SCREENING AND TREATMENT OF ALZHEIMER'S DISEASE

The invention relates to a method and means for diagnosing the susceptibility to, or existence of, Alzheimer's Disease (AD) and also the use of information gained by the aforementioned method to select a treatment program for treating Alzheimer's Disease.

Higher cortical functions such as memory, thinking and orientation are severely affected during a chronic progressive mental disorder such as dementia. Alzheimer's Disease is the most common form of dementia and is a degenerative cerebral disease with a characteristic pathology and biochemistry.

Alzheimer's disease, predominantly, affects the elderly producing a characteristic deterioration in the ability to think, conceive and reason and also a similar deterioration in the ability to carry out motor or coordination skills. Additionally, Alzheimer's Disease can cause behavioural and mood changes and it is not uncommon for sufferers to become aggressive.

Individuals suffering from Alzheimer's Disease, typically, lose the ability to carry out routine daily functions and therefore require a high level of care which is often provided by an elderly relative. This can have a significant and, sometimes, depressing effect on the outlook of the elderly carer.

In the UK around 300,000 people are affected with the disease and because of

the continuous declining brain function, prior to death, victims often experience a prolonged phase where they are unable to look after their basic needs.

Several different methods are used to assess the severity of Alzheimer's Disease. These include: the clinician's interview-based impression of change (CIBIC), the progressive deterioration scale (PDS) for functional/quality of life scales, the Alzheimer's Disease assessment scale (ADAS) and the mini mental state examination (MMSE). A MMSE score denotes the severity of cognitive impairment as follows: Mild AD: MMSE 21-26 Moderate AD: MMSE 10-20 Moderately Severe AD: MMSE 10-14 Severe AD: MMSE less than 10.

Specific treatments that interfere with the progression of Alzheimer's Disease are limited in number and efficacy. Moreover, recently, the National Institute of Clinical Excellence (UK) has provisionally recommended that there ought to be a high cost benefit ratio before drugs to treat the disease are prescribed.

As mentioned above, Alzheimer's Disease has a characteristic biochemistry and so neurochemical features. Indeed, the disease is thought to impair cholinergic transmission in the central nervous system. To combat this pathology a number of acetylcholinesterase (AChE) inhibitors have been prescribed. These inhibitors raise the concentration of acetylcholine at sites of neurotransmission

by preventing the natural degradation of acetylcholine by the enzyme acetylcholinesterase. Unfortunately, though, not all patients respond to this treatment and therefore, although there are a number of acetylcholinesterase inhibitors available on the market, such as Donepezil, Rivastigmine and Galantamine, their success rate varies.

A considerable amount of research has been carried out over a number of decades in order to identify a marker for the disease.

We have previously investigated the role of the choline acetyltransferase gene (CHAT) in Alzheimer's Disease. The enzyme encoded by this gene is responsible for the biosynthesis of acetylcholine at neural endings. We therefore considered that sequence variations _. in the gene encoding CHAT might be responsible for the disease and so serve as a marker for same. However, although following our investigations we were able to detect a total of 17 sequence variants in this gene, 3 of which gave rise to non-synonymous variants in the CHAT enzyme, none of the 17 sequence variants showed any association with Alzheimer's Disease. (Human Genetics (2003) Vol. 113, pages 258-267).

Given this lack of success, we were surprised, when conducting further experiments, to discover that there was a marker on the CHAT gene which showed an association with Alzheimer's Disease and, in particular, the response of patients to acetylcholinesterase inhibitor treatment in Alzheimer's Disease. We were even more surprised to discover that the marker was not part of the

coding sequence of the gene but was, rather, upstream thereof in a sequence of nucleic acids designated as one- probable promoter region of the several promoters for this gene.

The marker is an SNP, designated rs733722, which is located in the sequence of the CHAT locus 1587 base pairs 5' of the SLC18A3 gene which is, in turn, upstream of the putative exon R of CHAT (see Figure 1).

It is predicted that the marker, rs733722, resides in a transcription factor binding site. The number and nature of transcription factors is illustrated in Figure 2. The data was derived using transcription elements search software available on the world-wide web via the Computational Biology and Information Laboratory School of Medicine University of Pennsylvania URL: http://www.cbil.upenn.edu/tess.

