Factor XI Deficiency Reduces The Inflammatory Response To Polymicrobial Sepsis In Mice

Sepsis is a systemic host inflammatory response to entry and dissemination of an infectious agent. Interconnected pathways including inflammation, coagulation, and complement that serve protective functions when activated locally, contribute to morbidity and mortality when activated systemically during sepsis. Factor XIa (fXIa), the active form of the zymogen factor XI (fXI), contributes to thrombin generation in response to tissue injury. Prior work using a model of polymicrobial sepsis induced by cecal ligation and puncture (CLP) in mice suggests that fXI contributes to the pathology of sepsis by promoting disseminated intravascular coagulation (DIC) and inflammation (Tucker et al. J Infect Dis 2008;198:271 and Blood 2012;119:4762). We investigated the role of fXI in the CLP model, concentrating on the first 24 hours post-injury, using a high-level ligation that affects ≥75% of the length of the cecum. This level of injury causes near-100% mortality in wild type (WT) C57Bl/6 mice (median survival 42 hrs). Survival 7 days post-CLP was significantly greater for fXI-deficient (fXI^-/-, 30% p=0.002) and partially deficient (fXI^+/-, 17% p=0.003) mice compared to WT littermates (3%). For animals dying after CLP, median time to death was similar in the three groups. Four hours post-CLP, there were peaks in TNFα and IL-10 plasma levels in WT mice that were significantly smaller in fXI^-/- mice (TNFα 76 ± 44 and 24 ± 22 pg/mL [p=0.009]; IL-10 10,000 ± 11,000 and 1,000 ± 900 pg/mL [p=0.0003], respectively), with levels also higher in WT mice 24 hrs post-CLP. There was significant white blood cell accumulation in the peritoneum 24 hrs post-CLP in WT and fXI^-/- mice, with the lymphocyte fraction significantly elevated only in WT mice (p=0.008). The major rodent acute phase protein serum amyloid P was 5-fold higher in WT mice than in sham-treated mice, while levels in fXI^-/-mice were indistinguishable from shams, 24 hrs post-CLP.

In contrast to the indicators of inflammation, evidence for DIC was not strong for either genotype 24 hours post-CLP. Platelet counts were reduced similarly in both groups from a baseline mean of ∼600,000/μL, but mean counts remained above 300,000/μL. In both groups, plasma thrombin-antithrombin complex levels did not increase above baseline over the first 24 hrs post-CLP, and fibrinogen levels, which often decrease during DIC, increased modestly. No evidence of gross hemorrhage was noted at necropsy 4, 8, or 24 hrs post-CLP, and there was no histologic evidence of microvascular thrombosis (considered a “gold standard” for diagnosing DIC in laboratory animals) in kidney, liver or brain.

The contact system zymogens factor XII (fXII) and prekallikrein (PK) are activated on bacterial surfaces during infection, resulting in cleavage of high molecular weight kininogen (HK) to produce antimicrobial peptides and the pro-inflammatory peptide bradykinin. We measured plasma fXII, PK, and fXI using western blots and densitometry. FXI decreased by ∼30% on average 24 hrs post-CLP in WT mice, consistent with protease activation. PK levels decreased by ∼50% in WT mice 24 hrs post-CLP (p 0.007), but were unchanged in fXI^-/- mice. FXII levels decreased slightly in WT mice post-CLP, but increased in fXI^-/-mice 4 hrs after CLP, returning to baseline 8 hrs post-CLP. FXI, PK and fXII levels were unchanged in sham-treated animals. The data suggest that fXI contributes to contact activation after CLP, and its absence leads to reduced PK and fXII activation.

Previously, we reported that fXI deficiency reduced mortality after CLP (Tucker et al. J Infect Dis 2008;198:271) and infection with the gram-positive organism Listeria monocytogenes (Luo et al., Infect Immun. 2012;80:91), and proposed that this was in part related to down-regulation of coagulation, mitigating consumptive coagulopathy. The data presented here suggest that the high-level injury employed in the CLP-model produces an intense early systemic inflammatory response that leads to early death, before changes characteristic of DIC become obvious. The survival benefit conferred by fXI deficiency in this model then, may be primarily due to early modulation of inflammation, perhaps by altering contact activation, rather than to a prominent antithrombotic effect. Given the relatively mild bleeding diathesis associated with fXI deficiency, pharmacologic inhibition of fXIa may be a safe option for reducing the inflammatory response in septic patients.