Molecular genetics of Alzheimer's disease

Genetic factors are involved in the aetiology of Alzheimer's disease (AD) in 25-40% of the cases. In some cases AD clearly segregates as an autosomal dominant trait in families. Three genes have been identified which, when mutated, cause AD: the Abeta amyloid precursor protein gene (APP), and the presenilin-1 (PSEN1) and presenilin-2 (PSEN2) genes. Together, these mutations are responsible for 30-50% of the cases with autosomal dominant AD, and for about 5% of AD in general. In cases where the inheritance pattern is unclear and in sporadic cases the epsilon4 allele of the apolipoprotein E gene (APOE) has been identified as a major risk factor contributing to the pathogenesis of AD in about 20% of the cases. Although mutations in the known genes are a rare cause of AD they are useful for the purposes of presymptomatic diagnostics in autosomal dominant AD families that segregate these mutations. Also, the identification of these genes and mutations has been extremely important to the recent evolution in the understanding of the biology of the disease. However, other causative and risk genes are involved in AD and need to be identified in order to fully elucidate the biology of AD. This will ultimately lead to the development of effective therapies for this major disease.

Risk estimates of dementia by apolipoprotein E genotypes from a population-based incidence study: the Rotterdam Study

OBJECTIVES

To provide risk estimates of dementia and Alzheimer disease as a function of the apolipoprotein E (APOE) genotypes and to assess the proportion of dementia that is attributable to the APOE genotypes.

DESIGN

Case-control study nested in a population-based cohort study with a mean (SD) follow-up of 2.1 (0.9) years.

SETTING

General population in Rotterdam, the Netherlands.

PARTICIPANTS

A total of 134 patients with incident dementia and a random sample of 997 nondemented control subjects. No participant had dementia at baseline.

MAIN OUTCOME MEASURES

Odds ratios for dementia and Alzheimer disease, the fraction of dementia attributable to the APOE epsilon4 allele, and the proportion of the variance in age at the onset of dementia explained by the APOE genotypes.

RESULTS

Persons with the epsilon4/4 genotype had a more than 10-fold higher risk of dementia (odds ratio, 11.2; 95% confidence interval, 3.6-35.2), and subjects with the epsilon3/4 genotype had a 1.7-fold increased risk of dementia (95% confidence interval, 1.0-2.9) as compared with persons with the epsilon3/3 genotype. The proportion of patients with dementia that is attributable to the epsilon4 allele was estimated to be 20%. The APOE genotypes explained up to 10% of the variance in age at the onset of dementia. The association between the epsilon4 allele and dementia was strongest in the youngest age category and in those with a family history of dementia.

CONCLUSIONS

The APOE genotype is an important determinant of the risk of dementia. At a population level, however, other factors than the APOE genotype may play an important role in the cause of dementia.

Genetic association of apolipoprotein E with age-related macular degeneration

Age-related macular degeneration (AMD) is the most common geriatric eye disorder leading to blindness and is characterized by degeneration of the neuroepithelium in the macular area of the eye. Apolipoprotein E (apoE), the major apolipoprotein of the CNS and an important regulator of cholesterol and lipid transport, appears to be associated with neurodegeneration. The apoE gene (APOE) polymorphism is a strong risk factor for various neurodegenerative diseases, and the apoE protein has been demonstrated in disease-associated lesions of these disorders. Hypothesizing that variants of APOE act as a potential risk factor for AMD, we performed a genetic-association study among 88 AMD cases and 901 controls derived from the population-based Rotterdam Study in the Netherlands. The APOE polymorphism showed a significant association with the risk for AMD; the APOE epsilon4 allele was associated with a decreased risk (odds ratio 0.43 [95% confidence interval 0.21-0. 88]), and the epsilon2 allele was associated with a slightly increased risk of AMD (odds ratio 1.5 [95% confidence interval 0.8-2. 82]). To investigate whether apoE is directly involved in the pathogenesis of AMD, we studied apoE immunoreactivity in 15 AMD and 10 control maculae and found that apoE staining was consistently present in the disease-associated deposits in AMD-maculae-that is, drusen and basal laminar deposit. Our results suggest that APOE is a susceptibility gene for AMD.

