Dipeptide vinyl sultams: Synthesis via the Wittig-Horner reaction and activity against papain, falcipain-2 and Plasmodium falciparum

Article abstract of DOI:10.1016/j.bmcl.2006.04.079

The synthesis of phosphonate derivatives of N-phenyl- and N-benzyl-γ- and δ-sultams, and their application in the Wittig-Horner reaction with N-Boc-l-phenylalanine aldehyde to afford E- and Z-isomers, are described. These compounds were further processed to provide five dipeptide vinyl sultams, which were found to be inactive against papain at concentrations up to 50 μM. In contrast, vinyl sultams demonstrated weak activity against recombinant falcipain-2 and Plasmodium falciparum W2.

Full text of DOI:10.1016/j.bmcl.2006.04.079

Bioorganic & Medicinal Chemistry Letters 16 (2006) 4115–4119
Dipeptide vinyl sultams: Synthesis via the Wittig–Horner reaction
and activity against papain, falcipain-2 and Plasmodium falciparum
a
a
a,
b,
*
*
´
Claudia Valente, Rita C. Guedes, Rui Moreira, Jim Iley,
Jiri Gutc and Philip J. Rosenthalc
aCECF, Faculty of Pharmacy, University of Lisbon, Avenida das Forc¸as Armadas, Lisboa 1600-083, Portugal
bChemistry Department, The Open University, Milton Keynes MK7 6AA, UK
cDepartment of Medicine, San Francisco General Hospital, University of California, San Francisco, CA 94143-0811, USA
Received 13 March 2006; revised 25 April 2006; accepted 26 April 2006
Available online 12 May 2006
Abstract—The synthesis of phosphonate derivatives of N-phenyl- and N-benzyl-c- and d-sultams, and their application in the
Wittig-Horner reaction with N-Boc-L-phenylalanine aldehyde to afford E- and Z-isomers, are described. These compounds were
further processed to provide five dipeptide vinyl sultams, which were found to be inactive against papain at concentrations up to
50 lM. In contrast, vinyl sultams demonstrated weak activity against recombinant falcipain-2 and Plasmodium falciparum W2.
Ó 2006 Elsevier Ltd. All rights reserved.
Malaria is the major life-threatening parasitic disease in
tropical and sub-tropical regions. Worldwide, there are
at least 300 million acute cases of malaria and more than
1 million deaths each year, mostly young children infect-
ed with Plasmodium falciparum.1 With the rapid spread
of multidrug-resistant P. falciparum strains, the develop-
ment of safe and effective antimalarials has become an
important strategy towards achieving effective control
of malaria. Cysteine proteases regulate a broad spec-
trum of physiological functions in mammals, parasitic
protozoa, plants and yeast. In humans, elevated levels
of these enzymes can lead to disease states such as oste-
oporosis, rheumatoid arthritis and cancer.2,3 In para-
sites, they play a crucial role in metabolism and
reproductive function.2,3 Falcipain-2 is a cysteine prote-
ase stored in the acidic food vacuole of P. falciparum
and is likely involved in the hydrolysis of haemoglobin
that produces free amino acids required for parasite sur-
vival.4–6 Disruption of this haemoglobin degradation
pathway is lethal to the parasite and thus compounds
designed to inhibit falcipain-2 present excellent opportu-
nities for developing antimalarial drug candidates.
Vinyl sulfones (VS) and their analogues, such as sulfon-
amides and sulfonates, have been reported as a promis-
ing class of inhibitors for parasitic cysteine proteases.
They act by irreversibly alkylating the active site cys-
teine residue via conjugate addition.2,3,7
Developing conformationally restricted analogues is a
commonly used approach to probe the binding site of
receptors and to further improve activity.8 Despite the
recent finding that cyclic vinyl sulfones derived from
the Bsmoc amino-protecting group are irreversible
inhibitors of papain and cathepsin B,9 the design of con-
formationally restricted analogues of sulfones as cys-
teine protease inhibitors is still an underexplored field.
