Skip Navigation


Human Molecular Genetics Advance Access originally published online on December 21, 2006
Human Molecular Genetics 2007 16(3):295-306; doi:10.1093/hmg/ddl463
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Supplementary Data
Right arrow All Versions of this Article:
16/3/295    most recent
ddl463v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (5)
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Sundar, P. D.
Right arrow Articles by Schellenberg, G. D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sundar, P. D.
Right arrow Articles by Schellenberg, G. D.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

Published by Oxford University Press 2006

Two sites in the MAPT region confer genetic risk for Guam ALS/PDC and dementia

Purnima Desai Sundar1,5, Chang-En Yu1,5, Weiva Sieh2, Ellen Steinbart1,5, Ralph M. Garruto6, Kiyomitsu Oyanagi7, Ulla-Katrina Craig8, Thomas D. Bird4,5, Ellen M. Wijsman2,3, Douglas R. Galasko9 and Gerard D. Schellenberg1,4,5,*

1 Department of Medicine, Division of Gerontology and Geriatric Medicine, 2 Department of Medicine, Division of Medical Genetics, 3 Department of Biostatistics, 4 Department of Neurology and Pharmacology, University of Washington, Seattle, WA 98195, USA, 5 Geriatric Research Education and Clinical Center, Veterans Affairs Puget Sound Health Care System, Seattle Division, Seattle, WA 98108, USA, 6 Department of Anthropology, Laboratory of Biomedical Anthropology and Neurosciences, Binghamton University, SUNY-Binghamton, NY 13902-6000, USA, 7 Department of Neuropathology, Tokyo Metropolitan Institute for Neuroscience, Tokyo, Japan, 8 University of Guam, Mangilao, Guam 96923, USA and 9 Department of Neurosciences, University of California, San Diego, La Jolla CA 92093-0662, USA

* To whom correspondence should be addressed at: GRECC S-182B, Veterans Affairs Puget Sound Health Care System, 1660 S. Columbian Way, Seattle, WA 98108, USA. Tel: +1 2067642701; Fax: +1 2067642569; Email: zachdad{at}u.washington.edu

Received October 7, 2006; Accepted December 6, 2006


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 RESULTS
 DISCUSSION
 METHODS
 SUPPLEMENTARY MATERIAL
 REFERENCES
 
Unusual forms of amyotrophic lateral sclerosis (ALS-G), Parkinsonism dementia complex (PDC-G) and Guam dementia (GD) are found in Chamorros, the indigenous people of Guam. Neurofibrillary tangles composed of hyperphosphorylated tau are a neuropathologic feature of these closely related disorders. To determine if variation in the gene that encodes microtubule-associated protein tau gene (MAPT) contributes to risk for these disorders, we genotyped nine single nucleotide polymorphism (SNP) sites and one insertion/deletion in the 5' end of MAPT in 54 ALS-G, 135 PDC-G, 153 GD and 258 control subjects, all of whom are Chamorros. Variation at three SNPs (sites 2, 6 and 9) influenced risk for ALS-G, PDC-G and GD. SNP2 acts through a dominant mechanism and is independent of the risk conferred by SNPs 6 and 9, the latter two acting by a recessive mechanism. Persons with the high-risk SNP6 and SNP9 AC/AC diplotype had an increased risk of 3-fold [95% confidence interval (CI)=1.10–8.25] for GD, 4-fold (95% CI=1.40–11.64) for PDC-G and 6-fold (95% CI=1.44–32.14) for ALS-G, compared to persons with other diplotypes after adjusting for SNP2. Carriers of the SNP2 G allele had an increased risk of 1.6-fold (95% CI=1.00–2.62) for GD, 2-fold (95% CI=1.28–3.66) for PDC-G, and 1.5-fold (95% CI=0.74–3.00) for ALS-G, compared to non-carriers after adjusting for SNPs 6 and 9. Others have shown that SNP6 is also associated with risk for progressive supranuclear palsy. These two independent cis-acting sites presumably influence risk for Guam neuro-degenerative disorders by regulating MAPT expression.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
RESULTS
 DISCUSSION
 METHODS
 SUPPLEMENTARY MATERIAL
 REFERENCES
 
In the 1950s, amyotrophic lateral sclerosis (ALS) was highly prevalent (143 per 100 000) in Chamorros, the indigenous people of Guam (1). ALS in Guam (ALS-G) is clinically similar to ALS in other populations (1,2). However, unlike typical ALS, where neurodegeneration is confined to spinal cord motor neurons, in ALS-G, central nervous system changes occur including numerous neurofibrillary tangles (NFTs) typically in the neocortex, hippocampus and subcortical regions (2–6). A related disorder, parkinsonism dementia complex (PDC), is also highly prevalent in the same Chamorro population (2,7). The clinical symptoms of PDC in Guam (PDC-G) are parkinsonism (with bradykinesia, gait impairment, rigidity and often resting tremor), accompanied by progressive dementia. The neuropathologic features of PDC-G overlap with ALS-G (4,5,8,9) with a higher NFT burden in brain regions corresponding to the respective clinical phenotypes of the two disorders. Spinal cord NFTs are found in both ALS-G and PDC-G (10,11). More recently, late-life dementia without parkinsonism in Chamorros, referred to here as Guam dementia (GD), has been studied (12). The relationships between GD and ALS-G, and PDC-G and Alzheimer's disease (AD) are unclear. Preliminary neuropathology studies suggest that GD has prominent AD-type tangles with less prominent amyloid or neuritic pathology than that found in typical AD (13).

The etiology of ALS-G, PDC-G and GD, collectively referred to here as Guam neurodegenerative disorders, is unknown, though there is substantial evidence supporting a genetic hypothesis. Early work noted highly variable prevalence rates in different Guam villages. Families with multiple cases of ALS-G and PDC-G were common, numerous multigenerational pedigrees with multiple cases of ALS-G, PDC-G or both were described, and the affected father–son pairs were documented (7,14–18). In several studies, 35–40% of probands have relatives with ALS-G/PDC-G (14,17). In the village of Umatac, which had the highest prevalence of the disease (250/100 000) (1), segregation analysis suggested a major gene with penetrance possibly affected by an environmental factor(s) (19). More recently, Plato et al. (20), who followed the relatives of a cohort of ALS-G/PDC-G cases and controls for over 40 years, found more new cases in the offspring of affected subjects compared to relatives of controls. Another cluster of ALS and PDC exists on the Kii peninsula of Japan, where the disease is strongly familial (21–25). An environmental hypothesis is also plausible, supported by the decline in incidence of ALS-G over the past 40 years and an increase in onset ages for both ALS-G and PDC-G (2,12,26). No specific environmental factor has been identified though several have been proposed including ß-methylamino-alanine (BMAA), a neurotoxin found in cycad seeds (27). The most parsimonious hypothesis is that there are both genetic and environmental components to Guam neurodegenerative disorders.

