BackgroundFamilial Mediterranean fever (FMF) is a recessively inherited disorder characterized by self-limited episodes of fever typically accompanied by serosal, synovial, or cutaneous inflammation. Less commonly, patients may develop a range of inflammatory disorders, including Henoch Schonlein purpura, aseptic meningitis, and orchitis. Systemic AA amyloidosis may also develop, the risk of which appears to be influenced both by genetic and environmental determinants. FMF has been most often reported in non-Ashkenazi Jews, Armenians, Arabs, and Turks. In 1992 our laboratory mapped the gene causing FMF to the subtelomeric region of chromosome 16p; in 1997 we identified the FMF gene (designated MEFV) by positional cloning. MEFV is comprised of 10 exons spanning approximately 14 kb of genomic DNA; it was not present in any of the existing protein or expressed sequence tag databases. The first 3 mutations we identified were conservative missense mutations clustered in exon 10. The predicted protein, which we denoted pyrin, bears strong homology in its C-terminal half to the ribonucleoprotein Ro52 and to several other nucleic acid-binding proteins. Northern analysis demonstrated MEFV expression primarily in granulocytes. During the 1998 reporting period we concentrated on characterizing MEFV and its mutations in more detail. Among the major advances were: 1) identification of 6 novel disease-associated mutations (bringing the total number of known mutations to 16); 2) documentation of substantial numbers of American FMF patients with Ashkenazi Jewish and Italian ancestry, and demonstration of a 20% carrier frequency in the U.S. Ashkenazi population; 3) demonstration of MEFV expression in granulocytic bone marrow precursors, mature eosinophils, and activated monocytes; 4) development of polyclonal rabbit antibodies against pyrin; and 5) cloning and characterization of the mouse homolog of MEFV, and development of a targeting construct for a knockout mouse model. Objective of Present StudiesDuring the 1999 reporting period we have focused on 4 objectives: a) to extend mutational analyses and studies of gene and carrier frequencies in high-risk populations; b) to complete studies on the developmental regulation of MEFV expression in leukocytes; c) to study the expression, subcellular localization, structure, and function of the pyrin protein; and d) to characterize in more detail the rodent homologs of MEFV, and to develop relevant knockout and knockin mice.Results of the Last YearMutational analysis and population genetics: We have identified 3 new disease-associated mutations in MEFV: R42W (exon 1), R408Q (exon 4), and R653H (exon 10). Screening for all 19 presently known MEFV mutations, there are still a substantial number of individuals meeting clinical criteria for FMF in whom we have found only one disease associated mutation. For several of these individuals, we performed genomic sequencing of the complete coding region, splice junctions, and approximately 1 kb of the 5 prime untranslated region, and did not find a second mutation. Moreover, in a panel of Israeli patients with one M694V mutation and no identifiable second mutation, we found that the likelihood of developing FMF symptoms was markedly increased in patients who also had Behcet disease. These observations suggest that, on some permissive genetic backgrounds, a single MEFV mutation may be associated with clinical manifestations of FMF. We have also performed mutational screens on panels of DNA obtained from asymptomatic individuals, in order to determine population-specific carrier frequencies. In southern Italians, we found a carrier frequency for MEFV mutations of 12%, with E148Q being the most frequent at 8%. In the Armenian population, the carrier frequency was 47%, with an M694V carrier frequency of 19%. Figures of this magnitude strongly suggest the possibilty of heterozygote selection. Expression studies in leukocyte subpopulations: During the last year we have confirmed and extended our earlier observations on the expression of MEFV in mononuclear cells. In studies of cultured cell lines, we found MEFV expression in the monocytic cell lines U937 and THP-1. In monocytes from normal donors, we have found variable levels of message in unstimulated cells, and induction of message by interferons (IFNs) alpha and gamma, tumor necrosis factor alpha, and bacterial lipopolysaccharide (LPS). Kinetic experiments demonstrated MEFV expression within 30 minutes of IFN gamma or LPS stimulation. Taken together with our observation that IFN-gamma induction is not inhibitable by cycloheximide, these data indicate that MEFV is an IFN-gamma immediate early gene. MEFV induction can be blocked by the anti inflammatory cytokines IL-4, IL-10, and transforming growth factor beta. Message expression in mature granulocytes could be induced by IFN alpha and by the combination of