In this context, we now report the synthesis of the vinyl
sultam scaffold as a cyclic isostere of vinyl sulfonamides,
and thus as a building block for a new family of irrevers-
ible cysteine protease inhibitors. Vinyl sultams contain-
ing an exocyclic double bond are also structurally
similar to a-methylene lactones, which are known Mi-
chael acceptors for thiols.10,11 With the aim of designing
effective inhibitors for falcipain-2, the vinyl sultams were
attached to a peptide sequence. The chosen sequence
was Mu-Leu-Phe (Mu = 4-morpholinecarbonyl), by
analogy with the vinyl sulfone Mu-Leu-Hph-VS-Ph,
which was considered an optimal inhibitor for this en-
zyme (IC50 0.003 lM).2 A vinyl sultam with the Ac-
Phe-Phe moiety was also prepared based on the success
of the vinyl sulfones Cbz-Phe-Hph-VS-R against
cruzain, a related cysteine protease from Trypanosoma
Keywords: Wittig–Horner reaction; Vinyl sultams; Papain;
Antimalarials; Cysteine proteases; Falcipain-2; Plasmodium falciparum;
Molecular modelling.
*
Corresponding authors. Tel.: +351 217946477; fax: +351
217946470; e-mail: rmoreira@ff.ul.pt
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.04.079

4116
C. Valente et al. / Bioorg. Med. Chem. Lett. 16 (2006) 4115–4119
cruzi.2 The sultam nitrogen atom carried either a benzyl
or phenyl group. The evaluation of the target com-
pounds against falcipain-2, papain and the chloro-
quine-resistant P. falciparum strain W2 is also reported.
conditions B and C were investigated. In general, we
found LDA at À78 °C to afford higher yields of the sul-
tam phosphonates than BuLi either at À78 °C or room
temperature. Phosphonates 3 and 4 were condensed
with N-Boc-L-phenylalanine aldehyde (5), prepared
from N-Boc-L-phenylalanine via formation of the N,O-
dimethylhydroxamate.19 The reaction provided E- and
Z-vinyl sultams 6–9 (Table 2). The E/Z ratio is consis-
tent with the general observation that phosphonates
with alkyl groups give mostly the thermodynamically
more stable E-olefins.20,21
Synthesis involved the preparation of the c- and d-sul-
tam cores 1 and 2 (Schemes 1 and 2), using appropriate
modifications of previously described methods.12,13
Incorporation of the peptide moiety proceeded accord-
ing to the general sequence reported for the analogous
vinyl sulfones,14 which involve as the key step a
Wittig-Horner reaction (Scheme 3). This method has
been widely applied to the synthesis of a,b-unsaturated
sulfonate derivatives,6,14,15 but to the best of our knowl-
edge, not to sultams; reported synthesis of vinyl deriva-
tives of the latter involves: (i) the direct coupling by an
aldol-like reaction, followed by dehydration16 and (ii)
the intramolecular Heck reaction.17 In order to follow
the Wittig-Horner approach, the anion of 2a was gener-
ated (2 equiv n-BuLi, dry THF, rt, N2) and allowed to
react with diethyl chlorophosphate to afford compound
4a (method A).15,18 However, when this procedure was
applied to the other sultams it furnished either decom-
position or poor yields (Table 1). For these reasons,
The geometry of the double bond comes from the exam-
ination of the chemical shifts of vinyl and methine pro-
tons (Table 3). This process has been reported as a
reliable criterion for stereochemical assignment of a,b-
unsaturated sulfonates:15 as a result of the deshielding
effect of the sulfonyl group, the vinyl proton resonance
Table 1. Conditions for the synthesis of sultam phosphonates 3 and 4
Phosphonate
Yield (%)
Method B
Method A
Method C
3a
3b
4a
4b
27
0
56
—
—
—
65
43
45
34
48
15
(i)
(ii)
Cl(CH2)3SO2Cl
Cl(CH2)3SO2NHR
SO2
N
R
Method A: n-BuLi, dry THF, rt, N2; method B: n-BuLi, dry THF,
À78 °C, N2; method C: LDA, dry THF, À78 °C, N2.
