A pan-European study of the C9orf72 repeat associated with FTLD: geographic prevalence, genomic instability, and intermediate repeats.

We assessed the geographical distribution of C9orf72 G(4) C(2) expansions in a pan-European frontotemporal lobar degeneration (FTLD) cohort (n = 1,205), ascertained by the European Early-Onset Dementia (EOD) consortium. Next, we performed a meta-analysis of our data and that of other European studies, together 2,668 patients from 15 Western European countries. The frequency of the C9orf72 expansions in Western Europe was 9.98% in overall FTLD, with 18.52% in familial, and 6.26% in sporadic FTLD patients. Outliers were Finland and Sweden with overall frequencies of respectively 29.33% and 20.73%, but also Spain with 25.49%. In contrast, prevalence in Germany was limited to 4.82%. In addition, we studied the role of intermediate repeats (7-24 repeat units), which are strongly correlated with the risk haplotype, on disease and C9orf72 expression. In vitro reporter gene expression studies demonstrated significantly decreased transcriptional activity of C9orf72 with increasing number of normal repeat units, indicating that intermediate repeats might act as predisposing alleles and in favor of the loss-of-function disease mechanism. Further, we observed a significantly increased frequency of short indels in the GC-rich low complexity sequence adjacent to the G(4) C(2) repeat in C9orf72 expansion carriers (P < 0.001) with the most common indel creating one long contiguous imperfect G(4) C(2) repeat, which is likely more prone to replication slippage and pathological expansion.

Guanosine triphosphate cyclohydrolase 1 promoter deletion causes dopa-responsive dystonia.

BACKGROUND: Autosomal dominant dopa-responsive dystonia (AD-DRD) is caused by a biochemical defect primarily resulting from guanosine triphosphate cyclohydrolase 1 gene (GCH1) mutations. Few families have been reported without mutations in GCH1. METHODS: Genome-wide linkage analysis and positional cloning to identify the genetic defect in a Belgian AD-DRD family was carried out. RESULTS AND CONCLUSION: In this study, we report on the identification and characterization of a novel 24-kb deletion spanning exon 1 and the 5' regulatory region of GCH1 causing a wide spectrum of motor and nonmotor symptoms in a large Belgian AD-DRD family. This large-scale deletion of regulatory sequences leads to decreased GCH1 activity in all carriers, most probably resulting from allelic loss of transcription. We mapped the breakpoints of this deletion to the nucleotide level, allowing the development of a straightforward polymerase chain reaction assay for fast, efficient detection of this large deletion, which will prove valuable for preimplantation genetic diagnosis.

Pilot Testing of the "Turbidimeter", a Simple, Universal Reader Intended to Complement and Enhance Bacterial Growth Detection in Manual Blood Culture Systems in Low-Resource Settings.

Bloodstream infections and antimicrobial resistance are an increasing problem in low-income countries. There is a clear need for adapted diagnostic tools. To address this need, we developed a simple, universal reader prototype that detects bacterial growth in blood culture bottles. Our "turbidimeter" evaluates bacterial growth, based on the turbidity of the broth and the color change of the colorimetric CO2 indicator in commercially available blood culture bottles. A total of 60 measurements were performed using 10 relevant microbial species, spiked in horse blood, to compare the turbidimeter's performance with that of an automatic reference system. The turbidimeter was able to detect growth in all but one of the spiked blood culture bottles. In the majority (7/10) of the species tested, time-to-detection of the turbidimeter was shown to be non-inferior to the reference automated time-to-detection. This was, however, only the case when both the turbidity and color change in the colorimetric CO2-indicator were used to evaluate growth. We could not demonstrate the non-inferiority of the turbidity measurement alone. Overall, the turbidimeter performed well, but we also identified some improvements that will be implemented in the next version of the prototype.

Relative contribution of simple mutations vs. copy number variations in five Parkinson disease genes in the Belgian population.

