089-1100.
(34) Foran, P.; Shone, C. C.; Dolly, J . O. Differences in the protease
activities of tetanus and botulinum B toxins revealed by the
cleavage of vesicle-associated membrane protein and various
sized fragments. Biochemistry 1994, 33, 15365-15374.
(35) Keller, J . E.; Nowakowski, J . L.; Filbert, M. G.; Adler, M. Rapid
microplate assay for monitoring botulinum neurotoxin B cata-
lytic activity. J . Appl. Toxicol. 1999, 19 Suppl 1, S13-17.
(36) Adler, M.; Nicholson, J . D.; Cornille, F.; Hackley, B. E., J r.
Efficacy of a novel metalloprotease inhibitor on botulinum
neurotoxin B activity [published erratum appears in FEBS Lett.
1998, 437, 164]. FEBS Lett. 1998, 429, 234-238.
(37) de Paiva, A.; Ashton, A. C.; Foran, P.; Schiavo, G.; Montecucco,
C.; Dolly, J . O. Botulinum A like type B and tetanus toxins fulfils
criteria for being a zinc-dependent protease. J . Neurochem. 1993,
61, 2338-2341.
(38) Shone, C. C.; Quinn, C. P.; Wait, R.; Hallis, B.; Fooks, S. G.;
Hambleton, P. Proteolytic cleavage of synthetic fragments of
vesicle-associated membrane protein, isoform-2 by botulinum
type B neurotoxin. Eur. J . Biochem. 1993, 217, 965-971.
(39) Adler, M.; Nicholson, J . D.; Starks, D. F.; Kane, C. T.; Cornille,
F.; Hackley, B. E. J r. Evaluation of phosphoramidon and three
synthetic phosphonates for inhibition of botulinum neurotoxin
B catalytic activity. J . Appl. Toxicol. 1999, 19 Suppl. 1, S5-
S11.
(12) Scientific and Therapeutic Aspects of Botulinum Toxin; Brin, M.
F., Hallett, M., J ankovic, J ., Eds.; Lippincott Williams
&
Wilkins: Philadelphia, 2002; pp 189-371.
(13) Borodic, G.; J ohnson, E.; Goodnough, M.; Schautz, E. Botulinum
toxin therapy, immunologic resistance and problems with avail-
able materials. Neurology 1996, 46, 26-29.
(14) So¨llner, T.; Bennett, M. K.; Whiteheart, S. W.; Scheller, R. H.;
Rothman, J . E. A protein assembly disassembly pathway in vitro
that may correspond to sequential steps of synaptic vesicle
docking, activation, and fusion. Cell 1993, 75, 409-418.
(15) Anne, C.; Cornille, F.; Lenoir, C.; Roques, B. P. High-throughput
assay determination of botulinum type B neurotoxin protease
activity. Anal. Biochem. 2001, 291, 253-261.
(16) Martin, L.; Cornille, F.; Turcaud, S.; Meudal, H.; Roques, B. P.;
Fournie´-Zaluski, M. C. Metallopeptidase inhibitors of tetanus
toxin: A combinatorial approach. J . Med. Chem. 1999, 42, 515-
525.
(17) Chauvel, E. N.; Llorens-Cortes, C.; Coric, P.; Wilk, S.; Roques,
B. P.; Fournie-Zaluski, M.-C. Differential inhibition of ami-
nopeptidase A and aminopeptidase N by new beta-amino thiols.
J . Med. Chem. 1994, 37, 2950-2957.
(18) Martin, L.; Cornille, F.; Coric, P.; Roques, B. P.; Fournie´-Zaluski,
M.-C. Beta-amino-thiols inhibit the zinc metallopeptidase activ-
ity of tetanus toxin light chain. J . Med. Chem. 1998, 41, 3450-
3460.
