Erythrophagocytosis in autoimmune immunoglobulin A-mediated hemolysis.

INTRODUCTION: Warm antibody-mediated autoimmune hemolysis (WAIHA) is most often due to immunoglobulin G (IgG) antibodies and is detected by direct antiglobulin test (DAT). However, about 10% cases of hemolytic anemia are DAT negative. Herein, we describe a patient with DAT-negative hemolytic anemia due to an anti-IgA antibody. We investigate if isolated IgA can promote erythrophagocytosis. METHODS: We isolated patient and control IgA on Jacalin agarose. Isolated IgA was used to sensitize healthy ABO/Rh-matched donor red blood cells (RBC). Antibody binding was examined by flowcytometry. The effect of IgA on erythrophagocytosis was evaluated using Macrophage colony stimulating factor 1 (M-CSF)-differentiated autologous macrophages by Giemsa staining and immunofluorescence microscopy. RESULTS: We show that isolated IgA from the serum can bind to red cells. In addition, RBCs were phagocytosed efficiently by autologous macrophages in the presence of patient IgA. CONCLUSION: Our results show that IgA antibodies are capable of inducing erythrophagocytosis like IgG antibodies in the absence of complement fixation.

ß2-glycoprotein I promotes the clearance of circulating mitochondria.

ß2-glycoprotein I (ß2-Gp1) is a cardiolipin-binding plasma glycoprotein. It is evolutionarily conserved from invertebrates, and cardiolipin-bound ß2-Gp1 is a major target of antiphospholipid antibodies seen in autoimmune disorders. Cardiolipin is almost exclusively present in mitochondria, and mitochondria are present in circulating blood. We show that ß2-Gp1 binds to cell-free mitochondria (CFM) in the circulation and promotes its phagocytosis by macrophages at physiological plasma concentrations. Exogenous CFM had a short circulation time of less than 10 minutes in mice. Following infusion of CFM, ß2-Gp1-deficient mice had significantly higher levels of transfused mitochondria at 5 minutes (9.9 ± 6.4 pg/ml versus 4.0 ± 2.3 pg/ml in wildtype, p = 0.01) and at 10 minutes (3.0 ± 3.6 pg/ml versus 1.0 ± 0.06 pg/ml in wild-type, p = 0.033, n = 10). In addition, the splenic macrophages had less phagocytosed CFM in ß2-Gp1-deficient mice (24.4 ± 2.72% versus 35.6 ± 3.5 in wild-type, p = 0.001, n = 5). A patient with abnormal ß2-Gp1, unable to bind cardiolipin, has increased CFM in blood (5.09 pg/ml versus 1.26 ± 1.35 in normal) and his plasma induced less phagocytosis of CFM by macrophages (47.3 ± 1.6% versus 54.3 ± 1.3, p = 0.01) compared to normal plasma. These results show the evolutionarily conserved ß2-Gp1 is one of the mediators of the clearance of CFM in circulation.

Developmental endothelial locus-1 (Del-1) mediates clearance of platelet microparticles by the endothelium.

BACKGROUND: Phosphatidylserine-expressing microparticles circulate in blood with a short half-life of <10 minutes. We tested the role of an endothelium-derived phosphatidylserine-binding opsonin, developmental endothelial locus-1 (Del-1), in the uptake of platelet microparticles. METHODS AND RESULTS: Cultured human umbilical vein and microvascular endothelial cells avidly engulf BODIPY (4,4-difluoro-4-bora-3a,4a-diaza-s-indacene)-maleimide-labeled platelet microparticles. Microparticle uptake was inhibited by a monoclonal antibody to Del-1 (P=0.027) and by annexin A5 (P=0.027), abciximab (P=0.027), a monoclonal antibody to integrin αVß3 (P=0.027), and chlorpromazine (P=0.027). These results suggest that Del-1 mediates phosphatidylserine- and integrin-dependent endothelial uptake of microparticles by endocytosis. To assess the in vivo significance, we infused fluorescent platelet microparticles into the inferior vena cava of mice and harvested endothelial cells from the pulmonary and systemic circulation. Compared with their wild-type littermates, Del-1-deficient mice had decreased uptake in endothelial cells in lung (3.07±1.9 versus 1.09±1.3, P=0.02) and liver (2.85±1.1 versus 1.35±0.92, P=0.01). Furthermore, after endotoxin administration, Del-1-deficient mice displayed an increase in the level of microparticles compared with wild-type mice (P=0.02). CONCLUSIONS: These studies show a physiological role for Del-1 in the clearance of phosphatidylserine-expressing microparticles by endothelium.

Platelet senescence and phosphatidylserine exposure.

