Polyreactive IgM initiates complement activation by PF4/heparin complexes through the classical pathway

Abstract

The mechanisms by which exposure to heparin initiates antibody responses in many, if not most, recipients are poorly understood. We recently demonstrated that antigenic platelet factor 4 (PF4)/heparin complexes activate complement in plasma and bind to B cells. Here, we describe how this process is initiated. We observed wide stable variation in complement activation when PF4/heparin was added to plasma of healthy donors, indicating a responder "phenotype" (high, intermediate, or low). Proteomic analysis of plasma from these healthy donors showed a strong correlation between complement activation and plasma immunoglobulin M (IgM) levels (r = 0.898; P < .005), but not other Ig isotypes. Complement activation response to PF4/heparin in plasma displaying the low donor phenotype was enhanced by adding pooled IgM from healthy donors, but not monoclonal IgM. Depletion of IgM from plasma abrogated C3c generation by PF4/heparin. The complement-activating features of IgM are likely mediated by nonimmune, or natural, IgM, as cord blood and a monoclonal polyreactive IgM generate C3c in the presence of PF4/heparin. IgM facilitates complement and antigen deposition on B cells in vitro and in patients receiving heparin. Anti-C1q antibody prevents IgM-mediated complement activation by PF4/heparin complexes, indicating classical pathway involvement. These studies demonstrate that variability in plasma IgM levels correlates with functional complement responses to PF4/heparin. Polyreactive IgM binds PF4/heparin, triggers activation of the classical complement pathway, and promotes antigen and complement deposition on B cells. These studies provide new insights into the evolution of the heparin-induced thrombocytopenia immune response and may provide a biomarker of risk.

Graphical abstract

Figure 1.

Complement activation in response to PF4/heparin complexes among healthy donors defines a donor phenotype. (A) Plasma from healthy donors (n = 10) was incubated with buffer, PF4 ± heparin, or heparin alone, and binding of C3c to PF4/heparin complexes was measured by antigen-C3c capture ELISA assay. The graph shows the anti-C3c absorbance of donor plasma incubated with different antigens. Each symbol represents an individual donor. Results are shown from a representative experiment performed a minimum of 3 times, with multiple donors in each experiment. (B) Complement activation by PF4/heparin of donors (1, 2, 3, and 4) was determined during a period of 626 days (∼1.7 years) and normalized to donor 1, who was studied at all points. The graph shows percentage PF4/heparin complement activation relative to donor 1 on different days. **P < .0001.

Figure 2.

Complement activation by PF4/heparin correlates with plasma/serum IgM levels. (A) Proteomic analysis of plasmas with a high, intermediate, or low complement response phenotype shows strong correlation with plasma IgM. Graph shows IgM protein intensity determined by proteomic analysis (x-axis) and complement activation response to PF4/heparin, as measured by the antigen-C3c capture ELISA assay (y-axis). (B-D) Serum immunoglobulin levels (IgM, IgG, and IgA) from 29 healthy donors were measured in the clinical laboratory and correlated with an individual’s complement activation response to PF4/heparin, as measured by the antigen-C3c capture ELISA assay. Graphs show correlation of complement activation (y-axis) as a function of immunoglobulin levels (x-axis). Each symbol in the graph represents an individual donor. Complement activation values were normalized to an intermediate donor studied in parallel.

Figure 3.

Plasma IgM mediates complement activation by PF4/heparin complexes. (A) Commercial IgM (0-1000 μg/mL; filled symbols) or IgG (0-5000 μg/mL; open symbols) or monoclonal myeloma IgM (0-1000 μg/mL; hatched symbols) was added to the plasmas of 2 donors with a low complement response type (circle/square), and complement activation by PF4/heparin was measured by the antigen-C3c capture ELISA assay. Graph shows complement activation (y-axis) as a function of added immunoglobulin concentration. (B) Plasma with an intermediate donor phenotype was incubated with anti-IgM or control beads, followed by the addition of buffer, PF4/heparin, or PF4/heparin +400 µg/mL IgM, and complement activation was measured by the antigen-C3c capture ELISA assay. Graph shows complement activation (y-axis) in control or IgM depleted plasma (x-axis). (C) Plasma with a low donor phenotype was incubated with varying antigen concentrations (PF4; 0-25 μg/mL + heparin; 0-0.25 U/mL) and IgM (0-800 μg/mL), and complement activation was measured by the antigen-C3c capture ELISA assay. Graph shows complement activation response at varying IgM concentrations (y-axis) as a function of PF4/heparin concentrations (x-axis). *P < .005; **P < .0001. Results are shown from a representative experiment involving 3 donors tested on 3 different occasions.

