Longitudinal follow-up of late-onset Alzheimer disease families.

Historically, data for genetic studies are collected at one time point. However, for diseases with late onset or with complex phenotypes, such as Alzheimer disease (AD), restricting diagnosis to a single ascertainment contact may not be sufficient. Affection status may change over time and some initial diagnoses may be inconclusive. Follow-up provides the opportunity to resolve these complications. However, to date, previous studies have not formally demonstrated that longitudinally re-contacting families is practical or productive. To update data initially collected for linkage analysis of late-onset Alzheimer disease (LOAD), we successfully re-contacted 63 of 81 (78%) multiplex families (two to 17 years after ascertainment). Clinical status changed for 73 of the 230 (32%) non-affected participants. Additionally, expanded family history identified 20 additional affected individuals to supplement the data set. Furthermore, fostering ongoing relationships with participating families helped recruit 101 affected participants into an autopsy and tissue donation program. Despite similar presentations, discordance between clinical diagnosis and neuropathologic diagnosis was observed in 28% of those with tissue diagnoses. Most of the families were successfully re-contacted, and significant refinement and supplementation of the data was achieved. We concluded that serial contact with longitudinal evaluation of families has significant implications for genetic analyses.

Covariate analysis of late-onset Alzheimer disease refines the chromosome 12 locus.

Alzheimer disease (AD) is a progressive neurodegenerative disorder of later life with a complex etiology and a strong genetic component. Several genomic screens have suggested that a region between chromosome 12p13 and 12q22 contains at least one additional locus underlying the susceptibility of AD. However, localization of this locus has been difficult. We performed a 5 cM microsatellite marker screen across 74 cM on chromosome 12 with 15 markers in 585 multiplex families consisting of 994 affected sibpairs and 213 other affected relative pairs. Analyses across the entire data set did not reveal significant evidence of linkage. However, suggestive linkage was observed in several subsets. In the 91 families where no affected individuals carry an ApoE varepsilon4 allele, an HLOD score of 1.55 was generated at D12S1042. We further examined the linkage data considering the proposed linkages to chromosome 9 (D9S741) and chromosome 10 (alpha-catenin gene). There was a modest (P=0.20) increase in the LOD score for D12S368 (MLOD=1.70) when using the D9S741 LOD scores as a covariate and a highly significant (P<0.001) increase in the MLOD score (4.19) for D12S1701 in autopsy-confirmed families (n=228) when using alpha-catenin LOD scores as a covariate. In both cases, families with no evidence of linkage to D9S741 or alpha-catenin demonstrated most of the evidence of linkage to chromosome 12, suggesting locus heterogeneity. Taken together, our data suggest that the 16 cM region between D12S1042 and D12S368 should be the subject of further detailed genomic efforts for the disease.

Maternal lineages and Alzheimer disease risk in the Old Order Amish.

Old Order Amish, founded by a small number of Swiss immigrants, exist in culturally isolated communities across rural North America. The consequences of genetic isolation and inbreeding within this group are evident by increased frequencies of many monogenic diseases and several complex disorders. Conversely, the prevalence of Alzheimer disease (AD), the most common form of dementia, is lower in the Amish than in the general American population. Since mitochondrial dysfunction has been proposed as an underlying cause of AD and a specific haplogroup was found to affect AD susceptibility in Caucasians, we investigated whether inherited mitochondrial haplogroups affect risk of developing AD dementia in Ohio and Indiana Amish communities. Ninety-five independent matrilines were observed across six large pedigrees and three small pedigrees then classified into seven major European haplogroups. Haplogroup T is the most frequent haplogroup represented overall in these maternal lines (35.4%) while observed in only 10.6% in outbred American and European populations. Furthermore, haplogroups J and K are less frequent (1.0%) than in the outbred data set (9.4-11.2%). Affected case matrilines and unaffected control lines were chosen from pedigrees to test whether specific haplogroups and their defining SNPs confer risk of AD. We did not observe frequency differences between AD cases compared to controls overall or when stratified by sex. Therefore, we suggest that the genetic effect responsible for AD dementia in the affected Amish pedigrees is unlikely to be of mitochondrial origin and may be caused by nuclear genetic factors.

An autosomal genomic screen for dementia in an extended Amish family.

