WO2000079000A1 - Method for detection of early-onset alzheimer's disease - Google Patents

Abstract

The present invention relates generally to the field of neurological and physiological dysfunctions associated with Alzheimer's disease (AD). More particularly, the invention concerns the use of a variant in the preseniline-1 (PSEN1) regulatory region to determine in a human cell or tissue obtainable from a human being whether said human being has or is at risk for developing early-onset AD or AD. The present invention also relates to methods to screen for molecules that inhibit the reduction of PSEN-1 levels associated with polymorphisms in the PSEN-1 promoter region and molecules obtained via said methods.

Description

Method for detection of early-onset Alzheimer's disease.

Field of the invention The present invention relates generally to the field of neurological and physiological dysfunctions associated with Alzheimer's disease (AD). More particularly, the invention concerns the use of a variant in the preseniline-1 (PSEN1 ) regulatory region to determine in a human cell or tissue obtainable from a human being whether said human being has or is at risk for developing early-onset AD or AD. The present invention also relates to methods to screen for molecules that inhibit the reduction of PSEN-1 levels associated with polymorphisms in the PSEN-1 promoter region and molecules obtained via said methods.

Background of the invention In recent years important progress has been made in uncovering genes implicated in Alzheimer's disease. Three genes have been associated with autosomal dominant eariy-onset AD (1 ). In the β amyloid precursor protein (APP) gene located on chromosome 21 (2), six mutations were identified leading to early-onset AD with disease onset between 50 and 60 years of age (3). However, these mutations were found to be rare (1 ). Furthermore, two homologous genes, presenilin 1 (PSEN1) on chromosome 14 and presenilin 2 (PSEN2) on chromosome 1 , were identified (4-6). At present, 43 mutations are known in PSEN1 that lead to AD, in most cases before age 65 years (7). So far, only three mutations in PSEN2 were identified in AD cases (7). Screening a population-based series of early-onset AD patients (onset before 65 years of age) provided mutation frequencies for PSEN1 of 6% overall, 9% in cases with at least one first degree relative with dementia and 18% in autosomal dominant families (8). In this series only one PSEN2 mutation (1%) was found in a sporadic AD case (8). In addition to the autosomal dominant AD mutations in APP, PSEN1 and PSEN2, genetic association was consistently reported between the ε4 allele of the apolipoprotein E (APOE) gene and both late- and early-onset AD (9-10). Also, genetic association was reported between a di-allelic polymorphism in intron 8 of PSEN1 and late-onset AD, in which a twofold increase of the 11 genotype was observed (11 ). Since the association was present in a Caucasian North-American population but not in an African-American one, the possibility of linkage disequilibrium was put forward (11 ). In line with this explanation, the association was confirmed in case-series from Japan and the United Kingdom (12-16), while other populations of Caucasian descent failed to show association (17-23). Furthermore, no evidence was found for an influence of the PSEN1 intron 8 polymorphism on the amount or molecular form of β amyloid deposition, one of the main pathological characteristics of AD, suggesting that this polymorphism is not functionally related to AD pathogenesis (24). Until now, there is little evidence for genetic variability in PSEN1 that could account for the association with the PSEN1 intron 8 polymorphism (8). The only other polymorphism at 14q24.3 that was associated with early-onset AD is the D14S43 locus (25) located telomeric of PSEN1 (26), but this finding has never been confirmed.

Aims and summary of the invention It is clear that uncovering variants, mutations or polymorphisms of genes implicated in Alzheimer's disease is crucial for developing new diagnostic tools and therapeutic compounds. It is therefore an aim of the present invention to provide diagnostic methods, research tools such as transgenic animals and therapeutic compounds against AD. The present invention is based on a study of polymorphisms within and flanking the PSEN1 gene in 102 patients with early-onset AD and 118 age and sex matched controls. It was found that two di-allelic PSEN1 polymorphisms were associated with AD: a promoter polymorphism (p=0.02) and an intron 8 polymorphism (p=0.05). Association was also found with D14S1028, a simple tandem repeat (STR) polymorphism located upstream of PSEN1 (p=0.04). When excluding PSEN1 mutation cases (n=6), the intron 8 association was explained by linkage disequilibrium to the dominant PSEN1 mutations. In the non-mutation cases, the regulatory region of PSEN1 and D14S1028 associations remained significant. The 11 genotype of the promoter polymorphism (-48 C/T) was found 2.8 times more frequent in AD patients (95% confidence interval (CI) 1.2-6.6; p=0.01). The frequency of the 44 genotype of D14S1028 was 2.9 times increased (95% CI 1.1-7.9; p=0.005). The association of PSEN1 with AD was independent of the apolipoprotein E (APOE) genotype. The present invention also discloses that the PSEN1 promoter polymorphisms -2154 G/A, -2818 A G and -2823 l/D are associated with an increased risk for AD. The present invention further discloses that, especially with regard to the PSEN1 promoter variation/polymorphism -48 C/T, a neuron-specific twofold decrease in PSEN1 promoter activity for the -48C risk allele can in homozygous individuals lead to a critical decrease in PSEN1 expression. In addition, the PSEN1 promoter polymorphism/mutation -280 C/G was also shown to be associated with a significant decrease in promoter activity, in view of these data, the present invention aims at providing a method for determining whether a human subject has or is at risk for developing early-onset Alzheimer disease or Alzheimer disease comprising the step of detecting the presence or absence of a genetic lesion in the presenilin-1 gene of said subject, wherein said genetic lesion comprises a polymorphism in the presenilin-1 promoter region or regulatory region upstream of presenilin-1 and wherein the presence of said genetic lesion identifies a subject that has or is at risk for developing early-onset Alzheimer disease or Alzheimer disease. More specifically, the present invention aims at providing a method as described above, wherein said polymorphism in the presenilin-1 promoter region is chosen from the group comprising: -48 C/T; -280 C/G; -2154 G/A, -2818 A/G and -2823 l/D and/or wherein said polymorphism in the regulatory region upstream of presenilin-1 is a simple tandem repeat polymorphism at D14S1028. The present invention also aims at providing a transgenic non-human animal comprising in its genetic material a human presenilin-1 variant/mutation/polymorphism (the latter three terms can be used interchangeably) in the presenilin-1 promoter region and/or in the presenilin-1 regulatory region upstream of presenilin-1 and to provide a usage of said transgenic non-human animal to screen for therapeutic molecules to treat Alzheimer disease. The present invention further aims at providing a method to screen for molecules that inhibit the reduction of presenilin-1 levels induced by a polymorphism in the preseniiin-1 promoter region (specifically with regard to the polymorphisms in the presenilin-1 promoter region -48 CT and/or -280 C/G) and/or regulatory region upstream of presenilin-1 comprising:

-exposing said polymorphism in the presenilin-1 promoter region or regulatory region upstream of preseniiin-1 to said molecules, and -monitoring said presenilin-1 levels. The present invention finally relates to molecules obtainable by using a transgenic animal and/or a method as described above. Table and Figure legends

Table 1. Allele and genotype frequencies of polymorphisms in PSEN1 in early-onset AD cases and controls.

Table 2. Allele frequencies of STR polymorphisms flanking PSEN1 in early-onset AD cases and controls.

Table 3. Genotype frequencies in early onset AD cases stratified by the presence of PSEN1 mutations. Table 4. Allele and genotype frequencies of polymorphisms in PSEN2 in early-onset AD cases and controls.

Table 5: Sequence variations between the Neurogen sequence (N) and AF109907. Table 6: Overlapping primer sets of the PSEN1 promoter region Table 7: Polymorphisms detected in the PSEN1 5' upstream region. The nucleotide positions are relative to the uppermost transcription initiation site of exon 1A. Allele frequencies were determined in the 118 control individuals using SSCP.

Figure 1. Map of the chromosome 14q24.3 region containing PSENL The position of the STR polymorphisms and the 5'→3' orientation of PSEN1 are as published by Cruts et al. (37). Distances are in centimorgan (cM) according to the Genethon sex averaged genetic map of chromosome 14 (39), or in kilobases (kb) defined by restriction mapping of PSEN1 and YAC fragmentation. (Del-Favero,J et al., Gene,1999;193-201 ). Figure 2: Functional analysis of polymorphisms in the PSEN1 5' upstream region. Bars represent firefly luciferase/renilla luciferase ratios for the different constructs. Transcriptional activity data are presented as a percentage of the activity of the KHCCG construct with the C allele for the -280 C/G, the C allele for -48 C/T and the G allele for the + 522 C/G/A polymorphisms. Values are the mean ± SEM of in duplo determinations of three experiments using three independent DNA preparations in each of the experiments. Figure 3: Deletion mapping of functional PSEN1 promoter elements, a) The upper horizontal line represents a detailed restriction map of plasmid B22 (13). Exons 1A and 1B of PSEN1 are represented by shaded boxes. The positions of the PSEN1 -48C>T and -280C>G promoter polymorphisms are shown by arrows. Restriction enzyme recognition sites used for subcloning into the pGL3 basic vector are represented in bold. The deletion constructs in the pGL vector are represented by horizontal lines. Position of the 5' and 3' end of each construct is given based on the exonlA transcriptional start site t2 reported by Rogaev et al. (12) (L76518). Total length is given in bp. The relative luciferase activity of each construct in N2A and HEK293 cells is presented as a percentage of the SV40 promoter construct (%RLA). Figure 4:. Transcriptional activity of the PSEN1 -48C/T variants in transient transfection experiments. Bars represent firefly/reniila luciferase ratios for the different constructs (relative luciferase activity, RLA). Transcriptional activities are presented as percentage of the activity of the wild type construct (-280C, white bars). Values are the mean ± SEM of in duplo determinations of at least 3 experiments of 3 independent DNA preparations each.

Figure 5: EMSA analysis of allele-specific effect of the -280C>G mutation, on the interaction of nuclear protein complexes present in HEK293 cells with the (-291/-268) region. The same effect is seen for mouse (N2A) and human neuroblastoma (Kelly) cells (data not shown). 1-5 ng ³²P- labelled double-stranded probes were incubated with 10 μg nuclear extract from HEK293 cells. Lanes 1-6: include WT -280C construct. Lanes 7-16: include mutant -280G construct. In competition experiments 100 or 200 ng unlabeled probe was added before the labelled probe. In supershift experiments 400ng α-NF-1 was pre-incubated with the extract prior to incubation with labelled probe. Specific complexes are indicated with filled arrows. Supershifted bands are indicated with open arrows.