Identification of the genes that influence susceptibility to Guam neurodegenerative disorders is important to understanding the etiology of these disorders. ALS-G, PDC-G and GD are tauopathies, a class of diseases where aggregated tau (e.g. NFTs) is a prominent neuropathologic feature. For some tauopathies, genetic variation in MAPT, the gene that encodes for tau, causes or contributes to the disease. Mutations in MAPT cause frontotemporal dementia (FTD) with parkinsonism—chromosome 17 type (FTDP-17) (28–30), an autosomal dominant group of tauopathies with variable phenotypes. However, no MAPT mutations have been identified in subjects with Guam disorders (31,32) and the major locus responsible for this disease remains to be found (33). More subtle genetic changes in MAPT, presumably in regulatory regions, increase susceptibility to progressive supranuclear palsy (PSP) (34–36), FTD (37–39), corticobasal degeneration (CBD) (40,41) and possibly AD (42), all of which are tauopathies. Although the polymorphic site responsible for elevated risk has been difficult to identify definitively, recent work suggests that alleles at a single nucleotide polymorphism (SNP) in the first intron of tau (intron 0) influence MAPT expression and may be the causative site (43). We previously examined 49 ALS-G cases, 82 PDC-G cases and 78 Chamorro controls and reported significant association with the dinucleotide repeat marker CA3662, located 23.9 kb upstream of the transcription start site of MAPT (32) (Fig. 1). Since the CA3662 association is likely due to linkage disequilibrium (LD) with the causative variant in the surrounding region, and because recent work on the association between MAPT genotypes and PSP implicated SNPs in the 5' end of MAPT (35,43,44), we examined SNPs in corticotrophin-releasing hormone receptor 1 (CRHR1) and intramembrane protease 5 (IMP5), genes that are upstream of MAPT, and SNPs within the 5' end of MAPT (Fig. 1) in a larger case–control dataset of 54 ALS-G, 135 PDC-G, 153 GD and 258 Chamorro controls. We identified two independent regions that contribute to Guam neurodegenerative disorder risk. One is in the IMP5 region upstream of MAPT and acts in a dominant mechanism. The second is in MAPT, acts by a recessive mechanism, and possibly corresponds to the same site(s) in MAPT that confer risk for PSP.


Figure 4631
View larger version (15K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 1. The chromosome 17 MAPT region. The distances shown are proportional to the actual sizes of the different genomic elements. Numbers in boxes refer to the SNPs used in this study. The direction of transcription is indicated by the direction of the arrows used to represent CRHR1, IMP5, aMAPT and MAPT. Alternatively spliced MAPT exons (solid boxes) are indicated by connecting lines above each exon.

 

   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
DISCUSSION
 METHODS
 SUPPLEMENTARY MATERIAL
 REFERENCES
 
In the previous work, we showed an allelic association between ALS-G, PDC-G and an allele of a polymorphic site that is 23.9 kb from the 5' end of MAPT (32). Other work showed that for PSP, a tauopathy with some overlapping features with ALS-G and PDC-G, SNP sites within MAPT are associated with disease risk (35,43). To determine which polymorphic site(s) in the MAPT region contributes to risk of ALS-G and PDC-G, we genotyped eight markers within MAPT between the promoter and intron 9, and two in the flanking region 5' of MAPT (Table 1, Fig. 1). Also, we examined the same loci in a GD population to determine if the same alleles confer risk to this related disorder. The cases and controls used were Chamorros ascertained in Guam. Controls were 258 Chamorro subjects with mean age 66.5±14.0 years (35% male) who were cognitively normal, and did not have other neurologic disorders. The case groups were 54 ALS-G subjects (mean age at onset 51.1±11.6 years; 65% male), 135 PDC-G subjects (mean age at onset 64.2±10.1 years; 53% male) and 153 GD cases (mean age at onset 74.9±7.2 years; 27% male).


View this table:
[in this window]
[in a new window]

  		Table 1. Polymorphic sites

 
Allele frequencies for the markers genotyped are given in Table 2. We used marker del-in9 to tag the H1/H2 haplotype system, an inversion polymorphism where approximately 900 kb of chromosome 17 spanning MAPT are in opposite orientations in the two haplotypes (45). All SNPs genotyped in this study are within the chromosome 17 segment inverted in this polymorphism. The H1 haplotype is associated with increased risk for PSP and is more common than the H2 haplotype. In our Chamorro samples, the frequency of the H1 haplotype ranged from 95.9% in the PDC-G group to 91.3% in controls (Table 2), which is higher than in typical Caucasian controls (74–80%) (42,45,46). Consistent with previous reports in Caucasians, the H2 haplotype in Chamorros was much less diverse than the H1 haplotype (43,45); nearly all of the SNPs we examined were monomorphic in our sample of H2 haplotypes and we observed strong pairwise LD between the H1/H2 inversion polymorphism and MAPT region SNPs in Chamorro subjects. |D'| for the H1/H2 inversion polymorphism and nine MAPT SNPs, respectively, was 1.0 for SNPs 2 and 3; >0.8 for SNPs 1, 4–6 and 9; and >0.6 for SNPs 6 and 7.


View this table:
[in this window]
[in a new window]

  		Table 2. Allele frequencies by C17 inversion status and diagnosis group

 
We restricted all subsequent analyses to individuals who are homozygous for the H1 lineage, for two reasons. First, the Chamorro people of Guam have a much higher incidence of ALS than Caucasians and PDC-G does not have a clinico-pathological equivalent in Caucasians. Since Chamorros and Caucasians also have a different H1 frequency, a disease association involving SNPs that are associated with the inversion polymorphism could reflect the confounding effect of population admixture rather than a true disease association. Second, since inversions result in suppression of recombination, the H1 and H2 lineages represent evolutionarily and reproductively isolated regions of the genome even within a panmictic population. Therefore, analysis of just one of these lineages is necessary to permit evaluation of individual SNP effects separately from effects attributable to the inversion polymorphism. LD patterns among the nine MAPT region SNPs in 214 H1/H1 controls are shown in Table 3. LD was strongest (|D'|≥0.8 and/or r2≥0.4) between SNPs 1–2, 3–5, 6–7 and 8–9, and was relatively weak between the other SNPs.


View this table:
[in this window]
[in a new window]

  		Table 3. Pairwise LD among SNPs in Guam H1/H1 controls

 
Global differences in allele frequencies for SNPs 2 (P=0.007), 6 (P=0.011), 7 (P=0.055) and 9 (P=0.032) were observed among H1 homozygotes in the ALS-G, PDC-G, GD and control groups (Table 2). Allele frequencies did not differ significantly among the three case groups. Compared to controls, cases tended to have higher frequencies of the minor ‘A’ allele at SNP2, major ‘A’ allele at SNP6, major ‘A’ allele at SNP7 and major ‘C’ allele at SNP9. There was no evidence of Hardy–Weinberg disequilibrium (HWD) at any of the nine SNPs in H1/H1 controls or cases with PDC-G or GD. However, Hardy–Weinberg equilibrium in the 49 H1/H1 ALS-G cases was rejected at SNPs 3 (P=0.004), 4 (P=0.010) and 5 (P=0.023). Whereas allelic association tests are susceptible to false-positive results as a consequence of HWD, genotypic association tests remain valid (47). Global genotypic (Table 4) and allelic (Table 2) association screening tests yielded comparable results for SNPs 2, 6, 7 and 9. Exploratory analyses were also conducted for SNPs 3 and 5 based on the global genotype association test results, and possible univariate associations with ALS-G or PDC-G (data not shown, Supplementary Material, Table S1).


