The tumor-necrosis-factor receptor-associated periodic syndrome: new mutations in TNFRSF1A, ancestral origins, genotype-phenotype studies, and evidence for further genetic heterogeneity of periodic fevers.

Mutations in the extracellular domain of the 55-kD tumor-necrosis factor (TNF) receptor (TNFRSF1A), a key regulator of inflammation, define a periodic-fever syndrome, TRAPS (TNF receptor-associated periodic syndrome [MIM 142680]), which is characterized by attacks of fever, sterile peritonitis, arthralgia, myalgia, skin rash, and/or conjunctivitis; some patients also develop systemic amyloidosis. Elsewhere we have described six disease-associated TNFRSF1A mutations, five of which disrupt extracellular cysteines involved in disulfide bonds; four other mutations have subsequently been reported. Among 150 additional patients with unexplained periodic fevers, we have identified four novel TNFRSF1A mutations (H22Y, C33G, S86P, and c.193-14 G-->A), one mutation (C30S) described by another group, and two substitutions (P46L and R92Q) present in approximately 1% of control chromosomes. The increased frequency of P46L and R92Q among patients with periodic fever, as well as functional studies of TNFRSF1A, argue that these are low-penetrance mutations rather than benign polymorphisms. The c.193-14 G-->A mutation creates a splice-acceptor site upstream of exon 3, resulting in a transcript encoding four additional extracellular amino acids. T50M and c.193-14 G-->A occur at CpG hotspots, and haplotype analysis is consistent with recurrent mutations at these sites. In contrast, although R92Q also arises at a CpG motif, we identified a common founder chromosome in unrelated individuals with this substitution. Genotype-phenotype studies identified, as carriers of cysteine mutations, 13 of 14 patients with TRAPS and amyloidosis and indicated a lower penetrance of TRAPS symptoms in individuals with noncysteine mutations. In two families with dominantly inherited disease and in 90 sporadic cases that presented with a compatible clinical history, we have not identified any TNFRSF1A mutation, despite comprehensive genomic sequencing of all of the exons, therefore suggesting further genetic heterogeneity of the periodic-fever syndromes.