Chemical synthesis of biologically active monoglycosylated GM2-activator protein analogue using N-sulfanylethylanilide peptide

Article abstract of DOI:10.1002/anie.201303390

Going to SEA(lide): Total chemical synthesis of a 162-residue glycoprotein analogue of the monoglycosylated human GM2-activator protein (GM2AP) was achieved. Key steps were the use of N-sulfanylethylanilide (SEAlide) peptides in the kinetic chemical ligation synthesis of a large peptide fragment, and a convergent native chemical ligation for final fragment assembly. Copyright

Full text of DOI:10.1002/anie.201303390

Angewandte
Chemie
DOI: 10.1002/anie.201303390
Synthesis Design
Chemical Synthesis of Biologically Active Monoglycosylated GM2-
Activator Protein Analogue Using N-Sulfanylethylanilide Peptide**
Kohei Sato, Akira Shigenaga, Keisuke Kitakaze, Ken Sakamoto, Daisuke Tsuji, Kohji Itoh, and
Akira Otaka*
Protein therapeutics containing glycoproteins have been
attracting increasing attention as potential agents for treating
diseases that were once thought to be incurable.[1] In
therapeutics development using proteins, a critical issue is
how protein molecules are produced. Genetic engineering
procedures represent a significant advance in the production
of naturally occurring proteins, however, medicinal-chemis-
try-oriented examination of protein therapeutics incorporat-
ing unnatural structural units is far from satisfactory using
genetic protocols. One alternative to genetic protocols is
chemical synthesis of proteins in which the protein backbone
is constructed using ligation chemistry such as native chemical
ligation (NCL).[2] Chemically synthesized proteins could be
useful in medicinal chemistry.[3] Recent advances in NCL have
placed the chemical synthesis of small- and medium-sized
proteins within our reach. In particular, combinations of
sequential and convergent NCL protocols provide technical
improvements for the protein synthesis.[4] A representative
sequential NCL protocol uses a kinetically controlled NCL
(KCL)[5] developed by Kent and co-workers, in which
a peptide chain can be elongated based on the differences
between the reactivities of aryl and alkyl thioesters in a one-
pot, N-to-C-directed process.[6,7] The KCL protocol has
enjoyed great success in providing the large N-segment
alkyl thioesters for convergent synthesis. However, KCL
reactions using highly reactive alkyl thioesters are ineffi-
cient.[5c]
in the presence of phosphate salts and participates in NCL to
yield 3 (Scheme 1).[7,9]
Scheme 1. Reactivity of SEAlide peptide as crypto-thioester is tuned by
phosphate salts.
Herein, we address the chemical synthesis of the mono-
glycosylated GM2-activator protein (GM2AP) analogue 4
using SEAlide peptides. GM2AP is an essential cofactor for
the lysosomal degradation of ganglioside GM2 by b-hexosa-
minidase A (HexA).[10] Functional deficiencies of GM2AP
result in a fatal neurological disease, and thus we explored
a convergent synthesis platform compatible with various
GM2AP analogues to develop a medical treatment. Gener-
ally, synthetic strategies based on the incorporation of a sugar
chain acceptor residue, such as monoglycosylated asparagine,
and subsequent site-selective transfer of a sugar chain on the
acceptor site have been used for glycoprotein synthesis.[11,12]
GM2AP consists of 162 residues with N-linked glycans at
Asn32 (Figure 1). Our route to a monoglycosylated GM2AP
analogue relies on replacement of Asn32 by Cys, on which an
N-acetylglucosamine moiety is incorporated by an S alkyla-
tion of Cys32 with iodoacetyl-N-acetylglucosamine.[13] The
use of the S-alkylation protocol, which provides one addi-
tional ligation site, facilitates the SPPS of the peptide
fragments. The NCL-mediated construction of naturally
occurring GM2AP requires the synthesis of a thioester frag-
ment of at least 67 residues (8–74). Straightforward SPPS
does not always guarantee successful synthesis of such large
peptides (over 40–50 residues). Our convergent synthetic
strategy for the GM2AP analogue 4 is shown in Scheme 2.
