Synthesis of gold nanoparticles bearing the Thomsen-Friedenreich disaccharide: A new multivalent presentation of an important tumor antigen

Article abstract of DOI:10.1016/j.tetasy.2004.12.003

Herein we describe the synthesis gold nanoshells encapsulated with up to 90 units of the Thomsen-Friedenreich (TF) tumor-associated carbohydrate antigen (TACA) disaccharide (Galβ1-3GalNAc-α-O-Ser/Thr) as well as the assembly of a suitably linked designer

Full text of DOI:10.1016/j.tetasy.2004.12.003

Tetrahedron:
Asymmetry
Tetrahedron: Asymmetry 16 (2005) 587–598
Synthesis of gold nanoparticles bearing the
Thomsen–Friedenreich disaccharide: a new multivalent
presentation of an important tumor antigen
Sergei A. Svarovsky,a Zoltan Szekelyb and Joseph J. Barchi, Jr.a,*
aLaboratory of Medicinal Chemistry, Center for Cancer Research, National Cancer Institute at Frederick,
376 Boyles Street, Frederick, MD21702, USA
bStructural Biophysics Laboratory, Center for Cancer Research, National Cancer Institute at Frederick, 376 Boyles Street,
Frederick, MD21702, USA
Received 22 November 2004; accepted 2 December 2004
Abstract—Herein we describe the synthesis gold nanoshells encapsulated with up to 90units of the Thomsen–Friedenreich (TF)
tumor-associated carbohydrate antigen (TACA) disaccharide (Galb1-3GalNAc-a-O-Ser/Thr) as well as the assembly of a suitably
linked designer glycopeptide as a precursor to similar multivalent presentations on gold. The TF-coated nanoparticles are highly
stable, water soluble, and easily handled. Improvements in the linker technology used to attach the disaccharide to the particles
led to a robust multivalent platform for the presentation of this important carbohydrate. The antigen retains all recognition char-
acteristics while displayed on this template as shown by several in vitro assays. This area of research could lead to the development
of novel therapeutic agents that inhibit protein–carbohydrate interactions.
Published by Elsevier Ltd.
1. Introduction
sis of glycopeptide-based immunogens would be an
important strategy to raise strong immune responses
that more closely resemble the actual components pre-
sented in vivo.5–7
The Thomsen–Friedenreich disaccharide is a human
tumor-associated carbohydrate antigen1 present in about
90% of carcinomas but is rarely expressed in normal tis-
sues.2 Presentation of this antigen derives primarily
from the truncation of O-linked glycans of the normal
phenotype that are attached to the extracellular domain
of cell surface (often mucin) proteins. The presence of
smaller O-linked glycans on surface of carcinomas cell
serves to reveal underlying peptide (and other) epitopes
that would be masked in normal tissue. Not only is a
powerful immune response mounted against these ꢀnon-
selfꢁ antigens on tumors, the TF antigen is also used as a
diagnostic marker for preclinical tumor detection.2
These properties have prompted several groups to devel-
op immunotherapies and vaccine constructs based on
the TF structure as well as its immediate precursor,
TN antigen (GalNAca-O-Ser/Thr).3,4 In addition, the
unmasking of peptide epitopes suggests that the synthe-
Recent studies by the Glinsky group have unequivocally
proved the involvement of tumor cell TF structures in
invasion and metastasis of breast and prostate tumors.8
In particular, docking of tumors bearing TF structures
to the microvascular endothelium is mediated by the
b-galactosyl-binding protein galectin-3. These interac-
tions are proposed to cause the critical heterotypic adhe-
sion and homotypic aggregation of tumor cells that are
important steps in the metastatic cascade as well as
being required for cell survival and growth.9–11 Both
the adhesion events and cell growth are inhibited by
TF function-blocking or anti-galectin-3 antibodies.
Based on these results, the TF disaccharide and sur-
rounding in vivo peptide sequences are prime candidates
for the development of inhibitors of tumor cell growth
and metastasis.12
The well known cluster glycoside effect13 has spawned
the construction of myriad multivalent glycopolymers
for the study of carbohydrate–protein interactions and
*
Corresponding author. Tel.: +1 301 846 5905; fax: +1 301 846
6033; e-mail: barchi@helix.nih.gov
0957-4166/$ - see front matter Published by Elsevier Ltd.
doi:10.1016/j.tetasy.2004.12.003

588
S. A. Svarovsky et al. / Tetrahedron: Asymmetry 16 (2005) 587–598
the development of infinitely more potent inhibitors of
these processes as compared to monomeric sugars.14,15
Since clustering of galectins while binding cellular sugar
ligands seems to be critical for their function,10 multiva-
lent templates displaying galectin ligands could be extre-
mely useful tools for unraveling separate biochemical
functions. We have recently been interested in develop-
ing new multivalent systems to display TF antigen.16,17
Applying the concept of ꢀglyconanotechnologyꢁ as orig-
inally put forth by the group of Penades and co-work-
ers18–21 and others,22–25 we prepared 3-dimensional self
assembled monolayers (3D-SAMs) of gold atoms stabi-
lized by appropriately linked TF molecules. These
assemblies present the antigen in a manner that is recog-
nized by both lectins and antibodies, and they show
marked antimetastatic activity in vivo. We have also im-
proved on our synthesis of TF–glycoamino acid build-
ing blocks and applied these to the synthesis of a
designer glycopeptide with a linker compatible with pro-
ducing clusters of glycopeptides on 3D SAMs. This re-
port describes the synthesis and some of the properties
of these systems.
cursor for the synthesis of the gold nanoparticles. Inter-
mediate was treated as previously described
1
(glycosylation with trichloroacetimidate 527 and azide
reduction) to give 2.26 Reaction of 2 with thiolacetic acid
in the presence of AIBN smoothly and quantitatively
gave thioacetate 3, which upon removal of the protect-
ing groups afforded disulfide dimer 4 as a precursor to
nanoparticle synthesis.
Self-assembly to TF-coated gold particles (a-TFC5Au; a-
refers to the stereochemistry of the reducing end and C5
to the pentane linker) was carried out by slight modifica-
tions to now well documented procedures for the prep-
aration of water soluble carbohydrate-encapsulated
gold nanoparticles.21–24 Briefly, a solution of disulfide
(or corresponding monomeric thiol) and tetrachloro-
auric acid (HAuCl4) in water is treated at 0 ꢀC with an
excess of NaBH4 and stirred for a short time (1–2 h).
Purification is performed by ultrafiltration or centrifuga-
tion with methanol (gold SAMs are insoluble in MeOH
but freely soluble in water). Figure 1A shows a compar-
1
ison of the H NMR spectrum of a-TFC5Au with the
dimer TF-linked thiol (prepared by DTT treatment of
disulfide 4). All the features of the monomeric sugar
are present in a-TFC5Au but exhibit T2 broadening
due to the size of the particle and clustering of the sug-
ars. The UV spectrum of a-TFC5Au in Figure 1B shows
the characteristic plasmon band at 520nm and selected
transmission electron micrographs (TEM) that reveal
that the size of the particles is between 1 and 2 nm are
shown in Figure 1C. Particles can be fine tuned to vari-
ous sizes from 1 to 12 nm depending on reaction condi-
2. Results and discussion
2.1. Synthesis and evaluation of TF–AU particles
We have previously reported the synthesis of pentenyl
glycosides from the 2-azido-1-nitrate of 3,4,6-triacetyl
galactopyranose that were transformed to protected
TF disaccharide analogues and TF glycoamino acids
through simple and high-yielding protocols.26 Scheme
1 briefly depicts the key intermediates involved and
shows the transformation of the pentenyl group to a pre-
tions
transformations from 1 led to TN-coated gold nanopar-
(reductant,
temperature,
time).
Similar
ticles (data not shown, to be published elsewhere).
Scheme 1. Reagents and conditions: (i) 7, TMSOTf/CH2Cl2/molecular sieves, 90%; (ii) Zn, AcOH, Ac2O, THF, 90%; (iii) 80% AcOH, 83%; (iv)
AcSH/AIBN/1,4-dioxane, quant.; (v) cat NaOMe/MeOH, 97%; (vi) NaBH4, HAuCl4, H2O, 0 ꢀC.