Detailed description of the invention A genetic association study of polymorphisms in PSEN1 was conducted in a population-based series of 102 patients with early-onset AD and 118 community controls. All patients were previously subjected to mutation analysis of exons 16 and 17 of APP (amyloid precursor protein gene)(10) and all exons of PSEN1 and PSEN2 (8). No APP mutations were found, while six patients carried a PSEN1 mutation and one patient a PSEN2 mutation (8). Also, the frequency of the APOE ε4 allele was 2.3 times increased in patients compared to controls (27). In this invention, di-allelic PSEN1 polymorphisms in the promoter region, the 5' untranslated region (UTR) (8), intron 8 (11) and the 3' UTR were analysed. Furthermore linkage disequilibrium with STR polymorphisms flanking PSEN1 was examined. For PSEN2 three polymorphisms in exons 3 and 4 and intron 11 (8) were analysed, since no polymorphisms in the promoter region or UTRs were described. According to the current invention inside the PSEN1 gene, di-allelic polymorphisms were examined (Table 1 and Table 7). In the population studied, the allele (p=0.05) and genotype (p=0.03) distributions of the intron 8 polymorphism in patients differed significantly from that in controls. Unexpectedly the promoter polymorphism -48 C/T was associated with AD when comparing allele (p=0.02) and genotype (p=0.05) distributions between patients and controls. No significant differences were observed for the 5'UTR and 3'UTR polymorphisms. When analysing STR markers upstream of PSEN1 (Figure 1 ), the allele distribution of D14S1028 was also unexpectedly significant different between AD cases and controls (p=0.04) (Table 2). No significant association was detected with D14S77 located 30 kb upstream of PSEN1 (Figure 1 ). Downstream of PSEN1 , no association was found with D14S1004 or the more distantly located STRs D14S43 and D14S61 or the di-allelic polymorphisms in FOS and DLST (Figure 1 ). The significant differences in genotype distributions are explained by an increase of the 22 genotype for the intron 8 polymorphism, the 11 genotype for the promoter polymorphism and the 44 genotype for D14S1028. Genotype frequencies were stratified for the presence of mutations in PSEN1 (Table 3). Five out of the six patients with a PSEN1 mutation carried the promoter genotype 11 (83% compared to 80% in controls), while four patients carried the intron 8 genotype 22 (67% compared to 21% in controls; p=0.02) and D14S1028 genotype 44 (67% compared to 6% in controls; p=0.0002). In patients without PSEN1 mutations, the association remained significant for the promoter genotype 11 (OR 2.8; 95% CI 1.2-6.6; p=0.01) and the D14S1028 genotype 44 (OR 2.9; 95% CI 1.1-7.9; p=0.005), but not for the intron 8 polymorphisms. Similar genotype frequencies were found for the promoter and D14S1028 in APOE ε4 carriers and non-carriers. Allele and genotype distributions of the PSEN2 polymorphisms in exons 3 and 4 and intron 11 were similar in cases and controls (Table 4). Furthermore the allele distribution of the STR polymorphism at D1S479 near PSEN2 (5) was similar in cases and controls (p=0.57). Also, stratification for PSEN1 and/or PSEN2 mutations or the presence of an APOE ε4 allele failed to show evidence for association of PSEN2 with early-onset AD. Thus, a significant association is found with the promoter polymorphism -48C>T. The most frequent allele -48C is associated with an increased risk for early onset AD (EOAD) (OR=2.6) due to a significant overrepresentation of the CC genotype in the cases, independent of the APOEΛ allele. Moreover, in a systematic screen of the PSEN1 regulatory region 2 additional polymorphisms (-2154G>A and -2823I/D) that associated with increased risk for EOAD were found (table 7). All three polymorphisms showed a high degree of linkage disequilibrium allowing the identification of a risk haplotype (-48C/-2154G/-2823D). Homozygosity for the risk haplotype is significantly higher in cases than controls (p=0.03). In addition, we found 2 heterozygous mutations (-280C>G and -2818A>G) in 2 patients homozygous for the risk haplotype (table 7). Each polymorphism and mutation involved consensus sequences of transcriptional regulatory elements. Luciferase reporter gene data demonstrated that the -280G mutation decreased PSEN1 transcriptional activity by 30% in neuroblastoma cells (Fig. 2). The present invention also discloses data regarding the functionality of the - 280C>G and -48C>T variants in relation to EOAD showing a neuron-specific twofold decrease in promoter activity for the -48C risk allele, which in homozygous individuals can lead to a critical decrease in PSEN1 expression. Despite the neuron-specific effects on transcriptional activity, no cell-type specific binding of nuclear factors is detected for these variations. However, for the -280C>G mutation an allele-specific differential binding of nuclear proteins, one of which is a member of the NF-1 family, known to be involved in neuron-specific transcription is demonstrated. Hence the invention concerns a method for determining whether a human subject has or is at risk for developing early-onset Alzheimer disease (AD) or Alzheimer disease comprising the step of detecting the presence or absence of a genetic lesion in the presenilin-1 gene of said subject, wherein said genetic lesion comprises a polymorphism in the presenilin-1 (PSEN-1 ) promoter region or regulatory region upstream of PSEN-1 and wherein the presence of said genetic lesion identifies a subject that has or is at risk for developing early-onset AD or AD. The PSEN1 promoter polymorphism as such has recently been identified while performing a systematic mutation analysis of the coding and non-coding exons of PSEN1 in this same series of early-onset AD cases (8). The polymorphism is a C to T nucleotide change at nucleotide position -48, a C to G nucleotide change at nucleotide position -280, a G to A nucleotide change at nucleotide position -2154, a A to G nucleotide change at nucleotide position -2818, and a deletion (D)/insert (I) of 13 bp (5'-GCATGTCCTGGGT-3') at position -2823, upstream of exon 1A (8). PSEN1 has 13 exons of which exons 1A, 1 B and 2 are 5' non-coding exons (33). The allele 1 (C) is the most frequent with a frequency of 88% in the control individuals. In the cases, the allele 1 frequency is increased and the genotype 11 is 2.8 times more frequent than in the controls (p=0.01 ). Differential expression of APOE in brain of AD patients with specific APOE promoter polymorphisms has been associated with an increased risk for late-onset AD independent of the APOE ε4 genotype (33-35). Recently two early-onset AD cases were identified wherein the first out-frame PSEN1 mutation in intron 4 leads to deleted transcripts and early truncation of the PSEN1 protein (36). In these cases it was hypothesised that AD would result from haplo-insuflϊciency of wild type PSEN1 protein. Another possibility is that other, not yet identified, polymorphisms in the PSEN1 promoter region are responsible for the observed association with PSENL In this respect the finding with D14S1028 is important. According to the physical mapping data using yeast artificial chromosomes (YACs), D14S1028 is located upstream of the PSEN1 gene (see Figure 1 ) (37). With D14S1028 a 2.9 times increased frequency of the 44 genotype (p=0.005) was found. Also, allele 4 is in linkage disequilibrium with allele 1 of the promoter polymorphism since all carriers of the 44 genotype were homozygous for allele 1 at the promoter polymorphism. In contrast, no association was found of early-onset AD with the STR marker D14S77, also located upstream of PSEN1 at a distance of 20 kb from exon 1A (Figure 1 ). The lack of association with D14S77 may be explained by the limited power of this study in examining such a highly polymorphic marker with 25 alleles observed in our sample of cases and controls. Alternatively, it might be explained if D14S77 has a high mutation rate, as suggested by the large number of alleles, which would prevent detecting shared alleles. In line with this, no linkage disequilibrium was detected between D14S77 and the PSEN1 promoter polymorphism or D14S1028. Genetic association was also absent with the polymorphism located in exon 1 B (8). Exons 1A and 1B are alternatively used exons and sequence analysis of PSEN1 clones from a hippocampal cDNA library suggested that only exon 1A containing transcripts are present in brain (38). Also, there is no linkage disequilibrium between the exon 1 B polymorphism and the promoter or D14S1028 polymorphism. No association was found with polymorphisms in the 3'UTR or downstream of PSENL

Alternatively, the present invention concerns the use of a variant in the PSEN1 regulatory region to determine in a human cell or tissue obtainable from a human being whether said human being has or is at risk for developing early-onset AD or AD. Individuals with a high risk for AD (present in family pedigree) or, individuals not previously known to be at risk, or people in general may be screened routinely using for instance probes to detect the presence of a promoter PSEN 1 polymorphism by a variety of techniques. Genomic DNA used for the diagnosis may be obtained from body cells, such as those present in the blood, tissue biopsy, surgical specimen or autopsy material. The DNA may be isolated and used directly for detection or may be amplified by known techniques prior to analysis. In accordance with another aspect of the invention a transgenic mouse model for AD has incorporated in its mouse genome the human promoter region or regulatory regions upstream PSEN1 mutated to manifest the symptoms. The transgenic mouse exhibits symptoms of cognitive memory loss or behavioural disturbances. Said transgenic mouse can be used as starting point for rational drug design to provide therapeutic drugs or other types of small chemical molecules. The current invention discloses a first study that examines several polymorphic markers spanning the whole gene as well as flanking regions and shows association to multiple markers in the region upstream of PSEN1. In contrast to the study of Wragg et al. (11 ), who reported an increased homozygosity of the genotype 11 of the intron 8 polymorphism in late-onset AD patients, a statistically significantly increased frequency of genotype 22 has now been found. However, when we excluded the patients with a PSEN1 mutation, no association could be detected indicating that in our sample the intron 8 association is explained by linkage disequilibrium with PSEN1 mutations. These findings are in line with those of Sorbi et al. (20), who found evidence for association with the PSEN1 intron 8 polymorphism in patients of families segregating PSEN1 mutations, but not in patients with sporadic forms of AD. All six patients with a PSEN1 mutation in our series had a positive family history of AD and in two of them the mutation was proven to segregate in their family in an autosomal dominant manner (8,27). Three of the four patients with a PSEN1 mutation and the intron 8 genotype 22 carried the same PSEN1 Ala79Val mutation and allele sharing studies with STR polymorphisms flanking PSEN1 indicated that these patients received the mutation from a common ancestor (8). In contrast to the intron 8 association, the associations with the PSEN1 promoter and D14S1028 polymorphisms remained statistically significant when PSEN1 mutation carriers were excluded. Also, when stratifying for the presence or absence of the APOE ε4 allele the genetic association remained significant and was not different in both groups, suggesting that the association of PSEN1 with early-onset AD is independent of APOE. This is in agreement with the observation that the APOE genotype does not modulate the age of onset in patients of autosomal dominant AD families segregating a PSEN1 mutation (32). Significant association of early-onset AD with APOE ε4 in this series of population based early- onset AD patients had previously been demonstrated (10). The APOE ε4 allele and genotype distributions in cases and controls were not different from that reported for other Caucasian populations. The above-captioned data show that the association of early-onset AD with PSEN1 results from polymorphism(s) in the promoter region or regulatory regions upstream of PSEN1. Only partial information has been published on the PSEN1 promoter sequence (38) and no studies are available that demonstrate how PSEN1 expression is regulated in different tissues including brain. Therefore in the context of the current invention several clones have been isolated covering 28 kb upstream of exon 1A of PSENL Sequence analysis of these clones as well as functional studies of the regulatory sequences elucidate the genetic risk factor underlying the association observed between PSEN1 and early-onset AD in this invention. Thus approximately 7 kb upstream of the second non-coding exon of PSEN1 , exon 1B, was sequenced first and accordingly primers were designed to gradually screen this region for sequence variations associated with AD in the same series of EOAD patients (early-onset AD) and controls. Seven sequence variations were found (Table 7), two of which are rare polymorphism's occurring only in one patient. Three polymorphisms are shown to be associated with EOAD. The first variation, a C to T transition, corresponds to the promoter polymorphism mentioned above and is located 48 bp upstream of exon 1 A. The other two variations are located more upstream of exon 1A and are in linkage disequilibrium with the -48 variation. Previously we determined the genomic organisation of PSEN1 , constructing and characterising a 140 kb clone contig spanning the coding and regulatory region of PSEN1 using fiber-FISH, long-range PCR and restriction mapping. This clone contig contains 28 kb upstream of the first, non-coding exon, exon 1A, in BAC 16. Plasmid B22 contains the sequence from 6.5 kb upstream of exonlA to 2.6 kb downstream of exon 1 B, and was used as the starting clone in this invention. The present invention also relates to a functional analysis of the PSEN1 promoter region in relation to the EOAD related variations in the PSEN1 proximal promoter: - 280C>G and -48C>T. Deletion analysis of the PSEN1 upstream region showed that the smallest promoter fragment, extending from -47 to +823 though still containing the most upstream TSS (54), confers maximum promoter activity in both N2A and HEK293 cells even exceeding SV40 promoter activity in N2A cells. We detected the largest reduction in transcriptional activity when deleting the —47/+15 fragment, containing the major exon 1A TSSs (51 , 54, 66), suggesting that this fragment harbours the core promoter elements. Matlnspector v2.2 predicted three major elements with a 100% core and at least 95% matrix similarity in this fragment: Ets-1 at position -40, Myb at position -28 and SP1 at position -12. Both the Ets-1 and SP1 , strictly conserved between human and mouse sequences, site were recently reported to be functional PSEN1 promoter elements in SK-N-SH and HepG2 cells (54, 55). All larger fragments show lower promoter activity suggesting that both -3525Λ48 and +824/+1386 confer an overall inhibitory effect on PSEN1 promoter activity. 3' deletion of the +824/+ 1225 region increases promoter activity more than 4-fold, either due to the presence of negative regulators or to distance effects. 5' deletion analysis suggests the presence of negative regulatory elements in fragments -1218/— 515, - 320/-142 and -59/-48. The effect of these elements is most pronounced in N2A cells except for -59/-48, which probably contains either a HEK293 cell-specific inhibitor or a neuron-restrictive silencer element (NRSE; (56, 57)). Although we did not detect the 21 bp NRSE consensus sequence in the human PSEN1 upstream region, we can not rule out the presence of other elements with similar function. Matlnspector v2.2 indicated the presence of a high affinity binding site for NF-1 , member of a family of DNA binding proteins, some of which have been described to be involved in neuron- specific regulation of transcription (58). In the -320/-142 fragment we detected two binding sites for the δEF1 transcriptional repressor in tandem. Recently, it was shown that vertebrate members of the 5EF1 family need a bipartite element of the 6EF1 binding site CACCT for correct binding and functioning (54). Although the overall effect of the -3525/-48 region is negative, deletion mapping also suggests the presence of positive elements in fragments -514/-321 and -141/-59. Potential promoter elements in -141/-59 have no effect in N2A cells but deletion decreases promoter activity with 30% in HEK293 cells, suggesting the presence of a cell-type specific activator element. Pastorcic also reported no effect for human neuroblastoma cells but an even larger effect for HepG2 cells and demonstrated binding of an SP1 transcription factor in this region by DNase I footprinting (55). Matlnspector v2.2 detected besides this SP1 site, 2 potential binding sites for the zinc finger protein MZF1.

