Toward creating cell membrane glyco-landscapes with glycan lipid constructs

Article abstract of DOI:10.1016/j.carres.2012.03.044

Synthetic glycolipid-like constructs dispersible in biological media and capable of incorporating into cell membranes have the ability to create novel artificial glyco-landscapes on living cells. Using a variety of different glycans ranging from disaccharides to polysaccharides, together with different lengths and high hydrophilicity spacers, we created a series of synthetic glycolipid-like constructs. Contacting these constructs with live cells gave modified cells with controlled glycan density and/or altered biological function. The ability to also use these constructs as solutions to inhibit antibodies, toxins, and virions extends the potential diagnostic and therapeutic uses for these synthetic glycolipid-like constructs.

Full text of DOI:10.1016/j.carres.2012.03.044

Carbohydrate Research 356 (2012) 238–246
Contents lists available at SciVerse ScienceDirect
Carbohydrate Research
journal homepage: www.elsevier.com/locate/carres
Toward creating cell membrane glyco-landscapes with glycan lipid constructs
Elena Korchagina a, Alexander Tuzikov a, Andrey Formanovsky a, Inna Popova a, Stephen Henry b,
Nicolai Bovin a,
⇑
a Shemyakin Institute of Bioorganic Chemistry RAS, Miklukho-Maklaya 16/10, 117997 Moscow, Russia
b Biotechnology Research Institute, Scott Laboratory Building, 19 Mount Street Mail, No. C-52, AUT University, Private Bag 92006, Auckland 1142, New Zealand
a r t i c l e i n f o
a b s t r a c t
Article history:
Synthetic glycolipid-like constructs dispersible in biological media and capable of incorporating into cell
membranes have the ability to create novel artificial glyco-landscapes on living cells. Using a variety of
different glycans ranging from disaccharides to polysaccharides, together with different lengths and high
hydrophilicity spacers, we created a series of synthetic glycolipid-like constructs. Contacting these con-
structs with live cells gave modified cells with controlled glycan density and/or altered biological func-
tion. The ability to also use these constructs as solutions to inhibit antibodies, toxins, and virions extends
the potential diagnostic and therapeutic uses for these synthetic glycolipid-like constructs.
Ó 2012 Elsevier Ltd. All rights reserved.
Received 31 January 2012
Received in revised form 30 March 2012
Accepted 31 March 2012
Available online 7 April 2012
Keywords:
Neoglycolipids
Glyco-landscape
Antibodies
Spacers
Glycosylation
Cell modification
1. Introduction
lation of glycoconjugates at the cell surface. This paper describes in
detail the synthesis of FSL constructs, including variations of spacer
Glycosphingolipids by nature of their lipid tail have a natural
ability to self-insert into cell membranes,1–3 but often the require-
ment to use solvents to maintain them in a solution-phase suitable
for insertion can limit their compatibility with live cells. Equally
important is the complex chemical steps needed to synthesize
glycosphingolipids bearing ceramide lipid tails. Recently synthetic
glycolipid-like constructs, which have a spacer between a synthet-
ically friendly lipid (as a rule, dioleoylphosphatidylethanolamine,
DOPE) and the glycan head have been constructed (Chart 1).4 These
constructs are now known as Function-Spacer-Lipid constructs
(FSLs) while construct modified cells are termed ‘kodecytes’.5–9
All FSLs have been designed to disperse as a solution directly in or-
ganic solvents free water, saline and biological media (one function
of the selected spacers).5 Like natural glycolipids,1–3 all FSLs have
the ability to self-insert into cell membranes.5 Because FSLs are
synthetically made, their structure can be fully controlled and
engineered and they can be designed to carry glycans not naturally
present as glycolipids, for example hyaluronic acid. Since their
introduction in 2003 a range of biological applications for carbohy-
drate based FSL constructs have been reported (Table 1). These
constructs can be attached to live cells in a controlled and precise
manner5–9 making them powerful tools for the study and manipu-
arms between glycan and lipid tails, and gives new evidence of bio-
logical activity/applications relating to these synthetic glycolipid-
like constructs.
2. Experimental procedures
2.1. Materials
2.1.1. Synthesis
1,2-O-Dioleoyl-sn-glycero-3-phosphatidylethanolamine (DOPE)
was from Northern Lipids Inc. (NLI, Vancouver, Canada). All CMG
(N-carboxymethyglycine) linkers, MCMG2, CMG2 and CMG4, were
prepared as described.15 Hyaluronic acids HA5kDa, and HA17kDa were
obtained from Lifecore Biomedical, USA, HA8kDa was from Sigma.
Immunological reagents were monoclonal anti-A (Epiclone, CSL,
Melbourne) and affinity purified polyclonal sheep anti-hyaluronic
acid (Biogenesis 5029-990, Poole UK). TLC chromatography was per-
formed using Silica Gel 60 F254 aluminum sheets with visualization
via either 7% H3PO4 or ninhydrin.
2.2. Synthesis of activated Spacer-Lipid units
Various activated spacers conjugated to DOPE were
synthesized.
⇑
Corresponding author. Tel.: +7 495 330 71 38; fax: +7 495 330 55 92.
E-mail addresses: stephen.henry@aut.ac.nz (S. Henry), bovin@carb.ibch.ru (N.
Bovin).
Structures of all three linkers are shown on Scheme 2.
0008-6215/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.carres.2012.03.044

E. Korchagina et al. / Carbohydrate Research 356 (2012) 238–246
239
OH
HO
HO
O
O
HO
OH
OH
O
HN
O
O
O
O
HO
HO
OH
OH
OH
OH
HO
HO
O
HO
OH
OH
O
O
O
O
OH
O
O
O
H
N
P
O
O
O
O
HO
HO
N
H
OH
OH
O
O
O
Chart 1. Natural glycosphingolipid versus its FSL synthetic analog. Schematic diagram contrasting: (upper image) a natural Gb3 globotriaosylceramide (lignoceric acid) with
(lower image) FSL-Gb3 with the same glycan conjugated to spacer-DOPE.
vacuum (solid foam). The obtained mixture was separated on silica
gel column (2.8 ꢁ 33 cm, ꢀ200 mL of silica gel in CHCl3/MeOH
5:1). The mixture was placed on column in CHCl3/MeOH (5:1)
and the components were eluted with CHCl3/MeOH (1:1) and then
with alteration of MeOH/CHCl3/water composition from 7:7:1 to
6:6:1. The first elute was DOPE-Ad-MCMG2-Ad-DOPE (Rf = 0.72,
MeOH/CHCl3/water 6:6:1) while the second was aimed MCMG2-
Ad-DOPE (Rf = 0.34, MeOH/CHCl3/water 6:6:1). MCMG2 was eluted
with MeOH/1 M PyꢂHOAc (3:2, Rf = 0.47). Fractions, containing
pure MCMG2-Ad-DOPE amine were combined, evaporated, and
dried in vacuum. The residue was dissolved in 5% AcOH and
freeze-dried. Yield of MCMG2-Ad-DOPE (Scheme S1, Supplemen-
tary data) was 132 mg (26% on MCMG2). MS, m/z: 1887 [M+H],
1909 [M+Na], 1925 [M+K] and 1H NMR according to Supplemen-
tary data.