View this table:
[in this window]
[in a new window]

  		Table 4. Univariate logistic regression models of the risk of GD, PDC-G, or ALS-G associated with variants among chromosome 17 H1 homozygotes on Guam

 
Univariate logistic regression analyses showed that the risk of GD and PDC-G in Guam was significantly associated with SNPs 2, 6, 7 and 9 (Table 4). Evidence for association was also found for SNP3 and ALS-G (data not shown, Supplementary Material, Table S1), although the number of ALS-G cases was relatively small. The G allele of SNP2 in IMP5 conferred increased risk in a dominant manner, whereas the SNP5–7 A and SNP3 and 9 C alleles appear to confer risk in a recessive manner (Table 4, data not shown, Supplementary Material, Table S1). MAPT region SNPs were fit as either dominant or recessive genetic effects in higher order models for the sake of model parsimony. No significant evidence of an association was found between Apolipoprotein E (APOE) and ALS-G, PDC-G or GD in Guam (Table 2).

A fortuitous aspect of the data allowed us to infer that the associations observed for SNPs 6 and 7 could largely be explained by SNP6. All 181 H1/H1 individuals with the high-risk A/A genotype at SNP6 also had the high-risk A/A genotype at SNP7 with the exception of two PDC-G cases, who had the low-risk G/A genotype at SNP7. The SNP6 A allele was therefore a nearly perfect marker for the AA haplotype at SNPs 6 and 7 in this population. In contrast, 39/218 (17.9%) H1/H1 individuals with the high-risk A/A genotype at SNP7 had the low-risk G/A or G/G genotypes at SNP6. If the homozygous AA diplotype at SNPs 6 and 7 were truly responsible for disease, then using the SNP7 rather than SNP6 A/A genotype as a marker would lead to an attenuated estimate of the genetic effect because of the noise introduced by including 39 individuals with the low-risk GA/AA or GA/GA diplotypes at SNPs 6 and 7 into the ‘exposed’ category. Furthermore, including both SNPs 6 and 7 in the model would reduce the statistical significance of either effect because they are measures of the same underlying haplotype. The magnitude of the estimated risks of GD, PDC-G or ALS-G associated with SNP6 were stronger than SNP7 in both univariate (Table 4) and higher order models, and neither SNP was statistically significant when both were included in the model. These results indicated that the homozygous AA diplotype at SNPs 6 and 7, almost perfectly captured by the SNP6 A/A genotype, was associated with disease risk.

Combinations of two or more MAPT SNPs were analyzed to identify independent risk factors and interactions between genetic variants. Possession of the ‘G’ allele at SNP2 was an independent risk factor for both GD and PDC-G as well as for all cases combined (Table 5). Additionally, a significant interaction between the high-risk A/A and C/C genotypes at SNPs 6 and 9 was found for all three case groups and the combined set of cases. No significant three-way interaction was found, and APOE did not confound or weaken the MAPT effects.


View this table:
[in this window]
[in a new window]

  		Table 5. Risk of GD, PDC-G or ALS-G associated with MAPT region variants on Guam

 
The final genetic model for GD and PDC-G included SNP2 and the main effects and interaction of SNPs 6 and 9 (Table 5). When the final genetic model for GD and PDC-G was fit for ALS-G, a similar risk pattern emerged (Table 5). H1/H1 persons with the homozygous AC diplotype at SNPs 6 and 9 had a 3-fold [95% confidence interval (CI)=1.10–8.25] increased risk of GD, 4-fold (95% CI=1.40–11.64) increased risk of PDC-G and 6-fold (95% CI=1.44–32.14) increased risk of ALS-G, compared to persons who were not homozygous for this diplotype after adjusting for SNP2. H1/H1 carriers of the SNP2 G allele had a 1.6-fold (95% CI=1.00–2.62) increased risk of GD, 2-fold (95% CI=1.28–3.66) increased risk of PDC-G and 1.5-fold (95% CI=0.74–3.00) increased risk of ALS compared to non-carriers after adjusting for SNPs 6 and 9. In an analysis of all cases combined, the risk of developing GD, PDC-G or ALS-G was independently increased by 1.7-fold (95% CI=1.18–2.60) in carriers of the SNP2 G allele, and 3.8-fold (95% CI=1.64–8.78) in persons with the homozygous AC diplotype at SNPs 6 and 9.

The high-risk AC diplotype, or haplotype, defined by SNPs 6 and 9 is unlikely to represent the H1c haplotype reported to be associated with PSP and CBD (36). The frequency of H1 haplotypes containing this AC haplotype was significantly higher in the case samples (0.484–0.525) than in the control samples (0.355), as was the 3-SNP haplotype, TAC, that also includes SNP4 (frequency 0.415–0.437 in the case samples versus 0.296 in the control samples) (Supplementary Material, Table S3). Although the A allele of SNP6 is the allele on the H1c haplotype, estimation of 5-SNP haplotypes (including the three SNPs used to tag H1c) in the JPT HapMap sample shows that the frequency of the H1c haplotype is only 2.4% (Supplementary Material, Table S4). Also, the C allele of SNP9 perfectly predicts the existence of allele C for SNP rs2471738. Since the H1c haplotype has allele T for rs2471738, based on the JPT 5-SNP haplotype frequencies, the conditional probability that the SNP6–9 AC haplotype (or the SNP4–6–9 TAC haplotype) represents that the H1c haplotype is 0.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
METHODS
 SUPPLEMENTARY MATERIAL
 REFERENCES
 
The results described above show that genetic variation in the MAPT region contributes to risk for ALS-G, PDC-G and GD. Our previous work (32) showed an association between a single polymorphism (CA3662, Fig. 1) and ALS-G and PDC-G. Here, we confirm and extend our original findings by showing risk associated with additional SNPs in a larger sample. Also, these results extend our findings to GD, which was not previously studied. An intriguing novel finding is that the risk for Guam neurologic disorders is independently determined by SNPs in two different locations, one within MAPT and another upstream of MAPT in or near IMP5.

Genotypes within MAPT contribute to susceptibility to Guam neurologic disorders by a recessive mechanism. For SNPs 6, 7 and 9, homozygous genotypes (A/A, A/A and C/C, respectively) confer risk, whereas the other 2 genotypes for each site do not (Table 4). In studies of PSP, SNP6 genotypes also show a statistically significant association with risk (43,44), and in both our study and the work on PSP, the A allele is the high-risk allele. SNP6, located in the first intron of MAPT (Fig. 1), is in a region conserved across a number of mammalian species (43,48) suggesting a functional role in MAPT regulation. Rademaker et al. (43) showed that the 182 bp region containing SNP6 can act as a cis-acting transcriptional enhancer in a cell culture model and that the enhancer activity was greater when the SNP6 G allele was present compared to the A allele. Thus, for both Guam neurologic disorders and PSP, SNP6 may be the site, or at least one of the sites that determines disease risk, as opposed to some other close marker in LD with SNP6. This hypothesis is strengthened by the observation that even though the high-risk SNP6 allele is the same in the Chamarro as in other populations, the high-risk haplotype in the Chamarro population does not appear to be the H1c haplotype, previously reported to be associated with other neurological diseases (36). In fact, despite the higher frequency of ALS/PDC, the H1c haplotype appears to be much rarer in this population than in populations of European descent, which is inconsistent with a major role of H1c as a contributor to risk of ALS/PDC.

