Thrombin-activatable fibrinolysis inhibitor (TAFI) is an important regulator in the balance of coagulation and fibrinolysis. TAFI is a metallocarboxypeptidase that circulates in plasma as zymogen. Activated TAFI (TAFIa) cleaves C-terminal lysine or arginine residues from peptide substrates. The removal of C-terminal lysine residues from partially degraded fibrin leads to reduced plasmin formation and thus attenuation of fibrinolysis. TAFI also plays a role in inflammatory processes via the removal of C-terminal arginine or lysine residues from bradykinin, thrombin-cleaved osteopontin, C3a, C5a and chemerin. TAFI has been studied extensively over the past three decades and recent publications provide a wealth of information, including crystal structures, mutants and structural data obtained with antibodies and peptides. In this review, we combined and compared available data on structure/function relationships of TAFI.

BACKGROUND:Portal vein embolization (PVE) is used to increase future remnant liver size in patients requiring major hepatic resection. PVE using permanent embolization, however, predisposes to complications and excludes the use of PVE in living donor liver transplantation. In the present study, an absorbable embolization material containing fibrin glue and different concentrations of the fibrinolysis inhibitor aprotinin was used in an experimental animal model.METHODS:PVE of the cranial liver lobes was performed in 30 New Zealand White rabbits, which were divided into five groups, fibrin glue + 1000, 700, 500, 300 or 150 kunits/ml aprotinin, and were compared with a previous series of permanent embolization using the same experimental set-up. Caudal liver lobe hypertrophy was determined by CT volumetry, and portal recanalization was identified on contrast-enhanced CT images. Animals were killed after 7 or 42 days, and the results were compared with those of permanent embolization.RESULTS:PVE using fibrin glue with aprotinin as embolic material was effective, with 500 kunits/ml providing the optimal hypertrophic response. Lower concentrations of aprotinin (150 and 300 kunits/ml) led to reduced hypertrophy owing to early recanalization of the embolized segments. The regeneration rate over the first 3 days was higher in the group with 500 kunits/ml aprotinin than in the groups with 300 or 150 kunits/ml or permanent embolization. In the 500-kunits/ml group, four of five animals showed recanalization 42 days after embolization, with minimal histological changes in the cranial lobes following recanalization.CONCLUSION:Fibrin glue combined with 500 kunits/ml aprotinin resulted in reversible PVE in 80 per cent of animals, with a hypertrophy response comparable to that achieved with permanent embolization material. Surgical relevance Portal vein embolization (PVE) is used to increase future remnant liver volume in patients scheduled for major liver resection who have insufficient future remnant liver size to perform a safe resection. The current standard is PVE with permanent embolization materials, which renders patients found to have unresectable disease prone to complications owing to the permanently deportalized liver segments. Absorbable embolization might prevent the PVE-associated morbidity and lower the threshold for its application. In this study, PVE using fibrin glue and aprotinin resulted in an adequate hypertrophy response with 80 per cent recanalization after 42 days. Considering the minor histological changes following recanalization of embolized segments and potentially preserved function, reversible PVE might also be applied in living donor liver transplantation.