Table 1
Published biological applications of FSL constructs
Function-Spacer-Lipida construct
Biomedical application(s)
Functional group
GalNAc 1-3
Galb
Spacer
Ad
a
ABO quality control4–6
Flow cytometry controls7
Fuc
a
1-2
Mimicking ABO transfusion reactions8,9
In vivo ABO antibody neutralization9
Gal
a
1-3
Ad
ABO quality control4–6
Galb
Flow cytometry controls7
Fuc
a1-2
Gal
Gal
Gal
a
a
a
1-4Galb1-4Glcb
1-4Galb1-4Glcb
1-3
Ad
Ad
Ad
HIV virus inhibition with Gb310
Toxin neutralization with Gb310
Lewis antibody identification4
Galb
Fuca1-4
Galb1-3
Ad
Acquired-B determination4
2.2.3. CMG2-Ad-DOPE
Galb
Fuc
Gal
a
1-2
1-3Galß1-
To an intensively stirred solution of CMG2 (425 mg, 0.44 mM of
internal salt) in isopropyl alcohol (IPA)/water mixture (IPA/water
3:2, 10 mL) were added the 1 M aq solution of NaHCO3 (0.44 mL,
0.44 mM) and then the solution of DOPE-Ad-ONSu (211 mg,
0.22 mM) in dichloroethane (0.4 mL). The reaction mixture was
stirred for 2 h, acidified with 0.2 mL of AcOH and evaporated in va-
cuo to minimal volume at 35 °C. The solid residue was dried in vac-
uum (solid foam) and then thoroughly extracted with CHCl3/MeOH
mixture (CHCl3/MeOH 4:1, several times with 10 mL, TLC control).
The extracted residue consisted of unreacted CMG2 and salts (about
50% of CMG2 was recovered by desalting of combined residue and
fractions after chromatography on silica gel according to the proce-
dure described in the CMG2 synthesis.). The combined CHCl3/MeOH
extracts (solution of CMG2-Ad-DOPE, DOPE-Ad-CMG2-Ad-DOPE, N-
oxysuccinimide and some CMG2 were evaporated in vacuum and
dried. The obtained mixture was separated on silica gel column
(2.8 ꢁ 33 cm, ꢀ200 mL of silica gel in CHCl3/MeOH 5:1). The mix-
ture was placed on column in MeOH/CHCl3/water mixture
(MeOH/CHCl3/water 6:3:1 + 0.5% of pyridine) and the components
were eluted in a stepwise ternary gradient: MeOH/CHCl3/water
composition from 6:3:1 to 6:2:1 and then to 6:2:2 (all with 0.5%
of pyridine). DOPE-Ad-CMG2-Ad-DOPE was eluted first (Rf = 0.75,
MeOH/CHCl3/water 3:1:1), followed by desired DOPE-Ad-CMG2
(Rf = 0.63, MeOH/CHCl3/water 3:1:1), last eluted was CMG2
(Rf = 0.31, MeOH/CHCl3/water 3:1:1). Fractions, containing pure
CMG2-Ad-DOPE amine were combined and evaporated to dryness.
a
Ad
Galili xenotransplant mimics11
4GlcNAcß
Biotin
Fluorophores
Peptides
CMG2
Biotinylation of cells5,8,9 and virions5,12
Various Labeling of cells4,5 and virions5,12
CMG2
Peptide modification of cells13,14
a
All constructs have DOPE as their lipid tail.
2.2.1. DOPE-Ad-ONSu
To a solution of bis(N-hydroxysuccinimidyl) adipate (700 mg,
2.05 mM) in dry N,N-dimethylformamide (30 mL) were added
DOPE (300 mg, 0.403 mM) in chloroform (15 mL) and triethylamine
(50 lL) (Scheme 1). The mixture was stirred for 2 h at room temper-
ature, then neutralized with acetic acid and concentrated in vacuo.
Column chromatography (Sephadex LH-20, 1:1 chloroform–metha-
nol, 0.2% acetic acid) of the residue yielded the activated lipid
(370 mg, 95%) as a colorless syrup; TLC (chloroform–methanol–
water, 6:3:0.5): Rf = 0.5. 1H NMR according to Supplementary data.
2.2.2. MCMG2-Ad-DOPE
To an intensively stirred solution of MCMG2 (378 mg of wet
material with ꢀ10% of water, ꢀ0.27 mM) in pyridine (5 mL) were
added Et3N (0.1 mL, 0.72 mM) and a solution of DOPE-Ad-ONSu
(174 mg, 0.18 mM) in dichloroethane (0.87 mL). The reaction mix-
ture was stirred for 2 h and then acidified with 0.2 mL of AcOH and
evaporated to minimal volume at 35 °C. The residue was dried in

240
E. Korchagina et al. / Carbohydrate Research 356 (2012) 238–246
O
O
O
O
N
CH3
CH3
O
O
O
O
H2N
+
O
N
O
O
O
O
P
O
OH
O
Ad(ONSu)2
DOPE
OH
HO
O
O
OH
O
O
HO
N
O
HO
O
O
HN
CH3
CH3
O
O
NH2
O
O
O
N
H
H3C
O
+
O
O
O
O
H3C
O
OH
P
OH
OH
O
HO
DOPE-Ad-ONSu
Atri-sp
OH
HO
O
O
OH
HO
O
HO
O
H
N
CH3
CH3
HN
O
O
O
O
O
N
H
H3C
O
O
O
O
O
H3C
O
P
OH
OH
O
OH
HO
FSL-Atri
Scheme 1. Synthesis of DOPE-Ad-ONSu and FSL-Atri
.
