A complete FXII deficiency does not affect coagulopathy, inflammatory responses, and lethality, but attenuates early hypotension in endotoxemic mice

The contact coagulation system is activated in lethal endotoxemia, and levels of coagulation Factor XII (FXII) decline in the consumptive phases of this disease. Thus, we examined the effects of a total inactivation of the endogenous F12 gene in the major phases of endotoxemia using a continuous infusion lethal dose model of lipopolysaccharide (LPS) administration in mice. Blood levels of coagulation and inflammatory parameters were measured, along with BK levels and BP parameters.

For this study, we employed 7–8 wk of age male wild-type (WT) mice and mice with a total targeted inactivation of the FXII gene (F12^−/−) that were fully backcrossed into the C57BL/6 strain [1]. These mice were subjected to continuous challenge with LPS (Escherichia coli; Serotype O111; B4, Sigma, St. Louis, MO) in saline (5 μg/ml) that was placed in osmotic pumps and the latter implanted into the peritoneal cavities [2]. The pumps release their contents at the rate of 1 μl/hr. Under these conditions, WT mice showed approximately 50% lethality at 96 hr. No survival differences were observed between WT and F12^−/− mice under these conditions (Figure 1A).

Figure 1.

responses of WT and F12^−/− mice to LPS challenge. (A) Survival rates of WT and F12^−/− mice after implantation of LPS containing osmotic pumps. The solid line depicts the survival curve of WT mice (N=19). The survival at 96 hr was 52.6% of the mice administered LPS. The broken line indicates the survival curve of F12^−/− mice (N=30), where 53.3% of the mice survived at 96 hr. The difference of the survival rates between the genotypes was compared using the log-rank and Wilcoxon tests, and no statistical differences were detected. (B) Plasma BK levels after LPS challenge. Plasma levels of BK in WT (black bars, N = 6 mice at each timepoint) and F12^−/− mouse plasmas (grey bars, N = 6 mice at each timepoint) at various times after initiation of LPS challenge. * p < 0.05. The minimum detectable concentration was 10 pg/ml, which is the level indicated by the gray bars. (C) Heart rates after LPS challenge. (D) Mean arterial blood pressures after LPS challenge. The black bars indicate WT mice (N=9) and the gray bars depict F12^−/− mice (N=9). Pairwise comparisons with statistical significance are indicated as * (p<0.05).

Coagulation parameters and inflammation markers were then determined on 6–12 mice at 0 time, and at 3, 6, 9, 12, 18, and 24 hr post-LPS challenge. ELISA-based measurements of thrombin-antithrombin (TAT) levels in WT and F12^−/− mice were evaluated with a TAT ELISA Kit (ERL, South Bend, IN, USA) during continuous LPS challenge. The plasma levels of TAT in F12^−/− mice (76 ± 16 pM) were significantly (p < 0.05, pairwise comparison) lower than those same values in WT mice (125 ± 15 pM) under resting conditions (time = 0). The TAT levels rapidly increased to 276 ± 67 pM for WT and 308 ± 44 pM for F12^−/− mice at 3 hr post-LPS, and progressively decreased at later time points. TAT levels reached baseline at approximately 12 hr, and declined further to approximately one-half of the pre-LPS challenge values at 24 hr, likely because of a consumptive coagulopathy. Plasma fibrinogen levels, determined with the Fibri-Prest Automate Kit (Diagnostica Stago, Asnieres-Sur-Seine, France) continuously increased during the challenge, from 197 ± 12 at time = 0 to 440 ± 21 at 24 hr subsequent to LPS challenge for WT mice and 197 ± 5 at rest to 536 ± 55 at 24 hr post-LPS challenge, for F12^−/− mice. This likely reflects an acute phase response. Platelet counts, analyzed using an automated CBC analyzer continually decreased from 687 × 10³/μl to 303 ± 40 × 10³/μl for WT mice and 683 ± 28 × 10³/μl to 284 ± 66 × 10³/μl during this same 24 hr time period. Thus, a significant coagulopathy was found in each genotypic group under this LPS dosage regimen, but no significant differences were found between the WT and F12^−/− groups in these markers at any post-LPS timepoint examined.