A new thrombin-activatable fibrinolysis inhibitor (TAFI) deletion mutant, constructed by Zhou et al. 1, provides us with new clues regarding the mysterious mechanism of spontaneous activated TAFI (TAFIa) self-destruction. TAFI, also known as procarboxypeptidase U, procarboxypeptidase R, and procarboxypeptidase B2, is encoded by the CPB2 gene 2. TAFI is synthesized in the liver, and circulates in plasma as a proenzyme. During coagulation, TAFI is activated by a single proteolytic cleavage at Arg92 that releases the activation peptide from the catalytic domain 3. Once active, TAFIa prevents accelerated plasmin formation by removing C-terminal lysines from partially degraded fibrin. As long as the TAFIa concentration stays above a certain threshold, fibrinolysis will not occur 4, 5. There are no known physiologic inhibitors of TAFIa, but instead TAFIa ‘self-destructs’ in a matter of minutes. The mechanism of this instability has been a mystery since its discovery 6. At body temperature, TAFIa has a half-life of ~ 10 min. At room temperature, TAFIa is more stable, with a half-life of 2 h, and at 0 °C TAFIa is almost completely stable 7. The marked influence of temperature has led to the hypothesis that TAFIa instability is the result of a conformational change rather than proteolytic cleavage. This was confirmed with a TAFI mutant resistant to proteolysis by thrombin (Arg302Gln) that, after activation, had the same half-life as wild-type TAFI 8. Apart from the effect of temperature on TAFIa stability, mutations have been identified that influence the half-life of the enzyme. In 2001, a naturally occurring polymorphism (Thr325Ile) was found that led to a two-fold increase in half-life 9. Since then, many stabilizing mutations have been created in mutagenesis studies 10-13. Strikingly, all of these stabilizing mutations are located between amino acids 300 and 330. In 2008, the crystal structures of human 14 and bovine TAFI 15 were elucidated, and revealed a highly dynamic region consisting of amino acids 296–350. Therefore, it was hypothesized that the flexibility of this region induces a conformational change that ‘destroys’ the enzyme. Before activation, TAFI is stable, because the activation peptide prevents this conformational change. It was suggested that this stabilization is the result of hydrophobic interactions between Val35 and Leu39 in the activation peptide and Tyr341 in the dynamic region 14. However, a study published in this issue of the Journal of Thrombosis and Haemostasis by Zhou et al. 1 sheds new light on this issue. Zhou et al. constructed a TAFI mutant that lacks the first 73 amino acids of the activation peptide. In TAFI, these 73 amino acids form a globular domain that shields the catalytic domain from (large) substrates (Fig. 1). The 19 remaining amino acids (Ala74–Arg92) of the activation peptide link this globular domain to the catalytic domain. Zhou et al. found that, with deletion of this globular domain but not the linker segment, small substrates can enter the active site and be processed just as in activated TAFI. However, surprisingly, their mutant was stable, similarly to zymogen TAFI. After removal of the remaining 19 amino acid linker segment with thrombin, thus ‘activating’ it, the mutant was destabilized in the same way as normal TAFIa. This provides evidence that it is not the interaction between Val35, Leu39 and Tyr341 that stabilizes TAFI zymogen, but rather the Ala74–Arg92 segment. But how does the linker segment stabilize TAFI? From the crystal structure, this is not clear. This segment is not adjacent to the dynamic region, and nor does it interact with it. A crystal structure of the deletion mutant might answer this question. It would also be interesting to determine whether more amino acids can be deleted or mutated before this segment starts to lose its stabilizing properties. It has to be noted that the 73 amino acid deletion was introduced in a 180-fold more stable TAFI mutant (CIIYQ) and that this mutant was used as a control. Would the results be similar if the deletion were introduced in wild-type TAFI? Furthermore, in this study, only the effects on small substrates were reported, and it remains to be determined whether this new mutant shows the same effects with larger, more physiologic substrates. The effect of the deletion mutant on physiologic substrates is also interesting with regard to another long-lasting TAFI debate: does the activation peptide remain attached to TAFIa after cleavage, and thereby affect TAFIa activity? Activated TAFI can be purified with a concanavalin A Sepharose column, which interacts with the sugar residues on the activation peptide 16. This would not be possible if the activation peptide and TAFIa became separated after activation. Also, western blot experiments with mAbs directed against the activation peptide