Toward fully synthetic glycoproteins by ultimately convergent routes: A solution to a long-standing problem

Article abstract of DOI:10.1021/ja0491836

A method is disclosed for the convergent synthesis of multiply glycosylated peptides. The approach centers on a convergent technique for generating masked, complex glycopeptide-containing C-terminal acyl donors. Activation of the latent donor in situ and

Full text of DOI:10.1021/ja0491836

Published on Web 05/07/2004
Toward Fully Synthetic Glycoproteins by Ultimately Convergent Routes: A
Solution to a Long-Standing Problem
J. David Warren,† Justin S. Miller,† Stacy J. Keding,† and Samuel J. Danishefsky*,†,‡
Laboratory for Bioorganic Chemistry, Sloan-Kettering Institute for Cancer Research, 1275 York AVenue,
New York, New York 10021, and Department of Chemistry, Columbia UniVersity, HaVemeyer Hall, 3000 Broadway,
New York, New York 10027
Received February 13, 2004; E-mail: s-danishefsky@ski.mskcc.org
Glycoproteins are important biomacromolecules that are bio-
synthesized through posttranslational glycosylation of newly fash-
ioned proteins emerging from the ribosome. A long-term goal of
the growing field of chemistry-based glycobiology is the delineation
of the consequences of glycosylation on critical properties such as
protein folding, proteolytic stability, and cell adhesion.1 Such
insights could explain why nature bothers to glycosylate otherwise
functional proteins. Moreover, glycoproteins have potentially
important clinical roles in the contexts of vaccines,2 diagnostics,3
and therapeutics.4 Indeed, erythropoietin (1) (see Figure 1), albeit
a heterogeneous glycoprotein,5 is clinically valuable as a treatment
for anemia, among other indications.4a
In Scheme 1 we sketch out a milestone challenge en route to
these long-term goals. In this synthesis, we hope to generate
glycopeptides 4 and 7 in a convergent way from their respective
glycan and peptide precursors, and to join them with full conver-
gence to reach 8.
Scheme 1. Convergent Synthesis of Bifunctional Glycopeptides
Figure 1. Erythropoietin contains an O-linkage at Ser126 and N-linkages
at Asn24, Asn38, and Asn83.6 Coordinates obtained from the Protein Data
Bank (PDB accession no. 1BUY)7,8 were used to render the protein backbone
with RasMol; schematic depictions of the glycans were added by the authors.
The carbohydrate structures are representative,9 as glycan compositions can
vary significantly in different glycoforms of erythropoietin.5
A series of technical obstacles would have to be overcome to
accomplish such a goal. A particularly serious issue is the synthesis
of a suitable glycopeptide acyl donor for ligation with 7. The first
possibility that comes to mind as an appropriate acyl donor is one
in which 4 is a thioester. However, many difficulties would have
to be surmounted to reach a thioester such as 4 by convergent
means.16
It was in pondering this problem that an attractive possibility
presented itself (Scheme 2). Our subgoal would be the synthesis
of a phenolic ester equipped with an unsymmetrical disulfide (cf.
10). The phenolic ester linkage would in itself probably not manifest
sufficient reactivity to serve as a viable acylating agent.17 However,
equipped as it would be with an ortho disulfide moiety, the phenolic
ester could operate as an incipient acyl donor. Thus, reduction of
the disulfide in 12, as shown, might well set the stage for in situ
elaboration into a thioester (cf. 13f14) by OfS acyl migration.
Thioester 14, present in an unfavorable, but dynamic equilibrium,
could perhaps be interdicted by the machinery appropriate to native
chemical ligation (NCL),12c ultimately culminating in 8.
The isolation of homogeneous glycoproteins from natural sources
in significant quantity is extremely difficult.10 Formidable as the
challenges are, we felt it worthwhile to explore the prospects of
reaching homogeneous glycopeptides and, in time, glycoproteins
through the medium of chemical synthesis. Since many naturally
occurring, medicinally important glycoproteins (cf., for example,
erythropoietin4a and gp12011) display multiple glycosylation sites
containing large oligosaccharide domains, we insisted on stringent
standards of convergence and stereoselectivity.
