Design and synthesis of functionalized trisaccharides as p53-peptide mimics

Article abstract of DOI:10.1016/j.tetlet.2010.05.012

Oligosaccharides represent potentially useful scaffolds for the development of peptidomimetics. We report here the design and synthesis of functionalized trisaccharides modeled after an α-helical 15-mer peptide region of p53 which binds to its cellular regulator MDM2. The trisaccharide scaffold was obtained efficiently by applying the sulfoxide glycosylation reaction as a key methodology.

Full text of DOI:10.1016/j.tetlet.2010.05.012

Tetrahedron Letters 51 (2010) 3724–3727
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Design and synthesis of functionalized trisaccharides as p53-peptide mimics
*
Kaori Sakurai , Daniel Kahne
Department of Chemistry and Chemical Biology, Harvard University, 12 Oxford St., Cambridge, MA, USA
Department of Biological Cellular and Molecular Pharmacology, Harvard Medical School, Longwood, Boston, MA, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Oligosaccharides represent potentially useful scaffolds for the development of peptidomimetics. We
Received 3 April 2010
Revised 30 April 2010
Accepted 7 May 2010
Available online 15 May 2010
report here the design and synthesis of functionalized trisaccharides modeled after an
a-helical
15-mer peptide region of p53 which binds to its cellular regulator MDM2. The trisaccharide scaffold
was obtained efficiently by applying the sulfoxide glycosylation reaction as a key methodology.
Ó 2010 Elsevier Ltd. All rights reserved.
Keywords:
Oligosaccharide
Scaffold
Peptidomimetics
p53-MDM2 interaction
Oligosaccharides offer attractive scaffolds for the development of
peptdomimetics.1–5 Monosaccharides possess polyhydroxyl groups
which can be readily functionalized and their well-defined
conformation allows the presentation of the functionalities to be
designed in a predictable fashion. There are a variety ofmonosaccha-
rides, which could be built into various architectures of polymers by
employing different types of glycosidic linkages. However, oligosac-
charides had not been extensively exploited to mimic relatively
large protein secondary structures. We previously demonstrated
Trp23, and Leu26, corresponding to i, i + 4, i + 7 residues of the helix,
make hydrophobic contacts to the MDM2 pocket (Fig. 1a and b).7 It
suggested that the side chain functionalities of the three residues
represent the minimal structural elements critical for binding
MDM2. To test this notion, we designed a p53-peptide mimic based
on a trisaccharide scaffold as shown in Figure 1c. Based on the de-
sign strategy developed in our prior studies,5 an
a-1,4-linked tri-
2-deoxygalactose was employed as a suitable scaffold to match
the molecular dimension of the MDM2-binding region of the p53
peptide.10 It has been proposed earlier that overall conformation
of oligosaccharides could be predicted qualitatively through
conformational analysis of glycosidic bonds, which are primarily
governed by an exo-anomeric effect and non-bonded steric effects
between the adjacent sugars.4,8 It was thus predicted that the glyco-
the use of a pentasaccharide as a scaffold for an
mic.5 A functionalized pentasaccharide was designed based on an
helical DNA-binding peptide region of a bZIP protein and was shown
to indeed bind DNA. -Helical peptides are also commonrecognition
a-helical peptide mi-
a
-
a
elements in protein-protein interactions such as in p53-MDM2 and
BAD-BcLxL, which are important chemotherapeutic targets. Because
of the potential utility as therapeutics and biological probes, devel-
sidic bonds in the a-1,4-linked oligogalactose system would favor
an arrangement where monomer sugars are stacked on top of each
other, resulting in an overall rod-shaped molecule.5,9 In this pro-
posed conformation, C3 and C6 hydroxyl groups of the adjacent
monosaccharide units would emerge from the same face of the
rod-shaped trisaccharide (Fig. 1c). By comparing a computer-gener-
opment of synthetic
interactions has been an active area of research.6 To expand the
scope of oligosaccharides as suitable scaffolds for the -helix mim-
a-helix mimics as inhibitors of protein–protein
a
ics, we next explored the design of oligosaccharides with protein-
binding functionalities. Here, we report the design and synthesis of
a trisaccharide scaffold bearing the MDM2-binding functionalities
of an a-helical peptide region of p53.
An X-ray structure of the N-terminal domain of MDM2 bound to
a p53-derived 15-residued peptide fragment has shown that Phe19,
ated model for an
a-1,4-linked tri-2-deoxygalactose scaffold and
the crystal structure of the MDM2-bound p53 peptide (Fig. 1b and
d), C6–OH of A ring, C3–OH of B ring, and C6–OH of C ring were cho-
sen to bear the side chain analogues for Phe, Trp, and Leu residues
(Fig. 1c). To endow the molecule with water solubility, amino
groups were incorporated to the A and C ring monomers.11 The
MDM2-bound p53 peptide has distinct amphiphilic faces with
hydrophobic residues lined up on one side of the helix and hydro-
philic residues on the opposite side. It was anticipated that mimick-
ing such an amphiphilic property may also help enhance the MDM2
binding ability of the trisaccharide.
* Corresponding author. Tel.: +1 617 496 0208; fax: +1 617 496 0215.
E-mail address: kahne@chemistry.harvard.edu (D. Kahne).
Present address: Department of Biotechnology and Life Science, School of
Engineering, Tokyo University of Agriculture and Technology, Koganei-shi, Tokyo
184-8488, Japan.
0040-4039/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2010.05.012