S. A. Svarovsky et al. / Tetrahedron: Asymmetry 16 (2005) 587–598
589
Figure 1. (A) Comparison of the 1H NMR spectra of compound 4 and a-TFC5Au. Details are observable for most protons in the spectrum of a-
TFC5Au. (B) UV spectrum of a-TFC5Au showing the characteristic plasmon band at 520nm. (C) Transmission electron micrographs of a-TFC5Au
showing the uniform size and dispersion of the particles.
2.2. Refinement of the linker
cluded lectin affinity chromatography (LAC),
agglutination of rhodamine-labeled peanut agglutinin
(PNA), and aggregation of a TF-specific antibody33 (see
Experimental section for details). The TF-coated parti-
cles were retained on agarose-bound PNA affinity gels
but did not interact with agarose-bound Pisum sativum
agglutinin (PSA), an a-mannose-specific lectin. Either
a-TFC5Au or a-TFPEG6Au could be eluted from the gel
with competing amounts of galactose. As a testament to
the recognition capacity and specificity of PNA to the
antigen on the particles, the b-TFPEG6Au (vide supra)
did not bind to the column, confirming the requirement
of the natural a-stereochemistry at the reducing end of
the antigen for interaction with the protein (Fig. 2A).
Strong agglutination of rhodamine-labeled PNA was ob-
served by confocal microscopy after treatment with a-
TFPEG6Au whereas no change was observed when rhod-
amine-labeled PSA was used in the same assay (Fig. 2B).
These results were confirmed by agglutination of the
divalent monoclonal IgG(3) antibody F11 with the a-
TFPEG6Au (data not shown). We have performed both
anti-tumor and anti-metastatic bioassays with TFPEG6Au
and have shown that these particles have both in vitro
activity against tumor growth at specific doses and almost
completely inhibit lung metastasis in vivo against an im-
planted metastatic breast cancer cell line (Heimburg, Bar-
chi, Svarovsky, Rittenhouse-Olson, unpublished results).
Although various tests (vide infra) confirmed the func-
tionality and recognition of the antigen in the pentanet-
hiol-linked particles, we reasoned that a simple pentyl
chain may not be the optimum linker for future biologi-
cal studies. In addition, there is evidence that the linker
can play a role in the increased potency observed with
specific multivalent carbohydrate templates.28 Hence
we prepared other particles based on a synthetic linker
that retained the C5 chain and added a polyethylene gly-
col (PEG) spacer between this chain and the antigen.
PEGylation of peptides, proteins, and orally available
drugs can impart highly desirable effects such as longer
circulating half lives, protection from immunogenicity,
and increased water solubility.29–31 Following the work
of Palegrosdemang et al.,32 we prepared a hexa-PEG
(PEG6) precursor 6 armed with a masked thiol on one
end and a free acceptor hydroxyl group at the other
(Scheme 2). Glycosylation of 6 with the known trichlo-
roimidate 7 afforded a 79% yield of a separable a/b
(4:1) mixture of the glycoconjugate 8, which was pro-
cessed as before through 9 to the thiol 10 in high overall
yield. Gold particles were assembled with both anomers
of this ligand identically as for a-TFC5Au. These particles
(a-TFPEG6Au, bTFPEG6Au) were extremely stable for
months at room temperature. All further studies were
performed with PEG6-linked particles.
2.4. TF-glycopeptide precursors for multivalent assembly
on gold
2.3. In vitro functional assays
Several assays were performed that validated the func-
tional nature of the antigen on the particles. These in-
As mentioned briefly above, TF and other O-linked
TACAꢁs are covalently attached to specific cell surface

590
S. A. Svarovsky et al. / Tetrahedron: Asymmetry 16 (2005) 587–598
O
50 % NaOH
Br
HO
PEG6
H
6
AcSH, AIBN
MeOH/h
HO
O
HO
O
O
S
6
6
6
O
O
OAc
AcO
AcO
TMS-OTf, 6
O
OAc
O
O
AcO
AcO
O
O
O
N
O
DCM/THF
H
O
O
N
OAc
3 O
CCl3
NH
OAc
79%
3O
O
S
α:β = 4:1
7
6
8
Zn, Ac2O,
AcOH, THF
OH
HO
O
OH
HO
HO
O
O
OAc
O
O
AcO
AcO
1. Cat. AcCl/MeOH
2. NaOMe/MeOH
O
HN
O
OH
O
O
O
SH
O
6
O
OAc
NH
O
10
O
O
S
6
9
Scheme 2.
Figure 2. (A) Structures a- and b-TFPEG6Au highlighting the anomeric stereochemistry and the binding propensity toward PNA. (B) TEMꢁs of
Rhodamine-PNA and Rhodamine-PSA with (right) and without (left) addition of a-TFPEG6Au. Also shown is the definition of the schematic for the
gold particles.
proteins such as mucins. Hence, when an immune re-
sponse is mounted against a particular TACA, the gly-
cans are presented in the context of the protein
backbone and its surrounding sequence at the point of
attachment. It seems reasonable then to assume that
the ideal construction to develop as an immunogen in
a putative vaccine preparation would be a combination
of both the sugar and the peptide sequence, that is, a

S. A. Svarovsky et al. / Tetrahedron: Asymmetry 16 (2005) 587–598
591
glycopeptide. As shown above, thiol-stabilized gold
nanoparticles are extremely useful templates for multi-
valent presentation. It was our thought that relevant
tumor-associated glycopeptides could similarly be
displayed on these particles and we sought to capitalize
on this strategy by constructing suitably linked glycosyl-
ated peptides to be displayed on gold 3D-SAMꢁs.
derivatives 15a and 15b44 proceeded with excellent yield
and stereoselectivity, but under strictly optimized condi-
tions. Reactions in CH2Cl2 at ꢀ40 ꢀC gave a 1:1 mixture
of anomers 17a or 17b, whereas reaction at the same
temperature in THF gave exclusively a-product, but in
low yield (40%). Raising the temperature to 25 ꢀC afford-
ed a 2:1 mixture of a/b products if the reaction was run
in CH2Cl2; however, in a 3:1 mixture of CH2Cl2 to THF,
the a/b ratio increased to 20:1 with yields consistently in
the 90–95% range. Although more steps are involved in
contrast to previous syntheses, the overall yield from
simple D-galactose and the ease of most synthetic
manipulations makes this scheme attractive for large
syntheses of these critical building blocks. In addition,
the use of the phenacyl ester for carboxyl protection
on the amino acid allowed us to remove this group
and reduce the azide in one simple high-yielding step
to give the glycopeptide precursors 17a,b.
In our previous work that outlined the preparation of 2,
we also reported the synthesis of serine and threonine
glycoamino acid building blocks for use in glycopeptide
synthesis.26 Although 2 is a versatile intermediate for
both glycosylations and linker chemistry, the pentenyl
group proved to be partly problematic as a glycosyl do-
nor to amino acid acceptors with respect to yields (43–
64% with threonine and serine donors, respectively).
Over the past 20years, there have been a host of pub-
lished efficient syntheses of TF antigen neoglycoconju-
gates14,16,34–39 and TF glycoamino acids.5,7,36,40–44 In
our continuing efforts toward high-yielding syntheses
of these building blocks, we revisited some of the work
of Schmidt39,45 and found some marked improvements
in specific steps. Most notably, we have succeeded in
obtaining extremely high yields and stereoselectivities
in the glycosylation of Fmoc-serine and threonine
phenacyl esters prepared according to Luning
et al.44,46 (Scheme 3). Compound 1147 (derived from D-
galactose through, (1) triacetyl galactal azidonitration,
(2) removal of the anomeric nitro group with hydrazine
hydrate, (3) silylation, (4) deacetylation, and (5) benzyli-
denation in 30% overall yield) was glycosylated with 5 at
ꢀ45 ꢀC to give the known48 disaccharide 12 quantita-
tively and with exclusive b-stereochemistry. When the
reaction was run at 0 ꢀC, a 1:1 mixture of anomers
was formed. The once troublesome anomeric desilyl-
ation step,39,49 that was later optimized on a related ana-
logue by Schmidtꢁs group,45 was refined by the use of
triethylamine/trihydrofluoride complex in THF,50 which
removed the TBDMS group cleanly and quantitatively
to give 13, which was subsequently converted to the tri-
chloroacetimidate 147 in 95% yield by the procedure of
Ren and Liu.27 Glycosylations with serine and threonine
2.5. Attempt at glycopeptide-coated nanoparticle
synthesis
We proceeded to prepare a model glycopeptide precur-
sor for presentation on gold nanoparticles (schematic
in box, Fig. 3). This glycopeptide 18 was previously de-
signed to bind to MHC-I isotypes where the carbohy-
drate was placed in an anchor position known to bind
in a pocket of the T-cell receptor (TcR).51 Glycopeptide
18 was shown to inhibit binding of a well-characterized
Kb-binding peptide with a 4 nM IC50 and elicits a strong
cytotoxic T-lymphocyte (CTL) response against the TF
disaccharide.51 We reasoned that a multivalent presen-
tation of this glycopeptide may lead to an enhanced
immune response and possibly more powerful
glycopeptide-restricted CTL. For the synthesis of gold
particles, we had the requirement that the peptide must
be extended by a linker that could be processed to a ter-
minal thiol group. We prepared the nona-glycopeptide
18 by standard peptide synthesis on 2-chlorotrityl resin
with HATU/HOAt activation of individual amino acids.
We succeeded in extending the N-terminus as shown in
Figure 3B that led to a precursor to particle assembly
Ph
O
Ph
Ph
O
O
OAc
OAc
O
O
O
OAc
O
OAc
O
O
O
O
ii
OTBDMS
i
OH
OTBDMS
AcO
O
AcO
O
HO
OAc
N3
OAc
N3
N3
13
12
11
iii
Ph
Ph
Ph
O
O
O
O
O
OAc
OAc
O
O
OAc
OAc
O
OAc
O
OAc
O
iv
v
O
O
O
H
R
O
O
AcO
AcO
O
AcO
NHFmoc
OH
O
NHFmoc
O
NHFmoc
O
O
15a R= H
15b R= Me
OAc
OAc
N
NH
OAc
N
3O
O
3O CCl3
NH
HO
R
O
R
14
O
Ph
O
Ph
17a R=H
17b R= Me
16a R= H
16b R=Me
Scheme 3. Reagents and conditions: (i) 5, TMSOTf/CH2Cl2/molecular sieves, ꢀ45 ꢀC, 100%; (ii) Et3NÆ3HF, THF, 100%; (iii) CH2Cl2, CCl3CN,
DBU, 0 ꢀC, 95%; (iv) 3:1 CH2Cl2:THF, 15a,b, rt, 90–95%; (vi) 2:2:1 THF/AcOH/Ac2O, Zn, rt, 79%.