Following extensive earlier studies, wherein our chemistry was
interfaced with critical findings from other laboratories,12 we
demonstrated the ability to ligate fully synthetic N-linked glyco-
peptides containing various complex oligosaccharides to larger
polypeptide domains.13 We were thus able to obtain major glyco-
sylated subsections of prostate specific antigen (PSA)3a and gp120.14
Our next goal, with an eye toward erythropoietin, was the
achievement of a convergent approach to reach polypeptides that
are multiply glycosylated at fixed sites.15
While the proposal for the intermediacy of a discrete thioester
14 is certainly attractive, it is not necessarily an obligatory event
en route to ligation. Thus, as suggested in structure 13, the presence
of a free ortho benzene thiol function might well enhance the
acylating facility of the phenolic ester,18 perhaps through in situ
† Sloan-Kettering Institute for Cancer Research.
‡ Columbia University.
9
6576
J. AM. CHEM. SOC. 2004, 126, 6576-6578
10.1021/ja0491836 CCC: $27.50 © 2004 American Chemical Society

C O M M U N I C A T I O N S
Scheme 4. Convergent Coupling of Functionalized Glycopeptides
hydrogen bonding or anchimeric assistance. In either case, the
sequence is set into motion only upon reductive cleavage of the
disulfide.
The synthesis of the masked glycopeptide acyl donor was
accomplished as follows (Scheme 2). Commercially available
2-mercaptophenol (9) was oxidized to its symmetrical disulfide.
Subsequent exchange with excess ethyl disulfide19 furnished an
unsymmetrical, aryl-alkyl disulfide. Acylation of the phenolic
hydroxyl group with Boc-phenylalanine followed by acidic cleavage
of the Boc protecting group afforded phenylalanine derivative 10.
Standard peptide coupling of 10 with a fully protected peptide acid20
followed by global deprotection provided 11, which contains a free
aspartate residue. Aspartylation of a chitobiose-derived glycosyl-
amine with 11 afforded the virtual acyl donor 12, corresponding
to our target structure 4 (Scheme 1).
concomitant increase in the desired product 19. Only one compound
of the appropriate mass was observed by LCMS and H NMR,
again indicating a lack of epimerization.
1
Scheme 2. Structure, Synthesis, and Proposed Mechanism of a
We then directed our attention to evaluation of the generality of
our findings by first establishing applicability to incorporation of
O-linked domains,22 as shown through the synthesis of 20. Notably,
we also assembled the histidine-containing system 21 (see Figure
2). Since activated C-terminal histidine residues are known to be
particularly susceptible to epimerization,23 we employed a prophy-
lactic dinitrophenyl (DNP) protecting group24 for the imidazole
τ-nitrogen during the synthesis. As expected, the DNP group was
cleaved concurrently during the mildly nucleophilic coupling
reaction. Gratifyingly, no indications of epimerization were observed
Masked Glycopeptide Acyl Donora
1
by LCMS or H NMR.
a Reagents and conditions: (a) I2, MeOH, H2O; (b) BF3‚OEt2, EtSSEt,
CH2Cl2, 99%, 2 steps; (c) Boc-Phe-OH, EDCI, DMAP, CH2Cl2/THF, 93%;
(d) 4 N HCl/dioxane, 94%; (e) Fmoc-Arg(Pbf)-Asp(tBu)-Arg(Pbf)-Ser(tBu)-
Gly-OH, HATU, DIEA, DMF, 61%; (f) TFA/phenol/Et3SiH/H2O,
35:2:1:1, 60%; (g) GlcNAcâ1f4GlcNAcâ1-NH2, HATU, DIEA, DMSO,
52%.
Figure 2. Bifunctional glycopeptides containing O-linkages.
As a preliminary probe of the practicability of this scheme,
phenylalanine derivative 10 was subjected to the neutral, aqueous,
reducing reaction conditions described above, along with free
cysteine (Scheme 3). The desired free dipeptide Phe-Cys (17) was
obtained in 78% isolated yield. Only a single product was observed
by LCMS and 1H NMR, strongly suggesting that epimerization at
the R-carbon of phenylalanine did not occur.