Apolipoprotein E (APOE) is the only universally confirmed susceptibility gene for late-onset Alzheimer disease (LOAD), although many loci are believed to modulate LOAD risk. The genetic homogeneity of isolated populations, such as the Amish, potentially provide increased power to identify LOAD susceptibility genes. Population homogeneity in these special populations may reduce the total number of susceptibility genes contributing to the complex disorder, thereby increasing the ability to identify any one susceptibility gene. Dementia in the Amish is clinically indistinguishable from LOAD in the general population. Previous studies in the Amish demonstrated a significantly decreased frequency of the APOE-4 susceptibility allele, but significant familial clustering of dementia [M.A. Pericak-Vance, C.C. Johnson, J.B. Rimmler, A.M. Saunders, L.C. Robinson, E.G. D'Hondt, C.E. Jackson, J.L. Haines, Alzheimer's disease and apolipoprotein E-4 allele in an Amish population, Ann. Neurol. 39 (1996) 700-704]. These data suggested that a genetic etiology independent of APOE may underlie the dementia observed in this population. In the present analysis, we focused on a large, multiplex, inbred Amish family (24 sampled individuals; 10 of whom are affected). We completed a genomic screen to identify novel LOAD loci (n=316 genetic markers), using both model-dependent "affecteds-only" analysis (dominant and recessive) and model-independent affected relative pair analysis. Interesting results (lod>1.5 or p<0.01) were obtained for markers on eight chromosomes (2q, 5q, 6q, 7p, 8p, 8q, 11p, 18p, 18q, and 19q). The highest overall score was a multipoint lod score of 3.1 on chromosome 11p. Most regions we identified were not previously detected by genomic screens of outbred populations and may represent population-specific susceptibilities to LOAD. These loci are currently under further investigation in a study of LOAD including additional Amish families.

Fine mapping and genetic heterogeneity in the pure form of autosomal dominant familial spastic paraplegia.

We evaluated seven families segregating pure, autosomal dominant familial spastic paraplegia (SPG) for linkage to four recently identified SPG loci on chromosomes 2q (1), 8q (2), 12q (3), and 19q (4). These families were previously shown to be unlinked to SPG loci on chromosomes 2p, 14q, and 15q. Two families demonstrated linkage to the new loci. One family (family 3) showed significant evidence for linkage to chromosome 12q, peaking at D12S1691 (maximum lod = 3.22). Haplotype analysis of family 3 did not identify any recombinants among affected individuals in the 12q candidate region. Family 5 yielded a peak lod score of 2.02 at marker D19S868 and excluded linkage to other known SPG loci. Haplotype analysis of family 5 revealed several cross-overs in affected individuals, thereby potentially narrowing the SPG12 candidate region to a 5-cM region between markers D19S868 and D19S220. Three of the families definitively excluded all four loci examined, providing evidence for further genetic heterogeneity of pure, autosomal dominant SPG. In conclusion, these data confirm the presence of SPG10 (chromosome 12), potentially reduce the minimum candidate region for SPG12 (chromosome 19q), and suggest there is at least one additional autosomal dominant SPG locus.

Identification and expression analysis of spastin gene mutations in hereditary spastic paraplegia.

Pure hereditary spastic paraplegia (SPG) type 4 is the most common form of autosomal dominant hereditary SPG, a neurodegenerative disease characterized primarily by hyperreflexia and progressive spasticity of the lower limbs. It is caused by mutations in the gene encoding spastin, a member of the AAA family of ATPases. We have screened the spastin gene for mutations in 15 families consistent with linkage to the spastin gene locus, SPG4, and have identified 11 mutations, 10 of which are novel. Five of the mutations identified are in noninvariant splice-junction sequences. Reverse transcription-PCR analysis of mRNA from patients shows that each of these five mutations results in aberrant splicing. One mutation was found to be "leaky," or partially penetrant; that is, the mutant allele produced both mutant (skipped exon) and wild-type (full-length) transcripts. This phenomenon was reproduced in in vitro splicing experiments, with a minigene splicing-vector construct only in the context of the endogenous splice junctions flanking the splice junctions of the skipped exon. In the absence of endogenous splice junctions, only mutant transcript was detected. The existence of at least one leaky mutation suggests that relatively small differences in the level of wild-type spastin expression can have significant functional consequences. This may account, at least in part, for the wide ranges in age at onset, symptom severity, and rate of symptom progression that have been reported to occur both among and within families with SPG linked to SPG4. In addition, these results suggest caution in the interpretation of data solely obtained with minigene constructs to study the effects of sequence variation on splicing. The lack of full genomic sequence context in these constructs can mask important functional consequences of the mutation.

