Analysis of human fibrinogen by scanning transmission electron microscopy

A proposed stoichiometrical calibration procedure to achieve transferability of D-dimer measurements and to characterize the performance of different methods

BACKGROUND

There is little transferability between D-dimer levels obtained by different reagents today. This makes it difficult to compare results from different clinical studies.

OBJECTIVES

We give a comprehensive proposal for calibration of D-dimer assays. All crucial steps and underlying assumptions are made explicit.

METHODS

The new approach is based on using a set of fibrinolysates of patient samples clotted and treated with tPA to obtain maximal conversion to D-dimers. Their expected maximal D-dimer concentrations are calculated stoichiometrically from their different fibrinogen values and the published molecular masses of fibrinogen and average D-dimer. The characteristics of five latex enhanced D-dimer immunoassays were also tested against early and late fibrin fragments using this procedure. These were produced by prolonged fibrinolysis of a set of patient samples of varying fibrinogen concentrations.

RESULTS

These varied typically between methods and lysis times. One of the methods showing the highest yield irrespective of lysis time was used for calibration. A linear standard curve with zero intercept and R2 = 0.95 was obtained.

CONCLUSION

Following this procedure will allow better transferability of D-dimer in future clinical trails.

The structure and function of the alpha C domains of fibrinogen

The alpha C domains have been localized on fibrinogen and fibrin. Several model systems have been developed to study their functions. Analysis of the amino acid sequence of the alpha C domains suggested that each is made up of a globular and an extended portion. Microcalorimetry confirmed this result and showed that the two alpha C domains interact intramolecularly. Electron microscopy of fibrinogen with a monoclonal antibody to the alpha C domains demonstrated that these regions normally interact with the central portion of the molecule. In the conversion from fibrinogen to fibrin there is a large scale conformational change, such that the alpha C domains dissociate from the central region and are available for intermolecular interaction. Experiments with highly purified and well characterized fragment X monomer, missing either one or both of the alpha C domains, indicate that intermolecular interactions between alpha C domains are important for the enhancement of lateral aggregation during fibrin polymerization. Isolated alpha C fragments polymerized at neutral pH and interacted with the alpha C domains of fibrin monomer to influence clot formation. Several dysfibrinogenemias in which there are amino acid substitutions in, or truncations of, the alpha C domains revealed that these changes can have dramatic effects on polymerization and clot structure. The polymerization of A alpha 251 recombinant fibrinogen, that contains A alpha chains truncated at residue 251, was altered, as were the mechanical properties and the rate of fibrinolysis of the clots. Altogether, these results help to define the role of the alpha C domains in determining the structure and properties of clots.

alpha IIb beta 3 redistribution triggered by receptor cross-linking

Fibrinogen binding to alpha IIb beta 3 on adherent, spread platelets triggers active, cytoskeletally-directed redistribution of fibrinogen/alpha IIb beta 3 complexes on the platelet surface. Gold-conjugated fibrinogen, unlabeled, soluble fibrinogen, and individual fibrinogen molecules have been demonstrated to trigger receptor redistribution. Here we examine the respective roles of receptor cross-linking and ligand occupancy of receptors in initiating this movement. Monovalent, alpha IIb beta 3-binding fibrinogen fragments RGDS and HHLGGAKQAGDV did not trigger receptor redistribution, suggesting that ligand binding to a single receptor is an insufficient stimulus. Binding of monoclonal antibodies 10E5, AP2, and AP3 to the receptor did not trigger receptor movement. However, cross-linking these receptor-bound monoclonal antibodies by polyclonal anti-mouse IgG or by conjugation of the anti-receptor antibody to large colloidal gold particles triggered receptor redistribution identical in rate, pattern, and final distribution to that previously seen with fibrinogen binding. We conclude that receptor cross-linking provides the signal for initiation of fibrinogen/alpha IIb beta 3 complex redistribution on platelet surfaces.

Intramolecular crosslinking of monomeric fibrinogen by tissue transglutaminase

In addition to generating polymeric products from human fibrinogen, human erythrocyte transglutaminase (protein-glutamine:amine gamma-glutamyltransferase, EC 2.3.2.13) was shown to catalyze the intramolecular reaction of crosslinking two of the constituent chains within monomeric fibrinogen itself. This internally fused protein derivative contains appreciable amounts of the N epsilon-(gamma-glutamyl)lysine bridge peptide and displays the A alpha.gamma hybrid chain pattern of crosslinking, characteristic for the actions of tissue transglutaminases on fibrinogen. Diagnostic analysis in pathological situations, where such enzymes might have escaped from cells into the plasma environment, should include a search for the internally crosslinked soluble fibrinogen monomer.

