Synthetic N -Acetyl- d -glucosamine based fully branched tetrasaccharide, a mimetic of the endogenous ligand for CD69, activates CD69+ killer lymphocytes upon dimerization via a hydrophilic flexible linker

Article abstract of DOI:10.1021/jm100055b

On the basis of the highly branched ovomucoid-type undecasaccharide that had been shown previously to be an endogenous ligand for CD69 leukocyte receptor, a systematic investigation of smaller oligosaccharide mimetics was performed based on linear and bra

Full text of DOI:10.1021/jm100055b

4050 J. Med. Chem. 2010, 53, 4050–4065
DOI: 10.1021/jm100055b
Synthetic N-Acetyl-D-glucosamine Based Fully Branched Tetrasaccharide, a Mimetic of the
Endogenous Ligand for CD69, Activates CD69þ Killer Lymphocytes upon Dimerization via
a Hydrophilic Flexible Linker
†
†
†
‡
ꢀ †
ꢁ
ꢀ ‡
ckova, Lukas Zı
ꢀꢁ
ꢀ §,
nkova, Martina Libigerova, Ljubina Ivanova,
ꢀꢁ ‡
´
r Sklenar,
ꢁ
ꢁ
Anna Kovalova, Miroslav Ledvina, David Saman, Daniel Zyka, Monika Kubı
´
´
dek, Vladimı
ꢀ §,
´
r Kren, and Karel Bezouska*
Petr Pompach,§, Daniel Kavan,§, Jan Bı
´
´
§,
ly, Ondrej Vanek, Zuzana Kubı
ꢀ §
ꢁ
ꢁ
ꢀ §,
§,
,§,
ꢀ
Maria Antolı
ꢀ §,
kova, Hynek Mrazek, Daniel Rozbesky, Katerina Hofbauerova, Vladimı
ꢀ
ꢀ §,
ꢁ
ꢀ
ꢁ
ꢁ
´
†Institute of Organic Chemistry and Biochemistry, V.V.I., Academy of Sciences of the Czech Republic, Flemingovo Naꢀm. 2, 16610 Prague,
Czech Republic, ‡National Centre for Biomolecular Research, Faculty of Science, Masaryk University, Kamenice 2, 61137 Brno,
Czech Republic, §Department of Biochemistry, Faculty of Science, Charles University Prague, Hlavova 8, 12840 Prague, Czech Republic, and
´
ꢁ ꢀ
Institute of Microbiology, V.V.I., Academy of Sciences of the Czech Republic, Vıdenska 1083, 14220 Prague, Czech Republic
Received January 14, 2010
On the basis of the highly branched ovomucoid-type undecasaccharide that had been shown previously to
be an endogenous ligand for CD69 leukocyte receptor, a systematic investigation of smaller oligosacchar-
ide mimetics was performed based on linear and branched N-acetyl-D-hexosamine homooligomers
prepared synthetically using hitherto unexplored reaction schemes. The systematic structure-activity
studies revealed the tetrasaccharide GlcNAcβ1-3(GlcNAcβ1-4)(GlcNAcβ1-6)GlcNAc (compound
52) and its R-benzyl derivative 49 as the best ligand for CD69 with IC50 as high as 10-9 M. This compound
thus approaches the affinity of the classical high-affinity neoglycoprotein ligand GlcNAc23BSA.
Compound 68, GlcNAc tetrasaccharide 52 dimerized through a hydrophilic flexible linker, turned out
to be effective in activating CD69þ lymphocytes. It also proved efficient in enhancing natural killing
invitro, decreasing the growth oftumorsinvivo, andactivatingthe CD69þ tumor infiltratinglymphocytes
examined ex vivo. This compound is thus a candidate for carbohydrate-based immunomodulators with
promising antitumor potential.
Introduction
cells is now much better understood because of the identifica-
tion of various surface NK receptors contributing to the
process of NK cell activation or inactivation.4-6 The activa-
tion and triggering of natural killing are under the control of a
complex signaling machinery to which many receptors, com-
ponents of the cell surface “receptor zipper”, are known to
contribute.7 In this study, we focused our attention on a group
of calcium-dependent animal lectins with binding affinity for
carbohydrate structures. The complex saccharide structures
are involved in many biologically important signal transduc-
tion processes, and thus, they play a key role in molecular
recognition events contributing to cell-cell, cell-bacteria,
and cell-virus interactions.8-10 The lectin receptors are able
to recognize oligosaccharide structures present on the surface
of tumor cells and initiate their lysis by cells of the immune
system.11,12 We are interested in two NK cell lectin activation
receptors, rat NKR-P1 and human CD69, unique for their
ability to distinguish between closely related carbohydrate
structures and to recognize the N-acetyl-D-hexosamines
(HexNAc) in both gluco and galacto configurations.13,14
Carbohydrates interact with these lectins over an extensive
surface area, but the structure and position of the oligosac-
charide binding sites are unique for each of the two receptors.
Rat NKR-P1 has a binding groove that accommodates the
linear oligosaccharides,15 whereas sugar-binding sites in hu-
man CD69 are at three separate locations, and thus bra-
nched carbohydrates seem to be preferred.16,17 Our previous
systematic study of the activating lectin receptor NKR-P1
In view of the recent increasing incidence of pathological
states characterized by secondary immune deficiencies in the
general population, considerable attention has been given to
the investigation of effective ways to modulate the activities of
the individual components of the immune system. Natural
killer (NKa) cells, which form 5-10% of mononuclear cells in
the blood, have many unique immunological activities and
thus provide an obvious target for immune activation. NK
cells play a crucial role in innate immunity, as they are
characterized by fast and strong cytolytic response against
tumor or virally infected cells. They also have the ability to
release cytokines and chemokines mediating inflammatory
responses and to influence hematopoiesis and the adaptive
immune response.1-3 Despite their role in immune defense
mechanisms, major questions regarding their therapeutical
potential remained unanswered.4,5 The ability of NK cells to
discriminate between normal and tumor or virally infected
*To whom correspondence should be addressed. Phone: þ420-2-
4106-2383. Fax: þ420-2-2195-1283. E-mail: bezouska@biomed.cas.cz.
aAbbreviations: 3N, GlcNAcβ1-3GlcNAc; 6N, GlcNAcβ1-6GlcNAc;
A, GalNAc; BN, GlcNAcβ1-3(GlcNAcβ1-4)GlcNAc; CB, GlcNAcβ1-
4GlcNAc; CT, GlcNAcβ1-4GlcNAcβ1-4GlcNAc; HexNAc, N-acetyl-D-
hexosamine; ION, ionomycin; LAC, lactose dimer; mAb, monoclonal
antibody; N, GlcNAc; NG, GlcNAc23BSA neoglycoprotein; NK, natu-
ral killer; OM, [GlcNAcβ1-2(GlcNAcβ1-4)(GlcNAcβ1-6)ManR1-6]-
[GlcNAcβ1-2(GlcNAcβ1-4)ManR13][GlcNAcβ1-4]-
Manβ1-4GlcNAcβ1-4GlcNAc.
r
pubs.acs.org/jmc
Published on Web 04/30/2010
2010 American Chemical Society

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 10 4051
Chart 1. Structures of the Synthesized HexNAc Based Oligosaccharides Used in the Study
demonstrated the binding hierarchy of HexNAc type oligo-
saccharides. The binding affinity increases in the group of
monosaccharides from GlcNAc and GalNAc to ManNAc
and in the order of chitooligomers ([-β-D-GlcNAc(1f4)]n)
with elongation of the oligosaccharide chain up to four
sugar units. In linear oligosaccharides of the N-acetylgluco-
samine type, the β(1f4) glycosidic bond is preferred over
β(1f6)- and β(1f3)-linked regioisomers of chitobiose ([-β-D-
GlcpNAc(1f4)]2) showing lower binding affinity.18,19
In the case of the human CD69 receptor, the physiological
ligands are not yet known. Using recombinant CD69 protein,
we have previously identified three separate binding sites for
GlcNAc in the monomeric unit of this receptor.20 Moreover,
we have shown thatthe complex pentaantennary bisecting un-
decasaccharide from egg white glycoprotein ovomucoid with
the structure [GlcNAcβ1-2(GlcNAcβ1-4)(GlcNAcβ1-6)-
ManR1-6][GlcNAcβ1-2(GlcNAcβ1-4)ManR13][GlcNA-
cβ1-4]Manβ1-4GlcNAcβ1-4GlcNAc (OM, Chart 1), is
one of the best ligands of natural origin identified so far.17
The affinity of this natural oligosaccharide for CD69 is in the
low nanomolar range, thus competing successfully with the
artificial high affinity ligand, GlcNAcBSA neoglycoprotein
(NG, Chart 1). This natural structure has thus provided an
important paradigm for the development of smaller synthetic
oligosaccharide mimetics that would be accessible for large
scale synthesis and further potential use for in vivo experi-
mental tumor therapies. The total chemical synthesis of
complex-type N-linked oligosaccharides identical to natural
ones has been achieved,21 but this involves a number of
chemical protection/deprotection steps and remains a time-
consuming and costly task. Additional approaches have been
developed to supply the glycobiology community with a
sufficient amount of pure complex oligosaccharides, but
these procedures have not yet been sufficiently adapted
for robust and inexpensive synthesis.9,18 Yet another possi-
bility is represented by de novo chemical syntheses of
branched homooligosaccharides, which might afford new
oligosaccharides rarely occurring in nature in pure form
and in sufficient amounts.22,23 Recently, we synthesized
the branched homotrisaccharide β-D-GlcNAc-(1f3)[β-D-
GlcNAc-(1f4)]-D-GlcNAc and found that its binding affi-
nity for NK cell lectin receptors competed successfully with
oligosaccharides of much greater complexity.19
Here we aimed to identify oligosaccharide ligands useful for
the modulation of activities of NK cells through their surface
receptors, NKR-P1 and CD69.11,17 We focused on assessing
the binding affinity of NKR-P1 receptor for linear homo-
oligosaccharides composed of β-linked HexNAc with several
types of glycosidic bonds. In the case of CD69, we tried to
select a small size oligosaccharide composed of highly
branched HexNAc that would mimic the natural oligosac-
charide from ovomucoid described above.17 For this purpose
we synthesized and biologically tested a group of model
oligosaccharides: (a) linear homooligosaccharides of N-acet-
ylgalactosamine type, i.e., [-β-D-GalNAc(1f4)]2 (21), [-β-D-
GalNAc(1f4)]3 (32), and [-β-D-GalNAc(1f3)]2 (37); (b)

