Combinatorial Synthesis of a Small-Molecule Library Based on the Vinyl Sulfone Scaffold

Article abstract of DOI:10.1021/ol035774+

(Equation presented) A 30-member library of small molecules based on the vinyl sulfone scaffold was prepared on rink amide resin, using solid phase-based reactions such as oxidation and Horner-Wadsworth-Emmons reaction. The library was designed such that

Full text of DOI:10.1021/ol035774+

ORGANIC
LETTERS
2003
Vol. 5, No. 23
4437-4440
Combinatorial Synthesis of a
Small-Molecule Library Based on the
Vinyl Sulfone Scaffold
,†,‡
Gang Wang† and Shao Q. Yao*
Departments of Chemistry and Biological Sciences, National UniVersity of Singapore,
3 Science DriVe 3, Singapore 117543, Republic of Singapore
chmyaosq@nus.edu.sg
Received September 14, 2003
ABSTRACT
A 30-member library of small molecules based on the vinyl sulfone scaffold was prepared on rink amide resin, using solid phase-based
reactions such as oxidation and Horner−Wadsworth−Emmons reaction. The library was designed such that three points of diversity were
readily introduced in the library to accommodate the S ′, S , and S binding pockets of different cysteine proteases, making the strategy
1
1
2
suitable for high-throughput generation of potential cysteine protease inhibitors.
Cysteine proteases have drawn extensive attention from
pharmaceutical industries in recent years since many enzymes
belonging to this class are critically involved in major human
diseases, including diseases related to viral infections (i.e.,
SARS), strokes, cancer, Alzheimer’s disease, inflammation,
arthritis, and many others,1 making them excellent therapeutic
targets. Over the last few decades, many research groups
have developed chemical approaches capable of generating
diverse small-molecule inhibitors that target different classes
of cysteine proteases with various degrees of efficacy.2
Among the different classes of inhibitors, molecules contain-
ing different types of Michael acceptors are extremely potent
and highly specific toward cysteine proteases. They work
by irreversibly inactivating the catalytic cysteine residue
within the active site of the enzyme (Figure 1). Many of
these inhibitors, including R,â-unsaturated ketones, acryl-
amides, vinyl sulfones, etc., have been successfully synthe-
sized, and some have been tested in clinical experiments.2-4
With the emergence of microarray technologies and
proteomics in the past few years,5 peptidyl vinyl sulfones,
in particular, have been shown to be extremely useful as
(3) (a) Palmer, J. T.; Rasnick, D.; Klaus, J. L.; Bromme, D. J. Med.
Chem. 1995, 38, 3193-3196. (b) Bogyo, M.; McMaster, J. S.; Gaczynska,
M.; Tortorella, D.; Goldberg, A. L.; Ploegh, H. Proc. Natl. Acad. Sci. U.S.A.
1997, 94, 6629-6634. (c) Bogyo, M.; Shin, S.; McMaster, J. S.; Ploegh,
H. L. Chem. Biol. 1998, 5, 307-320. (d) Nazif, T.; Bogyo, M. Proc. Natl.
Acad. Sci. USA 2001, 98, 2967-2972. (e) Groll, M.; Nazif, T.; Huber, R.;
Bogyo, M. Chem. Biol. 2002, 9, 655-662.
(4) (a) Wang, G.; Uttamchandani, M.; Chen, G. Y. J.; Yao, S. Q. Org.
Lett. 2003, 5, 737-740. (b) Overkleeft, H. S.; Bos, P. R.; Hekking, B. G.;
Gordon, E. J.; Ploegh, H. L.; Kessler, B. M. Tetrahedron Lett. 2000, 41,
6005-6009.
(5) Chen, G. Y. J.; Uttamchandani, M.; Lue, Y. P. R.; Lesaicherre, M.
L.; Yao, S. Q. Curr. Top. Med. Chem. 2003, 3, 705-724.
† Department of Chemistry.
‡ Department of Biological Sciences.
(1) (a) Otto, H. H.; Schirmeister, T. Chem. ReV. 1997, 97, 133-171. (b)
McKerrow, J. H.; James, M. N. G. Perspect. Drug Discuss. Des. 1996, 6,
1-125. (c) Anand, K.; Ziebuhr, J.; Wadhwani, P.; Mesters, J. R.; Hilgenfeld,
R. Science 2003, 300, 1763-1767.
(2) (a) Powers, J. C.; Asgian, J. L.; Ekici, O. D.; James, K. E. Chem.
