Cooperative effect between anti-PF4/H and anti-PF4 antibodies increases cell activation and thrombotic risk in HIT

Abstract

Heparin-induced thrombocytopenia (HIT) is a severe complication of heparin therapy, frequently associated with thrombosis. Immunoglobulin G (IgG) antibodies to heparin-platelet factor 4 (PF4/H) complexes play a central role in HIT by activating platelets and leukocytes via Fc gamma Receptor IIa (FcγRIIA). However, some patients also develop IgG against unmodified PF4 (anti-PF4), but their implication in the pathophysiology of HIT is unclear. Therefore, we assessed the impact of the joint presence of anti-PF4/H and anti-PF4 antibodies on cellular activation, platelet count, and thrombus formation, using chimeric monoclonal IgG1 antibodies specific for either PF4/H complexes (5B9) or PF4 alone (1E12). As expected, 5B9 coincubated with washed platelets without heparin did not induce platelet activation, but when a nonactivating concentration of 1E12 was present with 5B9, significant platelet activation was observed. This functional cooperation was Fc dependent and involved FcγRIIA receptors, given that it was no longer detectable with F(ab')2 fragments of 1E12 or 5B9 or with ibrutinib, which inhibits the FcγRIIA pathway. 5B9 at a nonactivating concentration of 1E12 also induced thrombus formation without heparin under flow conditions. Furthermore, when the 2 antibodies were injected together into human FcγRIIA/human PF4 transgenic mice, thrombocytopenia always occurred, with pulmonary thrombi in one-third of the injected mice, similar to that observed after injection of 5B9 and heparin. These results support that functional cooperation may exist between anti-PF4 antibodies of different specificity and promote cell activation, thrombocytopenia, and thrombosis. This process may also increase the risk of thrombosis in HIT even after heparin treatment has been discontinued.

Graphical abstract

Figure 1.

Synergistic platelet activation and aggregation induced by 5B9 and 1E12. (A) SRA performed with 5B9 and 1E12 incubated alone or together, without or with UFH, IV.3, or a control Ab. Data are the mean of serotonin release (n = 10). (B) SRA was performed with plasma samples from patients with HIT (n = 18) or with (C) nonpathogenic anti-PF4/H Abs (n = 18) and washed platelets from 4 healthy donors, without or with 1E12, control Ab, or UFH. (D) Whole blood platelet aggregation induced by 5B9 and 1E12 alone or together, without or with UFH or control Ab. Data are the mean of AUC (n = 23). Mann-Whitney U test was used to compare the different conditions tested. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001. The dotted line represents the cutoff value for SRA in panel A, B and C or whole blood platelet aggregation in panel D. AUC, area under the curve; ELISA, enzyme-linked immunosorbent assay.

Figure 2.

Cross-linking of 5B9 and 1E12 with FcγRIIA is essential for their synergistic Ab effect. (A) Platelet activation measured by SRA (n = 10) and (B) flow cytometry assay (CD62P expression, n = 13). 5B9 and 1E12 were tested alone or together, without or with IV.3 (10 μg/mL) or their F(ab’)₂ fragments. (C) Platelet activation (CD62P expression, n = 4) induced by 5B9 and 1E12 alone or together, without or with ibrutinib. 5B9 with UFH and PF4 was tested as a positive control. Mann-Whitney U test was performed to compare the different conditions tested. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001.

Figure 3.

A 2-way cooperative effect of 5B9 and 1E12 enhances their binding to platelets. Platelet binding of (A) labeled-DG-5B9 (n = 5) and (C) labeled-DG-1E12 (n = 8) alone or in the presence of 1E12, 5B9, or their F(ab’)₂ fragments or IV.3. Inhibitory effect of ibrutinib on the platelet binding (B) of labeled-DG-5B9 with or without 1E12 or UFH + PF4 (n = 7) (D) of labeled-DG-1E12 with or without 5B9 (n = 7). Data are the median (minimum to maximum) of the MFI ratio. Mann-Whitney U test was used to compare the different conditions tested. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001.

Figure 4.

Synergistic effect between 5B9 and 1E12 on TF synthesis and thrombus formation in vitro. (A) Relative TF mRNA synthesis after addition of 5B9 and/or 1E12 or UFH to whole blood from healthy donors (mean ± standard error of the mean [SEM], n = 6 donors). (B) Representative images of thrombus formation in von Willebrand factor–coated microfluidic channels perfused (8 minutes, 500 s^–1) with recalcified whole blood incubated with 5B9 and/or 1E12 with or without control Ab or UFH. Images corresponding to areas of 0.1 mm² are shown with platelets in green (DiOC6), fibrin(ogen) in red (Alexa Fluor 647–labeled fibrinogen), and leukocytes in blue (Hoechst 33342, DNA dye). (C) The mean areas covered by platelets (green), leukocytes (blue), or fibrin(ogen) (red) were calculated for each condition (mean ± SEM; n = 12 independent experiments), by measuring using ImageJ software the surface covered by large aggregates (>100 μm²) in 30 different areas. For each experiment, the highest concentration of 5B9 and 1E12 that did not induce thrombus formation when incubated alone in the absence of UFH was selected (50 or 100 μg/mL for 5B9 and 1 or 2 μg/mL for 1E12) and used for subsequent synergy experiments. Mann-Whitney U test was performed to compare the different conditions tested. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001. IVIg, IV immunoglobulins.

Figure 5.

5B9 and 1E12 induce thrombocytopenia and thrombosis in vivo. (A) Evolution of platelet count (PC) after injection of 5B9 (1 μg/g body weight; n = 4), 1E12 (0.3 or 1 μg/g; n = 3) or both Abs (n = 3) in transgenic HIT mice (human FcγRIIA, human G6b-B, human PF4 transgenic, mouse PF4 knockout). At days 1, 2, 3 and 4, 1.4 IU/g of UFH was injected in mice treated with 5B9 (n = 3). Data are represented as mean ± SEM. Mann-Whitney U test was performed to compare the different conditions tested. ∗P < .05. (B-E) Sections of lungs with intravascular thrombi from mice injected with 1E12 1 μg/mL + 5B9 1 μg/mL (B-C) or 5B9 1 μg/mL + UFH 0.1 IU/mL (D-E). (F-K) Sections of lungs without thrombi (controls) in mice injected with 5B9 1 μg/mL (F-G), 1E12 1 μg/mL (H-I), or PBS (J-K). Sections were stained using hematoxylin, phloxine, and saffron (HPS) in panels B,D,F,H,J or immunofluorescence in panels C,E,G,I,K. Scale bar, 100 μm (for HPS staining and 150 μm for immunofluorescence).