Myotilin is mutated in limb girdle muscular dystrophy 1A.

We have identified a mutation in the myotilin gene in a large North American family of German descent expressing an autosomal dominant form of limb girdle muscular dystrophy (LGMD1A). We have previously mapped this gene to 5q31. Symptoms of this adult onset disease are progressive weakness of the hip and shoulder girdles, as well as a distinctive dysarthric pattern of speech. Muscle of affected individuals shows degeneration of myofibers, variations in fiber size, fiber splitting, centrally located myonuclei and a large number of autophagic vesicles. Affected muscle also exhibits disorganization and streaming of the Z-line similar to that seen in nemaline myopathy. We have identified a C450T missense mutation in the myotilin gene that is predicted to result in the conversion of residue 57 from threonine to isoleucine. This mutation has not been found in 396 control chromosomes. The mutant allele is transcribed and normal levels of correctly localized myotilin protein are seen in LGMD1A muscle. Myotilin is a sarcomeric protein that binds to alpha-actinin and is localized in the Z-line. The observed missense mutation does not disrupt binding to alpha-actinin.

Lewy body and Alzheimer pathology in a family with the amyloid-beta precursor protein APP717 gene mutation.

Mutations in the amyloid precursor protein (APP) gene cause one form of early onset familial Alzheimer's disease (AD). One such family has been studied genetically and neuropathologically and represents the basis of the present report. Four siblings with the APP717 Val to Ile mutation, aged 59, 65, 61 and 64 years, apolipoprotein E (APOE) genotyped 2,4 (first three) and 2,3 respectively, had severe AD, Braak stage VI with frequent neurofibrillary tangles in the primary visual cortex, Brodmann area 17. The first one also met McKeith criteria for the limbic stage of dementia with Lewy bodies but did not have substantia nigra Lewy bodies. The second two met McKeith criteria for the neocortical stage of dementia with Lewy bodies and both had substantia nigra Lewy bodies. The fourth had AD but no Lewy bodies. A cousin without the APP717 mutation who was APOE 3, 4, developed dementia at age 60 and died at age 75. She had severe cerebrovascular atherosclerosis, less severe AD, Braak stage V, with sparing of area 17. She also had Lewy bodies in the substantia nigra and in the cortex and met McKeith criteria for neocortical stage of dementia with Lewy bodies. Extrapyramidal features were present in all five. Lewy bodies have been described in 53% of reported autopsies on individuals with the APP717 Val to Ile mutation coincident with dementia and AD neuropathologic changes. These observations suggest an association between the chromosome 21 APP mutation and Lewy body formation, possibly mediated by other environmental or genetic factors.

Neuropathological features of frontotemporal dementia and parkinsonism linked to chromosome 17q21-22 (FTDP-17): Duke Family 1684.

Frontotemporal dementia with parkinsonism (FTDP-17) is an autosomal dominant disorder that presents clinically with dementia, extrapyramidal signs, and behavioral disturbances in mid-life and progresses to death within 5 to 10 years. Pathologically, the disorder is characterized by variable neuronal loss and gliosis in the frontal and temporal lobes, limbic structures, and the midbrain. Autopsied individuals from some kindreds display abundant neurofibrillary change while others, including a single affected individual from Duke Family 1684, lack distinctive histological features and exhibit only mild neuronal loss and gliosis in limbic structures and subcortical nuclei when examined by routine silver stain. Recently, mutations in the microtubule associated protein tau have been shown to segregate with the disease in this family and in many other affected kindreds. In order to examine the distribution of tau deposits, we performed tau immunohistochemistry, immunoblotting, and immunoelectron microscopy of tau-containing filaments. Immunohistochemistry revealed numerous tau deposits within glial cells and within neurons. Twisted ribbon-like filaments observed by immunoelectron microscopy were immunodecorated with tau AT8 antibody. Sarkosyl-insoluble tau extracted from the hippocampus and cortex migrated as 2 major bands at 64 and 68 kilodaltons and a minor band at 72 kilodaltons, which after alkaline phosphatase treatment appeared to contain mainly tau isoforms with 4 repeats. Furthermore, the ratio of soluble tau with 4 to 3 microtubule-binding repeats was increased. The role of tau mutations in this disorder is discussed in this paper.