Although we detected the largest reduction in transcriptional activity when deleting -47/+15, promoter activity is not completely abolished. The remaining promoter activity can be due to transcription from the more downstream located TSSs t3, t4, t5 reported by Rogaev et al. (51 ). Another, more likely scenario, is that transcription initiates from exon 1 B driven by promoter elements located downstream of the major exon 1A TSSs. The latter hypothesis is favoured since transcriptional activity increases with deletion of the +16/+202 region, deleting all reported TSSs of the exon 1A transcripts. This effect is the most pronounced in HEK293 cells with a nearly 3-fold increase, hereby reaching expression levels corresponding to one third of the maximal promoter activity. In the N2A cells the increase is not that drastic but the transcriptional activity of the fragment is still higher than 50% of the SV40 promoter activity, also corresponding to one third of maximal promoter activity. Matlnspector v2.2 detected in the close proximity of exon 1B, the same combination of an Ets-1/2 binding site spanning the TSS and a number of GC-rich elements including an SP1 binding site, as was shown for exon 1A.

When removing fragment +204/+508, including the start of exon 1B, promoter activity drops to background levels. The present invention thus discloses that the region downstream of the major exon 1 A TSSs contains functional promoter elements, most probably driving transcription from exon 1B both in neuronal and non-neuronal cells. To study the effect of the promoter deletions on the alternative usage of the first exons and the activity of the exon 1 B promoter in the presence of the exon 1A promoter, we examined the occurrence of exon 1A and 1B transcripts in cells transfected with the PSEN1 promoter deletion constructs. Most of the PSEN1 transcripts found in brain or colon cDNA libraries contain exon 1A (51 ). We found by RT-PCR using exon-specific primers that both transcripts are present in lymphoblasts, brain, liver, testis and colon. Sequence analysis of the RT-PCR products showed that none of the transcripts contained both exon 1A and exon 1 B. These data show that the only exon 1 B transcript ever reported is not an artefact and that exon 1 B transcription is driven by its own promoter. Why exon 1 B transcripts are rarely found in cDNA libraries but are readily detected by RT-PCR is not clear. Our RT-PCR analysis was not quantitative, so we can not exclude that concentration differences between the two transcripts lead to differences in cloning. Another reason for the underrepresentation of exon 1B transcripts in cDNA libraries might be differences in efficiency of reverse transcription of exons 1A versus 1 B. Also the difference in length between the exon 1A and exon 1 B transcripts could lead to preferential cloning of the shorter 1 A transcript.

The overall transcription level is on average twice as high in N2A cells as compared to HEK293 cells. Northern blot analysis does not show a significant difference in mRNA levels between human brain and kidney (51 ). A recent analysis of the mouse PSEN1 promoter shows preferential transcriptional activity in neuron-like cells, supporting a cell-type-specific pattern of PSEN1 expression. To determine whether the observed difference in human PSEN1 expression levels is species- or tissue-specific we study PSEN1 promoter activity in human neuroblastoma cells.

We herein also report a neuron-specific 30% decrease in transcriptional activity of the PSEN1 promoter for the mutant -280G allele as compared to the wild type - 280C allele. These data suggest that -280C>G alters or creates a cis element important for transcriptional activity in neuron-like cells, but to a lesser extend in kidney cells. 5' deletion of the -320/-142 (-280C) fragment gives an increase in expression most pronounced in N2A cells. We herein also disclose that the two alleles of - 280C>G have different binding properties for specific nuclear factors. EMSA revealed that the nuclear extracts obtained from Kelly, HEK293 and N2A cells all contained nuclear proteins specifically recognising this region of the PSEN1 promoter resulting in two specific DNA-protein complexes A and B. Incubation of the wild type -280C allele with nuclear extracts leads predominantly to the formation of complex B. For the mutant -280G probe we demonstrated that a major part of the DNA is involved in the formation of an alternative complex A which migrates slower. We showed that the formation of complex B is competed more efficiently with mutant than wild type probe and vice versa for complex A. Since no third complex was detected, binding sites are most probably partially or completely overlapping, therefore proteins have to compete for binding and can never bind together. However, binding of these proteins does not seem to be exclusive for either of the alleles. Binding of specific nuclear factors to DNA is a dynamic process influenced by the binding affinities of the DNA sequence. Overlapping binding sites for different TF are known to be important for tissue-specific gene regulation (60-62).

Matlnspector v2.2 predicted the creation of a potential NF1 binding site (core similarity 1 / matrix similarity 0.872) for the -280G mutation. Competition with the NF-1 consensus sequence alters the ratio between the two complexes significantly. Since the predicted binding affinities for NF-1 were higher for -280G than for -280C we expected the upper complex A to be competed by the NF-1 consensus probe. However, the presence of unlabeled NF-1 probe decreases the formation of complex B in favour of the formation of complex A. Also, pre-incubation of the nuclear extracts with anti-NF-1 antibody gives rise to a supershift of the lower band, confirming the competition data. However, competition with mutant NF-1 has the same effect, suggesting that not NF-1 but another family member binds to the -291 /-268 fragment. Therefore, we are studying the competitive effect of other NF-1 binding sites found in vivo. Also, supershift experiments using labelled WT and Mt NF-1 consensus probes in stead of PSEN1 promoter fragments will give insight into the mobility and specificity of these NF-1 complexes. Important to note is that all three extracts contained proteins specifically recognising the PSEN1 promoter fragment. Although particular NF1 isoforms (NF1B and NF1 -related genes) appear to be enriched in the cerebellum and brain (64, 65), they are also expressed in many other tissues and cell lines. Indeed, NF1 proteins exhibit ubiquitous patterns of expression, and it is therefore unlikely that they play a primary role in determining cell- or tissue-specific transcription. However, NF1 proteins have been implicated in brain-specific gene expression (66, 67) and more recently in human neuroblastoma-specific expression of the human 3 nAchR subunit gene (58). The obvious lack of restricted expression for the NF1 gene family has led to the hypothesis that NF1 may either activate or silence gene expression in a cell-specific manner by participating in a combinatorial code involving cofactors, best exemplified by liver-specific vitellogenin gene expression (68). This same type of mechanism may account for cell type-specific expression of PSENL Alternatively, the nature of the binding to the PSEN1 NF1 site by these NFI-like proteins may be different and therefore the effect of -280C>G on transcriptional activity may differ between different cell types. The present invention demonstrates that the -48C>T polymorphism, located in the PSEN1 core promoter, even has a larger effect on PSEN1 promoter activity than -280C>G and provides evidence that -48C>T modifies transcriptional activity of the PSEN1 promoter in a cell-type specific manner. In mouse neuroblastoma cells, we detected a 2-fold decrease in transcriptional activity for the risk (C) allele as compared to the wild type (T) allele for both the -320/ +1225 and -59/+823 PSEN1 promoter fragments. In human kidney cells we were only able to detect this 2-fold drop in promoter activity for the small promoter fragment, while no difference in transcriptional activity could be detected for the -320/ +1225 fragment. The -59/+823 fragment containing -48T shows expression levels comparable to those seen for the most active promoter fragment -47A+823. Also, our 5' deletion analysis showed that deletion of the -59/-48(C) fragment significantly increases promoter activity. These data show that a strong negative regulator binds to the sequence surrounding position -48 with the highest affinity for the C-allele. This might be either a real inhibitor or a weak activator competing for a binding site with a stronger activator. In this view it is interesting that - 48 is located at the 5' end of the core promoter, shown to contain a functional Ets-1 like element at -38 (Pastorcic and Das, 1999). Since the effect of -48C>T in HEK293 cells is restricted to the small promoter fragment, we suggest the presence of a cell- type specific element in the -318/-59 or +824/+1225 region that masks the effect of - 48C>T in HEK293 cells in vivo. Therefore we can conclude that the -48C allele, associated with increased risk for EOAD, decreases PSEN1 promoter activity 2-fold in neuronal cells. Despite the major effect on transcriptional activity of PSEN1 no significant differences in complex mobility or binding affinity were detected between - 48C and -48T using the -43/-59 probe. In conclusion, we detected maximal PSEN1 promoter activity for the smallest fragment -47/+823 still spanning all exon 1A TSSs in both N2A and HEK293 cells and delineated the PSEN1 core promoter to the -47/+15 region. Also, we provided evidence for transcription initiation from exon 1 B driven by its own promoter and the presence of exonl B transcripts in different tissues. The -48C>T polymorphism, associated with EOAD, is located just upstream of the PSEN1 core promoter. We showed that the 12 bp fragment ending at -48 (-59/-48) contains a strong negative regulatory element involved in cell-type-specific inhibition of PSEN1 transcription. We demonstrated a twofold neuron-specific decrease for the -48C allele associated with EOAD using reporter gene analysis. Previously, it was shown that a 50% reduction of PSEN1 levels significantly increased Aβ42 production (70). Since most of our EOAD patients are homozygous for -48C, the corresponding 50% decrease in promoter activity will have a major effect on PSEN1 expression levels, possibly leading to increased Aβ42 production. Combination with other promoter variations, as there is the -280C>G mutation, might even lead to enhanced transcriptional effects. Moreover, since for this heterozygous EOAD-related mutation we demonstrated a neuron-specific 30% decrease. Despite the neuron-specific effects on transcriptional activity no cell- type specific band shifts were detected for either of these promoter variations. However, for the -280C>G mutations we showed differential binding of nuclear proteins to the alternative alleles, one of which is a member of the NF-1 family. Together, these studies show an important role for variability in regulatory elements in the genetic predisposition for developing AD. In this regard, the present invention relates to a method to screen for molecules which inhibit the reduction of presenilin-1 levels induced by a polymorphism in the presenilin-1 promoter region, in particular induce by the polymorphisms -48 C/T and/or -280 C/G, and/or regulatory region upstream of presenilin-1 comprising 1 ) exposing said polymorphism in the presenilin-1 promoter region or regulatory region upstream of presenilin-1 to said molecules, and 2) monitoring said preseniiin-1 levels. The present invention also relates to molecules obtainable by using a transgenic animal as described above and/or a method as described above. These methods are also referred to as 'drug screening assays' or 'bioassays' and typically include the step of screening a candidate/test compound or agent for the ability to interact with (e.g. bind to) the presenilin-1 promoter region and/or regulatory region upstream of presenilin-1 and/or factors interacting with said regions such as NF-1 as indicated above. Candidate compounds or agents, which have this ability, can be used as drugs to combat or prevent pathological conditions of AD. Candidate/test compounds such as small molecules, e.g. small organic molecules, and other drug candidates can be obtained, for example, from combinatorial and natural product libraries as described above. Typically, the assays are cell-free assays which include the steps of combining the presenilin-1 promoter region and/or regulatory region upstream of presenilin-1 and/or factors interacting with said regions such as NF-1 and a candidate/test compound, e.g., under conditions which allow for interaction of (e.g. binding of) the candidate/test compound with the presenilin-1 promoter region and/or regulatory region upstream of presenilin-1 and/or factors interacting with said regions such as NF-1 to form a complex, and detecting the formation of a complex, in which the ability of the candidate compound to interact with ) the presenilin-1 promoter region and/or regulatory region upstream of presenilin-1 and/or factors interacting with said regions such as NF-1 is indicated by the presence of the candidate compound in the complex. Formation of complexes between the presenilin-1 promoter region and/or regulatory region upstream of presenilin-1 and/or factors interacting with said regions such as NF-1 and the candidate compound can be quantitated, for example, using standard immunoassays. The presenilin-1 promoter region and/or regulatory region upstream of presenilin-1 and/or factors interacting with said regions such as NF-1 employed in such a test may be free in solution, affixed to a solid support, borne on a cell surface, or located intra-cellularly. To perform the above described drug screening assays, it is feasible to immobilize the presenilin-1 promoter region and/or regulatory region upstream of presenilin-1 and/or factors interacting with said regions such as NF-1 or its (their) target molecule(s) to facilitate separation of complexes from uncomplexed forms, as well as to accommodate automation of the assay. Interaction (e.g., binding of) of presenilin-1 promoter region and/or regulatory region upstream of presenilin-1 and/or factors interacting with said regions such as NF-1 presenilin-1 promoter region and/or regulatory region upstream of presenilin-1 and/or factors interacting with said regions such as NF-1 to a target molecule, can be accomplished in any vessel suitable for containing the reactants. Examples of such vessels include microtiter plates, test tubes, and microcentrifuge tubes. For sake of clarity the current invention is further described and explained by way of the following non-limiting disclosure like materials and methods used and the results obtained thereof.