IFN alpha and colchicine. It was not induced by C5a, TNF alpha, GM-CSF, G-CSF, LPS, or IL-4.Studies of the pyrin protein: Using polyclonal rabbit antibodies developed during the 1998 reporting period, we have demonstrated induction of pyrin protein by Western blot and immunoprecipitation in monocytes stimulated with LPS and/or IFN gamma. Pyrin could not be detected with these antibodies in resting or stimulated granulocytes, despite the relative abundance of MEFV message in these cells. In vivo labeling studies in MEFV-transfected cells demonstrated that pyrin is phosphorylated, regardless of the activation status of the cells. We have studied the subcellular localization of pyrin by utilizing fusion constructs with a green fluorescent protein (GFP) tag. Both N terminal and C-terminal GFP fusion proteins localized to the cytoplasm of transfected 293T, THP-1, NIH3T3, and HeLa cells by fluorescence microscopy. Stimulation by LPS and IFN gamma did not alter localization. Subcellular localization of deletion mutants will be tested to determine whether pyrin has a cytoplasmic retention domain. After unsuccessful attempts to express soluble pyrin in bacterial and baculovirus expression systems, we have developed a mammalian expression system that produces myc- and his-tagged and untagged full length pyrin in a soluble form. We have also developed a purification scheme for tagged recombinant pyrin, which we plan to use in crystallographic studies during the next year. Development of animal models: Part of our efforts in this area have centered on characterizing the rodent homologs of MEFV. During the 1998 reporting period we had cloned the mouse homolog, and expression studies carried out in the last year indicate that the mouse gene, like its human counterpart, is expressed in peripheral blood granulocytes but not lymphocytes, and can be induced by pro-inflammatory cytokines in monocytes. Consistent with expression in granulocytes, we found high levels of expression in the primary follicles and marginal zones of the splenic white pulp. We have recently identified the rat homolog of MEFV. It has 9 exons, with a coding sequence of 2253 bp (751 aa). The sequence is 82.1% similar to the mouse, and 73.5% identical to the mouse sequence. The predicted rat protein contains a B-box zinc finger domain, Robbins Dingwall nuclear localization signal, and coiled-coil domain, but, like the mouse, does not contain a B30.2 domain. Since developing a knockout construct during the last reporting period, we have now introduced this recombinant DNA into mouse embryonic stem cells and have developed lines with the null mutation. In collaboration with investigators at NHGRI, we performed blastocyst injections, and subsequently confirmed germline transmission from wild type/knockout chimeric mice. This summer we interbred heterozygous knockout mice, and our first litters with homozygous knockouts were born within the last several weeks. The mice have no gross developmental abnormalities, and more detailed phenotypic studies are now under way. Constructs have also been developed for wild-type, E148Q, M680I, and V726A knockin mice. Embryonic stem cells have been transfected, and clones are currently being screened for the desired genotype. Conclusions and SignficanceDuring the last year we have extended our understanding of the FMF gene and its protein product. We have increased the number of known disease-associated mutations to 19, documented high carrier frequencies in the southern Italian and Armenian populations, and examined the possibility that in some circumstances a single mutation at MEFV may be sufficient for disease. Our investigations of MEFV expression indicate that the gene is under the regulation of pro- and anti-inflammatory cytokines in both granulocytes and monocytes, and that MEFV is an IFN gamma immediate early gene. Studies of the pyrin protein are consistent with induction by proinflammatory mediators, but call into question our working hypothesis that pyrin acts directly as a transcription factor. The recent development of knockout mice will provide a powerful tool to understand the function of pyrin in health and disease. During the next year, our objectives will be:1) to perform mutational and genotype-phenotype studies on selected referrals to the Clinical Center; 2) to develop a clinical protocol to evaluate alternative treatments (such as interferon alpha and etanercept) in FMF patients who do not respond to, or cannot tolerate, colchicine; 3) to identify proteins that interact with pyrin by the yeast two-hybrid assay and/or immunoprecipitation; 4) using cDNA microarrays, to determine the downstream consequences of pyrin expression; 5) to begin crystallographic studies of pyrin; 6) to characterize the phenotype of knockout mice; and 7) to continue development of knockin mice.
Showing the most recent 10 out of 35 publications