No association of alpha1-antichymotrypsin flanking region polymorphism and Alzheimer disease risk in early- and late-onset Alzheimer disease patients.

The alpha1-antichymotrypsin (AACT)-155 allele was found elsewhere to have a significant effect on Alzheimer disease (AD) risk in individuals with at least one APOE-4 allele. We compared AACT genotypes of 284 cases of sporadic AD and 172 controls. The frequency of the AACT-155 allele did not differ significantly between cases and controls, either overall or when restricted to subjects with at least one APOE-4 allele. Logistic regression controlling for age and sex failed to show an effect due to AACT either alone or acting with APOE. There was no evidence of an interaction between APOE-4 and the AACT-155 allele to reduce age at onset. Thus, our data do not support an association of AACT-155 with risk or age at onset in AD.

Evidence for anticipation in autosomal dominant limb-girdle muscular dystrophy.

Anticipation, an increase in severity or decrease in age of onset (AO) inherent in the transmission of the disease gene from affected parent to affected child, has been increasingly described in human disease. To assess anticipation in a large kindred in which autosomal dominant limb-girdle muscular dystrophy (LGMD1A) is segregating, age of disease onset was collected from patient interviews of affected family members. A total of 25 parent-offspring pairs, in which the parents are three (3R), four (4R), or five (5R) generations removed from a common founding ancestor, were available for analysis. Life table analyses showed significant decreases in age at first reported symptoms in the offspring of the 3R (chi2=5.55, p=0.02) and 4R (chi2=7.81, p=0.005) parents. Pairwise analyses confirmed this decrease with a median decrease of 13 years in transmission to offspring from 3R parents and 18 years in transmission to offspring from 4R parents. The finding of anticipation in this pedigree suggests that the mutation in LGMD1A may be the result of the expansion of an unstable trinucleotide repeat.

Complete genomic screen in late-onset familial Alzheimer's disease.

Alzheimer's disease (AD) is a complex genetic disorder. Linkage analysis has helped unravel a portion of the genetic component of AD by identifying four loci that play a role in the genetics of AD (amyloid precursor protein, presenilin 1, presenilin 2, and apolipoprotein E). These loci account for approximately 50% of the genetic etiology of AD. A total genomic screen is an efficient way to identify additional genetic effects in AD. A series of multiplex late-onset (>60 years) AD families were ascertained (NINDS-ADRDA diagnostic criteria) and sampled. A subset (n = 16) of the largest families (52 affecteds with DNA, 83 unaffecteds with DNA) were used to rapidly screen the genome (n = 280 markers) for additional major genetic effects. Critical values for regional follow-up were p < or =0.05 for SimIBD or sibpair analysis and/or a LOD score > or = 1.00. Fifteen regions warranted initial follow-up based on these criteria. An additional screening set was used (n = 38 families, 89 affecteds with DNA, 216 unaffecteds with DNA) for the follow-up analysis. These analyses revealed four regions of continued interest on chromosomes 4, 6, 12, and 20. Chromosome 12 presented the strongest results. Peak two point "affecteds only" LOD scores were 1.3, 1.6, 2.7, and 2.2 and (affected relative pair SimIBD) p values were 0.04, 0.03, 0.14, and 0.04 for D12S373, D12S1057, D12S1042, and D12S390, respectively. These markers span approximately 30 cm near the centromeric region of chromosome 12. Sibpair analysis resulted in two point Maximum Lod Score (MLS) results of 0.4, 1.2, 3.2, and 1.0 for the above markers. Multipoint MLS analysis supported these findings. Saturation mapping of all available markers in the chromosome 12 region as well as further investigation of the regions on 4, 6, and 20 is ongoing with candidate gene analysis to follow.

No genetic association between the LRP receptor and sporadic or late-onset familial Alzheimer disease.

The low-density lipoprotein receptor-related protein gene (LRP1) is often mentioned as a candidate gene for Alzheimer disease (AD) because of its role as a receptor for apolipoprotein E (apoE), a major genetic risk factor for late-onset familial and sporadic AD. A recent association study of a tetranucleotide repeat polymorphism located 5' to the LRP1 gene detected an increase in the 87 base pair allele in AD cases compared to unaffected controls. Additionally, an independent study involving a genomic screen for genes associated with late-onset AD identified a region as a possible location of a late-onset AD gene on chromosome 12p between D12S373 and D12S390, about 10 cM proximal to LRP1. We examined 144 late-onset multiplex AD families, 436 sporadic AD cases, and 240 controls and found no evidence of linkage or association of LRP1 and AD. Our data indicate that genetic variation of the LRP1 gene is not a major risk factor in the etiology of AD.