Examples

1. Association of polymorphisms with AD - Patients were derived from a population-based epidemiological study of early-onset AD (27). Within two areas of the Netherlands, the study aimed at a complete ascertainment of all AD patients in whom the disease onset was at or before the age of 65 years. Age at onset was defined as the age at which memory loss or changes in behaviour were first noted. For this study, the clinical diagnosis of AD was independently confirmed by two neurologists using a standardised protocol according to the NINCDS-ADRDA criteria for AD. Of the 201 eligible patients, 198 (99%) participated in the study. Family history of dementia addressed all first, second and third degree relatives of the patients. Patients who were known to be related were excluded from the association studies presented here. Blood samples for DNA extraction were collected of 102 (52%) of the participating patients. The mean age at onset of the patients was 56.7 ± 5.4 years and the mean attained age at the time of the study was 63 ± 4.4 years. Patients were compared to an age-matched control series (n=118; mean age 63 ± 4.4 years) that was drawn randomly from the Rotterdam Study (27, 28). Based on family history (up to 2 degrees), these subjects were not related. None of the control subjects showed symptoms of dementia and none had cognitive test scores suspect for dementia (27,28). Twelve patients and ten controls were random selected from these samples for the initial screening for polymorphisms of the PSEN1 5' upstream region.

-DNA analysis: a polymerase chain reaction (PCR) was performed using 200 ng genomic DNA as template in a 25 μl reaction mixture containing 20 pmoi of each PCR primer, 0.2 units (U) Taq DNA polymerase (Gibco BRL, Gaithersburg, USA), 1.0 mM MgCI₂, 75 mM Tris-HCI pH9.0, 20 mM (NH₄)₂S0₄, and 0.01 % Tween-20. The PCR amplification consisted of 30 cycles of 90 sec at 94°C, 90 sec at the empirically defined optimal annealing temperature, and 90 sec at 72°C. The STR polymorphisms at D1S479, D14S1028, D14S77, D14S1004, D14S43 and D14S61 were PCR amplified using published primers of which one was fluorescently labelled. The alleles were separated on a 6% polyacrylamide gel containing 8M urea using an ABI373A automated DNA sequencer (Applied Biosystems, Foster City, USA) and analysed using the GeneScan 672 software (Applied Biosystems). Alleles were numbered as in the CEPH Genotype Database. When additional alleles were observed, new numbers were assigned (Table 2). The Bst \ polymorphism 0.9 kb downstream of the c-fos oncogene (FOS)(29), the BsfXi polymorphism in exon 1 B of PSEN1 (5'UTR), and the Ddel, BbtP\ and PvuW polymorphisms in exons 3 and 4 and intron 11 of PSEN2 (8) were detected by restriction fragment length polymorphism (RFLP) analysis of the respective PCR amplified exons. PCR amplification was performed using published primers (8, 29) and the products were digested overnight with 5 U of the appropriate restriction enzyme. Also, PCR primers 3UTR7 (5'-AAACAGTACAGCTATTTCTCATC- A-3') and 3UTR8 (5'-GCTTCAACAGCCATTTTACTC-3') were developed to amplify an A to T polymorphism at nucleotide 952 in the 3'UTR of PSEN1 that recently has been identified. The amplification products were digested with 5 U Λ//alll resulting in a constant fragment of 37 bp and 2 alleles of 308 bp (T-allele) and 162 + 146 bp (A-allele) respectively. The alleles were separated by agarose gel electrophoresis and visualised on a UV transiliuminator after ethidium bromide staining. For each PCR- RFLP, the longer restriction fragment was assigned allele number 1 , while the shorter one was assigned allele number 2. Also, the polymorphism in intron 8 of PSEN1 was analysed by PCR-RFLP analysis using a mismatch primer and BamHI digestion as described (11 ). Allele numbering was in accordance with the original publication (11 ). The polymorphism at codon Gly366 in exon 14 of the dihydrolipoyl succinyltransferase (DLST) gene (30) and the PSEN1 promoter polymorphism (8) were scored by single-strand conformational polymorphism (SSCP) analysis. PCR was performed using published primers (8,30). PCR amplification products of DLST exon 14 were loaded on a 1x HydroLink MDE gel (J.T. Baker, Phillipsburg, USA) and separated by electrophoresis for 20 h at 800 V and at room temperature. The promoter polymorphism of PSEN1 was analysed on precast ExcelGel gels (Pharmacia Biotech, Uppsala, Sweden), electrophoresis was for 3.5 h at 600 V using the MultiPhorll electrophoresis system (Pharmacia Biotech). After electrophoresis, the SSCP bands were visualised by silver staining. The allele corresponding to the higher band in the SSCP analysis was assigned number 1 , while the lower allele was assigned number -Statistical analysis: the significance of the association between early-onset AD and the polymorphisms studied was assessed using the likelihood ratio test or the Fisher exact test when appropriate. For di-allelic markers, genotype and allele frequencies were compared between cases and controls. For the STR markers, comparisons were restricted to allele frequencies because of the large number of (rare) genotypes. When the overall likelihood ratio test was significant, the strength of association between the alleles or genotypes and early-onset AD was estimated with the odds ratio (OR) presented with 95% confidence intervals (CI). Linkage disequilibrium was tested using the EH program as described by Terwilleger and Ott (31 ). Genotype and allele frequencies were in Hardy-Weinberg equilibrium. Stratified analyses were conducted based on the presence of known PSEN1 mutations (8) or an APOE ε4 allele (10).

- Transposon-based and direct sequencing of the PSEN1 5' upstream region: Plasmid B22 was subcloned into the appropriate restriction sites of the pOCUS-2 vector (Novagen Inc., Madison, WI). The pOCUS-2 constructs were transferred into chemically competent γδ donor cells, which carry the γδtransposon on an F factor, and one of the resulting colonies was mated with the γδ recipient cells according to the manufacturers protocol. For each subclone 96 resulting colonies were randomly selected and stored in a 96 well plate containing LB medium with 20% glycerol. The transposition site of the different clones was mapped by colony PCR combining one of the two vector-specific primers (POCUSUP or POCUSDOWN) with a phosphorylated transposon-specific primer (GDIR) in separate reactions (Strathmann et al., 1991 ). Colonies were selected based on the length of their PCR products and DNA was prepared for sequencing using the Wizard Plus SV Minipreps DNA Purification System. Plasmid sequencing was accomplished using the Thermo Sequenase™ II Dye Terminator Cycle Sequencing Kit (Amersham Life Science, Cleveland, USA) and the supplier's protocol using the published primers GD1 and GD2 (Strathmann et al., 1991 ). The sequences were assembled using the Lasergene software for Windows (DNASTAR inc., Madison, USA). To close the remaining gaps, primers flanking the gaps were designed with the Lasergene software and used in cycle sequencing to complete the sequence.

-Sequence analysis of the PSEN1 5' upstream region: results

Four fragments of plasmid B22, ranging from 2 to 3.2 kb, were subcloned in the appropriate restriction sites of the pOCUS-2 vector based on the detailed restriction map of plasmid B22 (Theuns et al., 1999). The first 3 fragments did not overlap, while the third and the fourth overlap with approximately 350bp. Following the subcloning and mating procedure 96 clones were randomly selected and their transposition site was determined by colony PCR. Based on the lengths of their PCR products at least 5 clones per kb were selected for sequencing. In regions that were difficult to sequence, a second series of clones was selected. Remaining gaps were closed by direct sequencing using primers flanking the gaps. Direct sequencing was also used for linking the first three fragments. In total we sequenced 9643 bp, with 6698 bp upstream of exon 1B with an average coverage higher than 4. 83% of this sequence was sequenced at least twice and on both strands, 13% at least twice but only on one strand, and only 4% was sequenced only once. In the latter case however, the sequencing results were optimal. The percentage of the bases Guanine and Cytosine in the total sequence lies within the normal range. The GC content does not change in the 6.7 kb fragment upstream of exon 1 B nor in the 6.4 kb upstream of exon 1 A. In the fragment upstream of exon 1A it starts rising above 55% however, from 500 bp upstream of exon 1A towards 73% in the last 100 bp and with a maximum of 100% in the 10 bp fragment right upstream of exon 1A. Intron 1A still has a very high GC content of 74 % while in exon 1 B a decrease of the GC content towards intron 1 B is detected. In the region starting from 200 bp upstream of exon 1 A the GpC versus CpG ratio is shifted from the normal 5:1 ratio towards a 1 :1 ratio, giving rise to clusters or islands of CpG dinucleotides. This 1 :1 ratio is maintained through exon 1A and intron 1A, while in exon 1B the CpG content decreases again. Using the Matlnspector V2.2 program (Quandt et al., 1995) we identified in the 200 bp region upstream of exon 1A a number of potential transcription factor binding sites with a core similarity of 1. Four potential binding sites with a matrix similarity higher than 0.95 were detected in the first 100 bp upstream of exon 1A: MZF1 (0.982), NF1 (0.958), TCF11 (0.967), VMYB (0.966). Nine binding sites were detected within the next 100bp: twice DELTAEF1 (0.981/0.961 ), twice MYOD (0.982/0.987), twice MZF1 (0.996/0.975) and once a GC box (0.994), SP1 (0.999) and LMO2COM (0.977) site. For the exon 1B 200 bp upstream region no transcription factor binding sites were found with a core similarity of 1 and matrix similarity higher than 0.95. Three binding sites were found in the first 100 bp with a matrix similarity higher than 0.90: IK2 (0.918/0.925), AP1 (0.930) and AP4 (0.905). One site was detected in the next 100 bp, CEBPB (0.918). The frequency of Alu repeats was found to be relative high in the 9.6 kb sequence. Several repeats were found covering more then 2 kb in total, which is about 22 % of the complete sequence, with hot spots at 1 or 2 kb upstream of exon 1A and in intron 1 B. Shortly after we finished the sequence of the upstream region of PSEN1 , Rowen and his co-workers published the complete PSEN1 sequence in Genbank (AF109907). Comparison of our sequence (Neurogen, N) with the sequences available in Genbank revealed a small number of variations (Table 5). Five sequence variations were found in the region upstream of exon 1 A, four downstream of exon 1 B. -DNA analysis:results

