Molecular design of glycoprotein mimetics: Glycoblotting by engineered proteins with an oxylamino-functionalized amino acid residue

Article abstract of DOI:10.1002/chem.200500531

The general and efficient method for the site-directed glycosylation of proteins is a key step in order to understand the biological importance of the carbohydrate chains of proteins and to control functional roles of the engineered glycoproteins in terms

Full text of DOI:10.1002/chem.200500531

DOI: 10.1002/chem.200500531
Molecular Design of Glycoprotein Mimetics: Glycoblotting by Engineered
Proteins with an Oxylamino-Functionalized Amino Acid Residue
Naoki Matsubara,[a] Kei Oiwa,[a] Takahiro Hohsaka,[b] Reiko Sadamoto,[a]
Kenichi Niikura,[a] Norio Fukuhara,[a] Akio Takimoto,[c] Hirosato Kondo,[c] and
Shin-Ichiro Nishimura*[a]
Abstract: The general and efficient
method for the site-directed glycosyla-
tion of proteins is a key step in order
to understand the biological impor-
tance of the carbohydrate chains of
proteins and to control functional roles
of the engineered glycoproteins in
terms of the development of improved
glycoprotein therapeutics. We have de-
veloped a novel method for site-direct-
ed glycosylation of proteins based on
chemoselective blotting of common re-
ducing sugars by genetically encoded
proteins. The oxylamino-functionalized
l-homoserine residues, 2-amino-4-O-
(N-methylaminooxy) butanoic acid and
2-amino-4-aminooxy butanoic acid,
were efficiently incorporated into pro-
teins by using the four-base codon/anti-
codon pair strategy in Escherichia coli
in vitro translation. Direct and chemo-
selective coupling between unmodified
simple sugars and N-methylaminooxy
group displayed on the engineered
streptavidin allowed for the combinato-
rial synthesis of novel glycoprotein
mimetics.
Keywords:
glycoprotein
carbohydrates
· glycosylation
·
·
mass spectrometry
Introduction
tion difficult. Therefore, the general and efficient method
for the site-directed glycosylation of proteins is a key step
both to understand biological importance of the carbohy-
drate chains of proteins and to control functional roles of
the engineered glycoproteins in terms of the development of
improved glycoprotein therapeutics.[2] Recently, a pioneering
workby Schultz and co-workers provided a general strategy
for the cotranslational synthesis of selectively glycosylated
proteins in which the modified amino acids carrying b-
GlcNAc and a-GalNAc residues are genetically encoded in
Escherichia coli.[3,4] This method requires the evolutional
procedure for the isolation of an orthogonal Methanococcus
jannaschii tyrosyl tRNA synthetase (MjTyrRS) that specifi-
cally charges the corresponding M. jannaschii suppressor
(mutRNATyrCUA) with these sugar amino acids in response to
the amber codon (TAG) in E. coli.
Our interest has been focused on the high-throughput
generation of glycoprotein and glycopeptide mimetics in
order to investigate the effect of the site-specific glycosyla-
tion on the protein structure and functions. This will become
a key and rate-limiting step for the discovery of the new
generation glycoprotein drugs. We have established a facile
method for the site specific introduction of the w-aminoalk-
yl glycosides based on the transamination reaction by trans-
glutaminases at designated glutamine residues of the syn-
thetic peptides[5] or the mutant proteins.[6] A keto-containing
Glycosylation is one of the most important processes for
post-translational modifications in proteins, because this
modification step affects protein folding and stability, modi-
fies the intrinsic biological activity of proteins, and regulates
their target molecules/cells to be recognized in the biological
systems.[1] However, it is well known that naturally occurring
glycoproteins are often present as a population of many dif-
ferent and heterogeneous glycoforms that makes the investi-
gation of glycosylation effects on protein structure and func-
[a] N. Matsubara, K. Oiwa, Dr. R. Sadamoto, Dr. K. Niikura, Prof.
Dr. N. Fukuhara, Prof. Dr. S.-I. Nishimura
Laboratory for Bio-Macromolecular Chemistry
Division of Biological Sciences, Graduate School of Science
Frontier Research Center for Post-Genome Science and Technology
Hokkaido University, Sapporo 001-0021 (Japan)
Fax : (+81)11-706-9042
E-mail: shin@glyco.sci.hokudai.ac.jp
[b] Dr. T. Hohsaka
School of Material Science
Japan Advanced Institute of Science and Technology
1-1 Asahidai, Tatsunokuchi Ishikawa 923-1292 (Japan)
[c] Dr. A. Takimoto, Dr. H. Kondo
Discovery Research Laboratories
Shionogi & Co. Ltd.
Osaka 553-0002 (Japan)
6974
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2005, 11, 6974 – 6981

FULL PAPER
amino acid, p-acetyl-l-phenylalanine, has been successfully
incorporated in response to the amber nonsense codon[7]
and this unnatural amino acid can be used for the subse-
quent coupling with aminooxy saccharide derivatives[8] to
generate glycoprotein mimetics. Although these approaches
made the incorporation of carbohydrates into the engi-
neered proteins possible, both strategies definitely need
chemically modified sugar derivatives having an appropri-
ately functionalized group at each anomeric position. In the
present study, we report the feasibility of the frameshift sup-
pression strategy on the basis of
functional group can be used for generating neoglycopepti-
des.[13,14] Recently, it was also reported that combined use of
the glycoblotting by oxylamino-containing polymers and
MALDI-TOF/TOF mass spectrometry allows for both facile
purification and precise analysis of common oligosacchar-
ides and glycopeptides from human serum or mouse skin
glycoproteins.[11,15] In this study, two unnatural amino acids,
2-amino-4-O-(N-methylaminooxy) butanoic acid (1) and 2-
amino-4-aminooxy butanoic acid (2), were employed for the
synthesis of novel aminoacyl-tRNA (Figure 1). In order to
the four-base codon/anticodon
system[9] in the production of
glycoprotein mimetics. In case
of the strategy using expanded
genetic codes, it has been sug-
gested that many kinds of un-
natural amino acids can be in-
corporated into proteins by
using the chemically aminoacy-
lated frameshift suppressors in
in vitro translation systems in
response to both various four-
and five-base codons.[10] We
thought that this approach can
be used for the production of
the specific precursor proteins
having unnatural amino acids
that allow the direct glycosyla-
tion (glycoblotting)[11] with
a
variety of unmodified sugars
(naturally occurring oligosac-
charides) at the desired posi-
tions. It should be noted that
this strategy permits combina-
torial synthesis of glycoprotein
mimetics from a single precur-
sor mutant protein. In addition,
combined use of this method
with the common procedure
using the noncoding triplet
codons[12] will allow the incor-
poration of two or more differ-
ent carbohydrates into distinct
sites of a single polypeptide.