Confirmation of linkage of oculopharyngeal muscular dystrophy to chromosome 14q11.2-q13 in American families suggests the existence of a second causal mutation.

Oculopharyngeal muscular dystrophy (OPMD) is a late-onset, autosomal dominant disorder characterized by progressive ptosis, dysphagia, and extremity weakness. Linkage of OPMD to 14q11.2-q13 has been reported in a series of French-Canadian families. Tightly linked markers have been defined and haplotype analysis in these data show a single segregating disease chromosome throughout the OPMD French-Canadian families. We have ascertained and sampled five multigenerational outbred American OPMD families. Four of the five families have known French-Canadian ancestry while the fifth is of English/Scottish origin. Linkage analysis was performed using standard likelihood methods. A peak multipoint lod score of 6.30 was obtained for the marker MYH7.1 in the OPMD families. The English/ Scottish family exhibited a different chromosomal haplotype for the OPMD alleles than the families of French-Canadian origin. These data suggest this family may represent a second, possibly independent mutation in this disorder. Linkage was confirmed to chromosome 14q11.2-q13 with no evidence of genetic heterogeneity.

Apolipoprotein E epsilon2 does not increase risk of early-onset sporadic Alzheimer's disease.

We examined the association of apolipoprotein E (ApoE) genotype and the risk of early-onset Alzheimer's disease (AD) in 209 white early-onset sporadic cases (43% male) and 303 white controls (48% male) of similar age distribution. The risk of AD was significantly increased, relative to the 3/3 genotype, in people with the 4/4, 3/4, and 2/4 genotypes, controlling for age at time of examination and sex. The 2/3 genotype reduced slightly the risk of AD, although the effect was not statistically significant. We conclude, contrary to some previous reports, that the ApoE epsilon2 allele does not increase the risk of early-onset sporadic AD.

Complete genomic screen in late-onset familial Alzheimer disease. Evidence for a new locus on chromosome 12.

CONTEXT: Four genetic loci have been identified as contributing to Alzheimer disease (AD), including the amyloid precursor protein gene, the presenilin 1 gene, the presenilin 2 gene, and the apolipoprotein E gene, but do not account for all the genetic risk for AD. OBJECTIVE: To identify additional genetic risk factors for late-onset AD. DESIGN: A complete genomic screen was performed (N=280 markers). Critical values for chromosomal regional follow-up were a P value of .05 or less for affected relative pair analysis or sibpair analysis, a parametric lod score of 1.0 or greater, or both. Regional follow-up included analysis of additional markers and a second data set. SETTING: Clinic populations in the continental United States. PATIENTS: From a series of multiplex families affected with late-onset (> or =60 years) AD ascertained during the last 14 years (National Insititute of Neurological Disorders and Stroke-Alzheimer's Disease and Related Disorders Association diagnostic criteria) and for which DNA has been obtained, a subset of 16 families (135 total family members, 52 of whom were patients with AD) was used for the genomic screen. A second subset of 38 families (216 total family members, 89 of whom were patients with AD) was used for the follow-up analysis. MAIN OUTCOME MEASURES: Linkage analysis results generated using both genetic model-dependent (lod score) and model-independent methods. RESULTS: Fifteen chromosomal regions warranted initial follow-up. Follow-up analyses revealed 4 regions of continued interest on chromosomes 4, 6, 12, and 20, with the strongest results observed forchromosome 12. Peak 2-point affecteds-only lod scores (n=54) were 1.3, 1.6, 2.7, and 2.2 and affected relative pairs P values (n=54) were .04, .03, .14, and .04 for D12S373, D12S1057, D12S1042, and D12S390, respectively. Sibpair analysis (n=54) resulted in maximum lod scores (MLSs) of 1.5, 2.6, 3.2, and 2.3 for these markers, with a peak multipoint MLS of 3.5. A priori stratification by APOE genotype identified 27 families that had at least 1 member with AD whose genotype did not contain an APOE*4 allele. Analysis of these 27 families resulted in MLSs of 1.0, 2.4, 3.7, and 3.3 and a peak multipoint MLS of 3.9. CONCLUSIONS: A complete genomic screen in families affected with late-onset AD identified 4 regions of interest after follow-up. Chromosome 12 gave the strongest and most consistent results with a peak multipoint MLS of 3.5, suggesting that this region contains a new susceptibility gene for AD. Additional analyses are necessary to identify the chromosome 12 susceptibility gene for AD and to follow up the regions of interest on chromosomes 4, 6, and 20.