1a R = Ph, 49% (2 steps)
1b R = Bz, 79% (2 steps)
Table 2. Yields and ratio of E/Z isomers
Scheme 1. Reagents and conditions: (i) PhNH2, TEA, DCM, reflux,
1 h or BzNH2, DCM, À10 °C, 1 h; (ii) NaOH, TEA, MeOH, reflux.
Phosphonate
Product
E/Z
Yield (%)
E/Z ratio
3a
6a
6b
E
Z
60
16
79:21
(i,ii,iii)
3b
4a
4b
7a
7b
E
Z
58
29
67:33
90:10
86:14
SO2
SO2
O
N
R
8a
8b
E
Z
65
7
2a R = Ph, 43% (3 steps)
2b R = Bz, 44% (3 steps)
9a
9b
E
54
9
Scheme 2. Reagents and conditions: (i) RNH2, 100 °C; (ii) POCl3,
reflux; (iii) K2CO3, AcOEt, rt.
Z
(EtO)2(O)P
BocNH
BocNH
Ph
(
)
n
(
)
n
SO2
(ii)
(i)
1 or 2
O2S
+
R
N
O2S
N
R
(
)
n
N
Ph
R
n=1: 6b (Z) R = Ph; 7b (Z) R = Bz
n=2: 8b (Z) R = Ph; 9b (Z) R = Bz
n=1: 3a R = Ph; 3b R = Bz
n=2: 4a R = Ph; 4b R = Bz
n=1: 6a (E) R = Ph; 7a (E) R = Bz
n=2: 8a (E) R = Ph; 9a (E) R = Bz
(iii)
73-84%
H2N
H2N
SO2
(
)
n
+
N
R
O2S
(
)
n
N
Ph
Ph
R
n=1: 10b (Z) R = Ph; 11b (Z) R = Bz
n=2: 12b (Z) R = Ph; 13b (Z) R = Bz
n=1: 10a (E) R = Ph; 11a (E) R = Bz
n=2: 12a (E) R = Ph; 13a (E) R = Bz
Scheme 3. Reagents and conditions: (i) LDA, THF, À78 °C, (EtO)2POCl, N2; (ii) n-BuLi, THF, Boc-Phe-CHO (5), À78 °C to rt, N2; (iii) TFA 50%
(w/v DCM solution), rt, 20 min.

C. Valente et al. / Bioorg. Med. Chem. Lett. 16 (2006) 4115–4119
4117
Table 3. Chemical shifts d (parts per million) and coupling constants J
(hertz) in CDCl3 for vinyl and methine protons of E/Z isomers
Table 4. Biological activity of dipeptide vinyl sultams
Papaina
IC50 (lM)
c
Compound
Compound
E/Z
d CH@
d CHCH2Ph
[I] lM % of inhibition rec-FP-2 P. falciparum W2
6a
6b
7a
7b
8a
8b
9a
9b
E
Z
E
Z
E
Z
E
Z
6.31 br d (8.8)
6.35 br s
4.53 br s
4.91 m
4.48 m
9a
14
50
50
—
—
50
50
—
NI
NIb
—
—
—
26.5
48.4
31.4
13.7
23.6
>10
>10
>10
6.29 br d (9.2)
6.26 br s
21a
22a
23a
23b
24d
4.96 m
4.71 m
—
6.24 br d (8.8)
6.12 br s
NI
NI
—
7.8
>10
0.2f
5.10 br s
4.68–4.74 m
5.17 br s
6.21 d (9.6)
5.98 br s
0.003e
a Assays with papain using the incubation method.24
b No inhibition after 30 min of incubation.
c Assays of falcipain inhibition and parasite development were deter-
mined as described earlier.6
is found downfield in the E-isomer compared to the
Z-isomer, whereas methine proton signal appears up-
field. Stereochemistry is further supported by means of
NOESY spectra, on the basis of a NOE cross-peak
between the vinyl and @CCH2 protons in the Z-isomer
9b.
d Compound 24 = Mu-Leu-hPh-VSPh.
