Factor XII-driven coagulation traps bacterial infections

Abstract

Blood coagulation is essential for stopping bleeding but also drives thromboembolic disorders. Factor XII (FXII)-triggered coagulation promotes thrombosis while being dispensable for hemostasis, making it a potential anticoagulant target. However, its physiological role remains unclear. Here, we demonstrate that FXII-driven coagulation enhances innate immunity by trapping pathogens and restricting bacterial infection in mice. Streptococcus pneumoniae infection was more severe in FXII-deficient (F12-/-) mice, with increased pulmonary bacterial burden, systemic spread, and mortality. Similarly, Staphylococcus aureus skin infections and systemic dissemination were exacerbated in F12-/- mice. Reconstitution with human FXII restored bacterial containment. Plasma kallikrein amplifies FXII activation, and its deficiency aggravated S. aureus skin infections, similarly to F12-/- mice. FXII deficiency impaired fibrin deposition in abscess walls, leading to leaky capsules and bacterial escape. Bacterial long-chain polyphosphate activated FXII, triggering fibrin formation. Deficiency in FXII substrate factor XI or FXII/factor XI co-deficiency similarly exacerbated S. aureus infection. The data reveal a protective role for FXII-driven coagulation in host defense, urging caution in developing therapeutic strategies targeting this pathway.

Graphical abstract

Figure 1.

S. pneumoniae–induced lethality is increased in F12 ^−/− mice. WT and F12^−/− mice were intranasally inoculated with 5.0 × 10⁶ CFUs of S. pneumoniae (PN122) Xen10 (n = 15 per group comprising 10 male and 5 female mice). (A) Representative S. pneumoniae in vivo bioluminescence on a pseudocolor scale overlaid on top of a greyscale image of WT and F12^−/− mice. (B–D) (B) Bacterial counts as measured by in vivo bioluminescence of S. pneumoniae demonstrated as mean total flux (photons per second) ± SEM presented on a logarithmic scale. Bacterial loads observed in (C) lungs and (D) bronchoalveolar lavage (BAL). Filled and empty symbols represent male and female mice, respectively. (E) Number of secondary infections detected via S. pneumoniae in vivo bioluminescence. (F) Survival analysis of WT and F12^−/− mice challenged with S. pneumoniae. *P < 0.05 and **P < 0.01 F12^−/− versus WT. P values were determined using one-way ANOVA (B) or Student’s t test (C and D).

Figure 2.

Accelerated S. aureus–induced skin infection in FXII-deficient mice is rescued by human FXII. WT, F12^−/− mice, and F12^−/− mice reconstituted once with a dose of 2 µg human FXII per gram of body weight (F12^−/−+FXII) were injected subcutaneously with 1 × 10⁹ CFU S. aureus bioluminescent strain Xen29 (n = 10 per group). (A) Representative S. aureus in vivo bioluminescence on a pseudocolor scale overlaid on top of a greyscale image of WT, F12^−/−, and F12^−/−+FXII mice. (B) Bacterial counts as measured by in vivo bioluminescence of S. aureus demonstrated as mean total flux (photons per second) in a logarithmic scale. (C) Mean total size of the skin lesion in cm² ± SEM. (D–G) Bacterial burdens observed in kidneys (D), spleens (E), lungs (F), and livers (G) of WT, F12^−/−, and F12^−/−+FXII mice at day 5 after infection. Values are CFUs ± SEM within the entire organ as determined by serial dilutions of tissue homogenates. *P < 0.05 and **P < 0.01 versus WT (B and C) or F12^−/− (D–G). P values were determined using Student’s t test (B and C) or one-way ANOVA (D–G).

Figure 3.

