Prevalence and Clinical Impact of Reduced Coagulation Factor XII Activity in Patients Receiving Extracorporeal Membrane Oxygenation.

OBJECTIVES: Extracorporeal membrane oxygenation provides large surface exposure to human blood leading to coagulation activation. Only limited clinical data are available on contact activation and coagulation factor XII activity in extracorporeal membrane oxygenation patients. DESIGN: Prospective cohort study. SETTING: Three medical ICUs at the Medical University of Vienna. PATIENTS: Adult patients receiving venovenous or venoarterial extracorporeal membrane oxygenation. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: The primary outcome was the change in coagulation factor XII activity in response to extracorporeal membrane oxygenation. Secondary outcomes included the prevalence of reduced coagulation factor XII activity (< 60%) among patients receiving extracorporeal membrane oxygenation and association of coagulation factor XII activity with thromboembolic and bleeding complications. An exploratory endpoint was the association of coagulation factor XII activity and activated partial thromboplastin time in heparinase-treated samples in vitro. Fifty-one patients with a total of 117 samples were included in the study between July 2018 and February 2020. Fifty patients (98%) had reduced coagulation factor XII activity at any timepoint during extracorporeal membrane oxygenation. Median coagulation factor XII activity during extracorporeal membrane oxygenation treatment was 30% (interquartile range, 21.5-41%) and increased after discontinuation (p = 0.047). Patients with thromboembolic complications had higher median coagulation factor XII activity during extracorporeal membrane oxygenation (34% vs 23%; p = 0.023). The odds of a thromboembolic event increased by 200% per tertile of median coagulation factor XII activity (crude odds ratio, 3.034; 95% CI, 1.21-7.63). No association with bleeding was observed. In heparinase-treated samples, coagulation factor XII activity correlated well with activated partial thromboplastin time (r = -0.789; p = 0.007). CONCLUSIONS: We observed a high prevalence of reduced coagulation factor XII activity in adult patients on extracorporeal membrane oxygenation, which may confound activated partial thromboplastin time measurements and limit its clinical usefulness for monitoring and titrating anticoagulation with unfractionated heparin. Lower coagulation factor XII activity was associated with less thromboembolic complications, which may highlight the potential of coagulation factor XII to serve as a target for anticoagulation in extracorporeal membrane oxygenation.

Synthesis and release of Hageman factor (Factor XII) by the isolated perfused rat liver.

The site of synthesis of Hageman factor (HF, Factor XII) has not been previously demonstrated with certainty. We have studied the production and release of HF in the isolated perfused rat liver and have compared rates of synthesis in this system with absolute rates of degradation measured in vivo. Rat livers, perfused for 5 h with a recycling fluid consisting of a perfluorochemical emulsion (Fluosol 43), were used to demonstrate a cumulative increase of HF in the perfusate as measured by a specific and sensitive radioimmunoassay. The rate of increase in the perfusate pool of HF during the final 4 h of perfusion yielded a mean synthetic rate of 3.5 micrograms/h per 100 g body wt, which was approximately 0.2% of the synthetic rate of albumin in the same system. The cumulative appearance of albumin and transferrin was linear after 1 h and calculated rates of synthesis were 2,012 micrograms/h per 100 g and 263 micrograms/h per 100 g body wt, respectively. De novo synthesis of HF was confirmed by demonstrating incorporation of [14C]lysine into specific immunoprecipitates of HF, and by the observations that both specific incorporation of labeled amino acid and net release of immunoassayable HF were inhibited by the administration of cycloheximide. Finally, it was evident that the rates of synthesis observed in the isolated perfused liver agreed closely with absolute rates of degradation of HF measured in vivo with 125I-rat HF (4.0 micrograms/h per 100 g). From these data we conclude that the liver is the principal site of synthesis of HF.

The intrinsic coagulation-kinin pathway, complement cascades, plasma renin-angiotensin system, and their interrelationships.

