Selectively Deoxyfluorinated N-Acetyllactosamine Analogues as 19F NMR Probes to Study Carbohydrate-Galectin Interactions

Article abstract of DOI:10.1002/chem.202101752

Galectins are widely expressed galactose-binding lectins implied, for example, in immune regulation, metastatic spreading, and pathogen recognition. N-Acetyllactosamine (Galβ1-4GlcNAc, LacNAc) and its oligomeric or glycosylated forms are natural ligands of galectins. To probe substrate specificity and binding mode of galectins, we synthesized a complete series of six mono-deoxyfluorinated analogues of LacNAc, in which each hydroxyl has been selectively replaced by fluorine while the anomeric position has been protected as methyl β-glycoside. Initial evaluation of their binding to human galectin-1 and -3 by ELISA and ¹⁹F NMR T₂-filter revealed that deoxyfluorination at C3, C4′ and C6′ completely abolished binding to galectin-1 but very weak binding to galectin-3 was still detectable. Moreover, deoxyfluorination of C2′ caused an approximately 8-fold increase in the binding affinity towards galectin-1, whereas binding to galectin-3 was essentially not affected. Lipophilicity measurement revealed that deoxyfluorination at the Gal moiety affects log P very differently compared to deoxyfluorination at the GlcNAc moiety.

Full text of DOI:10.1002/chem.202101752

Full Paper
doi.org/10.1002/chem.202101752
Chemistry—A European Journal
Selectively Deoxyfluorinated N-Acetyllactosamine
Analogues as 19F NMR Probes to Study Carbohydrate-
Galectin Interactions
Martin Kurfiřt,[a, b] Martin Dračínský,[c] Lucie Červenková Šťastná,[a] Petra Cuřínová,[a]
Vojtěch Hamala,[a, b] Michaela Hovorková,[d] Pavla Bojarová,[d] and Jindřich Karban*[a]
Abstract: Galectins are widely expressed galactose-binding
lectins implied, for example, in immune regulation, metastatic
spreading, and pathogen recognition. N-Acetyllactosamine
(Galβ1-4GlcNAc, LacNAc) and its oligomeric or glycosylated
forms are natural ligands of galectins. To probe substrate
specificity and binding mode of galectins, we synthesized a
complete series of six mono-deoxyfluorinated analogues of
LacNAc, in which each hydroxyl has been selectively replaced
by fluorine while the anomeric position has been protected
as methyl β-glycoside. Initial evaluation of their binding to
human galectin-1 and -3 by ELISA and 19F NMR T2-filter
revealed that deoxyfluorination at C3, C4’ and C6’ completely
abolished binding to galectin-1 but very weak binding to
galectin-3 was still detectable. Moreover, deoxyfluorination of
C2’ caused an approximately 8-fold increase in the binding
affinity towards galectin-1, whereas binding to galectin-3 was
essentially not affected. Lipophilicity measurement revealed
that deoxyfluorination at the Gal moiety affects log P very
differently compared to deoxyfluorination at the GlcNAc
moiety.
Introduction
ral ligands also include lactose, N-acetyllactosamine type I (also
termed lacto-N-biose, Galβ1-3GlcNAc), and other glycans con-
taining a galactose moiety at the non-reducing end. Immune
cells, including activated T and B cells, dendritic cells, macro-
phages or neutrophils, express galectins and use them to fine-
tune immune and inflammatory responses.[3] Moreover, various
types of cancer are associated with altered, usually elevated,
expression of galectins in affected tissues,[4] where individual
galectins play complex and complementary roles in neoplastic
transformations, tumor-cell survival, angiogenesis, metastasis
and tumor immune escape.[5] Therefore, galectins have become
valid targets for therapeutic intervention against cancer or
inflammation-related diseases.[4] Despite significant progress in
design and preparation of selective galectin inhibitors,[6] which
resulted in the discovery of promising inhibitors such as TD139
(introduced by Galecto Inc.),[7] development of molecules
showing high affinity and selectivity to particular galectins
remains a challenging task.[4]
The fluorination of carbohydrate ligands is an effective
approach to study protein-carbohydrate interactions as clearly
demonstrated in the last two decades.[8] Fluorine efficiently
mimics a hydroxyl group, and has highly advantageous proper-
ties as an NMR probe that permits to employ T2-filter,[9]
STDreF[10] or recently developed 2D STD-TOCSYreF[11] experi-
ments. These methods enable straightforward epitope-mapping
with an unprecedented accuracy and provide precise structural
information inaccessible by other techniques.[12] For example,
Lin and coworkers utilized 19F NMR to identify novel interactions
within the binding subsite E of galectins between arginine
residues Arg73GalÀ 1, Arg186GalÀ 3 and arene groups of TD139,
which helped to explain the origin of its selectivity to galectin-3
over galectin-1 and -7.[13] Oscarson et al. published the synthesis
Galectins (Gals) are a family of small β-galactoside binding
proteins found both extracellularly and intracellularly.[1] Galec-
tins bind carbohydrate ligands with
a highly conserved
carbohydrate-recognition domain (CRD) consisting of approx-
imately 130 amino-acid residues.[2] Despite similarities in their
CRDs, 15 mammalian galectins with subtle variations in
functions and substrate specificities were identified, including
biomedically highly significant human galectin-1 (Gal-1) and
human galectin-3 (Gal-3). Galectin extracellular functions de-
pend on their ability to specifically recognize and cross-link
membrane-bound N- or O-glycans containing type II mono- or
poly-N-acetyllactosamine (LacNAc, Galβ1-4GlcNAc). Their natu-
[a] M. Kurfiřt, Dr. L. Červenková Šťastná, Dr. P. Cuřínová, V. Hamala,
Dr. J. Karban
Department of Bioorganic Compounds and Nanocomposites
Institute of Chemical Process Fundamentals of the Czech Academy of
Sciences
Rozvojová 135, 16502 Prague 6 (Czech Republic)
E-mail: karban@icpf.cas.cz
[b] M. Kurfiřt, V. Hamala
University of Chemistry and Technology Prague
Technická 5, 16628 Prague 6 (Czech Republic)
[c] Dr. M. Dračínský
Institute of Organic Chemistry and Biochemistry of the Czech Academy of
Sciences
Flemingovo náměstí 542/2, 160 00 Prague 6 (Czech Republic)
[d] M. Hovorková, Dr. P. Bojarová
Institute of Microbiology of the Czech Academy of Sciences
Vídeňská 1083, 142 20 Prague 4 (Czech Republic)
Supporting information for this article is available on the WWW under
https://doi.org/10.1002/chem.202101752
Chem. Eur. J. 2021, 27, 1–13
1
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Paper
doi.org/10.1002/chem.202101752
Chemistry—A European Journal
of LacNAc oligomers labeled with 19F in the CF3CONH function-
ality and intended for recognition studies of human galectins
by 19F NMR methods.[14] Mono-deoxyfluorinated carbohydrate
ligands usually have diminished or comparable affinity com-
pared to natural ligands in non-covalent recognition by lectins
because the fluorinated position can no longer act as a
hydrogen bond donor. However, instances of improved affinity
have been reported. For example, immobilized 3’-sialylated
LacNAc deoxyfluorinated at C2’ of the Gal moiety (NeuAcα2-
3(2F)Galβ1-4GlcNAc) showed an about 4-fold higher affinity to
Toxoplasma gondii adhesive protein TgMIC1 than natural
sialylated LacNAc.[15] Very recently, Gibson et al. reported on the
chemoenzymatic synthesis of a series of deoxyfluorinated lacto-
N-biose analogues and their immobilization on PHEA (poly
(hydroxyethyl acrylamide)) coated gold nanoparticles (Fig-
ure 1).[16] The analogue 2 deoxyfluorinated at the 3-position of
the galacto moiety showed an enhanced affinity to Gal-3 and
Gal-7 in comparison with the non-fluorinated ligand 1, whereas
the analogue 3, deoxyfluorinated at the 6-position of the
galacto moiety displayed abrogated binding to both Gal-3 and
Gal-7. Importantly, multiply fluorinated LacNAc analogue 4
showed reversal of the selectivity in comparison with the non-
fluorinated standard 1, strongly favoring binding to Gal-7 over
Gal-3, which highlighted deoxyfluorination as a promising
strategy in the development of selective galectin inhibitors.
An important aspect in the development of efficient
fluorinated glycomimetics is the impact of regio- and stereo-
whereas the lipophilicity of deoxyfluorinated disaccharides or
amino sugars has not been systematically investigated.
As the aforementioned studies demonstrated the utility of
fluorinated carbohydrates as probes or potent selective lectin
ligands, we decided to prepare a complete series of mono-
deoxyfluorinated LacNAc analogues 5–10 (Figure 2) for a de-
tailed chemical and NMR epitope mapping of the galectin
binding site, and for the assessment of the impact of
deoxyfluorination on the lipophilicity of this important ligand.
Systematic deoxyfluorination of all possible LacNAc hydroxyls
should increase the likelihood of identifying a more potent or
selective inhibitor among the fluorinated ligands. β-Methyl
glycosides were chosen for the protection of the anomeric
1
position to simplify H and 19F NMR spectra and to maintain
high solubility in water, permitting application of highly
accurate liquid-state epitope-mapping techniques such as 2D
STD-TOCSYreF.[11] These techniques should elucidate the struc-
tural basis of galectin ligand specificity, and may explain why
different galectins display subtle variations in ligand selectivity.