4052 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 10
Kovalovaꢀ et al.
branched homooligosaccharides of N-acetyl-D-glucosamine
type, i.e., triantennary tetrasaccharideβ-D-GlcNAc-(1f3)-[β-
D-GlcNAc-(1f6)]-[β-D-GlcNAc-(1f4)]-D-GlcNAc (52) and
biantennary trisaccharides β-D-GlcNAc-(1f3)-[β-D-GlcNAc-
(1f6)]-D-GlcNAc (53) and β-D-GlcNAc-(1f4)-[β-D-GlcN-
Ac-(1f6)]-D-GlcNAc (54); and (c) biantennary homooligo-
saccharide of N-acetyl-D-galactosamine type, β-D-GalNAc-
(1f3)-[β-D-GalNAc(1f4)]-D-GalNAc (61). Dimerization of
oligosaccharide 52, which proved to be one of the best ligands
for human CD69 using well established chemistry used in
peptide and protein chemistry,24 provided a series of com-
pounds of which 68 proved to be an efficient activators of
the effector cells of the immune system when tested both in
vitro and in vivo. Thus, the ability of the dimerized oligosac-
charide 68 to efficiently activate cells of the immune system
parallels that found previously for the dimerized peptide
ligand for CD69.24
Scheme 1a
Results
We prepared three series of carbohydrate ligands for NK
cell receptors NKR-P1 and CD69 (Chart 1). The first series
consisted of linear GalNAc based oligosaccharides 37, 21, and
32 used together with commercially available (chitobiose, CB;
chitotriose, CT) and previously described (3N, ref 19) oligo-
saccharides in the GlcNAc (N) format. The second series
contained biantennary GalNAc based oligosaccharide 61
intended to provide an equivalent to the corresponding
GlcNAc based oligosaccharide BN described previously.19
The third, most important, series was prepared (based on the
results of biological tests with two previous series) only in the
GlcNAc (N) format and included the biantennary oligosac-
charides 53 and 54 constituting (together with BN) the
complete series of trisaccharide GlcNAc (N) isomers, as well
as the fully branched triantennary tetrasaccharide 52.
Chemistry. In the synthesis of the target oligosaccharides,
we were often confronted with sterically hindered systems,
especially in the case of highly branched structures and
oligosaccharide structures consisting of N-acetylgalactosa-
mine units. In general, the axially oriented C(4)-OH group
on a galactopyranose skeleton is the least reactive secondary
OH group.25 To overcome this fact, the phthalimide glyco-
sylation method was applied. This procedure is considered to
be one of the most efficient 1,2-trans-stereoselective glyco-
sylation processes for low-reactive secondary OH groups.26
Nevertheless, we recently observed reduced stereoselectivity
of this method when very low reactive aglycons were used.27
The parallel approach enabling simultaneous glycosyla-
tion of several hydroxyl groups of the glycosyl acceptor was
employed for the preparation of branched oligosaccha-
rides.19 In addition, a silver perchlorate promoted glycosyla-
tion procedure was applied using glycosyl bromides as
glycosyl donors in the presence of silver carbonate. This
was selected because of relatively good efficiency and stereo-
selectivity in the case of low reactive, axially oriented OH(4)
groups in comparison to other standard glycosylation
methods.27 Silver carbonate acts as a scavenger of perchloric
acid and was used instead of a commonly used organic base,
which can inhibit glycosylation reaction.28,29 Appropriate
glycosyl donors and glycosyl acceptors of galacto configura-
tion were obtained via inversion of the configuration at the
C(4) carbon of the corresponding synthons with gluco con-
figuration by nucleophilic displacement of the mesyl group
by sodium acetate.27
a Reagents and conditions: (a) MsCl, pyridine, room temp; (b)
NaOAc, DMSO, 130 °C; (c) MeONa, MeOH, room temp; (d) BnBr,
NaH, DMF, room temp; (e) (Ph3P)3RhCl, toluene-EtOH-H2O, reflux,
and then HCOOH, reflux.
Synthesis of Linear Oligosaccharides of the D-Galactosa-
mine Type. In contrast to the large number of studies devoted
to the synthesis of linear oligosaccharides with β(1f4)-
linked 2-amino-2-deoxy-D-glucopyranose units, efficient
and practical methods for the synthesis of analogous oligo-
saccharides with β(1f4)-linked 2-amino-2-deoxy-D-galacto-
pyranose units remain scarce.26 This unsatisfactory situation
was apparently due to problems with formation of a β(1f4)
glycosidic bond between two D-galactosamine units. Only
few reports on disaccharide syntheses of this type have been
published so far, and these used a glycosyl acceptor
having the azido group as the masked amino function at
the C(2) position and 2-deoxy-2-phthalimido-D-galactopyr-
anosyl bromide as the glycosyl donor.30,31
Our primary goal in the synthesis of the target β(1f4)- and
β(1f3)-linked linear oligosaccharides of the D-galactosa-
mine type (21, 32, and 37) was to prepare appropriate
glycosyl acceptors 7, 8, 10 (Scheme 1), and 13. Compound
7 was obtained from benzyl 2-acetamido-3,6-di-O-benzyl-2-
deoxy-4-O-methanesulfonyl-R-D-glucopyranoside32 (3) upon
treatment with anhydrous sodium acetate in dimethyl sulfoxide
at 130 °C to give the galacto-derivative 5, which afforded 7 by
ꢀ
Zemplen deacetylation. The same synthetic approach was
applied for the preparation of compound 8starting from benzyl
2-acetamido-3-O-allyl-6-O-benzyl-2-deoxy-R-D-glucopyrano-
side28 (2). O-Benzylation of 8 and subsequent deallylation of
compound 9 so obtained by catalytic isomerization of the
protecting allyl group to prop-1-enyl group using Wilkinson’s
catalyst (Ph3P)3RhCl, followed by acid hydrolysis, led to
glycosyl acceptor 10. Benzyl 3,6-di-O-benzyl-2-deoxy-2-ph-
thalimido-β-D-galactopyranoside (13) was prepared from ethyl
4-O-acetyl-3,6-di-O-benzyl-2-deoxy-2-phthalimido-1-thio-β-
D-galactopyranoside27 (11) in two steps using a procedure
described by Westerlind et al.33
The use of 3,6-di-O-benzylderivative 7 as a starting glyco-
syl acceptor in preparation of the target β(1f4)-linked
oligosaccharides 21 and 32 was not successful. Glycosylation

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 10 4053
Scheme 2a
ꢀ
pyridine to give 19. Zemplen O-deacetylation of 19 yielded
benzyl glycoside 20. Hydrogenolysis of benzyl groups over
Pd/C catalyst afforded the target unprotected β(1f4)-linked
disaccharide 21.
The problem of effectively synthesizing β(1f4)-linked
galactosamine type oligosaccharides was solved by employ-
ing compound 13 as the starting glycosyl acceptor (Scheme 3).
Compound 13, bearing a phthalimido group at position C(2)
instead of an acetamido group, was chosen to decrease the
negative influence of intermolecular H-bonds between the
OH group and the NH group of the amide function in the
glycosylation process.36-38 Glycosylation of 13 with glycosyl
bromide 22 under the above-described conditions gave
the required disaccharide 23 in a satisfactory yield of 48%.
Furthermore, 4-O-acetyl-1,5-anhydro-3,6-di-O-benzyl-2-deoxy-
2-phthalimido-D-lyxo-hex-1-enitol (24) and 4-O-acetyl-3,6-
di-O-benzyl-2-deoxy-2-phthalimido-β-D-galactopyranosyl-4-
O-acetyl-3,6-di-O-benzyl-2-deoxy-2-phthalimi-do-β-D-galac-
topyranoside (25) were isolated as the reaction side products,
i.e., the products of elimination and cross-reaction of a
glycosyl donor.39,40 Formation of glycal and symmetrical
disaccharide with the β head to head glycosidic bond similar
to that found in β,β-trehalose from glycosyl donors of
2-deoxy-2-phthalimido-D-glucopyranose type has been
described.39,40
ꢀ
Zemplen deacetylation of disaccharide 23 followed by
glycosylation of the obtained compound 26 with glycosyl
donor 15 afforded trisaccharide 27 in an overall yield of 51%.
Furthermore, the β,β-trehalose type product 28 was also
isolated.19 The alkali-labile protecting groups were removed
by treatment with hydrazine hydrate in boiling ethanol, and
the obtained crude amine 29 was acetylated by acetic anhy-
ꢀ
dride in pyridine to give compound 30. Zemplen deacetyla-
tion of 30 gave benzyl glycoside 31, and its benzyl groups
were subsequently hydrogenolysed over Pd/C catalyst to
give the target unprotected trisaccharide 32.
a Reagents and conditions: (a) Br2, CH2Cl2, 0 °C; (b) AgClO4,
Ag2CO3, CH2Cl2, -15 °C; (c) (Ph3P)3RhCl, toluene-EtOH-H2O, re-
flux, and then HCOOH, reflux; (d) BuNH2, MeOH, reflux; (e) Ac2O,
pyridine, room temp; (f) MeONa, MeOH, room temp; (g) H2, Pd/C,
AcOH-H2O, room temp.
For the preparation of β(1f3)-linked galactosamine dis-
accharide 37 (Scheme 4) a synthetic approach analogous to
the one used for the synthesis of linear oligosaccharides
containing β(1f4)-linked N-acetyl-D-galactosamine was
employed. Silver perchlorate in the presence of silver carbo-
nate promoted glycosylation of glycosyl acceptor 10 with
galactopyranosyl bromide 15 to provide β(1f3)-linked dis-
accharide 33 in a good yield of 70%. This was then depro-
tected to obtain the final disaccharide 37.
Synthesis of Branched Oligosaccharides of the D-Glucosa-
mine Type. Synthesis of the target triantennary tetrasacchar-
ide 52 and biantennary trisaccharides 53 and 54 of D-
glucosamine type was carried out starting from multiple
glycosylation of the glycosyl acceptor benzyl 2-acetamido-
2-deoxy-R-D-glucopyranoside (39) (Scheme 5). Glycosyla-
tion of 39 with 3,4,6-tri-O-acetyl-2-deoxy-2-phthalimido-β-
D-glucopyranosyl bromide28 (38), promoted with silver per-
chlorate in the presence of silver carbonate, afforded the
tetrasaccharide 40 (15%) and deletion trisaccharides 41 (8%)
and 42 (17%). The obtained oligosaccharides were converted
to unprotected oligosaccharides 52, 53, and 54 as described
above.
of 7 with 4-O-acetyl-3,6-di-O-benzyl-2-deoxy-2-phthalimido-
D-galactopyranosyl bromide27 (22) promoted by silver perchlo-
rate in the presence of silver carbonate gave a complex mixture
of reaction products. Any attempts to isolate the expected
disaccharide benzyl 4-O-acetyl-3,6-di-O-benzyl-2-deoxy-2-
phthalimido-β-D-galactopyranosyl-(1f4)-2-acetamido-3,6-di-
O-benzyl-2-deoxy-R-D-galactopyranoside failed.
The disaccharide 21 was successfully prepared when the
less sterically hindered 3-O-allyl derivative 8, as glycosyl
acceptor, and glycosyl bromide 15 protected with less bulky
and more electronegative O-acetyl groups as glycosyl donor
were used at the glycosylation step (Scheme 2). Their reac-
tion under the same conditions afforded the required
β(1f4)-linked disaccharide 16 in a yield of 37%. The glyco-
syl donor 15 described by Nilsson et al.34 was prepared by an
alternative way, i.e., from ethyl 3,4,6-tri-O-acetyl-2-deoxy-2-
phthalimido-1-thio-β-D-galactopyranoside35 (14) in reaction
with bromine.
Synthesis of Branched Trisaccharide of the D-Galactosa-
mine Type. Similar to the synthesis of branched oligosacchar-
ides containing D-glucosamine units, the approach based on
multiple glycosylation was applied as well for D-galactosa-
mine type oligosaccharides. The glycosyl acceptor benzyl 2-
acetamido-6-O-benzyl-2-deoxy-R-D-galactopyranoside (55),
Deallylation of the protected disaccharide 16 using the
aforementioned procedure gave compound 17. The alkali-
labile protecting groups were removed by treatment with
n-butylamine in boiling methanol, and the obtained amine 18
was peracetylated by reaction with acetic anhydride in