Overlapping primer sets were designed based on the 7kb sequence upstream of exon 1B using the Lasergene software. A polymerase chain reaction (PCR) was performed using 200 ng genomic DNA as template in a 25 I reaction mixture containing 25 pmol of each PCR primer, 1 unit (U) Taq DNA polymerase (Gibco BRL, Gaithersburg, USA), 0.2 mM of each dNTP (Pharmacia Biotech, Uppsala, Sweden), 0.05 % W1 and 1x Taq buffer (GIBCO BRL). In order for some primers to work, the PCR Optimizer™ Kit (Invitrogen Corporation, CA, USA) was used. Deviating buffer conditions, additives and the MgCI₂ conditions are described in table 2. The PCR amplification consisted of 30 cycles of 60 sec at 94°C, 90 sec at the empirically defined optimal annealing temperature, and 120 sec at 72°C (table 5). In the single-strand conformational polymorphism (SSCP) analyses overlapping primer sets were used to amplify fragments of the PSEN1 5' upstream region. PCR products were analysed on precast ExcelGel gels (Pharmacia Biotech), electrophoresis was for 3.5 h at 600 V using the MultiPhorii electrophoresis system (Pharmacia Biotech). After electrophoresis, the SSCP bands were visualised by silver staining. When aberrant SSCP patterns were observed the sequences of the fragments were determined using cycle sequencing. The PCR amplification products were pre-treated with 10 U exonuclease I and 2 U shrimp alkaline phosphatase to remove excess PCR primers and nucleotides. PCR amplification product (5 μl) was used as template in the cycle sequencing reaction using the ABI PRISM Dye Terminator Cycle Sequencing Core Kit according to the supplier's protocol using the same primers as in the PCR amplification. Comparison of the sequences was done using the Lasergene software. Sequence variations in the PSEN1 5' upstream region were analysed by restriction enzyme digestion of the amplified products when they involved the creation or abolition of a restriction enzyme recognition site (Restriction fragment length polymorphism, RFLP). If no suitable restriction enzyme was found, mismatch primers were developed to amplify and detect the polymorphism. Genomic PCR amplification products were digested over night (ON) using 10 U of the corresponding restriction enzyme (table 6) at the appropriate reaction temperature. The restriction fragments were separated on a 1.5-3% agarose gel, depending on the allele sizes, and visualised on a UV trans-illuminator after EtBr staining. Finally the allelic distribution of the different polymorphisms was compared with the distribution of the alleles of the exon 1A polymorphism reported above. At first, overlapping primer sets were designed to screen the 3.3 kb sequence upstream of and including exon 1A, for polymorphisms (Table 6). PCR-SSCP analysis was performed on a selection of 12 patients and 10 controls of the Rotterdam EOAD sample (van Duijn et al., 1994). Altered SSCP patterns were observed for 5 primer sets (Table 7). Cycle sequencing of the different patterns revealed 7 sequence variations in total (Table 7). The different patterns found for primer set Prom2 could be explained by the C/T variation at 48 bp upstream of exon 1A (-48) reported earlier (Cruts et al., 1998b). For primer set Proml only patient 1093-000 showed an altered SSCP pattern. This patient has an onset-age of 56 years and a familial background. Cycle sequencing revealed a C to G transition at 280 bp upstream of exon 1A (-280), creating an Ncol site. The analysis of this sequence variation was extended to the complete Rotterdam EOAD sample (van Duijn et al., 1994) using SSCP analysis. The SSCP pattern of primer set Proml 3 was extremely complex and five samples, giving different patterns, were sequenced. The complex SSCP patterns were due to two different sequence variations present on the fragment amplified by the respective primer set. The first sequence variation was a G/A variation at position -2154 bp, which can be detected by BsmBI digestion; when A is present the site disappears. This polymorphism is in complete linkage disequilibrium with the exon 1A polymorphism in the samples examined. G is the most frequent allele and is associated with a higher risk for EOAD. This polymorphism is studied in more detail in the remaining patients and controls of the Rotterdam EOAD sample. The second variation is a variable T-stretch at position -2319, which has only been examined in the five samples that have been sequenced. Thus far, four different alleles have been found, ranging from 12 to 15 T's. All samples examined are heterozygous with a combination of two subsequent alleles e.g. 12 and 13 T's. The fourth primer set, Proml 5, showed four different SSCP patterns in the selection of 22 samples. Cycle sequencing revealed two different sequence variations in the fragment amplified with the respective primers. The first variation is a 13 bp insert or deletion of the sequence 5'-GCATGTCCTGGGT-3' at position -2823. This polymorphism was further examined in all 102 EOAD patients and 118 controls by SSCP and was in complete linkage disequilibrium with the exon 1A polymorphism except for three control samples and one patient. Therefore the exonlA polymorphism was re-examined using Hgal and the insert/deletion polymorphism using Nlalll RFLP analysis. It turned out that the exon 1A SSCP analyses of two out of three control samples were wrongly interpreted and that the RFLP results affirmed the linkage disequilibrium between these 2 polymorphisms. The difference observed for patient 1092-000 is due to a second variation present in the Proml 5 fragment, an A to G transition located at position -2818. This variation is only present in this patient and the rest of the 22 samples examined are homozygous for the A allele. This variation can be detected by Banl digestion since the restriction site is present when A is substituted by G. The RFLP analysis is extended to the Rotterdam EOAD sample series (van Duijn et al., 1994). The 3.2 kb upstream of the former fragment was screened in a more direct way by designing primer sets covering the regions containing the two variations between our sequence and AF109907. -Eukaryotic cell culture and transient transfection Mouse Neuro2a-neuroblastoma (N2a) cells were routinely propagated in a minimal essential medium with Earie's salt (Life technologies), 10% foetal bovine serum (Life technologies), 2mM L-glutamine (Life technologies), 200IU/ml Penicillin, 200μg/ml Streptomycin (Life technologies) and 0.1 mM nonessential amino acids (Life technologies). Human embryonic kidney cells (HEK293) were routinely propagated in Optimem (Life technologies) with 10% foetal bovine serum (Life technologies), 200 lU/ml Penicillin and 200 μg/ml Streptomycin (Life technologies). For transient transfection N2a and HEK293 cells were seeded in 6-well tissue culture dishes, at 9 X 10⁴ and 7 X 10⁵ cells/well respectively, and allowed to recover for 24 hours. Cells were co-transfected with 1 μg of one of the PSEN1 promoter constructs, pGL3 basic vector or pGL3 promoter plasmid (containing an SV40 promoter upstream of the firefly luciferase gene, Promega) and 20 ng of pRL-TK plasmid (which contains a herpes simplex virus thymidine kinase promoter upstream of the renilla luciferase gene, Promega), using the Lipofectamine procedure (Life technologies) as described in the manufacturer's protocol. - Relative luciferase activity measures Transfected cells were cultured for 48 hours, washed with 1 ml phosphate-buffered saline (PBS, Life Technologies), and lysed with Passive lysis buffer (Promega). Firefly luciferase activities (LA_F) and renilla luciferase activities (LA_R) were measured sequentially using a Dual-Luciferase reporter assay system (Promega) and a model TD-20E Luminometer (Turner design). To correct for transfection efficiency and DNA uptake the relative luciferase activity (RLA) was calculated as: RLA= LA_F/LA_R. To compare the RLA in one cell line with another, relative RLA was calculated as a percentage of the RLA of the SV40 promoter plasmid: %RLA = (%RLA7RLA_SV40)^*100 or as a percentage of the promoter construct with the most frequent alleles (KHCCG): %RLA=(RLA/RLA_k)*100.

-Effect of the 5' upstream region polymorphisms on the transcriptional activity of the PSEN1 promoter Five different promoter reporter plasmids were constructed in the pGL3-basic vector. The first one was a Kpnl-H/ndlll subclone of the PSEN1 promoter plasmid B22. This clone contains the 5' upstream region of PSEN1 from -323 to +1231 and covers the polymorphic sites at positions -280, -48 and +522 (exon 1B polymorphism, reported by Cruts et al., 1998)). Sequencing of the insert revealed a C at position -280, a C at -48 and G at +522 (KHCCG) which is in agreement with our former sequencing results. In each of the remaining constructs one single nucleotide had been altered respectively at position -280 where C was replaced by G (-280 G), at -48 where C was replaced by T (-48 T) and at +522 where G was replaced by A (+ 522 A) or C (+ 522 C). The effect of the different alleles on the transcription activity of the PSEN1 promoter was studied in vitro in a transient transfection expression system. For each construct the transfection experiments were repeated three times, using three independent DNA preparations in each experiment. All luciferase assays were performed in duplo for each sample. In human embryonic kidney (HEK293) cells no significant differences in expression levels were detected for any of the polymorphisms studied (Figure 2). In mouse neuroblastoma (N2A) cells however, analysis of the -280 G construct revealed a 30% decrease in promoter activity compared to the 'wild type' C allele (Figure 2). No significant differences were detected for the other polymorphisms. Interestingly, the overall promoter activity in N2Acells is not much higher than the activity in HEK293 cells. 2. Functionality of the -280 C/G and -_^48 C/T variants in relation to AD

- Deletion analysis of the PSEN1 upstream region

Starting from plasmid B22, we constructed a deletion panel of the PSEN1 upstream region from covering nucleotides -3525 to +1386 (numbering according to the exon 1A transcription start site (TSS) t2 reported by Rogaev (L76518) (51 ). Plasmid B22, containing exons 1A and AB of PSEN1 and approximately 6.5 kb of upstream sequences, was previously restriction mapped (52) and sequenced (AF205592) (50). Genomic fragments obtained from plasmid B22 by restriction digestion or PCR amplification, were subcloned upstream of the firefly luciferase gene in the promoteriess pGL3-basic vector. Transcriptional activity was analysed in transiently transfected HEK293 and N2A cells. To correct for transfection efficiency and DNA uptake, we co-transfected the promoter constructs with plasmid pRL-TK, containing a Herpes Simplex virus thymidine kinase promoter driving Renilla luciferase. The relative luciferase activity (RLA) of the deletion constructs was calculated as the ratio of firefly luciferase (LA_F) to renilla luciferase activity (LA_R) (Figure 3). We defined the RLA of the SV40 promoter construct, which serves as a positive control vector, as 100% activity. All other transcriptional activities are presented as a percentage of the SV40 promoter construct. The overall transcription level was about twice as high in N2A cells as compared to HEK293 cells, however relative changes in activity between the PSEN1 promoter constructs were in general similar for most deletions (Figure 3). The smallest deletion fragment (-47/+823) still containing the most upstream TSS (54), showed the highest promoter activity in both cell types, exceeding the SV40 promoter activity in N2A cells with 50%. All larger fragments showed lower promoter activity suggesting that both the -3525/-48 and the +824/+1386 region contain predominantly negative regulatory elements. In both cell types, deletion of the +824/+1225 region increased promoter activity more than 4-fold. Deletion of the -3525/-1219 region had no effect on promoter activity. Deletion of -1218/-515 increased transcriptional activity 3-fold in N2A and 2-fold in HEK293 cells. However, deleting -514/-321 decreased promoter activity 2-fold in N2A and with 30% in HEK293 cells. Further deletion of the -320/-142 sequence increased activity again, 3-fold in N2A and nearly 2-fold in HEK293. Deletion of -141/-59 had no effect on promoter activity in N2A cells but decreased expression significantly in HEK293 cells. Deletion of the -59/-48 fragment increased promoter activity more than 3-fold in HEK293 cells and with 50% in N2A cells. Deletion of - 47/+15 resulted in a 3-fold reduction in N2A and a 7-fold in HEK293 cells. However, promoter activity was not completely abolished and further deletion of +16/+202 resulted in increased transcriptional activity. In the HEK293 cells this effect was the most pronounced with a nearly 3-foid increase resulting in one third of maximal promoter activity in these cells. In the N2A cells the increase was not that drastic but the transcriptional activity of the fragment is still higher than 50% of the SV40 promoter activity. Deleting +204/+508, the start of exon 1 B inclusive, promoter activity dropped to background levels again.

