Adhesion of Staphylococcus aureus to the vessel wall under flow is mediated by von Willebrand factor-binding protein

Abstract

Adhesion of Staphylococcus aureus to blood vessels under shear stress requires von Willebrand factor (VWF). Several bacterial factors have been proposed to interact with VWF, including VWF-binding protein (vWbp), a secreted coagulase that activates the host's prothrombin to generate fibrin. We measured the adhesion of S aureus Newman and a vWbp-deficient mutant (vwb) to VWF, collagen, and activated endothelial cells in a microparallel flow chamber. In vivo adhesion of S aureus was evaluated in the mesenteric circulation of wild-type (WT) and VWF-deficient mice. We found a shear-dependent increase in adhesion of S aureus to the (sub)endothelium that was dependent on interactions between vWbp and the A1-domain of VWF. Adhesion was further enhanced by coagulase-mediated fibrin formation that clustered bacteria and recruited platelets into bacterial microthrombi. In vivo, deficiency of vWbp or VWF as well as inhibition of coagulase activity reduced S aureus adhesion. We conclude that vWbp contributes to vascular adhesion of S aureus through 2 independent mechanisms: shear-mediated binding to VWF and activation of prothrombin to form S aureus-fibrin-platelet aggregates.

Figure 1

vWbp mediates shear-dependent adhesion of S aureus to VWF. (A) Microparallel flow chamber perfusion over coated VWF (50 μg/mL) with fluorescently labeled wild-type (WT) and vwb strains at shear rates of 500 seconds⁻¹ and 1000 seconds⁻¹ in medium (n ≥ 6). (B) Perfusion over coated collagen with WT and vwb strains at 500 seconds⁻¹ or 1000 seconds⁻¹. VWF (60 μg/mL) was present in the medium where indicated (n ≥ 4). (C) Perfusion over coated collagen with WT or vwb strains in the presence of VWF-deficient plasma at 1000 seconds⁻¹. VWF (60 μg/mL) was added where indicated (n ≥ 4). All results are expressed as mean ± SEM. *P < .05, **P < .01, ***P < .001.

Figure 2

Coagulase activity increases S aureus adhesion to collagen under shear stress. (A) Microparallel flow chamber perfusion over coated collagen with fluorescently labeled WT, vwb, and coa/vwb strains in plasma with or without preincubation (37°C for 15 minutes) at 1000 seconds⁻¹. The ability of bacteria to generate coagulase-mediated fibrin during the preincubation phase in plasma increases subsequent bacterial adhesion. Where indicated, dabigatran (500 nm) was added to the plasma (preincubation 37°C for 15 minutes) (n ≥ 5). (B) Perfusion over coated collagen with fluorescently labeled fibrinogen and WT, vwb, and coa/vwb strains in the presence of plasma and platelet-rich plasma (preincubation 37°C for 15 minutes) (n ≥ 4). All results are expressed as mean ± SEM. *P < .05, **P < .01, ***P < .001.

Figure 3

Platelets increase S aureus adhesion under shear stress. (A) Microparallel flow chamber perfusions over coated collagen with fluorescently labeled WT, vwb, and coa/vwb strains in the presence of plasma and platelet-rich plasma (preincubation 37°C for 15 minutes) at 1000 seconds⁻¹. Addition of dabigatran (500 nM) or eptifibatide (7.5 μg/mL) where indicated (n ≥ 5). (B) Perfusion over coated collagen with WT, vwb, and coa/vwb strains in platelet-rich plasma at 1000 seconds⁻¹. Platelets were labeled with rhodamine-G (preincubation 37°C for 15 minutes). Adhesion of platelets was lower when perfused with a strain lacking vWbp compared with WT and was further reduced when perfused together with a mutant strain lacking both coagulases (vwb/coa) (n ≥ 4). All results are expressed as mean ± SEM. *P < .05, **P < .01. (C) SEM image (×20 000) of WT perfusion in PRP over collagen at 1000 seconds⁻¹, illustrating the interactions among bacteria, fibrin, and platelets. Bar represents 1 μm. (D) SEM image (×5000) of WT perfusion in PRP over collagen at 1000 seconds⁻¹. Bar represents 2 μm. Full white arrows indicate bacteria; dotted white arrows indicate platelets.

Figure 4

Blocking A1-domain of VWF inhibits S aureus binding to collagen under shear stress. (A) Perfusion over coated collagen with fluorescently labeled WT strain and 60 μg/mL VWF in medium at a shear rate of 1000 seconds⁻¹. Anti-A1 VWF domain antibody (Ab) 6D1 (final concentration 10 μg/mL) or anti-tPa monoclonal immunoglobulin G1 Ab (10 μg/mL) were added where indicated (n ≥ 5). (B) Perfusion over coated collagen at 1000 seconds⁻¹ with WT strain in PPP and PRP (PPP or PRP preincubation of 15 minutes at 37°C). 6D1 (final concentration 10 μg/mL) was added where indicated (n ≥ 4). All results are expressed as mean ± SEM. *P < .05, **P < .01.

Figure 5

vWbp mediates bacterial adhesion to activated HUVECs under flow conditions. (A) Microparallel flow chamber perfusions. ECs were activated with the Ca²⁺-ionophore A23187 (0.1 mM) followed by a 10-minute perfusion of fluorescently labeled WT and vwb strains at a shear rate of 1000 seconds⁻¹ in DMEM. Where indicated, the anti-A1 VWF domain antibody 6D1 (final concentration 10 μg/mL) was present. No difference was observed in the presence or absence of the anti-tPa monoclonal immunoglobulin G1 antibody. All results are expressed as mean ± SEM. *P < .05, **P < .01, n ≥ 5. (B) Image of microparallel flow chamber perfusion over activated HUVECs with WT at a shear rate of 1000 seconds⁻¹. S aureus forms strings on VWF over a distance of >200 microns.

Figure 6

Bacterial adhesion to activated endothelium in vivo is VWF and vWbp -mediated. (A) In vivo venous mesenteric perfusion model with C57Bl/6-Vwf^+/+ and C57Bl/6-Vwf^−/− mice. A total of 5 µL of the Ca²⁺-ionophore A23187 (10 mM) was applied to the region of the visualized vascular bed to trigger EC activation and VWF release. A suspension of carboxy-fluorescein–labeled WT, vwb, or coa/vwb strains was injected through the jugular catheter. Where indicated, bacterial inoculation was preceded by a bolus of 50 µL of dabigatran (10 µM). All results are expressed as mean ± SEM. **P < .01, ***P < .001, n ≥ 7. (B) Image of in vivo venous mesenteric perfusion model with C57Bl/6-Vwf^+/+ mice.