Results and Discussion
Synthesis of aminoacyl-tRNA
carrying unnatural amino acids:
As a versatile nonproteinogenic
functional group with a unique
reactivity, we selected the oxyl-
amino groups because it has
been demonstrated that the
synthetic peptides bearing this Figure 1. Chemical structures of the key compounds used in this study.
Chem. Eur. J. 2005, 11, 6974 – 6981
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6975

S.-I. Nishimura et al.
prepare the aminoacyl-tRNACCCG charged with these un-
natural amino acids, key intermediates 3 and 4 were synthe-
sized from l-homoserine according to the synthetic proce-
dures indicated in Schemes 1 and 2. Compounds 3 and 4
were coupled with pdCpA[16] to give the aminoacyl-pdCpA
derivatives 5 and 6. Next, they
tavidin having an unusual amino acid residue 1 for further
modification study.
Glycoblotting by engineered streptavidin: The N-methyl
oxylamino group incorporated into the mutant streptavidin
were employed for the conjuga-
tion with tRNACCCG in the pres-
ence of T4 RNA ligase accord-
ing to the method reported by
Hecht[17] and the subsequent
deprotection
target unnatural amino acid-
tRNACCCG Hser(ONHMe)-
afforded
the
,
tRNACCCG (7) and Hser-
(ONH2)-tRNACCCG (8) in good
yield.
In vitro transcription/transla-
tion: Position specific incorpo-
ration of the unnatural amino
acids by in vitro transcription/
translation was preliminarily
carried out according to the
protocols reported previously.[10]
As expected, in vitro transcrip-
tion/translation using E. coli
S30 extract in the presence of 7
or 8 proceeded smoothly and
permitted position specific in-
corporation of these two un-
natural amino acids into T4
phage lysozyme and streptavi-
din. Figure 2a shows the West-
ern blot analysis of the incorpo-
ration of 1 and 2 into the Gln45
position of the T4 phage lyso-
zyme. These results suggested
that the efficiency of the incor-
poration was estimated to be
17% (compound 1) and 12%
(compound 2), respectively.
Similarly, these amino acids
were incorporated into the
Tyr83 position of the full-length
streptavidin with 21% (com-
pound 1) and 15% (compound
2) as shown in Figure 2b. Con-
sidering the expression levels of
wild-type T4 phage lysozyme
and streptavidin produced by
the present in vitro transcrip-
tion/translation system (Fig-
ure 2c) and the satisfactory
purity of the protein checked
by a silver staining (Figure 2d),
Scheme 1. a) 4-Pentenoic anhydride, Et3N, H2O/MeOH 3:2; b) BTEAC, K2CO3, tBuBr, DMF (53% over two
steps); c) MsCl, Et3N, CH2Cl2; d) LiBr, acetone (81% over two steps); e) NaH, N-methyl-N-tert-butoxycarbo-
nylhydroxylamine, DMF (69%); f) TFA; g) NVOC-Cl, Et3N, H2O/dioxane 1:1; h) CNCH2Cl, Et3N, CH3CN
(82% over three steps); i) pdCpA, DMF (62%); j) tRNACCCG-CA, T4 RNA ligase; k) UV irradiation; l) I2.
Scheme 2. a) 4-Pentenoic anhydride, Et3N, H2O/MeOH 3:2; b) BTEAC, K2CO3, tBuBr, DMF (53% over two
steps); c) N-hydroxyphthalimide, Ph3P, DEAD, THF (84%); d) MeNH2/MeOH (2:3) (89%); e) TFA; f)
NVOC-Cl, Et3N, H2O/Dioxane (1:1); g) CNCH2Cl, Et3N, CH3CN (11% over three steps); h) pdCpA, DMF
we employed the mutant strep- (39%); i) tRNACCCG-CA, T4 RNA ligase; j) UV irradiation; k) I2. BTEAC = benzyltriethylammonium chloride.
6976
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2005, 11, 6974 – 6981

Glycoprotein Mimetics
FULL PAPER
19488.3, Figure 3g). Relatively low yields found for this en-
zymatic sialylation reaction might be due to the heterogene-
ous reaction conditions by employing precursor glycoprotein
(lactosylated streptavidin) immobilized on the magnetic Ni
beads. Table 1 summarizes the molecular weights character-
ized by MALDI-TOF MS of the glycoprotein mimetics stud-
ied in these experiments. Although further optimization
study must be carried out for achieving much higher modifi-
cation yield, the glycoblotting by the engineered proteins
having N-methyl oxylamino group will greatly contribute to
the construction of the library of artificially glycosylated
proteins in terms of the site-directed modifications with a
variety of carbohydrates.
