A new pedigree with thrombomodulin‐associated coagulopathy in which delayed fibrinolysis is partially attenuated by co‐inherited TAFI deficiency

Thrombomodulin‐associated coagulopathy (TM‐AC) is a rare bleeding disorder in which a single reported p.Cys537* variant in the thrombomodulin gene THBD causes high plasma thrombomodulin (TM) levels. High TM levels attenuate thrombin generation and delay fibrinolysis.


| INTRODUC TI ON
The transmembrane protein thrombomodulin (TM) critically regulates blood coagulation by localizing thrombin to the vascular endothelial surface enabling the activation of several key substrates.
The targets of the TM-thrombin complex include protein C, which after conversion to activated protein C (APC) limits further thrombin generation by inactivating coagulation factors Va and VIIIa. 1 TM also acts as a cofactor in thrombin-mediated activation of the procarboxypeptidase thrombin-activatable fibrinolysis inhibitor (TAFI) to activated TAFI (TAFIa). TAFIa attenuates the binding of tissue-type plasminogen activator (tPA) and plasminogen to fibrin by cleaving carboxyterminal lysines from partially degraded fibrin, thereby downregulating fibrinolysis. 2 The physiological importance of TM is illustrated by the newly recognized autosomal dominant bleeding disorder thrombomodulin-associated coagulopathy (TM-AC), which to date has been associated with a single p.Cys537* variant in the thrombomodulin gene THBD. [3][4][5][6] This truncation variant results in excessive shedding of large quantities of the functionally active TM extracellular domain into plasma. This results in a significant bleeding diathesis because the high TM levels promote excessive generation of APC, which suppresses normal thrombin generation. 3-6 TM-AC is also associated with delayed fibrinolysis that can be corrected by inhibition of TAFIa, suggesting that the surplus TM in plasma stimulates thrombin-mediated TAFI activation. 3 Here we report on a TM-AC pedigree with abnormal bleeding associated with a previously unreported THBD variant. We also describe how some pedigree members also harbor an independently inherited loss-of-function rare variant in CPB2 resulting in reduction in TAFI levels and TAFIa generation and causing amelioration of the delayed fibrinolysis associated with TM-AC.

| ME THODS
The study pedigree was identified in a systematic inspection of geno-

| Detection and annotation of the THBD and CPB2 variants
Within the study collection, there was a single index case (II.2, Figure 1A) with a previously unreported high-impact THBD variant, similar to the variant in previously reported pedigrees with TM-AC. [3][4][5][6] This was a monoallelic single nucleotide deletion annotated as • The complex hemostatic defect comprises reduced thrombin generation and delayed fibrinolysis.
• We report a pedigree with a new THBD variant indicating a common molecular pathogenesis of TM-AC.
• Coinherited TAFI deficiency results in attenuated delayed fibrinolysis but not reduced thrombin generation.

| Characteristics of the TM-AC cases
The adult cases with THBD p.Pro496Argfs*10 (II. 2 Figure 1D), predominantly after dental procedures and trauma similar to previously reported TM-AC pedigrees, 3-6 but also after childbirth. Abnormal bleeding was not reported for the two cases in generation IV who were all aged 14 years or younger at enrolment and who had not undergone invasive dental or surgical procedures. Coinheritance of CPB2 p.Arg114* (cases II.2 and II.3) had no discernible effect on the frequency or severity of bleeding. Plasma coagulation times, clotting factor levels, and platelet function testing in the THBD p.Pro496Argfs*10 cases were normal (data not shown). Consistent with previous reports of TM-AC, plasma TM levels were increased by at least two orders of magnitude in all pedigree members with THBD p.Pro496Argfs*10 ( Figure 1D). Plasma TAFI levels were  16 ETP, endogenous thrombin potential; TM, thrombomodulin; TAFI, thrombin-activatable fibrinolysis inhibitor; PPH, post-partum hemorrhage; V, variant allele; +, wild type allele; NT, not tested. *Indicates abnormal dental bleeding despite pro-hemostatic measures including antifibrinolytic and plasma treatment. Control ranges are shown in square brackets almost two-fold lower in the three cases harboring CPB2 p.Arg114* compared to cases without this genotype (mean ± standard error of the mean [SEM] 542 ± 81 versus 1025 ± 113 U/L, Figure 1D), consistent with absent expression of the CPB2 allele harboring the p.Arg114* variant.
To investigate potential interactions between the THBD and CPB2 genotypes, we compared thrombin generation in plasma from adult pedigree members and control plasma (National Institute for Biological Standards and Control standard plasma) as previously reported. 3 The TM-AC cases (II.2, III.2, and III.3) demonstrated a reduction in endogenous thrombin potential and reduced peak thrombin concentration ( Figure 1D). The differences in thrombin generation between cases and controls were smaller following initiation with 5 pmol/L tissue factor but the overall trend was the same (data not shown). These data with THBD p.Pro496Argfs*10 echo those of the previous descriptions of TM-AC associated with the p.Cys537* variant and are consistent with increased generation of APC and excessive suppression of thrombin generation. [3][4][5][6] The presence of the additional CPB2 p.Arg114* variant in the TM-AC case II.2 had no discernible effect on thrombin generation.