- Transcriptional effect of PSEN1 promoter variations

We introduced the T allele of -48C>T in the 1545 bp Kpn\-Hind\\\ fragment of B22, corresponding to the -320 to +1225 region, by in vitro mutagenesis and sequence integrity was confirmed by sequencing. In transiently transfected human embryonic kidney (HEK293) ceils no significant differences in expression levels were detected between -48T and -48C (Figure 4). In mouse neuroblastoma (N2A) cells however, a two-fold decrease was detected for -48C (Figure 4). We also introduced the -48T allele in the 882 bp construct -59A+823 by PCR amplification using an upper primer containing the T allele. Two-fold reduction in promoter activity for the C allele was detected in both transiently transfected N2A and HEK293 cells. -Electrophoretic mobility shift assays

We examined whether -280C>G and -48C>T affect the specific recognition of the PSEN1 promoter by nuclear factors (Figure 5) using electrophoretic mobility shift assays (EMSA). For -280C>G we used a double-stranded oligonucleotide probe spanning the region -291/-268 containing either the wild type (-280G) or mutant (- 280C) allele (Figure 5). All three nuclear extracts, obtained from human neuroblastoma (Kelly), human embryonic kidney (HEK293) and mouse neuroblastoma (N2A) cells, contained nuclear proteins specifically recognising this region of the PSEN1 promoter, resulting in two specific DNA-protein complexes. Complex B showed a higher mobility resulting in the lower band (B) on the gel in figure 5. Complex A migrated slower in the gel, resulting in the upper band (A) (Figure 5). Both complexes were present for both -280C and -280G alleles, however, the ratio of the amount of the 2 complexes differs significantly. The amount of complex B formed on the -280C probe is higher than the amount of complex A (intensity band B>A, Figure 5 lane 1 ), whereas a significant shift towards formation of complex A was observed for -280G (intensity band A>B, Figure 5 Iane7). Moreover, DNA binding of the protein(s) of complex A is competed more efficiently by, and at lower concentrations of unlabeled probe -280G (Figure 5 lanes 10-11 ) as compared to the -280C probe (Figure 5a lanes 8-9). The increased formation of complex A on -280G was even enhanced when the NF-1 consensus sequence was used as competitor (Figure 5a lanes 12-13). A similar though less pronounced effect was observed for competition with mutant NF- 1 probe (Figure 5a lanes 14-15). Incubation of the nuclear extracts with an antibody with broad specificity for NF-1 family members, gave rise to a supershift in mobility of complex B (Figure 5a lanes 6 and 16).

For -48C>T we started with double-stranded oligonucleotide probes spanning nucleotides -59/-43. All three nuclear extracts contained nuclear proteins binding specifically to this region of the PSEN1 promoter. However, no significant differences in complex mobility or binding affinity were detected between -48C and -48T using the -59/-43 probe. -Materials and methods

*PSEN1 promoter luciferase reporter constructs Genomic DNA fragments containing parts of the putative PSEN1 promoter were subcloned upstream of the firefly luciferase gene of the promoteriess pGL3-basic vector (Promega, Madison, WI, USA). Plasmid B22 DNA was digested with different combinations of restriction enzymes and fragments were separated on an agarose gel. The target fragments were extracted from gel and ligated into the corresponding restriction sites of the vector: the 4911 bp Bg/ll/BamHI fragment spanning nucleotides -3525/+1386, the 2443 bp H/noflll fragment spanning -1218/+1225, the 1739 bp Nnel/H/ndlll fragment spanning -514/+1225, the 1545 bp Kpnl/H/nαlll fragment spanning -320/+1225, the 1366 bp Sacl/H/ndlll fragment spanning -141/+1225 and the 625 bp Xnol/H/ndlll fragment spanning +603/+1225 (numbering according to the exon 1A transcription start site (TSS) t2 reported by Rogaev et al. (L76518) (51 ). Later this deletion panel was refined with PCR-based deletion clones (Figure 3). Genomic fragments were obtained by PCR amplification of plasmid B22 DNA (52) using the proof-reading Fu polymerase (Promega) and primers designed to our published sequence of the human PSEN1 upstream region (AF205592) (50). Primers were designed to incorporate restriction enzyme sites at the end of the amplified products to facilitate subcloning of the fragments in the pGL3-basic vector. The Quick-change in vitro mutagenesis kit (Stratagene, La Jolla, CA, USA) was used to introduce the T allele at position -48 in the 1545 bp Kpnl/H/^'ncflll fragment using primer 5'-gtgggccggccgccaacgaTgccagagccggaaatgacg-3' and its reverse complement. The upper primer designed to amplify the 882 bp -59/+823 fragment spans position -48, the T allele was introduced in this fragment using an upper primer corresponding to the T allele. -48T clones were selected by Hgal digestion. The integrity of all inserts was confirmed by sequence analysis using the Thermo Sequenase™ II Dye Terminator Cycle Sequencing Kit (Amersham Pharmacia Biotech inc., NJ, USA) using vector- and PSEΛ/7-specific primers designed for screening of the PSEN1 5' upstream region (50). Sequences were assembled using the Lasergene software (DNASTAR inc., Madison, USA). ^*Eukaryotic cell culture and transient transfection Mouse Neuro2a-neuroblastoma (N2a) cells were propagated in a minimal essential medium with Earie's salt (Life Technologies, Gaithersburg, USA), 10% foetal bovine serum (Life technologies), 2mM L-glutamine (Life technologies), 200 lU/ml penicillin, 200 g/ml Streptomycin (Life technologies) and 0.1 mM nonessential amino acids (Life technologies). Human embryonic kidney cells (HEK293) were propagated in Optimem (Life technologies) with 10% foetal bovine serum (Life technologies), 200 lU/ml penicillin and 200 g/ml Streptomycin (Life technologies). For transient transfection N2a and HEK293 cells were seeded in 6-well tissue culture dishes, at 9 X 10⁴ and 7 X 10⁵ cells/well respectively, and allowed to recover for 24 hours. Cells were co-transfected with 20 ng of pRL-TK plasmid containing the herpes simplex virus thymidine kinase promoter upstream of the renilla luciferase gene (Promega) and 1 μg of either one of the PSEN1 promoter constructs or one of the control plasmids, using the Lipofectamine procedure (Life technologies) as described in the manufacturer's protocol. Empty pGL3-basic vector was used as a negative control, pGL3-promoter plasmid containing the SV40 promoter upstream of the firefly luciferase gene (Promega) as a positive control. "Luciferase activity

Transfected cells were cultured for 48 hours, washed with 1ml phosphate-buffered saline (PBS, Life Technologies), and lysed with Passive lysis buffer (Promega). Firefly luciferase activities (LA_F) and renilla luciferase activities (LA_R) were measured sequentially using a Dual-Luciferase reporter assay system (Promega) and a model TD-20E Luminometer (Turner design). To correct for transfection efficiency and DNA uptake, the relative luciferase activity (RLA) was calculated as RLA = LA_F/LA_R. To compare the RLA in one cell line with another, relative RLA was calculated as a percentage of the RLA of the SV40 promoter construct %RLA = (RLA/RLA_SV40)^*100 or the wild type construct %RLA = (RLA_mt/RLA ^*100. ^*Preparation of nuclear extracts

HEK293, N2A and Kelly (human neuroblastoma cells) were grown under normal growth conditions to a density of 0.5-1 ^*10⁶ cells/ml. Approximately 10⁹ cells were harvested in PBS (Life Technologies), washed twice in PBS, pelleted by centrifugation at 950 x g and 4°C for 15 min, frozen in liquid nitrogen and stored at -80°C. Nuclear extracts were prepared according to a modified Dignam et al. (14) procedure. Cells were thawed, placed on ice and lysed with a Dounce homogenizer, after 15 min of swelling in 5 packed cell volumes (pcv) of hypotonic buffer H containing 20mM Hepes pH 7.9, 1.5 mM MgCI₂, 10mM KCI and freshly added 1 mM DTT and 1 mM PMSF. Nuclei were collected by centrifugation at 3000 x g for 15 min and the nuclear pellet was resuspended in 5 pcv of buffer D containing 20 mM Hepes pH 7.9, 1.5 mM MgCI₂, 0.2 mM EDTA, 20% glycerol, 0.42 M KCI and freshly added 1 mM DTT and 1 mM PMSF. The suspension was stirred for 30 min at 4°C and insoluble debris were pelleted by centrifugation at 18000 x g for 30 min at 4°C. The supernatant crude nuclear extract was concentrated by ammonium sulphate precipitation. The resulting pellet was gently resuspended in 5ml HGED.1 containing 50 mM Hepes pH 7.9, 1 mM EDTA, 1 mM DTT, 20% glycerol and 0.1 M KCL, and dialysed overnight at 4°C in the same buffer. Nuclear extracts were divided in aliquots and stored at -80°C. The protein concentration of the extracts was determined using a Bradford assay (BIO- RAD, Munich, Germany). *Electrophoretic mobility shift assays

Single strand oligonucleotides were designed spanning either the PSEN1 proximal promoter sequence variations or the consensus sequence of the NF-1 binding site: - 48C, 5'-gccgccaacgaCgccagagccgga-3'; -48T, 5'-gccgccaacgaTgccagagccgga-3'; -280C, 5'-aggatggccatCgcttgtatgccg-3'; -280G, 5'-aggatggccatGgcttgtatgccg-3'; NF-1 , 5'-ttttggattgaagccaatatgataa-3', mNF-1 , 5'-ttttggattgaaTAAaatatgataa-3'. Blunt-ended double stranded probes were obtained by annealing of the oligonucleotides with their respective reverse complements and were labelled with [γ-³²P]dATP and T4 polynucleotide kinase (Life Technologies). All probes were purified on a non- denaturing 15% polyacrylamide gel in 0.5 x TBE. For the binding reactions 0.1-0.5 ng ³²P-labeled double stranded probe (10⁴ Cerenkov counts) was added to a total reaction volume of 20 μl containing 2-10 μg nuclear extract, 1x Hepes binding buffer [12% glycerol, 20 mM HEPES or TRIS, 50 mM KCI, 1 mM EDTA, 1 mM DTT, 1 mM PMSF] and 0.5 μg poly (dl-dC) (Boehringer-Mannheim, Germany). For competition assays unlabelled double stranded probes were added to the reaction mixture prior to addition of the labelled probe. In the supershift experiments the extracts were 20 min pre-incubated on ice with rabbit polyclonal anti-NF-1 (SC-87X; Santa Cruz inc., CA, USA). The binding reactions were incubated at room temperature for 15-35 min. Protein-DNA complexes were analysed by electrophoresis on non-denaturing 5% polyacrylamide gels in 0.5 X TBE and visualised by autoradiography.

References

1. Van Broeckhoven C. Molecular genetics of Alzheimer disease: Identification of genes and gene mutations. Eur.Neurol. 1995; 35: 8-19.

2. Goate A, Chartier-Harlin M-C, Mullan M, et al. Segregation of a missense mutation in the amyloid precursor protein gene with familial Alzheimer's disease. Nature 1991 ; 349: 704-706.

3. Hardy J. Amyloid, the presenilins and Alzheimer disease. TINS 1997; 20: 154-159.

4. Sherrington R, Rogaev El, Liang Y, et al. Cloning of a gene bearing mis-sense mutations in early-onset familial Alzheimer's disease. Nature 1995; 375: 754-760.

5. Levy-Lahad E, Wasco W, Poorkaj P, et al. Candidate gene for the chromosome 1 familial Alzheimer's disease locus. Science 1995; 269: 973-977.

6. Rogaev El, Sherrington R, Rogaeva EA, et al. Familial Alzheimer's disease in kindreds with missense mutations in a gene on chromosome 1 related to the Alzheimer's disease type 3 gene. Nature 1995; 376: 775-778. 7. Cruts M, Van Broeckhoven C. Presenilin mutations in Alzheimer's disease. Human Mutation 1998; 11 : 183-190.

8. Cruts M, van Duijn CM, Backhovens H, et al. Estimation of the genetic contribution of presenilin-1 and -2 mutations in a population-based study of presenile Alzheimer disease. Hum. Mol. Genet. 1998; 7: 43-51.

9. Strittmatter WJ, Saunders AM, Schmechel D, et al. Apolipoprotein E: High-avidity binding to -amyloid and increased frequency of type 4 allele in late-onset familial Alzheimer disease. Proc.Natl.Acad.Sci.USA 1993; 90: 1977-1981.

10. van Duijn CM, de Knijff P, Cruts M, et al. Apolipoprotein E4 allele in a population- based study of early-onset Alzheimer's disease. Nat.Genet. 1994; 7: 74-78.

H .Wragg M, Hutton M, Talbot C, et al. Genetic association between intronic polymorphism in presenilin-1 gene and late-onset Alzheimer's disease. The Lancet 1996; 347: 509-512.

12.Higuchi S, Muramatsu T, Matsushita S, Arai H, Sasaki H. Presenilin-1 polymorphism and Alzheimer's disease. The Lancet 1996; 347: 1186-1186.

13.Kehoe P, Williams J, Holmans P, et al. Association between a PS-1 intronic polymorphism and late onset Alzheimer's disease. Neuroreport. 1996; 7: 2155- 2158.