K. Sakurai, D. Kahne / Tetrahedron Letters 51 (2010) 3724–3727
3725
Figure 1. (a) N-terminal p53 peptide (15–29), (b) MDM2-bound conformation of 15-mer p53 peptide (side chains other than Phe, Trp, and Leu are omitted for clarity),7 (c)
functionalized trisaccharide as a p53-peptide mimic, (d) predicted conformation for the functionalized trisaccharide 1a (anomeric phenylsulfide group is omitted for
simplicity).
The synthesis of the target trisaccharides 1a–d was planned
as illustrated in Figure 2. As a preliminary effort to pursue a syn-
thetic route which provides access to several analogues of 1a, we
also designed derivatives 1b–d, which bear different Trp side
chain analogues. For the stepwise assembly of the monomer
sugars (3–5; A, B and C rings, respectively), the sulfoxide glyco-
sylation reaction was employed as an efficient coupling method-
ology.5,12–14
Each monomer (3–5) was first prepared with appropriate hydro-
phobic side chain functionalities and protecting groups (Scheme 1).
The A ring 3 and the C ring 5 required site-specific derivatization of
the C6 hydroxyl groups, whereas the B ring 4 required derivatization
Figure 2. Retrosynthesis of the trisaccharide-based p53 peptide mimic (1a–d).

3726
K. Sakurai, D. Kahne / Tetrahedron Letters 51 (2010) 3724–3727
Scheme 1. Synthesis of the monomers. Reagents and conditions: (a) H2O, reflux; (b) NaN3, 60% AcOH–H2O; (c) Ac2O, Py, (d) PhSH, BF3ÁEt2O, CH2Cl2, 60%; (e) NaOMe, MeOH;
(f) Ph3P, DEAD, p-nitrobenzoic acid, toluene; (g) NaOMe, MeOH, 80%; (h) BzCl, Py, CH2Cl2, À30 °C to À5 °C, 84%; (i) (CH3)2CHCH2COCl, Py, CH2Cl2, 73%; (j) TESOTf, 2,6-lutidine,
CH2Cl2, À20 °C, 96%; (k) mCPBA, CH2Cl2; (l) Ac2O, Py; (m) PhSH, BF3ÁEt2O, CH2Cl2; (n) NaOAc, MeOH, (
a/b = 6:1), 96%; (o) TBDPSCl, imidazole, DMF, 78%; (p) (Bu2Sn)2O,
benzene; (q) PMBCl, Bu4NI, benzene, 80%; (r) TMSOTf, Et3N, CH2Cl2, À78 °C to À20 °C, 89%; (s) mCPBA, CH2Cl2, À78 °C.
of the C3 hydroxyl group. Synthesis of the common precursor 9 for
the A and C rings started with peracetylated -glucal 6. Ferrier rear-
steps from 9 with selective C6–OH acylation with isovaleryl group (–
COCH2CH(CH3)2), and protection of C4–OHas a TESether and mCPBA
oxidation.
D
rangement of 6 followed by 1,4-addition of azide group and acetyla-
tion generated 3-azido-2,3-dideoxy-glucose 7.15,16 Installation of
anomeric phenylsulfide followed by deacetylation provided 8. The
C4 and C6 hydroxyl groups of 8 simultaneously reacted under the
Mitsunobuconditionto afford galactosederivative 9 after hydrolysis
of the resulting p-nitrophenyl groups. Selective benzoylation of C6–
OH of 9 then provided the A ring 3. The C ring 5 was obtained in three
The B ring 4 was prepared from commercially available 2-deoxy
D-galactose 12. Compound 12 was first peracetylated and then con-
verted into phenyl thioglycoside, which was followed by deacety-
lation and selective protection of the C6–OH as TBDPS ether to give
diol 13. The C3 hydroxyl group of 13 was selectively protected with
p-methoxy benzyl group via 3,4-stannylate intermediate.
Scheme 2. Assembly of functionalized trisaccharides. Reagents and conditions: (a) 4, Tf2O, DTBMP, CH2Cl2, 4-allyl-1,2-dimethylbenzene, À78 °C, 75%; (b) PPTS, THF–MeOH,
96%; (c) 5, Tf2O, DTBMP, CH2Cl2, 4-allyl 1,2-dimethylbenzene, À78 °C, 78%; (d) DDQ, CH2Cl2–H2O, 0 °C to rt, 80%; (e) R3CO2H, DIC, DMAP, CH2Cl2 (17a, 80%, 17b, 99%, 17c, 79%,
17d, 95%); (f) TASF, DMF (18a, 70%, 18b, 77%, 18c, 83%, 18d, 92%); (g) polymer-supported Ph3P, THF–H2O, 50 °C (1a, 22%, 1b, 14%, 1c, 55%, 1d, 26%).