ReV. 2002, 102, 4639-4750. (b) Maly, D. J.; Huang, L.; Ellman, J. A.
ChemBiochem 2002, 3, 16-37. (c) Hernandez, A. A.; Roush, W. R. Curr.
Opin. Chem. Biol. 2002, 6, 459-465.
10.1021/ol035774+ CCC: $25.00 © 2003 American Chemical Society
Published on Web 10/18/2003

Scheme 1. Solid-Phase Synthesis of Vinyl Sulfonesa
Figure 1. Binding of a vinyl sulfone to the active site of a cysteine
protease.
activity-based probes for high-throughput profiling of cys-
teine proteases in both gel- and microarray-based experi-
ments,4,6 largely due to their negligible cross-reactivity
toward other classes of enzymes.2b Under physiological
conditions, the electrophilic, vinyl sulfone moiety of the
inhibitor is not reactive toward most nucleophilic elements
present in a biological species (i.e., amine and thiol groups
in a protein).3a However, upon binding to the active site of
an enzyme having a catalytic cysteine residue, i.e., a cysteine
protease, the vinyl sulfone would react, at a highly specific
rate, with the thiol in the cysteine residue, which results in
the formation of a covalent enzyme-inhibitor adduct, leading
to subsequent irreversible inactivation of the enzyme.7
A number of strategies have been developed for the
chemical synthesis of both peptide- and nonpeptide-based
vinyl sulfone libraries.3,4 Traditionally, peptide vinyl sulfones
were prepared by methods based on conventional solution-
phase peptide synthesis.3a Recently, a number of solid-phase
strategies have been developed.3b-e,4 Overkleeft et al. first
synthesized the N-terminal peptide fragment of the vinyl
sulfone on a solid support, followed by nucleophilic cleavage/
ligation using a desired vinyl sulfone-containing, C-terminal
amino acid.4b Bogyo et al. reported solid-phase methods to
generate positional-scanning combinatorial libraries of pep-
tide vinyl sulfones, with variations at the P2-P4 positions
of the library,3b-e by attaching a vinyl sulfone-containing
aspartic acid at the P1 position onto a rink amide resin via
its side-chain carboxylic acid (Figure 1). However, this
strategy is limited only to synthesis of peptide vinyl sulfones
having carboxyl side chains at the P1 positions (e.g., Asp
and Glu). In our own work,4a we recently reported a solid-
phase strategy for the synthesis of peptide vinyl sulfones
having different P1 residues. By anchoring the vinyl sulfone-
derivatives onto 2-chlorotrityl resin via the phenolic alcohol
moiety at the P1′ position, we were able to install diversities
at any given P position of the inhibitor (i.e., P1, P2, etc.).
None of these methods, however, allows the facile solid-
phase synthesis of vinyl sulfone libraries having variations
at the P1′ position, which is known to be critical for binding
to the S1′ site of the enzyme active site. Herein, we report
the first strategy for solid-phase synthesis of vinyl sulfone
a Reagents and conditions: (a) Cs2CO3 (2.1 equiv) or DBU (2.1
equiv), DMF, 6-12 h at room temperature, 60-85%. (b) Rink
amide resin (0.7 mmol/g), DIC (3.3 equiv), HOBt (3.3 equiv), DIEA
(6 equiv), DMF (10 mL/g resin), 6 h at room temperature, 0.65-
0.7 mmol/g loading. (c) m-CPBA (3.5 equiv), DCM (10 mL/g
resin), 1-1.5 h at room temperature, 95-100%. (d) LHMDS (5
equiv; 1.0 M in THF), 30 min with resin, washed with THF, then
Fmoc-AA-CHO (4 equiv) in THF (20 mL/g resin), 2 h at room
temperature, 0.2-0.4 mmol/g loading. (e) DBU/HOBt/DMF (2:1:
97, 10 mL/g resin), 10 min at room temperature. (f) 7 (8 equiv),
DIC (7.5 equiv), HOBt (7.5 equiv), DIEA (8.5 equiv), DMF, 6 h
at room temperature. (g) TFA/DCM/H2O (50:45:5, 15 mL/g resin),
precipitate in cold ether.
molecules, which allows easy introduction of diversities at
not only the P positions (e.g., P1 and P2) but also the P1′
position of the inhibitor. As proof-of-concept experiments,
we have successfully applied this strategy to the synthesis
of a 30-member small-molecule-based (i.e., MW < 500)
vinyl sulfone library that installs three, five, and two different
variables at P1′, P1, and P2 positions of the inhibitor,
respectively (Scheme 1).