e Non-recombinant falcipain.25
f Itg2 strain.25
Overall, what is immediately clear from these results is
that dipeptide vinyl sultams are significantly less active
(ꢀ10À4-fold) than their acyclic analogues, vinyl sulf-
ones. This result was totally against expectation, espe-
cially given a recent report which shows that bicyclic
a-methylene-c-sultams are reactive towards sulfur nucle-
ophiles.17 A possible explanation is that the a-carbon of
Michael acceptors 14 and 21–23 incorporates the ring
moiety, interfering with enzyme binding and leading to
a substantial reduction in the activity when compared
to their acyclic counterparts, in which the a-carbon is
not substituted.23 To get some insight into the binding
mode of vinyl sultams to cysteine proteases, we studied
the molecular interactions between the more potent
inhibitors 23a,b and cruzain, using the program
GOLD.26 Falcipain-2 and cruzain share more than
40% homology in the mature domain and about 90%
for the binding site residues.27,28 We chose a cruzain
structure co-crystallized with vinyl sulfone CbzPhe-
hPhe-SO2CH2Ph, 25, via a covalent bond (PDB code:
1F2A). Each ligand was energy-minimized and then
To synthesize the dipeptide vinyl sulfones (14 and 21–
23), the N-Boc group of 6–9 was easily cleaved with tri-
fluoroacetic acid and the resulting amines were coupled
with Boc-protected amino acids under standard peptide
coupling conditions (Scheme 4).22 These compounds
were screened for effectiveness against papain, the arche-
type of cysteine proteases. No inhibition could be detect-
ed at concentrations up to 50 lM. However, when
evaluated against falcipain-2, they showed some activity,
albeit modest (Table 4). Compound 23a exhibited the
best IC50 value (13.7 lM), which was consistently
accompanied by the most favourable activity against
P. falciparum W2. Though the range of compounds is
limited, the data presented suggest that the ring size
(21a vs 23a), the geometry of the double bond (23a vs
23b) and the nitrogen substituent (22a vs 23a) may have
some impact on activity. As shown by the data for com-
pounds 14 and 22a substitution of the AcPhe by the Mu-
Leu moiety has little effect on activity.
Ph
H
O
O
H
H
H2N
H
S
Ph
N
SO2
N
(i)
(ii)
SO2
N
N
AcNH
BocNH
N
R
13%
R
54-70%
(
)
O
n
O
Ph
Ph
14 (E)
(
)
n
Ph
10-13
n = 1, R = Bz, 15a (E), 15b (Z)
n = 2, R = Ph, 16a (E), 16b (Z)
n = 2, R = Bz, 17a (E), 17b (Z)
(iii)
56-88%
O
H
H
H
H
N
N
SO2
SO2
N
N
(iv)
H2N
H
N
R
N
R
15-50%
O
O
O
(
)
n
(
)
n
Ph
Ph
n = 1, R = Bz, 21a (E)
n = 2, R = Ph, 22a (E)
n = 2, R = Bz, 23a (E), 23b (Z)
n = 1, R = Bz, 18a (E), 18b (Z)
n = 2, R = Ph, 19a (E), 19b (Z)
n = 2, R = Bz, 20a (E), 20b (Z)
Scheme 4. Reagents and conditions: (i) 12a, N-acetyl-L-phenylalanine, TEA, HOBt, DCC, DCM, rt, three days; (ii) N-Boc-L-leucine-N-
hydroxysuccimide ester, TEA, THF, rt, 14–48 h; (iii) TFA 50% (w/v DCM solution), rt, 20 min; (iv) 4-morpholinecarbonyl chloride, TEA, THF, rt,
14–16 h.

4118
C. Valente et al. / Bioorg. Med. Chem. Lett. 16 (2006) 4115–4119
subjected to 50 docking runs. The top 10 solutions (i.e.,
those with the highest fitness score) were visually ana-
lyzed for: (i) the hydrophobic and hydrophilic interac-
tions between the ligand and enzyme surfaces and (ii)
the distance between the b-carbon atom of the Michael
acceptor system and the catalytic cysteine residue. Mod-
elling the interaction of vinyl sulfone 25 with cruzain re-
vealed that seven conformations presented hydrogen
bonding interactions between the sulfonyl oxygen atoms
and the imidazole NH of His159, indole NH of Trp177
and one of the NH of Gln19, while all 10 inspected con-
formations present the Phe side chain sitting in the S2
pocket of the enzyme. Additional hydrogen bonding to
the carbonyl oxygen atom of Gly66 enables further sta-
bilization of the non-covalent complex. This conforma-
tion presents the b-carbon atom of the vinyl sulfone
˚
system in proximity (3.6–3.8 A) to the sulfur atom of
Cys25, as would be expected for Michael addition
chemistry.
