Cardiac troponin T forms a tetramer in vitro.

Cardiac troponin T (cTnT) is a myofibrillar protein essential for calcium-dependent contraction. This property has led to functional studies of developmentally expressed cTnT isoforms and mutants identified in patients with hypertrophic cardiomyopathy. The release of cTnT into the serum following myocardial infarction has led to the development of antibody-based assays for measuring cTnT serum concentration. We examined the behavior of cTnT in solution. Recombinant human cTnT3, the dominant isoform in the adult human heart, was used. The protein was pure and functional, as demonstrated by SDS-PAGE and surface plasmon resonance. cTnT3 was found to bind specifically and in a concentration-dependent manner to cTnC. Routine size exclusion chromatography suggested a higher-than-expected MW for cTnT. Using analytical ultracentrifugation, we found cTnT3 in solution to be mainly in the form of a tightly bound tetramer at concentrations as low as 4 micromol/L. Our sedimentation velocity and transmission electron microscopy results indicate that the tetramer's shape is elongated rather than globular. CTnT's self-association in solution is an important consideration in the design and interpretation of experiments with the aim of understanding the biochemical and biophysical properties of cTnT, its isoforms, and its mutants.

Probing the gamma2 calcium-binding site: studies with gammaD298,301A fibrinogen reveal changes in the gamma294-301 loop that alter the integrity of the "a" polymerization site.

To determine the significance of the gamma2 calcium-binding site in fibrin polymerization, we synthesized the fibrinogen variant, gammaD298,301A. We expected these two alanine substitutions to prevent calcium binding in the gamma2 site. We examined the influence of calcium on the polymerization of gammaD298,301A fibrinogen, evaluated its plasmin susceptibility, and solved 2.7 and 2.4 A crystal structures of the variant with the peptide ligands Gly-Pro-Arg-Pro-amide (GPRP) and Gly-His-Arg-Pro-amide (GHRP), respectively. We found that thrombin-catalyzed polymerization of gammaD298,301A fibrinogen was modestly impaired, whereas batroxobin-catalyzed polymerization was significantly impaired relative to normal fibrinogen. Notably, the influence of calcium on polymerization was the same for the variant and for normal fibrinogen. Fibrinogen gammaD298,301A was more susceptible to plasmin proteolysis in the presence of GPRP. This finding suggests structural changes in the near-by "a" polymerization site. Comparisons of the structures revealed minor conformational changes in the gamma294-301 loop that are likely responsible for the weakened "a" site. When considered altogether, the data suggest that the gamma2 calcium-binding site does not significantly modulate polymerization. We cannot, however, rule out the possibility that the weakened "a" polymerization site masks an important role for the gamma2 calcium-binding site in normal polymerization. Somewhat unexpectedly, the structure data showed that GPRP bound to the "b" site and induced the same local conformational changes as GHRP to this site. This structure shows that "A:b" interactions can occur and suggests that these may participate in normal polymerization.

Investigation of residues in the fibrin(ogen) gamma chain involved in tissue plasminogen activator binding and plasminogen activation.

In order to characterize tissue plasminogen activator (t-PA) binding to gamma-chain residues in fibrinogen, we generated variant fibrinogens substituting alanine for gamma D316, gamma D318, gamma D320, and gamma K321. We measured thrombin-catalyzed polymerization and found normal polymerization with gamma K321A, no polymerization with gamma D316A, and, as reported by Lounes et al. in 2002, impaired polymerization with gamma D318A and gamma D320A. We measured t-PA binding in a solid-phase assay, and t-PA activity by the generation of plasmin. Comparing normal fibrin with fibrinogen, we found a seven-fold increase in binding and a two-fold increase in activity. Binding to all variant fibrinogens was the same as normal. In contrast, t-PA binding to all variant fibrins was weaker than binding to normal fibrin, 2.5-fold for gamma K321A, seven-fold for gamma D320A and 10-fold for gamma D316A and gamma D318A. Plasmin generation in the presence of variant fibrinogens was similar, although not identical, to normal, and plasmin generation in the presence of variant fibrins was impaired for the Asp to Ala variants. As the three variants with the weakest t-PA binding and least activity also showed impaired polymerization, our results support previous findings demonstrating the DD:E complex, found in the normal fibrin polymer, is necessary for the fibrin enhanced binding of t-PA and activation of plasminogen.

Calcium-binding site beta 2, adjacent to the "b" polymerization site, modulates lateral aggregation of protofibrils during fibrin polymerization.