We have therefore, fortuitously, identified a marker in the CHAT gene that associates with the response to treatment by acetylcholinesterase inhibitors in Alzheimer's Disease.

Statements of Invention

According to a first aspect of the invention there is therefore provided a method of diagnosing the existence of, or susceptibility to, Alzheimer's Disease comprising: a) obtaining a nucleic acid sample from a subject;

b) examining said sample for the presence of the T allele at marker rs733722 in the promoter region of the CHAT gene; and, where the marker is present, c) concluding that said individual may benefit from acetylcholinesterase treatment for Alzheimer's Disease.

In a preferred method of the invention said examination involves a genetic test in order to identify the said marker and so involves, the use of at least one probe that is adapted to identify the existence of the T allele rather than the C allele at marker rs733722. The genomic DNA from a sample, containing the rs733722 locus is amplified using a polymerase chain reaction or other amplification method. In a preferred embodiment of the method of the invention step b) involves the use of a probe or oligonucleotide, ideally labelled, that is adapted to bind to either the rs733722 site and/or said site and a site upstream and/or downstream thereof. Accordingly, the probe comprises a sequence of DNA that is complementary to a region of the rs733722 site that comprises acttcctgcg gggtggggggc/tacaatgggag aagcatctgc.

Additionally or alternatively, suitable oligonucleotides for amplifying the rs733722 site include oligonucleotides that are complementary to regions of DNA either 5' and/or 3' of rs733722 and so include oligonucleotides that are complementary to any part of the following sequence 5' to rs733722 or 3' to rs733722.

5' flank: ctttcatctg ttcttcacaa gacccacaag tgaaaaatga ggataaaccg cacattctac acacgataac aacatagcaa gtccttatac tgctcttact gtgtgcccgg cccccttagc agtaggtact attattattc aattttacaa agaaactaac tgagggacag agaggcaaag taattccctg caggtactct aatgagtacg tggcagagct gggagccatc ctggttgtct gcctcgaaag accacactct ttgcactgca tcgcggcact tcctgcgggg tgggggg

Observed: Y(c/t)

3' flank: acaatgggag aagcatctgc gtctaatgct gctttacttt tgaggccaga aaaatgggaa ggctcccctc tgactctgga agagagacgc aaaccgtaat ctcaacaaca caatccccac ctccaacctc agccgccctg gagcctctct cccgccagtc cgcccactgg aacacgggtt ccatgtgcca tccagggtca acgccgctct ggggacgcgt caggcccagc gcacagcctg ggcagctcag cctgtca

Our preferred primers are

F: TCTAATGAGTACGTGGCAGAGC

R: CTGGATGGCACATGGAACC

Thus, the presence of the C and T alleles are detected at rs733722 by the use of a specific probe in either amplified or genomic DNA.

The presence of the rs733722 C allele is indicative of a lack of function of the choline acetyltransferase gene and its corresponding protein. This, in turn, means there will be a lack of biosynthesis of acetylcholine and therefore, it follows, that the use of acetylcholinesterase inhibitors to treat the disease is unlikely to be successful because the premise for using these inhibitors is based upon a reasonable supply of acetylcholine. Clearly, this will not be the case if choline acetyltransferase is impaired or non-functional.

The information derived from the diagnosis relating to the invention provides valuable information for the treatment of this condition. It follows, that an individual who carries the T allele for the rs733722 marker is more likely to benefit from a prescription of acetylcholinesterase inhibitors. The converse is true for an individual who has a C allele at this marker.

According to yet a further aspect of the invention there is therefore provided a method for treating an individual suffering from, or suspected to be suffering from, Alzheimer's Disease comprising: a) obtaining a nucleic acid sample from the subject; b) examining said sample for the presence of a T/C allele at marker rs733722 in a promoter region of the CHAT gene and where the marker is present; c) optionally, concluding, where C allele is present, that said individual is susceptible to, or suffering from, Alzheimer's Disease; and d) concluding, where T allele is present, that the individual is likely to benefit from treatment with at least one acetylcholinesterase inhibitor and so prescribing this treatment.

