Synthesis of 5-(mercaptomethyl)-3(E)-undecene-1,11-dioic acid, a non-peptide glutathione analog

Article abstract of DOI:10.1016/S0040-4039(01)00554-8

The key step in the synthesis of the title compound is a Julia reaction to create the E-double bond by introducing a functionalized C3 carbon chain.

Full text of DOI:10.1016/S0040-4039(01)00554-8

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 3709–3711
Synthesis of 5-(mercaptomethyl)-3(E)-undecene-1,11-dioic acid,
a non-peptide glutathione analog
Jean-Fre´de´ric Duckert, Laurence Balas* and Jean-Claude Rossi
UMR C.N.R.S. 5074, Universite´ Montpellier I, Faculte´ de Pharmacie, 15 Av. C. Flahault, F-34060 Montpellier Cedex, France
Received 30 March 2001; accepted 3 April 2001
Abstract—The key step in the synthesis of the title compound is a Julia reaction to create the E-double bond by introducing a
functionalized C3 carbon chain. © 2001 Elsevier Science Ltd. All rights reserved.
Glutathione (GSH) is a widely spread mercapto-con-
taining tripeptide. It is found in nearly all cells playing
various physiological and biochemical roles.1,2 As a
result, it is largely involved in human health and dis-
eases.3–6 Thus, GSH analogs, as mimics or enzymes
inhibitors, have often been studied as valuable tools for
a better knowledge of the catalytic sites of different
GSH-binding enzymes.
follows: (i) the analog should bear two terminal car-
boxylic functions;10 (ii) regardless of the carboxylic
acids, the chain lengths are best optimized with C3
atoms at one side of the ethylene thiol group and C5
atoms at the other.11 According to Toda,12 we decided
not to take into account the amino group of GSH since
it is not crucial for the binding to the LTC4 receptor.
From peptide literature,13 the presence of a trans-
alkene instead of an amide linkage usually provides
good bioisosteric mimics due to their close stereochem-
ical resemblance and their inertness to enzymatic
hydrolysis.
We are interested in GSH mimics as precursors for
suitable non-peptide leukotriene (LT) C4 probes to
further investigate the cys-LT2 receptor, implicated in
asthma. Although the literature on GSH derivatives is
well documented, few reports concern the synthesis of
non-peptide GSH analogs.7–9 Structure–affinity rela-
tionships observed in the literature can be featured as
Thus, on the basis of these observations, we report
herein the synthesis of a new GHS analog 1: we replace
the cysꢀgly amide bond of GSH by a E-double bond,
SH
O
O
O
H
N
Glutathione
N
H
OH
OH
HO
HO
NH2
O
SH
O
O
O
O
O
O
O
O
O
O
1
OMe
OMe
OMe
EtO
EtO
3
4
C3unsat
SH
C5sat
O
O
O
ref.11
2
OH
HO
MeO
5
C3sat
C5sat
Scheme 1.
* Corresponding author. Fax: 33-4-67-54-86-25; e-mail: balas@balard.pharma.univ-montp1.fr
0040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(01)00554-8