The relative contribution of simple mutations and copy number variations (CNVs) in SNCA, PARK2, PINK1, PARK7, and LRRK2 to the genetic etiology of Parkinson disease (PD) is still unclear because most studies did not completely analyze each gene. In a large group of Belgian PD patients (N = 310) and control individuals (N = 270), we determined the mutation frequency of both simple mutations and CNVs in these five PD genes, using direct sequencing, multiplex amplicon quantification (MAQ), and real-time PCR assays. Overall, we identified 14 novel heterozygous variants, of which 11 were absent in control individuals. We observed eight PARK2 (multiple) exon multiplications in PD patients and one exon deletion in a control individual. Furthermore, we identified one SNCA whole-gene duplication. The PARK2 and LRRK2 mutation frequencies in Belgian PD patients were similar to those reported in other studies. However, at this stage the true pathogenic nature of some heterozygous mutations in recessive genes remains elusive. Furthermore, though mutations is SNCA, PINK1, and PARK7 are rare, our identification of a SNCA duplication confirmed that screening of these genes remains meaningful.

Genetic variability in the mitochondrial serine protease HTRA2 contributes to risk for Parkinson disease.

In one genetic study, the high temperature requirement A2 (HTRA2) mitochondrial protein has been associated with increased risk for sporadic Parkinson disease (PD). One missense mutation, p.Gly399Ser, in its C-terminal PDZ domain (from the initial letters of the postsynaptic density 95, PSD-95; discs large; and zonula occludens-1, ZO-1 proteins [Kennedy, 1995]) resulted in defective protease activation, and induced mitochondrial dysfunction when overexpressed in stably transfected cells. Here we examined the contribution of genetic variability in HTRA2 to PD risk in an extended series of 266 Belgian PD patients and 273 control individuals. Mutation analysis identified a novel p.Arg404Trp mutation within the PDZ domain predicted to freeze HTRA2 in an inactive form. Moreover, we identified six patient-specific variants in 5' and 3' regulatory regions that might affect HTRA2 expression as supported by data of luciferase reporter gene analyses. Our study confirms a role of the HTRA2 mitochondrial protein in PD susceptibility through mutations in its functional PDZ domain. In addition, it extends the HTRA2 mutation spectrum to functional variants possibly affecting transcriptional activity. The latter underpins a previously unrecognized role for altered HTRA2 expression as a risk factor relevant to parkinsonian neurodegeneration.

Genetic variant in the HSPB1 promoter region impairs the HSP27 stress response.

The 27 kDa heat shock protein 1 (HSP27) is a member of the ubiquitously expressed small heat shock protein family and has pleiotropic cytoprotective functions. Since HSP27 may act as a motor neuron survival factor, we analyzed the genetic contribution of the human HSPB1 gene (HSPB1) to the etiology of amyotrophic lateral sclerosis (ALS). In a cohort of sporadic ALS patients, we identified three rare genetic variations and one of which (c.-217T>C) targeted a conserved nucleotide of the Heat Shock Element (HSE) in the HSPB1 promoter. Since binding of Heat Shock Factor 1 (HSF1) to this HSE is essential for stress-induced transcription of HSPB1, we examined the effect of the c.-217C allele on transcriptional activity and HSF binding. The basal promoter activity of the HSPB1 c.-217C mutant allele decreased to 50% as compared to the wild-type promoter in neuronal and non-neuronal cells. Following heat shock, the HSE variant attenuated significantly the stress-related increase in transcription. Electrophoretic mobility shift assays demonstrated a dramatically reduced HSF-binding to the c.-217C mutant allele as compared to the c.-217T wild-type allele. In conclusion, our study underscores the importance of the c.-217T nucleotide for HSF binding and heat inducibility of HSPB1. Therefore, our study suggests that the functional HSPB1 variant may represent a genetic modifier in the pathogenesis of motor neuron disease; however, it is necessary to confirm this HSPB1 variant in additional ALS patients.

Genetic risk and transcriptional variability of amyloid precursor protein in Alzheimer's disease.

It is well established that Alzheimer's disease causing mutations in APP, PSEN1 and PSEN2 lead to a relative increased production of Abeta42, thereby fostering its deposition in plaques. Recently others and we showed that amyloid precursor protein (APP) overproduction, either as a result of genomic locus duplication or altered regulatory sequences in the APP promoter region, leads to early-onset disease. Here, we have expanded our study of genetic variability in the APP promoter to a large group of well-documented Belgian patients (n = 750, mean onset age = 75.0 +/- 8.6, range = 37-96). We identified three different APP promoter mutations (-369C-->G, -534G-->A and -479C-->T) in seven patients. In patients with onset < or =70 years (n = 204), we identified one patient carrying the London APP V717I mutation while no patients carried an APP locus duplication, indicating that APP promoter mutations (n = 2) were more frequently associated with increased risk for early-onset Alzheimer's disease. The two mutations (-369C-->G and -534G-->A) increasing APP promoter activity by nearly 2-fold and mimicking an APP duplication, appeared in probands of families with multiple patients with dementia. The -479C-->T mutation that increased APP expression only mildly (1.2-fold), was observed in four patients with onset ages ranging from 62 to 79 years (mean 71.5 years), suggesting that its contribution to disease risk is more pronounced at later age due to modulating factors. In conclusion, we provided evidence that mutations in APP regulatory sequences are more frequent than APP coding mutations, and that increased APP transcriptional activity constitutes a risk factor for Alzheimer's disease with onset ages inversely correlated with levels of APP expression.