Conclusion
Figure 2. Incorporation of the unnatural amino acids into proteins using
in vitro transcription/translation strategy. a) Western blot analysis of the
expression of the T4 phage lysozyme containing CGGG at the Gln45:
lane M, molecular weight markers; lane wt, T4 phage lysozyme (wild-
type); lane 1, in the absence of aminoacyl-tRNACCCG ; lane 2, in the pres-
ence of compound 7 [Hser(ONHMe)-tRNACCCG]; lane 3, in the presence
of compound 8 [Hser(ONH2)-tRNACCCG]. b) Western blot analysis of the
expression of the streptavidin containing CGGG at the Tyr83: lane M,
molecular weight marker; lane wt, streptavidin (wild-type); lane 1, in the
absence of aminoacyl-tRNACCCG ; lane 2, in the presence of compound 7
[Hser(ONHMe)-tRNACCCG]; lane 3, in the presence of compound 8
[Hser(ONH2)-tRNACCCG]. c) Comparison of the expression level be-
tween T4 phage lysozyme and streptavidin. lane M, marker proteins;
lane 1, wild-type T4 phage lysozyme; lane 2, wild-type streptavidin. d)
Silver staining analysis of the mutant streptavidin containing unnatural
amino acid 1: lane M, molecular weight markers; lane 1, in vitro transla-
tion mixture; lane 2, a fraction obtained by T7-tag affinity chromatogra-
phy; lane 3, a fraction obtained by purification in combination with His-
tag affinity chromatography and T7-tag affinity chromatography.
We could successfully establish a general method for the
synthesis of homogeneous glycoprotein mimetics bearing
simple sugar chains at the desired position based on the
unique glycoblotting strategy. Since the availability of non-
coding triplet codons ultimately limits the number of amino
acids encoded by any organism, the present method might
become an alternative or a complementary way for the pro-
duction of a variety of glycoprotein mimetics. Although it
remains to be discussed whether this site-specific glycosyla-
tion through the non-natural linkage influences the structure
and functions of the glycoprotein mimetics, this method will
greatly contribute to a more rapid and high-throughput
assay system for searching new types of glycosylated protein
drugs in combinatorial manner. We are now investigating
the effect of this non-natural linkage on the conformation of
the glycoprotein by means of NMR/CD spectroscopy. The
results of both structural and biological characterization of
glycoprotein mimetics will be reported in the near future.
was tested for subsequent glycoblotting (chemoselective li-
gation with unmodified mono- and oligosaccharides). The
general protocols for the glycoblotting experiments are de-
scribed in the following: Sugar solutions (final concentra-
tion; 100–400 mm) were added to a solution of the mutant
protein (200 ng) dissolved in 100 mm acetate buffer (pH 4.3,
3.0 mL) and the reaction mixture was incubated at 378C for
44 h. As shown in Figure 3, glycoblotting by the engineered
streptavidin (Figure 3a) of Glc (d-glucose, Figure 3b),
Glca1,4Glc (maltose, Figure 3c), Glca1,4Glca1,4Glc (malto-
triose, Figure 3d), GlcNAc (N-acetyl-d-glucosamine, Fig-
ure 3e), and Galb1,4Glc (lactose, Figure 3f) were observed
by MALDI-TOF MS. In case of the wild-type streptavidin
as a negative control, no unfavorable side-reaction such as
non-specific glycation at lysine residues was detected. The
unmodified protein bearing N-methyl oxylamino group can
be removed by trapping with aldehyde-functionalized poly-
mers; this step will make a practical and efficient purifica-
tion of the target glycoprotein mimetics possible (data not
shown).
Experimental Section
General methods and materials: T7-Tag antibody, T7-Tag antibody agar-
ose was obtained form Novagen. E. coli T7 S30 extract system for circu-
lar DNA, ProtoBlot II AP System with stabilized substrate mouse, and
MagneHis Ni-particles were purchased from Promega Co. T4 RNA
ligase, protein markers, and prestained protein markers were from New
England Biolabs. Immun-Blot PVDF membrane for protein blotting was
obtained from Bio-Rad. A T4 phage lysozyme gene was purchased from
TAKARA Bio Inc. Plasmids (pGSH), encoding a sequences of T7 pro-
moter, a T7 tag at the N-terminal, a (His)6 tag at the C-terminal, and a
T7 terminator, using for expression of the streptavidin and T4 phage ly-
sozyme, each of wild-type and mutant including a CGGG codon, and
tRNACCCG(-CA) (the insertion of the CCCG anticodon at the anticodon
loop and the lackof a CA di-nucleotide unit at the 3 ’ end) were prepared
according to the methods reported previously [T. Hohsaka, D. Kajihara,
Y. Ashizuka, H. Murakami, M. Sisido, J. Am. Chem. Soc. 1999, 121, 34–
40]. Chemical reactions were monitored by thin-layer chromatography
(TLC) on pre-coated plates of Silica Gel 60F254 (E. Merck). Visualization
was accomplished with UV light (254 nm) and treatment with a solution
of (NH4)6Mo7O2·4H2O (20 g) and Ce(SO4)2 (0.4 g) in 10% sulfuric acid
(400 mL) and heating at 1508C or ninhydrin reagent (Wako Pure Chemi-
cal Industries, Ltd.). Enzymatic reactions in aminoacyl-tRNA synthesis
In addition, it was also revealed that the streptavidin
modified with lactose (m/z 19208.7) can be converted by re-
combinant a2,3-sialyltransferase into a novel glycoprotein
mimetic having GM3 trisaccharide at the position 83 (m/z
Chem. Eur. J. 2005, 11, 6974 – 6981
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6977

S.-I. Nishimura et al.
Figure 3. MALDI-TOF MS analysis of the molecular weight of mutant streptavidin and glycoprotein mimetics: a) mutant streptavidin having an unnatu-
ral amino acid 1; product obtained by glycoblotting of mutant streptavidin with b) glucose, c) maltose, d) maltotriose, e) N-acetylglucosamine, f) lactose;
g) modification of the mutant streptavidin with glycosyltransferase.
were monitored by 15% PAGE with 7m Urea. Purifications of the syn-
thetic product by column chromatography were carried out by using
silica gel 60N (KANTO Co., 63–210 nm mesh). All other reagents were
purchased from Sigma-Aldrich Co. Ltd., Wako Pure Chemical Industries,
Ltd., Tokyo Kasei Co. Ltd. and Nacalai Tesque, Inc.
