Synthesis of a molecular mimic of the Glc1Man9 oligoside as potential inhibitor of calnexin binding to ΔF508 CFTR protein

Article abstract of DOI:10.1016/S0960-894X(02)00151-8

Deletion of phenylalanine at position 508 of the CFTR protein is associated with a severe form of cystic fibrosis. Biosynthetic arrest of the misfolded ΔF508 CFTR protein in the endoplasmic reticulum is due to prolonged interaction with protein chaperones. In order to overcome this retention and thereby restore the delivery of the protein to the plasma membrane, a molecular mimic of the glycoprotein oligoside moiety has been designed and synthesized. Ability of this mimic to inhibit the binding of the natural Glc₁Man₉GlcNAc oligoside to calnexin has been measured.

Full text of DOI:10.1016/S0960-894X(02)00151-8

Bioorganic & Medicinal Chemistry Letters 12 (2002) 1237–1240
Synthesis of a Molecular Mimic of the Glc1Man9 Oligoside as
Potential Inhibitor of Calnexin Binding to DF508 CFTR Protein
Slim Cherif,a Michael R. Leach,b David B. Williamsb and Claude Monnereta,*
aUMR 176 CNRS-Institut Curie, Section de Recherche, 26 rue d’Ulm, 75248 Paris Cedex 05, France
bDepartment of Biochemistry, University of Toronto, Toronto, Canada M5S 1A8
Received 17 December 2001; revised 25 February 2002; accepted 28 February 2002
Abstract—Deletion of phenylalanine at position 508 of the CFTR protein is associated with a severe form of cystic fibrosis. Bio-
synthetic arrest of the misfolded ÁF508 CFTR protein in the endoplasmic reticulum is due to prolonged interaction with protein
chaperones. In order to overcome this retention and thereby restore the delivery of the protein to the plasma membrane, a molec-
ular mimic of the glycoprotein oligoside moiety has been designed and synthesized. Ability of this mimic to inhibit the binding of
the natural Glc1Man9GlcNAc oligoside to calnexin has been measured. # 2002 Elsevier Science Ltd. All rights reserved.
The autosomal recessive human genetic disorder cystic
fibrosis (CF) is caused by the loss or dysfunction of a
plasma membrane Clꢀ channel known as the CF trans-
membrane conductance regulator (CFTR).1 Among the
over eight hundred mutations that have been detected,
deletion of a phenylalanine at position 508 (>90% of
CF patients) is associated with a severe form of the dis-
ease.2 The ÁF508 mutation is hypothesized to cause
‘misfolding’ of the nascent chain and prevents delivery
of the cystic fibrosis transmembrane conductance reg-
ulator to the plasma membrane.3 Several data suggest
that at least three chaperones, Hsc70,4 Hsp90,5 and cal-
nexin,6 participate in CFTR biogenesis. Both chaper-
ones form transient complexes with nascent, immature
CFTR molecules. Because restoration of processing can
be partially obtained by incubation of cells between 25
and 29 ꢁC7,8 and, since the ÁF508 mutation does not
substantially decrease the chloride channel activity of
CFTR,9 any circumvention of the misprocessing of
ÁF508 CFTR provides an attractive therapeutic strat-
egy. For example, drugs which influence the interaction
of molecular chaperones with their substrates10 have
already been investigated with some success.
glycoproteins bind to certain molecular chaperones only
subsequent to the processing of the carbohydrate units
(a triglucosyl sequence) to the monoglucosylated state.
Binding studies to calnexin12 and to calreticulin13 have
indicated that a monoglucosylated minimum structure
like Glc1Man5GlcNAc with its a 1!6 branch point,
A–F, as underlined in Figure 1, is required for relevant
chaperone binding.
En route towards this oligosaccharide, we already
reported14 the synthesis of disaccharide A–B and tri-
saccharide A–B–C moieties and, more recently, the
synthesis of the methyl glycoside of the tetrasaccharide
A–B–C–D.15
It remained to determine whether a molecule containing
the determinant disaccharide, Glc-Man A–B and the
two terminal mannoses G and H with the same orien-
tation in space as in the parent structure, is still a rele-
vant ligand for calnexin binding. To progress in this
Any attempt to overcome the biosynthetic arrest of
ÁF508 CFTR protein requires knowledge of the
mechanism that causes this retention.11 Newly synthesized
*Corresponding author. Tel.: +33-1-4234-6655 fax: +33-1-4234-
6631; e-mail: claude.monneret@curie.fr
Figure 1. Oligoside core of CFTR.
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00151-8

