Quite Interesting PDB Structures

quips

An X(-mas) Factor that gets our vote
Quips Index

Christmas factor - a human protein involved in blood clotting

Christmas Factor isn't, as you might think, a seasonal version of the TV talent competition; nor is it named after the religious festival, despite being first described in the 1952 Christmas edition of the BMJ (ref. 1) In fact, it is one of the enzymes responsible for blood clotting, named after a child called Stephen Christmas. He was found to be deficient in this protein, otherwise known as Factor IX. Stephen Christmas had haemophilia B (sometimes called Christmas Disease after him), a condition which occurs in 1 in 50 000 men (the gene is on the X-chromosome).
Christmas Factor is a four-domain enzyme secreted into the bloodstream as a zymogen (an inactive precursor). The structure of the whole protein has been solved from a porcine source (PDB entry 1pfx) (view-1) but all domains of the human protein have been solved separately, or in pairs. Its structural coverage can be visualised with UniPDB. The role of Christmas Factor is to cleave an arginine-isoleucine peptide bond in the enzyme Factor X. In doing so, it activates this enzyme which is the next step of the clotting cascade that ultimately results in cross-linked fibrin and a blood clot.

Domains dissected

The N-terminal part of Christmas Factor is a gamma-carboxyglutamic acid (GLA) domain (view-2) which binds to platelet membranes in a calcium-dependent manner. GLA domains (Pfam family PF00594) are present in a number of human proteins involved in clotting, inflammation and cell death.

Chemical structure of gamma-carboxyglutamic acid (CGU).
These domains contain extensive post-translational modifications of glutamates to gamma-carboxyglutamates (abbreviated as GLA) that can interact with calcium ions. GLA is represented in the PDB as a modified residue with a different three-letter code CGU. As (view-2) shows there are a large number of these negatively-charged CGU residues in the sequence of the GLA domain. In fact, twelve out of the forty residues in this region of the protein are gamma-carboxyglutamic acid. The porcine Christmas Factor structure was determined in the absence of calcium so this domain is mostly disordered. As the exact conformation cannot be reliably determined this region has been given zero occupancy by the crystallographers.
The next two domains of Christmas Factor are epidermal growth-factor-like (EGF-like) domains (view-3). This type of domain (Pfam family PF00008) is found in a large number of secreted proteins and has a fold that is stabilised by three disulfide bonds. EGF-like domains bind to other proteins and they are likely to have that role in Christmas Factor as well. In addition they push the C-terminal catalytic domain away from the platelet surface. Christmas Factor has a very low catalytic rate on its own, but this is increased by five orders of magnitude when it binds to Factor VIII or with Factor VII and Tissue Factor. While no structures of these complexes have yet been solved, available data suggest that these other proteins do indeed bind to the EGF-like domains of Christmas Factor.

The Factor IX catalytic domain

After the second EGF-like domain we find the 'business end' of Christmas Factor, the catalytic domain, a trypsin-like serine protease that cleaves Factor X and activates it (view-4). However, before Christmas Factor becomes catalytically active, it must itself be activated by removal of the 35 residue peptide between the EGF-like domains and the catalytic domain. Other serine proteases, Factors XI or VII, are responsible for this cleavage, which causes a conformational rearrangement in which the new N-terminus becomes buried in the protein adjacent to the substrate-binding site. The two chains of the mature Christmas Factor (termed Factor IXa) remain covalently linked by a disulfide bond.
As in all serine proteases, the active site contains a catalytic triad of serine, histidine and aspartate (view-5) in the substrate-binding site. Recall that Christmas Factor cleaves an Arg-Ile bond in the substrate. The arginine residue is bound inside a deep pocket in the enzyme, which positions the following peptide bond so it can be attacked by the catalytic machinery. This pocket is the major determinant of substrate specificity and has an aspartate residue at its base that forms a salt bridge to the substrate arginine. In PDB entry 1pfx, the inhibitor D-Phe-Pro-Arg-chloromethylketone provides the arginine side-chain. This inhibitor mimics the substrate not only of Christmas Factor, but also of thrombin and some other serine proteases that cleave after an arginine residue. By binding covalently to the active-site residues, D-Phe-Pro-Arg-chloromethylketone prevents catalysis.

Poor rate explained

The poor catalytic rate of Christmas Factor on its own is explained by the structure of the catalytic and second EGF-like domains of the human protein (PDB entry 1rfn). In this structure, there are two loops, termed the '99' and '60' loops, that hinder the binding of polypeptide substrate to the active site. The conformational change required in these regions was highlighted by the structure of a 'super active' mutant of Christmas Factor (PDB entry 2wpi). This mutant, which differs in sequence from the normal protein at only two residues, has a thousand-fold higher activity than the wild type enzyme. Conformational changes in this mutant are observed in two '99' and '60' loops (view-6). In the normal pathway for clotting, a similar movement of these loops will be produced by binding of the other clotting factors and thereby increase the Christmas factor catalytic rate.

The original Christmas disease mutation

Stephen Christmas lived until 1993 when he died of AIDS, the result of being infected during a blood transfusion before screening for HIV contamination became routine practice. Despite his illness, he campaigned tirelessly for haemophilia sufferers and improvements in blood-product safety. In 1992, he made a further contribution to the understanding of haemophilia when he participated in a genotyping study to identify the mutation in his Factor IX gene. DNA sequencing revealed a guanine to cytosine substitution that produced an altered form of the protein with cysteine 252 mutated to a serine residue (ref. 2)
Although there are more than 140 known mutations in the human Factor IX protein that give rise to haemophilia B, this single change is the original Christmas disease mutation. In PDB entry 1rfn this cysteine has been renumbered to residue 42. As you can see in the structure (view-7) this Cys participates in a disulfide bond that stabilises the protein fold. The partner cysteine residue in this disulfide bond is in fact Cys 58 - so this is the immediate neighbour of the His 57 of the catalytic triad mentioned earlier.
So if you slip with the knife while carving the turkey, be very grateful that you have a fully functioning, stable Christmas Factor.

Further exploration

Because of its central role in coagulation, Christmas Factor is an attractive target for anti-coagulation drugs. Haemophilia B patients with reduced levels of Christmas Factor show only mild symptoms, suggesting that the enzyme could be partially inhibited without causing uncontrolled bleeding on injury. You can look at two structures of human Christmas Factor (PDB entries 3lc3 and 3lc5) that have been solved in complex with methanediamine derivatives of benzothiophene. These drugs mimic the substrate arginine side-chain which is naturally bound in the active site. This mini-tutorial shows how to use a fragment search to find small molecules in the PDB archive that contain benzothiophene. Christmas Factor is so similar to other trypsin-like proteases that cleave after an arginine residue (especially others in the clotting cascade) that finding inhibitors that are highly specific to this enzyme alone is a major challenge.
Disulfide bonds and their role in protein structure were described in more detail in the March 2011 Quips on NGF - you can re-visit that here.