User:Chase Haven/Sandbox 1

Basic Crystallographic Features

The heterogeneity of fibrin clots make them virtually impossible to crystallize, and fibrinogen itself has intrinsic properties that enabled it to elude crystallization efforts until 2009. Some of the interesting aspects of this crystal are discussed below.

Coiled-Coil Domains

The coiled-coil domains separate the two outer D domains from the central globular E domain. These coils are mostly helical, with a notable exception being the “out-loop” in gamma chains residues . Certain areas also have runs of polar amino acids, coinciding with a pronounced bending at this location. Three notable runs include Alpha: , Beta: , and Gamma: . These non-helical regions share structural similarity and conservation with that of previously-crystalized chicken fibrinogen, even though they have strikingly different amino acid sequences.^[5]

It has been proposed that the coiled coils are the source of (at least in part) fibrinogen's inherent "floppiness". The non-helical stretches may act as flexible hinges interspersed with the more rigid helical components. This also fits with data regarding the compressible and stretchable nature of fibrin gels. Also fitting to this "hinge hypothesis", are two known mutations that cause defective fibrin formation occur in the polar (nonhelical) regions.^[6]

Indiscernible Regions

The elusive alpha-C termini regions were unable to be resolved. This was actually expected, as these regions are thought to be intrinsically disordered.^[7] A study in 2006 used the PONDR VL-XT algorithm (Predictor Of Natural Disordered Regions)and found that of the 866 residues of fibrinogen, 397 or 46% were predicted to be disordered. The longest stretch being 129 consecutive disordered residues. Large portions of these figures correspond with the alpha-C terminal regions. Interestingly, these regions have been shown to be involved in binding numerous proteins including those involved in coagulation (FXIII, other fibrin molecules) fibrinolysis (plasminogen, tissue-plasminogen activator), and others.^[8] The intrinsic disorder could be beneficial for fibrinogen by allowing a broad binding diversity, while also stabilizing the protein in a variety of trauma-induced scenarios such as hypothermia and acidosis. The results of this algorithm also coincide with, up until the past decade, the distinct lack of available structural data concerning the different domains of fibrinogen, and likely contribute to the difficulty of crystalization attempts.

As it pertains to this crystal structure, it is also possible that these portions of the molecule were proteolytically removed sometime during purification, as has been observed in the literature. N-terminal segments of the alpha and B chains (source of fibrinopeptides) in the central globular region were also unresolved, but are known to be mobile.^[9]

Cation-Pi Interactions

When ionic interactions are evaluated in the gaseous phase, they are incredibly strong. However, in aqueous media the water molecules negates some of the advantage of an electrostatic interaction due to the energy expense required to shed the solvation shell surrounding each of the charged molecules in the bridge. Conversely, cation-pi interactions do not have such an extensive penalty to pay and can thus be highly favorable. In fibgrinogen there are an estimated 94 energetically significant cation-pi interactions, providing a considerable amount of stability to the molecule. The strongest interaction occurs between of the gamma chain, providing an interaction energy of -7.24 kcal/mol. Of those that are energetically significant, 13 cation-pi interactions are between arginine and phenylalanine, 27 between arginine and tryptophan, and only 7 between arginine and tyrosine. There are also 12 between lysine and phenylalanine, 15 for lysine to tyrosine, and 20 between lysine and tryptophan.^[10]

Disulfide Bonds

Each fibrinogen has bonds and no free thiol groups. All of these are considered structural; they hold the six peptide chains together and are crucial for proper folding and assembly.^[11] There are three obvious clusters of disulfide bonds in the molecule, located at the N-terminal region, C-terminal region and in an intermediate location. In the central E domain containing the N-termini of the chains, a series of five symmetrical disulfide bridges secure it's globular structure. In the alpha chain two reside at Aalpha28 and Aalpha36, in the Bbeta chain at position 65, and two more in the gamma chain at positions 8 and 9. Since the two at the gamma chain are reciprocal, they orient the chains in an antiparallel manner. Other non-symmetrical inter-chain disulfides in the E domain form what has been referred to as a disulfide "ring."^[12][13]

Sources of Flexibility

The function of fibrinogen requires an intrinsic level of flexibility; this protein has to form a dense yet resilient network that can withstand shear forces within the vasculature. A clot must be able to close gaps caused by injury, and for that reason the clot components must be capable of assuming a variety of shapes. Several studies, many utilizing atomic force microscopy, have demonstrated a remarkable ability of fibrinogen and fibrin fibers to stretch and bend to an extreme extent: under some conditions, nearly an additional 30% of it’s resting state length.^[14] A lot of work has been undertaken in order to better understand the origin this remarkable ability in the molecule and the clots that it makes up.