592
S. A. Svarovsky et al. / Tetrahedron: Asymmetry 16 (2005) 587–598
Figure 3. (Box) Schematic of a glycopeptide-coated gold particle. (A) Structure of the designer glycopeptide 18 and linker 19 used for the synthesis of
the precursor to gold self assembly. (B) Structure of final product 20 used in the particle synthesis.
making use of the known spacer52 19 that was processed
to a terminal thiolated glycopeptide 20 through imino-
thiolane53 chemistry. Our initial attempts to synthesize
glycopeptide-stabilized gold looked promising, but the
insolubility of the resulting particle precluded character-
ization of the assembly. A previous synthesis of the
homo serine analogue of 18 also reported solubility
problems with the deprotected glycopeptide.7 Our hope
was to add enough hydrophilicity in the linker to over-
come this drawback. Further research into solving this
problem and the synthesis of other glycopeptide systems
for multivalent presentations will be reported in due
course.
4. Experimental
Melting points were determined on Fisher–Johns melt-
ing point apparatus and are uncorrected. Rf values refer
to TLC performed on Analtech Uniplates GF pre-
coated with silica gel 60to a thickness of .025 mm.
The spots were visualized by charring with a solution
of ammonium molybdate (IV) tetrahydrate (12.5 g)
and cerium (IV) sulfate tetrahydrate (5.0g) in 10% aque-
ous H2SO4 (500 mL). Flash column chromatography
(FCC) was performed under medium pressure using sil-
ica gel 60 (230–400 mesh, E. Merck) and usually em-
ployed a stepwise solvent polarity gradient, correlated
with TLC mobility. Unless otherwise noted, chemicals
were purchased from Aldrich–Sigma (Milwaukee, WI)
and used without further purification.
3. Conclusion
NMR spectra were recorded on a Varian Inova-400
instrument with residual CHCl3 (7.26 ppm) as the inter-
nal standard at frequencies of 399.74 MHz for 1H
and 100.51 MHz for 13C. Assignments were based
on gCOSY, TOCSY, ROESY, and 13C/DEPT experi-
ments. 1H NMR data are tabulated in the order of multi-
plicity (s, singlet; d, doublet; dd, doublet of doublets; dt,
double of triplets; t, triplet; q, quartet; m, multiplet; br s,
broad signal), number of protons, and coupling con-
stant(s) in hertz. IR spectra were recorded on a JASCO
FT/IR-615 spectrophotometer. Specific optical rotations
were determined using JASCO-P1010 polarimeter in
0.5 dm cuvette at 589 nm in chloroform. Five consecutive
We have successfully prepared gold nanoparticles
coated with the important TF antigen in a robust and
highly active form. These particles have interesting
activities both in vitro and in vivo. The chemistry to pre-
pare the TF precursors was high yielding and stereose-
lective and the convergence of our routes allows the
synthesis of glycopeptide building blocks in an equally
efficient manner. We are using similar chemistry to ex-
ploit these useful templates for the display of other bio-
logically relevant glycans and glycopeptides. These
systems should find extended use both as chemical and
biological tools and have the potential to develop into
anticancer therapeutics.20,54,55

