Simultaneous activation of complement and coagulation by MBL-associated serine protease 2

Abstract

The complement system is an important immune mechanism mediating both recognition and elimination of foreign bodies. The lectin pathway is one pathway of three by which the complement system is activated. The characteristic protease of this pathway is Mannan-binding lectin (MBL)-associated serine protease 2 (MASP2), which cleaves complement proteins C2 and C4. We present a novel and alternative role of MASP2 in the innate immune system. We have shown that MASP2 is capable of promoting fibrinogen turnover by cleavage of prothrombin, generating thrombin. By using a truncated active form of MASP2 as well as full-length MASP2 in complex with MBL, we have shown that the thrombin generated is active and can cleave both factor XIII and fibrinogen, forming cross-linked fibrin. To explore the biological significance of these findings we showed that fibrin was covalently bound on a bacterial surface to which MBL/MASP2 complexes were bound. These findings suggest that, as has been proposed for invertebrates, limited clotting may contribute to the innate immune response.

Figure 1. Prothrombin fragmentation.

A) Prothrombin fragments generated during activation by factor Xa and cleavage by thrombin as would be observed on SDS-PAGE under reducing conditions (based on ref (5)). The arrows indicate the cleavage sites for factor Xa (fXa1 Arg273-Thr274 and fXa2 Arg322-Ile323) and thrombin (*1 Arg155-Ser156 and *2 Arg286-Thr287). The molecular mass of the fragments shown was calculated based on the primary sequence (Swiss-Prot entry P08709). B) The observed cleavage pattern of prothrombin incubated with factor Xa (FXa) (lane 1) or trMASP2 (lane 2) for 4 h at 37°C and analyzed by reducing SDS-PAGE and Coomassie Blue staining. The bands were identified by N-terminal sequencing together with theoretical and observed size. The 33.8 kDa fragment migrates faster than the 31.5 kDa fragment due to differences in the number of N-linked glycosylations. C) The sites of cleavage by factor Xa and trMASP2 are identical as confirmed by N-terminal sequencing as described in the text.

Figure 2. Thrombin activity generated by factor Xa and trMASP2 measured by cleavage of VPR-AMC.

Samples (100 µl) contained 10 µg prothrombin (PT)+25 ng factor Xa (fXa), 25 ng factor Xa alone, 10 µg prothrombin+100 ng trMASP2, 100 ng trMASP2 alone or 10 µg prothrombin alone. As can be observed the cleavage of the substrate by factor Xa, trMASP2 and prothrombin is negligible, but incubation of prothrombin with factor Xa or trMASP2 generates activity. The figure shows the development of the relative fluorescence at different time points. A typical result from 3 experiments is shown.

Figure 3. VPR-AMC cleavage (A), fibrin aggregation (B) and fibrinogen cleavage and cross-linking (C) by thrombin generated by mannan-bound MBL/MASP2K.

In fig 3A, thrombin generation is shown by cleavage of VPR-AMC. Column 1 shows the VPR-AMC turn-over when MBL/MASP2K was incubated with mannan in the presence of Ca²⁺ ions, followed by addition of prothrombin. Column 2 is the control for column 1 where MBL/MASP2K complexes were bound in the presence of EDTA. Column 3 is a further control with no rMASP2 added, and Column 4 is the same as column 1, but without prothrombin. The figure shows the cleavage of VPR-AMC by thrombin (as relative fluorescence) after 4h at 37°C. Fig 3B shows the MBL/MASP2K dose-dependent fibrin polymerization in a microtiter well after incubation with prothrombin and fibrinogen for 3.5 hours. The X-axis shows a 2-fold dilution series of the MBL/MASP2K complexes starting at 200 ng/well. The background was defined as the activity from a sample in which MBL/MASP2K complexes were prevented from binding by the presence of EDTA and subtracted. Fig 3C shows generation of fibrinogen γ chain dimer and α chain oligomers. Sample numbers correspond to those in fig 3A. Only in lane 1 have the γ chain dimer as well as α chain oligomers been formed. This arises from fibrinogen activation followed by factor XIIIa mediated cross-linking. This occurs only in the presence of activated MBL/MASP2K complexes bound to mannan, and in the presence of prothrombin. A typical result from 3 experiments is shown.

Figure 4. Reducing SDS-PAGE of prothrombin incubated at 37°C for 4 hours alone or with activated C1s, activated C1r,C1s mixture or factor Xa.

Neither C1s alone or C1r, C1s mixtures are capable of cleaving prothrombin. Factor Xa is included as a positive control.

Figure 5. Fibrin deposition on MBL target surface.

Fig 5A shows ¹²⁵I-fibrin deposition on a mannan/fibrinogen coated surface to which rMBL/rMASP2K complexes are bound. The background (a control in which rMBL/rMASP2K complexes were prevented from binding to the surface due to the presence of EDTA) has been subtracted from the remaining samples. The figure shows that deposition above background level only occurs if rMASP2K is present and if the incubation with fibrinogen is in the presence of prothrombin. Incubation was for 4h at 37°C. Fig 5B ¹²⁵I-fibrin deposition on S. aureus-derivatized beads. S. aureus-derivatized beads were incubated at 4°C with rMBL/rMASP2 complexes and subsequently with prothrombin and ¹²⁵I-fibrinogen for 7 h at 37°C. The figure shows that fibrin gets deposited on the beads upon activation of prothrombin by rMBL/rMASP2 complexes (columns 2 and 5, closed bars). The fibrin bound to the surface also gets covalently cross-linked to the bacteria since it was not removed by washing the beads with urea (column 2, open bar). If the deposition was done in the presence of iodoacetamide (IAM), which inhibits factor XIIIa, the radioactivity associated with the beads was reduced to background level upon urea extraction (column 5, open bar). In the controls in which no rMASP2 was added or in the presence of protease inhibitors (C1 inhibitor or Pefabloc SC) no deposition on the beads could be observed. Each sample was tested twice and the error bars represent one standard deviation from the mean.