Factor XI contributes to myocardial ischemia-reperfusion injury in mice

Factor XI (FXI) is the zymogen of a plasma protease (FXIa) that serves a limited role in formation of normal blood clots at sites of injury (hemostasis).^1,2  In vitro, FXI is converted to FXIa by the enzymes thrombin or factor XIIa (FXIIa).¹  Given the absence of abnormal bleeding in FXIIa-deficient individuals,²  FXI activation by thrombin is likely more relevant for hemostasis. FXIIa-dependent FXI activation, while not required for hemostasis, may contribute to coagulation during the host-response to various inflammatory stimuli.^3-5 

In humans, plasma FXI levels are associated with risk for venous thromboembolism and ischemic stroke.^6-12  Some data also support a link with myocardial infarction.^13-15  Work with rodent and primate models support an important role for FXI in thrombosis,^16-18  and strongly suggest that FXIIa-mediated FXI activation is a contributor.^17,18  In addition to its effects on thrombus growth, FXI contributes to inflammation. In mice, FXI deficiency or inhibition blunts the cytokine response to certain types of infections,^3-5  improving survival, and reduces cerebral ischemia-reperfusion injury after transient middle-cerebral artery occlusion.^19,20 

Preclinical and epidemiologic data suggest that interfering with the FXIIa-FXI interaction could inhibit thromboinflammatory processes without compromising hemostasis.^3-7  In support of this hypothesis, an anti-FXI antibody (14E11) that preferentially interferes with FXI activation by FXIIa is antithrombotic in mice and baboons^17-20  and reduces the cytokine response during polymicrobial sepsis in mice.^4,5  The dual roles of FXI in thrombosis and inflammation suggest that inhibiting the FXIIa-FXI interaction may be beneficial in acute myocardial infarction (MI). Using mice, we investigated whether inhibiting FXI activation with 14E11 reduces myocardial ischemia-reperfusion injury.

Human FXI and FXIa were from (Haematologic Technologies, Essex Junction, VT). Human FXII was from Enzyme Research Laboratory (South Bend, IN). Genomic DNA was isolated from human leukocytes. Generation and purification of the monoclonal anti-FXI antibody 14E11 has been reported.¹⁷ 

Purified human FXII (200 nM) and FXIa (10 nM) or FXI (30 nM) in the absence or presence of 5 µg/mL DNA were incubated with 14E11 (5-200 nM) in 20 mM N-2-hydroxyethylpiperazine-N′-2-ethanesulfonic acid, pH 7.4, 100 mM NaCl, 0.1% polyethylene glycol 8000, and ZnCl₂ (10 μM) at 37°C for 60 min. FXII activation was stopped by aprotinin (10 μM), and ΔOD_405nm in the presence of S-2302 (500 µM) was monitored. FXI activation was terminated by mixing aliquots with polybrene (0.2 mg/mL) and S-2366 (1 mM), and ΔOD_405nm was monitored. Results were compared with a standard curve prepared with purified FXIa.

We used an established murine model of nonthrombotic acute MI,^21,22  with minor modifications. Male C57BL/6 mice (17-20 weeks) were intubated and ventilated under isoflurane anesthesia, the jugular vein was catheterized, and body temperature was maintained at 37°C. Lead II electrocardiogram (ECG) was monitored. After thoracotomy, the left anterior descending artery (LAD) was gently ligated near the origin with a suture loop. Ischemia was verified by pale left ventricular myocardium (LVM) surface and ST-segment elevation. After 35 minutes, surviving mice received phosphate-buffered saline (100 µL) with or without 1 mg/kg 14E11 intravenously (n = 10, each). At 40 minutes, the ligature was loosened to allow LVM reperfusion, verified by ECG and LVM color.

After 2 hours of reperfusion, the LAD was religated, and an 800-µL suspension of fluorescent microspheres (4% [weight/volume], 2-8 µm diameter, in water, 0.01% polysorbate-80; Duke Scientific, Palo Alto, CA) were infused over 4 minutes into the LVM apex to occlude vessels in perfused capillary beds surrounding the zone of ischemia, without entering the ischemic tissue itself, to identify the area at risk (AAR) for infarction during LAD ligation.²² 

The heart was immediately excised, cut into 1-mm slices, and photographed under UV light to identify the AAR. Tissue slices were then stained for 10 min with 1% 2,3,5-triphenyltetrazolium chloride solution (TTC; Sigma) and placed in 10% neutral-buffered formalin. Myocardium not taking up TTC was considered infarcted. Tissue-section images were analyzed (Photoshop) in a blinded fashion to measure the infarcted region as a percentage of LVM. Infarct volume was normalized to the AAR in the corresponding slides. Statistical significance was calculated (P ≤ .05) by analysis of variance, and these are expressed as the means ± standard error of the mean (SEM) (Prism GraphPad). All animal experiments were approved by the institutional animal care and use committee of Oregon Health & Science University.

