AN AUTOSOMAL DOMINANT PERIODIC FEVER ASSOCIATED WITH AA AMYLOIDOSIS IN A NORTH INDIAN FAMILY MAPS TO DISTAL CHROMOSOME 1q

2034 ARTHRITIS & RHEUMATISM Vol. 43, No. 9, September 2000, pp 2034 2040 2000, American College of Rheumatology AN AUTOSOMAL DOMINANT PERIODIC FEVER ASSOCIATED WITH AA AMYLOIDOSIS IN A NORTH INDIAN FAMILY MAPS TO DISTAL CHROMOSOME 1q MICHAEL F. MCDERMOTT, EBUN AGANNA, GRAHAM A. HITMAN, B. WILLIAM OGUNKOLADE, DAVID R. BOOTH, and PHILIP N. HAWKINS Supported by the Research Advisory Committee of the Special Trustees of St. Bartholomew s and the Royal London Hospitals N.H.S. Trust, the Medical Research Council, and The Wellcome Trust. Michael F. McDermott, MRCPI, Ebun Aganna, BSc, Graham A. Hitman, MD, FRCP, B. William Ogunkolade, PhD: St. Bartholomew s and the Royal London School of Medicine, London, UK; David R. Booth, PhD, Philip N. Hawkins, PhD, FRCP: Royal Free Hospital, London, UK. Address reprint requests to Michael F. McDermott, MRCPI, 5th Floor, Alexandra Wing, The Royal London Hospital, Whitechapel, London E1 1BB, UK. Submitted for publication February 18, 2000; accepted in revised form April 27, 2000. Objective. To investigate genetic susceptibility in the first Indian family identified as having an autosomal dominantly inherited periodic fever syndrome. The inflammatory disease was characterized chiefly by arthralgia, skin rashes, and AA amyloidosis. Methods. Markers from known periodic fever susceptibility loci were investigated in 7 affected and 11 healthy members of a north Indian family. These included the TNFRSF1A locus (formerly known as TN- FRI), which is involved in autosomal dominant tumor necrosis factor receptor associated periodic syndrome on chromosome 12p13, the familial Mediterranean fever locus (MEFV) on chromosome 16p13, the hyperimmunoglobulinemia D and periodic fever syndrome (HIDS) locus on chromosome 12q24, and the Muckle-Wells syndrome/familial cold urticaria (MWS/FCU) locus on distal chromosome 1q44. Results. Linkage to both TNFRSF1A and MEFV was definitively excluded, and DNA sequencing of these genes revealed no mutations. Furthermore, there was no evidence of linkage to the HIDS locus. In contrast, significant logarithm of odds scores for 5 markers from the MWS/FCU region were obtained in this family, and the disease segregated with the same haplotype in all affected members. Conclusion. We have identified an inherited inflammatory disease in a north Indian family with clinical features overlapping some of those of MWS and FCU. The susceptibility gene maps to distal chromosome 1q44, a region already implicated in both MWS and FCU. Different mutations in the same (or a closely related) gene may be responsible for an inflammatory disease with a broad phenotype among diverse ethnic populations. The hereditary periodic fever syndromes are a group of genetic multisystem disorders characterized by recurrent episodes of fever in association with inflammation that variably affect many organ systems, most notably the joints, skin, and serosal linings. Although some of the individual features of these diseases are shared, there are clear distinctions in their mode of inheritance, overall pattern of clinical features, and frequency of symptoms. Autosomal dominant recurrent fever (ADRF) syndromes have been reported in several ethnic groups, the majority of which are northern European in origin (1). The most common and extensively studied is tumor necrosis factor receptor associated periodic syndrome (TRAPS) (2), which incorporates familial Hibernian fever (FHF; Mendelian Inheritance in Man [MIM] number 142680) (3); among the less common conditions in this category are Muckle-Wells syndrome (MWS; MIM191900) (4) and familial cold urticaria (FCU; MIM120100) (5). MWS is a rare disorder which presents with various combinations of urticaria and progressive sensorineural deafness, and may be complicated by reactive systemic (AA) amyloidosis; FCU is also a rare disease of unknown pathogenesis, characterized by urticarial wheals, swollen painful joints, and chills and fever after exposure to cold, with onset in early life. The FCU gene has been mapped to distal chromosome 1q44 in 5 North American families, 2 of whom had definite northern