Serum apolipoprotein E level is not increased in Alzheimer's disease: the Rotterdam study

The APOE*4 allele of the apolipoprotein E gene (APOE) is an important risk factor for Alzheimer's disease. It has been suggested that levels of apolipoprotein E (apoE) in plasma are increased in Alzheimer's disease. In this population-based study, we found that serum apoE levels were lower in Alzheimer patients compared to non-demented controls (0.75 micromol/l (SD 0.35), vs. 0.83 micromol/l (SD 0.40), P < 0.05). This finding is in accordance with lower serum apoE levels as observed in carriers of the APOE*4 allele, who are over-represented in Alzheimer's disease. After adjustment for age, sex, total protein, albumin level, body mass index and the APOE genotype, the difference in serum apoE levels largely disappeared. Our population-based study suggests that the differences in serum apoE level between Alzheimer patients and controls are mainly the result of differences in the distribution of the APOE genotype.

A presenilin-1 truncating mutation is present in two cases with autopsy-confirmed early-onset Alzheimer disease

We have examined genomic DNA from 40 cases of autopsy-confirmed early-onset Alzheimer disease (EOAD) (age at onset <=65 years) that were all unselected for family history. We have sequenced the 10 exons and flanking intronic sequences of the presenilin-1 (PS-1) gene for all 40 individuals. A single mutation, a deletion of a G from the intron 4 splice-donor consensus sequence, was detected in two individuals in this study. The mutation was associated with two shortened transcripts, both with shifted reading frames resulting in premature-termination codons. All the PS-1 mutations described elsewhere have been missense or in-frame splice mutations, and recent data suggest that these result in disease by gain-of-function or dominant-negative mechanisms. The mutation that we have identified is likely to result in haploinsufficiency and would be most consistent with other mutations acting in a dominant-negative manner. However, we cannot exclude the possibility that the small amounts of truncated transcripts exert a gain of function. Since no other mutations or polymorphisms were detected in our patients, mutations in the coding regions and splice consensus sequences of PS-1 are likely to be rare in EOAD cases unselected for family history.

Estimation of the genetic contribution of presenilin-1 and -2 mutations in a population-based study of presenile Alzheimer disease

Two closely related genes, the presenilins ( PS ), located at chromosomes 14q24.3 and 1q42.1, have been identified for autosomal dominant Alzheimer disease (AD) with onset age below 65 years (presenile AD). We performed a systematic mutation analysis of all coding and 5'-non-coding exons of PS -1 and PS -2 in a population-based epidemiological series of 101 unrelated familial and sporadic presenile AD cases. The familial cases included 10 patients of autosomal dominant AD families sampled for linkage analysis studies. In all patients mutations in the amyloid precursor protein gene ( APP ) had previously been excluded. Four different PS -1 missense mutations were identified in six familial cases, two of which where autosomal dominant cases. Three mutations resulted in onset ages above 55 years, with one segregating in an autosomal dominant family with mean onset age 64 years (range 50-78 years). One PS -2 mutation was identified in a sporadic case with onset age 62 years. Our mutation data provided estimates for PS -1 and PS -2 mutation frequencies in presenile AD of 6 and 1% respectively. When family history was accounted for mutation frequencies for PS -1 were 9% in familial cases and 18% in autosomal dominant cases. Further, polymorphisms were detected in the promoter and the 5'-non-coding region of PS -1 and in intronic and exonic sequences of PS -2 that will be useful in genetic association studies.

The presenilin genes: a new gene family involved in Alzheimer disease pathology

A positional cloning approach has led to the identification of two closely related genes, the presenilins (PS), for autosomal dominant presenile Alzheimer disease (AD): PS-1 at 14q24.3 and PS-2 at 1q31-q42. The PS-1 gene was identified by direct cDNA selection of yeast artificial chromosomes containing the candidate chromosomal region. Subsequently, the PS-2 gene was identified due to its high sequence homology with PS-1 and its location within the candidate region defined by linkage studies. To date, 30 different missense mutations and one in-frame splice site mutation were described in PS-1, while only two missense mutations were detected in PS-2, suggesting that PS-1 mutations are more frequently involved in familial presenile AD. The PS transcripts encode novel proteins that resemble integral transmembrane proteins of roughly 450 amino acids and at least seven transmembrane domains. The genomic organization of the PS genes is very similar showing that full length PS-1 and PS-2 are encoded by 10 exons. However, different alternative splicing patterns have been observed for PS-1 and PS-2 indicating that the corresponding proteins (ps-1 and ps-2) may have similar but not identical biological functions.