FXI activation by FXIIa triggers coagulation in the aPTT assay used in clinical practice.²  Despite this, the reaction is not required for hemostasis, because total absence of FXIIa does not cause abnormal bleeding, even with trauma or surgery.²  However, evidence from animal models support a role for FXIIa activation of FXI in thromboinflammatory conditions, including ischemia-reperfusion injury in the central nervous system.^19,20  Our goal was to test the ability of a monoclonal immunoglobulin G (IgG) (14E11) that specifically interferes with the FXIIa-FXI axis, to reduce myocardial ischemia-reperfusion injury.

The 14E11 antibody binds to an epitope on the FXI A2 domain and cross-reacts with FXI from most mammalian species.¹⁷  It preferentially interferes with FXI activation by FXIIa, prolonging the aPTT of human and mouse plasmas ∼2.5-fold. FXI activation by thrombin and FXIa-catalyzed activation of factor IX are not inhibited by 14E11. In most plasma coagulation models the interaction between FXIIa and FXI is unidirectional (FXIIa activates FXI).²  However, evidence from sepsis models supports the premise that FXIa can activate FXII, the precursor of FXIIa.⁵  The 14E11 antibody has a modest inhibitory effect on FXII activation by FXIa (Figure 1A) and slows FXI autoactivation in the presence of a polyanion (DNA) (Figure 1B). This may be relevant for FXI activation during inflammatory processes.²³ 

The 14E11 antibody attenuates thrombus formation in mouse and primate models¹⁷  and improves survival in a mouse sepsis model.⁴  Interestingly, during sepsis, the beneficial effect of FXI inhibition/deficiency may be largely due to blunting of cytokine-mediated inflammation, rather than an effect on coagulation.⁵  FXI may also play a role in leukocyte migration.²⁴  FXI deficiency reduced vessel wall infiltration by leukocytes in atherosclerotic (apolipoprotein E deficient) mice.²⁵  These preclinical data indicate 14E11 has anti-inflammatory properties in addition to antithrombotic properties, making it an attractive agent to test in myocardial ischemia-reperfusion injury.

In the present study, LAD occlusion was confirmed by sustained ST segment elevation and by observing regional pallor and ventricular wall motion abnormalities consistent with focal perfusion deficits. During reperfusion, natural color returned to the LVM distal to the ligation. Analysis of AAR (Figure 2A; defined as the area devoid of microspheres normalized to ventricular volume) revealed no significant difference between mice treated with 14E11 or vehicle (Figure 2B; 33.0% ± 5.5% vs 34.0% ± 7.6% of ventricular volume, respectively; P = .86). There was a 39% reduction in the volume of infarcted tissue within the ischemic zone (AAR) in 14E11-treated animals in comparison with vehicle-treated controls (Figure 2C; 37.8% ± 4.2% vs 62.2% ± 7.5% of the AAR; P ≤ .0001). This protective effect is comparable to that observed previously for 14E11 in a model of cerebral ischemia-reperfusion injury.²⁰ 

Our results suggest that contact activation contributes to myocardial ischemia-reperfusion injury and that injury can be partially mitigated by inhibiting FXIIa activation of FXI. Given the mechanism of action of 14E11, this beneficial effect could be mediated by reducing contact activation, thrombin generation, inflammation, or any combination of these. An advantage of therapeutic targeting of the FXIIa/FXI axis is that, unlike currently available anticoagulants, it should not compromise hemostasis. Given this, it may be possible to use this type of therapy in addition to standard of care to improve outcomes.

The authors gratefully acknowledge Donna Van Winkle for providing access to her laboratory and equipment.

This work was supported by National Institutes of Health, National Heart, Lung, and Blood Institute grants HL095315 (A.G. and E.I.T.), HL117589 (E.I.T. and N.G.V.), HL128016 (A.G. and E.I.T.), HL101972 (O.J.T.M. and A.G.), and HL081326 (D.G.).

Contribution: C.U.L., N.G.V., and A.G. designed the study, interpreted the data, and wrote the manuscript; L.L., I.I., and Z.C. executed the experiments; A.G., D.G., E.I.T., M.T.H., and O.J.T.M. oversaw the project, critically reviewed the data, and wrote the manuscript.

Conflict-of-interest disclosure: A.G., C.U.L., E.I.T., and N.G.V. are employees of Aronora, Inc., a company that may have a commercial interest in the results of this research. A.G., E.I.T., and Oregon Health & Science University (OHSU) have a financial interest in Aronora, Inc. This potential conflict of interest has been reviewed and managed by the OHSU Conflict of Interest in Research Committee. The remaining authors declare no competing financial interests.

Correspondence: András Gruber, Departments of Biomedical Engineering and Medicine, CH13B, Oregon Health & Science University, School of Medicine, 3303 SW Bond Ave, Portland, OR 97239; e-mail: grubera@ohsu.edu.