1238
S. Cherif et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1237–1240
knowledge, we have designed the molecular mimic 1 of
the oligoside core of CFTR including these different
elements separated by linkers. Simultaneously, we pos-
tulated that, by introducing carbamate linkages instead
of glycosidic bonds, higher affinity between such a
molecular mimic and the protein chaperone could be
achieved by improving hydrophobicity of the molecule,
and better stability could be observed (Fig. 2).
Conversion of 4 into the mannosyl acceptor 7 was con-
veniently achieved after debenzoylation (NaOMe–MeOH),
monobenzylidenation (HBF4, 42%),16 and regioselective
benzylation of 6 (55% yield) under phase transfer cata-
lysis17 (BnBr, NaOH, Bu4NHSO4). Condensation of 7 with
per-O-benzyl glucose trichloroacetimidate18 8 (TMSOTf,
dioxane, rt) afforded, with high stereoselectivity
(>95:5), the disaccharide 9, isolated in 71% yield. Next,
removal of the p-methoxybenzyl group led to 10 (42%
yield) which can be activated at the anomeric position to
subsequently afford linear tri or tetrasaccharide A–B–C
or A–B–C–D identical to those previously reported.14,15
To obtain the disaccharide determinant, a glucose-man-
nose-containing disaccharide selectively deprotected at
the anomeric position was first prepared. Thus the per-O-
benzoyl-d-mannose 2a was converted into the chloride
derivative 3a which, in turn, led to the p-methoxybenzyl
glycoside 4 under Koenigs–Knorr conditions (AgOTf,
sym-collidine, 62%). It must be noticed that, under the
same conditions, the corresponding O-acetylated man-
nose compound 2b and thereby the chloride derivative 3b
led to the 1,2-ortho-ester 5 exclusively (Scheme 1).
However, subsequent progress in the present synthesis
involved the conversion of 9 into the corresponding per-
O-acetylated compound 11 by hydrogenolysis (Pd/C
10%, MeOH, quantitative) and pyridine acetylation.
After selective anomeric deprotection, compound 12 was
activated as the phenylcarbonate 1319 before coupling
with 2-(2-aminoethoxy)ethanol to afford 14 which was
subsequently transformed into the activated species 15
(Scheme 2).
On the other hand, the di-mannose fragment 19 was
readily prepared (Scheme 3) from per-O-acetylated
d-mannose 16. Thus, 16 was regioselectively deacety-
lated (H2N-NH2, AcOH, 65%) at the anomeric carbon
to afford 17, which was activated as the 4-nitrophenyl-
carbonate (90%) 18 and condensed with 2-(2-amino-
ethoxy)ethanol in the presence of Et3N (90%).
The next step involved formation of the 4-nitro-
phenylcarbonate of 19 followed by condensation with
diaminopropanol to give 2120 (74% yield) which, in
turn, was once again, activated and condensed with
mono-Boc-1,3-diaminopropanol leading to 23 (59%)
and to 24,21 after deprotection under acidic conditions.
.
Synthesis of 1 [HRMS: (Maldi-TOF, DHB, EtOH H2O,
TFA 0.1%). Calculated: m/z 1314.4327 (M+ Na+).
Figure 2. Target molecule.
Scheme 1. (a) a,a-Dichloromethyl methylether (5 equiv), ZnCl2–Et2O (1 M in Et2O) (0.1 equiv), Ch2Cl2, 16 h, 87%; (b) p-MBnOH (4 equiv),
CF3SO3Ag (1.3 equiv), sym-collidine (0.8 equiv), CH2Cl2, 4 A, molecular sieves, 15 h, rt, 62% for 4 and 85% for 5; (c) NaOMe–MeOH, rt, 24 h,
then benzylidene dimethylacetal (1.1 equiv), HBF4 in Et2O (1 equiv), DMF, 24 h, 42%; (d) BnBr, aq NaOH, Bu4NHSO4, CH2Cl2, 55%; (e)
TMSOTf, dioxane, 8 (2 equiv), rt, 30min, 71%; (f) CAN, CH 3CN–H2O (4:1), 0 ꢁ to rt, 42%.