4054 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 10
Kovalovaꢀ et al.
Scheme 3a
a Reagents and conditions: (a) AgClO4, Ag2CO3, CH2Cl2, -15 °C; (b) MeONa, MeOH, room temp; (c) N2H4.H2O, EtOH, reflux; (d) Ac2O, pyridine,
room temp; (e) H2, Pd/C, AcOH - H2O, room temp.
obtained by deallylation of compound 8, was glycosylated by
3 molar excess of 3,4,6-tri-O-acetyl-2-deoxy-2-phthalimido-
β-D-galactopyranosyl bromide34 (15), employing promotion
by silver perchlorate in the presence of silver carbonate
(Scheme 6). This reaction afforded not only branched trisac-
charide 57 (9%) but also disaccharide 56 (13%) as an
undesired side product resulting from the lower reactivity
of the axial C(4)-OH group of the glycosyl acceptor. After
deprotection as above we obtained the target biantennary
disaccharide 61.
Biological Evaluation. In order to evaluate the binding
affinity of the synthesized oligosaccharides for the target NK
cell receptors, we started with the recombinant rat NKR-
P1A receptor shown previously13,18 to have high affinity
for GlcNAc monosaccharide and related compounds.
Microtiter plates were coated with high affinity ligand
GlcNAc23BSA neoglycoprotein, and individual compounds
were tested as inhibitors of binding of the soluble radiola-
beled receptor to these plates. Results shown in Figure 1
indicate that the synthesized compounds were mostly aver-
age or poor ligands compared to the GlcNAc control. Here,
in the linear GlcNAc/GalNAc series, the β1-4 linkage is
preferred to other (β1-3 or β1-6) linkages (Figure 1A).
Branching of the oligosaccharides resulted in significant
decrease in the inhibitory potencies independently of the
series used (Figure 1B and Figure 1C).
More interesting results were obtained when the synthe-
sized compounds were tested as inhibitors of binding of
another lymphocyte receptor, human CD69, to its high
affinity GlcNAc23BSA ligands.14,20 This receptor has been

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 10 4055
Scheme 4a
its anomeric configuration is fixed in the R-position through
the benzyl residue. The initial confirmation of high affinity
binding of 49 to recombinant human CD69 was provided by
NMR titration (Figure 2A). Evaluation of the binding curve
revealed a complete disappearance of the acetate signals due
to specific binding until the saturation point corresponding
to one bound molecule of the oligosaccharide per receptor
subunit, after which a linear increase of the free (unbound)
ligand was observed. Evaluation of the binding curve also
provided a quantitative estimation for the value of Kd in the
nanomolar range, which is in complete agreement with the
inhibition data. Two additional binding techniques were
3
employed. Direct binding assay using H-labeled tetrasac-
charide 52 allowed us to establish the details of the binding
parameters: there was one high affinity binding site per
receptor subunit (two sites per receptor dimer) with Kd =
3.2 ꢀ 10-9 M (Figure 2B). Moreover, the binding to the
dimeric receptor displayed the same cooperativity reported
previously for the high affinity binding site for GlcNAc
with the Hill coefficient approaching the maximal theoretical
value for a two-subunit protein (htheor = 2, hexpt. = 1.98).41
The third direct binding assay based on tryptophan fluores-
cence quenching was performed using both the dimeric
CD69 receptor42 and its monomeric form obtained by site-
directed mutagenesis.41 This arrangement allowed us to
confirm the dependence of both affinity and ligand binding
cooperativity on the dimeric arrangement of the receptor:
the Kd value dropped from 3.4 ꢀ 10-9 to 1.12 ꢀ 10-7 M, and
only binding according to the single-site model could be
observed for the monomeric protein (Hill’s coefficients were
1.94 and 1.05 for the dimeric and for the monomeric protein,
respectively; Figure 2C and D).
Dimerized 52 Efficiently Precipitates Soluble CD69 De-
pending on Linker Chemistry. High affinity binding of 52 to
CD69 shown by both inhibition and direct binding tests
indicated that this oligosaccharide might represent a suit-
able minimum mimetic of the complex physiological ligand
for the receptor. In order to activate CD69þ immune cells
through an engagement of this antigen, however, the ligand
mimetic must be present in a multivalent (at least bivalent)
form. To achieve efficient dimerization of the tetrasacchar-
ide 52, we have adopted the standard chemistry used for
peptide and protein cross-linking24 and elaborated it for
efficient activation and coupling of oligosaccharides acti-
vated as β-glycosylamines (Scheme 7).43 The entire proce-
dure involves a conversion of the reducing oligosaccharide
52 into the corresponding β-glycosylamine 62, followed by
reaction with thiophosgene to produce the reactive isothio-
cyanate 63. This compound is then coupled to a series of
linear aliphatic diamines to produce the dimeric tetrasac-
charides 64-67 having a linker of varying length defined by
two to eight methylene groups. The synthesized and purified
GlcNAc tetrasaccharide dimers were evaluated for their
abilities to precipitate soluble CD69 protein in a process that
resembles the cross-linking of the cellular form of the re-
ceptor. While the monomeric tetrasaccharide 52 was not
active in this test, the dimeric compounds 64-67 proved to
be positive (Figure 3A). The best activity was achieved for
compound 65 bearing the butyl aliphatic linker.
a Reagents and conditions: (a) AgClO4, Ag2CO3, CH2Cl2, -15 °C; (b)
N2H4.H2O, EtOH, reflux; (c) Ac2O, pyridine, room temp; (d) MeONa,
MeOH, room temp; (e) H2, Pd/C, AcOH-H2O, room temp.
shown previously16,20 to contain multiple binding sites
for GlcNAc in its molecule and thus to prefer branched N-
acetyl-D-hexosamine sequences to linear ones.17 Indeed, only
minor differences have been found in the linear GlcNAc/
GalNAc series compared to the GlcNAc monosaccharide
control (Figure 1D). However, a notable hierarchical in-
crease in inhibitory potencies has been found in the branched
GlcNAc/GalNAc series (Figure 1E). This increase has been
considerably more profound in the GlcNAc series compared
to the GalNAc series (Figure 1E).
On the basis of these findings, detailed structure-activity
studies were performed in the branched GlcNAc series. Here,
an even more developed hierarchy of the gradually increas-
ing affinities could be seen (Figure 1F). The trisaccharide
BN described previously19 has already attained 10 times
better inhibitory potency compared to the GlcNAc control.
The newly synthesized compounds 53 and 54 have reached
100 times and 1000 times greater inhibitory activities, re-
spectively. The small, fully branched tetrasaccharide 52 was
an even better inhibitor, achieving 10 000 times better in-
hibitory activity than the GlcNAc control. Its corresponding
R-benzyl derivative was even 10 times more efficient, attain-
ing the potency of the much more complex natural oligo-
saccharide OM, and it was a 100 000 times more potent
inhibitor compared to the N-acetyl-D-glucosamine mono-
saccharide (Figure 1F). Thus, the tetrasaccharide 52 has been
selected for further work as an efficient mimetic of a much
larger ovomucoid undecasaccharide (OM, Chart 1) de-
scribed previously as natural ligand,17 as well as of the
classical artificial high affinity ligand, GlcNAc neoglycopro-
tein (NG, Chart 1) .14
Direct Binding Assay for 52 and 49. The high affinity
interaction of 49 with the soluble recombinant lymphocyte
receptor CD69 was further confirmed using several techni-
ques allowing us to observe the binding directly. Compound
49 appeared to be particularly suitable for these studies, since
Recently published work has provided a clear indication
that not only the length but also the chemical nature of a
linker, such as its hydrophobicity and flexibility, may sig-
nificantly influence the outcome of an interaction of a
bivalent artificial ligand with a corresponding receptor.44

4056 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 10
Kovalovaꢀ et al.
Scheme 5a
a Reagents and conditions: (a) AgClO4, Ag2CO3, CH2Cl2, -15 °C; (b) N2H4 H2O, EtOH, reflux; (c) Ac2O, pyridine, room temp; (d) MeONa, MeOH,
3
room temp; (e) H2, Pd/C, AcOH-H2O, room temp.

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 10 4057
Scheme 6a
Figure 1. Biological testing of the synthesized HexNAc based oligo-
saccharides using inhibition assay. Indicated compounds were tested
as the inhibitors of binding of the radiolabeled rNKR-P1A (left) or
hCD69 (right) to the high affinity GlcNAc23BSA ligand. From the
completeinhibition curves, IC% values were calculated. Shown are the
mean ( SD from at least three independent experiments.
fully glycosylated form.46 Such verification is essential be-
cause the activities using the soluble and the cellular forms
of the immune receptors may differ significantly.17 We used
the standard cellular activation assays based on the produc-
tion of inositol phosphates and monitoring of intracellular
calcium concentrations.11 We isolated CD69low (<5% sur-
face expression) and CD69high (>30% surface expression)
lymphocytes from human peripheral blood mononuclear
cells. Both monomeric tetrasaccharide 52 and four dimeric
tetrasaccharides described above (compounds 65, 68, 69, and
70) were dissolved in PBS for testing as activators of these
cellular populations and were compared to the effect of PBS
only (negative control), or of specific monoclonal antibodies
against CD69 as well as the N-acetyllactosamine disacchar-
ide described previously41,45 to be a potent activator of
CD69þ lymphocytes (positive controls). All tested com-
pounds were added at 10-fold molar excess over the esti-
mated amount of surface CD69, as well as over the measured
Kd concentration. The inositol phosphate production results
show that very little activation of CD69low lymphocytes
occurred despite the presence of a small amount of CD69þ
cells in this population (Table 1). On the other hand, the
dimerized (although not the monomeric) tetrasaccharides
65-70 all had notable effects on the production of inositol
phosphates with the best compound (68) exerting effects
comparable to those of the positive controls, monoclonal
antibody, and N-acetyllactosamine disaccharide (Table 1).
The effect of the tested compounds on an increase in intra-
cellular calcium levels followed the effects on inositol phos-
phates production. In CD69low cellular population only the
a Reagents and conditions: (a) (Ph3P)3RhCl, toluene-EtOH-H2O,
reflux, and then HCOOH, reflux; (b) AgClO4, Ag2CO3, CH2Cl2,
-15 °C; (c) N2H4.H2O, EtOH, reflux; (d) Ac2O, pyridine, room temp;
(e) MeONa, MeOH, room temp; (f) H2, Pd/C, AcOH-H2O, room temp.
Therefore, we took the dimeric tetrasaccharide 65 as the lead
compound and prepared three additional similar com-
pounds, 68, 69, and 70, differing in the hydrophobicity and
flexibility of the linker based on the starting divalent amine
(Scheme 8) used. Compounds 65, 68, 69, and 70 thus all
had the optimal length of the four-carbon linker, albeit in
four different chemical variants: hydrophobic flexible, hy-
drophilic flexible, hydrophobic rigid, and hydrophilic rigid,
respectively. These compounds, after synthesis and purifica-
tion, were again tested in the receptor precipitation assay.
The results show that optimal activity was achieved for
compound 68 containing the hydrophilic flexible linker
(Figure 3B). This activity was comparable to that of the
previously described lactose-di-N-acetyl dimer.45 The origi-
nal lead compound 65 with the hydrophilic flexible linker
was also quite active, but compounds 69 and 70 with the rigid
linker were much less active.
Dimerized Tetrasaccharides 68 Activate Lymphocytes Ex-
pressing High Levels of CD69. We decided to verify our
conclusions concerning the optimal immunostimulating
compound further using CD69high lymphocytes containing
the cellular form of the receptor in its natural dimeric and