NMR spectroscopy: 1H and 13C NMR spectra were recorded and mea-
sured at 600 and 120 MHz, respectively, with AVANCE 600 (Bruker)
using CDCl3 as a solvent. All signals of the new compounds synthesized
in this study were characterized with H,H COSY and HMQC techniques.
Mass spectrometry: FAB mass spectra were obtained with JMS-HX100
(JEOL) using glycerol or m-nitrobenzyl alcohol as a matrix. ESI mass
spectra were obtained with JMS-700TZ (JEOL). MALDI-TOF MS were
measured with ultraflex and biflex (Bruker). Samples were desalted and
concentrated using 10 mL C4 ZipTip (Millipore) according to the manu-
6978
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2005, 11, 6974 – 6981

Glycoprotein Mimetics
FULL PAPER
Table 1. Theoretical and observed molecular weights of glycosylated
streptavidin.
with brine, dried using MgSO4, filtrated, concentrated, and purified by
flash column chromatography (hexane/ethyl acetate 4:1) to give product
(72 mg, 69%). Rf =0.28 (hexane/ethyl acetate 3:1); 1H NMR (600 MHz,
CDCl3): d = 7.37 (s, 1H, NH), 5.88–5.81 (m, 1H, C=CH-), 5.08–4.97 (m,
2H, -C=CH2), 4.63–4.60 (m, 1H, a-H), 3.93–3.86 (m, 2H, g-H2), 3.07 (s,
1H, N-CH3), 2.43–2.33 (m, 4H, C=C-CH2-, C=C-C-CH2-), 2.13–2.02 (m,
2H, b-H2), 1.48, 1.44 (all s, 18H, 2”(CH3)3-C-); 13C NMR (120 MHz,
CDCl3): d = 172.42, 171.09, 157.00 (3C, 3”C=O), 137.27 (C=C-), 115.09
(C=C-), 81.69, 81.60 (2”(CH3)3-C-), 70.52 (g-C), 50.58 (a-C), 36.66 (N-
CH3), 35.60, 29.75 (2C, C=C-C-C), 29.60 (b-C), 28.20, 27.89 (2C, 2”
(CH3)3-C-); HR-FAB MS: m/z: calcd for C19H34N2O6: 386.2417; found:
387.2487 [M+H]+.
Additional carbohydrate
Calcd mass
Found
none
glucose
maltose
maltotriose
GlcNAc
lactose
18875.3
19037.5
19199.6
19361.7
19078.5
19199.6
19490.9
18871.5[a]
19034.3[a]
19206.8[a]
19368.5[a]
19073.1[a]
19208.7[a]
19488.3[b]
sialyllactose
[a] [M+H]+. [b] [MÀH]À.
2-(Pent-4-enoyl)amino-4-[O-[N-methyl-(N-6-nitroveratryloxycarbonyl)]-
amino]hydroxybutanoic acid cyanomethyl ester (3): 2-(Pent-4-enoyl)ami-
no-4-[O-(N-methyl-N-tert-butoxycarbonyl)amino]hydroxybutanoic acid
tert-butyl ester (70 mg, 0.18 mmol) was dissolved in trifluoroacetic acid
(1 mL) and stirred for 1 h. The reaction mixture was concentrated, then
dissolved in a solution of H2O (400 mL) and dioxane (400 mL), and
cooled at 08C. To the mixture was added triethylamine (88 mL,
0.63 mmol) and 6-nitroveratryloxycarbonyl chloride (NVOC-Cl; 60 mg,
0.22 mmol). The mixture was stirred at 08C for 2 h in the dark, then ex-
tracted with ethyl acetate. The organic layer was washed with 5% citric
acid and brine, dried using MgSO4, filtrated, concentrated, and purified
by flash column chromatography (chloroform/methanol 30:1). The con-
centrated residue was dissolved in acetonitrile (0.5 mL) and cooled at
08C under nitrogen atmosphere. Triethylamine (90 mL, 0.64 mmol) and
chloroacetonitrile (0.1 mL, 1.6 mmol) were added to the solution and the
reaction mixture was stirred at room temperature for 24 h. The mixture
was extracted with ethyl acetate and the extract was washed with brine,
dried using MgSO4, filtrated, concentrated, and purified by flash column
chromatography (hexane/ethyl acetate 2:3) to give compound 3 (68 mg,
82%). Rf =0.40 (hexane/ethyl acetate 1:2); 1H NMR (600 MHz, CDCl3):
d= 7.64 (s, 1H, aromatic), 6.92 (s, 1H, aromatic), 5.79–5.72 (m, 1H, C=
CH-), 5.54–5.48 (m, 2H, O-CH2-Ph), 5.01–4.91 (m, 2H, -C=CH2), 4.80–
4.76 (m, 1H, a-H), 4.72–4.65 (m, 2H, O-CH2-CN), 3.91, 3.90 (all s, 6H,
2”OCH3), 3.94–3.89 (m, 2H, g-H2), 3.14 (s, 3H, N-CH3), 2.35–2.29 (m,
4H, C=C-CH2-, C=C-C-CH2-), 2.24–2.19 (m, 1H, b-H2-a), 2.06–2.01 (m,
1H, b-H2-b); 13C NMR (120 MHz, CDCl3), d172.84, 170.66, 157.16 (3C,
3”C=O), 153.48, 148.54, 140.19, 126.49, 110.93, 108.32 (6C, aromatic),
137.04 (C=C-C), 115.38 (C=C-C), 114.03 (-CN), 70.06 (g-C), 65.12 (O-C-
Ph), 56.47, 56.45 (2C, 2”OCH3), 49.50 (a-C), 48.79 (O-C-CN), 36.54 (N-
CH3), 35.28, 29.38 (2C, C=C-C-C), 28.96 (b-C); HR-FAB MS: m/z: calcd
for C22H28N4O10: 508.1805; found: 509.1877 [M+H]+.
facturerꢁs instruction. Typically, samples were mixed with the 1 mL of 2,5-
dihydroxybenzoic acid (DHB) in 33% acetonitrile containing 0.1% tri-
fluoroacetic acid at the concentration of 10 mgmLÀ1. The theoretical
mass was calculated an average mass of a protein without methionine.