Standard NCL condensation of the alkylthioester Fr 1 (5),
which has an S-acetamidomethyl (Acm) protecting group on
Cys8, with Fr 2 (6), which is S-Acm-protected at Cys68 and
the SEAlide moiety,[14] followed by S glycosylation at Cys32
In the course of developing synthetic protocols for peptide
thioesters using Fmoc solid-phase peptide synthesis (Fmoc
SPPS),[8] we found that the N-sulfanylethylanilide (SEAlide)
peptides 1, as crypto-thioesters, can be successfully used for
one-pot, N-to-C-directed ligation under kinetically controlled
conditions. The high kinetic control is achieved by selecting
an appropriate buffer salt (phosphate) in the reaction
medium, that is, the SEAlide unit 1 remains intact under
NCL conditions with N-terminal cysteinyl peptides 2 in the
absence of phosphate salts, whereas it functions as a thioester
[*] K. Sato, Prof. Dr. A. Shigenaga, K. Kitakaze, K. Sakamoto,
Prof. Dr. D. Tsuji, Prof. Dr. K. Itoh, Prof. Dr. A. Otaka
Institute of Health Bioscience and Graduate School of Pharma-
ceutical Sciences, The University of Tokushima
Shomachi, Tokushima 770-8505 (Japan)
E-mail: aotaka@tokushima-u.ac.jp
[**] This research was supported in part by a Grant-in-Aid for Scientific
Research (KAKENHI) and research grants from the Takeda Science
and the Uehara Memorial Foundations. K.S. is grateful for
a scholarship from the Yoshida Scholarship Foundation.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201303390.
Angew. Chem. Int. Ed. 2013, 52, 7855 –7859
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
7855

.
Angewandte
Communications
except for Cys32, were protected by Acm groups. Peptides
corresponding to 6 and 8 were synthesized on Fmoc-Ser(tBu)-
incorporated-N-acetoamidomethylsulfanylethylaniline[16] and
Fmoc-Gly-incorporated-N-triphenylmethylsulfanylethylani-
line-linked resin, respectively, using Fmoc SPPS. In situ
neutralization protocols were used in the Boc SPPS.[17]
The progress of the reactions for the synthesis of 10 are
summarized in the Supporting Information (SI-Figures 1 and
2). NCL for the synthesis of 10 from 5 and 6 in 6m guanidine
(Gn)·HCl-0.2m 3-[4-(2-hydroxyethyl)piperazin-1-yl]propane-
1-sulfonic acid (HEPPS) in the presence of 3% (v/v)
thiophenol, proceeded smoothly to afford the Acm-protected
74-residue peptide 12 (Fr 1–Fr 2 (Cys(SAcm)8,68, Cys(SH)32,
SEA(SAcm))) (SI-Figure 1). Glycosylation of the Cys32
peptide 12 with iodoacetyl-N-acetylglucosamine in 6m
Gn·HCl-0.2m HEPPS yielded the monoglycosylated S-Acm
peptide 13 (Fr 1–Fr 2 (Cys(SAcm)8,68, Cys(SGlcNAc)32, SEA-
(SAcm))). in 90% yield upon isolation. The Acm groups on
13 were removed with AgOTf in the presence of anisole in
TFA at 48C for 24 h,[18] followed by incubation in the presence
of 3% (w/v) dithiothreitol (DTT), and subsequent HPLC
purification to afford the desired 10 in 42% yield upon
isolation (SI-Figure 2).