4058 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 10
Kovalovaꢀ et al.
Figure 2. Direct binding of compound 52 to human CD69 receptor: (A) NMR titration of 10 μM soluble CD69 with 49; (B) binding of
3H-labeled 52 to CD69 followed by equilibrium dialysis; (C, D) binding of 52 to the dimeric and monomeric form of CD69 followed by
tryptophan fluorescence quenching.
Scheme 7a
a Reagents and conditions: (a) sat. aq NH4HCO3, 30 °C, 7 days; (b) CSCl2, NaHCO3, acetone-H2O; (c) 2.4 equiv of 63 per 1 equiv of diamine,
CH2Cl2, room temp, 2 days.
addition of ionomycin resulted in a significant increase in
intracellular calcium (Figure 4A). In CD69high population, all
four dimeric tetrasaccharides were active, and the activity of
compound 68 was comparable with that of monoclonal
antibody positive control. Testing the compounds further at
10 times higher concentrations did not increase their activity
(data not shown).
Dimerized Tetrasaccharide 68 Increased the Killing of
Tumor Cell Lines in Vitro. To estimate the efficiency of the
synthesized compounds in enhancing the antitumor poten-
tial of the immune system, we have first tested their effects in
the standard short-term (4 h) cytotoxicity assays. Compound
68 and the lead compound 65 increased significantly the
killing of human erythroleukemic cell line K562, a standard
target cell line known to be sensitive for natural killing. The
effect of compound 65 was comparable to that of the two
monoclonal antibodies against CD69 used as positive con-
trols, while compound 68 had even much higher effect
enhancing the efficiency of natural killing about 4.5 times
under the given experimental conditions in vitro (Figure 5A).
Moreover, compound 68 was also active in the case of NK
resistant tumor cell line RAJI (Figure 5B) in the situation
where other compounds or monoclonal antibodies used as
positive controls had little effect. Under these experimental
conditions compound 68 enhanced natural killing about 3.5
times (Figure 5B).
Dimerized Tetrasaccharides 68 Suppresses the Growth of
Experimental Tumors and Activates Tumor Infiltrating Lym-
phocytes. An initial assessment of the antitumor efficacy
of the synthesized compounds was performed using the
experimental model of mouse B16 melanoma using a low
metastasis variant.47 In this assay, compound 68 was most
efficient decreasing the size of the tumors at day 26 and day
30 (Table 2). Compounds 65 and 70 also had some effects
seen at day 30 after the injection of tumor cells. Interestingly,
the two monoclonal antibodies against CD69 and the dimer-
ized N-acetyllactosamine had very little effect in this assay
(Table 2). In order to assess the possible effects of the
dimerized tetrasaccharides on the immune system, a large
number of immune system parameters must be monitored.
However, one of the most important parameter is the activity
of killer cells operating inside the tumors (especially tumor
infiltrating lymphocytes and dendritic cells). In order to
estimate this parameter, we performed cytotoxicity assay

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 10 4059
using tumor (melanoma) infiltrating lymphocytes isolated
from animals treated with the individual compounds and
assayed after isolation of these cells ex vivo. Notably, com-
pound 68 was the only compound among the tested set that
remained effective in this assay using both B16 melanoma
and NK resistant P815 mastocytoma (Figure 6A and
Figure 6B, respectively). Compound 65 also had a statisti-
cally significant effect but only in the case of the B16
melanoma (Figure 6A). We have initially tested the effect of
the injection of 1 μmol (approximately 2 mg) of the individual
compounds, a dose that was much smaller compared to the
tested antibodies that had to be injected in 10 mg amounts.
While increasing the amounts of the tested compounds had
very little benefit, injecting 0.1μmol amounts had verysimilar
effects, and 0.01 μmol amounts gave no effect.
not been identified, calcium and certain N-acetyl-D-hexosa-
mines have been shown previously to be specific ligands for
this receptor.14,17 Direct binding assays and molecular mod-
eling studies revealed the existence of three binding sites for
GlcNAc in the CD69 molecule, one of which represents the
high affinity binding site with an estimated value of Kd of
63 μM.20 Later, by use of a new construct for recombinant
Table 1. Activation of Purified CD69low and CD69high NK Cells Using
the Dimeric HexNAc Based Oligosaccharides
cell
control Mab LAC 52
65
68
65a
78b
69
70
CD69low
60a
82b
58a
76b
58a
77b
72a
75b
55a 70a
65b 71b
56a
72b
52a
75b
CD69high
1252a 1123a 58a 750a 1310a 520a 720a
1420b 1374b 60b 820b 1455b 610b 800b
a Amount of insP2 was measured 4 min after the addition of indivi-
dual compounds. bAmount of insP3 (cpm) was measured 2 min after the
addition of individual compounds.
Discussion
Although the physiological ligand for the widespread leuko-
cyte activation marker and triggering receptor CD69 has
Figure 4. Monitoring of intracellular calcium 3 min after the addi-
tion of tested compounds was performed in CD69low (A) and
CD69high (B) lymphocytes. ION is the ionomycin positive control.
All changes in (B) were statistically significant at p e 0.01 compared
to the PBS control (C). Values for the tested compounds and
ionomycin were taken from the complete curves at 3 and 11 min,
respectively.
Figure 3. Precipitation of soluble recombinant CD69 by equimolar
amounts of the tested ligand: (A) monomeric and dimerized 52 with
linker of different length; (B) monomeric and dimerized 52 with
linker having various chemical properties (hydrophobicity and
rigidity). LAC designates the lactose-di-N-acetyl dimer described
previously.45
Scheme 8a
a Reagents and conditions: (a) 1 equiv of 1,4-butanediamine-2,3-diol per 2.4 equiv of 63, CH2Cl2, room temp, 2 days; (b) 1 equiv of 1,4-
butanediamine-2-en per 2.4 equiv of 63, CH2Cl2, room temp, 2 days; (c) 1 equiv of 1,4-butanediamine-2,3-diol-2-en, CH2Cl2, room temp, 2 days.

4060 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 10
Kovalovaꢀ et al.
Figure 5. Natural killing assays in the presence of the tested
compounds using sensitive human cell line K562 (A) and resistant
human cell line RAJI (B). Statistical significance of changes com-
pared to PBS control is marked by asterisks as described in Experi-
mental Section.
Figure 6. Natural killing of tumor infiltrating lymphocytes isolated
from mice treated with the indicated compounds. Killing of B16
melanoma (A) and NK resistant mastocytoma P815 (B) targets is
shown. Statistical significance of changes compared to PBS control
is marked by asterisks as described in Experimental Section.
Table 2. Initial Assessment of the Antitumor Properties of the Synthe-
sized Compounds Using the Mouse B16S Melanoma Model
initial compound targeting killer lymphocytes including
natural killer cells to tumor sites and keeping the killer
lymphocytes active in the inhibiting microenvironment of
the tumors. Since this undecasaccharide turned out to be
available only in small amounts after complicated isolation
from biological material17 and since the total chemical synth-
esis of such complex oligosaccharide was shown to be very
complicated,21 considerable experimental efforts have been
aimed toward the synthesis and identification of smaller
oligosaccharide mimetics.
tumor size (cm2)
compd
day 26
day 30
PBS
MAb1
MAb2
LAC
65
0.6 ( 0.2
0.6 ( 0.1
0.4 ( 0.1
0.4 ( 0.1
0.5 ( 0.1
0.1 ( 0.1
0.6 ( 0.2
1.3 ( 0.3
1.1 ( 0.2
0.8 ( 0.2
0.7 ( 0.1
0.8 ( 0.2
0.2 ( 0.1
0.8 ( 0.2
68
70
Here we describe one approach toward the development
of efficient oligosaccharide mimetic for CD69 based on the
synthesis of N-acetyl-D-hexosamine-based homooligosaccha-
rides and their dimerization through several types of chemical
linkers in order to prepare bivalent compounds suitable for
CD69 cross-linking. The initial screening of GlcNAc and
GalNAc based homooligosaccharides, both linear and bran-
ched, identified the fully branched tetrasaccharide 52 as the
best ligand for CD69 with affinity in the nanomolar range,
thus approaching the affinity of the much more complex
ovomucoid derived undecasaccharide and that of the classical
high affinity ligand, GlcNAc23BSA neoglycoprotein. The R-
benzyl derivative of this compound was an even better
inhibitor, indicating the possibility of additional hydrophobic
interaction near the carbohydrate recognition site of CD69.18
This tetrasaccharide was then used as a lead for the develop-
ment of an efficient dimerization protocol and optimalization
of the linker used for dimerization (length, hydrophobicity,
rigidity). The previously developed assay measuring the
amount of soluble CD69 precipitated after interaction with
the bivalent ligand45 has been used to provide a simple test of
expression of CD69 optimized to have a high physical and
biochemical stability,42 direct binding assays including NMR
titrations, equilibrium dialysis, and fluorescence quenching
measurements revealed that binding of GlcNAc to the dimeric
CD69 protein proceeded in a cooperative fashion with Kd =
0.4 μM.41 However, after dissociation into the monomeric
subunits when no cooperativity in carbohydrate binding
could occur, the value of Kd dropped to about 16 μM using
the optimized monomeric protein.41 Recently, high affinity
physiological ligand for CD69 has been identified among the
highly branched ovomucoid type oligosaccharides.17 In parti-
cular, the pentabranched undecasaccharide having a structure
[GlcNAcβ1-2(GlcNAcβ1-4)(GlcNAcβ1-6)ManR1-6][Gl-
cNAcβ1-2(GlcNAcβ1-4)ManR1-3][GlcNAcβ1-4]Manβ1-
4GlcNAcβ1-4GlcNAc turned out to be an efficient ligand
for CD69 with Kd in the low nanomolar range. This ligand
may represent a target structure for CD69þ NK cells, since it
could be detected at the surface of NK sensitive targets cell
lines and could be induced by stress in NK resistant targets as
well.17 This undecasaccharide may thus represent a suitable