OH
OMe
O
OMe
O
O
HO
HO
H3C
O
O
O
O
O
O
O
O
O
H
N
H
N
H
N
H
N
CH3
CH3
H
N
OH
O
O
NO2
N
N
N
N
N
H
O
N
N
H
N
H
N
HO
H2N
MCMG2-Ad-DOPE
NH
H
N
H
H
O
O
O
O
OH
O
O
O
O
O
O
O
O
O
O
O
P
O
OMe
O
OMe
O
O
O
H3C
OH
Atri-sp-Ad-ONp
OH
Atri-sp-Ad-ONp
HO
OH
HO
O
OMe
OMe
O
OH
HO
O
O
O
HO
O
O
O
O
O
O
N
O
O
O
H
N
NH
H
N
H
N
H
N
H
N
H
N
CH3
CH3
H3C
O
O
O
O
N
N
O
N
Atri-sp-Ad-MCMG2-Ad-DOPE
N
H
N
H
N
H
N
H
N
N
H
O
H
O
O
O
O
O
O
O
O
OH
O
O
O
O
O
P
O
H3C
OMe
OH
OMe
OH
HO
OH
OH
O
O
O
O
O
O
O
N
O
O
O
O
H
N
H
N
H
N
H
N
H
N
CH3
CH3
O
O
N
N
N
N
H
N
H
N
H
N
H
N
H
H2N
CMG2-Ad-DOPE
O
O
O
O
O
OH
O
O
O
O
O
O
O
P
OH
OH
Atri-sp-Ad-ONp
OH
HO
O
OH
O
OH
OH
HO
O
O
HO
O
O
O
O
N
O
O
O
O
NH
H
N
O
H
N
H
N
H
N
H
N
H
N
H3C
O
CH3
CH3
O
O
O
O
Atri-sp-Ad-CMG2-Ad-DOPE
N
N
N
N
H
N
H
N
H
N
H
N
H
N
H
O
O
O
O
O
O
O
O
O
O
O
O
O
O
OH
O
P
O
H3C
OH
OH
OH
OH
HO
OH
OH
OH
OH
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
H
N
H
N
O
O
H
N
H
N
H
N
H
N
H
N
H
N
H
CH3
CH3
O
O
N
N
N
N
N
N
CMG4-Ad-DOPE
N
N
N
N
H
N
N
H
N
H
N
H
H2N
N
H
N
H
N
H
N
H
H
O
O
O
O
O
O
O
O
O
O
OH
O
O
O
O
O
O
O
O
O
O
P
OH
OH
OH
OH
Atri-sp-Ad-ONp
OH
HO
HO
H3C
Atri-sp-Ad-CMG4-Ad-DOPE
O
OH
OH
O
OH
OH
OH
O
O
HO
O
NH
O
O
O
H
N
O
O
O
O
O
O
O
O
O
O
O
H
N
H
N
O
O
O
O
H
N
H
N
H
N
H
N
H
N
H
N
H
N
CH3
CH3
O
O
O
N
N
N
N
N
N
H
N
N
H
N
N
H
N
N
H
N
H
N
H
N
H
N
H
N
H
N
H
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
OH
O
O
O
O
P
O
H3C
OH
OH
OH
OH
OH
OH
HO
Scheme 2. Synthesis of CMG variations of FSL-A.
To remove any low molecular weight impurities and solubilized sil-
ica gel the residue was dissolved in IPA/water 1:2 mixture (2 mL),
and was passed through a Sephadex LH-20 column (column volume
130 mL, eluent—IPA/water 1:2 + 0.25% of pyridine). Fractions con-
taining pure CMG2-Ad-DOPE were combined and evaporated
(ꢀ20% of IPA was added to prevent foaming) to dryness, the residue
was dissolved in water (ꢀ4 mL) and freeze-dried. Yield of CMG2-
Ad-DOPE (Scheme S1, Supplementary data) was 270 mg (68% on
DOPE-Ad-ONSu or 34% on CMG2). MS, m/z 1831 [M+H] and 1H
NMR according to Supplementary data.
2.2.4. CMG4-Ad-DOPE
To an intensively stirred solution of CMG4 (340 mg, 0.14 mM of
hexa-Et3N salt) in IPA/water mixture (IPA/water 3:2, 7 mL) were
added 1 M aq solution of NaHCO3 (0.23 mL, 0.23 mM) followed by
the solution of DOPE-Ad-ONSu (110 mg, 0.11 mM) in dichloroethane

E. Korchagina et al. / Carbohydrate Research 356 (2012) 238–246
241
OH
OH
HO
OH
O
COOH
O
HO
O
H
O
O
CH3
CH3
HO
AcHN
O
N
O
O
O
O
N
H
O
HO
NHAc
OH
HO
O
O
O
O
P
OH
FSL-3'SLN
OH
HO
COOH
HO
AcHN
O
O
HO
OH
O
O
O
HO
H
N
O
O
CH3
CH3
O
O
O
O
HO
N
H
HO
NHAc
OH
O
O
O
O
P
OH
FSL-6'SLN
Chart 2. Structures of two sialyl FSL constructs.
(0.55 mL). The reaction mixture was stirred for 2 h, acidified with
0.2 mL of AcOH and evaporated to minimal volume at 35 °C. The
solid residue was dried in vacuum (solid foam). The obtained mix-
ture was separated on silica gel column (2.6 ꢁ 25 cm, ꢀ130 mL of
silica gel in IPA/water 5:1). The mixture was placed on column in
IPA/water mixture (3:1) and the components were eluted with
alteration of IPA/water composition from 3:1 to 2:1 and then to
3:2 (all with 0.5% of pyridine). First eluted was DOPE-Ad-CMG4-
Ad-DOPE (Rf = 0.61, MeOH/CHCl3/water 3:1:1), while the second
elute was aimed at CMG4-Ad-DOPE (Rf = 0.38, MeOH/CHCl3/water
3:1:1), last eluted was CMG4 (Rf = 0.12, MeOH/CHCl3/water
3:1:1). Fractions, containing pure CMG4-Ad-DOPE amine were
combined and evaporated to dryness and purified as above. Yield
of CMG4-Ad-DOPE (Scheme S1, Supplementary data) was 108 mg
(29% on CMG4). MS, m/z 2748 [M+H] and 1H NMR according to
Supplementary data.
excess was needed). TLC: Rf = 0.48 (MeOH/CHCl3/water 6:6:1).
MS, m/z 2606 [M+Na], 2622 [M+K and 1H NMR according to Sup-
plementary data.
2.3.4. FSL-Atri-CMG2
Solvent for conjugation—pyridine/water (4:1), 5 equiv of Et3N
as base. Yield 78% on CMG2-Ad-DOPE or 39% on A trisaccharide gly-
coside (100% excess was needed). TLC: Rf = 0.28 (IPA/acetonitrile/
water 4:3:2). MS, m/z 2551 [M+Na], 2567 [M+K] and 1H NMR
according to Supplementary data.