3710
J.-F. Duckert et al. / Tetrahedron Letters 42 (2001) 3709–3711
expecting an improvement of the affinity for the cys-
LT2 receptor in comparison with the analog containing
the completely saturated C5satC3sat 2 structure.
would be obtained from a Julia condensation between a
sulfone 7 and an C3 carbon atoms aldehyde 8. The
sulfone 7 could be derived from 1,3-diol 9, easily pre-
pared in three steps from commercially available 6-
chlorohexan-1-ol. The protective group G of the
sulfone 7 should be resistant to the reaction conditions
used for the deprotection of the R and R% groups, and
compatible with the Julia reaction and Jones oxidation
conditions. Following these requirements, G must be a
MOM group, whereas R and R% were chosen as tert-
butyl diphenylsilyl groups.
The synthesis of target 1 was first run via functionaliza-
tion of the corresponding a,b-unsaturated E-ketone 3
as previously reported11 for compound 2 (Scheme 1).
However, this strategy failed due to an in situ isomer-
ization of ketone 3 to 4.
To avoid these difficulties, we applied a new strategy
based on the Julia olefination reaction.14,15 Our
approach is summarized in the retrosynthetic Scheme 2,
where it was envisioned that the access to the homo-
allylic thiol 1 could be reached via a Mitsunobu reac-
tion from its hydroxylated analog 6, which in turn
The synthetic route to compound 1 is depicted in
Scheme 3. Thus, after silylation of 6-chlorohexan-1-ol
with tert-butyl diphenylsilyl chloride, alkylation with
diethyl malonate provides in 61% yield the diester 10.
SH
OH
O
O
O
O
HO
OH
MeO
OMe
6
1
OH
OH
OG
O
Cl
SO2Ph
+
HO
H
RO
R'
O
RO
7
9
8
Scheme 2.
OH
OH
CO2Et
CO2Et
c
a, b
Cl
HO
TBDPSO
TBDPSO
11
10
d
OMOM
OH
OMOM
SO2Ph
OMOM
I
f
e
TBDPSO
OMOM
TBDPSO
TBDPSO
14
13
12
g
MOMO
OH
h
TBDPSO
OTBDPS
TBDPSO
OTBDPS
16 (Z/E) : 20/80
SO2Ph
15
i
OMOM
OMOM
OH
O
O
O
O
j
k
MeO
OMe
MeO
OMe
HO
OH
6
18
17
l
SAc
SH
O
O
O
O
m
HO
OH
MeO
OMe
1
19
Scheme 3. (a) TBDPSCl, NEt3, DMAP, CH2Cl2, 97%; (b) diethylmalonate, NaH, THF, 61%; (c) LiAlH4, THF, reflux, 71%; (d)
MOMCl (1 equiv.), NaH, THF, 69%; (e) PPh3, imidazole, I2, xylene, 80°C, 88%; (f) PhSO2Na, EtOH, reflux, 78%; (g) n-BuLi,
−78°C then 8, −78 to −50°C; (h) Na–Hg 5%, THF/MeOH, 0°C, 77% (E+Z) from 14; (i) (nBu)4NF, THF, 88% (E+Z); (j) Jones
reagent, acetone, −10°C then CH2N2, Et2O, 60% (E+Z); (k) TMSBr, CH2Cl2, −30°C, 69% (E+Z) then preparative HPLC; (l)
PPh3, DIAD, AcSH, THF, 67%; (m) KOH, EtOH/H2O, reflux, 60%.