This information may facilitate the development of novel
galectin inhibitors. Here we report on the synthesis of
deoxyfluorinated LacNAc analogues 5–10 by chemical glycosy-
lation, the determination of their lipophilicity and initial
evaluation of their affinity to Gal-1 and Gal-3 by ELISA and 19F
NMR T2-filter methods.
selective deoxyfluorination on their lipophilicity. Controlling Results
lipophilicity is critical in the drug development and carbohy-
drates have been traditionally regarded as compounds with a
poor therapeutic potential due to their very low lipophilicity,
rapid metabolic degradation, and low binding affinities. Deoxy-
fluorination can improve pharmacological profile of carbohy-
drate-based drug candidates including lipophilicity and meta-
bolic stability.[17] Recent studies on mono- and multifluorinated
monosaccharides demonstrated how an increase in lipophilicity
by deoxyfluorination depends on the fluorine position and the
relative stereochemistry of all stereocenters.[18] However, these
studies have so far been restricted to simple monosaccharides
The envisaged synthetic routes are summarized in Scheme 1.
The central criteria for the development of our synthetic
approach were (1) smooth and reliable glycosylation to obtain
protected precursors of the target LacNAc analogues, and (2)
good accessibility of fluorinated monosaccharide building
blocks. Therefore, we opted for deoxyfluorinated O-benzylated
methyl 2-azido-β-d-glucopyranosides as fluorinated glycosyl
acceptors because of their excellent compatibility with glyco-
sylation conditions.[14,19] Consequently, glycosylation should
Figure 1. Selected structures of deoxyfluorinated lacto-N-biose analogues
conjugated to PHEA coated gold nanoparticles and the binding affinities of
the resulting multivalent systems to Gal-3 and Gal-7 expressed as Kd.[16]
PHEA=poly(hydroxyethyl acrylamide). ND=observed affinity was too low to
be determined.
Figure 2. Target mono-deoxyfluorinated LacNAc (Galβ1-4GlcNAc) analogues
5–10 prepared in this work.
Chem. Eur. J. 2021, 27, 1–13
www.chemeurj.org
2
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Paper
doi.org/10.1002/chem.202101752
Chemistry—A European Journal
expected that the optimization of their glycosidation will be
necessary to achieve sufficient yield and stereoselectivity.
Therefore, we decided to prepare readily available known 2-
fluoro-trichloroacetimidates 25[26] and 26[27] as well as 2-fluoro-
thioglycosides 27 and 28 to explore their synthetic utility in
chemical glycosylation. We expected that thioglycosides 27 and
28 could be easily obtained from known 2-fluoro-bromide 29[28]
(Scheme 1C) by phase-transfer glycosylation and protecting
group manipulation.
Synthesis of carbohydrate acceptors
Our synthetic endeavor started by the preparation of key
carbohydrate acceptors 11, 13 and 30 (Schemes 1B and 2).
Non-fluorinated acceptor 30 was prepared from known methyl
2-phthalimido-glucoside 15[19,22,29] by the reaction sequence
comprising O-benzylation to 31, regioselective opening of the
benzylidene acetal,[22] deprotection of the phthalimide group
and diazo transfer,[19,30] using modifications of the published
procedures to improve reproducibility and purity of the
products (see the Supporting Information for details). Inter-
mediates 31 and 15 were conveniently used in the synthesis of
6-fluoro and 3-fluoro acceptors 11 and 13, respectively. To
obtain 11, derivative 31 was first converted to compound 12 by
acidic hydrolysis,[31] deprotection of the phthalimide group and
diazo transfer[30] in a very good overall 87% yield. Microwave-
assisted fluorination using
a
modified Hoffman-Röder
protocol,[20] (see the Supporting Information) regioselectively
produced 6-fluoro acceptor 11 in a high 87% yield with no
traces of the product of C4 deoxyfluorination.
Scheme 1. Outline of the synthesis of the target LacNAc analogues 5–10 and
fluorinated monosaccharide building blocks. PG=protecting group.
The synthesis of 3-fluoro acceptor 13 relied on the double
inversion at C3 reported by Huang[21a] and Dube[21b] for related
2-amino monosaccharides. Compound 15 was subjected to
phthalimide deprotection and subsequent diazo transfer to
provide derivative 14 in 93% yield. Next, 14 was converted to
the corresponding O3 triflate, which was without character-
ization treated with tetrabutylammonium acetate (TBAA) in
DMF to furnish 3-O-acetyl-d-allo-derivative 32 in an excellent
96% yield. Zemplén deacetylation[32] of 32 provided 3-hydroxy
derivative 33, whose allo-configuration was confirmed by the
yield protected fluorinated methyl 2-azido-β-d-lactosides
(LacN3β1-OMe) as precursors of fluorinated LacNAc analogues
(Scheme 1A). The key reaction to obtain 6-fluoro acceptor 11
should be regioselective O6-deoxyfluorination of diol 12 in
analogy to the published procedures,[20] whereas the synthesis
of 3-fluoro acceptor 13 will rely on the fluorination with double
inversion at C3 of alcohol 14 (Scheme 1B).[21] Precursors 12 and
14 could be prepared by protecting group manipulation and
diazo transfer from the known common intermediate 15[22] that
in turn is readily available from d-glucosamine. We envisaged
acetylated deoxyfluorinated thiogalactosides 16, 17 and 18 as
readily accessible and powerful glycosyl donors, which could be
prepared from known fluorinated galactosyl bromides 19,[20,23b]
20,[23] and 21[23,24] by phase transfer galactosylation of thiophe-
nol (Scheme 1C).[23b] They should be stable under protecting
group manipulations, if these are necessary, and should enable
reliable stereocontrol by stereodirecting participation of the C2
acetate. We devised a new synthesis of known bromide 21[24]
from in-house dianhydro pyranose 22[25] by the reaction
sequence including a regioselective opening of 2,3-epoxide 23
(providing 3-fluoro derivative 24) as the key step (Scheme 1C).
As 2-fluoro donors do not offer stereocontrol of glycosylation
by anchimeric assistance effect of the C2 protecting group, we
3
magnitude of the indirect vicinal coupling constants JH2-H3
3
(2.9 Hz) and JH3-H4 (2.5 Hz) corresponding to the expected
gauche relationship between protons H2-H3 and H3-H4,
respectively. An attempt to perform conversion of 14 to 33 by
treatment of the corresponding O3 triflate with KNO2, (Latrel-
Dax inversion[21b,33]) resulted in the formation of 33 along with a
chromatographically inseparable allo-configured 3-nitro by-
product 34 (33/34=2.9:1). Repeated recrystallization of the
mixture furnished 34 in 7% yield for characterization. Mother
liquors were acetylated and purified by chromatography,
providing 32 in 42% yield. The treatment of allo-derivative 33
with DAST afforded 3-fluoro-glucopyranoside 35 in 41% yield
accompanied by the mixture of elimination products 36 and 37
in 5% yield and 3-chloro-glucopyranoside 38 in 10% yield.
Eliminations are relatively frequent side reactions during
nucleophilic fluorination.[34] Chlorinated product 38 originated
Chem. Eur. J. 2021, 27, 1–13
www.chemeurj.org
3
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Paper
doi.org/10.1002/chem.202101752
Chemistry—A European Journal
from acidic work-up with HCl and its formation can be
prevented by avoiding HCl. The magnitude of indirect vicinal
3
3
3
coupling constants JH2-H3 (9.3 Hz), JH3-H4 (9.3 Hz) as well as JH2-F
(14.1 Hz) and 3JH4-F (11.4 Hz) in 3-fluoro compound 35 confirmed
the gluco-configuration and the fully stereospecific course of
deoxyfluorination at C3 with inversion of configuration. Regio-
selective reductive benzylidene opening of 35 with NaBH3(CN)/
HCl produced the desired 3-fluoro acceptor 13 in 64% yield.
Synthesis of deoxyfluorinated methyl 2-azido-lactosides
Deoxyfluorinated protected methyl 2-azido-lactosides (LacN3)
were prepared by glycosylation of acceptors 11, 13 and 30 with
galactoside thiodonors. Methyl glycosides of 6-fluoro-LacN3 39
and 3-fluoro-LacN3 40 were prepared by glycosylation of
acceptors 11 and 13 with readily accessible phenyl 2,3,4,6-tetra-
O-acetyl-1-thio-β-d-galactopyranoside 41[35] in 93% and 68%
yield, respectively (Scheme 3). Protected methyl glycoside of 6’-
fluoro-LacN3 42 was prepared in 75% yield by glycosylation of
30 with the thioglycoside 16 using NIS/TfOH activation[36]
(Scheme 3). Phenyl β-thioglycoside 16 was prepared by phase-
transfer glycosylation of thiophenol with 6-fluoro-α-galactosyl
bromide 19[23b,20] available from diacetone-d-galactose in three
steps (Scheme 3). The anomeric configuration of thioglycoside
16 was identified as β from the magnitude of the scalar vicinal
Scheme 3. Synthesis of O-protected 6-, 3-, 6’- and 4’-monofluorinated methyl
2-azido-lactosides. a) 41 (1.2 equiv), 11 (1.0 equiv), NIS (1.9 equiv), TfOH
°
(0.48 equiv), DCM, MS 3 Å, 0 C, 2 h; b) 41 (1.3 equiv), 13 (1.0 equiv), NIS
°
(1.9 equiv), TfOH (0.30 equiv), DCM, MS 3 Å, 0 C, 2 h; c) TBAHS (1.0 equiv),
PhSH (3.0 equiv), EtOAc, Na2CO3 (1 M aqueous solution), rt, 15 h; d) 30
(1.0 equiv), 16 (1.1 equiv), NIS (2.2 equiv), TfOH (0.24 equiv), DCM, MS 3 Å,
°
°
0 C, 2 h; e) Ac2O, FeCl3 (1.6 equiv), 80 C, 4 h; f) HBr/AcOH 33% (9.7 equiv),
°
DCM, 0 C-rt, 2 h; g) 30 (1.0 equiv), 17 (1.2 equiv), NIS (2.5 equiv), TfOH
°
(0.45 equiv), DCM, MS 3 Å, 0 C, 2 h. DCM=dichloromethane, TfOH=trifluor-
omethanesulfonic acid, NIS=N-iodosuccinimide, MS=molecular sieves,
TBAHS=tetrabutylammonium hydrogen sulfate.