Mean age-of-onset of familial alzheimer disease caused by presenilin mutations correlates with both increased Abeta42 and decreased Abeta40.

The varied ways in which mutations in presenilins (PSEN1 and PSEN2) affect amyloid b precursor protein (APP) processing in causing early-onset familial Alzheimer disease (FAD) are complex and not yet properly understood. Nonetheless, one useful diagnostic marker is an increased ratio of Ab42 to Ab40 (Ab42/Ab40) in patients' brain and biological fluids as well as in transgenic mice and cells. We studied Ab and APP processing for a set of nine clinical PSEN mutations on a novel and highly reproducible enzyme-linked immunosorbent assay (ELISA)-based in vitro method and also sought correlation with brain Ab analyzed by image densitometry and mass spectrometry. All mutations significantly increased Ab42/Ab40 in vitro by significantly decreasing Ab40 with accumulation of APP C-terminal fragments, a sign of decreased PSEN activity. A significant increase in absolute levels of Ab42 was observed for only half of the mutations tested. We also showed that age-of-onset of PSEN1-linked FAD correlated inversely with Ab42/Ab40 (r = -0.89; P = 0.001) and absolute levels of Ab42 (r = -0.83; P = 0.006), but directly with Ab40 levels (r = 0.69; P = 0.035). These changes also partly correlated with brain Ab42 and Ab40 levels. Together, our data suggested that Ab40 might be protective by perhaps sequestering the more toxic Ab42 and facilitating its clearance. Also, the in vitro method we describe here is a valid tool for assaying the pathogenic potential of clinical PSEN mutations in a molecular diagnostic setting.

Global investigation and meta-analysis of the C9orf72 (G4C2)n repeat in Parkinson disease.

OBJECTIVES: The objective of this study is to clarify the role of (G4C2)n expansions in the etiology of Parkinson disease (PD) in the worldwide multicenter Genetic Epidemiology of Parkinson's Disease (GEO-PD) cohort. METHODS: C9orf72 (G4C2)n repeats were assessed in a GEO-PD cohort of 7,494 patients diagnosed with PD and 5,886 neurologically healthy control individuals ascertained in Europe, Asia, North America, and Australia. RESULTS: A pathogenic (G4C2)n>60 expansion was detected in only 4 patients with PD (4/7,232; 0.055%), all with a positive family history of neurodegenerative dementia, amyotrophic lateral sclerosis, or atypical parkinsonism, while no carriers were detected with typical sporadic or familial PD. Meta-analysis revealed a small increase in risk of PD with an increasing number of (G4C2)n repeats; however, we could not detect a robust association between the C9orf72 (G4C2)n repeat and PD, and the population attributable risk was low. CONCLUSIONS: Together, these findings indicate that expansions in C9orf72 do not have a major role in the pathogenesis of PD. Testing for C9orf72 repeat expansions should only be considered in patients with PD who have overt symptoms of frontotemporal lobar degeneration/amyotrophic lateral sclerosis or apparent family history of neurodegenerative dementia or motor neuron disease.

DLB and PDD: a role for mutations in dementia and Parkinson disease genes?

Based on the substantial overlap in clinical and pathological characteristics of dementia with Lewy bodies (DLB) and Parkinson disease with dementia (PDD) with Alzheimer disease (AD) and Parkinson disease (PD) we hypothesized that these disorders might share underlying genetic factors. The contribution of both sequence and copy number variants (CNVs) in known AD and PD genes to the genetic etiology of DLB and PDD however is currently unclear. Therefore, we performed a gene-based mutation analysis of all major AD and PD genes in 99 DLB and 75 PDD patients, including familial and sporadic forms, from Flanders, Belgium. Also, copy number variants in APP, SNCA, and PARK2 were determined. In the AD genes we detected proven pathogenic missense mutations in PSEN1 and PSEN2, and 2 novel missense variants in PSEN2 and MAPT. In the PD genes we identified 1 SNCA duplication, the LRRK2 R1441C founder mutation and 4 novel heterozygous missense variants with unknown pathogenicity. Our results suggest a contribution of established AD and PD genes to the genetic etiology of DLB and PDD though to a limited extent. They do support the hypothesis of a genetic overlap between members of the Lewy body disease spectrum, but additional genes still have to exist.