suggested an interaction between TAFIa and the activation peptide 17. It was suggested that this interaction might affect TAFIa activity. However, this was challenged by Marx et al. 18, who separated TAFIa from the activation peptide, and demonstrated that the activation peptide had no effect on TAFIa's activity or stability. The deletion mutant constructed by Zhou et al. lacks most of the activation peptide, and makes it ideal for this debate. They demonstrated that the activity of this mutant, after incubation with thrombin, is similar to that of the TAFI control with a normal activation peptide, at least for a small synthetic substrate. If the same is true for physiologic substrates, then a role for the activation peptide in regulating TAFIa activity or stability after activation becomes less likely. However, a surprising result from this study is that the ‘activated’ TAFI deletion mutant appears to be more stable than the activated TAFI control. If this is the case, it means that the presence of a full-length activation peptide (Phe1–Arg92) destabilizes active TAFIa as compared with the presence of a small activation peptide fragment (Ala74–Arg92), which is the opposite of what was expected. A simpler explanation for the increased stability is that, after activation, stable ‘zymogen’ is still present. The zymogen form of the deletion mutant has the same activity as the enzyme form, but is stable. Furthermore, Zhou et al. demonstrated that thrombomodulin does not act as a cofactor for the deletion mutant, because it lacks the positively charged regions on the activation peptide 19, 20, which are involved in thrombomodulin-dependent activation of TAFI by thrombin. As thrombin by itself is a rather poor activator of TAFI, it is difficult to ‘activate’ the mutant, and a mixture of thrombin-cleaved and true zymogen TAFI could hamper correct determination of the half-life of the activated form. Not only is the presented TAFI deletion mutant interesting for understanding the (in)stability mechanism, but it also has practical uses. The instability of TAFIa can be a nuisance in the laboratory. Every experiment requires a freshly activated batch, and this will always contain (inactivated) activator (mostly thrombin). Purifying the unstable TAFIa after activation remains cumbersome. Also, cell culture or animal experiments with an unstable enzyme are difficult, as are crystallization studies, which may take weeks to months. These issues led researchers on a quest to construct more stable TAFI mutants, with the current (or previous) record for TAFIa-CIIYQ, which has a 180-fold increased half-life, although this is still < 20 h 13. With this new, active and almost completely stable TAFI deletion mutant, these issues may have been solved, and will boost the search for TAFIa inhibitors as novel profibrinolytic drugs. T. Plug and J. C. M. Meijers wrote the manuscript. The authors state that they have no conflict of interest.

Background: Thrombin-activatable fibrinolysis inhibitor (TAFI) is a risk factor for coronary heart disease. TAFI is proteolytically activated by thrombin, the thrombin-thrombomodulin complex and plasmin. Once active, it dampens fibrinolysis and inflammation. The aim of this study was to generate TAFI-derived peptides that specifically modulate TAFI activation and activity. Methods: Thirty-four overlapping TAFI peptides, and modifications thereof, were synthesized. The effects of these peptides on TAFI activation and TAFIa activity were determined. In addition, the binding of the peptides to thrombin were determined. Results: Four peptides (peptides 2, 18, 19 and 34) inhibited TAFI activation and two peptides (peptides 14 and 24) inhibited TAFIa activity directly. Peptide 2 (Arg12-Glu28) and peptide 34 (Cys383-Val401) inhibited TAFI activation by the thrombin-thrombomodulin complex with IC50 values of 7.3 1.8 and 6.1 +/- 0.9 m, respectively. However, no inhibition was observed in the absence of thrombomodulin. This suggests that the regions Arg12-Glu28 and Cys383-Val401 in TAFI are involved in thrombomodulin-mediated TAFI activation. Peptide 18 (Gly205-Ser221) and peptide 19 (Arg214-Asp232) inhibited TAFI activation by thrombin and the thrombin-thrombomodulin complex. Furthermore, these peptides bound to thrombin (K-D: 1.5 +/- 0.4 and 0.52 +/- 0.07 m for peptides 18 and 19, respectively), suggesting that Gly205-Asp232 of TAFI is involved in binding to thrombin. Peptide 14 (His159-His175) inhibited TAFIa activity. The inhibition was TAFIa specific, because no effect on the homologous enzyme carboxypeptidase B was observed. Conclusions: Thrombin-activatable fibrinolysis inhibitor-derived peptides show promise as new tools to modulate TAFI activation and TAFIa activity. Furthermore, these peptides revealed potential binding sites on TAFI for thrombin and the thrombin-thrombomodulin complex.