The loss of the N-terminal methionine is common in E. coli expression
system.
Synthesis of aminoacyl pdCpA derivatives
2-(Pent-4-enoyl)amino-4-hydroxybutanoic acid tert-butyl ester: 4-Pente-
noic anhydride (3.7 mL, 20.2 mmol) and triethylamine (2.8 mL,
20.2 mmol) was added to an ice-cooled solution of l-homoserine (2.0 g,
16.8 mmol) in H2O (6 mL) and methanol (4 mL). The reaction mixture
was stirred at 08C for 2 h and then concentrated. The crude residue was
purified by flash column chromatography (chloroform/methanol/H2O
65:15:1). The concentrated residue was dissolved in N,N-dimethylforma-
mide (25 mL) and added to benzyltriethylamine chloride (4.0 g,
17.6 mmol). After being stirred for 5 min, K2CO3 (22.1 g, 160 mmol) and
tert-butyl bromide (26.5 mL, 230 mmol) were added. The reaction mix-
ture was stirred at 558C for 6 h and then added ice water. The mixture
was extracted with ethyl acetate and the organic layer was washed with
brine, dried using MgSO4, filtrated, and concentrated. The residual syrup
was purified by flash column chromatography (hexane/ethyl acetate 3:2)
to give product (2.3 g, 53%). Rf =0.50 (hexane/ethyl acetate 1:2).
1H NMR (600 MHz, CDCl3): d = 6.39 (d, J=7.20 Hz, 1H, NH), 5.85–
5.78 (m, 1H, C=CH-), 5.11–5.02 (m, 2H, -C=CH2), 4.65–4.60 (m, 1H, a-
H), 3.80–3.77 (m, 1H, -OH), 3.71–3.65 (m, 1H, g-H2-a), 3.56–3.50 (m,
1H, g-H2-b), 2.45–2.35 (m, 4H, C=C-CH2-, C=C-C-CH2-), 2.21–2.15 (m,
1H, b-H2-a), 1.54–1.50 (m, 1H, b-H2-b), 1.48 (s, 9H, (CH3)3-C-); HR-
FAB MS : m/z: calcd for C13H23NO4: 257.1627; found: 258.1709 [M+H]+.
2-(Pent-4-enoyl)amino-4-bromobutanoic acid tert-butyl ester: A solution
of 2-(pent-4-enoyl)amino-4-hydroxybutanoic acid tert-butyl ester (85 mg,
0.33 mmol) in CH2Cl2 (1.5 mL) was cooled at 08C under nitrogen atmos-
phere. Triethylamine (55 mL, 0.40 mmol) and methanesulfonyl chloride
(27 mL, 0.35 mmol) were added, and the solution was stirred for 1 h. Lith-
ium bromide (207 mg, 2.39 mmol) and acetone (1.0 mL) were added to
the reaction mixture, and the mixture was stirred for 19 h. The mixture
was concentrated and extracted with ethyl acetate. The organic layer was
washed with sat. NaHCO3, brine, dried using MgSO4, filtrated, and con-
centrated. The residual syrup was subjected to the purification by flash
column chromatography (hexane/ethyl/acetate 4:1) to give product
(86 mg, 81%). Rf =0.50 (hexane/ethyl acetate 2:1); 1H NMR (600 MHz,
CDCl3): d = 6.17 (d, J=7.37 Hz, 1H, NH), 5.85–5.78 (m, 1H, C=CH-),
5.09–5.01 (m, 2H, -C=CH2), 4.59–4.56 (m, 1H, a-H), 3.41–3.32 (m, 2H,
g-H2), 2.45–2.31 (m, 5H, C=C-CH2-, C=C-C-CH2-, b-H2-a), 2.22–2.17 (m,
1H, b-H2-b), 1.47 (s, 9H, (CH3)3-C-); HR-FAB MS: m/z: calcd for
C13H22NO3Br: 319.0783; found: 320.0862 [M+H]+.
2-(Pent-4-enoyl)amino-4-[O-[N-methyl-(N-6-nitroveratryloxycarbonyl)]-
amino]hydroxybutanoic acid pdCpA ester (5): The aminoacylation of
pdCpA was carried out by adding the compound 3 (1.0 mg, 2.0 mmol) to
N,N-dimethylformamide solution of pdCpA tetra-n-butylammonium salt
(10 mL, 0.44 mmol) in a microtube, and the mixture was incubated at
308C for 2 h. The reaction was monitored by a reverse-phase HPLC (In-
ertsil ODS-3 150–4.6, 20–100% CH3OH/0.1m NH4OAc, pH 4.5, over
20 min at flow rate of 1.0 mLminÀ1, detection at 260 nm), the products
had the retention time of 14.8 and 15.2 min, for the two positional (2’,3’)
isomers. After 2 h, Et2O (1 mL) was added to the solution and then the
precipitation was collected by a centrifugation (20000 g, 48C, 5 min). The
precipitate was dissolved in acetonitrile (20 mL) and the solution was sub-
jected to the re-precipitation by adding Et2O (1 mL). The crude product
was again dissolved in acetonitrile (0.2 mL) and 0.1m NH4OAc, pH 4.5
(0.8 mL), and purified by a reverse-phase HPLC (Inertsil ODS-3 50–20,
20% CH3OH/0.1m NH4OAc, pH 4.5, 2 min then 20–100% CH3OH/0.1m
NH4OAc, pH 4.5, over 20 min at flow rate of 10 mLminÀ1, detection at
260 nm), the products with the retention time of 14.2 and 14.6 min were
2-(Pent-4-enoyl)amino-4-[O-(N-methyl-N-tert-butoxycarbonyl)amino]hy-
droxybutanoic acid tert-butyl ester: NaH (60% dispersion in mineral oil,
13 mg, 0.32 mmol) was added to a solution of 2-(Pent-4-enoyl)amino-4-
bromobutanoic acid tert-butyl ester (47 mg, 0.32 mmol) in N,N-dimethyl-
formamide (1 mL) and the mixture was stirred for 5 min under nitrogen
atmosphere, then cooled at 08C. N-Methyl-N-tert-butoxycarbonylhydrox-
ylamine (86 mg, 0.27 mmol) was added to the solution, and the mixture
was stirred at 08C for 1 h. To the mixture was added ice water and the
mixture was extracted with ethyl acetate. The organic layer was washed
obtained as two positional (2’,3’) isomers. Compound
5 (0.27 mmol,
62%). HR-ESI MS: m/z: calcd for C39H51N11O22P2: 1087.2685; found:
1086.2632 [MÀH]À.