In our study, genotypes at SNP9 interact with SNP6 genotypes to increase risk. SNP9 is a non-synonymous change in exon 6, which is a coding exon. This exon is alternatively spliced and is present in approximately 20–40% of transcripts in the hippocampus and cerebellum, and approximately 10% of transcripts in the cortex (49). However, exon 6 sequences are not present in the aggregated pathologic tau. Thus, it is not clear at this point whether SNP9 functions in a regulatory fashion, or if it simply marks some other polymorphic site that is the true functional polymorphism. The regulation of MAPT expression is certainly complex. In addition to the enhancer sequence noted above, mutations that affect the regulation of alternative splicing have a profound effect on disease risk (29,30,50). Thus SNPs 6 and 9 may be in two different regulatory regions of MAPT. The fact that the genotypes interact suggests a cis-acting mechanism.

A novel finding of our work is that there is a second region marked by SNP2 that is 5' to MAPT and that acts independently of sites within MAPT to confer risk. Unlike SNPs 6 and 9, for SNP2, the G allele acts by a dominant mechanism. There are several possible hypotheses as to how SNP2 could influence risk for Guam disease. Perhaps, the most likely hypothesis is that SNP2 or an SNP in LD with SNP2 is in a cis-acting regulatory sequence that influences MAPT expression. Another related hypothesis is that SNP2 is in LD with polymorphic site in this region that influences the expression of another gene, provisionally called antisense to MAPT (aMAPT). This gene is transcribed in the opposite direction as MAPT and IMP5, has a transcription start site in MAPT intron 0, and extends past IMP5 (Fig. 1). aMAPT has multiple alternatively spliced forms with one exon overlapping with exon 0 of MAPT. This gene may act as a natural antisense regulatory sequence that controls MAPT expression (Ian D'Souza, personal communication). It is not clear at present whether the aMAPT sequence is translated into a protein. Thus, polymorphic sites 5' to MAPT could influence aMAPT expression, which in turn would influence MAPT expression. Another hypothesis is that the gene product of IMP5 is involved in Guam disease pathogenesis. SNP2 is a non-synonymous (H303R) change in the potential coding region of IMP5, a gene predicted from a number of unspliced expressed sequences in public databases. The predicted amino acid sequence for IMP5 has strong homology to signal peptide peptidases and weaker homology to presenilin1 and 2 (PSEN1 and PSEN2) (51). Because mutations in PSEN1 and PSEN2 cause AD, which is a tauopathy, variability at SNP2 could alter the function of IMP5 and thus influence risk for Guam disease. PSEN1 and PSEN2 are components of a transmembrane protease that causes the {gamma}-secretase cleavage of amyloid precursor protein (APP). This cleavage contributes to the production of an APP fragment called Aß, which is the primary component of the amyloid plaques in AD. Mutations in PSEN1 or PSEN2 alter the {gamma}-secretase cleavage site resulting in a more toxic form of Aß. However, this hypothesis seems unlikely for the following reasons. First, Aß deposition is not a feature of ALS-G or PDC-G. Second, there is no data available on whether IMP5 transcripts are translated to produce a protein. Third, IMP5 is not strongly homologous to the presenilins. Fourth, the active site aspartates in IMP5 are in opposite orientation to the presenilins making it unlikely that IMP5 would cleave APP and thus would not be involved in Aß production (51). Thus, it is more likely that the risk allele of SNP2 or another SNP in LD with SNP2 influences the regulation of MAPT and thus influences the disease process.

The SNP alleles described here are not likely to cause ALS-G, PDG-G or GD, but rather increase risk in combination with other genetic and environmental factors. This conclusion is based on the fact that not all affected subjects have the high-risk alleles, and not all carriers of the high-risk allele have the disease. Also, an earlier analysis of a component of a large pedigree from the village of Umatac found evidence against linkage of MAPT and ALS-G/PDC-G (32). The drop in prevalence of ALS-G and the increase in onset ages for both ALS-G and PDC-G over the past few decades suggest a change in exposure to an unknown environmental factor(s), and are not consistent with outbreeding diluting a genetic factor(s). Dietary BMAA (27,52) found in cycad seeds is one candidate for an environmental toxin. Our finding that picking, processing and eating cycads elevate risk for Guam disease (Borenstein, Mortimer, Dahlquist, Wu, Salmon, Gamst, Olichney, Thal, Silbert, Kaye, Adonay, Craig, Schellenberg and Galasko, submitted for publication) supports the hypothesis that BMAA may be an environmental susceptibility factor that potentially interacts with MAPT region alleles to increase risk for Guam disease. However, attempts to model neurodegeneration in animals by feeding BMAA-containing flour derived from cycad seeds, as prepared by Chamorros in Guam, required unrealistically large doses to produce an effect (53–55). A major environmental change in Guam is the post-World War II improvement in general diet. Perhaps, comparison of the environment in Guam to that of the Kii peninsula of Japan, where a cluster of very similar familial ALS and PDC cases occurs (21–25) will help identify common environmental and genetic factors.

The work presented here is the first genetic study of GD. Our working hypothesis is that GD is a variation of the same disease as ALS-G and PDC-G, and that the symptoms observed are more a function of when onset occurs rather than the underlying genetic and environmental architecture of these disorders. The fact that the same pattern of risk is observed for multiple SNPs for GD, ALS-G and PDC-G argues for a common disease mechanism involving tau. However, the underlying primary genetic architecture and environmental influence for each of these three Guam disorders is not known and could still be distinct for each, even though a common neurodegenerative pathway is the end result. The APOE and MAPT region results certainly argue that GD and AD are distinct. In all forms of Guam disease, the {varepsilon}4 and {varepsilon}2 APOE allele frequencies are not significantly different between cases and controls (Table 2). The remarkably low {varepsilon}4 frequency for controls (5%, Table 2) is consistent with other Asian populations where the {varepsilon}4 frequency is typically lower than Caucasian frequencies (56). In almost all AD populations, with few exceptions, {varepsilon}4 frequencies are elevated relative to controls, even when the control frequency is low (56). In most AD studies to date (57–64), with some exceptions (42,65–69), there appears to be no association between polymorphic sites in MAPT and AD. Results from the SNP2 region in AD cohorts have not been described. Arguing against the hypothesis that GD is distinct from AD is that these two disorders are not clinically distinguishable. Additional autopsies for GD cases are needed to establish the relationships between GD and ALS-G, and PDC-G and AD.

The observation that the same SNP (SNP6) influences risk for Guam neurologic disorders and PSP (43) and that other tauopathies such as FTD (37) and CBD (40,41) also show an association with MAPT region alleles, argues that these disorders at least partially share a common disease mechanism. The common feature of these diseases is aggregated tau occurring as NFTs, and in the case of PSP, also in glial cells. MAPT genetic variation may also play a role in typical Parkinson disease (PD) (46), which is not a tauopathy. The differences between these disorders are intriguing since the localization of neurodegenerative changes and the associated clinical symptoms vary remarkably between ALS-G, PDC-G, GD, PSP, PD and CBD. Presumably, the difference is the result of other unknown environmental and genetic factors, though the difference could also be due to the interactions between genetic variations in the IMP5 region versus variation within MAPT.