In a preferred embodiment of the method of the invention step b) involves the use of a probe or oligonucleotide, ideally labelled, that is adapted to bind to either the rs733722 site and/or said site and a site upstream and/or downstream thereof. Accordingly, the probe comprises a sequence of DNA that is

complementary to a region of the rs733722 site that comprises acttcctgcg gggtggggggc/tacaatgggag aagcatctgc.

Additionally or alternatively, suitable oligonucleotides for amplifying the rs733722 site include oligonucleotides that are complementary to regions of DNA either 5' and/or 3' of rs733722 and so include oligonucleotides that are complementary to any part of the following sequence 5' to rs733722 or 3' to rs733722.

5' flank: ctttcatctg ttcttcacaa gacccacaag tgaaaaatga ggataaaccg cacattctac acacgataac aacatagcaa gtccttatac tgctcttact gtgtgcccgg cccccttagc agtaggtact attattattc aattttacaa agaaactaac tgagggacag agaggcaaag taattccctg caggtactct aatgagtacg tggcagagct gggagccatc ctggttgtct gcctcgaaag accacactct ttgcactgca tcgcggcact tcctgcgggg tgggggg

Observed: Y(c/t)

3' flank: acaatgggag aagcatctgc gtctaatgct gctttacttt tgaggccaga aaaatgggaa ggctcccctc tgactctgga agagagacgc aaaccgtaat ctcaacaaca caatccccac ctccaacctc agccgccctg gagcctctct cccgccagtc cgcccactgg aacacgggtt ccatgtgcca tccagggtca acgccgctct ggggacgcgt caggcccagc gcacagcctg ggcagctcag cctgtca

Our preferred primers are

F: TCTAATGAGTACGTGGCAGAGC

R: CTGGATGGCACATGGAACC

According to a yet further aspect of the invention there is provided a method for treating Alzheimer's Disease which comprises:

a) determining whether an individual suffering from said disease shows C or T allele at the rs733722 marker in a promoter region of choline acetyltransferase gene; and b) where the individual shows a T allele for this marker prescribing at least one acetylcholinesterase inhibitor.

According to a further aspect of the invention there is provided an apparatus or computer system for performing one or more of the aforementioned methods of the invention.

According to a further aspect of the invention there is provided an apparatus or computer system for performing a method according to any preceding' claim comprising: a) a means for sequencing from a subject a nucleic acid sample comprising at least the promoter region of the CHAT gene; b) a means for determining the presence of the T/C allele at SNP site rs733722; and c) where said T allele is present, a means for presenting information that the individual from whom said sample is taken is likely to benefit from treatment with at least one acetylcholinesterael inhibitor and so, optionally, prescribing this treatment; and/or d) where said C allele is present a means for presenting information that the individual from whom said sample is taken is susceptible to, or suffering from, Alzheimers Disease.

The invention will now be exemplified by reference to the following information wherein:

Figure 1 shows genomic structure of the CHAT locus showing the relationship of the CHAT and SLC18A3 genes and the complex 5' structure of CHAT. The five alternative ChAT encoding transcripts which are transcribed from the CHAT locus are shown: all variants encode the same 69Kda ChAT protein, but the M variant also encodes an 82KDa and the S variant a 74kDa protein (Oda et al.,

1992; Touissaint et al., 1992; Chireux et al., 1995; Hahm et al., 1997; Ohno et al., 2001). For CHAT, nucleotide numbers start from the translational start site of the 82 kDA splice variant (M) of the gene. Similarly, numbering starts from the translational start site of SLC18A3 for variants detected in this gene; and

Figure 2 shows predicted transcription factor binding sites around rs733722.

Prediction was carried out using TESS

(TESS: Transcription Element Search Software on the WWW,

Jonathan Schug and G. Christian Overton,

Technical Report CBI L-TR- 1997-1001 -vθ.0

Computational Biology and Informatics Laboratory

School of Medicine

University of Pennsylvania, 1997

URL: http://www.cbil.upenn.edu/tess)

CHAT, the gene encoding CHAT. CHAT was subjected to de novo

polymorphism discovery using DHPLC, a resequencing technique that looks for polymorphic variants in DNA (Harold et al 2003).