2.3.5. FSL-Atri-CMG4
Solvent for conjugation—pyridine/water (2:1), 5 equiv of aq 1 M
NaHCO3 as base. Yield 96% on CMG4-Ad-DOPE or 32% on A trisac-
charide glycoside (200% excess was needed). TLC: Rf = 0.49
(MeOH/CHCl3/water 3:1:1). MS, m/z 3467 [M+Na], 3483 [M+K]
and 1H NMR according to Supplementary data.
2.3. Synthesis of blood group related FSLs
2.4. Synthesis of sialic acid based FSLs
Although an almost unlimited number of glycans can be created
as FSL constructs, for the sake of brevity and comparative discus-
a
Derivatives of 60SLN (Neu5Ac 2-6Galb1-4GlcNAc), and 30SLN
sion, the blood group A trisaccharide, GalNAc
a
1-3(Fuca1-2)Galß,
(Neu5Ac 2-3Galb1-4GlcNAc) trisaccharides (Chart 2), with adipic
a
is described here as the benchmark construct. Other related blood
linker, were synthesized similarly to A trisaccharide construct
group FSL constructs are shown in Chart S1 Supplementary data.
(Scheme 1). NMR data are specified in Supplementary data.
2.3.1. Preparation of FSL-Atri (with adipic (Ad) linker)
2.5. Synthesis of FSL-HA (hyaluronic acid)
To a solution of aminopropyl glycoside of A trisaccharide
(176 mg, 0.30 mM) in dry N,N-dimethylformamide (3 mL) were
added 50 lL of triethylamine and the solution of DOPE-Ad-ONSu
(320 mg, 0.33 mM) in dichloromethane (1 mL) (Scheme 1). The
mixture was stirred for 2 h at room temperature. Column chroma-
tography (the first on Sephadex LH-20, 1:1 chloroform–methanol,
and the second on silica gel, dichloromethane–ethanol–water,
6:5:1) of the mixture yielded 390 mg (90%) of the synthetic mole-
cules designated FSL-Atri. MS, m/z 1442 [M+H] and 1H NMR accord-
ing to Supplementary data.
A heterogeneous range of hyaluronic acid (HA) fragments were
prepared from chromatographic purified hyaluronidase digested
hyaluronic acid16 and used to manufacture FSL-HA constructs
(Chart 3). HA oligomers were converted to HA-glycamines by
reductive amination. Acylation of HA-glycamines with N-oxysuc-
cinimide ester DOPE-Ad-ONSu and subsequent purification re-
sulted in FSL-HA-Ad-DOPE (Scheme 3).
2.5.1. Fsl-HA8kda
Hyaluronic acid oligomer HA ꢀ8 kDa (36 mg) was dissolved in
5 M NH4OAc (3.6 mL), and the solution was kept for 21 h at
40 °C. After addition of aqueous 2 M NaCNBH3 in five sequential
2.3.2. Preparation of FSL-Atri-MCMG2, FSL-Atri-CMG2 and FSL-
A
tri-CMG4 (Scheme 2)
FSL-A-trisaccharide constructs based on a variety of carbo-
portions the mixture was kept at 40 °C (40
lL of 2 M NaCNBH3—
yxymethylglycine spacers were synthesized by the conjugation of
GalNAc 1-3(Fuc 1-2)Galb-O(CH2)3NHCO(CH2)4COO(p-C6H4)NO2
(Atri-sp-Ad-ONp) glycoside with corresponding amine (MCMG2-
Ad-DOPE, CMG2-Ad-DOPE or CMG4-Ad-DOPE) (Scheme 2). The
desired pure products were isolated by gel-permeation chromatog-
raphy (Sephadex LH-20, IPA/water 1:2 + 0.25% pyridine).
for 3 h, 80 L—18 h, 160 L—8 h, 160 L—21 h, 160 l
l
l
l
L—21 h).
a
a
Desalting of the reaction mixture by gel-permeation chromatogra-
phy on Sephadex G-10 column (1.8 ꢁ 40 cm, eluent—aqueous
0.1 M PyꢂAcOH) and freeze-drying gave 32.8 mg of HA-glycamine
(with admixture having no amine group HA oligomer). TLC (eluent
IPA/MeOH/MeCN/water 4:3:6:4): HA-glycamine (ninhydrine-posi-
tive) Rf = 0.2; HA oligomer Rf = 0.31. HA-glycamine (32.8 mg) was
dissolved in the mixture of IPA (1.5 mL) and water (1.5 mL). To
the rapidly stirred solution of HA-glycamine the solution of N-oxy-
2.3.3. FSL-Atri-MCMG2
DMF was used as solvent for conjugation (+4 equiv of Et3N).
Yield 85% on MCMG2-Ad-DOPE or 65% on A trisaccharide (30%
succinimide ester DOPE-Ad-ONSu (34 mg, 35 lM) in CH2Cl2

242
E. Korchagina et al. / Carbohydrate Research 356 (2012) 238–246
A -Ad-DOPE
tri
OH
O
H
OH
O
O
O
H
O
H O
H
N
O
AcHN
O
O
O
O
N
H
O
O
O
O
O
OH
O
H 3
C
O
P
OH
OH
H O
OH
OH
H
OH
C H 3
O
O
OH
H N
OH
OH
O
H O
OH
C H 3
O
O
OH
OH
O
O
O
O
O
O
O
O
H O
O
O
O
O
O
O
O
O
O
H O
N H
O
O
H
O
O
O
O
OH
OH O
H
O
O
O
H O
H N
H
H O
N
O
O
OH
OH
H
O
OH
O
O
C H
H 3
C H 3
OH
O
N
H O
N
H O
O
O
C H3
H O
O
C H 3
H O
H
H
O
OH
O
H
O
O
O
H O
O
H O
H O
H O
H O
H O
H
O
OH
O
O
OH
O
O
OH
O
H O
O
OH
O
O
H O
O
OH
O
O
OH
H O
O
N
H
O
O
O
O
O
O
N
HA20mer-DOPE
O
O
N
H3
C
H O
H O
O
N
H O
O
O
H
O
O
H
N
H
O
O
O
O
OH
O
H
H
H
N
O
O
H O
H
H
H 3
C
H 3
C
H
O
C H 3
H 3
C
C H3
O
OH
C H 3
O
N
H O
OH
OH
H
O
H O
H O
O
H O
H
O
O
O
OH
OH
O
H O
O
O
O
OH
H
O
O
O
O
H O
O
H O
H O
H
O
O
H O
OH
O
O
OH
H
O
O
H O
O
O
O
OH
OH
O
H
N
H O
O
O
O
O
OH
OH
OH
H O
H
N
O
O
OH
OH O
OH
OH
OH O
OH
N
O
O
OH
O
O
O
OH
OH
H O
OH
O
O
O
O
O
O
H O
C H 2
O
N
H
P
O
H
O
N
H
O
O
H O
H 3
C
O
O
O
N
H
O
O
O
O
OH
H O
N H
H 3 C
O
H3 C
N H
C H 3
H O
H3
C
N H
O
C H 3
O
O
C H3
Chart 3. Scale comparison of FSL-A with FSL-HA. FSL-HA is 40 oligosaccharide (ꢀ8 kDa), whereas FSL-Atri is a trisaccharide.