Activation of the classical complement pathway is initiated by immune complexes consisting of IgM antibody or IgG subclasses 1, 2, and 3. Binding to Clq leads to activation of C1s and digestion of C4 and C2 to yield a C3 convertase. The alternative complement pathway is initiated by complex polysaccharides as well as immune complexes of the IgA class which interact with Factors B, D, C3, and properdin to yield a stabilized C3 convertase consisting of PC3Bb. Cleavage of C3 and C5 by either pathway yields the C3a and C5a anaphylatoxins which cause histamine release from mast cells and formation of the C5b6789 attack complex causes cell lysis. Both immunologic and nonimmunologic tissue damage can initiate the surface dependent pathways of coagulation, fibrinolysis, and kinin formation. Surface bound Hageman Factor interacts with complexes of prekallikrein and HMW-kininogen as well as Factor XI and HMW-kininogen to form activated Hageman factor, kallikrein, and Factor XIa. Factor XIa continues the coagulation pathway, kallikrein and Factor XIa convert plasminogen to plasmin and kallikrein digests HMW-kininogen to yield bradykinin. The Cl inhibitor, which inactivates Cls is the major plasma inhibitor of activated Hageman factor and kallikrein. In its absence, a potentially fatal form of angioedema is seen. The inactivator of the C3a and C5a anaphylatoxins is identical to carboxypeptidase N, the major plasma inactivator of bradykinin thus demonstrating the common control mechanisms which regulate the complement and kinin-forming pathways.

Orphan receptor hepatocyte nuclear factor-4 antagonizes estrogen receptor alpha-mediated induction of human coagulation factor XII gene.

Factor XII (FXII) is a liver-specific zymogen involved in the regulation of hemostasis, particularly in the activation of fibrinolysis. Transcription of the FXII gene is stimulated by estrogens through specific interaction of the estrogen receptor alpha (ER alpha) with an estrogen response element present on FXII promoter. Interestingly, the magnitude of ER alpha induction in liver HepG2 cells is much lower than in NIH3T3 fibroblasts, suggesting that cell-specific factors may modulate ER alpha-dependent trans-activation. Comparative footprinting analysis of FXII promoter (from nucleotides -181 to +49) in liver vs. non-liver cell environments allowed identification of four deoxyribonuclease I-protected sites only in the presence of HepG2 nuclear extracts. Computerized homology search identified sites III and IV as consensus binding sequences for the liver-enriched transcription factor hepatocyte nuclear factor-4 (HNF-4), formerly an orphan receptor belonging to the superfamily of steroid/thyroid hormone nuclear receptors. In transient transfection assays in NIH3T3 cells, HNF-4 significantly inhibited (70%) estrogen induction of FXII promoter while not affecting basal promoter activity. Conversely, HNF-4 did not inhibit estrogen inducibility of FXII promoter in HepG2 cells due to the high endogenous levels of HNF-4 protein. In gel shift assays, HNF-4, either present in HepG2 nuclear extracts or generated by in vitro transcription/translation, specifically bound FXII promoter. This interaction is strictly required in eliciting the antagonistic effect because in NIH3T3 cells, selective mutations of sites III and IV abrogated HNF-4 inhibitory properties. In the liver-specific environment, the same mutant construct exhibited higher estrogen-dependent inducibility compared with native promoter. Rescue of estrogen responsiveness was also achieved using a dominant negative HNF-4, which counteracted endogenous HNF-4 activity. In conclusion, our findings address a direct role for HNF-4 in modulating estrogen-dependent transcription of the FXII gene promoter.

The effect of natural habituation on coagulation responses to acute mental stress and recovery in men.

Blood coagulation activation might be one mechanism linking acute mental stress with coronary events. We investigated the natural habituation of coagulation responses and recovery to short-term mental stress. Three times with one-week intervals, 24 men (mean age 47 +/- 7 years) underwent the same 13-min stressor (preparation, job interview, mental arithmetic). During each visit venous blood was obtained four times (baseline, immediately post-stress, 45 min of recovery, 105 min of recovery). Eight blood coagulation parameters were measured at weeks one and three. Acute stress provoked increases in von Willebrand factor antigen, fibrinogen, clotting factor FVII activity (FVII:C), FVIII:C, FXII:C (p's < or = 0.019), and D-dimer (N.S.). All coagulation parameters experienced full recovery except FVIII:C (p = 0.022). Stress did not significantly affect activated partial thromboplastin time and prothrombin time. At all time points FVIII:C and FXII:C levels were significantly higher at week one compared to week three (p's < or = 0.041). Before catheter insertion, systolic blood pressure (p = 0.001) and heart rate (p = 0.026) were relatively higher at week one. Unlike the magnitude of systolic blood pressure response to stress (p = 0.007) and of cortisol recovery from stress (p = 0.002), the magnitude of all coagulation responses to stress and the recovery from stress were similar in week one and week three. Sympathetic activation with anticipatory stress best explained increased baseline activity in FVIII and FXII at week one. An incapacity of the coagulation system to adapt to stress repeats is perhaps a consequence of evolution, but might also contribute to increased coronary risk in some individuals, particularly in those with cardiovascular diseases.