Select inflammation markers, viz., tumor necrosis factor-alpha (TNF-α), interleukin-6 (IL-6), and macrophage inflammatory protein-2 (MIP-2), were measured in plasmas of WT and F12^−/− mice at various times during continuous LPS challenge using Quantikine-M Immunoassay kits (R&D Systems, Minneapolis, MN, USA). Plasma levels of TNF-α in resting WT (12.7 ± 1.5 pg/ml) and F12^−/− (9.6 ± 0.3) mice significantly increased to 408 ± 51 pg/ml and 357 ± 66 pg/ml, respectively, at 3 hr post-LPS, and gradually declined through the next 24 hr, but did not reach resting levels at that later timepoint. Similarly, while IL-6 was undetectable in plasmas of both resting WT and F12^−/− mice, the peak values of 37.6 ± 1.6 ng/ml and 36.1 ± 2.5 ng/ml, respectively, at 3 hr, also gradually declined through 24 hr, but remained above resting levels in each case. Lastly, the MIP-2 levels paralleled the IL-6 data. MIP-2 was undetectable in plasma of WT and F12^−/− mice, reached peak levels at 3 hr post-LPS (51.6 ± 6.3 ng/ml for WT and 41.8 ± 6.1 ng/ml for F12^−/− mice), and gradually declined to just above detection levels at 24 hr after LPS administration. No statistical differences were found between the WT and F12^−/− groups in these particular inflammatory markers at any timepoint.

Plasma bradykinin (BK) levels were evaluated at various times post-LPS by RIA [1]. The plasma levels of BK in WT mice were significantly higher those in F12^−/− mice under resting conditions (Figure 1B), the latter of which were below levels of detection at all challenge times. This further illustrates the importance of the FXII-dependent pathway in BK generation. The BK levels in WT mice were increased at 3 hr and 6 hr after challenge, and then diminished after 9 hr (Figure 1B). On the other hand, the BK levels in F12^−/− mice were below detection in all time points, and did not change over time (Figure 1B).

Because of the lack of response of BK to LPS-administration in F12^−/− mice, we examined heart rates (HR) and central arterial BPs of the mice during endotoxemia [3]. HRs of both WT and F12^−/− mice were slightly elevated from 0 to 2 hr after LPS-containing pump implantation, and then decreased from 2–6 hr, reaching a nadir of approximately 500 bpm (Figure 1C). Although the drop in heart rates of F12^−/− mice was less severe than WT mice from 2–6 hr, statistical significance between the groups was not achieved. The mean arterial BPs (MAP) of WT mice showed patterns similar to their heart rates (Figure 1D). On the other hand, such drops of the MAP of F12^−/− mice were not observed from 2–6 hr, and the differences between WT and F12^−/− mice were statistically significant. After these initial periods, the parameters gradually decreased throughout the time period measured (Figure 1D).

Recent studies have indicated that activation of the FVII/TF pathway [4, 5], and downregulation of the anticoagulant Protein C pathway [2, 6–8], most likely play key roles in spontaneous bleeding and induction of coagulopathy in inflammatory diseases, such as endotoxemia, while the contact activation pathway does not appear to be as important in provoking or further establishing DIC in sepsis [9, 10] despite the pronounced activation of this system in human meningococcal disease [11] and in experimental models of bacteremia in baboons [12]. These conclusions correlate with the findings that patients with diminished levels of contact factors do not display spontaneous hemostasis defects in the absence of challenge. However, protection from fibrin and thrombus formation is afforded in various challenge models by deficiencies of components of the contact system, e.g., FXII [13, 14] and kininogen-1 [15].

In summary, the absence of FXII did not contribute to the coagulopathy, inflammatory changes, and mortalities of endotoxemic mice, but the absence of FXII attenuated early hypotensive changes that correlated with BK stimulation. Therefore, we conclude that the initiation of coagulopathy in endotoxemia is primarily triggered by the extrinsic pathway. Activation of the contact pathway in endotoxemia did not substantially contribute to the coagulopathy and inflammatory changes observed, but played a major role in hemodynamic changes via BK production.