   METHODS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 METHODS
SUPPLEMENTARY MATERIAL
 REFERENCES
 
Subjects
The Chamorro subjects included in this study were recruited through a series of studies in Guam. Archival clinical data and DNA samples were obtained from the National Institutes of Neurological and Communicative Disorders and Stroke Intramural Research program in Guam (1956–83, Ralph Garruto) and from Kiyomitsu Oyanagi. Diagnoses of ALS and PDC were confirmed by examination of case records including autopsy reports, which were available on many patients. More recently, subjects in Guam were evaluated through the University of Guam–University of California, San Diego Consortium (1995–2006) as part of an NIH Program Project Grant. All subjects (or proxies in the case of demented patients who lacked decisional capacity) provided written informed consent to participate in the study, and the study was approved by Institutional Review Boards at the University of Guam, UCSD and University of Washington. Details regarding subject recruitment and medical evaluation were described previously (12). Briefly, all subjects were first evaluated by medical history, physical examination, and cognitive testing using the Cognitive Abilities Screening Instrument (CASI). Subjects who failed the CASI (cutoff score ≤75/100) or with a history or screening physical examination suggestive of neurologic disease underwent a further structured neurological examination that included the Unified Parkinson Disease Rating Scale, a standardized psychometric test battery, and blood tests. Neuroimaging with head CT or MRI was obtained when possible. Consensus diagnoses were made by 3 neurologists after reviewing all relevant information and using standard clinical diagnostic criteria wherever possible. The diagnosis of PDC required insidious onset and gradual progression of primary parkinsonism and dementia, either of which could be the initial feature. However, the onset of parkinsonism during mild-to-moderate stage of dementia was necessary to diagnose PDC. Parkinsonism first noted in patients with severe or end-stage dementia was considered a secondary feature of GD. Dementia was diagnosed using the Diagnostic and Statistical Manual, fourth edition (DSM-IV) criteria (American Psychiatric Association, 1994). The pure dementia syndrome in Guam resembled AD clinically, and many patients met the NINCDS/ADRDA criteria for probable AD (70). Patients with this dementia profile who had additional factors such as stroke and depression met criteria for possible AD. Controls were subjects without any neurologic disorder or dementia. Subjects classified as controls had a CASI score ≥80, were functionally independent, and lacked motor weakness, tremor or gait difficulty on a brief screening exam. Subjects with a CASI score <80 were classified as controls only if detailed neurological and psychometric evaluation showed no evidence of dementia, mild cognitive impairment or any neurological disorder.

Polymorphic loci and genotyping
Genomic DNA was isolated either from available frozen brain tissue or fresh blood samples from case and control subjects. APOE genotypes were determined by a modification of the method of Hixson and Vernier (71). Some SNPs in the MAPT region were genotyped as restriction-fragment-length polymorphisms (Table 1). For these loci, short DNA fragments containing the polymorphic sites were produced by polymerase chain amplification (PCR), digested with the appropriate restriction enzyme, the resulting fragments resolved using agarose gel electrophoresis, and fragments visualized with ethidium bromide. Locus del-in9 is an insertion/deletion polymorphism. PCR products spanning this locus differ by 238 bp depending on the presence or absence of the insertion. Size differences were observed using agarose gel electrophoresis. The remaining SNPs were genotyped using TaqMan allele discrimination assays (Applied Biosystems, CA, USA) (Table 1). SNPs were selected after initially screening 18 SNPs in a panel of 16 Chamorro controls (Supplementary Material, Table S2). SNPs were selected as follows. The del-in9 polymorphism was selected because it is a tagging site for the H1/H2 haplotype. SNPs 1 and 2 in genes CRHR1 and IMP5, respectively, were selected to follow-up previous work that showed that polymorphism CA3662 is associated with ALS-G and PDC-G (32), because these SNPs had minor allele frequencies in Chamorros >5%, and based on LD, provided non-redundant information. SNPs 3–9 in MAPT were selected based on work by others on the association between MAPT alleles and PSP (35,43,44), and because these SNPs had minor allele frequencies of >5% in a panel of 16 Chamorro controls.

Statistical analysis
A {chi}2 goodness-of-fit test was used to assess departure from Hardy–Weinberg equilibrium of the genotype frequencies at each SNP in controls and cases. Pairwise LD between SNPs was measured by estimating |D'|, the normalized disequilibrium coefficient (72) and the squared correlation coefficient r2. Fisher's exact test was used to perform global tests to detect allele or genotype frequency differences across three or more groups. Odds ratios (ORs) and 95% CIs were estimated using logistic regression. All analyses were conducted using Stata 9.1 (StataCorp LP, College Station, TX, USA).

To identify potentially interesting SNPs, we performed global tests of allele and genotype frequency differences among the four diagnosis groups (controls, ALS-G, PDC-G and GD). Global tests allow multiple groups to be compared using a single statistical test, thereby protecting against inflated false-positive rates due to multiple comparisons. A global significance level of ~10% was used to select SNPs for more detailed comparisons of the controls and each case group separately. Disease-specific models of the combined effects of multiple SNPs were constructed using the following procedure. First, univariate logistic regression was performed to identify SNP genotypes associated (P≤0.05) with a particular disease. Pairs of associated SNPs and their interactions were then included in logistic regression models. SNPs involved in a significant interaction or that remained significant in two-locus models were then assessed in higher order models, and were included in the final model for the disease if they retained their significance. Age, gender and APOE genotype were assessed as potential confounders using a 10% change in the ORs for any genetic effect as the criterion for retention in the final model.

ALS-G, PDC-G and GD in Guam may represent pleiotropic effects or variable expressivity of a single altered gene as originally hypothesized for ALS-G and PDC-G by Plato et al. (18), or a single set of altered genes. The power of our analyses to detect associations with ALS-G alone was limited because DNA was available from only 54 ALS-G cases. Therefore, the best genetic model as determined by analyses of PDC-G and GD was also fit using only ALS-G cases, as well as all three disease groups combined, compared to the control group.

Haplotype frequencies were estimated with an EM algorithm by the program HAPFREQS (73) in individuals homozygous for the H1 inversion polymorphism. Frequencies were estimated separately in the controls, in each case sample, and in the combined sample, as well as, for comparison, in the HapMap Japanese (JPT) sample as the closest match to Guam in ethnic origin. Haplotype frequencies in the Guam samples were estimated for SNPs 6 and 9, and also for SNPs 4, 6 and 9. The latter 3-SNP haplotypes were used as a surrogate for the three SNPs previously reported to tag the H1c haplotype associated with PSP and CBD (36): rs1467967 (7.2 kb proximal to SNP4), rs242557 (SNP6) and rs2471738 (8.7 kb distal to SNP9). To compare haplotype status of the high-risk haplotype found in the current study with the H1c haplotype, haplotype frequencies were estimated in the JPT sample for all five SNPs. From these frequencies, we computed the conditional probability of haplotype H1c (AxAxT) as P(AxAxT | xTACx) and P(AxAxT | xxACx), where x indicates any SNP allele in a particular haplotype position, A, T,G and C are the particular SNP alleles, and the SNPs are indicated in their chromosomal order: rs1467967–SNP4–SNP6–SNP9–rs2471738. Similar analysis using the Caucasian (CEU) HapMap sample lead to the same conclusion regarding the conditional probability of the H1c haplotype (data not shown).


   SUPPLEMENTARY MATERIAL
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 METHODS
 SUPPLEMENTARY MATERIAL
REFERENCES
 
Supplementary Material is available at HMG Online.


   ACKNOWLEDGEMENTS
 
This work was supported by NIA grant PO1 AG14382, P50 AG 05136, and by the Department of Veterans Affairs. The authors are grateful for the efforts of research assistants and clinicians who have contributed to identifying and evaluating patients in Guam over the years.