Figure 3. Anomeric â-glycosylamines (22 and 23) that differ at a single
stereocenter (asterisks), and a bifunctional glycopeptide (24) containing one
of these structurally complex glycans.
Scheme 3. Demonstration of the Latent Acyl Donor Function
The next phase of our investigation involved validation of the
highly convergent methodology in the context of complex glycans.
Previously we had disclosed syntheses of glycosylamines 2213 and
2325 (Figure 3). It will be noted that these compounds differ in a
single stereogenic locus in the interior C-ring (see asterisks). Thus,
22 corresponds to a fully stereochemically competent N-linked high
mannose oligosaccharide. In contrast, its C-ring C-2 epimer is an
unnatural stereochemical mutant.
Preparation of pentasaccharide-derived glycopeptide coupling
partners proceeded via Lansbury aspartylation12b of 22 and 23 with
25 and 26, affording 27 and 28, respectively. As a prelude to the
finale, coupling of 28 with a relatively simple chitobiose-containing
acyl donor gave 24 (Figure 3) in 77% yield.
Given this encouraging result, we continued toward our goal of
bifunctional glycopeptides. Thus, 12 was treated with excess sodium
2-mercaptoethanesulfonate (MES-Na) at neutral pH in aqueous
phosphate buffered saline (PBS) in the presence of 18,21 (Scheme
4). As observed by LCMS during the reaction, 12 was almost
instantaneously (less than 3 min) converted to the thioester derived
from MES-Na (cf. 15); very little hydrolysis to the free carboxylic
acid was noted. Over the course of the next several hours, we
observed a gradual decrease in the amount of this thioester, with a
9
J. AM. CHEM. SOC. VOL. 126, NO. 21, 2004 6577

C O M M U N I C A T I O N S
Scheme 5. Convergent Coupling of Two Functionalized
Kenemans, P. Int. J. Biol. Markers 2000, 15, 343-356.
(3) (a) Dudkin, V. Y.; Miller, J. S.; Danishefsky, S. J. J. Am. Chem. Soc.
2004, 126, 736-738. (b) Peracaula, R.; Tabares, G.; Royle, L.; Harvey,
D. J.; Dwek, R. A.; Rudd, P. M.; de Llorens, R. Glycobiology 2003, 13,
457-470.
Glycopeptides Containing Distinct, Structurally Complex Glycansa
(4) (a) Ridley, D. M.; Dawkins, F.; Perlin, E. J. Natl. Med. Assoc. 1994, 86,
129-135. (b) Durand, G.; Seta, N. Clin. Chem. 2000, 46, 795-805. (c)
Koeller, K. M.; Wong, C. H. Nat. Biotechnol. 2000, 18, 835-841.
(5) Rush, R. S.; Derby, P. L.; Smith, D. M.; Merry, C.; Rogers, G.; Rohde,
M. F.; Katta, V. Anal. Chem. 1995, 67, 1442-1452.
(6) Lai, P. H.; Everett, R.; Wang, F. F.; Arakawa, T.; Goldwasser, E. J. Biol.
Chem. 1986, 261, 3116-3121.
(7) Berman, H. M.; Westbrook, J.; Feng, Z.; Gilliland, G.; Bhat, T. N.;
Weissig, H.; Shindyalov, I. N.; Bourne, P. E. Nucleic Acids Res. 2000,
28, 235-242.
(8) Cheetham, J. C.; Smith, D. M.; Aoki, K. H.; Stevenson, J. L.; Hoeffel, T.
J.; Syed, R. S.; Egrie, J.; Harvey, T. S. Nat. Struct. Biol. 1998, 5, 861-
866.
(9) Rahbek-Nielsen, H.; Roepstorff, P.; Reischl, H.; Wozny, M.; Koll, H.;
Haselbeck, A. J. Mass Spectrom. 1997, 32, 948-958.
(10) (a) Kornfeld, R.; Kornfeld, S. Annu. ReV. Biochem. 1985, 54, 631-664.
(b) Roth, J. Chem. ReV. 2002, 102, 285-303.
(11) Geyer, H.; Holschbach, C.; Hunsmann, G.; Schneider, J. J. Biol. Chem.
1988, 263, 11760-11767.
