Identification of independent APP locus duplication in Japanese patients with early-onset Alzheimer disease
- K Kasuga1,2,
- T Shimohata2,
- A Nishimura3,
- A Shiga1,2,
- T Mizuguchi3,
- J Tokunaga2,
- T Ohno4,
- A Miyashita5,
- R Kuwano5,
- N Matsumoto3,
- O Onodera1,
- M Nishizawa2,
- T Ikeuchi1
- 1Department of Molecular Neuroscience, Bioresource Science Branch, Center for Bioresources, Brain Research Institute, Niigata University, Niigata, Japan
- 2Department of Neurology, Bioresource Science Branch, Center for Bioresources, Brain Research Institute, Niigata University, Niigata, Japan
- 3Department of Human Genetics, Yokohama City University, Graduate School of Medicine, Yokohama, Japan
- 4Department of Neurology, Nagaoka Chuo General Hospital, Nagaoka, Japan
- 5Department of Molecular Genetics, Bioresource Science Branch, Center for Bioresources, Brain Research Institute, Niigata University, Niigata, Japan
- Correspondence to Dr T Ikeuchi, Department of Molecular Neuroscience, Brain Research Institute, Niigata University, 1 Asahimachi, Niigata 951-8585, Japan;
- Received 28 August 2008
- Revised 7 October 2008
- Accepted 7 October 2008
Background: The occurrence of duplications of the amyloid precursor protein gene (APP) has been described in European families with early-onset familial Alzheimer disease (EO-FAD) and cerebral amyloid angiopathy. However, the contribution of APP duplication to the development of AD in other ethnic populations remains undetermined.
Methods: The occurrence of APP duplication in probands from 25 families with FAD and 11 sporadic EO-AD cases in the Japanese population was examined by quantitative PCR and microarray-based comparative genomic hybridisation analyses. APP expression level was determined by real-time quantitative reverse-transcription (RT) PCR analysis using mRNA extracted from the peripheral blood of the patients.
Results: We identified APP locus duplications in two unrelated EO-FAD families. The duplicated genomic regions in two patients of these families differed from each other. No APP duplication was found in the late-onset FAD families or sporadic EO-AD patients. The patients with APP duplication developed insidious memory disturbance in their fifties without intracerebral haemorrhage and epilepsy. Quantitative RT-PCR analysis showed the increased APP mRNA expression levels in these patients compared with those in age- and sex-matched controls.
Conclusions: Our results suggest that APP duplication should be considered in patients with EO-FAD in various ethnic groups, and that increased APP mRNA expression level owing to APP duplication contributes to AD development.
Missense mutations in three genes encoding amyloid precursor protein (APP), presenilin-1 (PSEN1) and presenilin-2 (PSEN2) have been shown to cause early-onset familial Alzheimer disease (EO-FAD: MIM 104300) with autosomal dominant inheritance. In addition, recent studies have shown that an APP locus duplication can cause EO-FAD and cerebral amyloid angiopathy (CAA) in a European population.1 2 3 The cardinal clinical presentation in patients with APP duplication is progressive dementia, frequently accompanied by intracerebral haemorrhage (ICH) and seizures.2 3 4 5 Neuropathological examination has revealed the accumulation of β-amyloid (Aβ) and neurofibrillar tangles as well as severe CAA,2 4 6 which are reminiscent of those reported in aged Down syndrome patients carrying trisomy 21.7
The contribution of APP duplication to the development of AD in non-European populations remains undetermined. We examined the occurrence and frequency of APP duplication in a Japanese AD cohort consisting of familial and early-onset sporadic cases. By performing a gene dosage analysis of the APP locus, we found independent APP duplications in two unrelated EO-FAD families.
Probands from 25 unrelated FAD families, consisting of 12 early-onset (EO, onset <65 years) and 13 late-onset (LO, onset ⩾65 years) families, and 11 sporadic EO-AD cases who were diagnosed on the basis of the NINDS-AIREN criteria,8 were recruited through Niigata University Hospital or its affiliated neurology clinic. Pedigrees were considered to have an FAD by at least one first-degree relative with clinical features indicative of dementia. Six of the 12 EO-FAD families in which missense mutations in PSEN1 (L286V, G378E, L381V and L392V) or MAPT (R406W) were found in our previous study9 were excluded from the further genetic study, because the genetic causes of their AD have been identified. This study was approved by the Institutional Review Board of Niigata University.
Genomic DNA was extracted from peripheral blood leucocytes according to a standard protocol after obtaining written informed consent. We examined microsatellite markers within or flanking the APP locus, which included D21S265, D21S1253, D21S1443, D21S1896 and D21S1235. PCR was performed in the log linear range, which enabled semiquantitative assessment. The ratio of integrated peak heights was standardised using an internal diploid PCR control as previously described.10 Gene dosage analysis was further performed using SYBR green real-time PCR assay as previously described.10 Primer sequences are available upon request. Human APP alleles were normalised for albumin. Genomic DNA from a Down syndrome patient with trisomy 21 (ATCC, CCL-66d) was included as the positive control.
We analysed the possible dosage alteration of APP in the probands from the FAD and sporadic EO-AD cases. We first performed a semiquantitative PCR analysis using microsatellite markers within or flanking the APP locus. The normal individuals showed comparable peak height ratios (1.1–1.2), whereas the patients from two EO-FAD families (PEDs 3281 and 2945, fig 1A) showed different ratios of integrated peak heights (ratios, 1.6–2.3) (supplementary fig 1). This result suggests that these EO-FAD patients carry an altered gene dosage of the APP locus. There was no evidence of gene dosage alteration in the APP locus in patients with LO-FAD or early-onset sporadic AD. To confirm the gene dosage alteration of the APP locus, we performed a microarray CGH analysis using a tiling-grade 33 K array with an 86 kb resolution.11 The log2 R ratios for the patients were abnormal for the BAC clones on chromosome 21q21 covering 4.2 Mb ranging from RP11-124K12 to RP11-341116 in PED 3281 (supplementary fig 2A) and for the BAC clones covering 0.7 Mb ranging from RP11-598G6 to RP11-640A14 in PED 2945 (supplementary fig 2B).
