Fig. 4. Galectin-1 binding to CA125-C-TERM largely depends on O-linked ß-galactose-terminated oligosaccharide chains. CHO_MCAT-TAM2 (wild-type background with regard to galactosylation of both proteins and lipids) and CHO_{clone 13} cells (deficient with regard to galactosylation of both proteins and lipids; (Deutscher and Hirschberg, 1986) stably expressing CA125-C-TERM were used to prepare cell-free detergent lysates followed by incubation with GST and GST—galectin-1 beads, respectively. Where indicated, CHO_MCAT-TAM2 were treated with 10 µg/ml tunicamycin for 18 hours at 37°C prior to cell lysis. In each experiment, the CA125-C-TERM signal derived from 0.2% of the input was compared with 4% of the material bound to either GST or GST—galectin-1 beads. Protein samples were separated on NuPage Bis-Tris gels followed by CA125-C-TERM immunoblotting employing the mAb OC125. (A) Lysates derived from CA125-C-TERM-expressing CHO_MCAT-TAM2 cells (lanes 1-3), CA125-C-TERM-expressing CHO_MCAT-TAM2 cells treated with tunicamycin (lanes 4-6), CA125-C-TERM-deficient CHO_MCAT-TAM2 cells (lanes 7-9), CA125-C-TERM-deficient CHO_MCAT-TAM2 cells treated with tunicamycin (lanes 10-12), CA125-C-TERM-expressing CHO_{clone 13} cells (lanes 14-16) and CA125-C-TERM-deficient CHO_{clone 13} cells (lanes 17-19). In lane 13, HeLa-derived CA125 eluted from GST—galectin-1 beads is shown as a control. (B) Quantification of the results shown in panel A. On the basis of the input signal (0.2% of starting material; panel A, lane 1), about 40% of CA125-C-TERM present in the cell lysate is recovered on GST—galectin-1 beads under the conditions used (panel A, lane 3, 4% of eluate), based on quantification employing Bio-Rad® QuantityOne® software. This value was set to 100% binding efficiency and compared with CA125-C-TERM—galectin-1 binding efficiencies measured with lysates either derived from tunicamycin-treated CHO_MCAT-TAM2 cells or from CHO_{clone 13} cells. The results shown represent mean values of two independent experiments.