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 10 4061
binding efficiency of the dimerized GlcNAc tetrasaccharides.
From these tests, dimerized tetrasaccharide 68 bearing a
hydrophilic flexible linker emerged as the most promising
compound. Further biological evaluation revealed this com-
pound to be effective in cellular activation of CD69high NK
cells, which led tothe abilitytoenhancenatural killing invitro,
to decrease the rate of tumor growth in vivo, and to keep the
tumor infiltrating lymphocytes activated even in aggressive
tumor microenvironment, as revealed after ex vivo examina-
tion of their cytotoxicity. Moreover, this compound com-
pared favorably with other reagents cross-linking the CD69
target receptor; it was active in much smaller concentration
than the specific monoclonal antibodies against this receptor,
and most probably it would also be much less immunogenic.
The exact role of CD69 antigen in NK cell biology is not
fully understood. First, CD69 is expressed on many leukocyte
subsets48 and thus cannot be considered as an NK cell specific
receptor. Second, the work by Sanchez-Madrid and collea-
gues using CD69-/- mice provided a clear indication for the
negative roleof this antigen intumor cellkilling, since the mice
lacking surface expression of CD69 were more resistant to
tumors.49 On the other hand, the work of Moretta et al.
emphasizedthatmonoclonalantibodies (and perhaps ligands)
that are strongly bound to CD69 can activate CD69 positive
cells even in the absence of other (antigen dependent) pro-
liferation pathways.50 Our present work provides a strong
support for such a possibility and advocates the dimerized
tetrasaccharide 68 as a strong candidate for further systematic
experiments looking at the details of its efficacy in other
experimental tumor models and the details of the molecular
mechanisms of its action.
spectra were measured by the HSQC technique51 if necessary.
Positive-ion FAB mass spectra were acquired on a BEqG geo-
metry mass spectrometer ZAB-EQ (VG Analytical). Complex
mixtures were analyzed using LCQ mass spectrometric detector
(Finnigan) coupled to a HPLC system for LC/MS applications,
equipped with an ion-trap analyzer. þAPCI ionization was
used for recording spectra. TLC was carried out on Merck
aluminum sheets silica gel 60 F254, and column chromatography
was carried out on Fluka silica gel 60 (40-63 μm). Analytical RP
HPLC was performed using Waters Alliance HPLC system
(PDA 996 detector) equipped with a column (150 mm ꢀ 3.9 mm)
filled with Nova-Pak C18 (4 μm, Waters). Purity of all synthe-
sized compounds was determined to be g95% by RP HPLC
using a water-methanol gradient. Preparative RP HPLC was
performed with a Knauer system equipped with a column
(250 mm ꢀ 25 mm) filled with LiChrosorb RP-18 (5 μm, Merck).
2-Acetamido-2-deoxy-β-D-galactopyranosyl-(1f4)-2-acetami-
do-2-deoxy-D-galactopyranose (21). Hydrogenolysis of com-
pound 20 (50 mg, 0.08 mmol) according to general procedure D
afforded 25 mg (74%) of an R/β-anomeric mixture of compound
21. [R]D -98 (c 0.2, H2O). FAB MS calcd for C16H28N2O11 424.2,
found m/z 425.1 [M þ H]þ. For C16H28N2O11 (424.4): calcd
45.28% C, 6.65% H, 6.60% N; found 45.13% C, 6.76% H,
6.51% N. For NMR spectra see Supporting Information.
2-Acetamido-2-deoxy-β-D-galactopyranosyl-(1f4)-2-acetami-
do-2-deoxy-β-D-galactopyranosyl-(1f4)-2-acetamido-2-deoxy-D-
galactopyranose (32). Hydrogenolysis of compound 31 (85 mg,
0.08 mmol) according to general procedure D afforded 42 mg
(85%) of an R/β-anomeric mixture of compound 32. [R]D þ10
(c 0.3, H2O). MS EI calcd for C24H41N3O16 627.2, found m/z
650.5 [M þ Na]þ. For C24H41N3O16 (627.6): calcd 45.93% C,
6.58% H, 6.70% N; found 46.08% C, 6.53% H, 6.73% N. For
NMR spectra see Supporting Information.
2-Acetamido-2-deoxy-β-D-galactopyranosyl-(1f3)-2-acetami-
do-2-deoxy-D-galactopyranose (37). Hydrogenolysis of com-
pound 36 (100 mg, 0.14 mmol) according to general procedure
D afforded 30 mg (50%) of an R/β-anomeric mixture of com-
pound 37. [R]D þ56 (c 0.1, H2O). FAB MS calcd for C16H28N2O11
424.2, found m/z 425 [M þ H]þ. For C16H28N2O11 (424.4): calcd
45.28% C, 6.65% H, 6.60% N; found 45.13% C, 6.76% H, 6.48%
N. For NMR spectra see Supporting Information.
Conclusions
We prepared efficient mimetics of natural oligosaccharide
ligands for CD69, a widespread receptor triggering leukocyte
activation. Homooligomeric fully branched GlcNAc tetra-
saccharide 52 proved to be the most efficient ligand and could
be used as a lead for the development of effective activators of
CD69þ NK cells. Compound 68, GlcNAc tetrasaccharide
dimerized through a hydrophilic flexible linker, turned out to
be active in enhancing natural killing in vitro, decreasing the
growth of tumors in vivo, and increasing cytotoxic activity of
tumor infiltrating lymphocytes examined ex vivo. This com-
pound thus represents a strong candidate for an efficient
carbohydrate-based immunomodulator with a promising
antitumor potential.
Benzyl-2-acetamido-2-deoxy-β-D-glucopyranosyl-(1f3)-[2-ace-
tamido-2-deoxy-β-D-glucopyranosyl-(1f4)]-[2-acetamido-2-deoxy-
β-D-glucopyranosyl-(1f6)]-2-acetamido-2-deoxy-r-D-glucopyrano-
ꢀ
side (49). Zemplen O-deacetylationof compound46 (225 mg, 0.17
mmol) according to general procedure A followed by RP HPLC
in in water afforded 132 mg (83%) of compound 49. [R]D -41
(c 0.2, H2O). MS EI calcd for C39H60N4O21 920.4, found m/z
943.4 [M þ Na]þ. For C39H60N4O21 (920.9): calcd 50.86% C,
6.57% H, 6.08% N; found 50.98% C, 6.64% H, 5.96% N. For
NMR spectra see Supporting Information.
2-Acetamido-2-deoxy-β-D-glucopyranosyl-(1f3)-[2-acetamido-
2-deoxy-β-D-glucopyranosyl-(1f4)]-[2-acetamido-2-deoxy-β-D-glu-
copyranosyl-(1f6)]-2-acetamido-2-deoxy-D-glucopyranose (52).
Hydrogenolysis of compound 49 (25 mg, 0.03 mmol) according
to general procedure D afforded 17 mg (66%) of an R/β-ano-
meric mixture of compound 52. [R]D -25 (c 0.2, H2O). FAB MS
calcd for C32H54N4O21 830.3, found m/z 853.5 [M þ Na]þ. For
C32H54N4O21 (830.8): calcd 46.26% C, 6.55% H, 6.74% N;
found 46.09% C, 6.53% H, 6.78% N. For NMR spectra see
Supporting Information.
2-Acetamido-2-deoxy-β-D-glucopyranosyl-(1f3)-[2-acetamido-
2-deoxy-β-D-glucopyranosyl-(1f6)]-2-acetamido-2-deoxy-D-gluco-
pyranose (53). Hydrogenolysis of compound 50 (90 mg, 0.13
mmol) according to general procedure D afforded 2 mg (15%)
of an R/β-anomeric mixture of compound 53. [R]D -93 (c 0.1,
H2O). MS EI calcd for C24H41N3O16 627.2, found m/z 650.3
[M þ Na]þ. For C24H41N3O16 (627.6): calcd 45.93% C, 6.58% H,
6.70% N; found 45.89% C, 6.53% H, 6.78% N. For NMR spectra
see Supporting Information.
Experimental Section
General Chemistry. All reagents and solvents were purchased
from Sigma-Aldrich and used as received. Analytical samples
were dried at 6.5 Pa and 25 °C for 8 h. Melting points were
determined with a Kofler apparatus and are uncorrected.
Optical rotations were measured on Rudolph Research Analy-
tical AUTOPOL IV polarimeter at 20 °C. [R]D values are given
in deg cm3 g-1. Elemental analyses were performed using a
3
3
PerkinElmer 2400 series II CHNS/O elemental analyzer. IR
spectra were recorded on a Bruker Equinox 55 FTIR spectro-
meter, and wavenumbers are given in cm-1. NMR spectra were
recorded using Bruker Avance spectrometer at 500.1 MHz (1H)
and 125.8 MHz (13C) in CDCl3, using TMS as an internal
1
standard for H NMR spectra and CDCl3 as a standard for
13C NMR spectra. Chemical shifts are given in ppm (δ scale) and
coupling constants (J) in Hz. For unambiguous assignment of
signals in 13C NMR spectra, the heterocorrelated 2D NMR