D-D Interface

One group analyzed their crystal structure of a fibrin D-Dimer that had been crosslinked by factor XIII. The junction between these two globular regions was not found to be tight; there were no salt bridges and minimal hydrogen bonding. In fact, there appeared to be a layer of solvent between them. This makes sense for the formation of a dynamic fibrin network; as too strong of an end to end interaction would make the clot too rigid. Also interesting about this particular crystal structure was the location of the gamma-gamma cross-link between the D regions: they weren’t visible. Density could not attribute these to a single location, meaning that the covalent bond between E396 and Q398 was highly mobile. This mobility was also observed by another group, who noted that there were several possible orientations in D-Dimer crystals. Taken altogether this provides evidence for the existence of weak interactions at the D-D interface that can slide around and “play,” potentially lending flexibility to the overall network. (CSF&F Doolittle, above paper as well)

A:a Interaction

The knob-hole interaction responsible for fibrin polymerization is non-covalent, but incredibly strong. In fact, it has been shown that regions of the protein actually unfold due to stress before this interaction is broken. A group found that when force was applied to small fibrin polymers, a stepwise unfolding of the gamma nodule occurred with increasing force prior to rupture of the knob-hole interaction. This means that the interaction is sufficiently strong and will persist as the fiber experiences localized unfolding that accounts for fiber stretching. The in-vivo role of this may not be significant however, as FXIII cross-links gamma chains diametrically opposite of the knob-hole interactions. The strength of this covalent bond (despite being mobile) would likely circumvent the utility of stretching due to gamma-nodule unfolding.^[15]

Alpha-c Regions

As previously mentioned, many studies have observed an inherent flexibility of fibrin fibers as well as individual molecules themselves, characteristic of other protein fibers such as elastin and spider silk. Some researchers have hypothesized that these mechanical properties may stem from the phenomena of straightening of unstructured peptide sequences. Fibrin contains two regions per molecule that possess such a significant amount of intrinsic disorder: the elusive alpha-c regions. It is already accepted that these regions play a key role in determining some of the mechanical properties of fibrin clots through work with recombinant protein lacking these regions, but more recent studies are pointing to specific portions of these regions and relating them to the extensibility of the fibers themselves.^[16][17]

In particular, the length of tandem repeats in the alpha-c regions varies from species to species: low in chicken fibrinogen, which has no repeats, to high in lampreys which have 20 repeats of 18 residues.^[18][19][20] Interestingly, the amino acid content of the repeats seen in lamprey (as well as human and others) are reminiscent of the sequences seen in previously mentioned protein fibers like elastin.^[21]

Further supporting a potential role for these tandem repeats was a study that looked at fibrin extensibility from a variety of species and found that the length of the repeating sequence correlated well. Chicken fibrinogen (no repeats) had an extensibility of 47%, mouse fibrinogen (5 repeats of ~13 residues each) could extend 187%, and that of human (10 repeats of ~13 residues each) could be extended 217%. This significant finding clearly supports the hypothesis that the alpha-c regions provide a structural basis for fiber flexibility, but it is limited in it’s interpretative value. For example, there are other significant differences in sequence and structure in the different species of fibrinogen, including even the terminal region of the alpha-c chain, not just the tandem repeats.^[22][23]

The interaction of Fibrinogen with Thrombin

Proteolytic cleavage of fibrinopeptide A, the N-terminal sixteen amino acid residues of fibrinogen alpha chains, initiates fibrin assembly and the formation of a clot. This reaction is performed by thrombin, s trypsin-like serine protease. Unlike trypsin however, thrombin shows a remarkable specificity for fibrinogen; a highly unique quality that is in contrast with the very broad substrate specificity of other canonical serine proteases. Of the 376 arginine or lysines located in fibrinogen, thrombin only cleaves four (corresponding with the two each of fibrinopeptide A and B). (STUBBS #6) A crystal structure depicting a ternary complex of thrombin with synthetic analogs of fibrinopeptide A and hirudin’s C-terminal domain was produced in attempt to elucidate the interactions necessary for the enzyme’s specificity.

The fibrinopeptide A analog showed little secondary structure, although the sections that did demonstrate some important features about the interaction of thrombin with fibrinogen. A short helical sequence of residues 7-10 show a hydrogen bond between the carboxylate oxygen of Asp7 to the amide nitrogen of Ala10. This alternative hydrogen bonding pattern is not characteristic of alpha helices, and is thought to help give rise to the chain reversal occurring after Ala10, which is further stabilized by main-chain hydrogen bonds in the brief helical region, the subsequent amino acids 11 and 12.

This folding pattern facilitates the the formation of a consisting of the side chains of Phe8, Leu9 and Val15. Analysis of this regions shows a considerable number of contacts occur as this hydrophobic face inserts into the apolar binding site on thrombin. There are however, 3 hydrogen bonds that form a canonical beta-ladder, which are believed to orient substrates into the binding site of the protease. In this case, the antiparallel beta ladder occurs between Gly14 of FPA and Gly216 of thrombin. Other polar contacts include a number of ionic interactions between These are believed to also facilitate proper substrate orientation into the active site.

Sequence comparisons of fibrinopeptide A with different species show conservation of the basic features of their interactions with thrombin. Residues 8 and 9, which form part of the hydrophobic patch, are well conserved and only found to be replaced with other hydrophobic residues. Residue ten (site of chain reversal) makes no contacts with the enzyme and therefore can accomadate several alternative amino acids. The highly conserved glycine at position 12 illustrates the importance of it’s rotational flexibility. In order for the peptide to fold back into the active site following this residue, the peptide has to adopt dihedral angles that no other amino acids are capable of. (Stubbs #5)

Looking at the binding of the C-terminal analog, we see an extended chain with a short 3-10 helix. It’s interaction with thrombin shows numerous ionic contacts with residues 54-65 of hirudin to the anion-binding site of the protease, which is flanked by two hydrophobic faces. Despite the large number of ionic amino acids in this region there are relatively few hydrogen or ionic bonds; rather binding may be mediated more by their complementary structure.