As shown in Scheme 1, steps in our synthesis include (a)
solution-phase nucleophilic reaction between a commercially
available thiol-containing acid, 1, and diethyl iodomethyl
phosphonate to generate the sulfide phosphonate 2; (b)
(6) Chen, G. Y. J.; Uttamchandani, M.; Zhu, Q.; Wang, G.; Yao, S. Q.
ChemBiochem 2003, 4, 336-339.
(7) (a) Thompson, S. A.; Andrews, P. R.; Hanzlik, R. P. J. Med. Chem.
1986, 29, 104-111. (b) Liu, S.; Hanzlik, R. P. J. Med. Chem. 1992, 35,
1067-1075.
4438
Org. Lett., Vol. 5, No. 23, 2003

loading of 2 onto rink amide resin to give 3; (c) solid-phase
oxidation of 3 to give sulfone 4; (d) solid-phase Horner-
Emmons reaction with an amino acid-derived, Fmoc-
protected aldehyde to give 6; and (e) deprotection of Fmoc
group under optimized conditions followed by (f) acylation
and (g) TFA cleavage to release the desired vinyl sulfone 8
in solution. In contrast to what we previously reported,4a our
current approach takes advantage of the successful imple-
mentation of both m-CPBA oxidation of sulfide 2 and its
subsequent Horner-Emmons reaction on solid support with
high yield and purity, thus making it possible to install three
points of diversity (i.e., P1′, P1, and P2) within the vinyl
sulfone scaffold.
Table 1. Horner-Emmons Reaction from 4 to 6 under
Different Conditions
base
LDA
LDA
LHMDS 30 min, 25 °C
DBU 90 min, 25 °C
KHMDS 60 min, 25 °C
NaH 60 min, 25 °C
conditionsa
equiv % conversionb 4 f 6 % purityb
50 min, -78 °C
30 min, 0 °C
6
4
5
8
5
5
30%
100%
100%
<10%
65%
ndc
64%
>95%
ndc
ndc
45%
38%
a LHMDS and LDA were 1 and 2 M solutions, respectively, in THF.
Time indicated is the incubation time of resin 4 and the base, before addition
of Fmoc-AA-CHO. b Determined by HPLC, including both epimers (see
main text for discussion). All conditions shown in Table 1 were optimized
to maintain epimerization to <10%. Other conditions caused substantial
epimerization (data not shown). c nd ) not determined.
Starting from three mercaptoacids (1a-c), the nucleophilic
reaction proceeded smoothly with diethyl iodomethyl phos-
phonate in the presence of a suitable base (i.e., DBU or NaH
for 1a and 1c, Cs2CO3 for 1b, respectively) to give 2a-c in
good yields (60-85%). When aliphatic, 11-mercapto un-
decanoic acid 1b was used with bases such as DBU, NaH,
or NaOEt, the major product isolated upon workup was the
disulfide, rather than the desired product 2b. This problem
was overcome by using Cs2CO3 as the base in the reaction,
producing the desired product in high yield (85%). The
sulfide phosphonates 2a-c were conveniently loaded onto
the rink amide resin using the standard DIC/HOBT/DIEA
coupling procedures to afford the resin-bound sulfide phos-
phonates 3a-c with high substitution levels (0.65-0.7 mmol/
g). It should be noted that, while we prefer to perform step
a in solution to generate 2a-c from 1a-c, which is
commercially available and requires no further protection,
one could alternatively (i) block the thiols in 1a-c with a
suitable protecting group, (ii) load them on the solid support,
(iii) deprotect the thiol groups, and (iv) react with diethyl
iodomethyl phosphonate on solid phase to give 3a-c (data
not shown). All subsequent steps (steps c-g in Scheme 1)
were performed on the solid support. To optimize our
synthetic conditions, a small portion of the resin following
each step of synthesis was cleaved, and the product was
analyzed directly by LC-MS to assess its % conversion
(judged by disappearance of the starting material) as well
as % purity (judged by the product HPLC profile; see
