Tissue factor needs a “complement”

Antiphospholipid antibody (aPL)–induced fetal loss remains an important clinical problem for which the pathogenesis is poorly understood.¹  Once thought to result exclusively from placental thrombosis, recent reports suggest that thrombosis is not universally present in placentae from affected women.²  In this issue of Blood, Redecha and colleagues define a critical role for tissue factor (TF) and complement in a murine model of aPL-induced fetal injury in which placental inflammation is the primary lesion.

In this model, fetal resorption occurred in 43% of pregnant mice that received aPL IgG. aPL treatment was associated with strong expression of TF in the decidua, and with decidual invasion by neutrophils. Blockade of TF with an anti–murine TF antibody, or the presence of only low levels of TF, prevented resorption. These results demonstrated a critical requirement for TF in aPL-mediated fetal resorption.

In additional studies, an IgG1κ murine aPL that does not activate complement did not cause fetal resorption, although an IgG2bκ aPL that does fix complement did. Resorption was prevented by the C3 convertase inhibitor, Crry-IgG, but not by C3 deficiency (C3^−/− mice). Because previous work demonstrated that aPL did not induce fetal resorption in C5-deficient mice, the authors speculated that C5b-stimulated assembly of the MAC (C5b-9) might be required; however, C6-deficient mice were not protected from resorption.

These results suggested that antibody binding to trophoblast cells was not sufficient to induce resorption, that complement activation was necessary, and that a role for the C5a-C5a receptor (C5aR) pathway was likely. The latter postulate was confirmed by the observation that C5aR deficiency prevented fetal resorption. Moreover, C5aR was expressed on neutrophils (but not trophoblasts), and C5a induced neutrophil TF expression. Both systemic neutrophil depletion and selective deficiency of myeloid-cell TF also protected mice from aPL-induced fetal resorption. Finally, TF expression contributed to a neutrophil oxidative burst with evidence of oxidative stress injury in the decidua of aPL-treated mice.

Taken together, these findings are consistent with a mechanism in which aPLs bind to trophoblast and activate complement via the classic pathway, leading to generation of C5a. In turn, C5a binds neutrophil C5aR and induces neutrophil TF expression, decidual invasion, and release of superoxides that cause oxidative stress and fetal resorption (see Figure 7 in the article by Redecha and colleagues).

Some caveats concerning this study are worth mention. First, the structure of human and murine placentae differs.³  Second, the majority of studies were performed with human aPLs, the reactivity of which with β₂GPI was not reported. Third, decreased endovascular trophoblast invasion may be a feature of aPL-associated fetal loss in humans, but was not addressed.⁴  Fourth, although inflammatory-cell infiltration of human placentae from aPL patients has been observed, some studies suggest that infiltrates are composed primarily of macrophages.⁵  Finally, the association between TF expression and superoxide release has not been defined molecularly. These concerns suggest that the applicability of this model to human aPL-mediated fetal loss requires further analysis. Nevertheless, this elegant study provides insight into a complex and poorly understood disorder.