Mutation detection

CHAT encodes five splice variants; M, N1, N2, R and S. SLC18A3 is located within the first intron of the R variant of CHAT and is itself uninterrupted by introns (Figure 1). The cDNA sequence for each splice variant/gene was obtained from the GenBank database at NCBI (http://www.ncbi.nlm.nih.gov). Determination of coding sequences, untranslated regions (UTRs) and intronic regions was based on alignment of the cDNA sequences with genomic clone sequences, using BLAST sequence homology searches (http://www.ncbi.nlm.nih.gov/blast/Blast.cgi ).

PCR fragments spanning exons, UTRs and limited 5' flanking regions were designed using Primer 3.0 (http://www-qenome.wi.mit.edu/cgi- bin/primer/primer3_www.cqi). Primer sequences and PCR conditions for markers other than rs733722 are available on request from the authors or at http://www.uwcm.ac.uk/studv/medicine/psychological medicine/pub data/chat.h trn. PCR amplification was performed under standard conditions of 1X PCR

buffer (Qiagen), 1.5 mM MgCI ₂ , 250 μM dNTPs, 0.5 μM of each primer, 0.6 units

Hot Start Taq (Qiagen) and 48 ng genomic DNA in a 24 μl reaction. Cycling was

conducted in a MJ Tetrad (MJ Research) with an initial denaturation of 94 ⁰ C for 15min, followed by 35 cycles of 94 ⁰ C for 30s, appropriate annealing temperature for 30s and 72 ⁰ C for 45s with a final extension step of 72 ⁰ C for 10 min.

Synthesis of appropriately sized PCR products was confirmed by electrophoresis on 2% agarose gels. The primer sequences for rs733722 were: F: TCTAATGAGTACGTGGCAGAGC R: CTGGATGGCACATGGAACC

Polymorphisms were identified by DHPLC using a Wave™ DNA Fragment Analysis System (Transgenomic) as previously described (Abraham et a)., 2001). The 14 screening samples were amplified as described above, except the final extension in the PCR protocol was followed by denaturation at 94 ⁰ C for 5min and then cooling to 65 ⁰ C over 30min, to allow heteroduplex formation. Column temperature and acetonitrile gradient were determined using the DHPLC Melt program (http://insertion.stanford.edu/melt1.html). To ensure maximum sensitivity, in addition to the temperature suggested by the software (n°C), each fragment was also run at n+2°C. Samples showing heteroduplex formation were sequenced to identify the variant.

PCR products were purified through QlAquick columns (Qiagen) to remove unincorporated primers and dNTPs. Purified products were then bidirectionally sequenced on an ABI 377 DNA Sequencer (Applied Biosystems) using the Big Dye Terminator (v2.0) Cycle Sequencing kit (Applied Biosystems). Sequence traces were subsequently exported to Sequencher™ (Applied Biosystems) to characterise polymorphisms.

Genotyping

Where a natural restriction site existed that could distinguish between the two alleles of a SNP, a restriction fragment length polymorphism (RFLP) assay was designed, using the original PCR primers designed for DHPLC. In some cases where no natural restriction site existed, it was possible to create an artificial restriction site, by primer generated mutagenesis. For each RFLP assay, the UK1 association sample was PCR amplified, then digested with 5 units of the appropriate restriction enzyme for rs733722 this was Rsa1 for 2 hours at 37 ⁰ C. A 296bp fragment produced the following signature with the C allele: 286bp + 10bp and the following signature with the T allele: 195bp + 91 bp + 10bp. (Table 1). Digested products were electrophoresed on 2.5% agarose gels.

The five SNPs were typed in a multiplex primer extension assay. Extension primers were designed to be 17, 27, 37, or 47 nucleotides long and directly adjacent to the polymorphism. For each SNP, the UK1 association sample was PCR amplified and then purified by incubation with 1 unit each of exonuclease I and shrimp alkaline phosphatase at 37 ⁰ C for 1 hour. Primer extension was then performed using the ABI SnaPshot™ ddNTP Primer Extension kit on the ABI 3100 Genetic Analyser. Data were analysed using Genotyper® 2.5.