4062 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 10
Kovalovaꢀ et al.
2-Acetamido-2-deoxy-β-D-glucopyranosyl-(1f4)-[2-acetamido-
2-deoxy-β-D-glucopyranosyl-(1f6)]-2-acetamido-2-deoxy-D-gluco-
pyranose (54). Hydrogenolysis of compound 51 (230 mg, 0.32
mmol) according to general procedure D afforded 133 mg (66%)
of an R/β-anomeric mixture of compound 54. [R]D -65 (c 0.2,
H2O). FAB MS calcd for C24H41N3O16 627.2, found m/z 628
[M þ H]þ. For C24H41N3O16 (627.6): calcd 45.93% C, 6.58% H,
6.70% N; found 45.79% C, 6.73% H, 6.59% N. For NMR spectra
see Supporting Information.
2-Acetamido-2-deoxy-β-D-galactopyranosyl-(1f3)-[2-acetamido-
2-deoxy-β-D-galactopyranosyl-(1f4)]-2-acetamido-2-deoxy-D-ga-
lactopyranose (61). Hydrogenolysis of compound 60 (28 mg,
0.32 mmol) according to general procedure D afforded 14 mg
(64%) of an R/β-anomeric mixture of compound 54. [R]D þ25.77
(c 0.3, H2O). FAB MS calcd for C24H41N3O16 627.2, found m/z
628 [M þ H]þ. For C24H41N3O16 (627.6): calcd 45.93% C, 6.58%
H, 6.70% N; found 45.89% C, 6.63% H, 6.78% N. For NMR
spectra see Supporting Information.
2-Acetamido-2-deoxy-β-D-glucopyranosyl-(1f3)-[2-acetamido-
2-deoxy-β-D-glucopyranosyl-(1f4)]-[2-acetamido-2-deoxy-β-D-glu-
copyranosyl-(1f6)]-2-acetamido-2-deoxy-D-glucopyranose Thiour-
ea Dimerized through Hydrophobic Linkers (64-70). Compound
52 (24 mg) was incubated in saturated aqueous NH4HCO3
(10 mL) at 30 °C for 7 days.52 The bulk of NH4HCO3 was
removed by lyophilization. Dowex 50W 1X2 Hþ (Fluka) was
added to a solution of the solid sugar amine in water until the pH
of the mixture dropped to 4. The resin with absorbed glycosyla-
mines was filtered off and washed with ice cold water. Pure
product was eluted with 2 M NH3 in MeOH, affording 20.6 mg
of the corresponding β-glycosylamine 62 (86%). This compound
was dissolved in 2 mL of water and added dropwise within 3 min
at room temperature to a stirred mixture of 400 μL of CSCl2 and
200 mg of NaHCO3 in 2 mL of acetone. After 45 min, water
(2 mL) was added and acetone was removed in vacuo at 30 °C.
Unreacted CSCl2 was extracted with chloroform, and the result-
ing solution of glycosyl isothiocyanate 63 was lyophilized and
immediately used in the subsequent coupling reaction. A solution
of 63 (2.4 equiv) in dichloromethane was slowly added to the
corresponding diamine (1 equiv) in dichloromethane. After 24 h
at room temperature the reaction mixture was purified by column
chromatography on silica gel (EtOAc-MeOH, 9:1). Thereafter,
the compounds were purified by chromatography on BioGel P2
(BioRad) as described previously.53 The yields of the individual
compounds were 2.1 mg (66%) for 64, 1.8 mg (62%) for 65, 2.0
mg (65%) for 66, 1.7 mg (58%) for 67, 1.5 mg (50%) for 68, 1.8
mg (60%) for 69, and 2.1 mg (67%) for 70. Analytical data for 64:
1H NMR 8.14 d (NH), 5.07 d (1H, J = 3.0, H-1), 4.82 d (1H, J =
8.3, H-10), 4.75 d (1H, J = 8.3, H-10), 4.70 d (1H, J = 8.4, H-100),
4.68 d (1H, J = 8.4, H-100), 4.68 d (1H, J = 8.4, H-1), 4.59 d (1H,
J = 8.5, H-1000), 4.58 d (1H, J = 8.5, H-1000), 4.26 m (1H, H-2),
4.06 m (2H, H-2, 3), 3.65-3.86 m (3H, H-2, 20, 200), 3.47-4.07 m
(18H, 30, 300, 3000, 40, 400, 4000, 50, 500, 5000, 60, 600, 6000), 2.07 m (CH2).
FAB MS calcd for C68H114N12O40S2 1802.3, found m/z 1825.3
[M þ Na]þ. For C68H114N12O40S2 (1802.3): calcd 45.27% C,
6.32% H, 9.32% N; 3.55% S; found 45.32% C, 6.43% H, 9.28%
N, 3.32% S. For NMR spectra see Supporting Information.
Analytical data for compounds 65-70 are given in the Support-
ing Information.
BSA (Sigma) in TBS þ C for 2 h at 4 °C and incubated with 125I-
NKR358 corresponding to half of the saturation amount
and the indicated dilutions of the tested compounds in a
total reaction volume of 100 μL. Plates were washed three times
with TBS þ C, drained, and dried, and 100 μL of a scintilla-
tion solution was added to each well. Radioactivity in wells
was determined in a β-counter Microbeta (Wallac). All experi-
ments were performed in duplicate. The results are average
values from duplicate experiments within the range indicated
by error bars.
NMR Titrations. All NMR experiments were run at 300 K in
a Bruker Avance 600 MHz spectrometer equipped with a
1
cryogenic H/C/N TCI probehead. H-15N HSQC spectra of
0.3 mM [15N]-labeled wild-type CD69 protein were used as a
routine check of protein folding and stability. The sample buffer
consisted of 10 mM MES, pH 5.8, with 49 mM NaCl, 1 mM
NaN3, and 10% D2O. During NMR titration, a 0.1 mM
solution of the unlabeled wild-type CD69 protein was titrated.
In an initial experiment, aliquots of the GlcNAc ligand corre-
sponding to 25%, 50%, 75%, 100%, 200%, and 500% of
saturation were added and signals of the free GlcNAc ligand
were observed at 2.2 ppm in the 1D proton spectra and used for
the estimation of the free ligand concentration. In a separate
experiment aimed at estimating the binding constant, smaller
ligand additions were used as equivalence was approached. The
protein was titrated to 75% of the estimated number of binding
sites, after which the amount of ligand was increased in incre-
ments of 5% of the estimated number of binding sites until
the equivalence point was reached. All spectra were processed
using the software NMRPIPE.54 The dissociation constant Kd,
defined as Kd = (cp - cL þ [L])[L]/(cL - [L]), was obtained by a
nonlinear fitting of the [L] vs cL titration curves (Figure 1A).
Volume changes during titration were accounted for.
Equilibrium Dialysis. Oligosaccharide 52 was labeled using
NaB3H4 (specific activity of 500 GBq/mmol) and was prepared
as described previously20 and diluted with the unlabeled com-
pound according to the required specific activity. To set up
equilibrium dialysis experiments, a rotating apparatus with
glass blocks containing separate sealable chambers with exter-
nal access was used as described previously (5). Then 200 μL
aliquots of 0.1 μM solutions of CD69 proteins in 10 mM MES,
pH 5.8, with 49 mM NaCl and 1 mM NaN3 were incubated with
varying amounts of ligand at 278 or 300 K for 48 h. After
equilibration, 100 μL aliquots were withdrawn from the control
and from the protein-containing chambers. The total ligand
concentration was determined by liquid scintillation, and the
bound ligand was calculated as the difference between the
amount of GlcNAc in the chamber containing the protein and
the control chamber. The results were calculated and plotted
according to Scatchard as described previously.11
Tryptophan Fluorescence Quenching. Tryptophan fluores-
cence quenching experiments were performed according to the
described methodology51 with minor modifications. In initial
experiments, 100 nmolar aliquots of CD69 protein were pipetted
into multiple wells of a UV Star plate (Greiner, Germany) and
mixed with 10-fold serial dilutions of the GlcNAc ligand.
Incubation proceeded for 1 h at room temperature, and then
the fluorescence of tryptophane residues was measured in
duplicate wells using the bottom fluorescence measurements
on a Safire2 plate reader (Tecan, Austria) with the following
settings: λex = 275 nm, λem = 350 nm, excitation and emission
slits were set to 5 and 20, respectively, and the fluorescence gain
was manually set to 66. When the lowest concentration of ligand
that still caused the quenching of tryptophane fluorescence was
found, detailed dilutions of the ligand by 10% saturation steps
were performed, and the concentrations of free and bound
ligand were calculated as described previously.55
Binding and Inhibition Experiments. Soluble dimeric rat
NKR-P1A and soluble dimeric human CD69 were expressed
in Escherichia coli, and purified essentially as described
previously.13,42 These proteins were radioiodinated as repor-
ted,13 with carrier-free Na125I (Amersham) to a specific activity
of 107 cpm per μg of protein. Binding and inhibition assays were
performed as described previously13 with minor modifications.
Briefly, 96-well poly(vinyl chloride) microplates (Titertek Im-
muno Assay-Plate, ICN Flow, Irvine, U.K.) were coated over-
night at 4 °C with 50 μL of GlcpNAc17BSA (10 μg/mL, Sigma) in
TBS þ C buffer (10 mM Tris-HCl, pH 8.0, with 150 mM NaCl, 1
mM CaCl2, and 1 mM NaN3). Plates were blocked with 1%
Precipitation Assays with the Soluble CD69 Receptor. Each
ligand was dissolved in water at 20 nM corresponding to 10-fold
the Kd value. The 125I-labeled protein (20 nM, 50 μL)13 was

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 10 4063
added to each sample (50 μL) in 96-well microtiter plates.
Mixtures were incubated at 4 °C for 30 min, and then a 20%
(v/v) solution of PEG 8000 was added (100 μL). The mixture was
left to precipitate for 1 h at 4 °C. After centrifugation (10 min,
4 °C, 1800gav), the supernatant was carefully removed and a
10% (v/v) solution of PEG 8000 (100 μL) was added. This
procedure was repeated three times to wash the precipitate.
After additional centrifugation and supernatant removal, the
precipitates were dried overnight at 37 °C.
10 min. At 10 min after the beginning of the experiment (8 min
after the addition of compounds), ionomycin (Sigma) was added
to a final concentration of 1 μM. Antibodies were added in
saturating concentrations (10 μg/mL).
Natural Killing. The standard 51Cr release test was performed
as described previously.11 Briefly, 104 chromium-labeled target
cells in 100 μL of complete RPMI 1640 were mixed in triplicate
with 150 nM solutions of the tested compounds, or control
compounds, in 50 μL of RPMI 1640 in round-bottomed 96-well
plates. Antibodies were added in saturating concentrations
(10 μg/mL). Thereafter, the appropriate amount of effector
cells (CD69high lymphocytes obtained by Ficoll-Isopaque se-
paration;12 see above) was added to 100 μL of complete RPMI
1640, and the plate was incubated at 37 °C for 3 h. Then 50 mL of
1% Triton X-100 was added into the maxima release wells,
and the incubation continued for another 1 h. Plates were cooled
on ice bath, and 100 μL of the supernatant was used for
radioactivity measurements. The percentage of specific lysis
was calculated using the formula % = [ (exp - spont)/(max -
spont)] ꢀ 100, where exp is the counts in experimental wells,
spont is the counts in wells containing medium instead of the
effector cells, and max is the counts in wells containing 1%
Triton X-100. Complete killing curves were constructed, from
which the lytic unit counts for individual experiments were
calculated. Lytic efficiency was defined as the inverse of the
lytic unit count.
Preparation of CD69low and CD69high Lymphocytes. Periphe-
ral blood mononuclear cells were obtained from standard blood
fraction enriched in leukocytes (buffy coats from the local blood
transfusion service) after dilution with RPMI1640 medium and
centrifugation over Ficoll-Paque. Cells were incubated over-
night in complete RPMI1640 in plastic cell culture dishes
to allow the adherent cells to attach. Collected nonadherent
fraction of PBMC (N-PBMC) contained mostly lymphocytes
(T, B, and NK cells). Lymphocytes from donors expressing less
than 5% of CD69 were designated CD69low. Lymphocytes from
donors with more then 20% CD69 positive cells were further
activated by incubation at a density of 2 ꢀ 106 cells/mL in
complete RPMI1640 medium for 4 h with PMA (50 ng/mL) and
ionomycin (500 ng/mL). This procedure increased the surface
expression of CD69 to 75-85%, as analyzed by flow cytometry
using monoclonal antibody against CD69 labeled with phyco-
erythrin. Such lymphocytes were designated as CD69high.
Inositol Phosphate Production. [3H]Inositol phosphates were
separated and quantified by the methods described previously.11
Incorporation of [3H]inositol into phospholipid was achieved by
incubating human CD69low or CD69high lymphocytes obtained
as described above (107 cells/ml) with 100 μL of [3H]inositol
(1.48 TBq/mol, 37 MBq/ml; GE Healthcare) for 3 h at 37 °C,
followed by extensive washing, and resuspension at 108 cells/mL.
An amount of 50 μL of this suspension containing 5 ꢀ 106 cells in
complete RPMI with 10 mM Hepes, pH 7.4, was mixed with
50 μL of the tested compounds (60 nM), and the mixture was
incubated at 37 °C for indicated times. Antibodies were added in
saturating concentrations (10 μg/mL). Reaction was stopped by
rapid transfer of the reaction mixture to 100 μL of 10%
trichloroacetic acid. Reaction was neutralized by the addition
of 50 μL of triethylamine, and 20 μL of 50% aqueous slurry of
Dowex 1X8, 100-200 mesh (Sigma), in formate form was
added. The supernatant was collected, and inositol bispho-
sphates and inositol trisphosphates were eluted by the addition
of 50 μL of 0.3 and 0.6 M ammonium formate, pH 7.0,
respectively. The eluant was dried in thin-walled 96-well plate,
and the radioactivity was counted in 100 μL of biodegradable
counting scintillant (GE Healthcare) using the Microbeta coun-
ter (Wallac).
Animal Tumor Therapies. Young (6-8 week old) female
C57BL/6 mice were purchased from Charles River (Montreal,
Quebec, Canada) and handled under the guidelines of Institute
Animal Care Protocol. Mice were shaved in the right flank
area and were given injections sc with 2.5 ꢀ 104 viable B16F1
cells (low metastasis variant, ref 56) in a final volume of 100 μL
of PBS. Ten days after the injection of tumor cells, 1 μmol of the
tested compound, or 10 mg of mAb, was injected. Tumor
growth was followed by Vernier caliper measurement every
other day from day 7 after injection. All of the experiments
included 10 mice/group. Tumor area was calculated according
to the formula A = (ab)/2, where a is the largest superficial
diameter and b is the smallest superficial diameter. The tumor
infiltrating lymphocytes were isolated from animals at day
30,45and their cytolytic activity was measured as described
above using the appropriate mouse tumor cells as targets.
Statistical Analyses. Statistical analyses were calculated by
Student’s t test. P values of e0.05 were considered as significant
(/, p e 0.05; //, p e 0.01).
ꢀ
ꢁ
Acknowledgment. The authors thank to Marketa Vancu-
rova for her helpful comments on the manuscript. This work
ꢀ
was supported by grants from Ministry of Education of
Czech Republic (Grants MSM_21620808, 1M0505, and
AVOZ50200510 to K.B., Grant Z4 055 0506 to M.L., and
Grants MSM0021622413 and LC06030 to V.S.), from Czech
Grant Agency (Grants 303/09/0477 and 305/09/H008 to K.B.
and Grant 203/09/P024 to P.P.), from Grant Agency of Czech
Academy of Science (Grants KAN200520703 and 200100801
to M.L.), and by EU Project Spine 2 (Contract LSHG-CT-
2006-02/220 to K.B.).
Monitoring of Intracellular Calcium. Human CD69low or
CD69high lymphocytes were loaded with the calcium-sensitive
fluor Indo-1 by incubating l07 cells/mL with 5 pM Indo-1AM
(Molecular Probes, Eugene, OR) in complete RPMI with 25 pM
2-ME at 37 °C. Cells were washed twice and resuspended at
5 ꢀ 106 cells/mL in medium. The fluorescence of the cell sus-
pension was monitored with a Safire2 spectrofluorimeter by
using an excitation wavelength of 349 nm and emission wave-
length of 410 nm. The setting of the instrument was calibrated
for each experiment by lysing the Indo-1-loaded cells with
Triton X-100 (0.07%) for maximum fluorescence. The mini-
mum fluorescence was determined after the addition of 10 mM
EGTA and sufficient Tris base to raise the pH to >8.3.
Intracellular calcium concentration was calculated using the
formula [Ca2þ] (nmoles) = 250[(F - Fmin)/(Fmax - F)] where F is
the measured fluorescence and 250 (nM) is the dissociation
constant of Indo-1. Signaling was measured in the absence of
extracellular calcium in medium containing 1 mM EGTA.
Tested compounds were added in 2 min by rapid mixing of
10 μL of 300 nM solutions of these compounds with 90 μL of cel-
lular suspensions, and monitoring continued for an additional
Supporting Information Available: Synthesis and analytical
data of compounds 4-10, 15-20, 23-31, 33-36, 40-48, 50, 51,
55-60 and analytical data for target compounds 21, 32, 37, 49,
52-54, 61, and 65-70. This material is available free of charge
via the Internet at http://pubs.acs.org.
References
(1) Biron, C. A.; Nguyen, K. B.; Pien, G. C.; Cousens, L. P.; Salazar-
Mather, T. P. Natural killer cells in antiviral defense: function
and regulation by innate cytokines. Annu. Rev. Immunol. 1999, 17,
189–220.