Molecular genetic analysis of familial early-onset Alzheimer's disease linked to chromosome 14q24.3

Genetic linkage studies have indicated that chromosome 14q24.3 harbours a major locus for early-onset (onset age <65 years) Alzheimer's disease (AD3). Positional cloning efforts have identified a novel gene S182 or presenilin 1 as the AD3 gene. We have mapped S182 in the AD3 candidate region between D14S277 and D14S284 defined by genetic linkage studies in the two chromosome 14 linked, early-onset AD families AD/A and AD/B. We have shown that S182 is expressed in lymphoblasts and have determined the complete cDNA in both brain and lymphoblasts by RT-PCR sequencing. S182 is alternatively spliced in both brain and lymphoblasts within a putative phosphorylation site located 5' in the coding region. We identified two novel mutations, Ile143Thr and Gly384la located in, respectively, the second transmembrane domain and in the sixth hydrophilic loop of the putative transmembrane structure of S182. As families AD/A and AD/B have very similar AD phenotype our observation of two mutations in functionally different domains suggest that onset age and severity of AD may not be very helpful predictors of the location of putative S182 mutations.

Mutation analysis of the chromosome 14q24.3 dihydrolipoyl succinyltransferase (DLST) gene in patients with early-onset Alzheimer disease

Linkage analysis studies have indicated that the chromosome band 14q24.3 harbours a major gene for familial early-onset Alzheimer's disease (AD). Recently we localized the chromosome 14 AD gene (AD3) in the 6.4 cM interval between the markers D14S289 and D14S61. We mapped the gene encoding dihydrolipoyl succinyltransferase (DLST), the E2k component of human alpha-ketoglutarate dehydrogenase complex (KGDHC), in the AD3 candidate region using yeast artificial chromosomes (YACs). The DLST gene is a candidate for the AD3 gene since deficiencies in KGDHC activity have been observed in brain tissue and fibroblasts of AD patients. The 15 exons and the promoter region of the DLST gene were analysed for mutations in chromosome 14 linked AD cases and in two series of unrelated early-onset AD cases (onset age < 55 years). Sequence variations in intronic sequences (introns 3, 5 and 10) or silent mutations in exonic sequences (exons 8 and 14) were identified. However, no AD related mutations were observed, suggesting that the DLST gene is not the chromosome 14 AD3 gene.

A yeast artificial chromosome contig from human chromosome 14q24 spanning the Alzheimer's disease locus AD3

Familial Alzheimer's disease has been previously linked to three genetic loci on chromosomes 21, 19 and 14. The AD3 locus on chromosome 14 has not been cloned and the molecular defect in chromosome 14-linked AD3 families has yet to be identified. Genetic linkage analysis has placed the AD3 locus in band 14q24 between the dinucleotide markers D14S61 and D14S289, a genetic distance of approximately 6.4 cM. We have constructed a yeast artificial chromosome (YAC) contig that covers the entire minimal region, encompassing all genetic markers that are non-recombinant for the disease in AD3-linked families. This contig, constructed by using a combination of YAC end sequence walking and sequence-tagged site (STS) mapping, consists of 63 YACs from three different libraries. The AD3 contig contains 12 polymorphic dinucleotide repeat markers from D14S61 to D14S251, as well as an additional 43 non-polymorphic STSs. This contiguous physical map of the region will allow the physical distances between the markers to be determined, as well as providing a framework for the identification of candidate genes.