Antibodies to factor XII: a possible predictive marker for recurrent foetal loss.

Antibodies to factor XII (FXIIabs) have been demonstrated in some patients with the anti-phospholipid syndrome (APS). The presence of these antibodies were shown to lead to statistically significantly reduced levels of FXII (p = 0.02). In an extension to this study forty female patients with either primary APS (n = 26) or systemic lupus erythematosus (APS positive) (n = 14) were investigated for levels of factor XII, the presence of lupus anticoagulant and antibodies to cardiolipin, beta 2-glycoprotein I and factor XII. Twenty one of the forty patients had a history of foetal loss (> 2, mean = 2.6). Lupus anticoagulant positivity showed a weak association with foetal loss (odds ratio = 1.1). While there was no association between the presence of antibodies to cardiolipin or beta 2-glycoprotein I with foetal loss, antibodies to factor XII showed a strong and statistically significant association (odds ratio = 5.4, p = 0.025).

Activation of factor XII and prekallikrein with polysaccharide sulfates and sulfatides: comparison with kaolin-mediated activation.

The activation of Factor XII and prekallikrein by polysaccharide sulfates and sulfatides in the presence of high-molecular-weight (HMW) kininogen was studied, and compared with the kaolin-mediated activation reaction. Among a variety of artificially-sulfated polysaccharides and native polysaccharide sulfates, amylose sulfate (M.W.= 380,000 and sulfur content, 19.1%) and sulfatide were found to have the most efficient ability to trigger the activation of prekallikrein by Factor XII. The effects of these two kinds of negatively-charged surfaces on the following three activation reactions were compared; the activation of prekallikrein by Factor XII (reaction 1), the activation of Factor XII by kallikrein (reaction 2) and the activation of prekallikrein by Factor XIIa (reaction 3). All three reactions mediated by the selected surfaces were strongly accelerated by HMW kininogen and its derivatives, kinin-free protein and fragment 1.2-linked light chain, like the kaolin-mediated activation. However, this accelerating effect of HMW kininogen on the amylose sulfate- and sulfatide-mediated activations (reaction 1) was diminished after treatment with fluorescein iso-thiocyanate, whereas the effect on the kaolin-mediated activation was not influenced by fluorescein-labeling. In addition, reaction 2 mediated by amylose sulfate and sulfatide was extremely slow even in the presence of HMW kininogen, and the results also differed from those with kaolin. The sulfatide-mediated activation of reaction 1 was not inhibited by fragment 1.2 (His-rich fragment), which is released from HMW kininogen by the action of kallikrein, and is known to be a potent inhibitor of the kaolin-dependent activation. These results indicate that the mechanisms responsible for surface activation triggered by soluble amylose sulfate, sulfatide micelles and kaolin differ from each other as regards the molecular interaction with the contact factors.

Hageman factor and its binding sites are present in senile plaques of Alzheimer's disease.

Hageman factor (HF) or factor XII participates in several defense systems of the body. These include coagulation, fibrinolysis and complement activation. We investigated the expression of HF and its mRNA in control and Alzheimer's disease (AD) brain, using immunohistochemistry and polymerase-chain reaction (PCR) techniques. HF mRNA was detected in control and AD brain extracts, indicating that HF can be produced by endogenous brain cells. HF-like immunoreactivity was present in residual serum of capillaries in both control and AD brain, consistent with its known presence in the circulation. In addition, AD senile plaques were stained. The staining was dramatically enhanced when AD sections were incubated with solutions containing HF, indicating that plaques contain not only HF but also binding sites for HF. The enhanced staining was eliminated by pretreatment of solutions with the HF-binding agent kaolin. It was also eliminated by pretreatment of sections with protamine, an agent which strongly binds to negative surfaces. These data suggest that negatively charged surfaces in plaques might bind HF in vivo. Since HF can be activated by contact with negative surfaces, locally released HF could be playing a role in initiating a variety of inflammatory responses in AD brain.