Monitoring of the reactions for the synthesis of the C-half
segment 11 is summarized in Figure 2. The first ligation of 7
with 8 in 6m Gn·HCl-0.1m HEPPS, 50 mm tris(2-carboxy-
ethyl)phosphine (TCEP),[19] and 100 mm (4-carboxymethyl)-
thiophenol (MPAA),[20] in the absence of phosphate salts,
proceeded under kinetic conditions to yield the desired
ligated peptide (Fr 3–Fr 4; Figure 2B). A chemeoselective
reaction between the thioester group in 7 and the N-terminal
cysteinyl residue in 8 was achieved because the anilide-type
SEAlide moiety remained intact in the absence of phosphate
salts. A solution of 9 in 0.4m phosphate buffer was added to
Figure 1. Entire amino acid sequences of GM2AP and the correspond-
ing Asn32-mutated analogue. Five peptide fragments for the conver-
gent synthesis of the mutated analogue.
and subsequent removal of the Acm protections should afford
the monoglycosylated N-half segment 10 of the SEAlide
peptide. For the preparation of the C-half segment 11, one-
pot, N-to-C-directed sequential ligation using the SEAlide
peptide was used. The first NCL of the Fr 3 (7) containing N-
terminal 1,3-thiazolidine-4-carbonyl (Thz) and C-terminal
alkyl thioester with the N-terminal cysteinyl SEAlide peptide
Fr 4 (8) in the absence of phosphate salts, followed by
a second ligation of Fr 5 (9) in the presence of phosphate salts
and subsequent opening of the 1,3-thiazolidine ring[15] in the
Thz residue, gives the desired C-half segment 11. Standard
ligation between 10 and 11 in the presence of phosphate salts
and subsequent folding would provide the desired product.
The requisite peptide fragments, Fr 1 (5) and Fr 3 (7), and
Fr 2 (6), Fr 4 (8), and Fr 5 (9), were prepared by Boc and
Fmoc SPPS, respectively. To achieve the selective glycosyla-
tion, Cys8, Cyc68, and the SEAlide moiety in Fr 1 and Fr 2,
Scheme 2. Convergent synthetic strategy for the preparation of monoglycosylated GM2AP analogue. a) NCL. b) Glycosylation using alkylation
protocol. c) Removal of Acm groups. d) NCL in the absence of phosphate salts. e) NCL in the presence of phosphate salts. f) Opening of
thiazolidine ring. g) NCL in the presence of phosphate salts. h) Folding.
7856
www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 7855 –7859

Angewandte
Chemie
Figure 2. HPLC monitoring of reactions for the synthesis of the C-half
segment 11. A) First NCL, kinetically controlled conditions (t<1 min):
components 7 (2.0 mm) and 8 (2.0 mm) were ligated in Gn·HCl [6m,
HEPPS (0.1m), TCEP (50 mm), MPAA (100 mm), pH 7.0, 378C].
B) First NCL (t=3 h). C) Second NCL (t<1 min): To the above
reaction mixture was added a solution of the segment 9 (1.1 equiv) in
Gn·HCl [6m, Na phosphate (0.4m)]. [Final concentrations: Gn·HCl
(6m), Na phosphate (0.32m), HEPPS (20 mm), MPAA (20 mm), TCEP
(10 mm), segments 7 and 8 (0.4 mm each)]. D) Second NCL (t=24 h).
E) Opening of 1,3-thiazolidine ring (t=3 h): to the above reaction
mixture, 0.2m NH2OH·HCl was added. HPLC conditions: Cosmosil
5C18 AR-II column (4.6ꢀ250 mm) with a linear gradient of MeCN/
Figure 3. HPLC monitoring of NCL of the N-half segment 10 with 11
(A and B) and subsequent folding reaction (C). A) NCL (t=0):
components 10 (0.5 mm) and 11 (0.5 mm) were ligated in Gn·HCl
[6m, Na phosphate (0.5m), TCEP (50 mm), MPAA (50 mm), pH 6.0,
378C]. B) NCL (t=24 h). C) HPLC-purified reduced 4 [8 Cys(SH) form]
was folded in Gn·HCl [1m, Na phosphate (16 mm), Tris·HCl (42 mm),
reduced form glutathione (1.7 mm), oxidized form glutathione
(0.17 mm), Tween 20 (0.0025%, v/v), pH 8.0, 0.1 mgmLÀ1 protein].