3
coupling constant JH1-H2 (9.9 Hz). Phenyl 4-fluoro-β-d-thioglyco-
side 17 (required for the synthesis of methyl glycoside of 4’-
fluoro-LacN3 43) was prepared from known methyl 4-deoxy-4-
fluoro-6-O-trityl-α-d-glucopyranoside 44[37] (Scheme 3) by
a
minor modification of the published procedure.[38] Parallel
acetolysis of the trityl ether and methyl glycoside in 44 was
accomplished by treatment with acetic anhydride/FeCl3 furnish-
ing acetate 45 quantitatively. The acetate 45 was converted
into the respective bromide 20 by treatment with HBr/AcOH.[23]
Phase-transfer glycosylation of thiophenol with 20 provided
thioglycoside 17 in 50% yield after recrystallization. Glycosyla-
tion of 30 with the thioglycoside 17 using NIS/TfOH activation
proceeded in a fully-stereoselective manner, providing methyl
glycoside of 4’-fluoro-LacN3 43 in an excellent 92% yield
(Scheme 3).
Synthesis of phenyl 3-fluoro-β-thioglycoside 18 required for
the preparation of methyl glycoside of 3’-fluoro-LacN3 46
(Scheme 4) was developed from the known 1,6:3,4-dianhydro-
2-O-tosyl-β-d-galactopyranose (22) available from levoglucosan
in two steps.[25b,c] Reaction of 22 with sodium naphthalenide in
THF according to the procedure by Linclau et al.[39] provided
galacto-configured 3,4-epoxide 47 in an excellent 95% yield.
Treatment of 47 with sodium hydride in anhydrous DMF for six
hours resulted in isomerization to a mixture of 2,3- and 3,4-
epoxides, in which 1,6:2,3-dianhydro-β-d-gulopyranose
prevailed.[40] Subsequent addition of 2-naphthylmethylbromide
provided 4-O-naphthylmethyl 1,6:2,3-dianhydropyranose 23 in
Scheme 2. Divergent synthesis of acceptors 11, 13 and 30 from known
15.[19,22,29] a) BnBr (1.2 equiv), NaH (1.3 equiv), DMF, 0 C-rt, 24 h; b) NaBH3CN
°
°
(10.0 equiv), HCl in CPME (11.0 equiv), MS 3 Å, THF, 0 C, 1 h; c) ethylenedi-
amine (50.0 equiv), MeOH, reflux, 5 h; d) TfN3 in PE (2.5 equiv), DMAP
°
(1.0 equiv), MeCN, DCM, rt, 17 h; e) AcOH, H2O, 100 C, 0.5 h; f) DAST
°
(1.2 equiv), 2,4,6-collidine (2.4 equiv), DCM, 65 C, MW irradiation, 1 h; g) Tf2O
°
(1.5 equiv), pyridine (3.0 equiv), DCM, 0 C, 1 h; h) TBAA (3.0 equiv), DMF, rt,
°
17 h; i) DAST (3.0 equiv), DCM, À 78 C-rt, 55 h; j) MeOH, MeONa (cat); k)
KNO2 (14.7 equiv), DMF, rt, 17 h; l) Ac2O, pyridine, rt, 17 h. CPME=cyclopen-
tyl(methyl) ether, DMAP=4-dimethylaminopyridine, MeCN=acetonitrile,
DCM=dichloromethane, TBAA=tetrabutylammonium acetate, DAST=die-
thylaminosulfur trifluoride, MS=molecular sieves,
TfN3 =trifluoromethanesulfonyl azide.
Chem. Eur. J. 2021, 27, 1–13
www.chemeurj.org
4
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Paper
doi.org/10.1002/chem.202101752
Chemistry—A European Journal
Peracetate 49 was converted into the corresponding bromide
21 in an excellent yield (98%). It was subsequently glycosylated
with thiophenol, providing β-thioglycoside 18 in a very good
79% yield (Scheme 4). The NIS/TfOH-promoted glycosylation of
acceptor 30 with donor 18 afforded methyl glycoside of 3’-
fluoro-LacN3 46 in 72% yield and complete 1,2-trans stereo-
selectivity.
The synthesis of the 1,2-trans glycosidic bond in methyl 2-
azido-2’-fluoro-lactoside 50β (Table 1) could not rely on the
anchimeric assistance of the C2 functionality.[42] Therefore, we
prepared a series of 2-fluoro-galactosyl donors and tested their
reactivity and stereoselectivity in the glycosylation of acceptor
30 (Table 1). The TMSOTf-promoted glycosylation using acety-
lated 2-fluoro-trichloroacetimidate donor 25[26] in a 1:1 DCM/
MeCN mixture (reported to promote β-selectivity)[26] required
prolonged reaction time and elevated temperature to achieve
completion, and ultimately led to the formation of a complex
mixture of products (Table 1, entry 1). Therefore, we decided to
explore more reactive benzyl-protected 2-fluoro-trichloroaceti-
midate donor 26.[27] Glycosylation of 30 with 26α in DCM/MeCN
1:1 mixture led to 2-fluoro-N-trichloroacetyl-β-d-galactopyrano-
sylamine 51 as a major product resulting from imidate
rearrangement[43] in 39% yield, accompanied with only 6% yield
of the desired β-glycoside 50β (Table 1, entry 2). When DCM
alone was used as a solvent and TfOH as a promotor (Table 1,
entry 3), the glycosylation provided a separable mixture of
anomers 50α (14%) and 50β (28%). A slight increase in
anomeric β-stereoselectivity was observed when the acceptor
concentration was doubled (Table 1, entry 4). The use of β-
configured trichloroacetimidate 26β caused reversal of stereo-
selectivity (Table 1, entry 5) confirming that the TfOH-catalyzed
glycosylation with trichloroacetimidates proceeds to a signifi-
cant extend with inversion of the anomeric configuration.[27,44]
Scheme 4. Synthesis of methyl 2-azido-3’-fluoro-lactoside 46. a) Sodium
°
naphthalenide (2.6 equiv), THF, À 78 C, 0.5 h; b) NaH (2.0 equiv), DMF, rt, 6 h;
°
c) addition of 2-napththylmethyl bromide (2.0 equiv), À 25 C-rt, 15 h; d) KHF2
°
(9.5 equiv), KF (9.5 equiv), ethylene glycol, 200 C MW irradiation, 3 h; e) Pd/C
°
(10 wt %), H2 (1 atm), MeOH, 2 h; f) Ac2O, TESOTf (cat), 0 C, 1 h; g) HBr/AcOH
33% (10.0 equiv), DCM, 0 C to rt, 2 h; h) TBAHS (1.0 equiv), PhSH (3.3 equiv),
EtOAc, Na2CO3 (1 M aqueous solution), rt, 15 h; i) 18 (1.2 equiv), NIS
(1.6 equiv), TfOH (0.30 equiv), DCM, MS 3 Å, 0 C, 2 h. TESOTf=triethylsilyl
trifluoromethanesulfonate, TBAHS=tetrabutylammonium hydrogen sulfate,
DCM=dichloromethane, TfOH=trifluoromethanesulfonic acid, NIS=N-iodo-
succinimide, MS=molecular sieves, Nap=2-naphthylmethyl.
°
°
67% yield in a mixture with chromatographically separable 2-O-
naphthylmethyl 1,6:3,4-dianhydropyranose 48 obtained in yield
27%. Regioselective trans-diaxial opening of the 2,3-epoxide 23
according to the Fürst-Plattner rule[41] using a 1:1 KHF2/KF
mixture[39] under microwave irradiation to 200 C furnished 1,6-
°
anhydro-3-fluoro-galactopyranose 24 in 54% yield. Deprotec-
tion of the 3-O-naphthylmethyl group via catalytic hydro-
genolysis followed by triethylsilyl triflate-catalyzed acetolysis of
the internal acetal gave acetylated 3-fluoro-galactopyranose 49.
Table 1. Glycosylation of acceptor 30 with of 2-fluoro galacto-configured donors 25, 26, 27, and 28.
