(page 2 of 2)

Ruggeri came to TSRI from Milan in 1978 for what he thought would be a brief, one-year stay as a research associate. At that time, he was a practicing physician who specialized in hematology, particularly in treating patients with hereditary bleeding disorders like von Willebrand disease.

"I came here to acquire knowledge that I could apply to techniques to improve the diagnosis of these patients," says Ruggeri.

He soon became interested in the mechanism of von Willebrand disease and wound up staying an extra year before returning to Milan. When he did return in 1980 and began working with patients again, he found it impossible to ignore the scientific implications of what he saw. As associate director of the Hemophilia and Thrombosis Center at Policlinico Hospital in Milan, Ruggeri noticed that the von Willebrand factor proteins behaved differently in different patients with different forms of the disease.

Von Willebrand factor, the largest protein in human plasma, is a polymeric protein that is produced by endothelial cells that line blood vessels. One subunit of von Willebrand factor is more than 2,000 amino acids, and dimers of these subunits form the building blocks of the polymer. The polymers are large chains of various numbers of these dimers, and the size of these polymers is related to the function of the protein, which is to mediate the attachment of platelets to areas in the circulation where there is a lesion. The larger the polymers, the more sticky they become.

To induce blood clotting, the polymers should be as long as possible, since the number of binding sites increases with length (something biologists refer to as multivalency, with each subunit of a multimer contributing an equal number of binding sites).

Ruggeri observed that, in some patients, the von Willebrand factor was not as large (and sticky) as it needed to be. In others, it grew as large as it needed to induce blood clotting, but it was hyper-vulnerable to attack by a blood protease and was degraded too quickly. And sometimes, the von Willebrand factor proteins in patients seemed to have multiple problems.

These differences in the proteins translated into differences in how the platelets behaved—for instance, how they stuck to one another, and how they stuck to the surfaces of blood vessels. "The tests we were doing in [the patients] were showing different results, and it was clear that there were different categories of the disease," he says.

Point of No Return

Ruggeri returned to TSRI in 1982, armed with these observations and determined to examine the causes at the molecular level. In the years since, he has made many important discoveries related to the disease and its various phenotypes—which, it turns out, are caused by mutations in different parts of the von Willebrand factor molecule.

Prior to the arrival of modern molecular biology and genomics, this sort of work relied largely on characterizing phenotypes and performing molecular studies that were, by today's standards, crude. Ruggeri and others purified the von Willebrand factor protein from plasma and digested it with enzymes into its various domains. With the advent of molecular biological techniques that enabled investigators to express domains of a protein in large amounts, by the early 1990s several laboratories, including Ruggeri's in collaboration with TSRI Associate Professor Jerry Ware, began a line of work that has yielded several high-resolution structures of von Willebrand factor domains.

Ruggeri, in collaboration with Celikel and Varughese, solved the first such structure, the A1 domain of von Willebrand factor, in 1999. This was an arduous task as the A1 domain was difficult to crystallize. As a protein, von Willebrand factor is naturally sticky, which leads to trouble with solubility. But once Ruggeri and his colleagues crystallized and solved this first domain, similar studies with mutants and other structural variants became easier.

The A1 von Willebrand factor domain has a core of beta strands surrounded by a number of alpha helices. It appears to be identical to domains used by a whole family of other proteins—intergrins, for instance.

In von Willebrand factor, the A1 domain is crucial for blood clotting because it interacts with platelets. Not surprisingly, many of the mutations that have been reported in von Willebrand factor protein from patients with the disease are in this domain.

Gain of Function—In a Bad Way

Ruggeri describes how one mutation to the A1 domain von Willebrand factor protein causes von Willebrand disease. Paradoxically, this mutation does not knock out the function of the domain, but causes it to become more active. The mutation causes the A1 domain to bind to platelets too tightly.

Normally, von Willebrand factor is essential for bringing platelets to lesions on the surface of vessels through which blood is flowing rapidly. In order to do this, the proteins need to bind to both the platelets and to the collagen and other components of the matrix, which are exposed when a vessel is cut.

"But," says Ruggeri, "they are only meant to interact [with platelets] where there is a lesion."

The gain-of-function von Willebrand factor mutations cause the circulating von Willebrand factor proteins to bind to the platelets avidly, even when the platelets are in normal circulation. The platelets, in turn, become coated with the von Willebrand factor protein, and thus exhaust the sort of long, sticky, multimeric polymers needed to induce clotting. In the end, patients bleed because the platelets don't stick.

"They fly over the surface," says Ruggeri.

Ruggeri notes that several other mutations in von Willebrand factor function in different ways to cause the bleeding disorder. Such mutations can make the protein lose the ability to multimerize or to bind to the platelets at all.

The Future

Knowing the molecular bases of the various causes of von Willebrand disease is extremely helpful in designing therapies.

Patients who have the protein, but in whom it is not released in the blood can be treated with a peptide after they are cut. This peptide forces their cells to release stores of the von Willebrand factor protein and corrects the bleeding problem. This type of treatment is worthless for patients with different kinds of mutations.

On the other hand, patients who have a malfunctioning form of the protein per se are best treated by the administration of normal von Willebrand factor isolated from donated blood.

Asked about the future, Ruggeri notes that there is still a lot to do in the present. Today, he points out, even though the A1 domain and another domain of the von Willebrand factor protein have been solved, these account for only about one third of the total protein. No structure for the entire protein monomer exists and no structure for the dimer exists either.

"There is still a lot of work to do," says Ruggeri. "We are slowly building up our knowledge."



A hypothetical model of the von Willebrand factor A1 domain and alpha-thrombin concurrently bound to one molecule of glycoprotein Ib. Click to enlarge.