HPLC conditions: Cosmosil Protein-R column (4.6ꢀ250 mm) with
a linear gradient of MeCN/0.1% aq. TFA (35:65 to 55:45 over 30 min)
at a flow rate of 1.0 mLminÀ1, detection at 220 nm. *=Ser74-epimer-
ized product (observed mass was identical to that of desired product).
0.1% aq. TFA (20:50–50:50 over 30 min) at a flow rate 1.0 mLminÀ1
detection at 220 nm.
,
the reaction mixture to initiate the second NCL of the
SEAlide moiety with the cysteinyl residue of 9 (Figure 2C).
This reaction also proceeded efficiently to yield the desired
Thz peptide (Fr 3–Fr 4–Fr 5; Figure 2D), which then under-
went opening of the 1,3-thiazolidine ring upon addition of
NH2OH·HCl, to give 11 in 47% yield after HPLC purification
(Figure 2E). Kinetically controlled one-pot, N-to-C sequen-
tial ligation was therefore successful in assembling three
peptide fragments (7, 8, and 9).
Convergent assembly of 10 and 11 was accomplished by
NCL in the presence of phosphate salts. The reaction of 10
with 11 in 6m Gd·HCl-0.5m Na phosphate, 50 mm TCEP,
50 mm MPAA, proceeded at a reasonable reaction rate and
almost went to completion within 24 hours to yield the
desired ligated 162-residue protein (reduced form of 4). After
HPLC purification, folding in the presence of reduced and
oxidized forms of glutathione was performed to yield the
folded monoglycosylated 162-residue GM2AP analogue 4.
The SEAlide-mediated ligation protocol is also undoubtedly
of use in convergent assembly, considering the successful
coupling of 74- and 88-residue peptides. HPLC analysis of the
coupling reaction indicated that a material with a molecular
weight identical to that of the desired ligation product was
obtained (Figure 3B, denoted by asterisk). From our previous
investigation of SEAlide-mediated coupling, ligation at the
Ser-SEAlide site was observed to accompany partial epime-
rization (9–10%).[16] Epimerization of a Ser thioester has also
been reported to occur under normal NCL conditions.[21]
Mass analysis of the synthetic material revealed that the
product has the expected molecular weight of the folded
monoglycosylated GM2AP analogue 4 (calcd 17830.4; found
17830.3). Circular dichroism (CD) analysis indicated that the
synthetic material has a spectrum similar to that reported in
the literature (see the Supporting Information).[22] In the
hydrolysis-assisted conversion of GM2 into GM3 in the
presence of Hex A, the synthetic sample was more active than
an E. coli-expressed carbohydrate-free sample (Figure 4).[23]
In conclusion, SEAlide peptides are versatile reagents in
chemical protein synthesis. In this study, SEAlide peptides
were successfully used in the preparation of the N-half and C-
half segments required for the convergent strategies for
chemical synthesis of proteins. The use of SEAlide peptides
allows one-pot, sequential NCL reactions to proceed with
high kinetic control. Successful convergent coupling indicated
that the SEAlide peptides function as useful synthetic units in
convergent synthesis. Establishment of a synthetic platform
for obtaining the Cys32 GM2AP analogue enables the
Angew. Chem. Int. Ed. 2013, 52, 7855 –7859
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7857

.
Angewandte
Communications
[5] a) D. Bang, B. L. Pentelute, S. B. H. Kent, Angew. Chem. 2006,
118, 4089 – 4092; Angew. Chem. Int. Ed. 2006, 45, 3985 – 3988;
b) T. Durek, V. Y. Torbeev, S. B. H. Kent, Proc. Natl. Acad. Sci.
USA 2007, 104, 4846 – 4851; c) D. Bang, J. Lee, Y. Kwon, B. L.
Pentelute, Bioconjugate Chem. 2011, 22, 1645 – 1649.
[6] a) N. Ollivier, J. Vicogne, A. Vallin, H. Drobecq, R. Desmet, O.
El Mahdi, B. Leclercq, G. Goormachtigh, V. Fafeur, O. Melnyk,
Angew. Chem. 2012, 124, 213 – 217; Angew. Chem. Int. Ed. 2012,
51, 209 – 213; b) R. Yang, W. Hou, X. Zhang, C.-F. Liu, Org. Lett.