Entry
Donor
Acceptor (30) [equiv]
Promoter
/solvent
Product
α/β[d]
Yield [%][e]
[f]
[f]
[f]
1
2
3
4
5
6
25
2.0
0.8
2.0
4.0
2.0
1.5
1.5
0.8
TMSOTf, DCM/MeCN 1:1
TMSOTf, DCM/MeCN 1:1
TfOH, DCM
TfOH, DCM
TfOH, DCM
Ph2SO, TTBP, Tf2O, DCM
Ph2SO, TTBP, Tf2O, DCM
Ph2SO, TTBP, Tf2O, DCM
[g]
26α
26α
26α
26β
27
28
28
51
50
50
50
52
50
50
45[g]
42
40
1:1.9
1:2.3
2.5:1
1:1.8
1:1.2
1:1.7
50
36[h]
80
7
8[i]
71
[a] TBAHS (1.0 equiv), PhSH (3.1 equiv), EtOAc, Na2CO3 (1 M aqueous solution), rt, 15 h. [b] MeOH, MeONa (cat, pH=9), rt, 3 h. [c] NaH (4.9 equiv), BnBr
(4.8 equiv), DMF, 0 C to rt, 15 h. [d] Determined by 19F NMR. [e] Isolated yield. [f] The reaction led to a complex mixture of products. [g] The major product
°
isolated in 39% yield was 3,4,6-tri-O-benzyl-2-deoxy-2-fluoro-N-trichloroacetyl-β-d-galactopyranosylamine 51. [h] Purification of the product required base-
catalyzed deacetylation (see the Supporting Information). [i] The reaction was performed at a 1.19 mmol scale (for donor 28). TBAHS=tetrabutylammonium
hydrogen sulfate, MeCN=acetonitrile, DCM=dichloromethane, Tf2O=trifluoromethanesulfonic anhydride, Ph2SO=diphenyl sulfoxide, TTBP=2,4,6-tri-tert-
butylpyrimidine.
Chem. Eur. J. 2021, 27, 1–13
www.chemeurj.org
5
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Paper
doi.org/10.1002/chem.202101752
Chemistry—A European Journal
To further explore glycosylation with 2-fluoro-galactosyl donors,
2-fluoro-thioglycosides 27 and 28 were prepared from known
2-fluoro-α-d-galactosyl bromide 29[28] (Table 1) by phase-trans-
fer glycosylation, deacetylation and O-benzylation. Both donors
were activated using powerful Ph2SO/TTBP/Tf2O activation
protocol[45] at low temperatures. Acetyl-protected donor 27 was
not fully activated under these conditions (residual donor was
identified in the crude reaction mixture by NMR), which led to
difficulties in separation of products and low reaction yield
(Table 1, entry 6). On the contrary, activation of benzyl-
protected donor 28 in Ph2SO/TTBP/Tf2O system proceeded
smoothly, leading to high yield of product 50, although with
minimal stereoselectivity (Table 1, entry 7). Repeating the
reaction at a 1.19 mmol scale of the thiodonor using a
substoichiometric quantity of the acceptor 30 (Table 1, entry 8)
slightly improved the β-selectivity (α/β 1:1.7). All glycosylations
employing benzyl-protected donors 26α or 28 were syntheti-
cally usable because anomers 50α and 50β were separable. The
use of trichloroacetimidates led to lower glycosylation yields
(Table 1, entries 3–5) in comparison with thioglycosides (Ta-
ble 1, entries 7 and 8).
Deprotection to the target fluorinated N-acetyllactosamine
analogues 5–10
Fluorinated methyl 2-azido lactosides 39, 40, 42, 43, 46 and 50β
were converted to the corresponding fluorinated methyl N-
acetyllactosaminides 53–58 by treatment with thioacetic acid in
pyridine.[46] Final deprotection of these 2-acetamido-precursors
was accomplished by Zemplén deacetylation in methanol and
subsequent palladium-catalyzed hydrogenolysis of the O-benzyl
groups in methanol/acetic acid 4:1 to give the target fluoro
analogues 5–10 (Table 2)
Determination of lipophilicity
To evaluate the influence of deoxyfluorination at each position
on the lipophilicity expressed as log P, we employed 19F NMR-
based method recently developed by Linclau[18a] and then
applied to
a
number of mono and multifluorinated
monosaccharides.[18b-d] Briefly, following thermodynamic parti-
tion of the measured compound and a fluorine-containing
internal standard with known log P between water and n-
octanol, each phase is analyzed by 19F NMR spectroscopy and
log P is calculated from the values of relative integrals of
fluorine resonances of the measured compound and internal
standard in both phases (see equations S1-S3 in the Supporting
Information). The log P values for the LacNAc analogues are
Table 2. Deprotection to the target N-acetyllactosamine analogues 5–10.
O-protected
LacN3β1-OMe
O-protected
LacNAcβ1-OMe
Yield [%][e]
Fluorinated
LacNAcβ1-OMe
Yield [%][f]
39 (6F)[d]
40 (3F)
42 (6’F)
43 (4’F)
79
89
76
75
95
85
94
88
46 (3’F)
84
74
86
50β (2’F)
95[g]
°
[a] AcSH (100 equiv), pyridine, 10 C to rt, 15 h. [b] MeOH/MeONa (cat.), rt, 1–5 h. [c] Pd/C (10 wt %), H2 (1 atm), MeOH/CH3COOH 4:1, rt, 1–3 h. [d] The
number in parentheses indicates the position of fluorine. [e] isolated yield of O-protected LacNAcβ1-OMe. [f] isolated yield of analogues 5–10. [g] only
conditions [c] were used in the transformation of 58 to 10.
Chem. Eur. J. 2021, 27, 1–13
www.chemeurj.org
6
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Paper
doi.org/10.1002/chem.202101752
Chemistry—A European Journal
reported in Figure 3 along with those previously reported for
monofluorinated glucose analogues (the complete set of values
for monofluorinated galactose analogues are not available).[18b-d]
Glucose analogues 59–61 with fluorine in the secondary
positions C2-C4 showed very similar values within a narrow
range of log P=À 2.14 to À 2.21 while 6-fluoro-glucose 62 with
fluorine at the primary position showed a subtle decrease in
lipophilicity with log P=À 2.36. Comparing these values with
log P of LacNAc analogues 5–10, it is immediately apparent that
all LacNAc analogues showed significantly lower lipophilicities
than monofluorinated glucose. Furthermore, the differences
between individual compounds are more noticeable (maximum
Δlog P=0.53 for LacNAc analogues, maximum Δlog P=0.22 for
glucose analogues). Similarly, as in the glucose series, deoxy-
fluorination at the primary position C6’ of the galactose unit
resulted in the lowest lipophilicity in comparison with the
lipophilicities resulting from deoxyfluorination at the secondary
positions C2’-C4’ of this unit. Fluorination at the secondary
position C2’ of the galactose unit resulted in one of the highest
lipophilicity values in the LacNAc series, which contrasted with
the results in the glucose series. Unexpectedly, the lipophilicity
trend for C6 deoxyfluorination observed in the glucose series
and the LacNAc galactose moiety did not extend to the
glucosamine moiety as deoxyfluorination at the primary C6
position also resulted in one of the highest values of log P in
the series. Conversely, deoxyfluorination at the C3 of the
glucosamine unit gave the lowest value of log P in the series.
We speculate that this high hydrophilicity of the 3F-LacNAc
analogue may result from disruption of the transient inter-
residual hydrogen bond between O3-H and the galactose ring
oxygen O5’ (O5’···HÀ O3) upon deoxyfluorination.[47] The hydro-
gen-bonded O3-H hydroxyl is probably less solvated by water
molecules than are other hydroxyls, because its hydrogen-
bonding capability towards water is reduced,[48] and its
contribution to hydrophilicity of LacNAc is therefore less
significant. Accordingly, deoxyfluorination of O3 hydroxyl to 3F-
LacNAc 6 will not increase lipophilicity to such extent as
deoxyfluorination at other positions, also because it will expose
the ring oxygen O5’ to increased solvation by water.
Binding of LacNAc analogues 5–10to Gal-1and
Gal-3determined by ELISA
The binding of the prepared fluorinated LacNAc analogues to
galectins was tested in a competitive ELISA with recombinant
human Gal-1 and Gal-3, using the glycoprotein asialofetuin
(ASF) as the immobilized competitive galectin ligand. The
results are expressed in the form of IC50 values. The IC50 is
defined as the ligand concentration, at which half-maximal
inhibition of binding to the standard competitor ligand
asialofetuin (ASF) is reached. A lower value indicates stronger
binding.
We found that the fluorine substitution of LacNAc hydroxyls
did not substantially affect the selectivity between Gal-1 and
Gal-3 in most cases (Table 3, column 6). Generally, the affinity of
the prepared compounds for Gal-1 was lower than for Gal-3,
except for 2’-fluoro analogue 10. Compounds 6 (3F), 7 (6’F), and
8 (4’F) showed strong or even complete loss of affinity to either
galectin. Whereas a practically complete loss of inhibitory
potential was observed for Gal-1, leading to unsaturated
binding curves with IC50 in at least high milimolar range for
analogues 6–8 (at least 300-times weaker binding than under-
ivatized LacNAc), in the case of Gal-3 we observed just an
approximately 100-fold loss of binding potency. The weakest
Table 3. Inhibitory potential of LacNAc analogues for Gal-1 and Gal-3.
Compound
Affinity for Gal-1
IC50 [μM]
Affinity for Gal-3
IC50 [μM]
Selectivity
rp[a]
rp[a] for Gal-3[b]
LacNAc-
(standard)
LacNAcβ1-
OMe (63)
5 (6F) [c]
6 (3F)
7 (6’F)
8 (4’F)
9 (3’F)
10 (2’F)
131�40
1.0
49.7�8.9
1.0
1.9
2.8
2.6
5.1
1.9
132�56
1.0
26.1�8.4
17.9�2.9
33.4�9.4 3.9
>30000[d] <0.004 4504�1872 0.01 >6.7
>50000[d] <0.003 5221�1940 0.01 >9.6
>40000[d] <0.003 6268�2684 0.01 >6.4
243�82
0.5
38�11
1.3
1.7
6.4
0.5
15.6�7.7 8.4
29.2�8.6
[a] Relative inhibitory potency was defined as rp=IC50 (LacNAc)/ IC50
(ligand). [b] Selectivity for Gal-3 calculated as the ratio of IC50 (Gal-1)/ IC50
(Gal-3) shows how many times the affinity of the compound is higher for
Gal-3 than for Gal-1 under the same conditions. [c] The number in
parentheses indicates the position of fluorine. [d] Estimated values of IC50
are presented since full saturation of concentration-dependent inhibition
curves could not be reached due to limited compound solubility.