GENETIC MARKERS FOR AUTOSOMAL DOMINANT PERIODIC FEVER 2035 European ancestry (6). Linkage analysis in an extensive French Canadian pedigree supported these findings (7) and also implicated the same chromosomal region in MWS in a German family it was suggested that the FCU gene might be allelic with the MWS gene. The chromosome 1q44 localization for the MWS gene has been replicated in a study of 1 British and 2 French families (8). Susceptibility genes for the 2 main autosomal recessive conditions, familial Mediterranean fever (FMF; MIM249100, chromosome 16p13) (9,10) and hyperimmunoglobulinemia D and periodic fever syndrome (HIDS; MIM260920), (chromosome 12q24), have also been identified (11,12). There has also been substantial progress in elucidating the molecular basis of FMF and HIDS. Mutations of the MEFV (Mediterranean Fever) gene, encoding the pyrin/marenostrin protein, which appears to be expressed only in neutrophils, cause FMF. Mutations in the mevalonate kinase gene underlie susceptibility to HIDS. It is therefore now possible by mutation detection and linkage studies to use molecular analysis for the diagnosis of periodic fevers. We have investigated a multiplex Indian family that includes 7 living and 3 deceased members who had intermittent prolonged attacks of fever ( 1 week in duration), urticarial skin rashes, and periorbital edema. The attacks were associated with moderate-to-severe arthralgia and an intense acute-phase plasma protein response. Since the disease phenotype of the more severely symptomatic proband resembled the reported symptoms of TRAPS, our initial strategy was to sequence the coding region of the TNFRSF1A gene in the proband. When we were unable to find a mutation, we genotyped all family members with TNFRSF1A locus markers and formally excluded this region from containing the disease gene. We also excluded the possibility of pseudodominant FMF (13) by sequencing the MEFV gene in the proband, in addition to using markers flanking this gene, which is mutated in FMF. In addition, we excluded linkage to the HIDS locus. The susceptibility gene in this Indian family mapped to the same region of chromosome 1q44 that has been linked with MWS and FCU (6 8), and the clinical features share some of the features of both of these diseases. PATIENTS AND METHODS The multiplex Indian family. The kindred, which is based in New Delhi, India, includes at least 25 living firstdegree family members and was identified when the male proband referred himself at age 30 years with nephrotic syndrome due to AA amyloidosis. The proband did not have a recognized chronic inflammatory disease, but on direct questioning he reported a 20-year history of self-limiting episodes of pain and swelling in the ankles and feet, often associated with fever of 1 week in duration and sometimes lasting up to 4 weeks, with small urticarial lesions on the arms and legs that responded partially to regular treatment with colchicine. Frequent additional features included periorbital edema and conjunctivitis. The proband s father, a paternal uncle, and a male cousin had died of AA amyloidosis at ages 35, 32, and 26 years, respectively, all having had a virtually identical clinical syndrome. Seven living first-degree relatives had similar but milder symptoms. Most affected individuals first developed joint pains at the age of 10 years, and these were typically confined to the knees and elbows initially, but included the wrists and finger joints at a later stage. The arthralgia built up slowly and generally resolved within 1 2 days, sometimes moving from one joint to another, and was frequently associated only with erythema and mild joint swelling. There was no clinical evidence of destructive joint changes in the proband. The rash comprised small, nonpruritic urticarial lesions confined to the arms and legs. Deafness was not a feature of the disease in any individual. On direct questioning, and after TRAPS had been excluded, most affected individuals reported that some of the symptoms, particularly the joint pains, might have been provoked