2012, 14, 374 – 377; c) L. Raibaut, N. Ollivier, O. Melnyk, Chem.
Soc. Rev. 2012, 41, 7001 – 7015.
[7] a) K. Sato, A. Shigenaga, K. Tsuji, S. Tsuda, Y. Sumikawa, K.
Sakamoto, A. Otaka, ChemBioChem 2011, 12, 1840 – 1844; b) A.
Otaka, K. Sato, H. Ding, A. Shigenaga, Chem. Rec. 2012, 12,
479 – 490.
[8] S. Tsuda, A. Shigenaga, K. Bando, A. Otaka, Org. Lett. 2009, 11,
823 – 826.
[9] Other N-substituted type N-to-S acyl transfer devises can be
converted into the corresponding thioesters under acidic con-
ditions. See Ref. [6] and a) C. Ozawa, H. Katayama, H. Hojo, Y.
Nakahara, Org. Lett. 2008, 10, 3531 – 3533; b) K. S. A. Kumar,
S. N. Bavikar, L. Spasser, T. Moyal, S. Ohayon, A. Brik, Angew.
Chem. 2011, 123, 6261 – 6265; Angew. Chem. Int. Ed. 2011, 50,
6137 – 6141; c) S. N. Bavikar, L. Spasser, M. Haj-Yahya, S. V.
Karthikeyan, T. Moyal, K. S. Ajish Kumar, A. Brik, Angew.
Chem. 2012, 124, 782 – 787; Angew. Chem. Int. Ed. 2012, 51, 758 –
763.
[10] For a review of GM2AP, see: T. Kolter, K. Sandhoff, Annu. Rev.
Cell Dev. Biol. 2005, 21, 81 – 103.
[11] For recent reviews of glycoprotein synthesis, see: a) D. P.
Gamblin, E. M. Scanlan, B. G. Davis, Chem. Rev. 2009, 109,
131 – 163; b) C. Unverzagt, Y. Kajihara, Chem. Soc. Rev. 2013,
42, 4408 – 4420.
Figure 4. Hydrolysis of GM2 to GM3 with b-hexosamidase A (Hex A)
in the presence of GM2APs. A) Conversion of GM2 into GM3.
B) Comparable hydrolysis of GM2 with Hex A in the presence of
varying amounts of synthesized or expressed GM2AP. Hydrolysis was
monitored by TLC of the reaction aliquots (solvent: CHCl3/MeOH
0.2% (w/v) CaCl2 =55:45:10, (v/v); detect: orcinol reagent, 1208C,
5 min).
[12] For recent achievements on glycoprotein synthesis, see: a) I.
Sakamoto, K. Tezuka, K. Fukae, K. Ishii, K. Taduru, M. Maeda,
M. Ouchi, K. Yoshida, Y. Nambu, J. Igarashi, N. Hayashi, T.
Tsuji, Y. Kajihara, J. Am. Chem. Soc. 2012, 134, 5428 – 5431;
b) M. Murakami, R. Okamoto, M. Izumi, Y. Kajihara, Angew.
Chem. 2012, 124, 3627 – 3632; Angew. Chem. Int. Ed. 2012, 51,
3567 – 3572; c) H. Hojo, H. Tanaka, M. Hagiwara, Y. Asahina, A.
Ueki, H. Katayama, Y. Nakahara, A. Yoneshige, J. Matsuda, Y.
Ito, Y. Nakahara, J. Org. Chem. 2012, 77, 9437; d) O. Boutureira,
G. J. L. Bernardes, M. Fernandez-Gonzalez, D. C. Anthony,
B. G. Davis, Angew. Chem. 2012, 124, 1461 – 1465; Angew.
Chem. Int. Ed. 2012, 51, 1432 – 1436; e) V. Ullmann, M. Rꢀdisch,
I. Boos, J. Freund, C. Pçhner, S. Schwarzinger, C. Unverzagt,
Angew. Chem. 2012, 124, 11734 – 11738; Angew. Chem. Int. Ed.