Figure 4.

Polyreactive IgM mediates complement activation by PF4/heparin. (A) Antigen-specificity of commercial IgM was determined using microtiter plates coated with various antigens. The graph shows the binding (y-axis) of various concentrations of commercial IgM (1.25-80.0 µg/mL) to different antigens. (B) Complement activation by polyreactive monoclonal IgM, 2E4, in the plasma of 2 donors (low complement activation phenotype; circle/square) in response to buffer (open symbols), PF4 alone (hatched symbols), heparin alone (half-filled symbols) and PF4/heparin (filled symbols) as measured by the antigen-C3c capture ELISA assay. Graph shows complement activation (y-axis) as a function of added polyreactive IgM concentrations. Results are shown from a representative experiment involving 3 donors tested on 3 different occasions. **P < .0001, relative to no polyreactive IgM added. (C) Whole blood from the cord blood of a baby was incubated with buffer or PF4 ± heparin and binding of PF4/heparin (KKO) or C3c to the B cells was determined by flow cytometry. Histograms show the representative results from 2 different experiments with 2 different cord blood samples.

Figure 5.

PF4/heparin activate complement by classical pathway. (A) Plasma from a healthy donor was incubated with EDTA (10 mM) or EGTA (10 mM) ± MgCl₂ (10 mM) or with buffer before incubating with PF4/heparin and complement activation was measured by the antigen-C3c capture ELISA assay. The y-axis shows the complement activation in different incubation conditions. ***P < .0001. (B) Plasma from a healthy donor was incubated with various concentration of anti-C1q antibody, anti-MBL antibody, or control antibody (0-100 μg/mL) before adding PF4/heparin and complement activation by PF4/heparin was determined by the antigen-C3c capture ELISA assay. The y-axis shows the complement activation in presence of various antibodies. *P < .05; **P < .001; ***P < .0001 compared with no antibody added condition. Results are shown from a representative experiment involving 3 donors tested on 3 different occasions.

Figure 6.

Plasma IgM colocalizes with PF4/heparin and C3 fragments on B cells in healthy donors and patients on heparin therapy. (A-B) Whole blood from a representative healthy donor was incubated with buffer or PF4 (25 μg/mL) ± heparin (0.25 U/mL) and binding of C3c, PF4/heparin, and IgM on B cells was determined by flow cytometry. Binding of C3c/anti-PF4/heparin (KKO)/IgM to B cells is shown with and without PF4 ± heparin by histogram overlays in normal and excess heparin wash conditions (A) and as mean fluorescent intensity (B). Results are shown from a representative experiment involving 3 donors tested on 3 different occasions. (C-D) Binding of C3c, PF4/heparin, and IgM on B cells in the circulation of heparinized patients was determined by flow cytometry. Binding of C3c/anti-PF4/heparin (KKO)/IgM to B cells is shown pre- and postheparin exposure in the patient by overlay histograms (C) and as mean fluorescence intensity (D). Results are shown from 1 representative patient out of 3 patients studied. *P < .005; **P < .0001.

Figure 7.

Mechanism of complement activation by PF4/heparin complexes. (A) Heparin displaces PF4 to form ULCs. (B) Polyreactive natural IgM from plasma binds to ULCs, changes conformation and binds C1q to activate the classical pathway of complement activation. (C) IgM and complement-coated antigen binds to B cells via complement receptor 2.