J.-F. Duckert et al. / Tetrahedron Letters 42 (2001) 3709–3711
3711
The later is reduced into 1,3-diol 11 by an excess of
References
lithium aluminium hydride in 71% yield. Selective
monoprotection with methoxymethyl chloride (1 equiv.,
69%) followed by a modified Garreg reaction afforded
the iodide 13, which was transformated in 78% yield into
the corresponding racemic sulfone 14 in the presence of
phenylsulfinic acid sodium salt. The lithiated sulfone 14
was condensed with aldehyde 8 (R%=OTBDPS, derived
from propan-1,3-diol)16 at −78 to −50°C. Reduction of
the resulting diastereoisomeric mixture of b-hydroxysul-
fones 15 with sodium amalgam at 0°C afforded the
expected alkenyl compound 16 in 77% yield from 14. The
addition of sulfone anions to aldehydes and the reduction
step of the resulting hydroxy sulfones are capricious
processes depending on several factors.15 The Kocienski–
Lythgoe modification17 of the Julia reaction or
activation18 of the aldehyde by complexation with
DIBAL–methoxide did not allow us to improve the yield.
Different reducting agents (SmI2-HMPA, Mg power,
Na–Hg amalgam) were compared. In all our attempts to
perform the Julia olefination sequence, we observed only
a moderate stereoselectivity (Z/E: 20/80, measured by 1H
NMR spectroscopy at 360 MHz).
1. Sies, H. Free Rad. Biol. Med. 1999, 27, 916–921.
2. Monks, T. J.; Ghersi-Egea, J.-F.; Philbert, M.; Cooper,
A. J. L.; Lock, E. A. Toxicol. Sci. 1999, 51, 161–177.
3. Lomaestro, B. M.; Malone, M. Ann. Pharmacother. 1995,
29, 1263–1273.
4. Kelly, F. J. Food Chem. Toxicol. 1999, 37, 963–966.
5. Schro¨der, C. P.; Godwin, A. K.; O’Dwyer, P. J.; Tew, K.
D.; Hamilton, T. C.; Ozols, R. F. Cancer Invest. 1996, 14,
158–168.
6. Nagasawa, H. T.; Cohen, J. F.; Holleschau, A. M.;
Rathbun, W. B. J. Med. Chem. 1996, 39, 1676–1686.
7. Luisi, G.; Calcagni, A.; Pinnen, F. Tetrahedron Lett.
1993, 34, 2391–2392.
8. Lucente, G.; Luisi, G.; Pinnen, F. Il Farmaco 1998, 53,
721–735 and references cited therein.
9. Vince, R.; Brownel, J.; Akella, L. B. Bioorg. Med. Chem.
Lett. 1999, 9, 853–856.
10. Adang, A. E. P.; Brussee, J.; Gen, A. V.d.; Mulder, G. J.
Biochem. J. 1990, 269, 47–54.
11. Klotz, P.; Slaoui-Hasnaoui, A.; Bane`res, J.-L.; Duckert,
J.-F.; Rossi, J.-C. J. Med. Chem. 1998, 41, 2278–2288.
12. Okuyama, S.; Miyamoto, S.; Shimoji, K.; Konishi, Y.;
Fukushima, D.; Niwa, H.; Arai, Y.; Toda, M.; Hayashi,
M. Chem. Pharm. Bull. 1982, 30, 2453–2462.
13. Hann, M. M.; Sammes, P. G.; Kennewell, P. D.; Taylor,
J. B. J. Chem. Soc., Perkin Trans. 1 1982, 307–314.
14. Julia, M.; Paris, J.-M. Tetrahedron Lett. 1973, 4833–4836.
15. Kocienski, P. Phosphorus Sulfur 1985, 24, 97–127.
16. Breen, A. P.; Murphy, J. A. J. Chem. Soc., Perkin Trans.
1 1993, 23, 2979–2990.
After deprotection of both terminal silyl ethers with
fluoride anions in 88% yield, followed by Jones oxida-
tion, the diacid was treated with diazomethane to afford
the diester 18 (60% yield from 17). Selective removal of
the MOM group with TMSBr at −30°C then produced
free homoallylic alcohol 6 (69% yield), which was func-
tionalized via a Mitsunobu reaction with thioacetic acid
in 67% yield. Finally, in the resulting compound 19, the
thioacetate group was cleaved together with ester func-
tions under basic conditions at EtOH–H2O reflux to give
thiol 119 in 60% yield.
17. Kocienski, P. J.; Lythgoe, B.; Ruston, S. J. Chem. Soc.,
Perkin Trans. 1 1978, 829–833.
18. Spaltenstein, A.; Carpino, P. A.; Miyake, F.; Hopkins, P.
B. J. Org. Chem. 1987, 52, 3759–3766.
19. Compound 1: 5-(Mercaptomethyl)-3(E)-undecen-1,11-
dioic acid: 1H NMR (360 MHz, CDCl3) l ppm: 1.20–1.40
(m, 6H, H6, H7, H8); 1.60–1.72 (m, 2H, H9); 2.17–2.23
(m, 1H, H5); 2.34 (t, 2H, J=7.7 Hz, H10); 2.51 (dd, 2H,
J=8.4 Hz and 6.4 Hz, CH2); 3.13 (td, 2H, J2–3=6.4 Hz,
Surprisingly, we found no step where the Z/E mixture
could be easily separated by flash column chromato-
graphy. Using different silica gel phases and eluent
conditions, isolation by HPLC of the pure E-isomer 19
or 1 failed since the Z- and E-isomers were co-eluted and
the thiol function readily dimerized. Obtention of the
pure E-homoallylic thiol 1 was successfully achieved
from pure E-homoallylic alcohol 6, which was separated
from its isomers 6-Z by preparative HPLC on reverse
phase (gradient H2O–MeCN, Lichrospher RP-18, 5 mm).
J
2–4=1.70 Hz, H2); 5.32 (ddt, 1H, J3–4=15.4 Hz, J4–5
8.8 Hz, J4–2=1.2 Hz, H4); 5.60 (td, 1H, J3–2=7.0 Hz,
3–4=15.4 Hz, H3); 9.50 (s broad, 2H, COOH). 13C
=
J
NMR (360 MHz, CDCl3) l ppm: 24.29 (C9); 26.52
(CH2); 28.70 (CH2); 29.64 (C12); 33.37 (CH2); 33.90
(C10); 37.77 (C2); 45.83 (C5); 123.38 (C3); 136.80 (C4);
178.22 (CꢁO); 180.30 (CꢁO). LC–MS analysis: tR=10.57
min; (ES+) m/z=260.64 with gradient 100% H2O to 100%
MeCN–0.1% TFA in 20 min; column: symmetry shield
RP-18, 4.6×50 mm (3.5 mm); 1 mL/min; at u=220 nm.
20. Compound 1 was prepared as a racemic mixture. We
failed to separate the enantiomers by chiral HPLC or
using a resoluting agent to create diastereomers. Hope-
fully, after coupling to LTA4, the separation of the
different diastereomers of the expected LTC4 analog 20
should be easier.
In conclusion, application of the Julia olefin synthesis
provided an efficient access to the required E-double
bond on the C3 carbon atom chain. The 13 steps leading
to target 1 were achieved in 2.5% overall yield.
Starting from leukotriene A4, chemical epoxide ring-
opening with thiol 120 is underway to provide a protease-
resistant leukotriene C4.
OH
*
O
*
Acknowledgements
OMe
S
O
O
We wish to thank Professor Marc Julia for strengthening
discussion.
20
*
OH
HO