OH
OH
OH
O
OH
OH
O
O
OH
O
NH4OAc,
NaCNBH3
HO
OH
OH
O
O
OH
O
OH
NH
HO
O
O
O
O
HO
NH
O
O
NH2
HO
HO
HO
DOPE-Ad-NOSu
m
HO
O
O
NH
CH3
O
CH2
OH
O
HO
HO
HO
m
O
CH3
OH
NH
OH
HA glycamine
O
OH
HA
O
CH3
n = m+1
n-mer
O
CH3
OH
OH
O
OH
O
O
O
OH
O
OH
O
OH
NH
P
O
O
H
HO
O
O
N
HO
O
N
H
O
O
O
CH2
O
HO
HO
HO
m
O
NH
OH
OH
O
O
CH3
HA
-DOPE
O
CH3
n-mer
n = m+1
Scheme 3. Synthesis of HA-Ad-DOPE.
(0.2 mL) and then aqueous 1 M Na2CO3 in two portions (85 and
45 L) with 45 min interval were added. The mixture was stirred
for 45 min then acidified with AcOH (30 L). Gel-permeation chro-
biotin N-oxysuccinimide ester (75.1 mg, 0.22 mM) in DMF
(1.5 mL). The reaction mixture was stirred for 2.5 h and then acid-
ified with 0.15 mL of AcOH and evaporated to minimal volume
(beginning of a precipitate formation) at 35 °C. The residue was di-
luted with water (1.5 mL) and 0.05 mL of pyridine was added. The
resulting solution was passed through Sephadex LH-20 column
(column volume 300 mL, eluent—IPA/water 1:2 + 0.5% of pyridine
and 0.25% of AcOH). Fractions, containing pure FSL-biotin-CMG2
were combined and evaporated (ꢀ20% of IPA was added to prevent
foaming) to dryness, the residue was dissolved in water (ꢀ4 mL,
containing 0.05 mL of pyridine) and freeze-dried. Yield of FSL-bio-
tin-CMG2 (pyridinium salt) was 384.5 mg (ꢀ0.18 mM). The prod-
uct was dissolved in water (3 mL) containing NaHCO3 (0.18 mL of
1 M aqueous solution, 0.18 mM) and freeze-dried. Yield of FSL-bio-
tin-CMG2 (sodium salt) was 374.3 mg (90% on CMG2-Ad-DOPE)—
Scheme 4. TLC: Rf = 0.64 (MeOH/CHCl3/water 6:2:1. MS, m/z:
2079 [M+Na], 2095 [M+K], 2101 [MNa+Na] and NMR data are as
specified in Supplementary data.
l
l
matography of reaction mixture on Sephadex LH-20 column
(1.8 ꢁ 35 cm, eluent—MeOH/water 2:1, 0.03 M PyꢂAcOH) gave
42.5 mg of FSL-HA8kDa (with admixture of HA oligomer).
A solution of the crude product in water was slowly loaded onto
a C18 reverse phase column (1.2 ꢁ 7 cm, water). Elution with water
and water/MeOH 10:1 gave no lipid HA oligomer (12.3 mg). Elution
with water/MeOH 1:3 and then with water/MeOH/CHCl3 5:15:1
gave FSL-HA8kDa with little admixture of eluted C18 phase. This
product was evaporated and the residue (thin film on the flask
walls) was extracted with hexane (2 ꢁ 2 mL) followed by ether
(2 ꢁ 2 mL), then dissolved in water (1.5 mL) and freeze-dried. Yield
of pure FSL-HA8kDa was 20.6 mg (ꢀ50%). TLC: Rf = 0.33, eluent IPA/
MeOH/MeCN/water 4:3:6:4. NMR data are as specified in Supple-
mentary data.
For stability and storage FSL-HA8kDa was converted to mono-Na
salt (by phosphate) and (NH4)n-salt (by glucuronic acids). FSL-HA
was dissolved in water and a calculated quantity of aqueous NaH-
CO3 and NH3 were added followed by freeze-drying.
2.7. Preparation of biotinylated blood group A trisaccharides
By the analogous procedure but starting from commercially
available fractions of hyaluronic acid (Lifecore Biomedical, USA)
FSL-HA5kDa (yield 49%) and FSL-HA17kDa (yield 66%) were
synthesized.
Two mono-biotinylated forms of blood group A trisaccharides,
one, Atri-sp-AC-biotin, with a short 3-carbon spacer and aminoca-
proic unit (AC), and the other on long polyacrylamide spacer,
A
tri-PAA-biotin (MW ꢀ30 kDa), were prepared as described.17,18
2.6. Synthesis of FSL-biotin-CMG2
2.8. FSL membrane insertion creating kodecytes
To an intensively stirred solution of CMG2-Ad-DOPE (366 mg,
0.20 mM) in IPA/water mixture (IPA/water 1:1, 15 mL) were added
1 M aq solution of NaHCO3 (1.2 mL, 1.20 mM) and a solution of
The generic established method for inserting FSLs into cell mem-
branes was used to create kodecytes. This involves simply contact-
ing one volume of cells (either 100% packed or in solution) with one

E. Korchagina et al. / Carbohydrate Research 356 (2012) 238–246
243
OH
OH
O
O
O
O
O
O
O
H
O
O
O
O
H
H
N
H
CH3
CH3
H
O
O
N
N
N
N
N
N
N
N
H
N
H
N
N
N
H
N
H
H2N
H
O
O
O
O
O
OH
O
O
O
O
O
O
O
P
OH
OH
O
HN NH
O
H
O
H
O N
O
S
O
OH
OH
HN NH
O
O
O
O
H
O
O
H
O
O
O
O
H
O
O
H
H
N
H
CH3
CH3
H
O
O
N
N
N
N
N
N
N
N
H
N
H
N
N
N
H
S
N
N
H
H
O
O
O
O
O
OH
H
O
O
O
O
O
O
O
P
OH
OH
Scheme 4. Synthesis of FSL-biotin.
volume of FSL solution (usual range 10–1000 lg/mL of FSL in phos-
3. Results and discussion
3.1. FSL synthesis
phate buffer saline (PBS)) and incubation at 37 °C for 1–2 h.4–9 Still
based primarily on creating kodecytes but this time glycoconjugates
were attached through an alternative ‘sandwich’ technique. First
FSL-biotin kodecytes were obtained with a solution of 100 lg/mL
A variety of different and successful approaches were taken to
build a range of FSL constructs bearing glycans (Schemes 1–3).