or exacerbated by exposure to cold or were certainly more frequent during winter months. However, there was no evidence of Raynaud s phenomenon. There was a strong suggestion that clinical manifestations of the disease were milder in females than in males, and that the severity of the symptoms often decreased after the age of 50 years. No affected female was known to have developed amyloidosis, and the proband s grandmother had lived until age 79 years with only very mild features of the disease. Eighteen members of the family, including 7 individuals with symptoms of the disease, provided blood samples for further investigation. DNA extraction and microsatellite genotyping. DNA was prepared from 200 l of whole anticoagulated blood by the salt precipitation procedure (14). Fluorescent-labeled primers were designed from published sequences for microsatellite marker amplification, and polymerase chain reaction (PCR) was performed using touchdown conditions as described (15). TNFRSF1A locus (chromosome 12p13). An informative microsatellite marker from the first intron of the TNFRSF1A gene (16), in addition to the D12S99 and D12S77 flanking markers, was used to genotype all available family members, both affected and unaffected. MEFV locus (chromosome 16p13.3). Four microsatellite markers providing comprehensive coverage of the MEFV locus over a 20-cM distance were used: D16S3070 (telomeric) and D16S3275 (centromeric), which tightly flank the MEFV locus; D16S283, 2 5 cm telomeric to MEFV; and D16S423, 8 15 cm centromeric to MEFV. MWS/FCU locus (chromosome 1q44). We used the D1S423 and D1S2836 microsatellites, which define the boundaries of the MWS critical interval (8). We also used 4 markers (D1S491, D1S459, D1S2785, and D1S2842) centromeric to the MWS locus, as well as the single telomeric marker (D1S2682).

2036 MCDERMOTT ET AL HIDS locus (chromosome 12q24). We chose the D12S79 and D12S306 markers, which flank the HIDS locus (11). Genotypings were performed using the 373A automated sequencer (Perkin Elmer ABI, Foster City, CA), and fragments were analyzed using Genotyper software (Perkin Elmer ABI) on a Macintosh Quandra 650 computer (Apple, Somerville, MA). A mean of 3 4 markers/lane were combined for electrophoretic multiplexing, and allele sizes were calculated on the basis of an internal size standard in each lane (Genescan 350 or 500 Tamra; Perkin Elmer ABI). All genotypings were scored by an investigator who was blinded to phenotype, and in the case of uncertainty about allele size, genotypings were performed at least twice. Allele classification and frequencies were taken from published sources (Généthon database, Evry, France). Mutation detection by fluorescent sequencing. Approximately 100 ng of genomic DNA template was used in separate PCR reactions to amplify exons 1 6 and flanking intronic sequences of the TNFRSF1A gene, as well as exons 1 10 of the MEFV gene as described (2,17). PCR products were purified, sequenced with dye primer chemistry (Amersham, Buckinghamshire, UK), and run on a 373A automated sequencer (Perkin Elmer ABI). Sequence data were analyzed with Sequence Navigator software (Perkin Elmer ABI). Linkage analysis. The feasibility of a linkage study in this Indian family was explored by use of the Slink program (18,19). Simulation analysis was performed using a dominanttransmission model, with age-specific penetrances of 85% and 95% for individuals of ages 40 years and 40 years, respectively. These values were based on our observation that all affected patients had disease onset by age 40, usually by age 25. The probability of a normal homozygote being diagnosed as affected was set at 0.00001 for each age group (to allow for the possibility of phenocopies), and the frequency of the abnormal allele was set at 0.00001 (in view of the relative rarity of the disease). Simulation analysis showed that detection of significant linkage (logarithm of odds [LOD] score 3.0) was feasible, given the pedigree structure in the family. For 2-point linkage analysis, LOD scores between the disease locus and each individual marker were calculated by the Mlink program of the Linkage 5.2 