2012, 51, 11566 – 11570; f) P. Wang, B. Aussedat, Y. Vohra, S. J.
Danishefsky, Angew. Chem. 2012, 124, 11739 – 11743; Angew.
Chem. Int. Ed. 2012, 51, 11571 – 11575.
[13] T. Ackrill, D. W. Anderson, D. Macmillan, Pept. Sci. 2010, 94,
495 – 503.
[14] For strategy using S protection of the N-to-S acyl transfer
devices and application to protein syntheses, see Ref. [6a,9b],
and [9c].
[15] M. Villain, J. Vizzavona, K. Rose, Chem. Biol. 2001, 8, 673 – 679.
[16] K. Sakamoto, K. Sato, A. Shiganaga, K. Tsuji, S. Tsuda, H.
Hibino, Y. Nishiuchi, A. Otaka, J. Org. Chem. 2012, 77, 6948 –
6958.
[17] M. Schnçlzer, P. Alewood, A. Jones, D. Alewood, S. B. H. Kent,
Int. J. Pep. Protein Res. 1992, 40, 180 – 193.
preparation of a wide range of GM2AP protein analogues,
thus leading to medicinal-chemistry-based evaluation of
glycoproteins.
Received: April 22, 2013
Revised: May 15, 2013
Published online: June 13, 2013
Keywords: glycoproteins · glycosylation · peptides ·
.
solid-phase synthesis · synthesis design
[1] a) B. Leader, Q. J. Baca, D. E. Golan, Nat. Rev. Drug Discovery
2008, 7, 21 – 39; b) P. J. Carter, Exp. Cell Res. 2011, 317, 1261 –
1269.
[2] a) P. E. Dawson, T. W. Muir, I. Clark-Lewis, S. B. H. Kent,
Science 1994, 266, 776 – 779; b) P. E. Dawson, S. B. H. Kent,
Annu. Rev. Biochem. 2000, 69, 923 – 960; c) S. B. H. Kent, Chem.
Soc. Rev. 2009, 38, 338 – 351; d) S. B. H. Kent, Y. Sohma, S. Liu,
D. Bang, B. Pentelute, K. Mandal, J. Pept. Sci. 2012, 18, 428 – 436.
[3] a) C. S. Zhang, L. H. Lai, Biochem. Soc. Trans. 2011, 39, 1382 –
1386; b) M. Zakrzewska, A. Szlachcic, J. Otlewski, Curr. Pharm.
Biotechnol. 2011, 12, 1792 – 1798.
[18] N. Fujii, A. Otaka, T. Watanabe, A. Okamachi, H. Tamamura, H.
Yajima, Y. Inagaki, M. Nomizu, K. Asano, J. Chem. Soc. Chem.
Commun. 1989, 283 – 284.
[19] H. Rohde, J. Schmalisch, Z. Harpaz, F. Diezmann, O. Seitz,
ChemBioChem 2011, 12, 1396 – 1400.
[4] J. Y. Lee, D. Bang, Pept. Sci. 2010, 94, 441 – 447.
7858
www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 7855 –7859

Angewandte
Chemie
[20] E. C. B. Johnson, S. B. H. Kent, J. Am. Chem. Soc. 2006, 128,
6640 – 6646.
[21] T. Kawakami, S. Aimoto, Tetrahedron 2009, 65, 3871 – 3877.
[22] D. Ravasi, M. Masserini, G. Vecchio, Y.-T. Li, S.-C. Li, Neuro-
chem. Res. 2002, 27, 785 – 792.
[23] The E. coli-expressed recombinant carbohydrate-free GM2AP
protein was already reported to be in every respect at least as
effective as the purified native glycosylated GM2AP. See: H.
Klima, A. Klein, G. van Echten, G. Schwarzmann, K. Suzuki, K.
Sandhoff, Biochem. J. 1992, 292, 571 – 576.
Angew. Chem. Int. Ed. 2013, 52, 7855 –7859
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7859