Supporting Information for details). Oxidation of 3a-c was
accomplished by treatments of the resin with m-CPBA (5
equiv) for 1-1.5 h, to give 4a-c in nearly quantitative yields
(95-100%). Shorter oxidation times gave rise to the forma-
tion of the partially oxidized products, the sulfoxide phos-
phonates. Next, the resin-bound sulfone phosphonates, 4a-
c, were used as the precursors in the Horner-Emmons
reaction. Horner-Emmons reaction is a valuable carbon-
carbon bond-forming reaction that has been successfully
applied in the solid-phase synthesis of olefins and some
peptidyl Michael acceptors.8 In our strategy, the reaction was
used to generate the critical vinyl sulfone scaffold and at
the same time introduce the P1 diversity in the library. Most
published Horner-Emmons protocols require the presence
of an excess of a strong base in order for the reaction to go
to completion. They are therefore not suitable for our
synthesis because the presence of an excess of a strong base
may (1) cause the Fmoc group in the aldehyde to be cleaved
off prematurely, (2) lead to severe epimerization at the
aldehyde chiral center, and (3) lead to potential epimerization
in the vinyl sulfone product (i.e., 6 in Scheme 1).9 Conse-
quently, it is necessary to optimize the Horner-Emmons
conditions in our strategy such that the conversion of 4 to 6
is highly quantitative, while maintaining premature Fmoc
cleavage, as well as minimal epimerization in the reaction
(Table 1).
We accomplished this by adopting a two-step procedure,
which involves treatments of the resin-bound sulfone phos-
phonates, 4a-c, first with an excessively strong base (i.e.,
LHMDS) and then, upon removal of the base, with a Fmoc-
protected amino acid aldehyde 5 (prepared on the basis of
procedures previously described4a) under Horner-Emmons
conditions. Typically, the reaction was performed under
anhydrous conditions at room temperature. Resin 4 was
treated with 1 M LHMDS solution (in THF) for 30 min and
then washed with excess anhydrous THF under N2. The
Fmoc-AA-aldehyde solution (in THF) was subsequently
added to the resin, and the reaction was further incubated
for 1-2 h. To accurately determine the enantiomeric purity
of the product generated from the Horner-Emmons reaction,
upon completion of the reaction (step d in Scheme 1), the
product, 6, was first deprotected and then coupled to an
optically pure amino acid. The resulting dipeptide was then
cleaved off the resin, and analyzed directly by HPLC. The
extent of epimerization in 6, as well as its purity under
different reaction conditions, was estimated from the resulting
HPLC profile, by comparing the ratio of two separable peaks
arising from the two diastereomeric dipeptides in the product
mixture (as a result of epimerization in the Horner-Emmons
(9) Epimerization of 6 could possibly occur in two ways. One, the chiral
center in the starting material, amino aldehyde 5, may be directly epimerized,
leading to generation of both isomers in product 6. Alternatively, decon-
jugation of the vinyl sulfone in product 6 to generate an allylic sulfone,
followed by reconjugation, could also lead to direct epimerization of 6. In
our studies here, we cannot rule out either possibility.
(8) (a) Salvino, J. M.; Kiesow, T. J.; Darnbrough, S.; Labaudiniere, R.
J. Comb. Chem. 1999, 1, 134-139. (b) Wipf, P.; Henninger, T. C. J. Org.
Chem. 1997, 62, 1586-1590. (c) Sammelson, R. E.; Kurth, M. J. Chem.
ReV. 2001, 101, 137-202. (d) Caulfield, T. J.; Patel, S.; Salvino, J. M.;
Liester, L.; Labaudiniere, R. J. Comb. Chem. 2000, 2, 600-603.