2-(Pent-4-enoyl)amino-4-[O-(N-phthalimidyl)]hydroxybutanoic acid tert-
butyl ester: N-Hydroxyphthalimide (0.32 g, 2.0 mmol), triphenylphos-
phine (0.52 g, 2.0 mmol) and diethyl azodicarboxylate (0.31 mL,
2.0 mmol) were added to a solution of 2-(pent-4-enoyl)amino-4-hydroxy-
Chem. Eur. J. 2005, 11, 6974 – 6981
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6979

S.-I. Nishimura et al.
butanoic acid tert-butyl ester (0.42 g, 1.64 mmol) in tetrahydrofuran
(10 mL). The reaction mixture was stirred for 2 h at RT under nitrogen
atmosphere. The solvent was evaporated, and the residue was purified by
flash column chromatography (toluene/ethyl acetate 5:1) to give product
(0.56 g, 84%). Rf =0.57 (toluene/ethyl acetate 1:1); 1H NMR (600 MHz,
CDCl3): d = 7.86–7.77 (m, 4H, aromatic), 7.13 (d, J=7.13 Hz, 1H, NH),
5.88–5.84 (m, 1H, C=CH-), 5.10–4.98 (m, 2H, -C=CH2), 4.73–4.70 (m,
1H, a-H), 4.31–4.29 (m, 2H, g-H2), 2.46–2.41 (m, 4H, C=C-CH2-, C=C-
C-CH2-), 2.33–2.28 (m, 2H, b-H2), 1.46 (s, 9H, (CH3)3-C-); 13C NMR
(120 MHz, CDCl3): d = 172.31, 170.64, 163.53 (3C, 3”C=O), 137.16 (C=
C-), 134.66, 128.81, 123.68 (3C, 3”aromatic), 115.34 (C=C-), 82.18
((CH3)3-C-), 75.30 (g-C), 50.35 (a-C), 35.61, 29.46 (2C, C=C-C-C), 29.85
(b-C), 27.91 ((CH3)3-C-); HR-FAB MS: m/z: calcd for C21H26N2O6:
402.1791; found: 403.1879 [M+H]+.
sitional (2’,3’) isomers. After 2 h, Et2O (1 mL) was added to the solution
and then the precipitation was collected by a centrifugation (20000 g,
48C, 5 min). The precipitate was dissolved in acetonitrile (20 mL) and the
solution was subjected to the re-precipitation by using Et2O (1 mL) as
poor solvent. The crude material was purified by a reverse-phase HPLC
to afford compound 6 (0.17 mmol, 39%). HR-ESI MS: m/z: calcd for
C38H49N11O22P2: 1073.2529; found: 1072.2435 [MÀH]À.
Synthesis of aminoacyl tRNA: A mixture containing tRNACCCG(-CA)
(0.3 nmol), aminoacyl pdCpA (6.0 nmol), 1 mm ATP, 3.3 mm DDT, 15 mm
MgCl2, 20 mgmLÀ1 BSA, 15% DMSO and T4 RNA ligase (40 U) in a
20 mL of 55 mm HEPES-Na (pH 7.5) was incubated at 378C for 20 min.
The mixture was treated with 3m potassium acetate (pH 5.5, 0.3m potas-
sium acetate as a final concentration) and extracted with phenol/chloro-
form/isoamylalcohol and chloroform. The solution was irradiated
1
2-(Pent-4-enoyl)amino-4-(O-amino)hydroxybutanoic acid tert-butyl ester:
2-(Pent-4-enoyl)amino-4-[O-(N-phthalimidyl)]hydroxybutanoic acid tert-
butyl ester (0.16 g, 0.39 mmol) was dissolved in 40% methylamine/metha-
nol solution (2 mL). The mixture was stirred for 45 min at RT, and
evaporated. The residue was purified by flash column chromatography
(toluene/ethyl acetate 1:5) to give the target compound (0.94 g, 89%).
Rf =0.51 (ethyl acetate); 1H NMR (600 MHz, CDCl3): d = 6.31 (d, J=
7.19 Hz, 1H, NH), 5.88–5.80 (m, 1H, C=CH-), 5.44 (s, 2H, O-NH2), 5.10–
5.00 (m, 2H, -C=CH2), 4.60–4.56 (m, 1H, a-H), 3.76–3.71 (m, 2H, g-H2),
2.43–2.31 (m, 4H, C=C-CH2-, C=C-C-CH2-), 2.11–1.96 (m, 2H, b-H2),
1.48 (s, 9H, (CH3)3-C-); 13C NMR (120 MHz, CDCl3): d = 171.54, 171.45
(2C, 2”C=O), 136.94 (C=C-), 115.56 (C=C-), 82.12 ((CH3)3-C-), 72.00 (g-
C), 50.42 (a-C), 35.74, 29.45 (2C, C=C-C-C), 31.19 (b-C), 27.97 ((CH3)3-
C-); HR-FAB MS: m/z: calcd for C13H24N2O4: 272.1736; found: 273.1814
[M+H]+.