K. Sakurai, D. Kahne / Tetrahedron Letters 51 (2010) 3724–3727
3727
Protection of the C4 hydroxyl group as TMS ether and then oxida-
tion with mCPBA provided the B ring 4.
ity to the hydrophobic pockets of MDM2 with rigid backbone
structures and precisely positioned hydrophobic side chains.19–23
Therefore, additional structure-activity studies on the side chain
functionalities for the trisaccharide scaffold would be desired to
achieve the steric requirements for binding the MDM2 pocket. De-
tailed structural analysis on the present trisaccharide scaffold
would also facilitate further rational optimization of the design
of a helical p53-peptide mimic.
The A–C rings (3–5) were then assembled stepwise from the
reducing end to the non-reducing end using the sulfoxide-medi-
ated glycosylation reaction (Scheme 2).5,12–14 Both steps of the gly-
cosylation were achieved in good yields with corresponding
desired products as a single stereoisomer (78% for the A–B disac-
charide 15 and 75% for the A–B–C trisaccharide 2). For the glycosyl-
ation reaction between the A ring 3 and the B ring 4, small amounts
of the A–B–B trisaccharide was also isolated. This observation was
consistent with the earlier finding in our laboratory that the TMS
group could be cleaved during the sulfoxide glycosylation reac-
tion.5,13,14 Oxidative removal of the PMB group provided the trisac-
charide precursor 16. The alcohol 16 was furnished with four types
of Trp side chain analogues (17a–d) via esterification, which were
synthesized with good yields (79–99%). The desired trisaccharides
were to be achieved simply by a removal of a TBDPS group and the
reduction of azide groups. These final two steps, however, were
met with unexpected difficulties. First, the standard deprotecting
conditions for a TBDPS group using HF or TBAF failed to give the
desired diol (18a–d) without significant decomposition. The pres-
ence of bulky groups at the B ring presumably prohibited access
of the reagents to TBDPS group for desilylation. TAS-F in DMF
was thus used to deprotect the TBDPS group, which was success-
fully achieved with good yields (70–92%).17 The final step for the
reduction of the azide groups was performed in the presence of
polymer-supported triphenyl phosphine in the aqueous THF. The
resulting crude product was purified by reverse-phase HPLC as
TFA salts. While the use of resin-bound reducing agent greatly
facilitated the subsequent purification step of the products, the
overall yields were modest (14–55%). This is presumably due to
either inefficient hydrolysis of the iminophosphorane intermedi-
ates on resin or absorption of diamine products to the resin.
The inhibitory activity of the functionalized trisaccharides
against the p53-MDM2 interaction was evaluated by an inhibition
ELISA. In this assay, the wt-p53 15-mer peptide inhibited the p53-
In conclusion, we designed and synthesized functionalized
a-
1,4-linked trisaccharides modeled after a 15-residued helical re-
gion of p53. The synthesis of the target trisaccharides was achieved
by using the sulfoxide glycosylation reaction as the key methodol-
ogy. Two of the functionalized trisaccharides were found to weakly
inhibit the p53-MDM2 interaction. Our synthesis demonstrated an
efficient assembly of a trisaccharide scaffold, which should be use-
ful for further development of trisaccharide-based
for MDM2 binding.
a-helix mimics
Acknowledgment
This work was supported by National Institute of Health Grants
69721.
References and notes