Conflict of Interest statement. None of the authors have any conflict of interest.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 METHODS
 SUPPLEMENTARY MATERIAL
 REFERENCES
 

  1. Kurland L.T. and Mulder D.W. (1954) Epidemiologic investigations of amyotrophic lateral sclerosis. 1. Preliminary report on the geographic distribution, with special reference to the Mariana Islands, including clinical and pathologic observations. Neurology 4:355–378.[Free Full Text]

  2. Rodgers-Johnson P., Garruto R.M., Yanagihara R., Chen K.-M., Gajdusek D.C., Gibbs C.J. (1986) Amyotrophic lateral sclerosis and parkinsonism-dementia complex on Guam: a 30-year evaluation of clinical and neuropathologic trends. Neurology 36:7–13.[Abstract/Free Full Text]

  3. Hirano A., Arumugasamy N., Zimmerman H.M. (1967) Amyotrophic lateral sclerosis. A comparison of Guam and classical cases. Arch. Neurol. 16:357–363.[ISI][Medline]

  4. Hirano A., Malamud N., Elizan T.S., Kurland L.T. (1966) Amyotrophic lateral sclerosis and Parkinsonism-dementia complex on Guam. Further pathologic studies. Arch. Neurol. 15:35–51.[ISI][Medline]

  5. Hof P.R., Nimchinsky E.A., Buee-Scherrer V., Buee L., Nasrallah J., Hottinger A.F., Purohit D.P., Loerzel A.J., Steele J.C., Delacourte A., et al. (1994) Amyotrophic lateral sclerosis/parkinsonism-dementia complex of Guam: quantitative neuropathology, immunohistochemical analysis of neuronal vulnerability, and comparison with related neurodegenerative disorders. Acta. Neuropathol. 88:397–404.[Medline]

  6. Malamud N., Hirano A., Kurland L.T. (1961) Pathoanatomic changes in amyotrophic lateral sclerosis on Guam. Neurology 5:401–414.

  7. Hirano A., Kurland L.I.T., Krooth R.S., Lessell S. (1961) Parkinson-dementia complex, an endemic disease on the island of Guam. I. Clinical features. Brain 84:642–661.[Free Full Text]

  8. Hof P.R. and Perl D.P. (2002) Neurofibrillary tangles in the primary motor cortex in Guamanian amyotrophic lateral sclerosis/parkinsonism-dementia complex. Neurosci. Lett. 328:294–298.[CrossRef][ISI][Medline]

  9. Hirano A., Malamud N., Kurland L.I.T. (1961) Parkinson-dementia complex, an endemic disease on the island of Guam. II. Pathological features. Brain 84:662–679.[Free Full Text]

  10. Matsumoto S., Hirano A., Goto S. (1990) Spinal cord neurofibrillary tangles of Guamanian amyotrophic lateral sclerosis and parkinsonism-dementia complex: an immunohistochemical study. Neurology 40:975–979.[Abstract/Free Full Text]

  11. Kato S., Hirano A., Llena J.F., Ito H., Yen S.H. (1992) Ultrastructural identification of neurofibrillary tangles in the spinal cords in Guamanian amyotrophic lateral sclerosis and parkinsonism-dementia complex on Guam. Acta. Neuropathol. 83:277–282.[CrossRef][Medline]

  12. Galasko D., Salmon D., Craig U.K., Perl D.P., Schellenberg G. (2002) Clinical features and changing patterns of neurodegenerative disorders on Guam, 1997–2000. Neurology 59:1121.[Free Full Text]

  13. Galasko D., Salmon D.P., Olichney J., et al. (2003) Diverse types of pathology underlie dementia in older Chamorros on Guam. Neurology 60:Suppl. 1, A329–A330.

  14. Koerner D.R. (1952) Amyotrophic lateral sclerosis on Guam: A clinical study and review of the literature. Ann. Int. Med. 37:1204–1220.[ISI][Medline]

  15. Kurland L.T. and Mulder D.W. (1955) Epidemiologic investigations of amyotrophic lateral sclerosis. 2. Familial aggregations indicative of dominant inheritance. Part II. Neurology 5:249–268.[Free Full Text]

  16. Kurland L.T. and Mulder D.W. (1955) Epidemiologic investigations of amyotrophic lateral sclerosis. 2. Familial aggregations indicative of dominant inheritance. Part I. Neurology 5:182–196.[Medline]

  17. Kurland L.T. and Mulder D.W. (1955) Epidemiologic investigations of amyotrophic lateral sclerosis. Neurology 4:438–448.[Medline]

  18. Plato C.C., Cruz M.T., Kurland L.T. (1969) Amyotrophic lateral sclerosis/Parkinsonism dementia complex of Guam: further genetic investigations. Am. J. Hum. Genet. 21:133–141.[ISI][Medline]

  19. Bailey-Wilson J.E., Plato C.C., Elston R.C., Garruto R.M. (1993) Potential role of an additive genetic component in the cause of amyotrophic lateral sclerosis and Parkinsonism-dementia in the Western Pacific. Am. J. Med. Genet. 45:68–76.[CrossRef][ISI][Medline]

  20. Plato C.C., Galasko D., Garruto R.M., Plato M., Gamst A., Craig U.K., Torres J.M., Wiederholt W. (2002) ALS and PDC of Guam—forty-year follow-up. Neurology 58:765–773.[Abstract/Free Full Text]

  21. Itoh N., Ishiguro K., Arai H., Kokubo Y., Sasaki R., Narita Y., Kuzuhara S. (2003) Biochemical and ultrastructural study of neurofibrillary tangles in amyotrophic lateral sclerosis/parkinsonism-dementia complex in the Kii peninsula of Japan. J. Neuropath. Exp. Neurol. 62:791–798.[ISI][Medline]

  22. Kokubo Y. and Kuzuhara S. (2004) Neurofibrillary tangles in ALS and parkinsonism-dementia complex focus in Kii, Japan. Neurology 63:2399–2401.[Abstract/Free Full Text]

  23. Kokubo Y., Kuzuhara S., Narita Y. (2000) Geographic distribution of amyotrophic lateral sclerosis with neurofibrillary tangles in the Kii peninsula of Japan. J. Neurol. 415:850–852.[CrossRef]

  24. Kuzuhara S., Kokubo Y., Sasaki R., Narita Y., Yabana T., Hasegawa M., Iwatsubo T. (2001) Familial amyotrophic lateral sclerosis and parkinsonism-dementia complex of the Kii peninsula of Japan: clinical and neuropathological study and tau analysis. Ann. Neurol. 49:501–511.[CrossRef][ISI][Medline]

  25. Yase Y. (1970) Neurologic disease in the western Pacific islands, with a report on the foci of amyotrophic lateral sclerosis found in the Kii peninsula, Japan. Am. J. Trop. Med. Hyg. 19:155–166.[ISI][Medline]

  26. Garruto R.M., Yanagihara R., Gajdusek D.C. (1985) Disappearance of high-incidence amyotrophic lateral sclerosis and parkinsonism-dementia on Guam. Neurology 35:193–198.[Abstract/Free Full Text]

  27. Spencer P.S., Nunn P.B., Hugon J., Ludolph A.C., Ross S.M., Roy D.N., Robertson R.C. (1987) Guam amyotrophic lateral sclerosis-parkinsonism-dementia linked to plan excitant Neurotoxin. Science 237:517–522.[Abstract/Free Full Text]