4064 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 10
Kovalovaꢀ et al.
ꢁ
ꢀ
ꢁ
(2) Moretta, A. The dialogue between human natural killer cells and
dendritic cells. Curr. Opin. Immunol. 2005, 17, 306–311.
(3) Trinchieri, G. Biology of natural killer cells. Adv. Immunol. 1989,
47, 187–376.
(4) Bottino, C.; Castriconi, R.; Moretta, L.; Moretta, A. Cellular
ligands of activating NK receptors. Trends Immunol. 2005, 26,
221–226.
(5) Moretta, A.; Bottino, C.; Vitale, M.; Pende, D.; Cantoni, C.;
Mingari, M. C.; Biassoni, R.; Moretta, L. Activating receptors
and coreceptors involved in human natural killer cell-mediated
cytolysis. Annu. Rev. Immunol. 2001, 19, 197–223.
(6) Lanier, L. L. NK cell recognition. Annu. Rev. Immunol. 2005, 23,
225–274.
(7) Moretta, L.; Bottino, C.; Pende, D.; Vitale, M.; Mingari, M. C.;
Moretta, A. Different checkpoints in human NK cell activation.
Trends Immunol. 2004, 25, 670–676.
(27) Vesely, J.; Ledvina, M.; Jindrich, J.; Trnka, T.; Saman, D. Synth-
esis of 2-amino-2-deoxy-β-D-galactopyranosyl(1 f 4)-2-amino-2-
deoxy-β-D-galactopyranosides: using various 2-deoxy-2-phthali-
mido-D-galactopyranosyl donors and acceptors. Collect. Czech.
Chem. Commun. 2004, 69, 1914–1938.
ꢁ
ꢁ
ꢁ
ꢁ
(28) Farkas, J.; Ledvina, M.; Brokes, J.; Jezek, J.; Zajıcek, J.; Zaoral, M.
´
The synthesis of O-(2-acetamido-2-deoxy-β-D-glucopyranosyl)-
(1f4)-N-acetylnormuramoyl-L-R-aminobutanoyl-D-isoglutamine.
Carbohydr. Res. 1987, 163, 63–72.
ꢁ
ꢁ
(29) Ledvina, M.; Zyka, D.; Jezek, J.; Trnka, T.; Saman, D. New
effective synthesis of (N-acetyl- and N-stearoyl-2-amino-2-deoxy-
β-D-glucopyranosyl)-(1f4)-muramoyl-L-2-aminobutanoyl-D-iso-
gluta-mine. Analogs of GMDP with immunopotentiating activity.
Collect. Czech. Chem. Commun. 1998, 63, 577–589.
(30) Wilstermann, M.; Kononov, L. O.; Nilsson, U.; Ray, A. K.;
Magnusson, G. Synthesis of ganglioside lactams corresponding
to G(M1)-ganglioside, G(M2)-ganglioside, G(M3)-ganglioside,
and G(M4)-ganglioside lactones. J. Am. Chem. Soc. 1995, 117,
4742–4754.
(31) Hansson, J.; Garegg, P. J.; Oscarson, S. Syntheses of anomerically
phosphodiester-linked oligomers of the repeating units of the
Haemophilus influenzae types c and f capsular polysaccharides.
J. Org. Chem. 2001, 66, 6234–6243.
(32) Berkin, A.; Szarek, W. A.; Kisilevsky, R. Synthesis of 4-deoxy-4-
fluoro analogues of 2-acetamido-2-deoxy-D-glucose and 2-aceta-
mido-2-deoxy-D-galactose and their effects on cellular glycosami-
noglycan biosynthesis. Carbohydr. Res. 2000, 326, 250–263.
(33) Westerlind, U.; Hagback, P.; Duk, M.; Norberg, T. Synthesis and
inhibitory activity of a di- and a trisaccharide corresponding to an
erythrocyte glycolipid responsible for the NOR polyagglutination.
Carbohydr. Res. 2002, 337, 1517–1522.
(8) Dwek, R. A. Glycobiology: toward understanding the function of
sugars. Chem. Rev. 1996, 96, 683–720.
(9) Seeberger, P. H. Exploring life’s sweet spot. Nature 2005, 437, 1239.
(10) Varki, A. Biological roles of oligosaccharides;all of the theories
are correct. Glycobiology 1993, 3, 97–130.
ꢁ
(11) Bezouska, K.; Yuen, C. T.; Obrien, J.; Childs, R. A.; Chai, W. G.;
ꢁ
ꢀ
Lawson, A. M.; Drbal, K.; Fiserova, A.; Pospısil, M.; Feizi, T.
Oligosaccharide ligands for Nkr-P1 protein activate Nk Cells and
ꢁ
´
cytotoxicity. Nature 1994, 372, 150–157.
ꢁ ꢀ ꢁ ꢀ ꢀ
(12) Pospısil, M.; Vannucci, L.; Horvath, O.; Fiserova, A.; Krausova,
´
ꢁ
K.; Bezouska, K.; Mosca, F. Cancer immunomodulation by
carbohydrate ligands for the lymphocyte receptor NKR-P1. Int.
J. Oncol. 2000, 16, 267–276.
ꢁ
ꢀ
ꢀ
ꢁ
ꢀ
(13) Bezouska, K.; Vlahas, G.; Horvath, O.; Jinochova, G.; Fiserova,
ꢁ
A.; Giorda, R.; Chambers, W. H.; Feizi, T.; Pospısil, M. Rat
´
(34) Nilsson, U.; Ray, A. K.; Magnusson, G. Efficient syntheses of 3,4,6-
tri-O-acetyl-2-deoxy-2-phthalimido-β-D-galactopyranosyl and 3,4,6-
tri-O-acetyl-2-deoxy-2-phthalimido-R-D-galactopyranosyl chloride.
Carbohydr. Res. 1990, 208, 260–263.
natural killer cell antigen, NKR-P1, related to C-type animal
lectins is a carbohydrate-binding protein. J. Biol. Chem. 1994,
269, 16945–16952.
ꢁ
ꢀ
ꢁ
(14) Bezouska, K.; Nepovı
´
m, A.; Horvath, O.; Pospısil, M.; Hamann,
´
ꢁ
ꢀ
ꢁ
(35) Vesely, J.; Ledvina, M.; Jindrich, J.; Saman, D.; Trnka, T. Im-
proved synthesis of 1,2-trans-acetates and 1,2-trans ethyl 1-thio-
glycosides derived from 3,4,6-tri-O-acetyl-2-deoxy-2-phthalimido-
D-hexopyranosides. Collect. Czech. Chem. Commun. 2003, 68,
1264–1274.
J.; Feizi, T. CD69 antigen of human lymphocytes is a calcium-
dependent carbohydrate-binding protein. Biochem. Biophys. Res.
Commun. 1995, 208, 68–74.
ꢀ
ꢁ
ꢀ
(15) Plı
´
hal, O.; Byrtusova, P.; Pavlı
´
cek, J.; Mihok, L.; Ettrich, R.; Man,
ꢁꢀ ꢀ ꢁ
cek, V.; Husakova, L.; Bezouska, K. The
ꢁ
P.; Pompach, P.; Havlı
´
(36) Crich, D.; Dudkin, V. Why are the hydroxy groups of partially
protected N-acetylglucosamine derivatives such poor glycosyl ac-
ceptors, and what can be done about it? A comparative study of the
reactivity of N-acetyl-, N-phthalimido-, and 2-azido-2-deoxy-glu-
cosamine derivatives in glycosylation. 2-Picolinyl ethers as reactiv-
ity-enhancing replacements for benzyl ethers. J. Am. Chem. Soc.
2001, 123, 6819–6825.
isoforms of rat natural killer cell receptor NKR-P1 display a
distinct binding of complex saccharide ligands. Collect. Czech.
Chem. Commun. 2004, 69, 631–644.
ꢁ ꢀ ꢁ
(16) Pavlıcek, J.; Kavan, D.; Pompach, P.; Novak, P.; Luksan, O.;
´
ꢁ
Bezouska, K. Lymphocyte activation receptors: new structural
paradigms in group V of C-type animal lectins. Biochem. Soc.
Trans. 2004, 32, 1124–1126.
(37) Fowler, P.; Bernet, B.; Vasella, A. A 1H-NMR spectroscopic
investigation of the conformation of the acetamido group in some
derivatives of N-acetyl-D-allosamine and -D-glucosamine. Helv.
Chim. Acta 1996, 79, 269–287.
ꢁ
(17) Bezouska, K. Carbohydrate and non-carbohydrate ligands for the
C-type lectin-like receptors of natural killer cells. A review. Collect.
Czech. Chem. Commun. 2004, 69, 535–563.
ꢀ
ꢀ
ꢀ
(18) Krist, P.; Herkommerova-Rejnochova, E.; Rauvolfova, J.;
Semenuk, T.; Vavruskova, P.; Pavlıcek, J.; Bezouska, K.; Petrus,
(38) Vasella, A.; Witzig, C. Glycosylidene carbenes. Part 22. R-D-
Selectivity in the glycosidation by carbenes derived from 2-acet-
amido-hexoses. Helv. Chim. Acta 1995, 78, 1971–1982.
(39) Kantoci, D.; Keglevic, D.; Derome, A. E. A convenient synthetic
route to the disaccharide repeating-unit of peptidoglycan. Carbo-
hydr. Res. 1987, 162, 227–235.
ꢁ
ꢁ
ꢀ
ꢁ ꢁ
´
ꢁ
L.; Kren, V. Toward an optimal oligosaccharide ligand for rat
natural killer cell activation receptor NKR-P1. Biochem. Biophys.
Res. Commun. 2001, 287, 11–20.
ꢁ
ꢀ
ꢁ
(19) Rohlenova, A.; Ledvina, M.; Saman, D.; Bezouska, K. Synthesis of
linear and branched regioisomeric chitooligosaccharides as poten-
tial mimetics of natural oligosaccharide ligands of natural killer
cells NKR-P1 and CD69 lectin receptors. Collect. Czech. Chem.
Commun. 2004, 69, 1781–1804.
(40) Maloisel, J. L.; Vasella, A. Synthesis of N-acetylallosamine-derived
disaccharides. Helv. Chim. Acta 1992, 75, 1491–1514.
(41) Kavan, D.; Kubıckova, M.:; Bıly, J.; Vanek, O.; Hofbauerova, K.;
´ ´
ꢀ
ꢁ ꢀ ꢁ ꢀ
ꢁ
ꢀ
ꢀ ꢀ ꢁ
´
Mrazek, H.; Rozbesky D.; Bojarova, P.; Kren, V.; Zıdek, L.;
ꢁ
ꢀ
´
(20) Pavlıcek, J.; Sopko, B.; Ettrich, R.; Kopecky, V., Jr.; Baumruk, V.;
ꢀꢁ
ꢁ
ꢁ
ꢀ
ꢀ
ꢁ
Sklenar, V.; Bezouska, K. Cooperation between subunits is essen-
tial for high affinity binding of N-acetyl-hexosamines to dimeric
soluble and dimeric cellular forms of human CD69. Biochemistry
[Online early access]. DOI: 10.1021/bi100181a, Published Online:
April 6, 2010.
Man, P.; Havlı
Pospı
´
cek, V.; Vrbacky, M.; Martı
´
nkova, L.; Kren, V.;
ꢁ
´
sil, M.; Bezouska, K. Molecular characterization of binding
ꢁ
of calcium and carbohydrates by an early activation antigen of
lymphocytes CD69. Biochemistry 2003, 42, 9295–9306.
(21) Eller, S.; Schuberth, R.; Gundel, G.; Seifert, J.; Unverzagt, C.
Synthesis of pentaantennary N-glycans with bisecting GlcNAc and
core fucose. Angew. Chem., Int. Ed. 2007, 46, 4173–4175.
(22) Bongat, A. F. G.; Demchenko, A. V. Recent trends in the synthesis
of O-glycosides of 2-amino-2-deoxysugars. Carbohydr. Res. 2007,
342, 374–406.
ꢁ
ꢀ
ꢀ
ꢁ
ꢀ
(42) Vanek, O.; Nalezkova, M.; Kavan, D.; Borovickova, I.; Pompach,
ꢀ
ꢁ
ꢀ
P.; Novak, P.; Kumar, V.; Vannucci, L.; Hudecek, J.; Hofbauerova,
ꢀ
ꢀ
ꢁ
K.; Kopecky, V., Jr.; Brynda, J.; Kolenko, P.; Dohnalek, J.;
ꢁ
ꢀꢁ
´
k, L.; Zıdek, L.; Sklenar, V.; Bezouska, K.
ꢁꢀ
ꢁ
Kaderavek, P.; Gorcı
´
Soluble recombinant CD69 receptors optimized to have an excep-
tional physical and chemical stability display prolonged circulation
and remain intact in the blood of mice. FEBSJ. 2008, 275, 5589–5606.
(23) Holeman, A.; Seeberger, P. H. Carbohydrate diversity: synthesis of
glycoconjugates and complex carbohydrates. Curr. Opin. Biotech-
nol. 2004, 15, 615–622.
ꢁ
ꢁ ꢁ
(43) Semenuk, T.; Krist, P.; Pavlıcek, J.; Bezouska, K.; Kuzma, M.;
´
ꢁ
ꢀ
ꢀ
ꢁ
ꢀ
ꢀ
ꢁ
(24) Bezouska, K.; Kavan, D.; Mrazek, H.; Manglova, D.; Basova, P.;
Novak, P.; Kren, V. Synthesis of chitooligomer-based glycoconju-
gates and their binding to rat natural killer cell activation receptor
NKR-P1. Glycoconjugate J. 2001, 18, 817–826.
ꢁ
Kren, V. Glycomic analysis of tumor cells reveals the molecular
determinants critical for their sensitivity for natural killing and
apoptosis. FEBS J. 2008, 275 (Suppl. 1), 81.
(44) Chung, S.; Parker, J. B.; Bianchet, M.; Amzel, L. M.; Stivers, J. T.
Impact of linker strain and flexibility in the design of a fragment-
based inhibitor. Nat. Chem. Biol. 2009, 6, 407–413.
(25) Paulsen, H. Advances in selective chemical synthesis of complex
oligosacchrides. Angew. Chem., Int. Ed. Engl. 1982, 21, 155–173.
(26) Banoub, J.; Boullanger, P.; Lafont, D. Synthesis of oligosacchar-
ides of 2-amino-2-deoxy sugars. Chem. Rev. 1992, 92, 1167–1195.
ꢀ
ꢁ
ꢀ
(45) Bojarova, P.; Krenek, K.; Wetjen, K.; Adamiak, K.; Pelantova, H.;
ꢁ
ꢁ
Bezouska, K.; Elling, L.; Kren, V. Synthesis of LacdiNAc-termi-

