Blood, 15 September 2001, Vol. 98, No. 6, pp. 1643-1644
The vWF-cleaving protease: new opportunities in TTP
Thrombotic thrombocytopenic purpura (TTP) is a rare but
increasingly encountered disorder associated with abnormalities of plasma von Willebrand factor (vWF). Moschcowitz reported the first case in 1924, and a major review of TTP in 1966 described the clinical
features of 271 patients. TTP was characterized by a classic pentad of
signs, present to variable degrees: microangiopathic hemolytic anemia,
thrombocytopenia, neurological dysfunction, renal failure, and fever.
No effective treatment was known, and mortality was an impressive 90%.
Today, plasma exchange therapy has reduced the mortality to about 20%,
although why it works is not certain.
Many etiologies for TTP have been proposed, including endothelial
damage and defects of platelets or plasma components. The first
association with vWF was made in 1982, when Moake discovered that
"ultra-large" vWF (UL-vWF) multimers occurred in patients with
relapsing TTP. The absence of large vWF multimers impairs platelet
adhesion and causes bleeding, particularly in von Willebrand disease
type 2A. Conversely, UL-vWF multimers were proposed to mediate platelet
aggregation, thrombosis, and microangiopathic hemolysis in TTP.
The UL-vWF theory of TTP gained considerable support with the discovery
in 1997 that most patients with congenital relapsing TTP lack a plasma
protease that cleaves vWF and converts large multimers into smaller
forms. The following year, 2 studies documented that acquired TTP is
associated with autoimmune IgG inhibitors of this protease. These
observations suggested that plasma exchange is beneficial because it
removes inhibitors (if present) and infuses a specific protease. But
plasma exchange is expensive and cumbersome and has significant risks.
Furthermore, the proposed role of UL-vWF multimers remained dependent
on correlation rather than direct evidence. Clearly, identifying the
vWF-cleaving protease would open new scientific, diagnostic, and
therapeutic opportunities for this enigmatic disorder.
In this issue, 2 groups report the purification and amino-terminal
sequence of the vWF protease. Gerritsen and colleagues (page 1654) used
the IgG fraction from a patient with TTP containing an antibody
protease inhibitor to purify the protease from plasma mainly by
immunoaffinity chromatography. Fujikawa and coworkers (page 1662) used
friendlier starting material than plasma, factor VIII/vWF concentrates,
and employed conventional chromatography on heparin, ion exchange, and
size exclusion columns. From the products of these independent
purification methods, both groups obtained the same protein sequence.
The vWF protease is a 150-kd single-chain glycoprotein that requires
divalent metal ions. Exploring the human genomic database, Fujikawa and
colleagues found a match between their N-terminal protein sequence and
an EST cluster from chromosome 9q34. This composite cDNA sequence
encodes a metalloprotease belonging to the ADAMTS subfamily, although
the expected thrombospondin domain is not part of this partial
sequence. ADAMTS proteases typically contain both zinc and calcium
ions, and the congruence between the known properties of the vWF
protease and ADAMTS proteases provides additional compelling evidence
that the vWF protease has been identified.
To what use might this initial information be applied? It is likely
that polyclonal and monoclonal antibodies can be prepared against the
deduced protein sequence, making quantitation of plasma protease
antigen and a rapid diagnostic test for TTP no more than a clinical
ELISA away. But a significant fraction of patients with microangiopathy
have normal levels of this protease, and new tools based on this new
protease should facilitate the dissection of these potentially distinct
entities that converge on a common TTP-like phenotype. Ultimately,
completion of the cDNA sequence will lead to production of recombinant
protein, which could enable specific therapy for the rare patient with
congenital TTP and possibly for patients with low-titer autoantibodies.
Eventually we should be able to prove or refute the UL-vWF multimer
theory of TTP, thereby moving us one step closer to a cure.
Kenneth Kaushansky
University of Washington