LacNAcβ1-OMe=methyl N-acetyl-β-d-lactosaminide.
Figure 3. The values of log P of N-acetyllactosamine analogues 5–10 and the
reported values for monofluorinated glucose analogues 59–62.[18d]
Chem. Eur. J. 2021, 27, 1–13
www.chemeurj.org
7
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Paper
doi.org/10.1002/chem.202101752
Chemistry—A European Journal
inhibitor of Gal-3 was compound 8 with IC50 of 6268 μM.
However, analogues 10 (2’F) and 5 (6F) showed approximately
8-fold and 4-fold stronger binding to Gal-1, respectively,
compared to underivatized LacNAcβ1-OMe[49] (63). This trend
did not reiterate for Gal-3 because only compound 5 (6F)
displayed slightly increased binding relative to underivatized
LacNAc1β-OMe (63).
Assessment of the binding affinity of the analogues 5–10to
Gal-1and Gal-3by 19F T2-filter measurement
The presence of fluorine in the target LacNAc analogues 5–10
also permits to evaluate their binding affinities to Gal-1 and
Gal-3 using 19F NMR experiments such as T2 filter screening
methods.[9,50] T2 filter screening exploits the change of the
hydrodynamic mobility of a low-molecular carbohydrate ligand
upon its binding to a protein. When the ligand is bound to the
protein, its hydrodynamic mobility is reduced, and its motional
correlation time increases towards that of the galectin. An
increase in the motional correlation time is accompanied by an
increase in the transverse relaxation rate (R2) and a decrease in
its reciprocal value, the transverse relaxation time (T2). Both
parameters R2 and T2 can be determined using the standard
Carr-Purcell-Meiboom-Gill (CPMG) spin-echo train sequence
experiment.[51] To explore the utility of the 19F NMR T2 filter
methods for rapid screening of galectin binding, we prepared a
mixture of LacNAc analogues 5–10, optimized the setting of the
CPMG spin-echo pulse sequence and performed a determina-
tion of R2 of individual LacNAc analogues 5–10 in their
equimolar mixture in the absence of any protein (Table 4, R2,free).
Due to a significant dispersion of 19F chemical shifts, 19F NMR
signals belonging to individual analogues did not overlap,
which enabled the determination of R2 for each compound in
the mixture of 5–10 in only one experimental setup. Once the
R2 values in the unbound state (R2,free) were determined, the
experiment was repeated under the same conditions in the
presence of increasing concentrations of recombinant Gal-1 or
Gal-3 (Figures 4A and B, respectively). Both tested human
galectins provided qualitatively similar results. The R2 relaxation
rates of analogues 6 (3F), 7 (6’F) and 8 (4’F) were almost
Figure 4. Dependence of 19F R2 relaxation rates of 5–10 on concentration of
Gal-1 (A) and Gal-3 (B). Concentrations of ligands were 0.5 mM each.
independent of protein concentration, conversely, compounds
5 (6F), 9 (3’F), and 10 (2’F) showed a clear positive correlation of
R2 relaxation rates with increasing concentrations of Gal-1 as
well as Gal-3 (Figures 4A and B, respectively). These findings
qualitatively agree with results of ELISA and confirm that
analogues 5 (6F), 9 (3’F) and 10 (2’F) are significantly stronger
Gal-1 and Gal-3 binders than compounds 6 (3F), 7 (6’F) and 8
(4’F).
Direct comparison of relative binding affinities of individual
analogues 5–10 to Gal-1 and Gal-3 is possible via calculation of
the relative increase of transversal relaxation rate (R2,increase), of
individual analogues 5–10 after addition of galectin from
measured R2,free and R2,obs values (Table 4 and Eq. (1)). Although
R
2,increase of analogues 5–10 in the mixture with galectin depends
Table 4. Measured values of 19F transversal relaxation rates of LacNAc
analogues 5–10 in the absence of any galectin (R2,free), in the presence of
galectin (R2,obs) and in the presence of galectin as well as competing
LacNAcβ1-OMe (63) (R2,obs+I).
on several factors, in a series of structurally similar low-
molecular ligands (such as our series), it is qualitatively related
with binding affinity.
R
2,obs [sÀ 1
Gal-1[b] Gal-3[b] Gal-1[b,c] Gal-1[b,d] Gal-3[b,c]
]
R ]
2,obs+I [sÀ 1
Analogue[a]
R ]
2,free [sÀ 1
R2; obs À R2; free
R2; increase
¼
� 100
(1)
R2; free
5 (6F)
6 (3F)
7 (6’F)
8 (4’F)
9 (3’F)
10 (2’F)
2.69
1.79
1.80
1.50
1.63
1.65
6.50
1.85
1.75
1.49
5.15
7.54
4.94
1.88
2.08
1.94
3.86
3.85
5.11
1.93
1.90
1.55
3.88
5.41
4.68
2.24
1.98
1.94
2.98
3.68
3.36
1.77
1.90
1.60
2.27
2.44
The calculated R2,increase are depicted in Figure 5, which
shows that compounds 5 (6F), 9 (3’F) and 10 (2’F) achieved
substantially higher values of R2, increase in the presence of Gal-1
(142%, 261% and 357%, respectively) relative to compounds 6
(3F), 7 (6’F) and 8 (4’F) where the values of R2,increase were close
to 0%. Dependence of R2 on concentration of Gal-3 was less
distinct (compare Figures 4A and B). Nevertheless, its presence
[a] Concentrations of ligands was 0.5 mM each and the number in
parentheses indicates the position of fluorine. [b] Concentration of
galectin was 50 μM. [c] 63 (1.5 mM) was added prior to the measurement.
[d] 63 (5.0 mM) was added prior to the measurement.
Chem. Eur. J. 2021, 27, 1–13
www.chemeurj.org
8
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Paper
doi.org/10.1002/chem.202101752
Chemistry—A European Journal
Figure 6. Displacement values F of ligands 5 (6F), 9 (3’F) and 10 (2’F). [a]
Addition of 63 (1.5 mM concentration). [b] Addition of 63 (5.0 mM
concentration).
Figure 5. R2,increase values of analogues 5 (6F), 6 (3F), 7 (6’F), 8 (4’F), 9 (3’F) and
10 (2’F) after addition of Gal-1 (50 μM concentration) or Gal-3 (50 μM
concentration).
ligand than in the case of Gal-3. The 19F NMR T2 filter method
thus enables a rapid qualitative identification of galectin ligands
in a mixture during one experimental setup.
also caused a noticeable rise in R2,increase values of compounds 5
(6F), 9 (3’F) and 10 (2’F) (84%, 137% and 133%, respectively) in
comparison with R2,increase of compounds 6 (3F), 7 (6’F) and 8
(4’F) (5%, 16% and 30% respectively), which enabled a clear
identification of ligands. Gal-3 also shows weak but still Discussion
detectable binding to deoxyfluorinated ligands 6 (3F), 7 (6’F)
and 8 (4’F) in qualitative agreement with the ELISA.
The fluorine substituent is a powerful probe to investigate
glycan-protein interactions. To explore its utility for lactos-
amine-binding lectins, we performed the synthesis of a full
series of mono-deoxyfluorinated methyl β-glycosides of LacNAc.
The use O-benzylated methyl 2-azido-glucopyranosides as
glycosyl acceptors and acetylated thiophenyl galactosides as
glycosyl donors was a viable approach giving 68–93% yields
and complete stereoselectivity except for 2’-fluoro-LacNAc. We
were unable to reproduce the fair yield and stereoselectivity
reported for 2-fluoro-trichloroacetimidate donor 25,[26] however
employing O-benzylated 2-fluoro-thiogalactoside donor 28 was
a usable alternative. It should be noted that a low quantity of
the 6’-fluoro analogue 7 was formerly prepared by enzymatic
glycosylation using UDP-6-fluoro-d-galactose.[53] The remaining
fluoro analogues are new compounds, although previously
reported syntheses of fluorinated H type 2 human blood group
antigen[54] and Lewis X trisaccharide[55] relied on appropriately
protected 6-fluoro-LacNAc 5 and 4’-fluoro-LacNAc 8, respec-
tively, as glycosyl acceptors. Also, O1-tert-butyldimethylsilyl
analogue of 2’-fluoro-LacNAc 10 was employed as a substrate
in enzymatic sialylation.[26]
The replacement of a hydroxyl group for fluorine brings
about minimal steric alterations because fluorine is considered
isosteric with oxygen.[56] Since organic fluorine cannot act as a
hydrogen bond donor and is usually a weak hydrogen bond
acceptor, mono-deoxyfluorinated sugars are useful probes of
the hydrogen bonding network of carbohydrate ligands.[23b,57]
ELISA and T2-filter screening showed that fluorine installation at
C3, C6’ and C4’ in LacNAc analogues 6–8 resulted in a large
(Gal-3) or complete (Gal-1) loss of affinity, confirming informa-
tion obtained from X-ray crystallography where the C3, C4’, and
C6’ hydroxyl groups form key hydrogen bonds in highly
conserved binding subsites C and D of galectin CRDs.[58] The
�
�
R2; obs À R2; obs þI
R2; obs À R2; free
F ¼
� 100
(2)
The addition of the non-fluorinated ligand methyl N-acetyl-
β-d-lactosaminide 63[49] to the mixtures of all analogues 5–10
with Gal-1 or Gal-3 caused a noticeable drop in 19F R2 relaxation
rates of analogues 5 (6F), 9 (3’F) and 10 (2’F) (Table 4, R2,obs+I).