(365 nm) at 08C for 90 min and then reacted with = volume of 50 mm I2
10
(H2O/THF 1:1) for 20 min at 08C. The reaction mixture was treated with
3 vol. ethanol to precipitate the aminoacyl-tRNA. The pellet was washed
with 70% ethanol, dried, and dissolved in 1 mm potassium acetate
(pH 5.5, 3 mL), and the solution was immediately added to the reaction
mixture of the in vitro translation system.
In vitro protein biosynthesis: In vitro translation/transcription were per-
formed by using E. coli T7 S30 extract systems for circular DNA kit
(Promega). For the expression analysis, the reactions were carried out in
a 10 mL of a reaction mixture containing plasmid (0.5 mg), 0.1 mm each
amino acids except arginine, 0.01 mm arginine, S30 premix without amino
acid (4 mL), S30 extract (3 mL). If necessary, 0.01 mm aminoacyl
tRNACCCG was also added. The mixture was incubated at 378C for
60 min. For the large-scale expression, the reaction volumes were in-
creased up to 200 mL.
2-(Pent-4-enoyl)amino-4-[O-(N-6-nitroveratryloxycarbonyl)amino]hy-
droxybutanoic acid cyanomethyl ester (4): 2-(Pent-4-enoyl)amino-4-(O-
amino)hydroxybutanoic acid tert-butyl ester (49 mg, 0.18 mmol) was dis-
solved in trifluoroacetic acid (1 mL) and the solution was stirred for 1 h.
The reaction mixture was concentrated, then the residue was dissolved in
H2O (500 mL) and dioxane (1.0 mL). To this solution was added triethyla-
mine (75 mL, 0.54 mmol) and 6-nitroveratryloxycarbonyl chloride (60 mg,
0.22 mmol). The mixture was stirred at 08C for 1.5 h in the dark, then ex-
tracted with ethyl acetate. The organic layer was washed with 5% citric
acid and brine, dried using MgSO4, filtrated, concentrated, and purified
by flash column chromatography (chloroform/methanol 25:1). The con-
centrated residue was dissolved in acetonitrile (0.7 mL) and cooled at
08C under nitrogen atmosphere. To the solution was added triethylamine
(27 mL, 0.20 mmol) and chloroacetonitrile (41 mL, 0.65 mmol) and the
mixture was stirred at room temperature for 20 h. The mixture was ex-
tracted with ethyl acetate and the extract was washed with brine, dried
using MgSO4, filtrated, concentrated, and purified by flash column chro-
Western blotting analysis: In vitro translation mixture (0.25 mL) was ap-
plied to 15% polyacrylamide gels followed by SDS-PAGE. All samples
were heated prior to application to SDS-PAGE wells. After electroblott-
ing to a PVDF membrane, the membrane was incubated with 3% BSA
in TBST (25 mm Tris-HCl (pH 7.5), 150 mm NaCl, and 0.02% Tween20),
T7-tag antibody and anti-mouse IgG (H+L) AP conjugate at 378C for
30 min. The development with Western Blue was performed at 378C for
20 min. The efficiency of incorporation of unnatural amino acids was esti-
mated by comparing the band intensity of the full-length polypeptide
with its wild-type expressed in vitro translation. The band intensity was
evaluated using the Scion Image programs (Scion Corporation).
Silver staining analysis: Each sample was applied to 15% polyacrylamide
gels followed by SDS-PAGE. All samples were heated prior to applica-
tion to SDS-PAGE wells. Silver staining was carried out by using Silver
Staining II Kit Wako (Wako Pure Chemical Industries, Ltd.) according to
the manufacturerꢁs instructions. Developing time was 5 min.
Purification of proteins by T7-tag affinity chromatography: The volume
of the reaction mixture (200 mL) of in vitro translation was adjusted to
0.5 mL with phosphate buffer (4.3 mm Na2HPO4, 1.5 mm KH2PO4, 2.7 mm
KCl, 137 mm NaCl, 0.1% Tween20, 0.002% sodium aside, pH 7.3). The
mixture was added to a suspension of T7 tag antibody agarose (50 mL),
then the mixture was incubated at RT for 60 min. The agarose gel con-
taining proteins was loaded to an empty spin column, washed with the
phosphate buffer (5 mL), and eluted with 100 mm citric acid (pH 2.2,
0.12 mL). The elution was neutralized with 2m Tris (pH 10.4, 180 mL) and
concentrated using Microcon YM-10 (millipore).
matography (hexane/ethyl acetate 1:5) to give compound
4 (10 mg,
11%). Rf =0.50 (hexane/ethyl acetate 1:5); 1H NMR (600 MHz, CDCl3):
d = 7.75 (s, 1H, NH), 7.73 (1H, aromatic), 6.98 (s, 1H, aromatic), 5.87–
5.80 (m, 1H, C=CH-), 5.63–5.57 (m, 2H, O-CH2-Ph), 5.09–4.99 (m, 2H,
-C=CH2), 4.89–4.86 (m, 1H, a-H), 4.80–4.72 (m, 2H, O-CH2-CN), 4.05–
3.99 (m, 2H, g-H2), 3.99, 3.97 (all s, 6H, 2”OCH3), 2.44–2.37 (m, 4H, C=
C-CH2-, C=C-C-CH2-), 2.33–2.28 (m, 1H, b-H2-a), 2.14–2.09 (m, 1H, b-
H2-b); 13C NMR (120 MHz, CDCl3): d = 1712.81, 170.79, 157.68 (3C, 3”
C=O), 153.62, 148.57, 140.03, 126.33, 110.67, 108.31 (6C, aromatic),
136.97 (C=C-C), 115.52 (C=C-C), 114.10 (-CN), 72.94 (g-C), 64.82 (O-C-
Ph), 56.57, 56.46 (2C, 2”OCH3), 49.50 (a-C), 48.86 (O-C-CN), 35.34,
29.38 (2C, C=C-C-C), 29.19 (b-C); HR-FAB MS: m/z: calcd for
C21H26N4O10: 494.1649; found: 495.1730 [M+H]+.
