Background Familial Mediterranean fever (FMF) is a recessively inherited disorder characterized by self-limted attacks of fever with serosal, synovial, or cutaneous inflammation. In 1992 we mapped the FMF locus to chromosome 16p13.3, and in 1997 we isolated the underlying gene, MEFV. We found that this gene is expressed predominantly in granulocytes, and computational analyses of the predicted protein, pyrin, suggested that it might be a leukocyte-specific transcription factor. During the three years leading up to the present reporting period, we have (1) performed mutational studies that have more clearly defined the nature and population distribution of MEFV mutations; (2) demonstrated MEFV expression in myeloid bone marrow precursors and in cytokine stimulated monocytes, but little if any expression in lymphoid cells; (3) shown in transfected cell lines that full-length pyrin does not localize in the nucleus, but instead associates with microtubules and the actin cytoskeleton; (4) identified two pyrin-interacting proteins by the yeast two-hybrid assay: apoptosis-associated specklike protein with a caspase recruitment domain (ASC), and proline, serine, threonine phosphatase interacting protein (PSTPIP1); (5) cloned the mouse and rat homologs of pyrin; and (6) bred mice with a targeted disruption of Mefv in exon 3, leading to a truncated mouse pyrin. Results of the Last Year Characterization of the pyrin protein: During the 2001 reporting period significant insight into the possible function of pyrin emerged from sequence comparisons with newly identified proteins in the public database. The N-terminal ~90 aa of pyrin define a motif (the pyrin domain) found at the N-terminus of several proteins implicated in apoptosis or inflammation. Moreover, a recently identified gene (CIAS1) causing two distinct autoinflammatory disorders (Muckle-Wells syndrome [MWS] and familial cold urticaria [FCU]) has been shown to encode a protein (cryopyrin) with an N-terminal pyrin domain. Computer modeling performed in other laboratories indicates that the pyrin domain belongs to a structural superfamily that includes death domains, death effector domains, and caspase recruitment domains. In addition to these in silico analyses, we have also developed the first polyclonal antiserum that detects pyrin by immunohistochemistry. The antiserum detects overexpressed pyrin in transfected cells and endogenous pyrin in macrophages and neutrophils. Staining is consistent with previous data indicating a predominantly cytoplasmic localization for pyrin. Characterization of the pyrin-PSTPIP1 interaction: One of the pyrin interactors identified through the yeast two-hybrid assay is PSTPIP1 (also known as CD2BP1). During the last year we have identified the contiguous B-box zinc finger and coiled-coil (BCC) domains of pyrin as the most likely region of interaction. This conclusion is supported by yeast two-hybrid studies, subcellular colocalization of GFP-tagged pyrin BCC with native or epitope-tagged PSTPIP1, and by coimmunoprecipitation (co-IP) in transfected cells. In the yeast two-hybrid assay, at least one FMF-associated mutation in the C-terminal B30.2 domain of pyrin (V726A) is associated with reduced binding to PSTPIP1. By yeast two-hybrid and co-IP, the C-terminal SH3 domain of PSTPIP1 interacts with pyrin. Within this SH3 domain, the Y367 residue is critical to pyrin binding; co-IPs of site-directed mutants, or of lysates from pervanidate-treated cells, indicate that phosphorylation of this residue increases binding to pyrin. Recently, two mutations in PSTPIP1 have been shown by others to cause a variant of pauciarticular juvenile rheumatoid arthritis. Our data predict that these PSTPIP1 mutations inhibit Y367 dephosphorylation, leading to increased pyrin binding. Downstream effects of MEFV expression in transfected human cell lines: We are developing several systems of transient and conditional MEFV expression in human cell lines to study the global effects of MEFV on gene expression in cDNA microarrays. Proof-of-principle studies are currently underway using a limited number of genes involved in inflammation and apoptosis. Studies of the murine homolog of pyrin: By Westerns, we have found two isoforms of mouse pyrin (differing by 30 aa) generated by alternative splicing within exon 2. In transfected cells, using both anti-pyrin antibodies and epitope tags, mouse pyrin localizes in the cytoplasm, and interacts with the mouse homologs of ASC and PSTPIP1. Pyrin expression in murine monocytes is inducible by IL-4, IL-10, TNF alpha, and LPS. Pyrin appears more highly inducible in male than female mice. Pyrin expression is not induced by TNF-alpha in TNFRSF1A-KO mice, by IL-4 or IL-10 in JAK3-KO mice, or by any stimuli in NF-kappaB-KO or LPS-unresponsive C3H/HeNHsd mice. Transient transfection of Mefv into the mouse monocytic cell line RAW264.7, which does not ordinarily produce pyrin, leads to DNA fragmentation characteristic of apoptosis. Characterization of pyrin truncation mice: Initial studies of mice with a targeted disruption of exon 3 of Mefv were begun in FY2000. Although intended to be null mutants, these mice in fact express a truncated form of pyrin that retains the N-terminal pyrin domain, but lacks the BCC domains important for interaction with PSTPIP1. We had previously found that these mice develop normally, and that some but not all have intermittent fevers. We have now begun studies of the sensitivity of these animals to LPS, and our preliminary data indicate increased mortality in males but not females. We have also found that the truncated pyrin redistributes from the cytoplasm to the nucleus, and that peritoneal macrophages from mutant mice are much less susceptible to the induction of apoptosis than their wild-type or heterozygous littermates. Western blots show that caspase 8 cleavage is markedly reduced in the mutant macrophages, whereas caspase 9 cleavage is normal. By RNase protection assays, caspase 11 message is increased in macrophages from the mutant animals. Conclusions and Significance Taken together, the newly derived sequence comparison data, the interaction of pyrin with the proapoptotic molecule ASC, and our data from Mefv truncation mice strongly suggest a role for pyrin in apoptosis. The precise details may in part be regulated by the interaction between pyrin and PSTPIP1, which we have now begun to elucidate. The broader clinical importance of these pathways is underscored by the recent identification of mutations in a novel pyrin homolog in two other autoinflammatory disorders, as well as the discovery of mutations in PSTPIP1 in a variant of juvenile rheumatoid arthritis. We propose that, in the disorders of the pyrin pathway, there may be a defect in leukocyte apoptosis, leading to the persistence of leukocytes in the early stages of the inflammatory cascade, thus permitting the amplification of what would ordinarily be innocuous stimuli. During the next year, our objectives will be: (1) further elucidation of the interaction of pyrin with PSTPIP1, with particular attention to the effects of disease-associated mutations; (2) a new clinical initiative to study patients with mutations in proteins in the pyrin pathway; (3) continuation of studies of apoptosis in pyrin-truncation mice; (4) microarray analysis of the downstream effects of MEFV expression; (5) studies of pyrin null mice now under development; (6) development and characterization of knockin mice with FMF-associated pyrin mutations; and (7) crystallographic studies of pyrin.
Showing the most recent 10 out of 35 publications