This measurement of R2,obs+I allowed to use Equation 2[52] for
calculation of the displacement value F describing relative
decrease of the transverse relaxation rate of ligands 5 (6F), 9
(3’F) and 10 (2’F) upon addition of the competing ligand.
This F value can be interpreted as an estimated extent of
displacement of a particular fluorinated ligand in the CRDs by
63. The addition of 63 (1.5 mM concentration) to the mixture of
analogues 5–10 and Gal-1 led to F values 37%, 36% and 36%
for ligands 5 (6F), 9 (3’F), and 10 (2’F), respectively (Figure 6).
The F values of these ligands further rose to 48%, 62%, and
66% with an increase in concentration of 63 to 5.0 mM. When
63 (1.5 mM concentration) was added to the mixture of 5–10
with Gal-3, the F values of ligands 5 (6F), 9 (3’F), and 10 (2’F)
achieved 70%, 71%, and 64%, respectively. Taking into
consideration the structural similarity between analogues 5–10
and standard 63, this observation indicates that compound 63
acts as a competitive ligand for compounds 5 (6F), 9 (3’F), and
10 (2’F) in recognition by both Gal-1 and Gal-3, and that natural
and fluorinated LacNAc bind to the same binding site. More-
over, we can compare F values achieved after the addition of
the same amount of 63 (1.5 mM concentration) to the mixture
of 5–10 with Gal-1 or Gal-3 (Figure 6). Higher F values in the
case of Gal-3 indicate that ligands 5 (6F), 9 (3’F) and 10 (2’F)
bind Gal-1 with a higher affinity relative to the non-fluorinated
Chem. Eur. J. 2021, 27, 1–13
www.chemeurj.org
9
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Paper
doi.org/10.1002/chem.202101752
Chemistry—A European Journal
information from X-ray diffraction analysis was also corrobo-
rated by a complementary group-selective STD NMR method
(GS-STD) used to map the contact points between Gal-1 and
LacNAc in solution.[59] GS-STD confirmed that the conformation
and position of the LacNAc ligand in the binding site of Gal-1 in
solution and the crystal structure are comparable.
two groups.[63] Spatial proximity of His52 to the C2’-H2’ bond
was also suggested on the basis of GS-STD measurement.[59]
Although His52 residue in the apo form of Gal-1 is predom-
inantly present in the neutral Nɛ2-H tautomer, the presence of
1
the protonated form was indicated by H-15N long-range HMQC
experiments.[64] Therefore, deoxyfluorination at C2’ could
strengthen binding to Gal-1 by dipole-dipole or dipole-ion
interaction with neutral or protonated forms of His52, respec-
tively. According to ab initio computation, dissociation energies
of N+À H···FÀ C complexes can reach up to 13.5 kcal/mol.[65]
Moreover, a similar interaction between fluorine and the two
terminal nitrogen atoms of Arg21 residue has been observed in
human neuraminidase hNeu2, trapped in the form of a covalent
intermediate with 3-fluoro-sialic acid possessing a C3 equatorial
fluorine.[66]
Both ELISA and T2 filter screening confirmed that deoxy-
fluorination of the hydroxyl groups at C6 and C3’ resulted in
ligands with the affinity comparable to the non-fluorinated
ligand for both Gal-1 and Gal-3. This is consistent with X-ray
crystallography results suggesting that C6 and C3’ hydroxyl
groups are not engaged in the hydrogen-bonding network.[63]
In this case the agreement between ELISA and T2 filter screen-
ing was only qualitative for Gal-1. In particular, 3’-fluoro
analogue 9 was about a 7-fold weaker ligand than compound 5
(6-fluoro) according to ELISA. However, binding affinities are
assay-dependent and while NMR screening reflects changes in
hydrodynamic mobility of the ligand upon binding to protein,
the ELISA-type assay is based on competition between mono-
valent fluorinated ligand in solution and multiple lactosamine
epitopes attached to asialofetuin.
The complete loss of affinity for Gal-1 corresponds to a
similar loss of affinity previously found for bovine heart
galectin-1 when probed with 4’-fluoro, and 6’-fluoro methyl β-
lactosides.[57a] It is noteworthy that LacNAcβ1-OMe retained an
approximately 2-orders of magnitude weaker but still detect-
able affinity to Gal-3 upon deoxyfluorination at the 3-, 4’- and
6’-positions (ligands 6, 8 and 7, respectively). Very weak binding
of 6, 7 and 8 to Gal-3 determined by ELISA was also
corroborated by NMR from a relative increase in R2 values upon
addition of galectins (Figure 5), underscoring the utility of this
NMR method for sensitive screening of ligand-galectin inter-
actions. Gal-3 has also been recently probed with nanoparticles
carrying deoxyfluorinated lacto-N-biose and the 6’-position has
been confirmed as critical because deoxyfluorination at C6’
abolished binding.[16]
The introduction of a highly polar CÀ F bond can strengthen
noncovalent ligand-protein interaction for example via electro-
static and dipolar interactions with cationic residues (e.g. the
guanidinium group) and protein backbone or side-chain amide
groups.[60] While this strategy to improve affinity of protein
ligands was generally exceptionally fruitful, including galectin
ligands fluorinated at aryl substituents[13,61] or polyfluorinated
carbohydrates,[62] mono-deoxyfluorinated oligosaccharides typi-
cally display similar, attenuated or abrogated affinity to lectins
in comparison with the natural ligands.[57a–d] Exceptions include
approximately 4-fold enhancement of the binding affinity of
immobilized clustered fluorinated trisaccharides NeuAcα2-3(2F) Conclusion
Galß1-4GlcNAc and NeuAcα2-3(2F)Galß1-3GlcNAc over the non-
fluorinated ligands towards adhesive protein TgMIC1 of proto-
zoan parasite Toxoplasma gondii.[15,26] In solution, however, the
fluorinated and non-fluorinated ligands gave comparable
affinities. Also, affinity enhancement reported for selected
fluorinated Galβ1-3GlcNAc (lacto-N-biose) analogues towards
Gal-3 and 7 (Figure 1) was achieved using multivalent presenta-
tion of the fluorinated ligands on nanoparticles. Here, we
observed an unprecedented approximately 8-fold affinity
enhancement for monovalent 2’-fluoro analogue 10 over the
non-fluorinated ligand 63 towards Gal-1 in solution using ELISA
(Table 3) and qualitatively corroborated by NMR screening,
which delivered the highest R2,increase value within the set of
analogues (Figure 5). This enhancement was specific for Gal-1
because the same analogue 10 binds Gal-3 with a similar
affinity to its non-fluorinated counterpart 63.
Despite the highly conserved structure of human galectins
CRDs, subtle structural differences between the binding sites of
Gal-1 and Gal-3 exist.[63] A notable difference is the presence of
the L4 loop containing His52 residue in human Gal-1, which is
absent in human Gal-3 and bovine Gal-1 and Gal-3.[58]
Interestingly, the crystal structure of Gal-1 complexed with
lactose confirmed the close spatial proximity of His52 and C2’-
OH and indicated a possible hydrogen bond between these
In summary, we reported the novel chemical synthesis of a
complete series of mono-deoxyfluorinated methyl N-acetyl-β-d-
lactosaminides and the determination of their affinity to human
galectin-1 and -3 by using 19F NMR T2 filter and ELISA. The set of
fluoro analogues displays a favorable distribution of 19F NMR
chemical shifts (Figure S1) and the results of the T2 filter assay
qualitatively correlate with the ELISA and reported X-ray
crystallography data, thus validating the T2 filtering method for
screening of other galectins with our set of fluoro analogues.
The advantage of this method is providing information on
binding of series of ligands in only one experiment. We also
showed that deoxyfluorination at the C2’ position created a
ligand with improved binding to Gal-1. To our knowledge, this
is the first mono-deoxyfluorinated oligosaccharide monovalent
ligand to exhibit significantly enhanced binding to a lectin.
Although deoxyfluorination of hydroxyls at the positions 3, 4’
and 6’ of the LacNAc scaffold led to the complete abrogation of
binding affinity to Gal-1, both analytical methods confirmed
weak binding of the respective analogues to Gal-3. We envisage
that aside from screening, binding preferences of other
galectins, analogues 5–10 may be employed to study ligand-
lectin interactions in detail by means of 19F-related saturation
transfer difference NMR experiments. This will be particularly
Chem. Eur. J. 2021, 27, 1–13
www.chemeurj.org
10
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Paper
doi.org/10.1002/chem.202101752
Chemistry—A European Journal
H.-J. Gabius, J. Jiménez-Barbero, F. J. Cañada, J. Org. Chem. 2020, 85,
16072–16081.
[10] T. Diercks, J. P. Ribeiro, F. J. Cañada, S. André, J. Jiménez-Barbero, H.-J.