Interestingly, docking E-isomer 23a to cruzain revealed
that the top ranking conformation sits in the active site
with an inverted orientation when compared to 25, the
N-benzyl-d-sultam moiety being accommodated in the
S2 subsite through stabilizing van der Waals interac-
tions with Leu67, Ala133 and Leu157 (Fig. 1A). The
most striking observation was that the sulfur atom of
˚
Cys25 is in the vicinity (3.4 A) of the b-carbon and posi-
tioned 20° out off of the plane defined by the carbon–
carbon double bond, thereby disfavouring Michael
addition of cysteine to the b-carbon atom of the vinyl
sultam system. This type of reaction requires the sulfur
nucleophile to approach the carbon atom almost per-
pendicularly to the double bond plane in order to
achieve optimal orbital overlap. In contrast, the pre-
ferred conformation of the Z-isomer 23b is positioned
in the active site of cruzain similarly to 25, that is, with
the Leu side chain occupying the S2 pocket (Fig. 1B).
However, the Cys25(S)-Cb distance is probably too
Figure 1. GOLD-generated active site interaction of cruzain and 23a
(A) and 23b (B); Cys25 and the S2 pocket residues are in magenta.
Figure generated using Pymol software: DeLano, W.L. The PyMOL
Molecular Graphics System (2002) DeLano Scientific, San Carlos, CA,
USA. http://www.pymol.org.
falcipain-2 over papain. Docking studies suggest that
two different binding modes of vinyl sultams to cysteine
proteases are available, neither of which is favourable
for the conjugate addition of Cys25 to the vinyl sultam
system.
˚
long (ca. 4.0 A) for nucleophilic attack of cysteine to
take place. Taken together, these docking results suggest
that the d-sultam ring increases the degrees of freedom
of ligand binding to cruzain when compared with the
acyclic vinyl sulfone analogue, leading to enzyme–inhib-
itor complexes that do not favour Michael addition
chemistry.
Acknowledgments
We thank FCT, POCTI, FEDER for a Grant to C.V.
(SFRH/BD/10815/2002), and the National Institutes of
Health for financial support.
Docking 23a to papain (PDB code: 1PPN) revealed
that seven of the top 10 solutions present the N-ben-
zyl-d-sultam moiety being incorporated in the S2 sub-
site defined by Pro68, Val133 and Val157, and a
˚
Supplementary data
Cys25(S)-Cb distance ranging from 4.6 to 4.9 A.
Finally, all docked conformations of 23b show the
N-benzyl-d-sultam in the S2 pocket. Clearly, these re-
sults suggest that molecular recognition of 23a,b by
papain leads to enzyme–inhibitor complexes that are
conformationally unfavourable to conjugate addition
of Cys25 to the vinyl sultam, as required for enzyme
inactivation.
Supplementary data associated with this article (synthe-
sis procedures, spectroscopic data and molecular model-
ling details) can be found, in the online version, at
doi:10.1016/j.bmcl.2006.04.079.
References and notes
In summary, this paper describes the synthesis of vinyl
sultam derivatives via the Wittig–Horner reaction.
Although weakly active, vinyl sultams are selective for
1. WHO Expert Committee on Malaria: 20th Report, WHO
Technical Report Series, Vol. 892: Geneva, 2000.

C. Valente et al. / Bioorg. Med. Chem. Lett. 16 (2006) 4115–4119
4119
¨
2. Powers, J. C.; Asgian, J. L.; Ozlem, D. E.; James, K. E.
Kobayashi, M.; Ohta, N.; Kamata, S.; Tsuri, T. J. Org.
Chem. 2002, 67, 125.
17. Merten, S.; Fro¨hlich, R.; Kataeva, O.; Metz, P. Adv.
Synth. Catal. 2005, 347, 754.
