PG-dependence of inhibition of TF-mediated coagulation and signaling. (A) TF expression was determined by fluorescence-activated cell sorting (FACS) analysis of stably expressing human TF in CHO-K1 (wild-type CHO/TF) and pgsA-745 (PG-deficient CHO/TF) cell lines. Isotype-matched immunoglobulin G (IgG) was used as control. MF indicates mean fluorescence values. (B) Xa generation on monolayers of confluent TF-expressing wild-type (□) and PG-deficient CHO cells (○). Cells were incubated at 37°C with 10 nM VIIa and 50 nM X, and Xa generation was determined by chromogenic assay at the indicated times. Results are mean ± SEM (n = 5). (C) rTFPI-1–mediated internalization of Xa. Confluent wild-type (□) and PG-deficient CHO cells (▪) were incubated with rTFPI-1 in PBS 5% BSA for 30 minutes on ice and washed with the same buffer. Internalization of 50 nM [¹²⁵I]-Xa was carried out for 15 minutes at 37°C, and internalized [¹²⁵I]-Xa was determined by counting trypsinized and extensively washed cell pellet with background subtraction of cell-free wells subjected to the same procedure. Results are expressed relative to cells receiving no rTFPI-1, mean ± SEM (n = 5). (D) Inhibition of TF-VIIa–dependent Xa generation was determined by adding VIIa (10 nM) and X (50 nM) in the presence of the indicated concentrations of rTFPI-1 for 15 minutes at 37°C and determination of Xa by chromogenic assay; mean ± SEM (n = 6). (E) Representative Western blots of wild-type and PG-deficient CHO cells stimulated with VIIa (10 nM) and X (100 nM) in the presence of the indicated concentrations of rTFPI-1 for 10 minutes at 37°C. Phosphorylated ERK1/2 was detected with phospho-specific ERK1/2 antibody (α-PERK1/2) with loading control with αERK1/2. (F) Densitometric quantitation of rTFPI-1–mediated reduction in ERK1/2 phosphorylation with normalization for loading; mean ± SEM (n = 4). For D and F, ▵ indicates wild-type; □, PG-deficient CHO cells.