(12) (a) Likhosherstov, L. M.; Novikova, O. S.; Derevitskaja, V. A.; Kochetkov,
N. K. Carbohydr. Res. 1986, 146, C1-C5. (b) Cohen-Anisfeld, S. T.;
Lansbury, P. T. J. Am. Chem. Soc. 1993, 115, 10531-10537. (c) Dawson,
P. E.; Muir, T. W.; Clark-Lewis, I.; Kent, S. B. H. Science 1994, 266,
776-779.
(13) Miller, J. S.; Dudkin, V. Y.; Lyon, G. J.; Muir, T. W.; Danishefsky, S. J.
Angew. Chem., Int. Ed. 2003, 42, 431-434.
(14) (a) Mandal, M.; Dudkin, V. Y.; Geng, X.; Danishefsky, S. J. Angew.
Chem., Int. Ed. 2004, 43, 2557-2561. (b) Geng, X.; Dudkin, V. Y.;
Mandal, M.; Danishefsky, S. J. Angew. Chem., Int. Ed. 2004, 43, 2562-
2575.
(15) For the state of the art in biological approaches toward multiply
glycosylated proteins, see: Zhang, Z.; Gildersleeve, J.; Yang, Y. Y.; Xu,
R.; Loo, J. A.; Uryu, S.; Wong, C. H.; Schultz, P. G. Science 2004, 303,
371-373.
a Reagents and conditions: (a) 25 or 26, HATU, DIEA, DMSO, 50%
for 27 (from 22), 71% for 28 (from 23); (b) 20% piperidine in DMF, 63%;
(c) 0.2 M phosphate, 0.2 M NaCl, pH ∼7.4, excess MES-Na, 75%.
With this demonstration secure, we now addressed the critical
coupling of 27 and 28, shown in Scheme 5. This reaction was
accomplished smoothly upon reductive cleavage of the disulfide
linkage in 27, affording 29 as shown. Aside from validating the
methodology in a striking way, the synthesis of 29 serves to pinpoint
the power of the total synthesis approach. Since one of its
carbohydrates is unnatural, 29 cannot readily be obtained from
natural sources or via enzymatic manipulations.
(16) Significant advances in this type of problem arose through the use of
Ellman’s Fmoc-based sulfonamide linker (Shin, Y.; Winans, K. A.; Backes,
B. J.; Kent, S. B. H.; Ellman, J. A.; Bertozzi, C. R. J. Am. Chem. Soc.
1999, 121, 11684-11689), which has been employed in the synthesis of
glycopeptide thioesters. The method we practice here, however, insists
on maximal convergence, as opposed to a “cassette” approach to glycan
incorporation. We note that certain glycosidic linkages (e.g., fucosidic
linkages in erythropoietin) are particularly unstable toward the acidic
conditions (TFA) required for protecting group cleavage. Other “cassette”-
based Fmoc solid-phase peptide synthesis (SPPS) techniques potentially
leading to glycopeptide thioesters also include acidic conditions at some
point. These methods involve alteration of the Fmoc deblocking conditions
[(a) Li, X. Q.; Kawakami, T.; Aimoto, S. Tetrahedron Lett. 1998, 39,
8669-8672. (b) Clippingdale, A. B.; Barrow, C. J.; Wade, J. D. J. Pept.
Sci. 2000, 6, 225-234. (c) Hojo, H.; Haginoya, E.; Matsumoto, Y.;
Nakahara, Y.; Nabeshima, K.; Toole, B. P.; Watanabe, Y. Tetrahedron
Lett. 2003, 44, 2961-2964], or direct conversion into thioesters of
C-terminal acids (von Eggelkraut-Gottanka, R.; Klose, A.; Beck-Sickinger,
A. G.; Beyermann, M. Tetrahedron Lett. 2003, 44, 3551-3554), esters
(Swinnen, D.; Hilvert, D. Org. Lett. 2000, 2, 2439-2442), or trithioortho
esters (Brask, J.; Albericio, F.; Jensen, K. J. Org. Lett. 2003, 5, 2951-
2953). Boc-based SPPS employs strongly acidic cleavage conditions
(liquid HF) that are incompatible with many glycosidic linkages.
