Article Text

Download PDFPDF

Saitohin gene is not associated with Alzheimer’s disease
  1. J R Streffer1,
  2. A Papassotiropoulos1,
  3. P Kurosinski1,
  4. A Signorell1,
  5. M A Wollmer1,
  6. M Tsolaki2,
  7. V Iakovidou2,
  8. F Hörndli1,
  9. J Bosset1,
  10. J Götz1,
  11. R M Nitsch1,
  12. C Hock1
  1. 1Division of Psychiatry Research, University of Zurich, Switzerland
  2. 2Third Department of Neurology, Aristotle University of Thessaloniki, Thessaloniki, Greece
  1. Correspondence to: Dr Johannes Streffer, Division of Psychiatry Research, University of Zürich, Lenggstrasse 31, 8029 Zürich, Switzerland; streffer{at}


Background: The deposition of tau protein in neurofibrillary tangles constitutes an important feature of many neurodegenerative disorders, including Alzheimer’s disease. A polymorphic gene, saitohin (STH), nested within the tau gene (microtubule associated protein tau, MAPT), was recently identified and an association of a non-synonymous polymorphism in STH with increased risk for Alzheimer’s disease was suggested.

Objective and methods: To test the above hypothesis in a case–control association study of two independent white populations within Switzerland and Greece, comparing genotype and allele frequencies from 225 Alzheimer’s disease patients and 144 healthy control subjects.

Results: No differences in allelic or genotypic distributions between Alzheimer’s disease patients and controls was found in the individual samples (Swiss/Greek) or in the combined sample. Stratification for the presence of apolipoprotein E (APOE) ε4 allele, sex, or age did not show significant effects in the populations studied, nor was there an effect on the age of onset.

Conclusions: No evidence was found for an association of the non-synonymous polymorphism (Q7R) in STH and Alzheimer’s disease. This finding is in line with earlier studies showing no association between MAPT and Alzheimer’s disease.

  • tau gene
  • Alzheimer’s disease
  • saitohin

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Recently, Conrad et al identified a gene nested within an intron of the tau gene (microtubule associated protein tau, MAPT), termed saitohin (STH).1 STH encodes for a 128 amino acid protein and is located in intron 9 of MAPT (2.5 kb downstream of exon 9), a critical region for alternative splicing of exon 10. Splicing in of exon 10 results in a protein with four repeated microtubule binding domains (4R), in contrast to tau without the exon 10, which displays only three microtubule binding domains (3R). Distribution of STH mRNA in human tissue is similar to MAPT, and high in muscle, placenta, and fetal and adult brain. To date, no specific function is known for STH. However, putative phosphorylation sites for protein kinase C and casein kinase II have been identified, which may be in line with a combined function of STH and MAPT1; a non-synonymous polymorphism (Q7R; CAA to CGA) was used to carry out a case–control association study, which suggested overrepresentation of the GG genotype, as well as the G allele, in Alzheimer’s disease patients compared with controls.

Deposition of hyperphosphorylated tau in neurofibrillary tangles is a central feature of Alzheimer’s disease,2 the most common form of dementia in the elderly,3 and other neurodegenerative diseases, termed tauopathies, including frontotemporal dementia with Parkinsonism linked to chromosome 17 (FTDP-17), Pick’s disease, progressive supranuclear palsy, and corticobasal degeneration.4 An imbalance of the different tau isoforms is associated with the pathological changes in these diseases. Familial forms of these tauopathies have been shown to be caused by intronic and exonic mutations in the MAPT gene5,6 affecting mainly the region around the alternatively spliced exon 10. So far, studies examining potential associations between Alzheimer’s disease and polymorphisms in MAPT remain controversial.7–13 In this study we tested whether the STH polymorphism was associated with risk for Alzheimer’s disease, because first, STH and tau are in close vicinity, and second, the allelic variance is positioned in a region involved in the alternate splicing of MAPT, and thus also involving the regulation of the ratio between 3R and 4R tau isoforms.