S. A. Svarovsky et al. / Tetrahedron: Asymmetry 16 (2005) 587–598
593
measurements were taken and the average value is re-
ported. Positive ion fast-atom bombardment mass spec-
tra (FABMS) were obtained at an accelerating voltage
of 6 kV and a resolution of 2000. Glycerol was used as
the sample matrix, and ionization was effected by xenon
atoms. Laser scanning confocal microscopy was per-
formed on a Zeiss 510confocal microscope (NCI-Fred-
erick, Confocal Microscopy Facility, Image Analysis
Lab). Elemental analyses were performed by Atlantic
Microlab, Inc., Norcross, GA. Analyses of some PEG-
containing compounds (viz., 8) did not meet the ex-
pected criteria but all other spectroscopic data were con-
sistent with the desired structure. Transmission electron
micrographs were performed on a Hitachi H-7000
microscope equipped with a Gatan digital camera oper-
ating at 75 kV.
4.51 (td, 1H, J = 3.90, 10.54 Hz, H0-2), 3.65–4.18 (m,
8H, H0-5, H0-3, H00-6, H00-6, H0-4, H00-5), 3.60–3.67 (m,
1H, –OCH2CH2CH2CH2CH2SAc), 3.31–3.40(m, 1H,
–OCH2CH2CH2CH2CH2SAc), 2.78–2.85 (m, 2H,
–OCH2CH2CH2CH2CH2SAc), 2.70(br s, 2H, OH),
2.28 (s, 3H, SAc), 1.92, 1.94, 2.00, 2.02, 2.11 (s, 15H,
OAc, NHAc), 1.50–1.62 (m, 4H, –OCH2CH2CH2-
CH2CH2SAc), 1.32–1.42 (m, 2H, –OCH2CH2CH2-
CH2CH2SAc), 13C NMR (100 MHz, CDCl3) d =
196.46, 170.61, 170.39, 170.32, 169.72, 169.55, 101.86,
97.87, 78.38, 71.07, 70.82, 69.69, 69.37, 68.83, 67.89,
67.10, 62.88, 61.52, 48.06, 30.86, 29.57, 29.03, 28.88,
25.61, 23.56, 20.87, 20.81, 20.79, 20.72, 20.67. FAB
MS m/z: 696.1 (MH+). Anal. Calcd for C29H45NO16S:
C 50.06; H 6.52; N 2.01. Found: C 49.55; H 6.53;
N 2.00.
4.2. TF-pentyl-disulfide 4
4.1. 5-O-[2-Acetamido-2-deoxy-3-O-(2,3,4,6-tetra-O-acet-
yl-b-D-galactopyranosyl)-b-D-galactopyranosyl]-1-a-
thioacetyl pentane 3
A solution of 300 mg (0.43 mmol) of 3 in 5 mL of
MeOH was treated with 30 lL of 25% (w/v) NaOMe/
MeOH. The reaction was stirred at rt and for 24 h
with a slow flow of air bubbling through the solution.
Initially, two spots were observed by TLC (Rf 0.13
and Rf 0.00 in 20% v/v MeOH/CH2Cl2). The higher
Rf spot corresponded to the monomeric thiol while the
lower spot corresponded to the desired dimer 4. This
mixture was completely converted to dimer 4 after
24 h. The reaction was carefully neutralized with
strongly acidic Amberliteꢁ-120, and evaporated under
reduced pressure at 50 ꢀC to give 200 mg (96%) of
NMR pure 4, which was further purified by reverse
phase (C18) chromatography with 10% ! 40% (v/v)
MeOH/H2O to give 187 mg (90%) of 4 as a white pow-
der. The compound was soluble in both water and meth-
anol. Mp 247–249 ꢀC; [a]D =+93.8 (c 1.6, MeOH); IR
(neat) 3379.64, 2944.77, 2827.13, 2112.64, 1746.23,
1218.79; 1H NMR (400 MHz, CD3OD) d = 4.14 (d,
2H, J = 1.95Hz, NH), 3.87 (dd, 2H, J = 3.12, 10.93Hz,
H0-2), 3.68 (m, 2H, [–OCH2–]2), 3.40(m, 2H, –OC H2–),
2.68 (t, 4H, J = 7.42 Hz, –CH2S–SCH2–), 1.94
(s, 6H, NHAc), 1.43–1.74 (m, 12H, [–OCH2CH2-
CH2CH2CH2S–]2); 13C NMR (100 MHz, CDCl3)
d = 172.77 (NHAc), 105.08 (C-1), 97.60, 77.76, 75.52,
73.52, 71.33, 70.85, 69.07, 68.83, 67.54, 61.64, 61.38,
49.19, 38.41, 28.89, 28.81, 24.93, 21.69. FAB MS m/z:
969.2 (MH+). Anal. Calcd for C38H68N2O22S2
(3 · H2O) C, 44.61; H, 7.29; N, 2.74. Found: C 44.24;
H 7.21; N 2.73.
A solution of 120mg (0.17 mmol) of 226 in 4 mL of
anhydrous 1,4-dioxane was purged with argon for
20min. To this deoxygenated solution 1.4 mL
(2.55 mmol, 15 equiv) of triply-distilled thiolacetic acid
was added followed by 30 mg (0.03 mmol) of AIBN.
The reaction was left to stir under an argon atmosphere
at 75 ꢀC for 1 h. The reaction was quenched with cyclo-
hexene (0.1 mL) and the solution was co-evaporated
with xylenes (3·) under reduced pressure. FCC on silica
gel with 3:1 EtOAc/hexanes provided 125 mg (99.8%) of
white solid (7:3 mixture of amide rotamers). Rf 0.19 (3:1
EtOAc/hexanes); [a]D =+157.8 (c 0.22, CHCl3); IR
1
(neat) 3099.05, 1752.01, 1689.34, 1368.25, 1220.72; H
NMR (400 MHz, CDCl3) d = 7.48–7.54 (m, 2H, Ph),
7.27–7.38 (m, 3H, Ph), 5.56 (d, 1H, J = 8.98Hz, NH),
5.52 (s, 1H, PhCH), 5.34 (d, 1H, J = 2.34 Hz, H00-4),
5.13–5.19 (m, 1H, H00-2), 4.92–4.98 (m, 2H, H00-3, H00-
1), 4.76 (d, 1H, J = 7.81 Hz, b H00-1), 4.62 (m, 1H, H00-
2), 3.90–4.30 (m, 6H, H0-6, H00-6, H0-4, H00-5), 3.88
(m, 1H, H0-3), 3.59 (br s, 1H, H00-5), 3.60–3.70 (m, 1H,
–OCH2CH2CH2CH2CH2SAc), 3.35–3.45 (m, 1H,
–OCH2CH2CH2CH2CH2SAc), 2.83 (t, 2H, J = 7.42 Hz,
–OCH2CH2CH2CH2CH2SAc), 2.30(s, 3H, SAc), 1.93,
1.96, 2.00, 2.01, 2.11 (s, 15H, OAc, NHAc), 1.52–1.62
(m, 4H, –OCH2CH2CH2CH2CH2SAc), 1.34–1.44 (m,
2H, –OCH2CH2CH2CH2CH2SAc). FAB MS m/z:
784.0(MH +). Anal. Calcd for C36H49NO16S: C 55.16;
H 6.30; N 1.79. Found: C 54.89; H 6.32; N 1.87. A solu-
tion of 110mg (0.14 mmol) of this product in 3 mL of
80% AcOH was stirred at 45 ꢀC for 16 h. The reaction
solution was concentrated at reduced pressure and co-
evaporated twice with xylenes. The residue was purified
by FCC on silica gel using 7% MeOH in CH2Cl2 to pro-
vide 69 mg (71% yield) of 3 as a white foam (7:3 mixture
of amide rotamers). Mp 95–97 ꢀC; Rf 0.46 (10% MeOH
in CH2Cl2); [a]D =+107.7 (c 0.15, CHCl3); IR (neat);
3552.24, 1750.08, 1688.37, 1656.55, 1545.67, 1370.18,
1220.72; 1H NMR (400 MHz, CDCl3) d = 5.33 (dd,
1H, J = 0.78, 3.51 Hz, H00-4), 5.11–5.18 (m, 1H, H00-2),
4.94 (dd, 1H, J = 3.51, 10.54 Hz, H0-3), 4.78 (d, 1H,
J = 3.51 Hz, H0-1), 4.61 (d, 1H, J = 8.20Hz, bH00-1),
4.3. General preparation of TF-linked gold nanoparticles
A solution of 1% HAuCl4 (1.9 mL, 0.046 mmol in Au3+
)
was added to 200 mL of degassed H2O. To this solution
was added 10mg (0.010mmol) of thiol 4 or disulfide
10 (0.020 mmol per thiol; ratio of reactants: TFSH/
Au = 1:2.3).
A
cooled (0 ꢀC) solution of 20mg
(0.53 mmol) of NaBH4 in 25 mL H2O was added drop-
wise to the reaction solution with rapid stirring. The
resulting purple solution was stirred for an additional
1 h, purified by ultrafiltration on Milliporeꢁs Centriplusꢁ
YM-30cartridges and freeze-dried.