Genetic and physical characterization of the early-onset Alzheimer's disease AD3 locus on chromosome 14q24.3

Genetic linkage studies have provided significant evidence that a major gene defect, AD3, for familial early-onset Alzheimer's disease (EOAD) is located at chromosome 14q24.3, between the short tandem repeat (STR) markers D14S52 and D14S53 defining a genetic size of 22.7 cM for the AD3 candidate region. We constructed a physical map of the AD3 region using yeast artificial chromosomes (YACs) selected from both the CEPH and megaCEPH YAC libraries using the AD3 linked STR markers as well as new sequence-tagged sites (STSs) designed based on YAC terminal sequences. The YAC map is contiguous in the region between D14S258 and D14S53, a region of 8.2 cM, and has an estimated physical size of 4-8 Mb. The YAC contig map was used as a framework to localize three known genes, a pseudogene and two brain expressed sequence tags (ESTs). Linkage analysis studies in two Belgian chromosome 14 EOAD families AD/A and AD/B, identified obligate recombinants in family AD/A with D14S289 and D14S61 reducing the genetic size of the candidate AD3 region substantially. The minimal AD3 candidate region measured 6.4 cM on the genetic map and is contained within six overlapping megaCEPH YACs that covered a physical distance estimated between 2 and 6 Mb. These YACs as well as other YACs in the YAC contig map are valuable resources in gene cloning efforts or genomic sequencing experiments aiming at isolating the AD3 gene.

A population-based study of familial Alzheimer disease: linkage to chromosomes 14, 19, and 21

Linkage of Alzheimer disease (AD) to DNA markers on chromosomes 14, 19, and 21 was studied in 10 families in which the disease was apparently inherited as an autosomal dominant trait. Families were derived from a Dutch population-based epidemiologic study of early-onset AD. Although in all probands the onset of AD was at or before age 65 years, the mean age at onset was after age 65 years in four families (referred to as "LOAD"). Among the six families with early-onset AD (referred to as "EOAD," i.e., mean age of onset of AD of relatives was at or before age 65 years), conclusive linkage to 14q24.3 was found in one family with a very early onset (around 47 years), while linkage to the same region was excluded in two other families. For the LOAD families, predominantly negative lod scores were obtained, and the overall lod score excluded linkage to chromosome 14. The results with markers on chromosome 19 and chromosome 21 were not conclusive for EOAD and LOAD. The findings of our study confirm genetic heterogeneity within familial EOAD.

Genetic analysis of the cellular oncogene fos in patients with chromosome 14 encoded Alzheimer's disease

A major gene for familial early-onset Alzheimer's disease (AD) has been localised to chromosome 14q24.3. The c-fos gene (FOS), localised to 14q24.3-q31, is a candidate for the AD gene since it may be involved in the transcription regulation of the amyloid precursor protein gene (APP). Part of APP codes for the beta A4 amyloid present in AD brain lesions. We analyzed linkage of AD in the 2 early-onset AD families. AD/A and AD/B, using a polymerase chain reaction (PCR) based assay for a restriction fragment length polymorphism (RFLP). The RFLP is detected in BstNI digested DNA and is located near the 3' end of FOS. No obligate recombinants were detected. The 4 exons of FOS were sequenced in one pathologically confirmed AD patient in each family. No exonic mutations were found. Two intronic sequence variations were observed, one in intron 2 and one in intron 3. The intron 2 variation did not segregate with AD. The intron 3 variation which is a single G insertion was used in linkage studies in families AD/A and AD/B and showed conclusive linkage in both families in the absence of recombinants. Therefore, FOS cannot yet be excluded as a candidate gene for AD in these families since mutations may be present in regulatory sequences.

Apolipoprotein E4 allele in a population-based study of early-onset Alzheimer's disease

Several studies have reported an association of the apolipoprotein E allele epsilon 4 (APOE*4) to familial and sporadic late-onset Alzheimer's disease (LOAD). Here we report on the relationship between APOE*4 and early-onset Alzheimer's disease (EOAD) in a Dutch population-based study. The frequency of the APOE*4 allele was 2.3 times higher among EOAD cases compared to controls. Among patients, the allele frequency was 1.6 times higher in those with a positive family history than in those without. A significant increase in risk of EOAD was found for subjects homozygous for APOE*4 regardless of family history of dementia, but an increase in EOAD risk for APOE*4 heterozygotes could only be shown in subjects with a positive family history. Our study demonstrates a significant association between APOE*4 and EOAD which is modified by family history of dementia.