package (15), using the same dominant-transmission model used for Slink analysis. The analysis was performed at the Human Genome Mapping Project (HGMP) Resources Centre (Hinxton, Cambridge, UK), using Fastlink version 3.0 (20). Multipoint analysis of the disease against the markers at individual loci was performed using Vitesse. Sequential 4-point multipoint analysis was performed (21), and results were plotted contiguous to each other (22). Marker order and intermarker distances were based on existing linkage maps. For the multipoint analysis at the MWS/FCU locus, the following map was used: D1S491 D1S459 (19 cm), D1S459 D1S2785 (18 cm), D1S2785 D1S2842 (7.6 cm), D1S2842 D1S423 (5.4 cm), D1S423 D1S2836 (7.4 cm), D1S2836 D1S2682 (2.6 cm) (see Figure 1). Haplotype analysis was also performed, by parsimoniously minimizing the number of recombinants, to determine the degree of haplotype sharing between affected family members. Soluble TNFR levels. Soluble TNFRSF1A and TNFRSF1B (formerly TNFRII) levels were measured in plasma by solid-phase enzyme-linked immunosorbent assay Figure 1. Schematic regional map of chromosome 1q44. Markers used for linkage analysis in this study are indicated, and respective intermarker distances are given in cm. MWS/FCU Muckle-Wells syndrome/familial cold urticaria. (R&D Systems, Abingdon, UK). Briefly, standards and diluted samples were pipetted into microtiter plates precoated with either TNFRSF1A- or TNFRSF1B-specific monoclonal antibodies. After washing, an enzyme-linked polyclonal antibody specific for the relevant receptor was added. After a second wash, substrate solution was added to the wells, producing a color change proportional to the amount of TNFRS1A or TNFRSF1B present. Optical densities were measured at dual wavelengths to correct for optical imperfections in the plate. The assays measure the total amount of receptor present, i.e., free receptor plus receptor bound to TNF. All samples were analyzed in duplicate. Amyloidosis studies and laboratory markers of inflammation. The proband s renal biopsy sample was obtained, and the amyloid deposits were typed immunohistochemically (23). The organ distribution of the proband s amyloid was determined using whole-body serum amyloid P (SAP) component scintigraphy (24). Objective evidence of inflammatory disease was determined in each of the 18 family members by estimating the plasma concentration of serum amyloid A (SAA) protein and C-reactive protein (CRP) every 2 weeks on 3 6 occasions in each subject (25). Suppression of the inflammatory disorder by regular prophylactic use of colchicine was determined in the proband by measuring his plasma SAA level every 2 weeks for 2 months while he was not taking the drug, followed by further measurements every 2 weeks after he restarted colchicine at 1 mg twice a day. RESULTS To test the hypothesis that 1 of the recognized periodic fever susceptibility genes might confer susceptibility in this family, we studied the inheritance of informative markers from these loci. The TNFRSF1A gene was excluded as the susceptibility locus in this family, since sequencing of exons 1 6 of the external domains of this gene (where all reported mutations have been found) revealed a normal sequence. Likewise, the MEFV gene was excluded since no mutations were found in the coding sequences. Furthermore, multipoint LOD scores for markers flanking the TNFRSF1A locus indicated that the exclusion region (LOD score less than 2) extended for 16 cm in the family; results of haplotype studies supported these findings (data not shown). Two-point LOD scores for the D16S423,