594
S. A. Svarovsky et al. / Tetrahedron: Asymmetry 16 (2005) 587–598
4.4. Preparation of linker 6
1H NMR (400 MHz, CD3Cl) d = 7.39–7.43 (m, 2H,
Ph), 7.20–7.28 (m, 3H, Ph), 5.43 (s, 1H, PhCH), 5.28
(m, 1H, H-400), 5.17 (dd, 1H, J 0= 7.88, 10.20 Hz, H-200),
4.98 (d, 1H, J = 3.71 Hz, H-1 ), 4.92 (m, 1H, H-300),
4.67 (d, 1H, J = 7.88 Hz, H-100), 4.26 (m, 1H, H-40),
3.8–4.15 (m, 8H, H-20, H-30, H-60, H-600, H-50, H-500),
3.52 (m, 24H, PEG), 3.32 (m, 2H, OCH2CH2CH2-
CH2CH2SAc), 2.74 (m, 2H, OCH2CH2CH2CH2CH2-
SAc), 2.19 (s, 3H, SAc), 2.04, 1.94, 1.92, 1.86 (s, 12H,
4 · Ac), 1.46 (m, 4H, OCH2CH2CH2CH2CH2SAc),
A solution of 533 lL of 50% NaOH (6.7 mmol) was
added to 10g (35.5 mmol) of hexa(polyethyleneglycol)
(Aldrich) at 100 ꢀC. The reaction was stirred for
30min at 100 ꢀC and 5-bromo-1-pentene (1 g, 6.7 mmol)
was quickly added and the reaction was stirred at 100 ꢀC
for 24 h. The reaction was diluted with water and ex-
tracted 6· with EtOAc. Combined organic extracts were
evaporated and separated by FCC with 10:1 CH2Cl2/
MeOH. This gave 1 g of monopentenylated hexa(poly-
ethyleneglycol) as a clear liquid. Rf (10:1, CH2Cl2/
MeOH) = 0.3; IR (neat); 3465 (br s), 2231, 2022, 1690;
1H NMR (400 MHz, CD3Cl) d = 5.73–5.83 (m, 1H,
CH@CH2), 4.90–5.02 (m, 2H, CH@CH2), 3.53–3.70
(m, 24H, PEG), 3.43 (t, 2H, J = 6.49 Hz, OCH2-
CH2CH2CH@CH2), 2.08 (m, 2H, OCH2CH2CH2-
CH@CH2), 1.65 (m, 2H, OCH2CH2CH2CH@CH2);
13C NMR (100 MHz, CD3Cl) d = 72.49, 70.66, 70.59,
70.55, 70.52, 70.33, 70.07, 61.68, 30.19, 28.75. FAB
MS m/z: 351.3 (MH+), 373.3 (M + Na+). Anal. Calcd
for C17H34O7: C, 58.26; H, 9.78. Found: C 58.30; H
9.78. A solution of 2.8 g (8 mmol) of this product in
30mL MeOH was treated with 30mmol (2.3 g,
2.3 mL) of freshly distilled HSAc. The solution was
purged with argon for 20min and 10mg of AIBN was
added. The reaction was irradiated with 360nm light
overnight, quenched with 1 mL of cyclohexene and
evaporated. The residue was purified by FCC with
15:1 CH2Cl2/MeOH to afford 3 g (88%) of thioacetate
8 as a clear liquid. IR (neat); 3465 (br s), 1689; 1H
NMR (400 MHz, CD3Cl) d = 3.52–3.72 (m, 24H,
PEG), 3.42 (t, 2H, J = 6.49 Hz, OCH2CH2CH2CH2-
CH2SAc), 2.84 (d, 2H, J = 6.95Hz, OCH2CH2CH2CH2-
CH2SAc), 2.60(br s, 1H, OH), 2.29 (s, 3H, SAc), 1.56
(m, 4H, OCH2CH2CH2CH2CH2SAc), 1.39 (m, 2H,
OCH2CH2CH2CH2CH2SAc); 13C NMR (100 MHz,
CD3Cl) d = 195.91, 72.47, 71.09, 70.61, 70.57, 70.55,
70.36, 70.09, 61.72, 30.62, 29.34, 29.10, 29.01, 25.36;
FAB MS m/z: 427.4 (MH+). Anal. Calcd for
C19H38O8S: C, 53.50; H, 8.98. Found: C 52.17;
H 8.91.
1.30(m, 2H, OCH CH2 CH2CH2CH2SAc); 13C NMR
2
(100 MHz, CD3Cl) d = 195.86, 170.28, 170.23, 170.11,
169.38, 137.73, 128.81, 128.22, 128.08, 126.50, 126.13,
102.41, 100.58, 98.87, 75.81, 71.06, 70.77, 70.55, 70.50,
70.22, 70.07, 69.19, 68.67, 67.51, 66.97, 63.02, 62.39,
61.37, 58.84, 30.61, 29.64, 29.34, 29.09, 28.99, 25.35,
20.69, 20.67, 20.54; FAB MS m/z: 1031.9 (MH+),
1069.9 (M+K+).
4.6. Thioacetic acid-[2-acetamido-2-deoxy-4,6-O-benzyl-
idene-3-O-(2,3,4,6-tetra-O-acetyl-b-D-galactopyranosyl)-
a-D-galactopyranosyl]-S-{5-[(penta-2-ethoxy)-2-hydroxy-
ethoxy]-pentyl ester 9
A solution of 1.5 g of 8 in 50mL of 6:3:1 THF/AcOH/
Ac2O was treated with 15 g of Zn dust. After 4 h the
reaction was filtered through a pad of Celiteꢁ and the
zinc cake was washed with EtOAc. The filtrate was
washed successively with water, satd NaHCO3 and
brine. The organic layer was dried over Mg2SO4, evap-
orated and purified by FCC with 20:1 CH2Cl2/MeOH
to afford 1.2 g (79%) of 9 as a glassy solid. Rf (20:1
1
CH2Cl2/MeOH) = 0.2; [a]D =+72.4 (c 1.37, CHCl3); H
NMR (400 MHz, CD3Cl) d = 7.52–7.57 (m, 2H, Ph),
7.30–7.40 (m, 3H, Ph), 5.96 (d, 1H, J = 9.27 Hz, NH),
5.56 (s, 1H, PhCH), 5.38 (m, 1H, H-400), 5.20(dd, 1H,
J = 7.88, 10.66 Hz, H-20), 4.99 (m, 2H, H-10, H-300),
4.77 (d, 1H, J = 7.88 Hz, H-100), 4.67 (m, 1H, H-20),
0
3.72–4.30(m, 8H, H-2 , H-30, H-60, H-600, H-50, H-500),
3.65 (m, 24H, PEG), 3.44 (t, 2H, J = 6.49 Hz,
OCH2CH2CH2CH2CH2SAc), 2.86 (t, 2H, J = 7.42 Hz,
OCH2CH2CH2CH2CH2SAc), 2.32 (s, 3H, NHAc),
2.15 (s, 3H, SAc), 2.05, 2.04, 1.98, 1.97 (s, 12H,
4 · Ac), 1.59 (m, 4H, OCH2CH2CH2CH2CH2SAc),
1.42 (m, 2H, OCH2CH2CH2CH2CH2SAc); 13C NMR
(100 MHz, CD3Cl) d = 170.27, 170.12, 169.53, 169.36,
137.74, 128.71, 128.10, 128.06, 126.62, 126.21, 126.06,
101.40, 100.59, 98.57, 75.70, 74.52, 71.05, 70.85, 70.74,
70.54, 70.52, 70.36, 70.06, 69.99, 69.27, 68.81, 67.23,
66.96, 63.13, 61.25, 18.12, 30.60, 29.32, 29.08, 28.98,
25.33, 23.39, 20.68, 20.53. FAB MS m/z: 1048.2
(MH+). Anal. Calcd for C48H73NO22S: C, 55.00; H,
7.02; N, 1.34; O, 33.58; S, 3.06. Found: C 54.57; H
7.06; N 1.43.
4.5. Thioacetic acid-[2-azido-2-deoxy-4,6-O-benzylidene-
3-O-(2,3,4,6-tetra-O-acetyl-b-D-galactopyranosyl)-a-D-
galactopyranosyl]-S-{5-[(penta-2-ethoxy)-2-hydroxy-eth-
oxy pentyl ester 8
A solution of 2.5 g (3.26 mmol) of 7 and 1 g (2.36 mmol)
of 6 in CH2Cl2 was evaporated and co-evaporated with
toluene and the residue was dried under vacuum over-
night. The dried mixture was dissolved in 3:1 mixture
of anhydrous CH2Cl2/THF (60:20 mL) and added
˚
via cannula to 2 g of flame-dried MS (4 A). After stirring
for 20min 15 lL TMSOTf was quickly injected via syr-
inge. After 30min the reaction was quenched with Et 3N,
filtered, and evaporated. The crude residue was purified
by FCC with 20:1 CH2Cl2/MeOH to afford 1.9 g of 8 as
a mixture of anomers (1.5 g of a-anomer and 400 mg of
b-anomer, a/b = 3.75). Total yield 79%. Desired a-ano-
mer (glassy solid) Rf (10:1 CH2Cl2/MeOH) = 0.8; 4:1
ratio of rotamers. NMR of the major rotamer is given.
4.7. 5-Mercapto-pentyloxy-O-[2-acetamido-2-deoxy-3-O-
(b-D-galactopyranosyl)-a-D-galactopyranosyl]-(penta-2-
ethoxy)-ethanol 10
A solution of 1.2 g (1.14 mmol) of 9 in 20mL of dry
MeOH was treated with a few drops of acetyl chloride