| The CPB2 p.Arg114* variant downregulates fibrinolysis
The effect of the THBD and CPB2 variants on tPA-mediated fibrinolysis was analyzed by monitoring the turbidity of plasma samples after clot formation with 0.1 U/mL thrombin and calcium. In vitro plasma clot lysis was significantly delayed in samples from TM-AC cases III.2 and III.3 harboring the THBD variant alone (mean ± SEM time to 50% lysis [CLT] 223 ± 5.2 and 221 ± 5.9 minutes respectively versus 85 ± 1.9 minutes in control; P < .0001; Figure 2A, B), similar to previously reported cases with TM-AC. 3 In the TM-AC case II.2, who also harbors the CPB2 p.Arg114* variant, fibrinolysis was delayed compared to control plasma, but to a lesser extent than the TM-AC cases without CPB2 p.Arg114* (CLT 127 ± 1.6 minutes, F I G U R E 2 Delayed fibrinolysis in thrombomodulin-associated coagulopathy. A, Representative turbidity curves showing clot lysis in platelet-free plasma from controls or thrombomodulin-associated coagulopathy (TM-AC) cases. B, 50% lysis times were calculated from triplicate plasma samples using Shiny App for calculating clot lysis times. 17 Experiments were performed with and without 25 µg/mL potato tuber carboxypeptidase inhibitor (PTCI, Sigma Aldrich) or 65 μg/mL MA-T12D11. 11 Data shown represent the mean ± standard error of the mean of the turbidity measurements. ***P < .001 ****P < .0001 case versus control in plasma without thrombin-activatable fibrinolysis inhibitor (TAFI) inhibitors. Statistical significance was determined by one-way analysis of variance with Bonferroni's post hoc test. C, Representative profiles of biphasic activated TAFI (TAFIa) generation during in vitro clot lysis from controls or TM-AC cases. Clot lysis was performed as in (A) and simultaneously a second identical experiment was performed where at defined time points samples were collected and aprotinin (130 µg/mL) and trifluoroacetate salt (PPACK; 5 µmol/L) were added to stop TAFIa generation, after which the samples were placed on ice. Peak 1 (P1) and peak 2 (P2) correspond to thrombin-thrombomodulin-mediated and plasmin-mediated TAFIa generation respectively. D, TAFIa activity levels for P1 and P2 of TAFI generation during in vitro clot lysis for all adult pedigree members. Data represents mean ± standard deviation (n = 6) P < .0001). Clot lysis was enhanced in case III.4 who had the CPB2 p.Arg114* variant but did not carry the THBD mutation (CLT 66 ± 0.4 versus 85 ± 1.9 minutes in control, P < .001; Figure 2A, B).
To explore whether the modulatory effect of CPB2 p.Arg114* on fibrinolysis was a consequence of a reduction in the TAFI level, TAFIa concentration was measured during in vitro clot lysis. This model enables resolution of two peaks of TAFIa formation, generated first by TM/thrombin (P1) and subsequently by plasmin (P2; Figure 2C). In  Figure 2B). The concentrations of inhibitors included here had previously been used to overcome TAFIa activity in plasma. 3,13,14 However, in this study the exceptionally high concentrations of TM in the plasma of the case studies precluded the complete correction of clot lysis time. Higher concentrations of inhibitor were able to reduce levels further (data not shown). In TM-AC case II.2, who also harbors CPB2 p.Arg114*, the same concentrations of inhibitor permitted more complete correction of the fibrinolytic abnormalities, with clot lysis times similar to those of control plasma ( Figure 2B).
These data are the first to describe a genetic deficiency of TAFI and emphasize the key role of TAFI in attenuating fibrinolysis in TM-AC. The marked elevation in plasma TM attenuates thrombin generation and delays fibrinolysis. This is a similar consequence to the THBD p.Cys537* variant associated with all previously described cases of TM-AC. [3][4][5][6] The reported results suggest a common pathogenic mechanism in both THBD variants in which the chain truncation promotes shedding of a functionally active TM extracellular domain into plasma. 15 Remarkably, some pedigree members also harbor a pathogenic variant in CPB2 predicting p.Arg114* in TAFI, resulting in partial deficiency of plasma TAFI levels. To our knowledge this is the first known case of a genetic deficiency in TAFI to be described in humans. We show that TAFI deficiency is clinically asymptomatic, but that the reduction of procarboxypeptidase activity in plasma accelerates fibrinolysis in vitro. Coinheritance of the CPB2 p.Arg114* with THBD p.Pro496Argfs*10 partially ameliorates the delayed fibrinolytic profile associated with TM-AC, clearly demonstrating a crucial role for TAFI in this laboratory feature of TM-AC. The effect of this variant was similar to pharmacological inhibition of TAFIa in members of the pedigree with TM-AC. Analysis of this pedigree, in which members have highly impactful variants affecting two interacting coagulation pathway genes, enhances our understanding of ultra-rare human hemostatic disorders. The different combinations of the variants in the pedigree family members is a unique platform to allow insights into the regulation of the TM/thrombin system and TAFI activation that is highly relevant to a broad range of hemostatic disorders and therapies. The asymptomatic nature of genetic depletion of human TAFI underscores the potential to exploit inhibition of TAFI pharmacologically without bleeding complications.