Free, monovalent SLeX (Neu5Ac alpha 2-3Gal beta 1-4(Fuc alpha 1-3)-Glc
NAc), SLn (Neu5Ac alpha 2-3Gal beta 1-4GlcNAc) and corresponding BSA-c
onjugated forms-displaying different ratios of SLeX and SLn to protein
-were tested for their ability to inhibit binding of HL-60 cells to im
mobilized E-selectin. Free SLeX and conjugated SLeX-BSA inhibited cell
binding in a dose-dependent manner. SLn and SLn-BSA did not inhibit b
inding. SLeX(16)BSA (16 mol tetrasaccharide/mol BSA) and monovalent SL
eX inhibited cell binding with measured inhibitory concentrations (IC5
0S) of 1 mu M and 1 mM, respectively, demonstrating a three-order-of-m
agnitude enhancement of inhibitory activity with the multivalent form
of SLeX. A SLex(7)BSA conjugate was 10-fold less potent than those wit
h 11 or 16 mol SLeX/mol BSA. An assay which measured neutrophil rollin
g on interleukin (IL)-1 beta-activated human umbilical vein endothelia
l cells (HUVECs) showed 50% reduction in the number of rolling neutrop
hils in the presence of 1 mu M SLeX(16)BSA, whereas the level of free,
monovalent SLeX oligosaccharide required to produce the same effect w
as similar to 0.3 mM. SLeX-BSA was found to be an excellent reagent fo
r staining endothelial cells expressing E-selectin. Biotinylated SLeX-
BSA in conjunction with Texas red avidin-stained lipopolysaccharide (L
PS)activated HUVECs, and co-incubation of activated cells with anti-E-
selectin, specifically blocked staining. The distribution of E-selecti
n, as determined by binding of SLeX-BSA, was virtually identical with
that obtained by binding of anti-E-selectin antibody. The pattern was
punctate in nature, rather than being diffuse, suggesting that E-selec
tin may be organized as clusters within the plasma membrane. The resul
ts suggest that multivalent forms of SLeX bind to E-selectin with high
er affinity than do monovalent glycans. Clustering of E-selectin in th
e membrane may be important for binding to counter-receptors on leukoc
yte cell surfaces.