S. A. Svarovsky et al. / Tetrahedron: Asymmetry 16 (2005) 587–598
595
for 1.5 h, quenched with pyridine, evaporated, and puri-
fied by FCC with 20:1 ! 10:1 CH2Cl2/MeOH to afford
700 mg (65%) of debenzylidenated 9 as a glassy solid.
Rf (10:1 CH2Cl2/MeOH)= ; [a]D =+47.6 (c 0.60,
EtOAc/hexanes gave 15 g (100%) of 1248 as a mixture
of a,b isomers.
1
MeOH); H NMR (400 MHz, CD3Cl) d = 5.90(d, 1H,
4.9. 2-Azido-4,6-O-benzylidene-2-deoxy-3-O-(2,3,4,6-
tetra-O-acetyl-b-D-galactopyranosyl)-a-D-galactopyrano-
side 13
J = 9.74 Hz, NH), 5.37 (m, 1H, H-400), 5.20(dd, 1H,
J = 7.88, 10.66 Hz, H-200), 5.00 (m, 1H, H-300), 4.86 (d,
1H, J = 3.71 Hz, H-10), 4.66 (d, J = 8.34 Hz, H-10),
4.57 (m, 1H, H-20), 3.76–4.22 (m, 7H, H-30, H-60, H-
600, H-500, H-50), 3.54–3.72 (m, 24H, PEG), 3.45 (t, 2H,
J = 6.49 Hz, OCH2CH2CH2CH2CH2SAc), 2.86 (m,
2H, OCH2CH2CH2CH2CH2SAc), 2.32 (s, 3H, NHAc),
2.16 (s, 3H, SAc), 2.08, 2.06, 1.99, 1.98 (s, 12H,
4 · Ac), 1.59 (m, 4H, OCH2CH2CH2CH2CH2SAc),
1.42 (m, 2H, OCH2CH2CH2CH2CH2SAc); 13C NMR
(100 MHz, CD3Cl) d = 195.90, 170.36, 170.17, 170.06,
169.59, 169.35, 101.63, 98.15, 78.10, 71.04, 70.77,
70.64, 70.56, 70.53, 70.48, 70.44, 70.20, 70.04, 69.97,
69.81, 69.10, 68.58, 67.05, 66.94, 62.51, 61.28, 47.81,
30.59, 29.31, 29.05, 28.96, 25.32, 23.30, 20.64, 20.60,
20.57, 20.50. FAB MS m/z: 961.4 (MH+). Anal. Calcd
for C41H69NO22S: C, 51.29; H, 7.24; N, 1.46. Found:
C 50.47; H 6.91; N 1.55. This compound (500 mg,
0.52 mmol) in 20 mL of dry MeOH was treated with 7
drops of 25% w/v NaOMe/MeOH and stirred at room
temperature for an hour. The reaction was quenched
with Amberliteꢁ-120ion-exchange resin, filtered, and
evaporated. The residue was cleanly purified with 20%
MeOH in CH2Cl2 to give 350mg (90%) of 10 as a glassy
solid. Rf (20% MeOH in CH2Cl2) = 0.20; [a]D =+24.4 (c
A solution of 15 g (20mmol) of 12 in 250mL of anhy-
drous THF was treated with 20g (124 mmol) of Et 3NÆ
3HF. The reaction was stirred overnight at rt, diluted
with EtOAc, washed with satd NaHCO3, brine, and
water. The organic layer was dried over MgSO4 and
evaporated. The residue was purified by FCC with 7:3
EtOAc/hexanes to give 12.5 g (100%) of hemiacetal
1356 as a mixture of a,b isomers.
4.10. Preparation of 14
A solution of 10.7 g (17.2 mmol) of 8 in 200 mL of anhy-
drous CH2Cl2 (Aldrich) was cooled to 0 ꢀC and 20mL
(29 g, 200 mmol) of trichloroacetonitrile was added.
After stirring for a few min, 0.5 mL of DBU was added.
The reaction was stirred overnight, evaporated, and the
residue was purified by FCC with 3:2 hexanes/EtOAc to
give 12.5 g (95%) of pure 14.7
4.11. Na-Fluoren-9-ylmethoxycarbonyl-3-O-[2-azido-4,6-
O-benzylidene-2-deoxy-3-O-(2,3,4,6-tetra-O-acetyl-b-D-
galactopyranosyl)-a-D-galactopyranosyl]-L-serine phen-
acyl ester 16a
1
0.24, MeOH); H NMR (400 MHz, D2O) d = 4.80(d,
1H, J = 3.71 Hz, H-10), 4.35 (d, 1H, J = 7.88 Hz, H-
10), 4.24 (dd, 1H, J = 3.71, 10.66 Hz, H-20), 4.12 (m,
1H, H-400), 3.94 (dd, 1H, J = 2.78, 11.13 Hz, H-30),
0
3.48–3.90(m, 33H, 24H PEG, NH, H-3 , H-40, H-500,
A mixture of 1.05 g (2.36 mmol) of 15a 2.05 g
H-500, H-60, H-600), 3.43 (m, 3H, OCH2CH2CH2CH2-
CH2SH, H-20); 2.44 (m, 2H, OCH2CH2CH2CH2-
CH2SH), 1.91 (s, 3H, NHAc), 1.50(m, 4H,
OCH2CH2CH2CH2CH2SH), 1.33 (m, 2H, OCH2CH2-
CH2CH2CH2SH); 13C NMR (100 MHz, D2O)
d = 174.44, 101.75, 97.42, 97.30, 77.25, 74.92, 72.51,
70.87, 70.71, 70.57, 69.57, 69.11, 68.78, 68.69, 68.59,
68.48, 66.40, 61.16, 60.91, 48.45, 32.75, 27.98, 24.02,
23.66, 22.06, 22.04; FAB MS m/z: 750.3 (MH+), 788.3
(M+K+). Anal. Calcd for C31H59NO17S: C, 49.65; H,
7.93; N, 1.87. Found: C 49.28; H 7.86; N 1.82.
(2.67 mmol) of 14 was dissolved in dry CH2Cl2, evapo-
rated, and dried under vacuum for 1 h. The dried mix-
ture was dissolved 3:1 CH2Cl2/THF (60mL) under
argon and this solution was added under argon to 2 g
of flame-dried molecular sieves (40). After stirring for
30min at rt 15 lL of TMSOTf was added at once. After
1 h, the reaction was quenched with few drops of Et3N,
filtered through Celiteꢁ, and evaporated. Purification by
FCC with 1:1 ! 2:1 EtOAc/hexanes gave 1.95 g (80%)
of the a-isomer as a glassy solid. Rf (1:1, EtOAc/hex-
anes) = 0.33; [a]D =+34.05 (c 0.6 in CHCl3); IR (neat)
2231.24, 2114.56, 2022.96, 1751.05 cmꢀ1 1H NMR
;
(400 MHz, CDCl3) d = 7.30–7.90 (m, 18H, aromatic),
5.98 (d, 1H, J = 8.25 Hz, NH-Fmoc), 5.54 (s, 1H,
PhCH), 5.30–5.50 (m, 3H), 5.28 (dd, 1H,
J1,2 = 7.90Hz, J2,3 = 10.39 Hz, H200), 5.17 (d, 1H,
J = 2.95 Hz, aH10), 4.99 (dd, 1H, J3,4 = 3.45 Hz,
J2,3 = 10.39 Hz, H30), 4.77 (m, 1H, NHCH), 4.73 (d,
1H, J1,2 = 7.94 Hz, bH100), 3.75–4.50(m, 13H), 2.15,
2.03, 1.98 (s, 12H, Ac); 13C NMR (100 MHz, CDCl3)
d = 170.34, 170.15, 169.43, 143.66, 141.31, 137.65,
134.24, 133.89, 129.06, 128.89, 128.15, 127.85, 127.71,
127.18, 127.13, 126.13, 125.05, 120.12, 102.53, 100.62,
100.09, 75.79, 71.03, 70.82, 69.08, 68.65, 67.34, 66.98,
63.64, 61.41, 58.80, 54.55, 47.07, 20.72, 20.66, 20.57;
FAB MS (m/z): 1051.1 (MH+), 1023.1 (MH+ꢀN2) Anal.
Calcd for C53H54N4O19: C, 60.57; H, 5.18; N, 5.33.
Found: C 60.26; H 5.21; N 5.16.
4.8. 2-Azido-4,6-O-benzylidene-2-deoxy-3-O-(2,3,4,6-
tetra-O-acetyl-b-D-galactopyranosyl)-1-tert-butyldimeth-
ylsilyl-a-D-galactopyranoside 12
A dried mixture of 8.3 g (20mmol) of 1147 and 12.7 g
(25.8 mmol) of 5 was dissolved in 150mL of anhydrous
CH2Cl2 and the solution was transferred to a flask con-
˚
taining 5 g of flame-dried 4 A MS via cannula under
argon. After stirring for an hour at rt the reaction was
cooled to ꢀ45 ꢀC and 200 lL of TMSOTf was added
in several portions via syringe. After 30min at ꢀ45 ꢀC
the cold bath was removed and reaction was slowly
allowed to warm to rt. The reaction was quenched with
0.3 mL Et3N, diluted with CH2Cl2, filtered through Cel-
iteꢁ, and concentrated. Purification by FCC with 2:1

