A novel role for pro-coagulant microvesicles in the early host defense against streptococcus pyogenes

Abstract

Previous studies have shown that stimulation of whole blood or peripheral blood mononuclear cells with bacterial virulence factors results in the sequestration of pro-coagulant microvesicles (MVs). These particles explore their clotting activity via the extrinsic and intrinsic pathway of coagulation; however, their pathophysiological role in infectious diseases remains enigmatic. Here we describe that the interaction of pro-coagulant MVs with bacteria of the species Streptococcus pyogenes is part of the early immune response to the invading pathogen. As shown by negative staining electron microscopy and clotting assays, pro-coagulant MVs bind in the presence of plasma to the bacterial surface. Fibrinogen was identified as a linker that, through binding to the M1 protein of S. pyogenes, allows the opsonization of the bacteria by MVs. Surface plasmon resonance analysis revealed a strong interaction between pro-coagulant MVs and fibrinogen with a KD value in the nanomolar range. When performing a mass-spectrometry-based strategy to determine the protein quantity, a significant up-regulation of the fibrinogen-binding integrins CD18 and CD11b on pro-coagulant MVs was recorded. Finally we show that plasma clots induced by pro-coagulant MVs are able to prevent bacterial dissemination and possess antimicrobial activity. These findings were confirmed by in vivo experiments, as local treatment with pro-coagulant MVs dampens bacterial spreading to other organs and improved survival in an invasive streptococcal mouse model of infection. Taken together, our data implicate that pro-coagulant MVs play an important role in the early response of the innate immune system in infectious diseases.

Figure 1. MVs bind to S. pyogenes.

A) Pro-coagulant MVs were gold-labeled with annexin V (5 nm) and incubated with S. pyogenes bacteria in the presence of 1% plasma. Samples were processed by negative staining and analyzed in a transmission electron microscope at two different magnifications. Arrows point to annexin V-positive MVs and scale bars represent 500 nm (left) or 100 nm (right). B) An overview of an S. pyogenes bacterium opsonized with ctrl. (upper panel left site) and pro-coagulant MVs (upper panel right site) is shown (Scale bar 500 nm). Ctrl. MVs (left side) and pro-coagulant MVs (right side) were immunostained with a gold-labeled antibody against CD14 (middle panel) or tissue factor (lower panel) and incubated with S. pyogenes. The scale bar indicates 100 nm.

Figure 2. Pro-coagulant MVs bound to S. pyogenes induce clotting in human plasma.

A) S. pyogenes were incubated with pro-coagulant MVs or buffer in the presence or absence of plasma for 30 min at 37°C. After washing, bacteria were added to recalcified plasma and clotting time was determined. B) S. pyogenes were incubated with pro-coagulant MVs in the presence of normal or fibrinogen (Fbg) depleted plasma for 30 min at 37°C. After washing, bacteria were added to recalcified plasma and clotting time was determined. C) Wildtype AP1 or MC25 bacteria were incubated with pro-coagulant MVs in the presence of plasma for 30 min at 37°C. After washing, bacteria were added to recalcified plasma and clotting time was determined. Clotting times were performed in triplicate. The data represent the means ± SD of 3 independent experiments. ***P<0.001.

Figure 3. Binding properties of MV surfaces measured by SPR.

Ctrl. MVs (A) or pro-coagulant MVs (B) were coupled to a sensor chip and subjected to injections with serial dilutions of fibrinogen (0.1–5 µM). Overlaid concentration-responses of the binding curves are shown. For both ligands, the data of the interaction with fibrinogen fit best to the heterogeneous binding model. The kinetic and affinity parameters are listed in (C). Pro-coagulant MVs were coupled to a sensor chip over which 100 µg/ml M1 protein was injected (D) or which was loaded sequentially, first with 5 µM fibrinogen, and then with 100 µg/ml M1 protein (E).

Figure 4. Pro-coagulant MV-derived clots prevent bacterial dissemination.

Recalcified plasma was incubated with 2×10⁵ CFU/ml S. pyogenes and clot formation was triggered with pro-coagulant PLs or MVs or 2 pM tissue factor. Tris-buffer (ctrl-buffer) or ctrl-MVs samples were used as negative controls. The clots were covered with 1% plasma, incubated at 37°C and supernatants were plated after 2 (A) and 4 hours (B). The data represent the means ± SD of 3 independent experiments, *P<0.05.

Figure 5. Determination of the antimicrobial activity of plasma clots induced by pro-coagulant MVs.

A) Recalcified plasma was incubated with 2×10⁵ CFU/ml S. pyogenes in the presence of ctrl. MVs, pro-coagulant MVs, PLs, or 2 pM tissue factor. Tris-buffer was used as a control (100%). B) Samples were prepared as described in A, but clot formation was prevented by omitting calcium, or the addition of 1.5 mg/ml Gly-Pro-Arg-Pro (C). Scanning electron microscopy of S. pyogenes (D), pro-coagulant MVs (E), and a plasma clot with S. pyogenes that was induced by the addition of pro-coagulant MVs (F). The data represent the means ± SD of 3 independent experiments.

Figure 6. Analysis of plasma from septic mice.

Mice were subcutaneously infected with 2×10⁷ CFU S. pyogenes bacteria and plasma samples were collected at 0, 24, 30, 42 and 48 hours after infection (n = 5–11/group). Tissue factor activity (A) and pro-coagulant MVs (B) expressed as PS equivalent (nM) were measured. *P<0.05, ***P<0.0001.

Figure 7. Treatment with pro-coagulant MVs in a mouse model of S. pyogenes sepsis dampens bacterial dissemination and improves survival.

A) Mice were subcutaneously infected with 2×10⁷ CFU S. pyogenes bacteria, immediately treated with 100 µl ctrl. or pro-coagulant MVs (150 MPs/µl, derived from 2×10⁶ PBMCs) and survival was monitored up to 120 h after infection. Results show 4 independent experiments with MVs from 4 different donors and a total amount of 20 animals per group. (B–D) Mice were infected as described above and were immediately treated with 100 µl pro-coagulant MVs (150 MPs/µl, derived from 2×10⁶ PBMCs), non-treated mice served as controls. Eighteen hours after infection, mice were killed and bacterial loads in the blood (B), spleen (C), and liver (D) were determined. Data are presented as means of 10 mice per group and were obtained from 2 independent experiments (*P<0.05).