Org. Lett., Vol. 5, No. 23, 2003
4439

reaction; see Supporting Information). As shown in Table
1, five different bases (e.g., DBU, LHMDS, KHMDS, LDA,
NaH) were tried with different reaction times and temper-
atures. Only LHMDS under optimized reaction conditions
(i.e., 5 equiv, 30 min, 25 °C) produced the best result, giving
rise to quantitative conversion from 4 to 6 (100% conversion)
and at the same time with nearly quantitative yield (>95%
including both epimers). The degree of epimerization in 6
under these conditions was determined to be <10% (Sup-
porting Information). Consequently, all of our subsequent
work was carried out using the optimized conditions, unless
otherwise specified. Different Fmoc-AA-aldehydes, 5a-e,
have been used together with different resins, 4a-c, to
generate the resin-bound vinyl sulfones, 6, with high yields
and purities and minimal epimerization.10 We also observed
only the trans isomers in all Horner-Emmons products 6.
After the Horner-Emmons reaction, we continued to steps
e-g, by deprotecting the Fmoc group in 6, acylating the
resulting free amine with different acids, 7a,b, followed by
TFA cleavage to generate the final products, 8. It was found
that, while the standard Fmoc cleavage conditions (20%
piperidine in DMF for 30 min) worked well in step e for
reactions involving resin 4b, they failed with resins 4b,c,
generating mostly side products, which were subsequently
determined to be adducts derived from the addition of
piperidine to the double bond of the vinyl sulfone (Supporting
Information). We reasoned that this may be due to the
electron-withdrawing nature of the aromatic substituents at
P1′ position in the two resins, which increase the reactivity
in the vinyl sulfone moiety, making it more susceptible to
attacks from nucleophiles such as piperidine. This is con-
sistent with previous observations,11 in which different
derivatives of vinyl sulfones having varying P1′ substituents
had different degrees of reactivity toward base-catalyzed
additions of thiol compounds. We therefore modified the
Fmoc cleavage conditions. By using a solution containing
2% DBU and 1% HOBt in DMF,12 we were able to
efficiently cleave Fmoc from different resins in 10 min while
maintaining the integrity of the vinyl sulfone in the products.
Upon Fmoc cleavage, the free amine was further acylated
with either anisic acid (7a) or isonicotinic acid (7b) using
standard DIC/HOBT/DIEA coupling methods. This installed
the P2 position in our vinyl sulfone library. The resulting
final products were subsequently deprotected, cleaved off
the resin using TFA/DCM/H2O (5/4.5/0.5), and precipitated
in cold ether. In all, a total of 30 different vinyl sulfones
were prepared. The purity and identity of each product was
assessed by direct analysis of the precipitate using LC-MS.
All desired products were unambiguously confirmed and
determined to have varying degrees of purity (40-95%; see
Supporting Information). For products generated using resins
3a and 3b, the purity of the desired products ranges from
70 to 95%, which may be used directly for subsequent
screenings without any further purification. However, for
those vinyl sulfones generated from resin 3c, the purity is
typically between 40 and 70%. This low purity in products
generated from resin 3c may be due to the enhanced
reactivity in the vinyl sulfone moiety in 3c, as mentioned
earlier, which could have caused side reactions in our
synthesis.
In summary, we have developed the first strategy for solid-
phase synthesis of a small-molecule library based on the vinyl
sulfone scaffold, which allows facile installation of diversities
not only at the P positions but, more importantly, at the P1′
position. This strategy utilizes multiple steps of solid-phase
reactions, facilitating the easy preparation of potentially large
arrays of small molecules that contain the important vinyl
sulfone pharmacophore. Given the enormous interest in
developing various protease inhibitors, this strategy may find
broad applications in the studies of cysteine proteases.
Acknowledgment. Funding support was provided by the
National University of Singapore (NUS) and the Agency for
Science, Technology and Research (A*STAR) of Singapore.
S.Q.Y. is the recipient of the 2002 Young Investigator Award
(YIA) from the Biomedical Research Council (BMRC).
(10) Typical yields were ∼90-95%, with high purity of the desired
products (90%) and minimal epimerization (<10%).
(11) Reddick, J. J.; Cheng, J.; Roush, W. R. Org. Lett. 2003, 5, 1967-
1970.
(12) Novabiochem 2000 product catalog; Novabiochem: Laufelfingen,
Switzerland.
Supporting Information Available: Experimental details
and characterizations of compounds. This material is avail-
able free of charge via the Internet at http://pubs.acs.org.
OL035774+
4440
Org. Lett., Vol. 5, No. 23, 2003