Genetic studies

We found two index patients with FALS (3.4%) who had the GGGGCC repeat expansion in C9ORF72 (figure 1A, table 1). Neither patients with SALS nor controls had the repeat expansion. We then investigated whether the patients with FALS shared the risk haplotype for c9FTD/ALS identified in white populations. Although we were unable to obtain DNA samples from their relatives, both patients had the same risk alleles associated with c9FTD/ALS in white subjects; one had all 20 SNPs and the other had all but one, including very rare alleles in Japanese (figure 1B). Based on haplotypes from rs1822723 to rs2477518 in Japanese controls, we estimated that each patient had the risk haplotype (figure 1B). In addition, sequence analyses for amplicons spanning the hexanucleotide repeat to rs1982915 revealed that the normal repeat and the rs1982915 A allele were located on the same chromosome, suggesting that their expanded repeats were located on the risk haplotype (figure 1C). To investigate whether the risk allele was associated with repeat length, we examined the number of repeats in Japanese controls and ALS subjects. Although there were no significant differences in the average number of repeats between groups, the individuals carrying at least one rs3849942 allele A had a significantly longer repeat than those without allele A (table 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

(A) Electropherograms show capillary-based sequence traces of the repeat-primed PCR. The typical sawtooth patterns are observed in the cases carrying the GGGGCC repeat expansion and a positive control.7 The inset shows fluorescent fragment length analyses of a genotyping PCR fragment containing the GGGGCC repeat in C9ORF72. The repeat expansion carriers show one peak due to the presence of an unamplifiable repeat expansion. Numbers under the peaks indicate the number of GGGGCC repeats. (B) Schematic representation of the C9ORF72 gene and surrounding associated 20 single nucleotide polymorphisms (SNPs).5 Japanese linkage disequilibrium blocks (indicated by black bars) and risk allele frequencies (RAF) of each SNP in Japanese are presented based on HapMap-JPT (http://www.ncbi.nlm.nih.gov/projects/SNP/). Genotypes of Japanese c9FTD/ALS cases are shown. We estimated the haplotype consisting of these SNPs from control using SNPAlyze V.8 (DYNACOM, Chiba, Japan), and then we estimated the haplotype for each case. Case 1 has the risk haplotype (shown in red) and the most frequent non-risk haplotype (shown in black). Case 2 also has the risk haplotype except rs2814707 (shown in red) and the non-risk haplotype (shown in black). (C) Direct sequence analysis for amplicons spanning the hexanucleotide repeat to rs1982915 from case 1. Approximately 7 kbp amplicons spanning from the hexanucleotide repeat to rs1982915 are shown in the left panel (primer sequences and condition are available on request). Chromatograms for the amplicons are shown in the right panel. Because the PCR cannot amplify the expanded repeat, the amplicons only have the normal repeat. Although both cases have G/A allele for rs1982915 in genotyping, the sequence analysis of the amplicons shows that the products have only the A allele for rs1982915. The same results were obtained for case 2. (D) The pedigree of the cases carrying the repeat expansion. These families had no history of dementia or psychosis. The parents were healthy and without dementia for more than 80 years in case 1 and for more than 65 years in case 2. Arrow, the probands; squares, men; circles, women; solid symbols, affected family members. Sex of the unaffected siblings is obscured to protect privacy. (E) In the thoracic cord, myelin pallor is evident in the anterior and lateral corticospinal tracts (Klüver–Barrera stain). (F) A Bunina body (arrow) is evident in a facial nucleus motor neuron (H&E stain). (G) TDP-43-positive neuronal cytoplasmic inclusions (NCIs) are evident in a cervical anterior horn cell (TDP-43 immunostaining). (H) Double-labelling immunofluorescence staining in the neurons of the cervical anterior horn. In contrast to the colocalisation of TDP-43 and pTDP-43 in sporadic ALS (SALS) without the repeat expansion, case 1 shows that TDP-43-positive skein-like inclusions are negative for pTDP-43. (I–K) Scattered p62-positive NCIs are evident in the cerebellar granule cells (I), hippocampal CA3 neurons (J) and motor cortex (K) (p62 immunostaining). Scale bars, 20 μm. ALS, amyotrophic lateral sclerosis; FTD, frontotemporal dementia.

Case reports

The patients were diagnosed with ALS without dementia. Both patients had a sibling who had also been diagnosed with ALS but their parents had been healthy (figure 1D). There was no family history of dementia such as FTD or other neurological diseases in either family. Case 1 (case 4 in a previous report9) noticed hand clumsiness at age 61 years, and progressively developed dysarthria, dysphagia and limb weakness. He died 20 months after onset due to respiratory failure. Case 2 noticed dysarthria at age 63 and gradually developed weakness and muscle atrophy in her tongue and extremities. MRI revealed mild diffuse brain atrophy but the mini-mental state examination score was 29/30. She died 35 months after onset.

Histopathological studies

The brain of the case 1 patient weighed 1340 g and had no apparent abnormalities in its external appearance. Histopathological examination showed degeneration in the spinal anterolateral columns (figure 1E) as well as evidence of neuronal loss and gliosis in the spinal anterior horn, brainstem lower motor neuron nuclei (V, VII and XII) and precentral cortex. Bunina bodies were a feature in the affected lower motor neurons (figure 1F). TDP-43-immunopositive neuronal cytoplasmic inclusions (NCIs) were observed in the brainstem lower motor neuron nuclei, spinal anterior horn (figure 1G) and precentral cortex. In addition, TDP-43-immunopositive oligodendroglial cytoplasmic inclusions (GCIs) were observed in the pyramidal tract. However, pTDP-43-immunopositive NCIs and GCIs were rare. Double-labelling immunofluorescence in the neurons of the cervical anterior horn revealed that TDP-43-immunopositive skein-like inclusions were pTDP-43-immunonegative (figure 1H).

p62-Immunopositive and TDP-43-immunonegative inclusions in the cerebellum and hippocampus have been reported as pathological hallmarks of c9FTD/ALS.10 In our case, we found p62-immunoreactive NCIs in the cerebellar granule cells (figure 1I) and in the granule cells and pyramidal CA4-CA2 neurons of the hippocampus (figure 1J), which were not recognised by either TDP-43 or pTDP-43 antibodies. p62-Immunopositive inclusions were also observed in the spinal anterior horn and motor cortex (figure 1K), which were recognised by a TDP-43 antibody, but not by a pTDP-43 antibody.