  28. Poorkaj P., Bird T.D., Wijsman E., Nemens E., Garruto R.M., Anderson L., Andreadis A., Wiederholt W.C., Raskind M., Schellenberg G.D. (1998) Tau is a candidate gene for chromosome 17 frontotemporal dementia. Ann. Neurol. 43:815–825.[CrossRef][ISI][Medline]

  29. Spillantini M.G., Murrell J.R., Goedert M., Farlow M.R., Klug A., Ghetti B. (1998) Mutation in the tau gene in familial multiple system tauopathy with presenile dementia. Proc. Natl. Acad. Sci. U.S.A. 95:7737–7741.[Abstract/Free Full Text]

  30. Hutton M., Lendon C.L., Rizzu P., Baker M., Froelich S., Houlden H., Pickering-Brown S., Chakraverty S., Isaacs A., Grover A., et al. (1998) Association of missense and 5'-splice-site mutations in tau with the inherited dementia FTDP-17. Nature 393:702–705.[CrossRef][Medline]

  31. Perez-Tur J., Buee L., Morris H.R., Waring S.C., Onstead L., Wavrant-De Vrieze F., Crook R., Buee-Scherrer V., Hof P.R., Petersen R.C., et al. (1999) Neurodegenerative diseases of Guam: analysis of Tau. Neurology 53:411–413.[Abstract/Free Full Text]

  32. Poorkaj P., Tsuang D., Wijsman E.M., Nemens E., Garruto R.M., Craig U., Anderson L.-J., Bird T.D., Plato C.C., Weiderholt W., et al. (2001) Tau is a susceptibility gene for amyotropic lateral sclerosis-parkinsonism dementia complex of Guam. Arch. Neurol. 58:1871–1878.[Abstract/Free Full Text]

  33. Morris H.R., Steele J.C., Crook R., Wavrant-De Vrieze F., OnsteadCardinale L., GwinnHardy K., Wood N.W., Farrer M., Lees A.J., McGeer P.L., et al. (2004) Genome-wide analysis of the Parkinsonism-dementia complex of Guam. Arch. Neurol. 61:1889–1897.[Abstract/Free Full Text]

  34. Conrad C., Andreadis A., Trojanowski J., Dickson D., Kang D., Chen X., Wiederholt W., Hansen L., Masliah E., Thal L., et al. (1997) Genetic evidence for the involvement of Tau in progressive supranuclear palsy. Ann. Neurol. 41:277–281.[CrossRef][ISI][Medline]

  35. Pastor P., Ezquerra M., Perez J.C., Chakraverty S., Norton J., Racette B.A., McKeel D., Perlmutter J.S., Tolosa E., Goate A.M. (2004) Novel haplotypes in 17q21 are associated with progressive supranuclear palsy. Ann. Neurol. 56:249–258.[CrossRef][ISI][Medline]

  36. Pittman A.M., Myers A.J., Duckworth J., Bryden L., Hanson M., Abou-Sleiman P., Wood N.W., Hardy J., Lees A., de Silva R. (2004) The structure of the tau haplotype in controls and in progressive supranuclear palsy. Hum. Mol. Genet. 13:1267–1274.[Abstract/Free Full Text]

  37. Hughes A., Mann D., Pickering-Brown S. (2003) Tau haplotype frequency in frontotemporal lobar degeneration and amyotrophic lateral sclerosis. Exp. Neurol. 181:12–16.[CrossRef][ISI][Medline]

  38. Baker M., Litvan I., Houlden H., Adamson J., Dickson D., Perez-Tur J., Hardy J., Lynch T., Bigio E., Hutton M. (1999) Association of an extended haplotype in the tau gene with progressive supranuclear palsy. Hum. Mol. Genet. 8:711–715.[Abstract/Free Full Text]

  39. Higgins J.J., Golbe L.I., Debiase A., Jankovic J., Factor S.A., Adler R.L. (2000) An extended 5'-tau susceptibility haplotype in progressive supranuclear palsy. Neurology 55:1364–1367.[Abstract/Free Full Text]

  40. Di Maria E., Tabaton M., Vigo T., Abbruzzese G., Bellone E., Donati C., Frasson E., Marchese R., Montagna P., Munoz D.G., et al. (2000) Corticobasal degeneration shares a common genetic background with progressive supranuclear palsy. Ann. Neurol. 47:374–377.[CrossRef][ISI][Medline]

  41. Houlden H., Baker M., Morris H.R., MacDonald N., Pickering-Brown S., Adamson J., Lees A.J., Rossor M.N., Quinn N.P., Kertesz A., et al. (2001) Corticobasal degeneration and progressive supranuclear palsy share a common tau haplotype. Neurology 56:1702–1706.[Abstract/Free Full Text]

  42. Myers A.J., Kaleem M., Marlowe L., Pittman A.M., Lees A.J., Fung H.C., Duckworth J., Leung D., Gibson A., Morris C.M., et al. (2005) The H1c haplotype at the MAPT locus is associated with Alzheimer's disease. Hum. Mol. Genet. 14:2399–2404.[Abstract/Free Full Text]

  43. Rademakers R., Melquist S., Cruts M., Theuns J., Del-Favero J., Poorkaj P., Baker M., Sleegers K., Crook R., De Pooter T., et al. (2005) High-density SNP haplotyping suggests altered regulation of tau gene expression in progressive supranuclear palsy. Hum. Mol. Genet. 14:3281–3292.[Abstract/Free Full Text]

  44. Pittman A.M., Myers A.J., Abou-Sleiman P., Fung H.C., Kaleem M., Marlowe L., Duckworth J., Leung D., Williams D., Kilford L., et al. (2005) Linkage disequilibrium fine mapping and haplotype association analysis of the tau gene in progressive supranuclear palsy and corticobasal degeneration. J. Med. Genet. 42:837–846.[Abstract/Free Full Text]

  45. Stefansson H., Helgason A., Thorleifsson G., Steinthorsdottir V., Masson G., Barnard J., Baker A., Jonasdottir A., Ingason A., Gudnadottir V.G., et al. (2005) A common inversion under selection in Europeans. Nat. Genet. 37:129–137.[CrossRef][ISI][Medline]

  46. Skipper L., Wilkes K., Toft M., Baker M., Lincoln S., Hulihan M., Ross O.A., Hutton M., Aasly J., Farrer M. (2004) Linkage disequilibrium and association of MAPT H1 in Parkinson disease. Am. J. Hum. Genet. 75:669–677.[CrossRef][ISI][Medline]

  47. Schaid D.J. and Jacobsen S.J. (1999) Biased tests of association: comparisons of allele frequencies when departing from Hardy–Weinberg proportions. Am. J. Epidemiol. 149:706–711.[Abstract/Free Full Text]

  48. Poorkaj P., Kas A., D'Souza I., Zhou Y., Pham O., Olson M.V., Schellenberg G.D. (2001) A genomic sequence analysis of the mouse and human microtubule-associated protein tau. Mamm. Genome 12:700–712.[CrossRef][ISI][Medline]

  49. Andreadis A. (2005) Tau gene alternative splicing: expression patterns, regulation and modulation of function in normal brain and neurodegenerative diseases. Biochim. Biophys. Acta. 1739:91–103.[Medline]

  50. D'Souza I., Poorkaj P., Hong M., Nochlin D., Lee V.M.Y., Bird T.D., Schellenberg G.D. (1999) Missense and silent tau gene mutations cause front temporal dementia with parkinsonism - chromosome 17 type by affecting multiple alternative RNA splicing regulatory elements. Proc. Natl Acad. Sci. USA 96:5598–5603.[Abstract/Free Full Text]

