Plasma and IgGs from patients with lupus anticoagulant induce tissue factor in monocytes: a possible risk factor for thrombosis

Tissue factor (TF) is a transmembrane protein belonging to the class 2 receptor family, and is the most important trigger of blood coagulation. Through its binding to factor (F)VIIa/FVII, TF exerts its effect by catalyzing the activation of FX and FIX by FVIIa. Under normal physiologic conditions, TF is almost entirely localized in the vessel wall in smooth muscle cells, fibroblasts, and pericytes, and only injury to the vessel wall can expose larger amounts of TF to the circulating blood (for review, see [1]). It is well documented that activated complement factors, such as C5a, are potent inducers of TF in monocytes [2,3]. However, Ritis et al. [4] demonstrated that antiphospholipid (aPL) antibody–complement activation and downstream signaling via C5a receptors induced TF expression in isolated neutrophils incubated with human serum. As it is quite difficult to obtain isolated neutrophils without some contamination of monocytes, this may be the explanation for the increase in TF mRNA. In their system, monocyte-derived TF rich microparticles from C5a activated monocytes may have been transferred to the neutrophils and accounted for the TF associated with the neutrophils in their fluorescence-activated cell sorting measurements. Almost simultaneously, Redecha et al. [5] reported that isolated neutrophils from these mice expressed TF in response to aPL antibody-generated C5a. This does not exclude the possibility that the monocytes release TF-positive microparticles that bind to activated neutrophils in the circulation and therebymay account for the TF associatedwith the neutrophils. As we recently definitely established that granulocytes do not synthesize TF but acquire it from monocytes [6,7], it seemed important to test the hypothesis that aPL antibody-generated C5a may have different properties from other agonists in inducing TF in neutrophils. It is well established that patients withlupus anticoagulant have frequent and severe thrombotic episodes. The present study was performed to find whether we could use the plasma from the patients to determine whether their plasma might contain antibodies with the potential to activate the complement system and subsequent induction of TF, and thereby account for the thrombosis in these patients. We also wanted to examine whether IgG isolated from their plasma might induce TF in blood of healthy individuals. The plasma samples were from patients referred to us from within our institution, as well as from other referrals who had clinical conditions leading the physicians to suspect the presence of lupus anticoagulant. All of the patients included in this study tested positive for both the screening dilute Russell s viper venom time (DRVVT) test and the confirmatory DRVVTconfirm test. Among the 106 patients, 56 had evidence of thrombosis, and 50 did not or were of unknown status. The clinical data were used to further divide the patients into two groups: those with thromboembolic diseases and those without at the time of testing. The results from TF activity induction data were also divided into two groups: the presence or absence of significant TF induction activity. Citrated plasma samples from patients were thawed, added fragmin (10 IE mL) and CaCl2 (final concentration 8 mM) and reconstituted with an equal volume of blood cells of a healthy volunteers. The blood cells were prepared by collecting the blood in fragmin (Pfizer Inc, New York, NY, USA); (10 IE mL) and centrifuged immediately at 2500 · g for 10 min. Cells were collected, washed once with 0.15 M NaCl, and then transferred to the plasmaof the patients and incubated for 2 h in a shaker (200 r.p.m.) at 37 C. After the incubation, 0.1 mL of 2% EDTA was added to stop the reaction, and the mononuclear cells were then isolated as described previously [8] and frozen at) 70 Cuntil further tested. TF activity in thawed cells wasmeasured in a two stage amidolytic assay based on the ability of TF to accelerate the activation of FX by FVIIa, as previously described [8]. A crude rabbit brain TF source was used as a standard reagent with an arbitrary potency of 1000 mU mL. Granulocytes were isolated from freshly drawn blood in heparin as described previously [9]. The mean TF activity detected with normal donors was 0.20 mU per 10 monocytes [standard deviation (SD) 0.20, range 0.02–0.5]. The cut-off value for positiveTF inductionwas Correspondence: Bjarne Østerud, Department of Medical Biology, Faculty of Health Sciences, University of Tromsø, 9037 Tromsø, Norway. Tel.: +47 77644730; fax: +47 77645350. E-mail: bjarne.osterud@uit.no