Gabius, Chem. Eur. J. 2009, 15, 5666–5668.
[11] T. Diercks, A. S. Infantino, L. Unione, J. Jiménez-Barbero, S. Oscarson, H.-
J. Gabius, Chem. Eur. J. 2018, 24, 15761–15765.
[12] A. Ardá, J. Jiménez-Barbero, Chem. Commun. 2018, 54, 4761–4769.
[13] T.-J. Hsieh, H.-Y. Lin, Z. Tu, T.-C. Lin, S.-C. Wu, Y.-Y. Tseng, F.-T. Liu, S.-
T. D. Hsu, C.-H. Lin, Sci. Rep. 2016, 6, 29457.
[14] C. Romanò, S. Oscarson, Org. Biomol. Chem. 2019, 17, 2265–2278.
[15] J. A. Garnett, Y. Liu, E. Leon, S. A. Allman, N. Friedrich, S. Saouros, S.
Curry, D. Soldati-Favre, B. G. Davis, T. Feizi, S. Matthews, Protein Sci.
2009, 18, 1935–1947.
important for galectins other than Gal-1 and Gal-3 where fewer
data on binding of glycomimetics are available. Future research
should also focus on the preparation of multivalent constructs
with fluorinated LacNAc to determine how binding character-
istics of multivalently displayed fluoro analogues relate to the
binding of monovalent ligands reported here. Recent results
with lacto-N-biose analogues highlight the potential of the
multivalent display of fluorinated sugars to identify highly
potent and/or selective inhibitors.[16]
[16] S.-J. Richards, T. Keenan, J.-B. Vendeville, D. E. Wheatley, H. Chidwick, D.
Budhadev, C. E. Council, C. S. Webster, H. Ledru, A. N. Baker, M. Walker,
M. C. Galan, B. Linclau, M. A. Fascione, M. I. Gibson, Chem. Sci. 2021, 12,
905–910.
Acknowledgements
[17] a) M. Johannes, M. Reindl, B. Gerlitzki, E. Schmitt, A. Hoffmann-Röder,
Beilstein J. Org. Chem. 2015, 11, 155–161; b) A. Axer, R. P. Jumde, S.
Adam, A. Faust, M. Schäfers, M. Fobker, J. Koehnke, A. K. H. Hirsch, R.
Gilmour, Chem. Sci. 2021, 12, 1286–1294.
[18] a) B. Linclau, Z. Wang, G. Compain, V. Paumelle, C. Q. Fontenelle, N.
Wells, A. Weymouth-Wilson, Angew. Chem. Int. Ed. 2016, 55, 674–678;
Angew. Chem. 2016, 128, 684–688; b) V. Denavit, D. Lainé, J. St-Gelais,
P. A. Johnson, D. Giguère, Nat. Commun. 2018, 9, 4721; c) J. St-Gelais, M.
Bouchard, V. Denavit, D. Giguère, J. Org. Chem. 2019, 84, 8509–8522;
d) J. St-Gelais, É. Côté, D. Lainé, P. A. Johnson, D. Giguère, Chem. Eur. J.
2020, 26, 13499–13506.
We thank Ministry of Education, Youth and Sport (INTERCOST,
no. LTC20052) and Czech Science Foundation (grant no. 20-
01472S) for supporting our research. P. Bojarová acknowledges
support from LTC19038 by the Ministry of Education, Youth and
Sport (COST Action CA17140). This publication reports research
conducted within the framework of COST Action CA18103
(INNOGLY) supported by European Cooperation in Science and
Technology.
[19] D. Crich, V. Dudkin, J. Am. Chem. Soc. 2001, 123, 6819–6825.
[20] S. Wagner, C. Mersch, A. Hoffmann-Röder, Chem. Eur. J. 2010, 16, 7319–
7330.
[21] a) G. Wasonga, Y. Tatara, I. Kakizaki, X. Huang, J. Carbohydr. Chem. 2013,
32, 392–409; b) D. A. Williams, K. Pradhan, A. Paul, I. R. Olin, O. T. Tuck,
K. D. Moulton, S. S. Kulkarni, D. H. Dube, Chem. Sci. 2020, 11, 1761–1774.
[22] J. Alais, S. David, Carbohydr. Res. 1990, 201, 69–77.
Conflict of Interest
The authors declare no conflict of interest.
[23] a) J. Xia, J. Xue, R. D. Locke, E. V. Chandrasekaran, T. Srikrishnan, K. L.
Matta, J. Org. Chem. 2006, 71, 3696–3706; b) V. Denavit, D. Lainé, C.
Bouzriba, E. Shanina, É. Gillon, S. Fortin, C. Rademacher, A. Imberty, D.
Giguère, Chem. Eur. J. 2019, 25, 4478–4490.
Keywords: carbohydrates
spectroscopy · T2 filter
·
ELISA
·
galectins
·
NMR
[24] a) A. Hoffmann-Röder, M. Johannes, Chem. Commun. 2011, 47, 9903–
9905; b) T. L. Lowary, O. Hindsgaul, Carbohydr. Res. 1993, 249, 163–195.
[25] a) T. Holmstrøm, D. Raydan, C. M. Pedersen, Beilstein J. Org. Chem. 2020,
16, 2788–2794; b) M. Černý, V. Gut, J. Pacák, Collect. Czech. Chem.
Commun. 1961, 26, 2542–2550; c) T. S. Rasmussen, H. H. Jensen, Org.
Biomol. Chem. 2010, 8, 433–441.
[26] S. A. Allman, H. H. Jensen, B. Vijayakrishnan, J. A. Garnett, E. Leon, Y. Liu,
D. C. Anthony, N. R. Sibson, T. Feizi, S. Matthews, B. G. Davis,
ChemBioChem 2009, 10, 2522–2529.
[27] E. Durantie, C. Bucher, R. Gilmour, Chem. Eur. J. 2012, 18, 8208–8215.
[28] M. Salvadó, B. Amgarten, S. Castillón, G. J. L. Bernardes, O. Boutureira,
Org. Lett. 2015, 17, 2836–2839.
[29] R. P. McGeary, K. Wright, I. Toth, J. Org. Chem. 2001, 66, 5102–5105.
[30] C. J. Cavender, V. J. Shiner, J. Org. Chem. 1972, 37, 3567–3569.
[31] D. A. Schwartz, H.-H. Lee, J. P. Carver, J. J. Krepinsky, Can. J. Chem. 1985,
63, 1073–1079.
[1] R. D. Cummings, F. T. Liu, in Essentials of Glycobiology, 3rd edition. (Eds.:
A. Varki, R. D. Cummings, J. D. Esko, et al.), Cold Spring Harbor
Laboratory Press, Cold Spring Harbor (NY), 2017, p. 469–480.
[2] H. Leffler, S. Carlsson, M. Hedlund, Y. Qian, F. Poirier, Glycoconjugate J.
2002, 19, 433–440.
[3] a) M. F. Brinchmann, D. M. Patel, M. H. Iversen, Mediators Inflammation
2018, 9186940; b) N. L. Perillo, K. E. Pace, J. J. Seilhamer, L. G. Baum,
Nature 1995, 378, 736–739.
[4] R. P. M. Dings, M. C. Miller, R. J. Griffin, K. H. Mayo, Int. J. Mol. Sci. 2018,
19, 905.
[5] FÀ T. Liu, G. A. Rabinovich, Nat. Rev. Cancer 2005, 5, 29–41.
[6] a) L. Ingrassia, I. Camby, F. Lefranc, V. Mathieu, P. Nshimyumukiza, F.
Darro, R. Kiss, Curr. Med. Chem. 2006, 13, 3513–3527; b) R. J. Pieters,
ChemBioChem 2006, 7, 721–728; c) V. Denavit, D. Lainé, T. Tremblay, J.
St-Gelais, D. Giguère, Trends Glycosci. Glycotechnol. 2018, 30, SE21-SE40;
d) C. T. Öberg, H. Leffler, U. J. Nilsson, Chimia 2011, 65, 18–23; e) H.
Leffler, U. J. Nilsson, ACS Symposium Series Vol. 1115, Galectins and
Disease Implications for Targeted Therapeutics (Eds.: A. A. Klyosov, P. G.
Traber), American Chemical Society, Washington, D. C., 2012, Chapter 2,
p. 47-59; f) K. H. Mayo, ACS Symposium Series Vol. 1115, Galectins and
Disease Implications for Targeted Therapeutics (Eds.: A. A. Klyosov, P. G.
Traber), American Chemical Society, Washington, D. C., 2012, Chapter 3,
p. 61-77.
[7] T. Delaine, P. Collins, A. MacKinnon, G. Sharma, J. Stegmayr, V. K. Rajput,
S. Mandal, I. Cumpstey, A. Larumbe, B. A. Salameh, B. Kahl-Knutsson, H.
van Hattum, M. van Scherpenzeel, R. J. Pieters, T. Sethi, H. Schambye, S.
Oredsson, H. Leffler, H. Blanchard, U. J. Nilsson, ChemBioChem 2016, 17,
1759–1770.
[8] B. Linclau, A. Ardá, N.-C. Reichardt, M. Sollogoub, L. Unione, S. P.
Vincent, J. Jiménez-Barbero, Chem. Soc. Rev. 2020, 49, 3863–3888.
[9] a) J. D. Martínez, P. Valverde, S. Delgado, C. Romanò, B. Linclau, N. C.