The VNTR was genotyped by PCR amplification followed by visualisation on a 1.5% agarose gel.

Statistical Analysis

All polymorphisms were tested for deviation from HWE independently in each population. Chi-square and Fisher's exact test were used to analyse SNP associations using the Simple Interactive Statistical Analysis pages (http://home.clara.net/sisa): Fisher's exact test was used for analyses where one or more cell had a count of <5. The VNTR (polymorphism 12) was tested for association with LOAD using CLUMP (Sham and Curtis, 1995). Haplotypic association was tested using EHPLUS (Xie and Ott 1993) with PMPLUS (Zhao et a/., 2000) implemented to obtain empirical significance levels (Terwilliger and Ott, 1994). Marker-marker linkage disequilibrium (LD) analyses were also undertaken, using HAPMAX. This program employs an EM algorithm to allow for phase unknowns. Estimated haplotype frequencies were used to calculate D' and r ² estimates of LD. Meta-analysis was carried out using an inverse variance method of weighting in a fixed effect model. Heterogeneity between studies was assessed with a Chi-square test (Cooper and Hedges, 1994).

SNPs with a minor allele frequency of >0.15 were genotyped. To reduce redundancy, where pairs of markers were in strong LD (r ² >0.9), only one was typed. We examined the effect of five single nucleotide polymorphisms (SNPs) that conformed to these criteria, on the effect of AChE inhibitors on mini mental state examination (MMSE) score, in the mild-moderate stages of the disease (Table 1), plus one further SNP identified in the SNP databases (rs8178981).

Participants were 121 white AD patients (mean age at treatment initiation 75.12 ± 7.37y; mean MMSE 20.90 ± 4.08, 53% female) living in N Ireland. All

individuals had MMSE measures at approximately six monthly intervals. At initiation, 70 were prescribed donepezil, 41 galantamine and 10 rivastigmine: 8 patients had their medication changed during the study but were included in the analysis as the change was to another AChE inhibitor. Response to treatment was measured by rates of decline in MMSE scores (points/y) from the first to the last available point: twelve individuals had measurements at two time points, 62 at three time points and 47 at four time points: the average duration in the study for all AD cases was ~15m. Analysis was carried out in qtphase (Dudbridge, 2003).

SNP rs733722 explained almost 6% of the variance in response (p = 0.0065), with the T allele associated with negligible decline of MMSE score compared with the C allele. Permutation analysis (20,000 iterations) to allow for multiple testing gave a empirical p value for the association of 0.0313. The additive effect of having one copy of the C allele at this locus is a decline of 1.66 MMSE points/y compared with having the T allele (95% Cl: 0.629-2.656 MMSE points/y). Thus a subject who carries a CC genotype at this locus had a mean MMSE score decline of 1.66x2 i.e. 3.24 MMSE points/y compared with a TT homozygote with a negligible decline in MMSE score.

To ensure that the association was to response to treatment, rather than a main effect on the natural history of this disorder, we genotyped 176 AD individuals from a second UK AD sample, diagnosed with probable AD according to NINCDS-ADRDA criteria (as detailed in Harold et al. ² ), who had not taken AChE

inhibitors and who had MMSE > 0 at collection. Decline in MMSE score was calculated by using MMSE score at collection subtracted from an assumed score of 29 at onset of disease (age at onset 77.22 ± 6.29y, age at collection 83.39 ± 5.69y;18.8% males, average disease duration ~6y). The power of this sample to detect the effect size seen at the 5% level is 89% ⁴ . No evidence for association was found between rate of decline in MMSE score and rs733722 (p = 0.4258), suggesting the association in the drug treated group was to drug response.

Our data therefore suggest that rs733722, which lies in a putative promoter region of CHAT, is a marker of response to AChE treatment.

TABLE 1 : Quantitative association analysis of ChAT genotype and MMSE score in AD patients treated with AChE inhibitors

Markers are numbered as in Harold et al. (2003).