14. lsoe K, Urakami K, Ji Y, Adachi Y, Nakashima K. Presenilin-1 polymorphism in patients with Alzheimer's disease, vascular dementia and alcohol-associated dementia in Japanese population. Acta Neurol. Scand. 1996; 94: 326-328.

15. Ezquerra M, Blesa R, Tolosa E, et al. The genotype 2/2 of the presenilin-1 polymorphism is decreased in Spanish early-onset Alzheimer's disease. Neurosci.Lett. 1997; 227: 201-204.

16. Matsushita S, Arai H, Muramatsu T, et al. Presenilin-1 polymorphism in Alzheimer's disease and vascular dementia. Exp. Neurol. 1997; 146: 567-569. 17. Perez-Tur J, Wavrant-De Vrieze F, Lambert JC, Chartier-Hariin M-C, the Alzheimer's Study Group. Presenilin-1 polymorphism and Alzheimer's disease. The Lancet 1996; 347: 1560-1561.

18. Scott WK, Roses AD, Haines JL. Presenilin-1 polymorphism and Alzheimer's disease. The Lancet 1996; 347: 1560-1560.

19.Cai X-D, Stanton J, Fallin D, et al. No assiciation between the intronic presenilin-1 polymorphism and Alzheimer's disease in clinic and population-based samples. Am J Med Genet (Neuropsy Genet) 1997; 74: 202-203.

20.Sorbi S, Nacmias B, Tedde A, et al. Presenilin-1 gene intronic polymorphism in sporadic and familial Alzheimer's disease. Neurosci.Lett. 1997; 222: 132-134.

2LTysoe C, Whittaker J, Cairns NJ, et al. Presenilin-1 intron 8 polymorphism is not associated with autopsy- confirmed late-onset Alzheimer's disease. Neurosci.Lett. 1997; 222: 68-69.

22. Singleton AB, Lamb H, Leake A, McKeith IG, Perry RH, Morris CM. No association between an intronic polymorphism in the preseniiin-1 gene and Alzheimer's disease. Neurosci.Lett. 1997; 234: 19-22.

23.Lendon CL, Myers A, Cumming A, Goate AM, St. A polymorphism in the presenilin 1 gene does not modify risk for Alzheimer's disease in a cohort with sporadic early onset. Neurosci.Lett. 1997; 228: 212-214.

24. Mann DM, Pickering-Brown SM, Bayatti NN, et al. An intronic polymorphism in the presenilin-1 gene does not influence the amount or molecular form of the amyloid beta protein deposited in Alzheimer's disease. Neurosci.Lett. 1997; 222: 57-60.

25.Brice A, Tardieu S, Campion D, et al. Allelic association at the D14S43 locus in early onset Alzheimer's disease. Am. J. Med. Genet. 1995; 60: 91-93.

26. Cruts M, Backhovens H, Theuns J, et al. Genetic and physical characterization of the early-onset Alzheimer's disease AD3 locus on chromosome 14q24.3. Hum.Mol.Genet. 1995; 4: 1355-1364. 27. van Duijn CM, Farrer LA, Backhovens H, et al. A population-based study of familial Alzheimer's disease: Linkage to chromosome 14,19 and 21. Am.J. Hum. Genet. 1994; 55: 714-727.

28. Hofman A, Grobbee DE, de Jong PT, van den Ouweland FA. Determinants of disease and disability in the elderly: the Rotterdam Elderly Study. Eur.J. Epidemiol.

1991 ; 7: 403-422.

29. Cruts M, Backhovens H, Martin J-J, Van Broeckhoven C. Genetic analysis of the cellular oncogene fos in patients with chromosome 14 encoded Alzheimer's disease. Neurosci.Lett. 1994; 174: 97-100.

30. Cruts M, Backhovens H, Van Gassen G, et al. Mutation analysis of the 14q24.3 dihydrolipoyl succinyltransferase (DLST) gene in patients with early-onset Alzheimer disease. Neurosci.Lett. 1995; 199: 73-77.

3LTerwilliger JD, Ott J: Handbook of human genetic linkage. New York, John Hopkins University Press, 1994.

32. Van Broeckhoven C, Backhovens H, Cruts M, et al. APOE genotype does not modulate age of onset in families with chromosome 14 encoded Alzheimer's disease. Neurosci.Lett. 1994; 169: 179-180.

33. Cruts M, Hendriks L, Van Broeckhoven C. The presenilin genes: a new gene family involved in Alzheimer disease pathology. Hum.Mol.Genet. 1996; 5: 1449-1455.

34. Lambert JC, Pasquier F, Cottel D, Frigard B, Amouyel P, Chartier-Hariin MC. A new polymorphism in the APOE promoter associated with risk of developing Alzheimer's disease. Hum.Mol.Genet. 1998; 7: 533-540.

35.Bullido MJ, Artiga MJ, Recuero M, et al. A polymorphism in the regulatory region of APOE associated with risk for Alzheimer's dementia. Nat.Genet. 1998; 18: 69-71.

36.Tysoe C, Whittaker J, Xuereb J, et al. A presenilin-1 truncating mutation is present in two cases with autopsy-confirmed early-onset Alzheimer disease. Am.J.Hum.Genet. 1998; 62: 70-76. 37. Cruts M, Backhovens H, Wang SY, et al. Molecular genetic analysis of familial early-onset Alzheimer's disease linked to chromosome 14q24.3. Hum.Mol.Genet. 1995; 4: 2363-2371.

38. Rogaev El, Sherrington R, Wu C, et al. Analysis of the 5^* sequence, genomic structure, and alternative splicing of the presenilin-1 gene (PSEN1 ) associated with early onset Alzheimer's disease. Genomics 1997; 40: 415-424.

39. Dib C, Faure S, Fizames C, et al. A comprehensive genetic map of the human genome based on 5,264 microsatellites. Nature 1996; 380: 152-154.

40.Slooter, A.J. and van Duijn, CM. (1997) Genetic epidemiology of Alzheimer disease. Epidemiol. Rev., 19, 107-119.

41. Cruts, M., van Duijn, CM., Backhovens, H., Van den Broeck, M., Wehnert, A.,

Semeels, S., Sherrington, R., Hutton, M., Hardy, J., St George-Hyslop, P.H.,

Hofman, A. and Van Broeckhoven, C. (1998) Estimation of the genetic contribution of presenilin-1 and -2 mutations in a population-based study of presenile Alzheimer disease. Hum.Mol.Genet, 7, 43-51.

42. Campion, D., Flaman, J.M., Brice, A., Hannequin, D., Dubois, B., Martin, C, Moreau, V., Charbonnier, F., Didierjean, O. and Tardieu, S. (1995) Mutations of the presenilin I gene in families with early-onset Alzheimer's disease. Hum Mol.Genet, 4, 2373-2377. 43. Goate, A., Chartier-Hariin, M.-O, Mullan, M., Brown, J., Crawford, F., Fidani, L., Giuffra, L., Haynes, A., Irving, N., James, L., Mant, R., Newton, P., Rooke, K., Roques, P., Talbot, C, Pericak-Vance, M., Roses, A., Williamson, R., Rossor, M., Owen, M. and Hardy, J. (1991 ) Segregation of a missense mutation in the amyloid precursor protein gene with familial Alzheimer's disease. Nature, 349, 704-706. 44. Hutton, M., Busfield, F., Wragg, M., Crook, R., Perez-Tur, J., Clark, R.F., Prihar, G., Talbot, C, Phillips, H., Wright, K., Baker, M., Lendon, C, Duff, K., Martinez, A., Houlden, H., Nichols, A., Karran, E., Roberts, G., Venter, J.C, Adams, M.D., Cline, R.T., Phillips, C.A., Fuldner, R.A., Hardy, J. and Goate, A. (1996) Complete analysis of the presenilin 1 gene in families with early onset Alzheimer's disease. Neuroreport, 7, 801-805. 45. Levy-Lahad, E., Wasco, W., Poorkaj, P., Romano, D.M., Oshima, J., Pettingell,

W.H., Yu, C.-E., Jondro, P.D., Schmidt, S.D., Wang, K., Crowley, A.C., Fu, Y.-H.,

Guenette, S.Y., Galas, D., Nemens, E., Wijsman, E.M., Bird, T.D., Schellenberg,

G.D. and Tanzi, R.E. (1995) Candidate gene for the chromosome 1 familial Alzheimer's disease locus. Science, 269, 973-977.

46.Slooter, A.J., Cruts, M., Kalmijn, S., Hofman, A., Breteler, M.M., Van Broeckhoven, C. and van Duijn, CM. (1998) Risk estimates of dementia by apolipoprotein E genotypes from a population-based incidence study: the Rotterdam Study. Arch. Neurol., 55, 964-968. 47. van Duijn, CM., de Knijff, P., Cruts, M., Wehnert, A., Havekes, L.M., Hofman, A. and Van Broeckhoven, C. (1994) Apolipoprotein E4 allele in a population-based study of early-onset Alzheimer's disease. Nat.Genet, 7, 74-78.

48. Cruts, M. and Van Broeckhoven, C. (1998) Presenilin mutations in Alzheimer's disease. Human Mutation, 11 , 183-190. 49. van Duijn, CM., Cruts, M., Theuns, J., Van Gassen, G., Backhovens, H., Van den Broeck, M., Wehnert, A., Semeels, S., Hofman, A. and Van Broeckhoven, C. (1999) Genetic association of the presenilin-1 regulatory region with early-onset Alzheimer's disease in a population-based sample. Eur.J.Hum Genet, 7, 801-806.

50. Theuns, J., Del-Favero, J., Dermaut, B., van Duijn, CM., Backhovens, H., Van den Broeck, M., Serneels, S., Corsmit, E., Van Broeckhoven, C. and Cruts, M. (2000)

Genetic variability in the regulatory region of presenilin 1 associated with risk for Alzheimer's disease and variable expression. Hum.Molec.Genet, (In Press)

51. Rogaev, E.I., Sherrington, R., Wu, C, Levesque, G., Liang, Y., Rogaeva, E.A.,

Ikeda, M., Holman, K., Lin, C, Lukiw, W.J., de Jong, P.J., Fraser, P.E., Rommens, J.M. and St George-Hyslop, P.H. (1997) Analysis of the 5' sequence, genomic structure, and alternative splicing of the presenilin-1 gene (PSEN1) associated with early onset Alzheimer's disease. Genomics, 40, 415-424.

52. Theuns, J., Cruts, M., Del-Favero, J., Goossens, D., Dauwerse, H., Wehnert, A., den Dunnen, J.T. and Van Broeckhoven, C. (1999) Determination of the genomic organization of human presenilin 1 by fiber-FISH analysis and restriction mapping of cloned DNA. Mamm.Genome, 10, 410-414. 53.Dignam, J.D., Lebovitz, R.M. and Roeder, R.G. (1983) Accurate transcription initiation by RNA polymerase II in a soluble extract from isolated mammalian nuclei. Nucleic.Acids.Res., 11 , 1475-1489.

54.Mitsuda, N., Roses, A.D. and Vitek, M.P. (1997) Transcriptional regulation of the mouse presenilin-1 gene. J.Biol.Chem., 272, 23489-23497.

55. Pastorcic, M. and Das, H.K. (1999) An upstream element containing an ETS binding site is crucial for transcription of the human presenilin-1 gene. J.Biol.Chem., 274, 24297-24307.

56.Schoenherr, C.J. and Anderson, D.J. (1995) The neuron-restrictive silencer factor (NRSF): a coordinate repressor of multiple neuron-specific genes. Science, 267,

1360-1363.

57.Chong, J.A., Tapia-Ramirez, J., Kim, S., Toledo-Aral, J.J., Zheng, Y., Boutros, M.C., Altshuller, Y.M., Frohman, M.A., Kraner, S.D. and Mandel, G. (1995) REST: a mammalian silencer protein that restricts sodium channel gene expression to neurons. Cell, 80, 949-957.

58. Fornasari, D., Battaglioli, E., Flora, A., Terzano, S. and Clementi, F. (1997) Structural and functional characterization of the human alpha3 nicotinic subunit gene promoter. Mol.Pharmacol., 51 , 250-261.

59. Remade, J.E., Kraft, H., Lerchner, W., Wuytens, G., Collart, C, Verschueren, K., Smith, J.C. and Huylebroeck, D. (1999) New mode of DNA binding of multi-zinc finger transcription factors: deltaEFI family members bind with two hands to two target sites. EMBO J., 18, 5073-5084.