The design of all FSL constructs was deliberately constrained by
an absolute requirement to be dispersible in water (without organ-
ic co-solvents); a feature, which it is believed will favor biocompat-
ibility and potential therapeutic use. Constructs were further
constrained by a requirement that they are able to spontaneously
and stably incorporate into cell membranes and that the spacer
be inert with serum. Enabling all of these features required the
designing of a series of different spacers, in particular CMG varia-
tions, and abandoning alternatives such as ethylene glycol, which
although able to bring about most features, was not able to achieve
inertness with undiluted serum particularly, binding of PEG with
serum antibodies of numerous healthy donors.19 The diacyl nature
of the lipid tail was also important for normal cell retention as
exemplified by a construct with a single lipid chain (octadecanoic
acid) which although dispersible in saline did not efficiently insert
and/or be retained in cell membranes (results not shown). Notably,
all synthesized glyco-FSL constructs were water dispersible and
capable of inserting in cell membrane quantitatively4 when con-
structed in the simplest version, with –OCH2CH2CH2NHC-
OCH2CH2CH2CH2CO– spacer (Scheme 1) between glycan and
DOPE fragments. Insertion of more sophisticated CMG or MCMG
spacers were also utilized to make glycan constructs (Scheme 2).
These spacers though not required for glycans (although they did
bring about increased sensitivity) were required in the case of
hydrophobic F groups, as exemplified here by biotin derivatives
(Scheme 4). The data presented here are primarily with a 1,2-O-
dioleoyl-sn-glycero-3-phosphatidylethanolamine (DOPE) as the
FSL lipid tail, however other lipids such as sterol instead of an
oleoyl moiety, can cause cell labeling, albeit much less efficiently
(not shown). The ability to change the lipid tail of FSLs, for instance
to ceramide, is expected to alter the function of some FSLs and
potentially determine where they may partition within the cell
membrane.20 Multiple DOPE tails on the polymeric glycoconju-
gates (based on polyacrylamide backbones) give rise to insertion
complications due to intramolecular assembling of the tails,21 so
only monomeric DOPE constructs were used here. As the lipid
DOPE was compatible with all current biological assays (Table 1)
it was chosen as the default lipid for FSL synthesis. Stability trials
have established that FSL constructs are stable for periods exceed-
of FSL-biotin. The resultant biotin kodecytes were then reacted with
an excess of avidin (1 mg/mL) and washed to give rise to avidin
kodecytes. Equal volumes of avidin kodecytes were then incubated
with an excess of biotinylated glycan at room temperature for
30 min. The resultant A kodecytes, with either Atri-sp-AC-biotin or
A
tri-PAA-biotin attached via an avidin bridge to the FSL-biotin kode-
cytes were then suspended in PBS and used in serologic assays.
2.9. Red cell serology
The detection of A antigen on the kodecytes was via direct sero-
logical agglutination against IgM monoclonal anti-A.4 The detec-
tion of hyaluronic acid by indirect (anti-IgG bridging)
agglutination of HA-kodecytes with polyconal anti-HA.
2.10. Influenza virus binding assay
Human red blood cells (native) were first desialylated with sial-
idase (neuraminidase from Vibrio cholera, Sigma), 0.1 UN for 1 mL
packed red cells in acetate buffer solution (pH 5.0), containing
0.15 M sodium chloride and 4 mM calcium chloride; 1 h, 37 °C.
Desialylated red blood cells were then inserted with FSL-sialyl con-
structs (1 mg/mL—2 h 37 °C) to produce 60SLN kodecytes and 30SLN
kodecytes. The capability of sialyl kodecytes to bind three different
strains of influenza viruses (avian H5N2/Mallard/10218, swine
H9N2/9/98, and human B/HK/54800) was determined in a hemag-
glutination reaction assay (50
wells microtiter round bottom plate, 50
l
L of double dilutions of virus in 96-
L of 0.5% cell suspension,
l
40 min at 4 °C). Native red blood cells were used as a positive con-
trol while the same desialylated cells, but FSL unmodified, were
tested in parallel as a negative control.
2.11. Supporting information available
NMR data of a number of synthesized compounds; Scheme S1.
Synthesis of DOPE with MCMG2, CMG2 and CMG4 Spacers;
Chart S1: Other Blood Group Related FSL Constructs; Chart S2.
Spacer Defined Distances Between Cell Membrane and Ligand. This
material is available free of charge via the Internet at http://
elsevier.com.

244
E. Korchagina et al. / Carbohydrate Research 356 (2012) 238–246
Table 2
Serologic reactions of A kodecytes created with biotinylated A trisaccharide constructs
Biotinylated saccharidea
Anti-A serology
A kodecytesd
Description
A trisaccharide valencyb
Spacer length nmc
Atri-sp-AC-biotin
1
ꢀ20
ꢀ1.4
ꢀ10
++
++++
A
tri-PAA-biotin
a
b
c
Two forms of biotinylated saccharide constructs of differing spacer length and valency loaded onto avidin kodecytes to create A kodecytes.
Number of A trisaccharide residues on each biotinylated construct.
Total length of the construct (taking into consideration coil conformation of PAA and stretched one for sp-AC moiety) excluding avidin and FSL-biotin.
Serologic scoring as per Table 4.
d
Table 3
Sialyl kodecyte differentiation of human, swine, and avian influenza viruses
Test cells
Agglutinationa of kodecytes with viruses
Avian H5N2/Mallard/10218
Human B/HK/54800
Swine H9N2/9/98
60SLN kodecytesb,c
30SLN kodecytesc
Desialylated red cellsd
Native red cellse
++++ (1:80)
—
—
—
++++ (1:25600)
++++ (1:1600)
—
++++ (1:800)
—
++++ (1:1600)
++++ (1:640)
++++ (1:25600)
a
b
c
Serologic scores as reported in Table 4.