APOE genotype does not modulate age of onset in families with chromosome 14 encoded Alzheimer's disease

A recent study has demonstrated an association of the apolipoprotein E allele epsilon 4 (APOE*4) to familial and sporadic late-onset Alzheimer's disease (AD). Also, in late-onset AD families linked to chromosome 19, the onset age decreased when the number of APOE*4 alleles increased. A similar effect of the APOE*4 genotype was observed in early-onset AD families linked to mutations in the APP gene located on chromosome 21. We assessed whether the APOE genotype had an influence on the age of onset of AD in chromosome 14 linked early-onset AD families. No significant effect of the APOE genotype on onset age could be detected.

Mapping of a gene predisposing to early-onset Alzheimer's disease to chromosome 14q24.3

Genetic linkage studies with chromosome 21 DNA markers and mutation analysis of the beta-amyloid protein precursor gene located in 21q21.3 have indicated that early-onset Alzheimer's disease (EOAD) is a heterogeneous disorder for which at least one other chromosomal locus exists. We examined two extended histopathologically confirmed EOAD pedigrees, AD/A and AD/B, with highly informative short tandem repeat (STR) polymorphisms and found complete linkage of the disease to a (CA)n dinucleotide repeat polymorphism at locus D14S43 in 14q24.3 (Zmax = 13.25 at theta = 0.0). Using additional chromosome 14 STR polymorphisms we were able to delineate the region containing the EOAD gene to an area of, at most, 8.9 centiMorgans between D14S42 and D14S53, flanking D14S43 on both sides.

Presenile dementia and cerebral haemorrhage linked to a mutation at codon 692 of the beta-amyloid precursor protein gene

Several families with an early-onset form of familial Alzheimer's disease have been found to harbour mutations at a specific codon (717) of the gene for the beta-amyloid precursor protein (APP) on chromosome 21. We now report, a novel base mutation in the same exon of the APP gene which co-segregates in one family with presenile dementia and cerebral haemorrhage due to cerebral amyloid angiopathy. The mutation results in the substitution of alanine into glycine at codon 692. These results suggest that the clinically distinct entities, presenile dementia and cerebral amyloid angiopathy, can be caused by the same mutation in the APP gene.

Dinucleotide repeat polymorphism at the D21S16 [correction of D1S16] locus

Images

Unique sequence homology in the pericentromeric regions of the long arms of chromosomes 13 and 21

We previously isolated two polymorphic chromosome 21q probes, pVC1.21c (D21S190) and pVC1.34a (D21S149), localized in 21qcen-21q21.2. In addition, pVC1.21c recognized a sequence in 21q22.1-q22.2 and both probes cross-hybridized with non-chromosome-21 sequences. In this study we refined the proximal 21q locations of probes pVC1.21c and pVC1.34a to 21q11.1 and demonstrated that they recognize sequences on chromosome 13 but not on chromosomes 14, 15, and 22. Furthermore, the polymorphisms associated with the two loci were assigned to pericentromeric 13q for pVC1.34a and distal 21q for pVC1.21c. Our results are indicative of a region of unique sequence homology in the pericentromeric region of the long arms of chromosomes 13 and 21.

Identification of chromosome 21 DNA polymorphisms for genetic studies in Alzheimer's disease and Down syndrome

Linkage studies in families with presenile onset of Alzheimer's disease (AD) indicated the presence of a predisposing gene on the proximal long arm of chromosome 21. We mapped four new loci in the candidate AD region using somatic cell hybrids. For three of the four loci, several restriction fragment length polymorphisms were found; for one locus, a multiallelic (CA)n dinucleotide polymorphism was detected. Preliminary genetic mapping of the new polymorphic loci relative to the AD-linked loci was obtained in a reference pedigree. In addition, we used the (CA)n dinucleotide polymorphism to reconstruct the non-disjunction event in a Down syndrome (DS) patient whose mother died of familial AD.

Subregional localization of the chromosome 21 loci D21S24 and D21S26 using physical mapping techniques

We were able to refine the chromosomal position of two existing marker loci, using an extended chromosome 21 somatic cell hybrid panel. The locus D21S26 mapped in the region 21q11.2-q21.1, and the locus D21S24 in 21q22.1-q22.2. Physical and genetic analysis indicated that D21S26 is tightly linked to D21S13 and D21S16, two markers previously linked to familial Alzheimer's disease.