Contribution of VPS35 genetic variability to LBD in the Flanders-Belgian population.

VPS35 was recently identified as a novel autosomal dominant gene for Parkinson disease. In this study, we aimed to determine the contribution of simple and complex VPS35 variations to the genetic etiology of the spectrum of Lewy body disorders (LBD) in a Flanders-Belgian patient cohort (n = 677). We identified 3 novel missense variations in addition to 1 silent and 1 intronic variation predicted to activate a cryptic splice site, but no copy number variations. Despite the absence of these rare variations in the control group (n = 800), we could not attain convincing evidence for pathogenicity by segregation analysis or in silico predictions. Hence, our data do not support a major role for VPS35 variations in the genetic etiology of Lewy body disorders in the Flanders-Belgian population.

A C9orf72 promoter repeat expansion in a Flanders-Belgian cohort with disorders of the frontotemporal lobar degeneration-amyotrophic lateral sclerosis spectrum: a gene identification study.

BACKGROUND: Amyotrophic lateral sclerosis (ALS) and frontotemporal lobar degeneration (FTLD) are extremes of a clinically, pathologically, and genetically overlapping disease spectrum. A locus on chromosome 9p21 has been associated with both disorders, and we aimed to identify the causal gene within this region. METHODS: We studied 305 patients with FTLD, 137 with ALS, and 23 with concomitant FTLD and ALS (FTLD-ALS) and 856 controls from Flanders (Belgium); patients were identified from a hospital-based cohort and were negative for mutations in known FTLD and ALS genes. We also examined the family of one patient with FTLD-ALS previously linked to 9p21 (family DR14). We analysed 130 kbp at 9p21 in association and segregation studies, genomic sequencing, repeat genotyping, and expression studies to identify the causal mutation. We compared genotype-phenotype correlations between mutation carriers and non-carriers. FINDINGS: In the patient-control cohort, the single-nucleotide polymorphism rs28140707 within the 130 kbp region of 9p21 was associated with disease (odds ratio [OR] 2·6, 95% CI 1·5-4·7; p=0·001). A GGGGCC repeat expansion in C9orf72 completely co-segregated with disease in family DR14. The association of rs28140707 with disease in the patient-control cohort was abolished when we excluded GGGGCC repeat expansion carriers. In patients with familial disease, six (86%) of seven with FTLD-ALS, seven (47%) of 15 with ALS, and 12 (16%) of 75 with FTLD had the repeat expansion. In patients without known familial disease, one (6%) of 16 with FTLD-ALS, six (5%) of 122 with ALS, and nine (4%) of 230 with FTLD had the repeat expansion. Mutation carriers primarily presented with classic ALS (10 of 11 individuals) or behavioural variant FTLD (14 of 15 individuals). Mean age at onset of FTLD was 55·3 years (SD 8·4) in 21 mutation carriers and 63·2 years (9·6) in 284 non-carriers (p=0·001); mean age at onset of ALS was 54·5 years (9·9) in 13 carriers and 60·4 years (11·4) in 124 non-carriers. Postmortem neuropathological analysis of the brains of three mutation carriers with FTLD showed a notably low TDP-43 load. In brain at postmortem, C9orf72 expression was reduced by nearly 50% in two carriers compared with nine controls (p=0·034). In familial patients, 14% of FTLD-ALS, 50% of ALS, and 62% of FTLD was not accounted for by known disease genes. INTERPRETATION: We identified a pathogenic GGGGCC repeat expansion in C9orf72 on chromosome 9p21, as recently also reported in two other studies. The GGGGCC repeat expansion is highly penetrant, explaining all of the contribution of chromosome 9p21 to FTLD and ALS in the Flanders-Belgian cohort. Decreased expression of C9orf72 in brain suggests haploinsufficiency as an underlying disease mechanism. Unidentified genes probably also contribute to the FTLD-ALS disease spectrum. FUNDING: Full funding sources listed at end of paper (see Acknowledgments).