GENETIC MARKERS FOR AUTOSOMAL DOMINANT PERIODIC FEVER 2037 Table 1. 1q44* Cumulative 2-point LOD scores for markers spanning the MWS/FCU locus on chromosome Recombination fraction ( ) Marker 0.0 0.01 0.05 0.10 0.20 Z max max D1S491 0.86 0.77 0.53 0.41 0.37 D1S459 3.41 3.87 2.74 1.85 0.92 D1S2785 2.28 2.83 2.64 2.39 1.84 2.28 0.00 D1S2842 3.05 2.99 2.78 2.50 1.90 3.05 0.00 D1S423 3.55 3.49 3.26 2.95 2.29 3.55 0.00 D1S2836 1.45 1.43 1.34 1.23 0.98 1.45 0.00 D1S2682 3.48 3.43 3.19 2.89 2.23 3.48 0.00 * Maximal logarithm of odds (LOD) scores (Z max ) at a recombination fraction ( ) of 0 in the region containing the MWS/FCU gene show that disease is linked in this Indian family. D16S3275, and D16S283 markers flanking the MEFV locus were all significantly negative ( 4.54, 4.41, and 9.71, respectively, at a recombination fraction [ ] of zero; D16S3070 was uninformative). Likewise, the HIDS locus was excluded on the basis of 2-point LOD scores of 4.98 and 6.12 at 0 for D12S79 and D12S306, respectively. Two-point LOD scores between the disease and 7 markers spanning the MWS/FCU locus were positive for the 5 most telomeric markers and significantly positive for D1S2842, D1S423, and D1S2682, as shown in Table 1. Genotypes obtained for the markers in this family are shown in Figure 2, arranged according to probable haplotypes. A single haplotype, which was not found in any unaffected individuals, was shared between affected subjects. Informative recombinants on this haplotype were seen between D1S459 and D1S2785 in 3 affected members, and therefore a centromeric limit for the Figure 2. Pedigree of an Indian family with Muckle-Wells syndrome/familial cold urticaria. Crosshatched symbols indicate affected individuals, and blackened symbols indicate those with amyloidosis. The haplotypes for 7 markers from chromosome 1q44 are included, and marker order is shown at the top. A single haplotype (boxed) segregates with disease. Letters atop columns of numbers indicate haplotype designation. The haplotype contained within the box indicates the particular markers that segregate with disease and also defines the haplotype breakpoints.

2038 MCDERMOTT ET AL disease susceptibility locus was defined between D1S459 and D1S2785 in this family; however, there were no informative recombinants involving the most telomeric markers. Multipoint analysis resulted in a maximal LOD score (Z max ) of 4.07 at the most telomeric marker (D1S2682). The 1-LOD support interval spanned over 30 cm, extending to the telomere, and due to the lack of breakpoints between 5 adjacent markers, the multipoint curve was flat over this interval (Figure 2). The critical interval was therefore wider than reported previously for the MWS/FCU locus, but did completely encompass the 13.9-cM critical interval reported by Cuisset et al (8) (Figure 1). Intermarker LOD scores were consistent with the published map. Soluble TNFRSF1A and TNFRSF1B levels. All members were found to have normal levels of soluble TNFRSF1A in their sera, except for 1 unaffected child who had a TNFRSF1A level that was 85% of normal. A second unaffected child had a soluble TNFRSF1B level that was 80% of normal, while TNFRSF1B levels were normal in all other family members. Amyloidosis studies and laboratory markers of inflammation. The renal amyloid deposits in the proband were confirmed immunohistochemically to be type AA. Whole-body SAP component scintigraphy demonstrated amyloid deposits in the proband s spleen, kidneys, and adrenal glands, a pattern of visceral organ involvement completely typical in type AA (23). Monitoring of the plasma concentrations of SAA protein and CRP every 2 weeks confirmed the presence of an intermittent, substantial acute-phase response in each of the 7 symptomatic individuals, and no evidence of inflammatory disease in the others. The plasma concentration of SAA protein exceeded 50 mg/liter on at least 1 occasion in all of the individuals with the syndrome, most of whom had peak levels of 200 mg/liter (median 3 mg/liter for healthy individuals). The median SAA protein concentration in the proband was 170 mg/liter in the absence of colchicine prophylaxis and 19 mg/liter when he was receiving the drug regularly. DISCUSSION There has been substantial recent progress in elucidating the molecular basis of hereditary periodic fever syndromes, which has proved to be very diverse and has opened several completely new avenues of study in inflammation and molecular diagnosis. All of these disorders are compatible with normal life expectancy. However, since they all can stimulate a substantial acute-phase plasma protein response, a proportion of patients develop AA amyloidosis, which is