Increased S. aureus–induced cutaneous infection in plasma kallikrein–deficient mice. WT and Klkb1^−/− mice were injected subcutaneously with 1 × 10⁹ CFU S. aureus bioluminescent strain Xen29 (n = 10 per group). (A) Representative S. aureus in vivo bioluminescence on a pseudocolor scale overlaid on top of a greyscale image of WT and Klkb^−/− mice. (B) Bacterial counts as measured by in vivo bioluminescence of S. aureus demonstrated as mean total flux (photons per second) ± SEM in a logarithmic scale. (C) Mean total size of the skin lesion in cm² ± SEM. (D–G) Bacterial burden observed in kidneys (D), spleens (E), lungs (F), and livers (G) of WT and Klkb1^−/− mice at 5 days after infection compared with F12^−/− mice (dashed lines). Values are CFUs ± SEM within the entire organ as determined by serial dilutions of tissue homogenates. *P < 0.05, **P < 0.01 versus WT mice. P values were determined using Student’s t test.

Figure 4.

Defective S. aureus polyP-mediated fibrin formation and impaired fibrous abscess capsule integrity in F12 ^−/− mice. (A–H) WT, F12^−/− mice were subcutaneously inoculated with 3 × 10⁶ CFU S. aureus, and abscesses were removed 30 h after infection. Representative photographs of skin sections stained with (A–D) H&E or (E–H) fibrin-specific antibody 59D8 are shown. Dashed lines and black arrows denote abscess capsule and fibrin staining around the abscess, respectively. Scale bars represent 50 µm. (I) Quantification of fibrin staining given in % relative to WT mice. (J) Integrity of abscess capsule measured from immunohistochemistry sections by gaps within the discontinuous fibrin layer of the abscess capsule given relative to the abscess perimeter. (K) Number of S. aureus detected in the peripheral zone of the abscess given in relative to bacteria counts detected in WT mice, set to 100%. Images were analyzed with ImageJ software. Each symbol represents an individual animal. (L) PolyP was extracted from S. aureus by anion exchanger chromatography, separated on polyacrylamide/urea gel and visualized by DAPI-negative staining. Synthetic polyP with mean chain lengths of 39, 97, and 383 phosphate monomers served as molecular size standard. Purified polyP was loaded without and after incubation with PPX (10 U/ml for 2 h). A representative gel of n = 3 is shown. (M) Real-time thrombin generation in normal human plasma stimulated with S. aureus–derived polyP in the absence or presence of PPX_Δ12 (1 mg/ml). Plasma deficient in FXII (FXII def.) or FXI (FXI def.) and buffer-stimulated normal plasma was blotted for comparison. Representative thrombin generation curves of a series of n = 3. (N) Citrated whole mouse blood, readjusted to physiological Ca²⁺ and Mg²⁺ concentrations, was perfused over a surface coated with 3 × 10⁶ CFU S. aureus at a venous (100 s⁻¹) shear rate. Representative phase-contrast images of thrombi formed during perfusion of WT, F12^−/−, F11^−/−, or PPX_Δ12 -treated WT blood are shown. Surface covered area by all thrombi in % is given in the lower right corner. A representative experiment of n = 3 is shown. Source data are available for this figure: SourceData F4.

Figure 5.

FXII-FXI–driven coagulation traps S. aureus in skin infection. (A) WT, F11^−/−, or F12^−/−/F11^−/− mice were injected subcutaneously with 1 × 10⁹ CFU S. aureus bioluminescent strain Xen29 (n = 10 per group). Representative S. aureus in vivo bioluminescence on a pseudocolor scale overlaid on top of a greyscale image of WT, F11^−/−, or F12^−/−/F11^−/− mice are shown. (B) Bacterial counts as measured by in vivo bioluminescence of S. aureus demonstrated as mean total flux (photons per second) ± SEM in a logarithmic scale. (C) Mean total size of the skin lesion in cm² ± SEM. *P < 0.05, **P < 0.01 F11^−/− versus WT mice and ^#P < 0.05, ^##P < 0.01 F12^−/−/F11^−/− versus WT mice in A. (D–G) Bacterial burden observed in kidneys (D), spleens (E), lungs (F), and livers (G) of WT, F11^−/−, and F12^−/−/F11^−/− mice compared with F12^−/− mice (dashed lines) at day 5 after infection. Values are CFUs ± SEM within the entire organ as determined by serial dilutions of tissue homogenates. *P < 0.05 versus WT mice. P values were determined using Student’s t test.