S. Cherif et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1237–1240
1239
Scheme 2. (a) H2, Pd/C, rt, 5 h; Ac2O, pyridine, 15 h, 80%; (b) H2N–NH2, DMF, 16 h, 50 ꢁC, 52%; (c) phenyl chloroformate (2.5 equiv), pyridine
(2.5 equiv), CH2Cl2, 3 h, rt, 79%; (d) 2-(2-aminoethoxy)ethanol (1.5 equiv), Et3N, CH2Cl2, 15 h rt, 74%; (e) 4-nitrophenyl chloroformate (2.5 equiv),
pyridine (2.5 equiv), CH2Cl2, 0 ꢁC to rt, 15 h, 79%.
Scheme 3. (a) H2N–NH2, AcOH, THF, DMF, 15 min, 50 ꢁC, 65%; (b) 4-nitrophenyl chloroformate, pyridine, CH2Cl2, 5 h, rt, 85%; (c) 2-(2-amino-
ethoxy)ethanol, Et3N, CH2Cl2, 15 h, rt, 90%; (d) 1,3-diaminopropanol (0.5 equiv), Et3N, 15 h, rt, 74%; (e) Mono-boc-1,3-diaminopropanol,
DMAP, THF, 15 h, 50 ꢁC, 59%; (f) 1 M HCl, EtOAc, 4.5 h, rt, 62%.
D. W.; White, G. A.; O’Riordan, C. R.; Smith, A. E. Cell
1990, 63, 8727.
4. Yang, Y.; Janich, S.; Cohn, J. A.; Wilson, J. M. Proc. Natl.
Found: 1314.4306) was conveniently achieved through
condensation of fragment 15 with fragment 24 followed
by deacylation under Zemplen conditions (37% overall
Acad. Sci. U.S.A. 1993, 90, 9480.
yield for the last two steps).
5. Loo, M. A.; Jensen, T. J.; Cui, L.; Hou, Y.; Chang, X. B.;
Riordan, J. R. EMBO J. 1998, 17, 6679.
Molecular mimic 1 was tested for its ability to inhibit
the binding of [3H] Glc1Man9GlcNAc2 to GST-fused
calnexin22 immobilized on glutathione-agarose. Com-
6. Pind, S.; Riordan, J. R.; Williams, J. M. J. Biol. Chem.
1994, 269, 12784.
7. Denning, G. M.; Anderson, M. P.; Amara, J. F.; Marshall,
pared to the Glc a (1!3) Man disaccharide, compound
1 was a less potent inhibitor, since concentrations pro-
ducing 50% inhibition were about 300 mM for com-
pound 1 and 160 mM for the disaccharide. That the
linear tetrasaccharide Glc1Man3 (A–B–C–D in Fig. 1) is
several hundred-fold more potent that the disaccharide
itself suggests that the linear disposal of the four sugar
units is more relevant for binding inhibition than the
additional terminal mannose residues that mimic the
other arms of the Glc1Man9GlcNAc2 oligoside. How-
ever, among the different hypotheses which cannot be
excluded to explain this relatively low activity, are the
nature of the linkers and scaffold and the length of the
linkers. Therefore, experiments are going on to synthe-
size other molecular mimics of the oligoside part of a
high-mannose type glycoprotein.
J.; Smith, A. E.; Welsh, M. J. Nature 1992, 358, 761.
8. Brown, C. R.; Hong-Brown, L. Q.; Welch, W. J. J. Clin.
Invest. 1997, 99, 1432.
9. Zeitlin, P. J. Clin. Invest. 1999, 103, 447.
10. Li, C.; Ramjeesingh, M.; Reyes, E.; Jensen, T.; Chang, X.;
Rommens, J. M.; Bear, C. E. Nat. Genet. 1993, 3, 312.
11. Helenius, A. Mol. Biol. Cell 1994, 358, 761.
12. Ware, F. E.; Vassilakos, A.; Peterson, P. A.; Jackson,
M. R.; Lehrman, M. A.; Williams, D. B. J. Biol. Chem. 1995,
270, 4697.
13. Spiro, R. G.; Zhu, Q.; Bhoyroo, V.; Soling, H.-D. J. Biol.
Chem. 1996, 271, 11588.
14. Cherif, S.; Clavel, J.-M.; Monneret, C. J. Carb. Chem.
1998, 17, 1203.
15. Cherif, S.; Clavel, J.-M.; Monneret, C. J. Carb. Chem.
2002, 12, 1.
16. Albert, R.; Dax, K.; Pleschko, R.; Stutz, A. E. Carbohydr.
Res. 1985, 137, 282.
17. Garegg, P. J.; Iversen, J.; Oscarson, S. Carbohydr. Res.
1976, 50, C12.
References and Notes
18. Schmidt, R. R.; Michel, J.; Roos, M. Liebigs Ann. Chem.
1984, 1343.
1. Welsh, M. J.; Smith, A. E. Cell 1993, 73, 1251.
2. Kopito, R. R. Physiol. Rev. 1999, 79, S167.
3. Cheng, S. H.; Gregory, R. J.; Marshall, J.; Paul, S.; Souza,
19. Compound 13: syrup; [a]2D2 +34.5ꢁ (c 1, chloroform); MS
1
(FAB) 779.05 (M+Na+), 794.9 (M+K+); H NMR: d 2.01,