potentially fatal unless the underlying inflammatory condition can be suppressed. This is the first report of autosomal dominant periodic fever in patients from the Indian subcontinent, and the clinical presentation in this particular family combined features of several ADRFs. When we compare our study family with the 5 FCU families described by Hoffman et al (6), the periorbital edema, joint swelling, early disease onset, response to colchicine, and apparent sex difference in the occurrence of amyloidosis are all specific to this Indian family. However, it is perhaps too simplistic to generalize from a single family, particularly since the phenotype is markedly variable even within that family. Molecular genetics and linkage analysis mapped the susceptibility gene to the same chromosomal region as the MWS/FCU locus, and therefore, the most likely diagnosis is an overlapping form of MWS/FCU. Early diagnosis is important because of high mortality from amyloidosis, as shown by the death from AA amyloidosis of 3 family members aged 26 35 years, coupled with the presence of AA amyloidosis in 1 surviving family member. One member of the FCU family described by Jung et al also died of renal amyloidosis (7). The key clinical features reported in MWS are urticaria, limb pains, progressive deafness, and amyloidosis; those of FCU are urticaria after exposure to cold, transient arthritis, and amyloidosis. Both are inherited dominantly, have a relapsing and remitting nature, and may be associated with fever. However, the features of each are reported to vary quite substantially between families, especially the deafness of MWS and susceptibility to amyloidosis. Phenotypic expression of FCU can presumably be influenced markedly by climatic conditions. In this respect, it is notable that the present family resided in a warm part of India (perhaps accounting for the very mild symptoms in many of the affected members), and that families with FCU have been reported from Canada and Scotland, which have cold winters. Due to the relatively mild, nonspecific nature of the symptoms, the present condition could well be underdiagnosed in the Indian community; a previous report of 43 north Indian patients with apparently acquired, cold-induced urticaria, representing 10 20% of urticaria patients attending a dermatology unit (26), raises the intriguing possibility that the genetic abnormality in the present family might be more prevalent but generally less penetrant. Periorbital edema is not a generally recognized feature of these conditions, but was present in several

GENETIC MARKERS FOR AUTOSOMAL DOMINANT PERIODIC FEVER 2039 members of the study family, while progressive nerve deafness, a distinctive but inconsistent feature of MWS, did not occur. Variable disease expression has already been observed in MWS families linked to this region (8); in particular, a 70-year-old woman who was a carrier for the disease was described as having arthralgia without deafness or amyloidosis, while several of her children, who were most likely to have the identical mutation, presented with the complete clinical phenotype. Although the underlying genetic defect in MWS/ FCU has not yet been identified, and no specific laboratory tests are available to confirm the diagnosis, it is now possible to look for specific haplotype segregation in sufficiently large kindreds, as we have done in this family. With identification of the specific gene(s) involved, it will be possible to screen smaller families and, possibly, sporadic cases. At present, there is no specific treatment for these disorders, but symptoms usually respond to corticosteroids and other immunosuppressive drugs. A therapeutic trial of colchicine is worthwhile, however, since it has been reported to be quite effective in a proportion of cases, and was certainly so in the proband of the present family. This study raises the question of whether the genetic defect in MWS/FCU might involve different mutations of the same or closely related genes producing the overlapping symptoms. There is a precedent for this even among the narrow confines of the periodic fever syndromes, since HIDS and