No association or linkage between an intronic polymorphism of presenilin-1 and sporadic or late-onset familial Alzheimer disease.

Recent reports have shown an association between an intronic polymorphism of the presenilin-1 (PSEN1) gene and late-onset (age at onset > 65) familial and sporadic (no family history) Alzheimer disease (AD). The reported association was independent of the effect of the only previously identified gene associated with late-onset AD, APOE. Blood samples were obtained from members of 122 multiplex AD families, 42 unrelated cases of AD with positive family histories of dementia, 456 sporadic cases of AD, and 317 controls of similar ages at examination to the cases. These samples were genotyped for an intronic polymorphism of the PSEN1 gene, located 3' to exon 8, and the data analyzed for evidence of association or linkage. The samples were also genotyped for APOE and the data analyzed to see if the association or linkage changed when controlling for APOE genotype. There was no statistically significant increase (at alpha = .01) in allele 1 (199 bp) or genotype 1/1 in the sporadic AD cases, or in a random sample of one affected from each multiplex family, compared to controls. When examining the effect of the PSEN1 polymorphism while controlling for APOE genotype, APOE genotype was strongly associated with AD, but the PSEN1 polymorphism genotype was not. Model-trait dependent (lod score) and independent (Sim1BD) methods detected no evidence of linkage between PSEN1 and AD. In this independent dataset, the previously reported association between the intronic PSEN1 polymorphism and AD cannot be confirmed, and the conclusion that PSEN1 is a major susceptibility gene for late-onset AD is not supported.

Locus heterogeneity, anticipation and reduction of the chromosome 2p minimal candidate region in autosomal dominant familial spastic paraplegia.

We examined 11 Caucasian pedigrees with autosomal dominant 'uncomplicated' familial spastic paraplegia (SPG) for linkage to the previously identified loci on chromosomes 2p, 14q and 15q. Chromosome 15q was excluded for all families. Five families showed evidence for linkage to chromosome 2p, one to chromosome 14q, and five families remained indeterminate. Homogeneity analysis of combined chromosome 2p and 14q data gave no evidence for a fourth as yet unidentified SPG locus. Recombination events reduced the chromosome 2p minimum candidate region (MCR) to a 3 cM interval between D2S352 and D2S367 and supported the previously reported 7 cM MCR for chromosome 14q. Age of onset (AO) was highly variable, indicating that subtypes of SPG are more appropriately defined on a genetic basis than by AO. Comparison of AO in parent-child pairs was suggestive of anticipation, with a median difference of 9.0 years (p<0.0001).

Linkage of frontotemporal dementia to chromosome 17: clinical and neuropathological characterization of phenotype.

Frontotemporal dementia is a behavioral disorder of insidious onset and variable progression. Clinically, its early features reflect frontal lobe dysfunction characterized by personality change, deterioration in memory and executive functions, and stereotypical and perseverative behaviors. Pathologically, there is degeneration of the neocortex and subcortical nuclei, without distinctive features such as plaques, neurofibrillary tangles, or Pick or Lewy bodies. Within-family variation in neuropathology and clinical phenotype is observed. In cases where family aggregation is observed, it is inherited as an autosomal dominant, age-dependent disorder. Family studies recently have identified two dementia loci: chromosome 17 for disinhibition-dementia-parkinsonism-amyotrophic complex and pallido-ponto-nigral degeneration and chromosome 3 for familial nonspecific dementia. We describe a family (DUK1684) with clinically and neuropathologically confirmed, autosomal dominant, non-Alzheimer disease dementia. Linkage analysis of this family showed evidence for linkage to chromosome 17q21, with a multipoint location score (log10) of 5.52. A comparison of the clinical and pathological features in DUK1684 with those of the other chromosome 17-linked families, together with the linkage data, suggests that these families are allelic. These studies emphasize that genetic linkage analysis remains a useful tool for differentiating disease loci in clinically complex traits.