1240
S. Cherif et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1237–1240
2.05, 2.06, 2.07, 2.10, 2.11, 2.29 (7s, 7Â3H, OAc), 4.05–4.31 (m,
6H, 3-H, 50-H, 60-H, 60-H), 4.34 (dd, 1H, J=9.5, J0=3.5 Hz), 4.76
(dd, 1H, J=3.5 Hz), 5.01 (dd, 1H, J=J0=9.5 Hz), 5.30–5.45 (m,
4H), 6.07 (d, 1H, J=1.8 Hz, 1-H), 7.18–7.44 (m, 5H, Ar).
4H, CH2OCO), 4.06–4.16 (m, 4H, 5-H, 60-H), 5.68 (m, 4H,
NH), 6.00 (s, 1H, 1-H).
21. Compound 24: syrup; [a]2D2 +34ꢁ (c 1, chloroform); MS
(FAB) 1129.37 (M+ H+), 1151.38 (M+Na+); HMRS m/z
20. Compound 21: syrup; [a]2D2 +46ꢁ (c 1, chloroform); MS
(FAB) 1013.39 (M+H+), 1035.3 (M+ Na+); (found C:
44.86; H, 5.51; N, 6.10). C39H56N4O27 requires C, 46.25; H,
1151.3662. calcd 1151.3615; H NMR: d 2.00, 2.05, 2.09, 2.18
1
(4s, 8Â3H, OAc), 4.77 (broad s, 1H, NH), 6.00 (s, 1H, 1-H),
6.20–6.70 (m, 10H).
22. Vassilakos, A.; Michalak, M.; Lehrman, M. A.; Williams,
D. B. Biochemistry 1998, 37, 3480.
1
5.57; N, 6.20; H NMR: d 1.82 (broad s, 1H, OH), 2.01, 2.05,
2.1, 2.18 (4s, 8Â3H, OAc), 3.28 (m, 4H, CH2NH), 3.48 (m,