Whale Hageman factor (factor XII): prevented production due to pseudogene conversion.

In Southern blot analysis of the Hind III-digested whale genomic DNA obtained from the livers of two individual whales, we detected a single band with a size of five kilobase pairs which hybridized to full length guinea pig Hageman factor cDNA. We amplified two successive segments of the whale Hageman factor gene by polymerase chain reaction (PCR), and sequenced the PCR products with a combined total of 1367 base pairs. Although all of the exon-intron assemblies predicted were identical to those of the human Hageman factor gene, there were two nonsense mutations making stop codons and a single nucleotide insertion causing a reading frame shift. We could not detect any message of the Hageman factor gene expression by northern blot analysis or by reverse transcription-polymerase chain reaction (RT-PCR) analysis. These results suggest that in the whale, production of the Hageman factor protein is prevented due to conversion of its gene to a pseudogene. The deduced amino acid sequence of whale Hageman factor showed the highest homology with the bovine molecule among the land mammals analyzed so far.

The relationship between factor VII coagulant activity and factor XII activation induced in plasma by endogenous or exogenously added contact surface.

The contribution of various enzymes in the activation of factor VII, determined from the increase in factor VII coagulant activity (VIIc), was investigated following the exposure of citrated plasma to low temperature. The contact system of coagulation was initiated either by the contact surface present in certain plasmas (i.e. plasma from women in late pregnancy) or by micellar stearate added to plasma diluted with an equal volume of buffer (plasma from normal healthy subjects or from women in late pregnancy). With either of the contact surfaces, increase of VIIc and the concentration of enzymes derived from factor XII (XIIa) depended on the potency of the contact surface. The stearate-induced VIIc in diluted plasmas from women in late pregnancy or from normal subjects was inhibited by 60-70% in the presence of anti-factor IX monoclonal antibody. VIIc was not increased in XII-deficient plasma following the addition of stearate. The addition of purified human factor XII to this plasma restored the increase in VIIc and the activation of factor XII. In factor IX-deficient plasma, the stearate-induced increase in VIIc was only 38% of that seen in normal plasma and was restored by the addition of purified factor IX. Similarly in factor XI-deficient plasma, the stearate-induced increase in VIIc and the factor XII activation were 48% and 69% of that found in normal plasma. The addition of EDTA (2 mM) did not alter the extent of factor XII activation induced by contact surface, but it did inhibit the rise in VIIc. It is concluded that in the presence of contact surface the activation of factor XII and the sequential activation of factor XI and of factor IX results in the activation of factor VII. Activated factor IX is responsible for the major part of the factor VII activation whereas the rest may be through the direct activation by XIIa.

Immunohistochemical studies on synthesis and distribution of Hageman factor and kininogen.

Immunohistochemical localization of Hageman factor and high molecular weight kininogen were investigated in liver, skin and kidney of guinea pig using a new conjugation method with maleimide derivative as the coupling reagent at the level of light and electron microscopes. In the liver, the positive reactions of both factors were observed in the rough and smooth endoplasmic reticulums and the Golgi apparatus in hepatocytes. In Kupffer cells and sinusoidal endothelial cells, the reaction products were visible only in the endocytotic vesicles. These findings suggested that guinea pig Hageman factor and high molecular weight kininogen were synthesized only in hepatocytes at least the liver. In the skin, positive reactions were seen in the intercellular space of epidermis, and along the basement membranes of epidermis and vessels, and among the collagen fibers in interstitial tissue particularly in papillary dermis. This interstitial presence of these molecules was also seen in liver and kidney, which suggested that Hageman factor and high molecular weight kininogen were widely distributed in the interstitial tissue space even under normal conditions. In addition, positive reactions were obtained in the reabsorption vesicles and the lysosomes of the proximal convoluted tubules of kidney and in the pinocytotic vesicles of the basal cells of epidermis. According to these findings, it was suggested that both factors were produced in hepatocytes, secreted into the blood stream, distributed in the interstitial tissue by vascular permeability even under normal condition, and reabsorbed and catabolized in the kidney and the skin.

Role of HMW kininogen in surface-mediated activation of Factor XII.