1. Meutermans, W.; Le, G. T.; Becker, B. ChemMedChem 2006, 1, 1164. See
references therein.
2. Jensen, K. J.; Brask, J. Biopolymers 2005, 80, 747. See references therein.
3. Suhara, Y.; Izumi, M.; Ichikawa, M.; Penno, M.; Ichikawa, Y. Tetrahedron Lett.
1997, 38, 7167.
4. Hoffmann, R. W. Angew. Chem., Int. Ed. 2000, 39, 2054.
5. Xuereb, H.; Maletic, M.; Gildersleeve, J.; Pelczer, I.; Kahne, D. J. Am. Soc. Chem.
2000, 122, 1883.
6. Davis, J. M.; Tsou, L. K.; Hamilton, A. D. Chem. Soc. Rev. 2007, 36, 326.
7. Kussie, P. H.; Gorina, S.; Marechal, V.; Elenbaas, B.; Moreau, J.; Levine, A. J.;
Pavletich, N. P. Science 1996, 274, 948.
8. Limieux, R. U.; Koto, S. Tetrahedron 1974, 30, 1933.
9. A crystal structure of galabiose (O-a-D-galactopyranosyl-1,4-D-galactopyranose)
shows an overall structure in support of our conformational analysis. See:
Svensson, G.; Albertsson, J.; Svensson, C.; Magnusson, G.; Dahmen, J. Carbohydr.
Res. 1986, 146, 29.
MDM2 interaction at an IC50 value of 4 lM, which was consistent
with the literature precedence.18 Two of the functionalized trisac-
charides (1c–d) showed weak inhibitory activity against the p53-
MDM2 interaction, while compounds 1a and 1b were inactive.
The quinoline-substituted trisaccharide 1d inhibited with an IC50
10. 2-Deoxy sugar was used for synthetic ease and for removal of potential steric
hinderance at the trisaccharide-MDM2 interface.
11. In our preliminary studies, related trisaccharides functionalized with
hydrophobic side chains exhibited solubility problems in aqueous solution.
12. Kahne, D.; Walker, S.; Cheng, Y.; Vanengen, D. J. Am. Chem. Soc. 1989, 111, 6881.
13. Gildersleeve, J.; Pascal, R. A., Jr.; Kahne, D. J. Am. Chem. Soc. 1998, 120, 5961.
14. Gildersleeve, J.; Smith, A.; Sakurai, K.; Rhagavan, S.; Kahne, D. J. Am. Chem. Soc.
1999, 121, 6176.
of 550
lM whereas the 3-bromoindole-bearing trisaccharide 1c
gave IC50 of 850
l
M. This finding was rather unexpected since 1a
and 1b display the same side chain functionalities as those of the
p53 peptide or its known high-affinity analogue.19 Our results sug-
gested that the presentation of the hydrophobic side chains on the
trisaccharide scaffold was not optimal for the interaction with
MDM2. This may be due to not only the scaffold conformation
but also the side chain conformation and positioning. The interac-
tion between the p53 peptide and MDM2 occurs through a double
induced-fit mechanism, which requires a highly complementary
interface to overcome the entropic penalty incurred. Recent struc-
tural studies on the p53 analogs complexed with MDM2 showed
that the inhibitor molecules fulfill the strict steric complementar-
15. Florent, J. C.; Monneret, C. J. Chem. Soc., Chem. Commun. 1987, 1171.
16. Daley, L.; Roger, P.; Monneret, C. J. Carbhydr. Chem. 1997, 16, 25.
17. Scheidt, K. A.; Chen, H.; Follows, B. C.; Chemler, S. R.; Coffey, D. S.; Roush, W. R.
J. Org. Chem. 1998, 63, 6436.
18. Sakurai, K.; Chung, H.-S.; Kahne, D. J. Am. Chem. Soc. 2004, 126, 16288.
19. Garcia-Echeverria, C.; Chène, P.; Blommers, M. J. J.; Furet, P. J. Med. Chem. 2000,
43, 3205.
20. Vassilev, L. T.; Vu, B. T.; Graves, B.; Carvajal, D.; Podlaski, F.; Filipovic, Z.; Kong,
N.; Kammlott, U.; Lukacs, C.; Klein, C.; Fotouhi, N.; Liu, E. A. Science 2004, 303,
844.
21. Grasberger, B. L. et al J. Med. Chem. 2005, 48, 909.
22. Sakurai, K.; Schubert, C.; Kahne, D. J. Am. Chem. Soc. 2006, 128, 11000.
23. Fasan, R.; Dias, R. L.; Moehle, K.; Zerbe, O.; Obrecht, D.; Mittl, P. R.; Grütter, M.
G.; Robinson, J. A. ChemBioChem 2006, 7, 515.