  51. Friedmann E., Lemberg M.K., Weihofen A., Dev K.K., Dengler U., Rovelli G., Martoglio B. (2004) Consensus analysis of signal peptide peptidase and homologous human aspartic proteases reveals opposite topology of catalytic domains compared with presenilins. J. Biol. Chem. 279:50790–50798.[Abstract/Free Full Text]

  52. Cox P.A. and Sacks O.W. (2002) Cycad neurotoxins, consumption of flying foxes, and ALS-PDC disease in Guam. Neurology 58:956–959.[Abstract/Free Full Text]

  53. Duncan M.W., Steele J.C., Kopin I.J., Markey S.P. (1990) 2-amino-3-(methylamino)-propanoic acid (BMAA) in cycad flour: an unlikely cause of amyotrophic lateral sclerosis and parkinsonism-dementia of Guam. Neurology 40:767–772.[Abstract/Free Full Text]

  54. Garruto R.M., Yanagihara R., Gajdusek D.C. (1988) Cycads and amyotrophic lateral sclerosis/parkinsonism dementia. Lancet ii:1079.

  55. Duncan M., Kopin I.J., Garruto R.M., Lavine L., Markey S.P. (1988) 2-Amino-3 (methylamino)-propionic acid in cycad-derived foods is an unlikely cause of amyotrophic lateral sclerosis/parkinsonism. Lancet ii:631–632.

  56. Farrer L.A., Cupples L.A., Haines J.L., Hyman B., Kukull W.A., Mayeux R., Myers R.H., Pericak-Vance M.A., Risch N., Van Duijn C.M. (1997) Effects of age, sex, and ethnicity on the association between apolipoprotein E genotype and Alzheimer disease: a meta-analysis. JAMA 278:1349–1356.[Abstract]

  57. Clark L.N., Levy G., Tang M.X., Mejia-Santana H., Ciappa A., Tycko B., Cote L.J., Louis E.D., Mayeux R., Marder K. (2003) The Saitohin ‘Q7R’ polymorphism and tau haplotype in multi-ethnic Alzheimer disease and Parkinson's disease cohorts. Neurosci. Lett. 347:17–20.[CrossRef][ISI][Medline]

  58. Russ C., Powell J.F., Zhao J.H., Baker M., Hutton M., Crawford F., Mullan M., Roks G., Cruts M., Lovestone S. (2001) The microtubule associated protein Tau gene and Alzheimer's disease - an association study and meta-analysis. Neurosci. Lett. 314:92–96.[CrossRef][ISI][Medline]

  59. Green E.K., Thaker U., McDonagh A.M., Iwatsubo T., Lambert J.C., Chartier Harlin M.C., Harris J.M., Pickering-Brown S.M., Lendon C.L., Mann D.M.A. (2002) A polymorphism within intron 11 of the tau gene is not increased in frequency in patients with sporadic Alzheimer's disease, nor does it influence the extent of tau pathology in the brain. Neurosci. Lett. 324:113–116.[CrossRef][ISI][Medline]

  60. Streffer J.R., Papassotiropoulos A., Kurosinski P., Signorell A., Wollmer M.A., Tsolaki M., Iakovidou V., Horndli F., Bosset J., Gotz J., et al. (2003) Saitohin gene is not associated with Alzheimer's disease. J. Neurol. Neurosurg. Psychiatry 74:362–363.[Abstract/Free Full Text]

  61. Verpillat P., Richard S., Hannequin D., Dubois B., Bou J., Camuzat A., Pradier L., Frebourg T., Brice A., Clerget-Darpoux F., et al. (2002) Is the saitohin gene involved in neurodegenerative diseases? Ann. Neurol. 52:829–832.[CrossRef][ISI][Medline]

  62. Oliveira S.A., Martin E.R., Scott W.K., Nicodemus K.K., Small G.W., Schmechel D.E., Doraiswamy P.M., Roses A.D., Saunders A.M., Gilbert J.R., et al. (2003) The Q7R Saitohin gene polymorphism is not associated with Alzheimer disease. Neurosci. Lett. 347:143–146.[CrossRef][ISI][Medline]

  63. Cook L., Brayne C.E., Easton D., Evans J.G., Xuereb J., Cairns N.J., Rubinsztein D.C. (2002) No evidence for an association between Saitohin Q7R polymorphism and Alzheimer's disease. Ann. Neurol. 52:690–691.[ISI][Medline]

  64. Kwon J.M., Nowotny P., Shah P.K., Chakraverty S., Norton J., Morris J.C., Goate A.M. (2000) Tau polymorphisms are not associated with Alzheimer's disease. Neurosci. Lett. 284:77–80.[CrossRef][ISI][Medline]

  65. Tanahashi H., Asada T., Tabira T. (2004) Association between tau polymorphism and male early-onset Alzheimer's disease. Neuroreport 15:175–179.[CrossRef][ISI][Medline]

  66. Combarros O., Rodero L., Infante J., Palacio E., Llorca J., Fernandez-Viadero C., Pena N., Berciano J. (2003) Age-dependent association between the Q7R polymorphism in the saitohin gene and sporadic Alzheimer's disease. Dement. Geriat. Cog. Disord. 16:132–135.

  67. Conrad C., Vianna C., Freeman M., Davies P. (2002) A polymorphic gene nested within an intron of the tau gene: implications for Alzheimer's disease. Proc. Natl Acad. Sci. USA 99:7751–7756.[Abstract/Free Full Text]

  68. Bullido M.J., Aldudo J., Frank A., Coria F., Avila J., Valdivieso F. (2000) A polymorphism in the tau gene associated with risk for Alzheimer's disease. Neurosci. Lett. 278:49–52.[CrossRef][ISI][Medline]

  69. Peplonska B., Zekanowski C., Religa D., Czyzewski K., Styczynska M., Pfeffer A., Gabryelewicz T., Golebiowski M., Luczywek E., Wasiak B., et al. (2003) Strong association between Saitohin gene polymorphism and tau haplotype in the Polish population. Neurosci. Lett. 348:163–166.[CrossRef][ISI][Medline]

  70. McKhann G., Drachman D., Folstein M., Katzman R., Price D., Stadlan E.M. (1984) Clinical diagnosis of Alzheimer's disease: report on the NINCDS-ADRDA work group under the auspices of the Department of Human Services Task Force on Alzheimer's disease. Neurology 34:939–944.[Abstract/Free Full Text]

  71. Hixson J.E. and Vernier D.T. (1990) Restriction isotyping of human apolipoprotien E by gene amplification and cleavage with HhaI. J. Lipid Res. 31:545–548.[Abstract]

  72. Lewontin R.C. (1988) On measures of gametic disequilibrium. Genetics 120:849–852.[Abstract/Free Full Text]

  73. Goddard K.A.B., Yu C.E., Oshima J., Miki T., Nakura J., Piussan C., Martin G.M., Schellenberg G.D., Wijsman E.M., Brown W.T., et al. (1996) Toward localization of the Werner syndrome gene by linkage disequilibrium and ancestral haplotyping: lessons learned from analysis of 35 chromosome 8p11.1-21.1 markers. Am. J. Hum. Genet. 58:1286–1302.[ISI][Medline]


Add to CiteULike CiteULike