596
S. A. Svarovsky et al. / Tetrahedron: Asymmetry 16 (2005) 587–598
4.12. Na-Fluoren-9-ylmethoxycarbonyl-3-O-[2-azido-4,6-
O-benzylidene-2-deoxy-3-O-(2,3,4,6-tetra-O-acetyl-b-D-
galactopyranosyl)-a-D-galactopyranosyl]-L-threonine
phenacyl ester 16b
evaporated, and the residue was purified by FCC with
75:25:1 CHCl3/MeOH/AcOH to give 210mg (77%) of
the final product as a mixture of three amide rotamers
(NHAc, Fmoc-NH).
A mixture of 340mg (0.74 mmol) of 15b and 450mg
(3.22 mmol) of 14 was dissolved in minimum amount
of dry CH2Cl2, evaporated and dried under vacuum
for 1 h. The dried mixture was dissolved in 8 mL of
3:1 CH2Cl2/THF and added via cannula under argon
4.15. Solid-phase glycopeptide synthesis
The synthesis of peptide 18 was carried out manually
using a disposable plastic syringe fitted with a Teflon fil-
ter and connected to a vacuum waste bottle via a 2-way
Teflon valve. The synthesis was carried out starting from
Leucine-derivatized 2-chlorotrityl resin (Novabiochem,
0.1 mmol/g). Subsequent amino acids were synthesized
using Na-Fmoc amino acids (5 equiv) in DMF with
the addition of 4.9 equiv of HATU and HOAt as an
acylation catalyst. Each coupling reaction was con-
ducted for 1 h (monitoring with bromophenol blue).
The Fmoc group was removed by treatment with 20%
piperidine in DMF (3 and 10min). Resin was washed
with DMF between each coupling and deprotection
step. The glycosyl amino acid building block 17a
(2 equiv) was coupled for 2 · 6 h. The N-terminal amino
acid was coupled with the Boc-protected linker 19. The
completeness of all reactions was monitored using the
Kaiser test.
0
to 550mg of flame-dried molecular sieves (4 ). After stir-
ring for 20min 20 lL of TMSOTf was added at once.
After 30min the reaction was quenched with few drops
of Et3N, filtered through Celiteꢁ, and evaporated. The
residue was purified by FCC with 3:2 EtOAc/hexanes
to give 600 mg (95%) of a-isomer, 30mg (5%) of b-iso-
mer (a/b = 19:1) as glassy solids. Rf (1:1 EtOAc/hex-
anes) = 0.33; [a]D =+25.6 (c 1.00, CHCl3); IR (neat)
2232.20, 2115.53, 2022.96, 1750.08 cmꢀ1 1H NMR
;
(400 MHz, CDCl3) d = 7.30–7.94 (m, 18H, aromatic),
6.02 (d, 1H, J = 9.31 Hz, NH-Fmoc), 5.58 (s, 1H,
PhCH), 5.48 (d, 1H, J = 3.73 Hz, a-H10), 5.41 (m, 1H,
H400), 5.4–5.6 (m, 4H), 5.31 (dd, 1H, J1,2 = 7.88 Hz,
J2,3 = 10.34 Hz, H200), 5.04 (dd, 1H, J2,3 = 3.39 Hz,
J3,4 = 10.37 Hz, H300), 4.84 (d, 1H, J = 7.92 Hz, b-H100),
4.55 (m, 1H, NHCH), 4.50–4.70 (m, 2H), 4.41 (m, 1H,
H40), 4.05–4.40 (m, 11H, H), 3.94 (m, 1H, H20), 3.78
(m, 1H, H50), 2.17, 2.03, 2.02, 1.99 (s, 12H, Ac), 1.42
(d, 3H, J = 6.39 Hz, Me-Thr); 13C NMR (100 MHz,
CDCl3) d = 170.33, 170.14, 169.41, 141.33, 137.64,
134.11, 133.91, 128.97, 128.91, 128.17, 127.77, 127.12,
126.16, 125.19, 125.10, 120.02, 102.42, 100.69, 99.21,
75.95, 75.79, 71.08, 70.85, 69.16, 68.73, 67.32, 66.98,
66.80, 63.46, 61.34, 59.68, 58.60, 47.17, 20.72, 20.56,
19.27; FAB MS (m/z): 1065.1 (MH+), 1037 (MH+ꢀN2)
Anal. Calcd for C54H56N4O19: C 60.90; H 5.30; N
5.26. Found: C 60.59; H 5.28; N 5.23.
The cleavage of the peptide from the resin with simulta-
neous Boc and benzylidene removal was achieved by the
treatment with 50% TFA in dichloromethane for 2 h.
The resin was washed (5 · 2 mL) with the cleavage cock-
tail, concentrated to ca. 1 mL and precipitated from cold
ether. The precipitate was centrifuged, dissolved in 1:1
H2O/MeCN and purified by preparative HPLC (eluant
linear gradient 95:5 ! 20:80 H2O/MeCN over 60min,
tR = 34 min) to give 28% of peptide the N-terminal-
linked peptide as the free amine. MALDI-TOF-MS:
1712.17 for C78H1126N12O30 (MW = 1711.90). The tetra-
acetylated glycopeptide was treated with 50mM
NaOMe in MeOH for 30min. The reaction was
quenched by addition of glacial acetic acid, the mixture
evaporated, dissolved in a mixture of acetonitrile water,
and purified by preparative HPLC (eluant linear gradi-
ent 95:5 ! 20:80 H2O/MeCN over 60min, tR = 30
min) to give 93% of the free N-terminal-linked glycopep-
tide. MALDI-TOF: 1543.13 for C70H118N12O26 (MW
1543.75). A solution of 5 mg (3.2 lmol) of this material
2.5 mL of 200 mM phosphate buffer at pH 8.0 was
degassed by three pump-freeze-thaw cycles. To this
solution was added 750 lL (1.9 mg) of a solution of
Trautꢁs reagent 2-iminothiolane in pH 8.0buffer
(10mg/4 mL). After 1 h under argon the solution was
concentrated and purified by preparative HPLC to af-
ford monomeric thiol-modified peptide and the corre-
sponding disulfide (eluant linear gradient 95:5 ! 20:80
H2O/MeCN over 60min, tR = 31 min for monomer,
tR = 32 min for dimer). MALDI-TOF-MS for mono-
mer: 1646.05 for C74H126N13O26S+ (MW = 1645.93).
4.13. Na-Fluoren-9-ylmethoxycarbonyl-3-O-[2-acetam-
ido-4,6-O-benzylidene-2-deoxy-3-O-(2,3,4,6-tetra-O-ace-
tyl-b-D-galactopyranosyl)-a-D-galactopyranosyl]-L-serine
17a43
A solution of 1.92 g (1.82 mmol) of 16a in 50mL of a
6:3:1 THF/AcOH/Ac2O was treated with 15 g of Zn
powder. In 4 h the reaction was filtered through a sin-
tered glass funnel, concentrated, and purified by FCC
with 92:7:1 CHCl3/MeOH/AcOH to give 1.35 g (79%)
of 17a as a glassy solid.
4.14. Na-Fluoren-9-ylmethoxycarbonyl-3-O-[2-acetam-
ido-4,6-O-benzylidene-2-deoxy-3-O-(2,3,4,6-tetra-O-acet-
yl-b-D-galactopyranosyl)-a-D-galactopyranosyl]-L-threo-
nine 17b15
A solution of 300 mg (0.28 mmol) 16b in 10mL of 2:2:1
THF/AcOH/Ac2O was flushed with argon and treated
with 800 mg of Zn powder at room temperature. The
progress of the reaction was monitored by HPLC–MS.
In 1 h ratio of the product to the starting material was
1:2. In 4 h reaction was complete and HPLC showed
the desired product in 93% purity. The reaction solution
was filtered through the sintered glass, concentrated,
MALDI-TOF-MS
for
dimer:
3288.08
for
2+
C148H250N26O52S2 (MW = 3287.71). Preparation of
glycopeptide-coated gold particles 20 (Fig. 3B) was car-
ried out as per the general procedure. The solution
turned deep purple as with other syntheses, but after

