C-Reactive Protein Plays a Role in Antibody-Mediated Platelet Destruction

Abstract 526

IgG-mediated disorders like fetal or neonatal alloimmune thrombocytopenia (FNAIT) and the autoimmune acute immune thrombocytopenia (ITP), are caused by platelet-IgG-antibodies targeting platelets, resulting in Fcγ-receptor (FcγR)-mediated phagocytosis of platelets and eventually in thrombocytopenia. Markers to predict severity of these diseases are not yet available. Platelet-antibodies levels are related with outcome, however lack acceptable accuracy to predict disease severity.

We hypothesized that a bioassay measuring antibody-mediated platelet destruction might correlate with clinical outcome and therefore we developed an in vitro respiratory burst assay, using polymorphonuclear neutrophils (PMN). We observed that platelets opsonized with antibodies purified from sera of alloimmunized women did not result in strong antibody-mediated phagocyte responses unless the reaction occurred in the presence of patient serum (p<0.001) or normal human serum (NHS, p<0.001), also when recombinant human antibodies (IgG1 anti-HPA-1a) were used (p<0.001). This indicated that a serum-derived factor was required for antibody-mediated phagocytic activity by PMN. Serum heat inactivation excluded the role of complement, but the effect was dependent on Ca²⁺, and the level or the activity of this factor varied between different NHS (n=14). We therefore postulated that C-Reactive Protein (CRP), a Ca²⁺-dependent acute phase protein, variably present in NHS and upregulated during infections, could be this factor. CRP-neutralization using pneumococcal cell-wall polysaccharide and depletion of CRP from serum using phosphorylcholine beads reduced the phagocytic activity towards IgG-opsonized platelets (p<0.01 and p<0.001, respectively) while addition of CRP at physiological concentrations to serum enhanced it (p<0.05). Furthermore, the level of respiratory burst towards opsonized platelets in presence of various NHS correlated positively with the serum CRP levels (n=14, R²=0.45, p=0.0084). Fluorescently labelled CRP bound to IgG-opsonized platelets in a Ca²⁺-dependent manner (p<0.001). We hypothesized that a ligand for CRP, phosphatidylcholine, might be exposed on platelets after activation of platelet-FcγRIIa. However, the phagocytosis-enhancing effect did not appear to be mediated through platelet-FcγRIIa, as FcγRII-blocking antibodies did not decrease the level of phagocytosis. Also when using anti-D opsonized red blood cells (RBCs), which are FcγR-negative, we observed that the phagocytic activity was higher in the presence of serum (p<0.001). Using radioactively-labelled CRP, we found CRP to only bind to the IgG-opsonized on surfaces, independently of FcγR, as this effect was apparent on RBCs (p<0.05). Importantly, the concentration of CRP in sera from neonates with severe FNAIT was significantly elevated compared to healthy cord blood samples (4.21 mg/L vs. 0.05 mg/L, n=21, p=0.0008), and also in sera from children with acute ITP compared to healthy children (1.98 mg/L vs. 0.26 mg/L, n=19, p=0.02), indicating that CRP may be a contributing factor for platelet clearance in the presence of anti-platelet antibodies. To also test this hypothesis in vivo, we utilized our previously established mouse model (Teeling et al. Blood. 2001; 98:1095-1099). We administered 0.75 μg of rat-anti-platelet antibody causing mean platelet counts to drop from 1050 × 10⁹/L to 645 × 10⁹/L in 16 hours (n=7, p=0.0001). When 200 μg CRP was co-administrated, the mean platelet counts dropped further to 345 × 10⁹/L (n=7, p=0.0060), while administration of CRP alone had no effect on platelet counts, demonstrating that CRP potentiates the degree of thrombocytopenia induced by an anti-platelet antibodies in vivo.

These findings support a novel fundamental mechanism in which CRP can aggravate IgG-mediated phagocyte responses in FNAIT and acute ITP and perhaps also in other IgG-mediated responses in both health and disease. This involvement of CRP may possibly explain the lack of clinical correlation between severity and anti-platelet antibody titers and the frequent occurrence of acute ITP upon infection. We conclude that CRP functions as a previously unknown pathogenic co-factor in phagocyte responses towards IgG-opsonized platelets through a direct binding to IgG, thereby contributing to platelet destruction.