mevalonic aciduria are both associated with different mutations of the mevalonate kinase gene, but the resulting phenotypes are markedly different; mevalonic aciduria is a much more severe multisystem disease (12). AA amyloidosis is the most significant complication of inherited periodic fever syndromes, since it is unpredictable, progressive, and potentially fatal. In general, AA amyloidosis occurs in only 1 2% of patients with chronic inflammatory disorders. The development of AA amyloidosis in 4 members of this family is therefore of interest, and supports the possibility that susceptibility to this complication may also have a genetic influence. A genetic influence is also supported by the differing incidence of AA amyloidosis generally among different ethnic groups, as well as by the high frequency of amyloidosis in some families with MWS and its absence in many others. AA amyloid fibrils are derived from a 76 amino acid N-terminal cleavage fragment of the 104-residue SAA protein, and sustained acute-phase production of SAA is the only absolute prerequisite for AA amyloid deposition. The magnitude and duration of the acutephase response are probably substantial determinants of the risk of developing amyloidosis, although different isoforms of SAA 1 (the predominant isoform deposited as AA amyloid fibrils) and homozygosity for the SAA 1 isoform have been shown to be susceptibility factors for developing AA amyloidosis (27). The prognosis is chiefly determined by the extent of amyloid at diagnosis and the effectiveness with which production of SAA can be suppressed, by colchicine therapy or otherwise. Proteinuria, the most common presenting feature of AA amyloidosis, should be determined routinely in patients with inherited periodic fever syndromes, and the suspicion of amyloid should be followed up by renal or rectal biopsy and Congo red staining. SAP component scintigraphy is a sensitive and specific, noninvasive method for imaging amyloid in visceral organs and can be used serially to monitor the deposits in a quantitative manner. Scintigraphic followup studies have systematically confirmed that the amyloid deposits frequently regress when production of SAA is reduced to normal healthy baseline levels. Routine monitoring of SAA production should be an integral part of the management of all patients with AA amyloid, and automated immunoassay systems for SAA are available standardized on a World Health Organization International Reference Standard (28). Clearly, the next step in the elucidation of this disorder is cloning the gene(s) in question. The most recent report places the critical interval containing the MWS locus at 14 cm between the D1S2811 and D1S2682 markers. Physical mapping is necessary at this stage because this is a poorly mapped interval region without any well-characterized candidate genes. In this single Indian family, due to the lack of informative recombinants in the critical interval, linkage analysis alone will not provide the necessary precision for gene localization. The positional candidate approach may be more useful, especially since gene maps for this region have been improved as part of the Human Genome Project. The eventual discovery of the underlying gene(s) for these disorders will provide physicians with an accurate diagnostic tool and may suggest new therapeutic approaches to this particular disease, and possibly to inflammatory diseases in general. ACKNOWLEDGMENTS We are grateful to the family members for agreeing to participate in the study, and to Drs. Chris Amas and Hal M. Hoffman for helpful discussions.