Red cells were de-sialylated and then converted into sialyl kodecytes with specific sialic acid bearing FSLs.
FSL sialyl constructs used to make the kodecytes are as shown in Chart 2.
The same desialylated cell as used to make the kodecytes as a negative control.
Untreated red cells as a positive control (also the same cell as used to make the kodecytes).
d
e
ing 2 years when dry or at least 1 year in sterile saline solutions at
4 °C (or frozen).
from birds bind more strongly to sialic acid in a 2–3 linkage
(e.g., 30-sialyl-N-acetyllactosamine, 30SLN).23 Native human red
cells react with all influenza species, but when desialylated they
become unreactive (Table 3). Re-introduction of specific sialogly-
cans with FSL constructs to desialylated red cells (Chart 2) creates
kodecytes able to differentiate the virus species by their specific
sialic acid residues. The 60SLN kodecytes were reactive with only
human and swine viruses while the 30SLN kodecytes reacted only
with avian and swine viruses (Table 3). Because the same desialy-
lated cell is used to create both kodecytes and the negative control,
reactivity can be unambiguously assigned to reactivity against the
specific sialyl-FSL construct.
3.2. Biological analysis
In addition to the already published applications for FSL con-
structs (Table 1) we extended the range of applications and oppor-
tunities by creating three new series of constructs; one utilizing
FSL-biotin + avidin + biotinylated saccharides, another based on
sialic acid residues and another utilizing polysaccharide fragments
of hyaluronic acid.
3.2.1. Biotin + avidin + biotinylated glycan composite
By synthesising FSL biotin, a construct successfully used in
other applications (Table 1), an alternative approach to glycosylat-
ing the surface of cells was enabled. In this approach a three-stage
method was used in attaching biotinylated glycans via avidin to
FSL-biotin kodecytes. We evaluated the serological reactions of
two forms of biotinylated blood group A trisaccharides attached
to cells via FSL-biotin + avidin and found them to be clearly recog-
nized by anti-glycan antibodies (Table 2). Increasing the number of
glycosylation points and consequent length of the spacer when
3.2.3. Hyaluronic acid FSLs
Hyaluronic acid (HA) is large polysaccharide and relatively large
fragments of it can be synthesized into FSL-HA constructs (Chart
3). Using specific antibodies and HA binding proteins it could be
established that FSL-HA inserted into the membranes of murine
embryos and red cells. Solutions of FSL-HA constructs over the
range of 10–250 lg/mL produced maximally strong serological re-
sults. HA has been claimed in humans24 (and mice25) to improve
embryo implantation. However when we used either free HA, or
a range of FSL-HA constructs, we were unable to improve subfertile
murine implantation rates, despite demonstrating increased
in vitro adherence of HA embryo kodecytes to endometrial cells
(unpublished). Although FSL-HA did not affect fertility it remains
a potentially promising construct for the modification of cells or
liposomes.26
A
tri-sp-AC-biotin is compared with polymeric end-monobiotinylat-
ed conjugate Atri-PAA-biotin,18 resulted in increased sensitivity of
the assay. Although compared to the single FSL construct route this
method is somewhat cumbersome, it does offer a rapid and flexible
alternative approach to ‘glycosylating’ the surface of cells. How-
ever, although glycosylation of the cell clearly occurs, the [biotinyl-
ated-spacer-glycan] + [avidin] + [FSL-biotin] glyco-complex is
significantly unnatural, and may not necessarily mimic a natural
reaction, or that of a single FSL construct.
3.3. Spacer length
The inclusion of a spacer is integral to the construction of the
FSL and brings several important features to the construct, includ-
ing water dispersibility, and spacing of the glycotope away from
the membrane. It had previously not been established what influ-
ence the length of the spacer would have on serological reactions.
To investigate this we built and tested the serology of the blood
3.2.2. Sialic acid FSLs
The sialic acid-containing glycans are expressed on host respi-
ratory epithelial cell surfaces and are involved in influenza virus
infection.22 It is known that viruses isolated from human and
swine sources prefer binding to sialic acid in a 2–6 linkage (e.g.,
60-sialyl-N-acetyllactosamine, 60SLN), whereas viruses isolated
group
A trisaccharide on three spacers of increasing length

E. Korchagina et al. / Carbohydrate Research 356 (2012) 238–246
245
Table 4
Serologic reactions of blood group A kodecytes created with different FSL spacers
a
FSL-Atri
Spacersc
Ad
FSL-A kodecyte serologic reactionsb against dilutions of monoclonal anti-A
l
M
1
2
4
8
16
32
64
128
++
256
512
1024
50
10
5
++++
+++
++
++++
+++
++
++++
+++
+
++++
++
++++
++
—
++++
++
—
+++
++
—
++
—
—
—
—
—
—
—
—
CMG2
CMG4
50
10
5
++++
++++
+++
++++
++++
+++
++++
++++
+++
++++
+++
+++
++++
+++
++
++++
+++
++
+++
+++
++
+++
++
+
++
+
+
—
—
—
+
50
10
5
++++
++++
+++
++++
++++
+++
++++
++++
+++
++++
+++
+++
++++
+++
++
++++
+++
++
+++
+++
++
+++
++
+
++
++
—
—
—
a
All FSL constructs have A trisaccharide and DOPE as the lipid tail.
Serologic reactions observed against dilutions of monoclonal anti-A in gel reaction cards and scored as ++++ (maximal agglutination), +++ (strong), ++ (medium), + (weak),
b
(equivocal) and — (unreactive).
c
Spacers are as described in Chart S2.
(Scheme 2 and Table 4). It was found that at the 5
l
M level blood
grant 10-04-01693a. The authors thank D Blake, D Heathcote & E
Eriksson for unpublished data.
group A trisaccharide with the 1.9 nm short-spacer showed med-
ium serological reactivity with monoclonal anti-A when diluted
1:2, while with the longer 7.2 nm spacing of FSL-Atri-CMG2 gave
equivalent reactivity at an antibody dilution of 1:64. Further
increasing the length of the spacer to 11.5 nm (e.g., Atri-CMG4)
did not improve serological reactivity (Table 4), and is interpreted
that a spacer of >7 nm was adequate to induce maximal serological
reactivity at the red cell membrane. Neutralizing the charge on the
spacer by per-methylation of –COOH groups did not influence the
serological results.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.carres.2012.
03.044.