BackgroundThrombin-activatable fibrinolysis inhibitor (TAFI) is a proenzyme that links coagulation and fibrinolysis. TAFI can be activated by thrombin, the thrombin-thrombomodulin complex and plasmin through cleavage of the first 92 amino acids from the enzyme. Insilico analysis of the TAFI sequence revealed a potential thrombin cleavage site at Arg12. The aim of this study was to determine whether TAFI can be cleaved at Arg12 and whether this cleavage plays a role in TAFI activation.MethodsA peptide based on the first 18 amino acids of TAFI was used to determine whether thrombin was able to cleave at Arg12. Mass spectrometry was performed to determine whether the Arg12-cleaved peptide was released from full-length TAFI. Furthermore, a TAFI mutant in which Arg12 was replaced by a glutamine (TAFI-R12Q) was constructed and characterized with respect to its activation kinetics.ResultsThe peptide and mass spectrometry data showed that thrombin was able to cleave TAFI at Arg12, but with low efficiency in full-length TAFI. Characterization of TAFI-R12Q showed no difference in thrombin-mediated activation from wild-type TAFI. However, there was an approximately 60-fold impairment in activation of TAFI-R12Q by the thrombin-thrombomodulin complex.ConclusionsArg12 of TAFI plays an important role in thrombomodulin-mediated TAFI activation by thrombin. Thrombin is able to cleave TAFI at Arg12, but it remains to be determined whether Arg12 is part of an exosite for thrombomodulin or whether cleavage at Arg12 accelerates thrombomodulin-mediated TAFI activation.

The aggregation of amyloid β (Aβ) peptide is important in Alzheimer’s disease. Shorter Aβ fragments may reduce Aβ’s cytotoxicity and are used in diagnostics. The aggregation of Aβ16 is controversial; Liu et al. (J. Neurosci. Res. 75:162–171, 2004) and Liao et al. (FEBS Lett. 581:1161–1165, 2007) find that Aβ16 does not aggregate and reduces Aβ’s cytotoxicity, Du et al. (J. Alzheimer’s Dis. 27:401–413, 2011) reports that Aβ16 aggregates and that Aβ16 oligomers are toxic to cells. Here the aggregation potential of two shorter fragments, Aβ15 and Aβ16, and their influence on Aβ40 is measured by electron paramagnetic resonance (EPR) spectroscopy and the ThioT fluorescence assay (ThioT). Continuous-wave, 9 GHz EPR measurements and ThioT results reveal that neither Aβ15 nor Aβ16 aggregate by themselves and that they do not affect Aβ40 aggregation.

BACKGROUND:Thrombin-activatable fibrinolysis inhibitor (TAFI) is a 56-kDa procarboxypeptidase. Proteolytic enzymes activate TAFI into TAFIa, an inhibitor of fibrinolysis, by cleaving off the N-terminal activation peptide (amino acids 1-92), from the enzyme moiety. Activated TAFI is unstable, with a half-life of approximately 10 min at 37 degrees C. So far, it is unknown whether the activation peptide is released or remains attached to the catalytic domain, and whether it influences TAFIa's properties. The current study was performed to clarify these issues. METHODS:TAFI was activated, and the activity and half-life of the enzyme were determined in the presence and absence of the activation peptide. RESULTS:TAFIa was active both before and after removal of the activation peptide, and the half-life of TAFIa was identical in the two preparations. Furthermore, we observed that intrinsically inactivated TAFIa (TAFIai) aggregated into large, insoluble complexes that could be removed by centrifugation. CONCLUSIONS:The data presented in this article show that the activation peptide of TAFI is not required for TAFIa activity and that the activation peptide has no effect on the stability of the enzyme. These results are in favour of a model in which the activation peptide solely stabilizes the structure of the proenzyme. After activation of TAFI and subsequent breakage of interactions between the activation peptide and the catalytic domain, the activation peptide is no longer capable of performing this stabilizing task, and the integrity of the catalytic domain is lost rapidly. The resulting TAFIai is more prone to proteolysis and aggregation.