Chem. Rev. 2002, 102, 4639.
3. Hans-Hartwig, O.; Schirmeister, T. Chem. Rev. 1997, 97,
133.
4. Rosenthal, P. J. Adv. Parasitol. 1999, 43, 105.
5. Shenai, B. R.; Sijwali, P. S.; Singh, A.; Rosenthal, P. J.
J. Biol. Chem. 2000, 275, 29000.
6. Shenai, B. R.; Lee, B. J.; Alvarez-Hernandez, A.; Chong,
P. Y.; Emal, C. D.; Neitz, R. J.; Roush, W. R.; Rosenthal,
P. J. Antimicrob. Agents Chemother. 2003, 47, 154.
7. Lecaille, F.; Kaleta, J.; Bro¨mme, D. Chem. Rev. 2002, 102,
4459.
8. Patani, G. A.; LaVoie, E. J. Chem. Rev. 1996, 96, 3147.
9. Iley, J.; Moreira, R.; Martins, L.; Guedes, R. C.; Soares,
C. M. Bioorg. Med. Chem. Lett. 2006, 16, 2738.
10. Schmidt, T. J. Bioorg. Med. Chem. 1997, 5, 645.
11. Steurer, S.; Podlech, J. Eur. J. Org. Chem. 2002, 899.
12. Erman, W. F.; Kretschmar, H. C. J. Org. Chem. 1961, 26,
4841.
18. Kaiser, E. M.; Knutson, P. L. A. J. Org. Chem. 1975, 40,
1342.
19. Fehrentz, J. A.; Castro, B. Synthesis 1983, 8, 676.
20. Motoyoshiya, J.; Kusaura, T.; Kokin, K.; Yokoya, Sei-
ichi; Takaguchi, Y.; Narita, S.; Aoyama, H. Tetrahedron
2001, 57, 1715.
21. Go¨tz, M. G.; Caffrey, C. R.; Hansell, E.; McKerrow, J. H.;
Powers, J. C. Bioorg. Med. Chem. 2004, 12, 5203.
22. Gomes, P.; Arau´jo, M. J.; Rodrigues, M.; Vale, N.;
Azevedo, Z.; Iley, J.; Chambel, P.; Morais, J.; Moreira, R.
Tetrahedron 2004, 60, 5551.
23. Scheidt, K. A.; Roush, W. R.; McKerrow, J. H.; Selzer, P.
M.; Hansell, E.; Rosenthal, P. J. Bioorg. Med. Chem. 1998,
6, 2477.
24. Zhao, G.; Zhou, Z. S. Bioorg. Med. Chem. Lett. 2001, 11,
2331.
25. Rosenthal, P. J.; Olson, J. E.; Lee, G. K.; Palmer, J. T.;
Klaus, J. L.; Rasnick, D. Antimicrob. Agents Chemother.
1996, 40, 1600.
13. Farbenfabriken Bayer Akt.-Ges. British Patent 810, 356,
1959; Chem. Abstr. 1959, 53, 16068.
14. Palmer, J. T.; Rasnick, D.; Klaus, J. L.; Bro¨mme, D.
J. Med. Chem. 1995, 38, 3193.
15. Carretero, J. C.; Demillequand, M.; Ghosez, L. Tetrahe-
dron 1987, 43, 5125.
26. CCDC Software Ltd. Cambridge, UK; for modelling
details, see Supplementary data.
16. Inagaki, M.; Tsuri, T.; Jyoyama, H.; Ono, T.;
Yamada, K.; Kobayashi, M.; Hori, Y.; Arimura, A.;
Yasui, K.; Ohno, K.; Kakudo, S.; Koizumi, K.;
Suzuki, R.; Kato, M.; Kawai, S.; Matsumoto, S.
J. Med. Chem. 2000, 43, 2040; Inagaki, M.; Haga, N.;
27. Sajid, M.; McKerrow, J. Mol. Biochem. Parasitol. 2002,
120, 1.
28. Desai, P. V.; Patny, A.; Sabnis, Y.; Tekwani, B.; Gut, J.;
Rosenthal, P.; Srivastava, A.; Avery, M. J. Med. Chem.
2004, 47, 6609.