2040 MCDERMOTT ET AL REFERENCES 1. McDermott MF. Autosomal dominant recurrent fevers clinical and genetic aspects. Rev Rhum Engl Ed 1999;10:484 91. 2. McDermott MF, Aksentijevich I, Galon J, McDermott E, Ogunkolade BW, Centola M, et al. Germline mutations in the extracellular domains of the 55kDa TNF receptor (TNF-R1) define a family of dominantly inherited autoinflammatory syndromes. Cell 1999;97:133 44. 3. Williamson LM, Hull D, Mehta R, Reeves WG, Robinson BHB, Toghill PJ. Familial hibernian fever. QJM 1982;204:469 80. 4. Muckle TJ, Wells M. Urticaria, deafness and amyloidosis: a new heredo-familial syndrome. QJM 1962;31:235 48. 5. Zip CM, Ross JB, Greaves MW, Scriver CR, Mitchell JJ, Zoar S. Familial cold urticaria. Clin Exp Dermatol 1993;18:338 41. 6. Hoffman HM, Wright FA, Broide DH, Wanderer AA, Kolodner RD. Identification of a locus on chromosome 1q44 for familial cold urticaria. Am J Hum Genet 2000;66:1693 8. 7. Jung M, Ross B, Wienker TF, Preuss G, Kuester W, Reis A. A locus for familial cold urticaria maps to distal chromosome 1q: familial cold urticaria and Muckle Wells syndrome are probably allelic [abstract]. Am J Hum Genet 1996;59 Suppl:A223. 8. Cuisset L, Drenth JPH, Berthelot J-M, Meyrier A, Vaudour G, Watts RA, et al. Genetic linkage of the Muckle-Wells syndrome to chromosome 1q44. Am J Hum Genet 1999;65:1054 9. 9. International FMF Consortium. Ancient missense mutations in a new member of the RoRet gene family are likely to cause familial Mediterranean fever. Cell 1997;90:797 807. 10. The French FMF Consortium. A candidate gene for familial Mediterranean fever. Nat Genet 1997;17:25 31. 11. Drenth JPH, Cuisset L, Grateau G, Vasseur C, van de Velde- Visser SD, de Jong JGN, et al. Mutations in the gene encoding mevalonate kinase cause hyper-igd and periodic fever syndrome. Nat Genet 1999;22:178 81. 12. Houten SM, Kuis W, Duran M, de Koning TJ, van Royen-Kerkhof A, Romeijn GJ, et al. Mutations in MVK, encoding mevalonate kinase, cause hyperimmunoglobulinemia D and periodic fever syndrome. Nat Genet 1999;22:175 7. 13. Aksentijevich I, Torosyan Y, Samuels J, Centola M, Pras E, Chae JJ, et al. Mutation and haplotype studies of familial Mediterranean fever reveal new ancestral relationships and evidence for a high carrier frequency with reduced penetrance in the Ashkenazi Jewish population. Am J Hum Genet 1999;64:949 62. 14. Talmud P, Tybjaerf-Hansen A, Bhatnagar D, Mbewu A, Miller JP, Durrington P, et al. Rapid screening for specific mutations with a clinical diagnosis of familial hypercholesterolaemia. Atherosclerosis 1991;89:137 41. 15. Don RH, Cox PT, Wainwright BJ, Baker K, Mattick JS. Touchdown PCR to circumvent spurious priming during gene amplification. Nucleic Acids Res 1991;19:4008. 16. Eskdale J, Turestkaya RL, Armstrong C, Kuprash DV, Nedospasov SA, Gallagher G. A polymorphic microsatellite marker in the human p55 receptor, CD120a. Gene Immun 2000;1:228 30. 17. Booth DR, Gillmore JD, Lachmann HJ, Booth SE, Bybee A, Soyturk M, et al. The genetic basis of autosomal dominant familial Mediterranean fever. QJM 2000;93:217 22. 18. Ott J. Computer simulation linkage methods in human linkage analysis. Proc Natl Acad Sci U S A 1989;86:4175 8. 19. Weeks DE, Ott J, Lathrop GM. SLINK: a general simulation program for linkage analysis [Abstract]. Am J Hum Genet 1990;47 Suppl 3:A204. 20. Lathrop GM, Lalouel JM, Julier C, Ott J. Multilocus linkage analysis in humans: detection of linkage and estimation of recombination. Am J Hum Genet 1985;37:482 98. 21. Cottingham RW Jr, Idury RM, Schäffer AA. Faster sequential genetic linkage computations. Am J Hum Genet 1993;53:252 63. 22. O Connell JR, Weeks DE. The VITESSE algorithm for rapid exact multilocus linkage analysis via genotype set-recoding and fuzzy inheritance. Nat Genet 1995;11:402 8. 23. Hawkins PN. Diagnosis and monitoring of amyloidosis. Baillieres Clin Rheumatol 1994;8:635 59. 24. Hawkins PN, Lavender JP, Pepys MB. Evaluation of systemic amyloidosis by scintigraphy with 123 I-labeled serum amyloid P component. N Engl J Med 1990;323:508 13. 25. Wilkins J, Gallimore JR, Tennent GA, Hawkins PN, Limburg PC, van Rijswijk MH, et al. Rapid automated enzyme immunoassay of serum amyloid A. Clin Chem 1994;40:1284 90. 26. Pasricha JS, Nayyar KC. Idiopathic acquired cold urticaria in north India. Ann Allergy 1975;34:363 5. 27. Booth DR, Booth SE, Gillmore JD, Hawkins PN, Pepys MB. SAA 1 alleles as risk factors in reactive systemic AA amyloidosis. Int J Exp Clin Invest 1998;5:262 5. 28. Poole S, Walker D, Gaines Das RE, Gallimore JR, Pepys MB. The first international standard for serum amyloid A protein (SAA): evaluation in an international collaborative study. J Immunol Methods 1998;214:1 10.