To determine the gene dosage alteration of the APP locus more accurately, we next performed a real-time PCR analysis of exons 5, 8, 12 and 18 of APP. The gene dosages of these exons in the patients and the Down syndrome patient were 1.5-fold higher than those of the normal individuals (fig 1D). These results suggest that these patients carry a heterozygous APP duplication. We performed an extended real-time PCR analysis of the surrounding genes and determined the genomic region of duplication. The duplicated genomic regions determined by quantitative PCR analysis were consistent with the results obtained by microarray CGH analysis. Different genomic regions of duplication were observed between PEDs 3281 and 2945: a region ranging from C21orf42 to ADAMTS5 is duplicated in PED 3281, and that ranging from APP to CYYR1 is duplicated in PED 2945 (fig 1D).
Clinical and radiological features of patients with APP duplication
The clinical characteristics of the patients are summarised in table 1.
The first patient (III-2, PED 3281) was a female who gradually developed memory impairment at the age of 52 years. As her intellectual activity gradually declined, she was admitted to our hospital at age of 53. Her mini-mental state examination (MMSE) score was 16/30. Neuropsychological assessment demonstrated cognitive decline with impaired memory and visuoperceptual skills. No clinical features suggestive of Down syndrome were observed. Brain MRI using T2*-weighted sequences revealed multiple low-signal changes in the cerebrum (fig 1B), suggesting the presence of prior microhaemorrhage. Extensive subcortical white-matter hyperintensities were also noted on a fluid-attenuated inversion recovery (FLAIR) image (fig 1C). The cortical atrophy was relatively mild, and the volume of the medial temporal areas including the entorhinal cortex was preserved. The quantification of Aβ40, Aβ42 and tau in the cerebrospinal fluid revealed a decrease in Aβ42 and an increase in tau, consistent with an AD-type pattern (table 1).
The second patient (II-3, PED 2945) was a female who developed memory loss at age of 53. From the results of her neurological examination at the age of 55, she was found to show disorientation with poor comprehension; her MMSE score was 16/30. A neuropsychological test demonstrated cognitive decline with marked impairment in new learning, short-term memory, planning and organisational activities. Brain MRI showed mild cortical atrophy, and relatively spared medial temporal areas. Neither microhaemorrhage nor subcortical white matter change was noted.
Expression levels of APP mRNA in peripheral blood
To address the question of whether increased gene dosage affects APP mRNA expression, we performed a quantitative RT-PCR analysis using total RNA prepared from the peripheral blood of the patients.12 Quantitative RT-PCR analysis revealed a significant increase in APP mRNA expression level in the patients with APP duplication, compared with that in the age-matched control subjects (supplementary fig 3).
Previous studies demonstrated that APP locus duplication can lead to EO-FAD in the European population. Eight families with APP locus duplication including five French,1 2 two Dutch3 and one Finnish families4 5 have been described. The frequency of APP duplication was reported to be 8% in the French dominant EO-FAD cohort1 and less than 2% in the Dutch EO-FAD cohort.3 Subsequent screening analysis of large Swedish and Finnish EO-AD cohorts, however, failed to detect APP duplication.13 We report here the case of two unrelated Japanese EO-FAD families with APP duplication. We estimate the frequencies of APP duplication to be 8% in our cohort of FAD and 18% in the EO-FAD families, although there may be an ascertainment bias due to the relatively small size of our cohort. Different frequencies of APP locus duplication among these studies might be explained by the different ethnic populations or inclusion criteria of AD cohorts.
The duplicated genomic regions in our patients differed from each other and from those reported in European families (fig 1D) except for the shared genomic region between patient 1 and the French F009 family.1 3 5 The finding that most of the families carry different sizes of APP locus duplication suggests that duplication has arisen de novo within each kindred. Multiple genomic architectural features such as homologous low copy repeats that act as recombination substrates might be present in the APP locus.14
Both patients with APP duplication in our study developed no symptomatic haemorrhagic stroke or epilepsy, either of which is a relatively frequently observed in previously reported families.2 3 5 MRI of patient 1 showed microbleeding on T2* images and white-matter ischaemic changes on the FLAIR image, which suggests the presence of CAA and microinfarcts. On the other hand, neither microbleeding nor white matter changes were apparent on the MR images of patient 2. Factors responsible for the occurrence of CAA in patients with APP duplication remain to be determined.
It is conceivable that the Aβ accumulation observed in previous pathological examinations of patients with APP duplication is a result of the enhanced APP expression; however, no study has yet been reported that examined APP mRNA expression in patients with APP duplication. We report here that the patients with APP duplication have a significantly higher APP mRNA expression level in their peripheral blood than the age- and sex-matched controls. In Down syndrome patients carrying three copies of the APP allele, increased APP mRNA expression levels in the brain have been implicated in the development of AD pathology.15 Taken together, an overdose of APP dosage by 50% and an increased mRNA expression level over time have a marked effect on β-amyloid burden in the brain. This notion indicates that lowering APP expression level may be a therapeutic approach for patients with APP duplication.
The authors wish to thank the patients and families for their cooperation.
Funding This study was supported in part by Grant-in-Aid for Scientific Research (20590990) from the MEXT, Japan.
Competing interests None.
Ethics approval Provided by Niigata University School of Medicine.
▸ Additional figures are published online only at http://jnnp.bmj.com/content/vol80/issue9
Patient consent Obtained.