(19) Chauvel, E. N.; Coric, P.; Llorens Corte`s, C.; Wilk, S.; Roques,
B. P.; Fournie´-Zaluski, M. C. Investigation of the active site of
aminopeptidase A using a series of new thiol-containing inhibi-
tors. J . Med. Chem. 1994, 37, 1339-1346.
(20) David, C.; Bischoff, L.; Meudal, H.; Mothe, A.; DeMota, N.;
Danascimento, C.; Llorens-Cortes, C.; Fournie´-Zaluski, M. C.;
Roques, B. P. Investigation of subsite preferences in aminopep-
tidase A (EC 3.4.11.7) led to the design of the first highly potent
and selective inhibitors of this enzyme. J . Med. Chem. 1999, 42,
5197-5211.
(21) Fournie´-Zaluski, M. C.; Coric, P.; Turcaud, S.; Bruetschy, L.;
Lucas, E.; Noble, F.; Roques, B. P. Potent and systemically active
aminopeptidase N inhibitors designed from active-site investiga-
tion. J . Med. Chem. 1992, 35, 1259-1266.
(40) Hanson, M. A.; Oost, T. K.; Sukonpan, C.; Rich, D. H.; Stevens,
R. C. Structural basis fof BABIM inhibition of botulinum
neurotoxin type B protease. J . Am. Chem. Soc. 2000, 122,
11268-11269.
(41) Garcia, G. E.; Moorad, D. R.; Gordon, R. K. Buforin I, a natural
peptide, inhibits botulinum neurotoxin B activity in vitro. J .
Appl. Toxicol. 1999, 19 Suppl. 1, S19-22.
(42) Shone, C. C.; Roberts, A. K. Peptide substrate specificity and
properties of the zinc-endopeptidase activity of botulinum type
B neurotoxin. Eur. J . Biochem. 1994, 225, 263-270.
(43) Roques, B. P. Novel approaches to targeting neuropeptide
systems. Trends Pharmacol. Sci. 2000, 21, 475-483.
(44) Roques, B. P., Anne, C. Turcaud, S., Fournie´-Zaluski, M. C.,
Inhibiteurs de la toxine botuligne de type B, Patent Fr 0104895;
PCT Fr 0201258, 2001.
(22) Yaozhong, J .; Changyou, Z.; Skengde, W.; Daimo, L.; Yonan, M.;
Guilar, L. Solid-liquid-phase transfer catalytic synthesis of
R-amino acid via alkylation and nucleophilic addition of ben-
zaldehyde amines. Tetrahedron 1988, 44, 5343-5353.
(23) Oppolzer, W. Camphor as
a natural source of chirality in
asymmetric synthesis. Pure Appl. Chem. 1990, 62, 1241-1250.
(24) J osien, H.; Martin, A.; Chassaing, G. Asymmetric synthesis of
L-pyrenylalanine and L-9-fluorenylglycine via room-temperature
alkylations of a sultam-derived glycine imine. Tetrahedron Lett.
1991, 32, 6547-6550.
(25) Bischoff, L.; David, C.; Martin, L.; Meudal, H.; Roques, B. P.;
Fournie´-Zaluski, M. C. 2,4-dinitrophenyl-4-methoxybenzyl dis-
ulfide: A new efficient reagent for the electrophilic sulfenylation
of â-amino ester enolates. J . Org. Chem. 1997, 62, 4848-4850.
(26) Swaminathan S.; Eswaramoorthy S. Structural analysis of the
catalytic and binding sites of Clostridium botulinum neurotoxin
B. Nat. Struct. Biol. 2000, 7, 693-9.
(27) Schmidt, J . J .; Bostian, K. A. Endoproteinase activity of type A
botulinum neurotoxin: substrate requirements and activation
by serum albumin. J . Protein Chem. 1997, 16, 19-26.
(45) Eswaramoorthy, S.; Kumaran, D.; Swaminathan, S. A novel
mechanism for clostridium botulinum neurotoxin inhibition.
Biochemistry 2002, 41, 9795-9802.
J M0300680