(17) A phenolic ester may function as an acylating agent when employed in
conjunction with an intramolecularly disposed amine equivalent that is
generated in situ via Staudinger chemistry. For examples, see: (a) Nilsson,
B. L.; Kiessling, L. L.: Raines, R. T. Org. Lett. 2000, 2, 1939-1941. (b)
Saxon E.: Armstrong, J. I.; Bertozzi, C. R. Org Lett. 2000, 2, 2141-
2143.(c) Bianchi, A.; Bernardi, A. Tetrahedron Lett. 2004, 45, 2231-
2234.
In summary, the goal set forth in Scheme 1sa convergent method
for the synthesis of bifunctional glycopeptidesshas been met. The
mechanistic rationale set forth in Scheme 2 is supported by the
identification of the MES-Na thioester as a reaction intermediate.
Furthermore, potential hydrolysis and C-terminal epimerization of
the glycopeptide acyl donor have been suppressed. It seems likely
that the method and logic set forth above will enjoy application in
the building of complex glycopeptides of biological and even
medicinal consequence.
Acknowledgment. This work was supported by the NIH
[CA103823 (formerly AI16943)]. We thank Dr. George Sukenick
1
(NMR Core Facility, CA02848) for help with H NMR analyses
and Ms. Anna Dudkina and Ms. Sylvi Rusli for help with mass
spectral analyses and HPLC separations. We also thank Dr. Vadim
Dudkin for helpful discussions. Postdoctoral fellowship support is
gratefully acknowledged by J.D.W. (NIH, CA62948), J.S.M. (U.S.
Army Prostate Cancer Research Program, PC020147), and S.J.K.
(NIH, AI051883).
Supporting Information Available: Experimental procedures and
compound characterization data, including LCMS and NMR data (PDF).
This material is available free of charge via the Internet at http://
pubs.acs.org.
(18) See: Verhaeghe, J.; Lacassie, E.; Bertrand, M.; Trudelle, Y. Tetrahedron
Lett. 1993, 34, 461-464 and references therein.
(19) Caserio, M. C.; Fisher, C. L.; Kim, J. K. J. Org. Chem. 1985, 50, 4390-
4393.
(20) The protected peptide acid was synthesized by Fmoc SPPS.
(21) Glycopeptide 18 was prepared as described previously (ref 13).
(22) O-Linked glycopeptide precursors to 20 and 21 were synthesized using a
cassette approach, in which the Fmoc serine monomers used in SPPS
contained pendant saccharides.
(23) Kemp, D. S. In The Peptides: Analysis, Synthesis, Biology; Gross, E.,
Meienhofer, J., Eds.; Academic Press: New York, 1979; Vol. 1, Part A,
pp 315-381.
References
(1) (a) Imperiali, B.; O’Connor, S. E.; Hendrickson, T.; Kellenberger, C. Pure
Appl. Chem. 1999, 71, 777-787. (b) Lis, H.; Sharon, N. Eur. J. Biochem.
1993, 218, 1-27. (c) Rudd, P. M.; Elliott, T.; Cresswell, P.; Wilson, I.
A.; Dwek, R. A. Science 2001, 291, 2370-2376. (d) Bertozzi, C. R.;
Kiessling, L. L. Science 2001, 291, 2357-2364.
(2) (a) Calarese, D. A.; Scanlan, C. N.; Zwick, M. B.; Deechongkit, S.;
Mimura, Y.; Kunert, R.; Zhu, P.; Wormald, M. R.; Stanfield, R. L.; Roux,
K. H.; Kelly, J. W.; Rudd, P. M.; Dwek, R. A.; Katinger, H.; Burton, D.
R.; Wilson, I. A. Science 2003, 300, 2065-2071. (b) von Mensdorff-
Pouilly, S.; Snijdewint, F. G.; Verstraeten, A. A.; Verheijen, R. H.;
(24) Shaltiel, S.; Fridkin, M. Biochemistry 1970, 9, 5122-5127.
(25) Wang, Z. G.; Zhang, X. F.; Live, D.; Danishefsky, S. J. Angew. Chem.,
Int. Ed. 2000, 39, 3652-3656.
JA0491836
9
6578 J. AM. CHEM. SOC. VOL. 126, NO. 21, 2004