We have shown that bovine HMW kininogen remarkably accelerates the activation of Factor XII and prekallikrein in the presence of kaolin, adsorbing on kaolin through the fragment 1.2 region and forming a complex with prekallikrein through the light chain region (Sugo et al., 1980; Ikari et al., 1981). The present study was undertaken to examine the role of HMW kininogen in the activation of Factor XII and prekallikrein with other negatively-charged surfaces. The activation system used here was as follows; (1) Activation of prekallikrein by Factor XII, (2) Activation of Factor XII by plasma kallikrein and (3) Activation of prekallikrein by Factor XIIa. Among a variety of foreign surfaces, amylose sulfate and sulfatide were the most efficient in the activation reaction of Factor XII and prekallikrein. Bovene HMW kininogen accelerated all the three reactions triggered by these surfaces. However, the accelerating effect of HMW kininogen on the activation of Factor XII by plasma kallikrein was very weak, when amylose sulfate or sulfatide was used as surface. The three reactions were highly dependent on the amounts of HMW kininogen and surfaces contained in the reaction mixtures. Excess amount of them inhibited these reactions. Among the various fragments, which were prepared from HMW kininogen digests with plasma and urinary kallikreins (Sugo et al., 1980), a large fragment consisting of fragment 1.2 and light chain accelerated the reactions. Thus both fragment 1.2 and the light chain region in HMW kininogen were essential for these activation reactions.

Factor XII (Hageman factor) is a missing link between stress and hypercoagulability and plays an important role in the pathophysiology of ischemic stroke.

A new hypothesis is presented on the function of factor XII, which is postulated to be a "missing link" between acute stress and transient hypercoagulability. The implications of this idea are developed to show how chronic stress, which involves activation of hypertension and migraine as well as hypercoagulability, can cause of cerebrovascular disease. "Acute stress" is defined as "the normal short-term physiological response to the perception of major threats or demands". "Chronic stress" is "the abnormal ongoing physiological response to the continuing perception of unresolvable major threats or demands". The factor XII hypothesis is as follows: Acute stress includes release of epinephrine by the adrenal medulla. Epinephrine activates platelets by binding to alpha-2A adrenergic receptors. Activated platelets convert pre-bound factor XII to its active form, which then initiates the intrinsic coagulation cascade. This can be called the "activated platelet initiation pathway" for coagulation. Neither tissue factor nor pre-formed thrombin is required. Thrombosis proceeds to completion, but only a minute amount of thrombin is formed, and the process normally stops at this point. In people who lapse into a state of chronic stress, essential hypertension, which is also a manifestation of stress, synergizes with hypercoagulability: there is both a baseline rise in blood pressure and systemic platelet activation as well as superimposed labile rises of both. Upregulation of these two stress parameters is atherogenic: epinephrine-activated platelets stimulating thrombin formation interact with endothelial cells activated by angiotensin II to cause, first, smooth muscle cell proliferation, which is a histological hallmark of atherosclerosis, and, lastly, a symptomatic thrombotic occlusion-the stroke. The migraine symptoms which often accompany this process are a marker of chronic stress and ongoing pathophysiologic damage. Therapeutic predictions are made regarding novel ways of blocking stress-induced hypercoagulability and hypertension. Hypercoagulability could be targeted by monoclonal antibodies directed against the platelet-specific alpha-2 adrenergic receptor or the (putative) platelet receptor for Factor XII; hypertension could be treated with monoclonal antibodies directed against the beta-adrenergic receptor in the juxtaglomerular apparatus or by surgical denervation of the kidneys, either of which would decrease the renin release which helps drive the hypertension.

Generation of coagulation factors V, XI, and XII by the isolated rat liver.

When isolated rat livers were perfused with platelet-free erythrocytes suspended in Tyrode's solution containing 6% bovine serum albumin, the generation of factors V, XI, and XII was clearly demonstrable. On average, after 5 h of perfusion of a single liver, the concentrations in the perfusate (as a percentage of normal rat plasma) were about 6% for factor V, 8% for factor XI, and 5% for factor XII, compared with 20% for factor VII, which was used as the reference standard. When two livers were perfused, approximately twice these concentrations were achieved. When the properties of these factors in perfusate and plasma were compared, they agreed well except for differences in the celite adsorbability of factors XI and XII.