BACKGROUND: The exposure of phosphatidylserine occurs during platelet (PLT) activation and during in vitro storage. Phosphatidylserine exposure also occurs during apoptosis after the release of mitochondrial cytochrome c. We have examined the role of cytochrome c release, mitochondrial membrane potential (ΔΨm), and cyclophilin D (CypD) in phosphatidylserine exposure due to activation and storage. STUDY DESIGN AND METHODS: The exposure of phosphatidylserine and the loss of ΔΨm were determined in a flow cytometer using fluorescein isothiocyanate-lactadherin and JC-1, a lipophilic cationic reporter dye. The role of CypD was determined with cyclosporin A and CypD-deficient murine PLTs. Cytochrome c-induced caspase-3 and Rho-associated kinase I (ROCK1) activation were determined by immunoblotting and using their inhibitors. RESULTS: Collagen- and thrombin-induced exposure of phosphatidylserine was accompanied by a decrease in ΔΨm. Cyclosporin A inhibited the phosphatidylserine exposure and the loss of ΔΨm. CypD(-/-) mice had decreased loss of ΔΨm and impaired phosphatidylserine exposure. Collagen and thrombin did not induce the release of cytochrome c nor the activation of caspase-3 and ROCK1. In contrast, in PLTs stored for more than 5 days, the phosphatidylserine exposure was associated with cytochrome c-induced caspase-3 and ROCK1 activation. ABT737, a BH3 mimetic that induces mitochondrial pathway of apoptosis, induced cytochrome c release and activation of caspase-3 and ROCK1 and phosphatidylserine exposure independent of CypD. CONCLUSION: These results show that in stored PLTs cytochrome c release and the subsequent activation of caspase-3 and ROCK1 mediate phosphatidylserine exposure and it is distinct from activation-induced phosphatidylserine exposure.

Antiplatelet factor 4/heparin antibodies in patients with gram negative bacteremia.

Heparin-induced thrombocytopenia (HIT) is an antibody-mediated syndrome of thrombocytopenia and prothrombotic state that follows exposure to heparin. However, spontaneous HIT has been described in the setting of infection, without evidence of previous heparin administration. Since PF4 binds to lipid A portion of lipopolysaccharide, we tested for the presence of antiPF4/heparin antibodies in patients with gram-negative bacteremia. Patients with bacteremia had higher titers of antiPF4/heparin antibodies compared to normal controls 26.3 ± SD 34 units, N=32 versus 6.3 ± SD 2.38 units, N=10, P=0.001. FITC-labeled PF4 interacted with lipopolysaccharide in a concentration-dependent manner as determined by quenching of the emission spectrum following excitation at λ 488. In addition, immunoaffinity purified antiPF4/Heparin antibodies from 3 patients with HIT cross-reacted with PF4/heparin complex. These results show that PF4/LPS complex is immunogenic and can elicit cross-reacting antibodies against PF4/Heparin, providing an explanation for the presence of these antibodies in individuals, who were never been exposed to heparin before. These antibodies may also be at least partly responsible for the thrombocytopenia associated with infection.

Inhibition of thrombin activity by prothrombin activation fragment 1.2.

Prothrombin is the precursor of thrombin, the central enzyme in coagulation. Prothrombin is activated in vivo by the prothrombinase complex to form fragment 1.2 and thrombin. Fragment 1.2 has an amino-terminal gla domain and two kringle domains. The second kringle domain (kringle 2) binds to the exosite II on thrombin. Nascent thrombin generated on platelet surface remains non-covalently bound to fragment 1.2 by kringle 2-exosite II interaction. To determine whether this interaction can modulate coagulant activity of thrombin, we labeled thrombin at the active site with fluorescein-Phe-Pro-Arg chloromethylketone and monitored the fluorescence changes upon ligand binding. Anionic phospholipid-bound fragment 1.2 and fragment 2 bound to FPR-thrombin and induced changes in the active site with half maximal effects at 7.2 microM and 8.8 microM, respectively. We also tested the effect of anionic phospholipid-bound fragment 1.2 (0-10 microM) on thrombin clotting activity. Phospholipid-bound fragment 1.2 inhibited fibrinogen clotting in a concentration-dependent manner but had no significant effect on amidolytic activity towards S2238, suggesting a competitive inhibition of the fibrinogen binding site. Furthermore, fragment 1.2 inhibited FPR-thrombin binding to platelet. Consistent with these findings fragment 1.2 inhibited thrombin-induced aggregation of gel filtered platelets in a concentration-dependant manner. These results suggest that the membrane-bound prothrombin fragment 1.2 may play a role in hemostasis by down regulating the procoagulant activity of newly formed thrombin.