The membrane-proximal intermolecular disulfide bonds in glycoprotein Ib influence receptor binding to von Willebrand factor

Abstract

Background: In the platelet glycoprotein (GP)Ib-IX complex, the binding site for its ligand von Willebrand factor (VWF) is restricted to the N-terminal domain of the GPIbalpha subunit. How the other subunits in the complex, GPIbbeta and GPIX, regulate the GPIbalpha-VWF interaction is not clear.

Objectives and methods: As GPIbalpha connects with two GPIbbeta subunits via disulfide bonds, we tested whether these intersubunit covalent links were important to the proper VWF-binding activity of the GPIb-IX complex by characterizing the structure and VWF-binding activity of a mutant GPIb-IX complex that lacked the GPIbalpha-GPIbbeta disulfide bonds.

Results: Mutating both Cys484 and Cys485 of GPIbalpha to serine prevents GPIbalpha from forming covalent disulfide bonds with GPIbbeta, while maintaining the integrity of the complex in the membrane. The mutations cause two GPIbbeta subunits to form a disulfide bond between themselves. As compared to Chinese hamster ovary (CHO) cells stably expressing the wild-type GPIb-IX complex at a comparable level, CHO cells stably expressing the mutant GPIb-IX complex bind to significantly less soluble VWF in the presence of ristocetin and roll on the immobilized VWF under flow at a higher velocity.

Conclusions: The disulfide bonds between GPIbalpha and GPIbbeta are necessary for optimal GPIbalpha binding to VWF. The structural plasticity around the disulfide bonds may also help to shed light on the inside-out mechanism underlying GPIbbeta modulation of VWF binding.

Figure 1. GPIbα_SS containing the C484S/C485S is well expressed and takes on largely native-like conformation in transfected CHO cells

(A) Comparable expression levels of the mutant GPIb-IX complex expressed in CHOα_SSβIX cells (dashed line) and the wild type complex in CHOα_CCβIX cells (solid line). Surface expression of GPIbα and GPIX were measured by flow cytometry with monoclonal antibodies WM23 and FMC25, respectively. Dark grey peak: CHO cells; light grey peak: CHOβIX cells. (B,C) Conformation of GPIbα_SS expressed in CHOα_SSβIX cells is largely native-like as the wild type in CHOα_CCβIX cells. GPIbα expressed in transfected CHO cells were stained with the indicated conformation-sensitive monoclonal antibodies and measured by flow cytometry. Representative histograms of antibody binding was shown in (B). The mean fluorescence intensities (MFI) were measured from 4–6 independent experiments, normalized to the wild type level and presented as the mean ± SD (C). Groups were compared using the non-paired t test; **, p < 0.01. Binding of mutant GPIbα_SS to AK2, another conformation-sensitive antibody, was significantly increased compared to the wild type GPIbα_CC, suggesting a local conformational change around GPIbα residues 35– 59, the epitope region recognized by AK2.

Figure 2. The C484S/C485S double cysteine mutation in GPIbα abolishes formation of disulfide bonds between GPIbα and GPIbβ in transfected CHO cells

Lysates from CHOα_CCβIX (CC) and CHOα_SSβIX (SS) cells were resolved in a SDS gel under non-reducing (N.R.) or reducing (R.) conditions, and the blotted membrane was probed with WM23. The figure is a representative of 5 independent experiments.

Figure 3. Structural integrity of the mutant GPIb-IX complex

Cell lysates from CHOα_CCβIX (CC) and CHOα_SSβIX (SS) cells were immunoprecipitated with FMC25 in the buffer containing either 1% digitonin (A) or 1% triton X-100 (B), and resolved in a SDS gel under reducing conditions. After transfer, the membrane was probed separately for GPIbα, or with polyclonal antibodies for GPIbβ or GPIX. Each figure is a representative of 3–4 independent experiments.

Figure 4. Formation of a disulfide-linked GPIbβ dimer in the GPIb-IX complex containing the C484S/C485S mutation

(A) GPIbβ in the mutant GPIb-IX complex is linked to another protein through a disulfide bond. Cell lysate from CHOα_SSβIX cells was first co-immunoprecipitated with antibodies against GPIbα (Ibα lane), GPIX (IX lane) or mouse IgG (IgG lane) in the lysis buffer containing 1% digitonin, and then resolved in a SDS gel under non-reducing conditions. For comparison, 20% of the cell lysate used in immunoprecipitation was loaded directly in the same SDS gel ( lane). After transfer, the membrane was probed with a polyclonal anti-GPIbβ antibody. Note that only a band with a molecular mass of approximate 45kDa was present in the “Ibα” lane, and the apparent mass of GPIbβ monomer is approximately 25kDa. (B–D) GPIbβ in the mutant GPIb-IX complex is linked to another GPIbβ via a disulfide bond. CHOα_SSβIX cell membrane including the embedded proteins was separated from the cytosolic fraction and co-immunoprecipitated by anti-GPIX antibody SZ1. The precipitated fraction was first resolved in a 4–12% Bis-Tris precast SDS gel (Invitrogen) under non-reducing conditions. The gel strip was either (B) directly stained by silver staining or extracted, incubated at 65 °C for 20 min in the SDS sample buffer containing 20% DTT and placed on top of a 12% Tris-glycine SDS gel for electrophoresis under reducing conditions. The proteins in the gel were either (C) visualized by silver staining or (D) transferred to a PVDF membrane and probed with anti-GPIbβ polyclonal antibody. The diagonal molecular weight markers in the 2-D gel were apparent in (C) and marked in (D). The dashed line highlights the positions of the 45-kDa protein complex in the 2-D gel. Silver staining showed that GPIbβ was the only protein in this 45-kDa complex. The protein complex with an apparent mass of approximate 150kDa in (B) is the antibody used for immunoprecipitation and used as the positive control for the 2-D gel. Each figure is a representative of at least 2 independent experiments.

Figure 5. Ristocetin-induced binding of soluble VWF to transfected CHO cells

The binding was detected by flow cytometry, and plotted as MFI versus VWF concentration. MFI values of CHOα_CCβIX and CHOα_SSβIX cells were normalized against the GPIbα expression level as indicated by WM23 binding. All data are presented as the mean ± SD from 4 independent experiments. Groups were compared using the nonpaired t test; *, p < 0.05; **, p < 0.01.

Figure 6. Rolling of CHOα_CCβIX (black bar) and CHOα_SSβIX (white bar) cells on the VWF-coated surface under flow conditions

The cells were perfused into the parallel-plate flow chamber and settled for 2 minutes on the VWF-coated coverslip before being subjected to various flow shear stresses. CHOβIX cells did not tether on the VWF-coated coverslip (not shown). The rolling velocity was defined as the distance a cell travels during a defined period. The data are presented as the mean ± SEM from measurements of 92–174 cells under each condition. Groups were compared using the nonpaired t test; **, p < 0.01.