Purification of proteins by His-tag affinity chromatography: To a solution
containing proteins in 100 mm HEPES-Na (pH 7.5, 50 mL) was added a
suspension of MagneHis Ni particles (5 mL). The mixture was incubated
at RT for 30 min and washed thoroughly with 100 mm HEPES-Na
(pH 7.5, 50 mL). Finally, the target proteins were obtained by desorbing
with elution buffer (100 mm HEPES-Na, (pH 7.5) containing 750 mm imi-
dazole, 10 mL).
2-(Pent-4-enoyl)amino-4-[O-(N-6-nitroveratryloxycarbonyl)amino]hy-
droxybutanoic acid pdCpA ester (6): The aminoacylation of pdCpA was
carried out by adding the compound 4 (0.80 mg, 1.6 mmol) to N,N-dime-
thylformamide solution of pdCpA tetra-n-butylammonium salt (10 mL,
0.44 mmol) in a microtube, and the reaction mixture was incubated at
308C for 2 h. The reaction was monitored by a reverse-phase HPLC (In-
ertsil ODS-3 150-4.6, 20–100% CH3OH/0.1m NH4OAc, pH 4.5, over
20 min at flow rate of 1.0 mLminÀ1, detection at 260 nm), the products
with the retention time of 14.2 and 14.5 min were collected as the two po-
Glycoblotting by an engineered protein of unmodified optional carbohy-
drates: Proteins purified by T7 tag affinity chromatography (approx.
200 ng) was dissolved in acetate buffer (pH 4.0, 3 mL) containing carbo-
hydrate to be introduced. The mixture was incubated at 378C and sub-
jected to the purification by His-tag affinity chromatography and the
6980
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2005, 11, 6974 – 6981

Glycoprotein Mimetics
FULL PAPER
product was identified and characterized by MALDI-TOF mass spec-
trometry.
[1] R. A. Dwek, Chem. Rev. 1996, 96, 683–720.
[2] S. Elliott, Nat. Biotechnol. 2003, 21, 414–420.
[3] Z. Zhang, Science 2004, 303, 371–373.
[4] R. Xu, J. Am. Chem. Soc. 2004, 126, 15654–15655.
[5] T. Ohta, Angew. Chem. 2003, 115, 5344–5347; Angew. Chem. Int.
Ed. 2003, 42, 5186–5189.
[6] M. Sato, J. Am. Chem. Soc. 2004, 126, 14013–14022.
[7] H. Liu, J. Am. Chem. Soc. 2003, 125, 1702–1703.
[8] E. C. Rodriguez, L. A. Marcaurelle, C. R. Bertozzi, J. Org. Chem.
1998, 63, 7134–7135.
[9] T. Hohsaka, M. Sisido, Curr. Opin. Chem. Biol. 2002, 6, 809–815.
[10] T. Hohsaka, Biochemistry 2001, 40, 11060–11064.
[11] S.-I. Nishimura, Angew. Chem. 2005, 117, 93–98; Angew. Chem. Int.
Ed. 2005, 44, 91–96.
[12] J. C. Anderson, Proc. Natl. Acad. Sci. USA 2004, 101, 7566–7571.
[13] S. E. Cervigni, P. Dumy, M. Mutter, Angew. Chem. 996, 108, 1325–
1328; Angew. Chem. Int. Ed. Engl. 1996, 35, 1230–1232.
[14] M. R. Carrasco, R. T. Brown, J. Org. Chem. 2003, 68, 8853–8858.
[15] K. Niikura, R. Kamitani, M. Kuogochi, R. Uematsu, Y. Shinohara,
H. Nakagawa, K. Degushi, K. Monde, H. Kondo, S.-I. Nishimura,
Chem. Eur. J. 2005, 11, 3825–3834.
Sugar elongation (sialylation) reaction of Hser(ONHMe) mutant of
streptavidin carrying a lactose residue: Hser(ONHMe) mutant of strepta-
vidin purified by T7 tag affinity chromatography (approx. 200 ng) was
dissolved in acetate buffer (pH 4.0, 10 mL) containing 400 mm lactose.
The solution was incubated at 378C for 44 h. To a solution diluted with
100 mm HEPES-Na (pH 7.5, 20 mL) was added a suspension of MagneHis
Ni particles (2 mL). The mixture was incubated at RT for 30 min, washed
twice with 100 mm HEPES-Na (pH 7.5, 100 mL) and 50 mm HEPES-Na
(pH 7.5, 100 mL), respectively. The magnetic beads bearing lactosylated
streptavidin were directly re-suspended by adding to 100 mL 50 mm
HEPES-Na (pH 7.5) containing 3 mm CMP-Neu5Ac, 0.01% BSA, 0.2%
Triton X-100, and 300 mUmLÀ1 a2,3-sialyltransferase. The mixture was
incubated at 378C for 22 h, then washed twice with 50 mm HEPES-Na
(pH 7.5, 100 mL) and 100 mm HEPES-Na (pH 7.5, 100 mL), respectively.
The product was obtained by desorbing with elution buffer (100 mm
HEPES-Na (pH 7.5) containing 750 mm imidazole, 6 mL) and analyzed by
MALDI-TOF mass spectrometry.
[16] M. Lodder, S. Golovine, S. M. Hecht, J. Org. Chem. 1997, 62, 778–
779.
Acknowledgements
[17] V. A. Karginov, J. Am. Chem. Soc. 1997, 119, 8166–8176.
We thankDr. M. Kurogochi for his help in measuring MALDI-TOF MS.
This workwas supported partly by a grant for “Research Project of the
Functional Glycomaterials” from the Ministry of Education, Culture, Sci-
ence, and Technology, Japan.
Received: May 13, 2005
Published online: September 6, 2005
Chem. Eur. J. 2005, 11, 6974 – 6981
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6981