S. A. Svarovsky et al. / Tetrahedron: Asymmetry 16 (2005) 587–598
597
15. Kiessling, L. L.; Pohl, N. L. Chem. Biol. 1996, 3, 71–77.
16. Roy, R.; Baek, M. G. Recog. Carbohydr. Biol. Syst. Pt A:
Gen. Proced. 2003, 362, 240–249.
1 h, centrifugation and attempted re-dissolution with
various solvents (H2O, MeOH, DMSO) failed.
17. Baek, M. G.; Roy, R. Bioorg. Med. Chem. 2002, 10, 11–
17.
4.16. Lectin affinity chromatography (LAC)
18. Rojas, T. C.; de la Fuente, J. M.; Barrientos, A. G.;
Penades, S.; Ponsonnet, L.; Fernandez, A. Adv. Mater.
2002, 14, 585–588.
19. Barrientos, A. G.; de la Fuente, J. M.; Rojas, T. C.;
Fernandez, A.; Penades, S. Chem.—A Eur. J. 2003, 9,
1909–1921.
20. Rojo, J.; Diaz, V.; de la Fuente, J. M.; Segura, I.;
Barrientos, A. G.; Riese, H. H.; Bernade, A.; Penades, S.
Chembiochem 2004, 5, 291–297.
21. de la Fuente, J. M.; Barrientos, A. G.; Rojas, T. C.; Rojo,
J.; Canada, J.; Fernandez, A.; Penades, S. Angew. Chem.,
Int. Ed. 2001, 40, 2258.
22. Lin, C. C.; Yeh, Y. C.; Yang, C. Y.; Chen, G. F.; Chen, Y.
C.; Wu, Y. C.; Chen, C. C. Chem. Commun. 2003, 2920–
2921.
23. Otsuka, H.; Akiyama, Y.; Nagasaki, Y.; Kataoka, K. J.
Am. Chem. Soc. 2001, 123, 8226–8230.
24. Nolting, B.; Yu, J. J.; Liu, G. Y.; Cho, S. J.; Kauzlarich,
S.; Gervay-Hague, J. Langmuir 2003, 19, 6465–6473.
25. Hone, D. C.; Haines, A. H.; Russell, D. A. Langmuir 2003,
19, 7141–7144.
26. Svarovsky, S. A.; Barchi, J. J. Carbohydr. Res. 2003, 338,
1925–1935.
Agarose-bound galactose-specific peanut agglutinin
(PNA) and mannose/glucose-specific Pisum sativum
agglutinin (PSA) were purchased from Vector Labs,
Burlingame, CA (www.vectorlabs.com). Equal amounts
(1 mL) of agarose-immobilized PSA and PNA were
loaded onto two separate columns and washed with 10
bed volumes of 1 · PBS at pH 7.4 to wash out the lec-
tin-stabilizing sugars. Solutions of the TF–gold particles
in water (50 lL) were loaded onto each column and fur-
ther washed with 150 lL of 1 · PBS. After 10min each
column was washed with 10bed volumes of 1 · PBS in
1 mL aliquots. Each aliquot was compared by UV–vis
spectroscopy to a solution of 50 lL of TF–gold particles
in 1 mL 1 · PBS. No particles were eluted from the PNA
column in any of the 10 · 1 mL aliquots while elution of
particles from the PSA column was complete in the first
1 mL aliquot. Further, washing of the PNA column with
several 1 mL aliquots of 200 mM galactose displaced the
particles bound to the PNA thus indicating reversibility
of the binding.
27. Ren, T.; Liu, D. X. Tetrahedron Lett. 1999, 40, 7621–
7625.
28. Corbell, J. B.; Lundquist, J. J.; Toone, E. J. Tetrahedron:
Asymmetry 2000, 11, 95–111.
Acknowledgements
We would like to thank Kunio Nagashima and Michelle
Gignac of the Electron Microscopy facility, NCI Freder-
ick for collecting the TEM data and James A. Kelley
and Cong Lai of the LMC for Mass spectra.
29. Caliceti, P.; Veronese, F. M. Adv. Drug Deliv. Rev. 2003,
55, 1261–1277.
30. Kochendoerfer, G. Expert Opin. Biol. Ther. 2003, 3, 1253–
1261.
31. Roberts, M. J.; Bentley, M. D.; Harris, J. M. Adv. Drug
Deliv. Rev. 2002, 54, 459–476.
32. Palegrosdemange, C.; Simon, E. S.; Prime, K. L.; White-
sides, G. M. J. Am. Chem. Soc. 1991, 113, 12–20.
33. Rittenhouse-Diakun, K.; Xia, Z.; Pickhardt, D.; Morey,
S.; Baek, M. G.; Roy, R. Hybridoma 1998, 17, 165–173.
34. Roy, R.; Tropper, F. D.; Romanowska, A.; Jain, R. K.;
Piskorz, C. E.; Matta, K. L. Bioorg. Med. Chem. Lett.
1992, 2, 911–914.
35. Gambert, U.; Thiem, J. Carbohydr. Res. 1997, 299, 85–89.
36. Arsequell, G.; Valencia, G. Tetrahedron: Asymmetry 1997,
8, 2839–2876.
References
1. Hakomori, S. I. Adv. Cancer Res. 1989, 52, 257–331.
2. Springer, G. F. J. Mol. Med. 1997, 75, 594–602.
3. Lloyd, K. O. Drug News Perspect. 2000, 13, 463–470.
4. Ragupathi, G.; Howard, L.; Cappello, S.; Koganty, R. R.;
Qiu, D. X.; Longenecker, B. M.; Reddish, M. A.; Lloyd,
K. O.; Livingston, P. O. Cancer Immunol. Immunother.
1999, 48, 1–8.
5. Kunz, H. J. Pept. Sci. 2003, 9, 563–573.
6. Allen, J. R.; Harris, C. R.; Danishefsky, S. J. J. Am. Chem.
Soc. 2001, 123, 1890–1897.
37. Baek, M. G.; Roy, R. Macromol. Biosci. 2001, 1, 305–311.
38. Baek, M. G.; Rittenhouse-Olson, K.; Roy, R. Chem.
Commun. 2001, 257–258.
7. Hilaire, P. M. S.; Cipolla, L.; Franco, A.; Tedebark, U.;
Tilly, D. A.; Meldal, M. J. Chem. Soc., Perkin Trans. 1
1999, 3559–3564.
8. Glinsky, V. V.; Glinsky, G. V.; Rittenhouse-Olson, K.;
Huflejt, M. E.; Glinskii, O. V.; Deutscher, S. L.; Quinn, T.
P. Cancer Res. 2001, 61, 4851–4857.
39. Grundler, G.; Schmidt, R. R. Liebigs Ann. Chem. 1984,
1826–1847.
40. Komba, S.; Meldal, M.; Werdelin, O.; Jensen, T.; Bock,
K. J. Chem. Soc., Perkin Trans. 1 1999, 415–419.
41. Schultz, M.; Kunz, H. Interf. Chem. Biochem. 1995, 73,
201–228.
9. Glinskii, O. V.; Turk, J. R.; Pienta, K. J.; Huxley, V. H.;
Glinsky, V. V. J. Physiol.–London 2004, 554, 89–99.
10. Glinsky, V. V.; Glinsky, G. V.; Glinskii, O. V.; Huxley, V.
H.; Turk, J. R.; Mossine, V. V.; Deutscher, S. L.; Pienta,
K. J.; Quinn, T. P. Cancer Res. 2003, 63, 3805–3811.
11. Khaldoyanidi, S. K.; Glinsky, V. V.; Sikora, L.; Glinskii,
A. B.; Mossine, V. V.; Quinn, T. P.; Glinsky, G. V.;
Sriramarao, P. J. Biol. Chem. 2003, 278, 4127–4134.
12. Glinsky, G. V. Cancer Metast. Rev. 1998, 17, 177–185.
13. Lundquist, J. J.; Toone, E. J. Chem. Rev. 2002, 102, 555–
578.
42. Jiaang, W. T.; Chang, M. Y.; Tseng, P. H.; Chen, S. T.
Tetrahedron Lett. 2000, 41, 3127–3130.
43. Qiu, D. X.; Gandhi, S. S.; Koganty, R. R. Tetrahedron
Lett. 1996, 37, 595–598.
44. Luning, B.; Norberg, T.; Tejbrant, J. Glycoconjugate J.
1989, 6, 5–19.
45. Rademann, J.; Schmidt, R. R. Carbohydr. Res. 1995, 269,
217–225.
46. Luning, B.; Norberg, T.; Tejbrant, J. J. Chem. Soc., Chem.
Commun. 1989, 1267–1268.
47. Grundler, G.; Schmidt, R. R. Carbohydr. Res. 1985, 135,
203–218.
14. Roy, R. Trends Glycosci. Glycotechnol. 2003, 15, 291–310.

598
S. A. Svarovsky et al. / Tetrahedron: Asymmetry 16 (2005) 587–598
48. Greilich, U.; Brescello, R.; Jung, K. H.; Schmidt, R. R.
Liebigs Ann. 1996, 663–672.
53. Tarentino, A. L.; Phelan, A. W.; Plummer, T. H.
Glycobiology 1993, 3, 279–285.
49. Murakata, C.; Ogawa, T. Carbohydr. Res. 1992, 234, 75–
91.
50. Pirrung, M. C.; Shuey, S. W.; Lever, D. C.; Fallon, L.
Bioorg. Med. Chem. Lett. 1994, 4, 1345–1346.
51. Xu, Y. F.; Gendler, S. J.; Franco, A. J. Exp. Med. 2004,
199, 707–716.
52. Trester-Zedlitz, M.; Kamada, K.; Burley, S. K.; Fenyo,
D.; Chait, B. T.; Muir, T. W. J. Am. Chem. Soc. 2003, 125,
2416–2425.
54. Hirsch, L. R.; Stafford, R. J.; Bankson, J. A.; Sershen, S.
R.; Rivera, B.; Price, R. E.; Hazle, J. D.; Halas, N. J.;
West, J. L. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 13549–
13554.
55. Paciotti, G. F.; Myer, L.; Weinreich, D.; Goia, D.; Pavel,
N.; McLaughlin, R. E.; Tamarkin, L. Drug Deliv. 2004,
11, 169–183.
56. Wang, Z. G.; Zhang, X. F.; Ito, Y.; Nakahara, Y.; Ogawa,
T. Carbohydr. Res. 1996, 295, 25–39.