Reichardt, S. Oscarson, A. Ardá, J. Jiménez-Barbero, F. J. Cañada,
Molecules 2019, 24, 2337; b) J. D. Martínez, A. I. Manzano, E. Calviño,
A. d. Diego, B. Rodriguez de Francisco, C. Romanò, S. Oscarson, O. Millet,
[32] Z. Wang, Comprehensive Organic Name Reactions and Reagents (Ed.: Z.
Wang), 2010, p. 3123–3128.
[33] a) R. Lattrell, G. Lohaus, Justus Liebigs Ann. Chem. 1974, 1974, 901–920;
b) H. Dong, Z. Pei, O. Ramström, J. Org. Chem. 2006, 71, 3306–3309.
[34] K. Dax, M. Albert, J. Ortner, B. J. Paul, Carbohydr. Res. 2000, 327, 47–86.
[35] a) J. Xie, A. Molina, S. Czernecki, J. Carbohydr. Chem. 1999, 18, 481–498;
b) A. K. Balcerzak, S. S. Ferreira, J. F. Trant, R. N. Ben, Bioorg. Med. Chem.
Lett. 2012, 22, 1719–1721.
[36] G. H. Veeneman, S. H. van Leeuwen, J. H. van Boom, Tetrahedron Lett.
1990, 31, 1331–1334.
[37] a) C. W. Somawardhana, E. G. Brunngraber, Carbohydr. Res. 1983, 121,
51–60; b) P. Valverde, J.-B. Vendeville, K. Hollingsworth, A. P. Mattey, T.
Keenan, H. Chidwick, H. Ledru, K. Huonnic, K. Huang, M. E. Light, N.
Turner, J. Jiménez-Barbero, M. C. Galan, M. A. Fascione, S. Flitsch, W. B.
Turnbull, B. Linclau, Chem. Commun. 2020, 56, 6408–6411.
[38] M.-C. Chapeau, P. A. Frey, J. Org. Chem. 1994, 59, 6994–6998.
[39] C. Q. Fontenelle, D. Shishmarev, P. W. Kuchel, B. Linclau, Trends
Carbohydr. Res. 2017, 9, 29–34.
[40] J. St-Gelais, V. Denavit, D. Giguère, Org. Biomol. Chem. 2020, 18, 3903–
3907.
Chem. Eur. J. 2021, 27, 1–13
www.chemeurj.org
11
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Paper
doi.org/10.1002/chem.202101752
Chemistry—A European Journal
[41] A. Fürst, P. A. Plattner, Helv. Chim. Acta 1949, 32, 275–283.
[42] L. K. Mydock, A. V. Demchenko, Org. Biomol. Chem. 2010, 8, 497–510.
[43] H. M. Christensen, S. Oscarson, H. H. Jensen, Carbohydr. Res. 2015, 408,
51–95.
[44] A. Rencurosi, L. Lay, G. Russo, E. Caneva, L. Poletti, J. Org. Chem. 2005,
70, 7765–7768.
Protein Struct. 1975, 379, 456–461; f) L. Bhattacharyya, C. F. Brewer, Eur.
J. Biochem. 1988, 176, 207–212.
[58] Y.-C. Chan, H.-Y. Lin, Z. Tu, Y.-H. Kuo, S.-T. D. Hsu, C.-H. Lin, Int. J. Mol.
Sci. 2018, 19, 392.
[59] K. E. Kövér, E. Wéber, T. A. Martinek, É. Monostori, G. Batta, Chem-
BioChem 2010, 11, 2182–2187.
[60] a) K. Müller, C. Faeh, F. Diederich, Science 2007, 317, 1881–1886; b) J.
Pollock, D. Borkin, G. Lund, T. Purohit, E. Dyguda-Kazimierowicz, J.
Grembecka, T. Cierpicki, J. Med. Chem. 2015, 58, 7465–7474; c) W. K.
Hagmann, J. Med. Chem. 2008, 51, 4359–4369.
[45] J. D. C. Codée, R. E. J. N. Litjens, R. den Heeten, H. S. Overkleeft, J. H.
van Boom, G. A. van der Marel, Org. Lett. 2003, 5, 1519–1522.
[46] N. Shangguan, S. Katukojvala, R. Greenberg, L. J. Williams, J. Am. Chem.
Soc. 2003, 125, 7754–7755.
[61] a) F. R. Zetterberg, K. Peterson, R. E. Johnsson, T. Brimert, M. Håkansson,
D. T. Logan, H. Leffler, U. J. Nilsson, ChemMedChem 2018, 13, 133–137;
b) K. Peterson, R. Kumar, O. Stenström, P. Verma, P. R. Verma, M.
Håkansson, B. Kahl-Knutsson, F. Zetterberg, H. Leffler, M. Akke, D. T.
Logan, U. J. Nilsson, J. Med. Chem. 2018, 61, 1164–1175.
[62] I. N’Go, S. Golten, A. Ardá, J. Cañada, J. Jiménez-Barbero, B. Linclau, S. P.
Vincent, Chem. Eur. J. 2014, 20, 106–112.
[63] a) M. F. López-Lucendo, D. Solís, S. André, J. Hirabayashi, K.-i. Kasai, H.
Kaltner, H.-J. Gabius, A. Romero, J. Mol. Biol. 2004, 343, 957–970; b) J. Su,
T. Zhang, P. Wang, F. Liu, G. Tai, Y. Zhou, Acta Biochim. Biophys. Sin.
2015, 47, 192–198.
[64] S. Bertuzzi, A. Gimeno, R. Núñez-Franco, G. Bernardo-Seisdedos, S.
Delgado, G. Jiménez-Osés, O. Millet, J. Jiménez-Barbero, A. Ardá, Chem.
Eur. J. 2020, 26, 15643–15653.
[47] J. Rönnols, O. Engström, U. Schnupf, E. Säwén, J. W. Brady, G. Widmalm,
ChemBioChem 2019, 20, 2519–2528.
[48] D. Chen, M. Zhao, W. Tan, Y. Li, X. Li, Y. Li, X. Fan, Eur. J. Pharm. Sci.
2019, 130, 100–106.
[49] J. Landsröm, M. Bergström, C. Hamark, S. Ohlson, G. Wildman, Org.
Biomol. Chem. 2012, 10, 3019–3032.
[50] C. Dalvit, A. Vulpetti, J. Med. Chem. 2019, 62, 2218–2244.
[51] S. Meiboom, D. Gill, Rev. Sci. Instrum. 1958, 29, 688–691.
[52] C. Dalvit, A. Parent, F. Vallée, M. Mathieu, A. Rak, ChemMedChem 2019,
14, 1115–1127.
[53] Y. Kajihara, T. Endo, H. Ogasawara, H. Kodama, H. Hashimoto, Carbohydr.
Res. 1995, 269, 273–294.
[54] E. Petrakova, U. Spohr, R. U. Lemieux, Can. J. Chem. 1992, 70, 233–240.
[55] C. J. Moore, F.-I. Auzanneau, Beilstein J. Org. Chem. 2012, 8, 1134–1143.
[56] M. Tredwell, V. Gouverneur, in Comprehensive Chirality (Eds.: E. M.
Carreira, H. Yamamoto), Elsevier, Amsterdam, 2012, p. 70–85.
[57] a) D. SOLÍS, J. JIMÉNEZ-BARBERO, M. MARTÍN-LOMAS, T. DÍAZ-MAURIÑO,
Eur. J. Biochem. 1994, 223, 107–114; b) D. Solís, A. Romero, H. Kaltner,
H.-J. Gabius, T. Díaz-Mauriño, J. Biol. Chem. 1996, 271, 12744–12748;
c) A. Tatami, Y.-S. Hon, I. Matsuo, M. Takatani, H. Koshino, Y. Ito,
Biochem. Biophys. Res. Commun. 2007, 364, 332–337; d) P. S. Vermersch,
J. J. G. Tesmer, F. A. Quiocho, J. Mol. Biol. 1992, 226, 923–929; e) J. P.
van Wauwe, F. G. Loontiens, C. K. de Bruyne, Biochim. Biophys. Acta
[65] L. Shimoni, J. P. Glusker, C. W. Bock, J. Phys. Chem. 1995, 99, 1194–1198.
[66] S. Buchini, F.-X. Gallat, I. R. Greig, J.-H. Kim, S. Wakatsuki, L. M. G. Chavas,
S. G. Withers, Angew. Chem. Int. Ed. 2014, 53, 3382–3386; Angew. Chem.
2014, 126, 3450–3454.
Manuscript received: May 17, 2021
Accepted manuscript online: July 3, 2021
Version of record online: ■■■, ■■■■
Chem. Eur. J. 2021, 27, 1–13
www.chemeurj.org
12
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

FULL PAPER
Six mono-deoxyfluorinated LacNAc
analogues were prepared, and their
affinity to galectin-1 and À 3 was
M. Kurfiřt, Dr. M. Dračínský, Dr. L. Čer-
venková Šťastná, Dr. P. Cuřínová, V.
Hamala, M. Hovorková, Dr. P. Bojarová,
Dr. J. Karban*
tested. Fluorination at positions 3, 4’,
and 6’ abrogated binding to galectin-
1 and substantially reduced binding
to galectin-3. Placement of fluorine at
the positions 6 and 2’ increased
1 – 13
Selectively Deoxyfluorinated N-Ace-
tyllactosamine Analogues as 19F NMR
Probes to Study Carbohydrate-
Galectin Interactions
affinity to galectin-1, whereas binding
to galectin-3 was unaffected. Fluorina-
tion at position 3’ did not significantly
influence binding to either galectin.