GIGYF2 has no major role in Parkinson genetic etiology in a Belgian population.

Missense mutations were identified in the Grb10-Interacting GYF Protein-2 gene (GIGYF2), located in the chromosomal region 2q36-q37, in familial Parkinson disease (PD) patients of European descent. To determine the contribution of GIGYF2 mutations in an extended (N=305) Belgian series of both familial and sporadic PD patients, we sequenced all 32 coding and non-coding exons of GIGYF2. In three sporadic PD patients we identified two novel heterozygous missense mutations (c.1907A>G, p.Tyr636Cys and c.2501G>A, p.Arg834Gln), that were absent from control individuals (N=360). However, since we lack genetic as well as functional data supporting their pathogenic nature, we cannot exclude that these variants are benign polymorphisms. Together, our results do not support a role for GIGYF2 in the genetic etiology of Belgian PD.

Comprehensive genetic and mutation analysis of familial dementia with Lewy bodies linked to 2q35-q36.

The second most frequent form of neurodegenerative dementia after Alzheimer's disease is dementia with Lewy bodies (DLB). Since informative DLB families are scarce, little is presently known about the molecular genetic etiology of DLB. We recently mapped the first locus for DLB on chromosome 2q35-q36 in a multiplex Belgian family, DR246, with autopsy-proven DLB pathology in a region of 9.2 Mb. Here, we describe the ascertainment of additional DR246 family members and significant finemapping of the DLB locus to 3.3 Mb based on informative meiotic recombinants. Extensive sequencing of the 42 positional candidate genes within the DLB region did not identify a simple pathogenic mutation that co-segregated with disease in family DR246. Also high resolution analysis of copy number variations in the DLB locus did not provide evidence for a complex mutation. In conclusion, we confirmed the DLB locus at 2q35-q36 as a genetic entity but candidate gene-based sequencing and copy number variation analysis did not identify the pathogenic mutation in family DR246. Other detection strategies will be needed to reveal the underlying mutation explaining the linkage of DLB to 2q35-q26. Possibly the disease mutation in this family acts through a more complex mechanism than generally envisaged for monogenic disorders. Nevertheless, identifying the first familial DLB gene is likely to contribute an entry point into the pathogenic cascades underlying DLB pathology.

Promoter mutations that increase amyloid precursor-protein expression are associated with Alzheimer disease.

Genetic variations in promoter sequences that alter gene expression play a prominent role in increasing susceptibility to complex diseases. Also, expression levels of APP are essentially regulated by its core promoter and 5' upstream regulatory region and correlate with amyloid beta levels in Alzheimer disease (AD) brains. Here, we systematically sequenced the proximal promoter (-766/+204) and two functional distal regions (-2634/-2159 and -2096/-1563) of APP in two independent AD series with onset ages < or =70 years (Belgian sample, n=180; Dutch sample, n=111) and identified eight novel sequence variants. Three mutations (-118C-->A, -369C-->G, and -534G-->A) identified only in patients with AD showed, in vitro, a nearly twofold neuron-specific increase in APP transcriptional activity, similar to what is expected from triplication of APP in Down syndrome. These mutations either abolished (AP-2 and HES-1) or created (Oct1) transcription-factor binding sites involved in the development and differentiation of neuronal systems. Also, two of these clustered in the 200-bp region (-540/-340) of the APP promoter that showed the highest degree of species conservation. The present study provides evidence that APP-promoter mutations that significantly increase APP expression levels are associated with AD.

Linkage and association studies identify a novel locus for Alzheimer disease at 7q36 in a Dutch population-based sample.

We obtained conclusive linkage of Alzheimer disease (AD) with a candidate region of 19.7 cM at 7q36 in an extended multiplex family, family 1270, ascertained in a population-based study of early-onset AD in the northern Netherlands. Single-nucleotide polymorphism and haplotype association analyses of a Dutch patient-control sample further supported the linkage at 7q36. In addition, we identified a shared haplotype at 7q36 between family 1270 and three of six multiplex AD-affected families from the same geographical region, which is indicative of a founder effect and defines a priority region of 9.3 cM. Mutation analysis of coding exons of 29 candidate genes identified one linked synonymous mutation, g.38030G-->C in exon 10, that affected codon 626 of the PAX transactivation domain interacting protein gene (PAXIP1). It remains to be determined whether PAXIP1 has a functional role in the expression of AD in family 1270 or whether another mutation at this locus explains the observed linkage and sharing. Together, our linkage data from the informative family 1270 and the association data in the population-based early-onset AD patient-control sample strongly support the identification of a novel AD locus at 7q36 and re-emphasize the genetic heterogeneity of AD.