We examined 225 patients with Alzheimer’s disease (according to NINCDS-ADRDA criteria14) and 144 controls (tested with CERAD battery and MMS15) in two white populations (Swiss: 91 Alzheimer patients and 92 healthy controls; Greece: 134 Alzheimer patients and 52 controls) and undertook a case–control association study for the non-synonymous polymorphism in STH (Q7R; CAA to CGA).

DNA was isolated from whole blood with a DNA extraction kit (Qiagen, Hilden, Germany). Genotyping for STH was done with the PSQTM 96 system (Pyrosequencing AB, Uppsala, Sweden) (primer forward, GCC CCT GTA AAC TCT GAC CA; primer reverse, GCT GAG GGT TCT GTC TCT GG, pyrosequencing primer, GTG AGG GTG GAG GC; for details see APOE genotypes were assessed using the LightCycler instrument (Roche Diagnostics Corporation, Mannheim, Germany).16 Pipetting was done with a Genesis Workstation 200/8 using Gemini v3.5 software (TECAN).


For statistical analysis, we used Fisher’s exact test, Pearson’s χ2 (testing association), and the Kolmogorov–Smirnov test (testing distributions between two groups). A probability (p) value of < 0.05 was regarded as significant. Non-significant p values are not shown. The study was approved by the local ethics committee and informed consent was obtained before the investigation.


The allelic distribution for STH in the combined sample was 78.59% (A) and 21.41% (G), and the genotypic distribution was 61.8% (A/A), 33.6% (A/G), and 4.6% (G/G) (table 1, which also includes single population data), following Hardy–Weinberg equilibrium in both populations. Based on the genotype frequencies reported by Conrad et al,1 our study had a statistical power of more than 95% for α = 0.05. In contrast to Conrad et al, we did not detect any significant difference in the distribution of genotypes or alleles between the patients with Alzheimer’s disease patients and the controls.

Table 1

Analysis of 225 patients with Alzheimer’s disease and 144 healthy control subjects

In previous studies associations of single nucleotide polymorphisms assigned to MAPT with Alzheimer’s disease were found in APOE ε4 carriers.7,11 We therefore stratified our sample for APOE ε4 carriers. As previously described, the APOE ε4 allele was significantly associated with the risk of Alzheimer’s disease in our samples (Pearson’s χ2 = 20.44; p = 2×10−6). Stratification in APOE ε4 carriers and non-carriers did not reveal any significant differences between the two groups, nor did stratification for age (≥ 70 years or < 70 years at the age of onset) or sex. Finally, to test whether STH may act as a progression marker, we determined the age of onset by Kaplan–Meier analysis, as well as the mean age of onset, and found no differences between the genotypes (AA: mean age 72 years; 95% confidence interval (CI), 71 to 73; AG: mean age 71 years; 95% CI, 69 to 73; GG: mean age 73 years; 95% CI, 70 to 76).


Our findings are in line with two recent reports,17,18 both showing no association between STH polymorphism and Alzheimer’s disease. In addition, Verpillat and colleagues18 found a complete linkage of the Q genotype (A allele) with the H1 tau haplotype in 91 patients with frontotemporal dementia. They even concluded that the STH Q genotype could be part of the extended H1 tau haplotype. In the controversial discussion of the association between Alzheimer’s disease and polymorphisms in MAPT,7–13 these three studies (Cook et al, 2002,17 Verpillat et al, 2002,18 and our own data) argue strongly in favour of the absence of an association between variations in MAPT and the risk of developing Alzheimer’s disease.


This study, like others, fails to support a significant association between STH and Alzheimer’s disease. Stratification did not reveal effects in subpopulations or special risk groups. The promising gene saitohin (STH) (by its proposed function and location) is not associated with the risk of Alzheimer’s disease. Further analysis should be directed towards sporadic tauopathies.


We thank Mr Jay Tracy for careful reading and preparation of the manuscipt, and Ms Esmeralda Garcia, Ms Christin Wilde, Ms Andrea Hauer, and Ms Estelle Obrist for patient care and sampling. This work was supported by grants of the National Center for Competence in Research (NCCR) “Neuronal Plasticity and Repair”, the Swiss Science National Foundation (32-65869.01), and the Roche Research Foundation (22-2001).