Electron microscope investigation of the early stages of fibrin assembly. Twisted protofibrils and fibers

Structures formed during the early stages of clot formation have been produced in a controlled manner by polymerization of soluble fibrin monomers prepared from dissolved normal clots that had been formed upon addition of thrombin to fibrinogen. In agreement with other studies using different approaches, electron microscopy of negatively contrasted or rotary-shadowed specimens of these preparations reveal two-stranded protofibrils, as well as shorter oligomers and fibrin monomers. Individual fibrin molecules are similar in appearance to fibrinogen, suggesting that no large-scale changes in conformation occur on removal of the fibrinopeptides. Moreover, these micrographs show details of the protofibril structure not previously seen. The visualization of clear cross-over points of the filaments making up the protofibril indicate that these structures are twisted. Diffraction patterns of electron micrographs of both protofibrils and fibers and computer modeling of protofibrils also suggest that these structures are twisted but not precisely ordered.

Molecular morphology of fibrin monomers and early oligomers during fibrin polymerization

The structural features of early fibrin oligomers produced during the initial stages of polymerization were investigated by rotatory shadowing after cryotechnical preparation. The building blocks of polymerization, namely fibrin monomer units (in analogy to fibrinogen itself), were found to exhibit a high degree of flexibility which is independent of fibrinopeptide A and B removal. Early polymers exhibited directed longitudinal growth and were frequently branched. Along the main oligomer axis, fibrin monomer units were randomly orientated. Within early oligomers, a given fibrin monomer unit was found to establish a single contact with each of its two neighbors, suggesting that during the early stages of polymerization, only one polymerization and one binding site are activated per fibrinogen molecule (becoming an AB2 fibrin monomer unit). This morphological feature was corroborated by the finding that early oligomer fractions are deficient in only 50% of releasable fibrinopeptide A. Early associations between AB2 fibrin monomer units were demonstrated to be reversible and to occur in the absence of direct domainal contact; interactions thus presumably occur via fine molecular protrusions on either D or E domains. The arrangement of AB2 fibrin monomer units within early oligomers suggests that, with respect to their structural organization, fibrinogen molecules are radially symmetrical through the E domain (implying an antiparallel organization of polymerization and binding sites). This pattern is inconsistent with a "top-bottom" model, and thus with "half-staggered double-stranded" polymer growth. The methodological problems responsible for the apparent conflict with previous morphological findings are discussed.

Mass loss rate in collodion is greatly reduced at liquid helium temperature

The mass thickness of collodion films has been monitored at several temperatures, under conditions typical of electron microscopy, through the use of an electron energy loss spectrometer. Compared to room temperature, only a five-fold reduction in the rate of mass loss was observed through the use of a commercial liquid nitrogen cooled stage; in contrast, the rate of mass loss was reduced more than one hundred fold when these films were held at liquid helium temperature.

A model for fibrinogen: domains and sequence

Electron microscopy of rotary-shadowed fibrinogen demonstrates that the molecules modified for crystallization by limited cleavage with a bacterial protease retain the major features of the native structure. This evidence, together with image processing and x-ray analysis of the crystals and of fibrin, has been used to develop a three-dimensional low resolution model for the molecule. The data indicate that the two large end domains of the molecule would be composed of the carboxyl-terminus of the B beta chain (proximal) and gamma chain (distal), respectively; the carboxyl-terminus of the A alpha chain would fold back to form an additional central domain. On this basis, the carboxyl-terminal region of each of the three chains of fibrinogen is folded independently into a globular domain.

Structure of the signal recognition particle by electron microscopy

The signal recognition particle (SRP) is a ribonucleoprotein consisting of six distinct polypeptide components and one molecule of small cytoplasmic RNA (7SL RNA). The particle was previously shown to function in protein translocation across and protein integration into the endoplasmic reticulum membrane. Homogeneous signal recognition particle preparations were visualized by electron microscopy (i) after negative staining, (ii) by dark-field imaging of unstained specimens, and (iii) by platinum-shadowing. The results of each of these different techniques indicate that the signal recognition particle is a rod-shaped particle 5-6 nm wide and 23-24 nm long.