Article Text
Abstract
Familial Mediterranean Fever (FMF) is a hereditary inflammatory disorder characterised by short self resolving attacks of fever and serositis, especially peritonitis and arthritis. The discovery of the FMF gene in the summer of 1997 has provided the first specific biological tool for FMF diagnosis, which can be pinpointed among various inflammatory disorders through mutation screening.
FMF is classically transmitted as an autosomal recessive trait. The FMF gene, MEFV, is located on chromosome 16p13.3 and comprises 10 exons. To date, 29 point-mutations have been detected in exons 1, 2, 3, 5, 9 and 10, but no gross rearrangement has been found. There are three mutational “hot spots”: two in exon 10 (at codons 680 and 694) and one in exon 2 (codon 148). The 5 most frequent mutation M694V (42%), V726A (9%), M680IG/C (8%), M694I (5%) and E148Q (5%) derive from a founder effect which likely accounts for their high prevalence. The other mutations are found in „T 1% FMF chromosomes, and about one third of the FMF alleles are to be identified. It is quite likely that a broad number of cases are not MEFV linked and suffer from “FMF like” symptoms. Four populations are commonly affected: Armenians, Arabs, Non-Ashkenazi Jews (NAJ) and Turks. Italians, Ashkenazi Jews and Spaniards should probably not be considered as rarely affected populations any longer, because the prevalence of FMF mutations is also relatively high in them. The disease is also sometimes suspected and genetically confirmed in patients with no classic ancestry (French, British¡K). Rare mutations are preferentially found in those cases.
The FMF clinical panel ranges from total absence of symptoms (incomplete penetrance) to amyloidosis. This relies at least partly on allelic heterogeneity. Homozygosity for M694V has been reproducibly demonstrated to be associated with a severe disease course. This genotype is practically fully penetrant. In contrast, about two third of E148Q homozygous individuals are asymptomatic. The trend is that genotypes including either E148Q or V726A are associated with low penetrance and low risk of amyloidosis, whereas genotypes including two of either M694V, M694I or M680I are associated with full penetrance and a high risk of amyloidosis. Modifiers also affect the FMF phenotype. Recently, a Turkish study showed that patients with the G138 polymorphism in exon 2 are more prone to develop amyloidosis. MICA, the Major Histocompatibility Complex class I chain-related gene A, and SAA are two non-MEFV modifier locus. The impact of M694V homozygosity on the age at onset is aggravated if patients also inherited MICA A9, whereas MICA A4 is associated with a milder form of the disease. Renal amyloidosis is enhanced in patients with the SAA-1 alpha/alpha genotype and in males, each factor acting independently.
The MEFV mRNA is about 4kb length. It is expressed in mature neutrophils, eosinophils, monocytes, and in some cancer cell lines. It is upregulated in vitro by some proinflammatory mediators such as IFN. The protein encoded by MEFV was called Marenostrin by the French group, and Pyrin by the International consortium. It is a 781 amino acid protein of a predicted 86 kd molecular weight. The predicted rodent proteins have considerable homology with humans, but do not have an intact C terminal domain, in which most FMF-associated mutations have been found in humans! It has been thought for a while that Marenostrin/Pyrin was a transcription factor, because of the presence of two nuclear localisation sites in the gene sequence, and because of the presence of a zinc-finger motif. But immunofluorescent staining studies showed that unexpectedly, the protein was cytoplasmic and colocalized with microtubules. These data have been challenged recently by the finding of a deleted isoform that specifically binds to the nucleus. Thus identification of the MEFV function remains a difficult task.