Structural analysis of recombinant fibrinogen fragment D revealed that the calcium-binding site (beta2-site) composed of residues BbetaAsp261, BbetaAsp398, BbetaGly263, and gammaGlu132 is modulated by the "B:b" interaction. To determine the beta2-site's role in polymerization, we engineered variant fibrinogen gammaE132A in which calcium binding to the beta2-site was disrupted by replacing glutamic acid at gamma132 with alanine. We compared polymerization of gammaE132A to normal fibrinogen as a function of calcium concentration. Polymerization of gammaE132A at concentrations of calcium

Severe hypodysfibrinogenemia in compound heterozygotes of the fibrinogen AalphaIVS4 + 1G>T mutation and an AalphaGln328 truncation (fibrinogen Keokuk).

Two siblings with hypofibrinogenemia have lifelong trauma-related bleeding. Recently, the brother experienced recurrent thrombosis after cryoprecipitate infusions following surgery. The sister had 6 miscarriages. Plasma clots in each were resistant to compression and fibrinolysis and were soluble in 5 M urea. Examination by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) revealed only the presence of crosslinked gamma-gamma fibrin chain dimers without high polymers of alpha n. Fibrin clots contained an abnormal 35-kDa constituent recognized by an antibody to the mature fibrinogen Aalpha-chain residues 241-476 but not by antibodies to Aalpha219-348 or Aalpha349-406. DNA analysis revealed a heterozygous CAA-->TAA mutation at the codon for amino acid 328 of the Aalpha gene in these siblings and 2 asymptomatic family members. The Gln328stop mutation (fibrinogen Keokuk) predicted a 46% truncation and the production of a 35-kDa Aalpha chain. Analysis of purified fibrinogen revealed expression of the abnormal Aalpha chain in 4 family members but found no normal fibrinogen in the 2 hypofibrinogenemic patients. This paradox was resolved when they and their asymptomatic mother were found to be heterozygous for a second Aalpha mutation, a GT-->TT splice site mutation in intron 4 (IVS4 + 1 G> T). However, compound heterozygosity for both mutations was required for the expression of severe hypodysfibrinogenemia and for clinical symptoms.

Reduced platelet adhesion in flowing blood to fibrinogen by alterations in segment gamma316-322, part of the fibrin-specific region.

The interaction of platelets with fibrinogen is a key event in the maintenance of a haemostatic response. It has been shown that the 12-carboxy-terminal residues of the gamma-chain of fibrinogen mediate platelet adhesion to immobilized fibrinogen. These studies, however, did not exclude the possibility that other domains of fibrinogen are involved in interactions with platelets. To obtain more insight into the involvement of other domains of fibrinogen in platelet adhesion, we studied platelet adhesion in flowing blood to patient dysfibrinogen Vlissingen/Frankfurt IV (V/FIV), to several variant recombinant fibrinogens with abnormalities in the gamma-chain segments gamma318-320 and gamma408-411. Perfusion studies at physiological shear rates showed that platelet adhesion was absent to gammaDelta408-411, slightly reduced to the heterozygous patient dysfibrinogen V/FIV and strongly reduced to the homozygous recombinant fibrinogens: gammaDelta319-320, gamma318Asp-->Ala and gamma320Asp-->Ala. Furthermore, antibodies raised against the sequences gamma308-322 and gamma316-333 inhibited platelet adhesion under shear conditions. These experiments indicated that the overlapping segment gamma316-322 contains amino acids that could be involved in platelet adhesion to immobilized fibrinogen under flow conditions. In soluble fibrinogen, this sequence is buried inside the fibrinogen molecule and becomes exposed after polymerization. In addition, we have shown that this fibrin-specific sequence also becomes exposed when fibrinogen is immobilized on a surface.

Analysis of engineered fibrinogen variants suggests that an additional site mediates platelet aggregation and that "B-b" interactions have a role in protofibril formation.

The C-terminal domain of the fibrinogen gamma-chain includes multiple functional sites that have been defined in high-resolution structures and biochemical assays. Calcium binds to this domain through the side chains of gammaD318 and gammaD320 and the backbone carbonyls of gammaF322 and gammaG324. We have examined variant fibrinogens with alanine at position gamma318 and/or gamma320 and found that calcium binding, fibrin polymerization, and fibrinogen-mediated platelet aggregation, but not FXIIIa-catalyzed cross-linking, were abnormal. When measured by turbidity, thrombin-catalyzed polymerization was severely reduced, and batroxobin-catalyzed polymerization was completely obliterated. Moreover, thrombin-catalyzed polymerization was abolished by the peptide GHRP, which binds to the polymerization site in the beta-chain but does not inhibit polymerization of normal fibrinogen. ADP-induced platelet aggregation was also severely impaired. In contrast, as measured by SDS-PAGE, FXIIIa introduced cross-links between gamma-chains for all three variants, as expected if the gamma-chain C-terminal sites were normal. In addition, binding of the monoclonal antibody 4A5, which recognizes the C-terminal residues, was not different from normal. These data suggest two specific conclusions: (1) a site in the gamma-module other than the C-terminus is critical for platelet aggregation and (2) "B-b" interactions have a role in protofibril formation.