Article
nated glycoconjugates by mutant glycosyltransferases. A way to
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 10 4065
cytes expressing T cell receptor R/β. J. Exp. Med. 1991, 174, 1393–
1398.
(51) Gronberg, G.; Nilsson, U.; Bock, K.; Magnusson, G. Nuclear-
magnetic-resonance and conformational investigations of the pen-
tasaccharide of the forssman antigen and overlapping disaccharide,
trisaccharide, and tetrasaccharide sequences. Carbohydr. Res.
1994, 257, 35–54.
(52) Likhosherstov, L. M.; Novikova, O. S.; Derevitskaya, V. A.;
Kochetkov, N. K. A new simple synthesis of amino sugar β-D-
glycosylamines. Carbohydr. Res. 1986, 146, C1–C5.
(53) Lindhorst, T. K.; Kieburg, C. Glycocoating of oligovalent amines.
Synthesis of thiourea-bridged cluster glycosides from glycosyli-
sothiocyanates. Angew. Chem., Int. Ed. Engl. 1996, 35, 1953–
1956.
(54) Delaglio, F.; Grzesiek, S.; Vuister, G. W.; Zhu, G.; Pfeifer, J.; Bax,
A. NMRPipe: a multidimensional spectral processing system based
on Unix pipes. J. Biomol. NMR 1995, 6, 277–293.
(55) Chipman, D. M.; Grisaro, V.; Sharon, N. The binding of oligo-
saccharides containing N-acetylglucosamine and N-acetylmuramic
acid to lysozyme. J. Biol. Chem. 1967, 242, 4388–4394.
(56) Poste, G.; Doll, J.; Hart, I. R.; Fidler, I. J. In vitro selection of
murine B16 melanoma variants with enhanced tissue-invasive
properties. Cancer Res. 1980, 40, 1636–1644.
new glycodrugs and materials. Glycobiology 2009, 19, 509–517.
(46) Gerosa, F.; Tommasi, M.; Scardoni, M.; Accolla, R. S.; Pozzan, T.;
Libonati, M.; Tridente, G.; Carra, G. Structural analysis of the
CD69 early activation antigen by two monoclonal antibodies
directed to different epitopes. Mol. Immunol. 1991, 28, 159–168.
ꢁ
ꢀ
(47) Vannucci, L.; Fiserova, A.; Sadalapure, K.; Lindhorst, T. K.;
ꢀ
Kuldova, M.; Rossman, P.; Horvath, O.; Kren, V.; Krist, P.;
Bezouska, K.; Luptovcova, M.; Mosca, F.; Pospısil, M. Effect of
ꢀ
ꢁ
ꢁ
ꢀ
ꢁ
´
N-acetylglucosamine-coated glycodendrimers as biological modu-
lators in the B16F10 melanoma model in vivo. Int. J. Oncol. 2003,
23, 285–296.
(48) Testi, R.; D’Ambrosio, D.; De Maria, R.; Santoni, A. The CD69
receptor: a multipurpose cell surface trigger for hematopoietic cells.
Immunol. Today 1994, 15, 479–483.
(49) Sancho, D.; Gomez, M.; Sanchez-Madrid, F. CD69 is an immu-
noregulatory molecule induced following activation. Trends Im-
munol. 2004, 26, 136–140.
(50) Moretta, A.; Poggi, A.; Pende, D.; Tripodi, G.; Orengo, A. M.;
Pella, N.; Augugliaro, R.; Bottino, C.; Ciccone, E.; Moretta, L.
CD69-mediated pathway of lymphocyte activation: anti-CD69
monoclonal antibodies trigger the cytolytic activity of different
lymphoid effector cells with the exception of cytolytic T lympho-