60.Schule, R., Umesono, K., Mangelsdorf, D.J., Bolado, J., Pike, J.W. and Evans, R.M. (1990) Jun-Fos and receptors for vitamins A and D recognize a common response element in the human osteocalcin gene. Cell, 61 , 497-504.

6LLemaigre, F.P., Lafontaine, D.A., Courtois, S.J., Durviaux, S.M. and Rousseau, G.G. (1990) Sp1 can displace GHF-1 from its distal binding site and stimulate transcription from the growth hormone gene promoter. Mol.Cell Biol., 10, 1811- 1814. 62. Hoch, M., Gerwin, N., Taubert, H. and Jackie, H. (1992) Competition for overlapping sites in the regulatory region of the Drosophila gene Kruppel. Science, 256, 94-97.

63. Bedford, F.K., Julius, D. and Ingraham, H.A. (1998) Neuronal expression of the 5HT3 serotonin receptor gene requires nuclear factor 1 complexes. J.Neurosci.,

18, 6186-6194.

64. lnoue, T., Tamura, T., Furuichi, T. and Mikoshiba, K. (1990) Isolation of complementary DNAs encoding a cerebellum-enriched nuclear factor I family that activates transcription from the mouse myelin basic protein promoter. J.Biol.Chem., 265, 19065-19070.

65.Sumner, C, Shinohara, T., Durham, L., Traub, R., Major, E.O. and Amemiya, K. (1996) Expression of multiple classes of the nuclear factor-1 family in the developing human brain: differential expression of two classes of NF-1 genes. J.Neurovirol., 2, 87-100. 66. Tamura, T., Sumita, K., Hirose, S. and Mikoshiba, K. (1990) Core promoter of the mouse myelin basic protein gene governs brain-specific transcription in vitro. EMBO J., 9, 3101-3108.

67. Tamura, T., Aoyama, A., Inoue, T., Miura, M. and Mikoshiba, K. (1990) A new transcription element in the JC virus enhancer. J.Gen.Virol., 71 , 1829-1833. 68.Cardinaux, J.R., Chapel, S. and Wahli, W. (1994) Complex organization of CTF/NF-I, C/EBP, and HNF3 binding sites within the promoter of the liver-specific vitellogenin gene. J.Biol.Chem., 269, 32947-32956.

69.Aldudo, J., Bullido, MJ. and Valdivieso, F. (1999) DGGE method for the mutational analysis of the coding and proximal promoter regions of the Alzheimer's disease presenilin-1 gene: two novel mutations. Hum Mutat, 14, 433-439.

70. Refolo, L.M., Eckman, C, Prada, CM., Yager, D., Sambamurti, K., Mehta, N., Hardy, J. and Younkin, S.G. (1999) Antisense-induced reduction of presenilin 1 expression selectively increases the production of amyloid beta42 in transfected cells. J.Neurochem., 73, 2383-2388. Tables

Table 1.

Cases Controls

Polymoφhism N % N % p-value

Promoter Allele 1 194 95 207 88 0.02

2 10 5 27 12

Genotype 11 93 91 93 79 0.05

12 8 8 21 18

22 1 1 3 3

5'UTR Allele 1 168 82 187 83 0.75

2 38 18 39 17

Genotype 11 3 3 3 3 0.95

12 32 31 33 29

22 67 66 77 68

Intron 8 Allele 1 90 45 127 54 0.05

2 110 55 107 46

Genotype 11 26 26 34 29 0.03

12 38 38 59 50

22 35 36 24 21

3'UTR Allele 1 172 86 195 84 0.57

2 28 14 37 16

Genotype 11 74 74 83 72 0.79

12 24 24 29 25

22 2 2 4 3 Table 2.

Cases 1 Controls

Polymoφhism Allele Size (bp) N % N % p-value

D14S1028 1 243 4 2 9 4 0.04

2 237 8 4 4 2

3 245 8 4 6 3

4 231 51 27 46 20

5 227 19 10 40 18

6 247 9 5 8 4

7 235 46 24 64 28

8 229 1 0.5 5 2

9 239 19 10 9 4

10 241 18 9 21 9

11 249 4 2 8 4

12 225 0 0 1 0.4

13 233 0 0 2 1

14 251 4 2 3 1

15 253 1 0.5 1 0.4

16 255 0 0 1 0.4

D14S77 1 235 6 3 3 1 0.59

2 207 25 13 37 16

3 239 6 3 13 6

4 203 8 4 17 8

5 241 5 3 5 2

6 227 7 4 9 4

7 215 17 9 16 7

8 211 4 2 5 2

9 223 9 5 12 5

10 219 7 4 5 2

11 209 12 6 17 8

12 249 10 5 7 3

13 221 3 2 3 1

14 245 6 3 7 3 15 251 0 0 2 1

16 231 5 3 3 1

17 217 9 5 5 2

18 247 10 5 6 3

19 233 8 4 6 3

20 229 10 5 12 5

21 243 6 3 4 2

22 255 2 1 0 0

23 237 10 5 6 3

24 213 3 2 9 4

25 225 12 6 14 6

26 201 0 0 3 1

D14S1004 11 119922 7711 3355 72 32 0.45

2 190 73 36 92 40

3 196 16 8 19 8

4 194 19 9 27 12

5 198 25 12 18

Table 3.

PSENl Mutations

No PSEN1 Mutations

Polymoφhism N % N % p-value

D14S1028 Genotype 44 4 67 13 14 0.05

Other 2 33 82 86

PSENl promoter Allele 1 1 1 92 181 95 0.02

2 1 8 9 5

Genotype 11 5 83 87 92 0.04

12 1 17 7 7

22 0 67 1 1

PSENl intron 8 Allele 1 4 33 84 46 0.10

2 8 66 100 54

Genotype 11 2 33 23 25 0.10

12 0 0 38 41

22 4 67 31 34

Table 4.

Cases 1 Controls

Marker N % N % p-value

Exon 3 Allele 1 163 80 184 79 0.47

2 41 20 48 21

Genotype 11 63 62 70 60 0.95

12 37 36 44 38

22 2 2 2 2

Exon 4 Allele 1 103 53 126 54 0.72

2 93 47 106 46

Genotype 11 23 24 32 28 0.75

12 57 58 62 53

22 18 18 22 19

Intron 11 Allele 1 108 53 127 55 0.72

2 94 47 105 45

Genotype 11 23 23 31 27 0.72

12 62 61 65 56

22 16 16 20 17

Table 5.

Position m N Position mAF 109907 Nucleotide change Polymoφhism Pπmers (N>AF 109907)

1500 49585 G>A ? prom20

2609 50694 />T ? prom21

4035 52120 />ττ T stretch variation prom 13

4840 52925 G>A G to A transition proml 1

6158 54261 G>C None (Only G in Rot22) prom3

7736 55839 G>C ?

(8620) 56752 A>G ?

(8696) 56783 />A ?

(9167) 57254 T>A ?

Table 7:

Location Vaπation Allele frequencies Restπction site Pπmer set change

(-48) C→T 0.88/0.12 Ugal, Hs/7961, S/aNI Prom2

(-280) C→G ? Ncol Proml

(-1789) G→A ? Mismatch pπmers Proml 1

(-2154) G→A ? BsmBl Prom 13

(-2319) variable T-stretch ? ? Prom 13

(-2818) A→G ? Banl Prom 15

(-2823) deletion/insert of 13 bp: ? N/αl Prom 15

5'-GCATGTCCTGGGT-3' Table 6.

Name Primer sequence (Mg ^t| Ta (°C) Length (bp) Additives

Proml F CGCCACCCACCAGAAGTTTTGATT 1 mM 64 248 Proml R AGAGCGCTGGCGTGGGTTGT

Prom2F AGCAGCCTCAGAACCCCGACAA 1 ,5 mM 64 344 Prom2R ACTCCCCATCACGCACATACG

PronϋF GCCGGGAGAAGCACACG 1 mM 60 398 Prom3R TCCCCATCACGCACATACG

Prom4F TGACCCAGCAAACCAGACAC 1 mM 56 229 Prom4R GTACCTATTTGCGAACCAG

PromόF GTTAGGCTGGTCTAGAACTCCCAACCTCAT 1 mM 62 336 PromόR ATCTGCCTGCTCTTTGTCCAGTTTTGGTAA

Prom7F GAGGAGGGATTGGAGATTGATGCGATAGG 1 mM 60 ₄00 Prom7R TGGGCAAGTTACATTCTTCTAAAGTTTCTT

Prom9F TAGATGGGTTTTAGCCTGTATTTGTT 1,5 mM 60 359 Prom9R TTCCTATCGCATCAATCTCCA

PromlOF TCAGATGAGCCGCCCACCTT I mM 63 382 PromlOR AAGAGTAGGGATTTTAGTTTGTTCATTTA

Prom 11 F TGCCTCAGCCTCCCAAGTA 1 mM 53 361 Proml 1R AACCCATCTATTTGTGAACTAC

Proml 3F AAGGCTGGTTATTCAATGTTAG 1,5 mM 62 398 Proml 3R GTCCACCCACCTCACAGAAT

Proml4F AAGTGGGTTGAGTCCGAAAAGAG 1 mM 56 395 Proml 4R TGGCAACAAAAAGGCAGTAAGATT

Proml 5F TTAAAGGGTAGTGAGAAGGCTGGAGAAGAG I mM 62 32₄ Proml 5R AACTGCCCCACCCCCATTTC

Prom20F AGTGGGTAGCCTCCGTATTGATGAAGAAGGGGATGGACTT "buffer I 55 350 ProπώOR GGGGCTCCTTCAAAACCTCCTCGTGGACCTTGCTTCAAAT

ProπώlF ATTAACCTTGACTGTGCCTTCAGCTCCAGCCACCTCTTTA 'buffer 55 337 Prom21R CCCTAGTCTCTGCTCCCAATGCAACTCGTTCCAAATCTTT

Prom22F TGAGATTACAGGCCCACACCAC "> "> 65 Prom22R GAAACCCCGTCTCTACTAAAAA

Prom23F CCGGAATTACAGGCGTGAGC 1/1,2 mM 66 416 10% DMSO Prom23R GTGGCGAGAACCTGAGAAACTGC

^♦Invitrogen PCR Optimizer Kit

Claims

Claims

1. A method for determining whether a human subject has or is at risk for developing early-onset Alzheimer disease or Alzheimer disease comprising the step of detecting the presence or absence of a genetic lesion in the presenilin-1 gene of said subject, wherein said genetic lesion comprises a polymorphism in the presenilin-1 promoter region or regulatory region upstream of presenilin-1 and wherein the presence of said genetic lesion identifies a subject that has or is at risk for developing early-onset Alzheimer disease or Alzheimer disease.

2. A method according to claim 1 , wherein said polymorphism in the presenilin-1 promoter region is chosen from the group comprising: -48 C/T; -280 C/G; -2154 G/A, -2818 A/G and -2823 l/D and/or wherein said polymorphism in the regulatory region upstream of presenilin-1 is a simple tandem repeat polymorphism at D14S1028.

3. Use of a variant in the presenilin-1 promoter region and/or the preseniiin-1 regulatory region upstream of presenilin-1 according to claims 1 and 2 to determine in a human cell or tissue obtainable from a human being whether said human being has or is at risk for developing early-onset Alzheimer disease or Alzheimer disease.

4. Transgenic non-human animal comprising in its genetic material a human presenilin-1 variant in the presenilin-1 promoter region and/or in the presenilin-1 regulatory region upstream of presenilin-1.

5. Use of a transgenic non-human animal according to claim 4 to screen for therapeutic molecules to treat Alzheimer disease.

6. A method to screen for molecules which inhibit the reduction of presenilin-1 levels induced by a polymorphism in the presenilin-1 promoter region and/or regulatory region upstream of presenilin-1 comprising:

-exposing said polymorphism in the presenilin-1 promoter region or regulatory region upstream of presenilin-1 to said molecules, and -monitoring said presenilin-1 levels.

7. A method according to claim 6, wherein said polymorphism in the presenilin-1 promoter region is -48 C/T and/or -280 C/G.

8. A molecule obtainable by using a transgenic animal according to claim 5 and/or a method according to any of claims 6 to 7.