Eosinophil-selective binding and proapoptotic effect in vitro of a synthetic Siglec-8 ligand, polymeric 6'-sulfated sialyl Lewis x

Abstract

The lectin Siglec-8 (sialic acid-binding, immunoglobulin-like lectin), which is selectively expressed on eosinophil surfaces and regulates eosinophil survival, preferentially binds to the glycan 6'-sulfo-sialyl Lewis X (6'-sulfo-sLe(x)). Antibody engagement of Siglec-8 on eosinophils causes their apoptosis, suggesting that engagement of Siglec-8 with its natural glycan ligands in vivo may control allergic inflammation. We report that a soluble synthetic polymer displaying 6'-sulfo-sLe(x) glycan selectively binds to human eosinophils and human embryonic kidney 293 cells expressing Siglec-8. Binding was inhibited by anti-Siglec-8 antibody. In whole blood, eosinophils were the only leukocyte subtype to detectably bind polymeric 6'-sulfo-sLe(x). Interleukin-5-primed eosinophils underwent apoptosis when incubated with either anti-Siglec-8 monoclonal antibody or polymeric 6'-sulfo-sLe(x), although the glycan polymer was less effective. These data demonstrate that a soluble, multivalent glycan selectively binds to human eosinophils and induces their apoptosis in vitro and provide proof-of-concept that such a reagent could be used to selectively target eosinophils.

Fig. 1.

HEK 293 cells stably transfected with Siglec-8 have the ability to bind 30-kDa 6′-sulfo-sLe^x-PAA polymer. Shown are flow cytometric histograms comparing mock-transfected HEK 293 cells that fail to bind Siglec-8 antibody and 6′-sulfo-sLe^x-PAA polymer. Also shown is the failure of binding of several structurally similar glycan-conjugated polymers. Results are representative of two experiments with similar results.

Fig. 2.

Effect of different Siglec-8 antibodies on 30-kDa 6′-sulfo-sLe^x-PAA polymer binding to HEK 293 cells stably transfected with Siglec-8. Black bars represent data with no added antibodies. A polyclonal sheep Siglec-8 antibody had blocking activity (open bar), whereas a combination of mouse monoclonal antibodies (50 μg/ml each of 2E2, 2C4, and 9G4), had no blocking activity (gray bar). Results are from a single experiment representative of three experiments with similar results. MFI, mean fluorescence intensity.

Fig. 3.

Adhesion of mock-transfected HEK 293 cells (open bars) and HEK 293 cells stably transfected with Siglec-8 (black bars) to various immobilized substrates. Attachment was only observed when using 30-kDa 6′-sulfo-sLe^x-PAA polymer and HEK 293 cells transfected with Siglec-8. Data are means ± S.E.M. of triplicate determinations and are from n = 2–4 experiments.

Fig. 4.

The 2000-kDa 6′-sulfo-sLe^x PAA polymer selectively binds to eosinophils among leukocytes in whole blood. Attachment of the 2000-kDa 6′-sulfo-sLe^x PAA polymer and the 2C4 Siglec-8 monoclonal antibody to eosinophils was readily apparent, whereas no labeling of monocytes, lymphocytes, neutrophils, or basophils was observed. Binding of the LacNAc control PAA polymer is seen for neutrophils and a subset of lymphocytes. Results are representative of at least eight experiments with similar results.

Fig. 5.

IL-5-primed eosinophils undergo apoptosis when exposed to Siglec-8 antibody or 6′-sulfo-sLe^x-PAA polymer. Results represent means ± S.D. from five experiments at 24 h and three experiments at 72 h. *, p < 0.02 compared with the CD44 control; **, p < 0.03 compared with the LacNac control; †, p < 0.005 compared with the CD44 control.