References
1. Sneath, J.; Sneath, P. Nature 1955, 176, 172.
2. Marcus, D.; Cass, L. Science 1969, 164, 553–554.
3. Renton, P.; Hancock, J. Vox Sang. 1962, 7, 33–38.
4. Final comments
4. Frame, T.; Carroll, T.; Korchagina, E.; Bovin, N.; Henry, S. Transfusion 2007, 47,
876–882.
5. Blake, D. A.; Bovin, N. V.; Bess, D.; Henry, S. M. J. Vis. Exp. 2011, 54, e3289.
doi:http://dx.doi.org/10.3791/3289. (http://www.jove.com/details.php?id=
3289).
The ability to modify the glyco-landscape of living cells is ex-
pected to be a useful tool in understanding the involvement of spe-
cific glycoconjugates in biological settings.27 The glyco-landscape
of cells can certainly be modified by the direct use of glycosidases28
and glycosyltransferases29 but the actual ability to control the gly-
co-landscape by these methods is limited, and to some extent
incompletely defined. Alternatively genetic manipulation of a cell
can either inhibit or induce new glyco-landscapes30 but as glyco-
sylation is probably the most complex secondary gene event in a
cell, the outcome is unpredictable and poorly controlled. Direct
covalent labeling of glycoconjugates to the cell surface although
possible has significant risks in affecting both the vitality and func-
tionality of the modified cells. The alternative approach as shown
here is to use FSL constructs. Because FSL constructs can be con-
trollable and are synthesized and designed to have specific fea-
tures, they are expected to provide a useful alternative method
for altering the glyco-landscape of living cells.
6. Henry, S. Curr. Opin. Hematol. 2009, 16, 467–472.
7. Hult, A. K.; Frame, T.; Chesla, S.; Henry, S.; Olsson, M. L. Transfusion 2012, 52,
247–251. http://dx.doi.org/10.1111/j.1537-2995.2011.03268.x.
8. Oliver, C.; Blake, D.; Henry, S. Transfusion 2011, 51, 1723–1730. http://
dx.doi.org/10.1111/j.1537-2995.2010.03034.x.
9. Oliver, C.; Blake, D.; Henry, S. Transfusion 2011, 51, 2664–2675. http://
dx.doi.org/10.1111/j.1537-2995.2011.03184.x.
10. Harrison, A. L.; Olsson, M. L.; Brad Jones, R.; Ramkumar, S.; Sakac, D.;
Binnington, B.; Henry, S.; Lingwood, C. A.; Branch, D. R. Glycobiology 2010,
27, 515–524.
11. Henry, S. M. Xenotransplantation 2005, 12, 356.
12. Hadac, E. M.; Federspiel, M. J.; Chernyy, E.; Tuzikov, A.; Korchagina, E.; Bovin, N.
V.; Russell, S.; Henry, S. M. J. Virol. Methods 2011, 176, 78–84. http://dx.doi.org/
10.1016/j.jviromet.32011.06.005.
13. Heathcote, D.; Carroll, T.; Wang, J. J.; Flower, R.; Rodionov, I.; Tuzikov, A.; Bovin,
N.; Henry, S. Transfusion 2010, 50, 635–641.
14. Georgakopoulos, T.; Komarraju, S.; Henry, S.; Bertolini, J. Vox Sang. 2011, 102,
72–78. http://dx.doi.org/10.1111/j.1423-0410.2011.01512.x.
15. Bovin, N. K.; Chinarev, A. A.; Tuzikov, A. B. Multiligand Constructs. World
International Property Organization, Publication Number WO/2010/043230,
2010.
16. Turner, R. E.; Lin, P. Y.; Cowman, M. K. Arch. Biochem. Biophys. 1988, 265, 484–
495.
17. Korchagina, E. Y.; Bovin, N. V. Bioorg. Khim. 1992, 18, 283–298.
18. Chinarev, A. A.; Galanina, O. E.; Bovin, N. V. Methods Mol. Biol. 2010, 600, 67–78.
19. Garratty, G. Vox Sang. 2008, 94, 87–95.
20. Holthius, C.; van Meer, G.; Huitema, K. Mol. Membr. Biol. 2003, 20, 231–241.
21. Kovalenko, E. I.; Khirova, E. V.; Khaidukov, S. V.; Molotkovskaya, I. M.;
Ovchinnikova, T. V.; Sapozhnikov, A. M.; Telford, W.; Vlaskin, P. A.; Bovin, N. V.
Histochem. Cell Biol. 2007, 127, 313–326.
22. Kelm, S.; Schauer, R. Int. Rev. Cytol. 1997, 175, 137–240.
23. Gambaryan, A. S.; Karasin, A. I.; Tuzikov, A. B.; Chinarev, A. A.; Pazynina, G. V.;
Bovin, N. V.; Matrosovich, M. N.; Olsen, C. W.; Klimov, A. I. Virus Res. 2005, 114,
15–22.
5. Associated content
Supporting Information. 1H NMR data for all FSL constructs.
Scheme S1 showing synthesis of DOPE with MCMG2, CMG2 and
CMG4 Spacers. Chart S1—Schematic diagrams of other FSL blood
group related structures and Chart S2 showing spacer defined dis-
tances between cell membrane and ligand. This material is avail-
able free of charge via the Internet at http://elsevier.com.
Acknowledgements
24. Bontekoe, S.; Blake, D.; Heineman, M. J.; Williams, E. C.; Johnson, N.; Cochrane
Database of Systematic Reviews, 2010, Issue 7. Art. No.: CD007421. doi:http://
dx.doi.org/10.1002/14651858.CD007421.pub2.
The study was supported in part by the Russian Academy of Sci-
ences Presidium Program ‘Molecular and Cell Biology’ and RFBR

246
E. Korchagina et al. / Carbohydrate Research 356 (2012) 238–246
25. Gardner, D. K.; Rodriguez-Martinez, H.; Lane, M. Hum. Reprod. 1999, 14, 2575–
2580.
28. Olsson, M. L.; Clausen, H. Br. J. Haematol. 2007, 140, 3–13.
29. Galili, U.; Anaraki, F. Glycobiology 1995, 5, 775–782.
30. Marionneau, S.; Moullac-Vaidye, B.; Le Pendu, J. Glycobiology 2002, 12, 851–
856.
26. Blundell, C. D.; Seyfried, N. T.; Day, A. J. Structural and Functional Diversity of
Hylauronan-Binding Proteins. In Chemistry and Biology of Hyaluronin; Garg, H.
G., Hales, C. A., Eds.; Elsevier Press: Netherlands, 2004; pp 189–204.
27. Chai, W.; Stoll, M. S.; Galustian, C.; Lawson, A. M.; Feizi, T. Methods Enzymol.
2003, 362, 160–195.