High-density SNP haplotyping suggests altered regulation of tau gene expression in progressive supranuclear palsy.

Two extended haplotypes exist across the tau gene-H1 and H2-with H1 consistently associated with increased risk of progressive supranuclear palsy (PSP). Using 15 haplotype tagging SNPs (htSNPs), capturing >95% of MAPT haplotype diversity, we performed association analysis in a US sample of 274 predominantly pathologically confirmed PSP patients and 424 matched control individuals. We found that PSP risk is associated with one of two major ancestral H1 haplotypes, H1B, increasing from 14% in control individuals to 22% in PSP patients (P<0.001). In young PSP patients, the H1B risk could be localized to a 22 kb regulatory region in intron 0 (P<0.001) and could be fully explained by one SNP, htSNP167, creating a LBP-1c/LSF/CP2 site, shown to regulate the expression of genes in other neurodegenerative disorders. Luciferase reporter data indicated that the 182 bp conserved regulatory region, in which htSNP167 is located, is transcriptionally active with both alleles differentially influencing expression. Further, we replicated the htSNP167 association in a second, independently ascertained US PSP patient-control sample. However, the htSNP association showed that H1 risk alone could not explain the overall differences in H1 and H2 frequencies in PSP patients and control individuals. Thus, risk variants on different H1 htSNP haplotypes and protective variants on H2 contribute to population risk for PSP.

Alzheimer-associated C allele of the promoter polymorphism -22C>T causes a critical neuron-specific decrease of presenilin 1 expression.

We, amongst others, have shown that CC homozygosity at the -22C>T promoter polymorphism in presenilin 1 (PSEN1) is associated with increased risk for Alzheimer's disease (AD). Also, studies in AD brains suggested that CC homozygosity increased the risk for AD by increasing the Abeta load. We characterized the PSEN1 promoter by deletion mapping, and analysed the effect of the -22C and -22T alleles on the transcriptional activity of PSEN1 in a transient transfection system. We showed a neuron-specific 2-fold decrease in promoter activity for the -22C risk allele, which in homozygous individuals would lead to a critical decrease in PSEN1 expression. The deletion mapping suggested that the 13 bp region (-33/-20) spanning the -22C>T polymorphism harbours a binding site for a negative regulatory factor. This factor has a higher affinity for the -22C risk allele and is strongly dependent on downstream sequences for cell-type-specific expression differences. Together, these studies provide evidence that the increased risk for AD associated with PSEN1 may result from genetic variations in the regulatory region, leading to altered expression levels of PSEN1 in neurons.

The gene encoding nicastrin, a major gamma-secretase component, modifies risk for familial early-onset Alzheimer disease in a Dutch population-based sample.

Nicastrin regulates gamma-secretase cleavage of the amyloid precursor protein by forming complexes with presenilins, in which most mutations causing familial early-onset Alzheimer disease (EOAD) have been found. The gene encoding nicastrin (NCSTN) maps to 1q23, a region that has been linked and associated with late-onset Alzheimer disease (LOAD) in various genome screens. In 78 familial EOAD cases, we found 14 NCSTN single-nucleotide polymorphisms (SNPs): 10 intronic SNPs, 3 silent mutations, and 1 missense mutation (N417Y). N417Y is unlikely to be pathogenic, since it did not alter amyloid beta secretion in an in vitro assay and its frequency was similar in case and control subjects. However, SNP haplotype estimation in two population-based series of Dutch patients with EOAD (n=116) and LOAD (n=240) indicated that the frequency of one SNP haplotype (HapB) was higher in the group with familial EOAD (7%), compared with the LOAD group (3%) and control group (3%). In patients with familial EOAD without the APOE epsilon4 allele, the HapB frequency further increased, to 14%, resulting in a fourfold increased risk (odds ratio = 4.1; 95% confidence interval 1.2-13.3; P=.01). These results are compatible with an important role of gamma-secretase dysfunction in the etiology of familial EOAD.