Introduction
Frontotemporal dementia (FTD) is the most common form of early-onset dementia, second only to Alzheimer’s disease, and is characterised clinically by changes in personality or language impairment and pathologically by the progressive degeneration of the frontal and anterior temporal lobes [
21,
24]. FTD shares numerous similarities at the genetic and neuropathological level with amyotrophic lateral sclerosis (ALS), a devastating neurodegenerative disorder in which the loss of motor neurons in brain and spinal cord causes progressive weakness and paralysis, ultimately leading to death from respiratory failure [
12]. ALS and FTD can co-occur, and they have been proposed to be part of the same spectrum of disease [
17].
An expanded hexanucleotide repeat in the
C9orf72 gene has been identified recently as the most common known genetic cause of both FTD and ALS [
9,
20,
25]. Whilst <33 hexanucleotide repeats occur in the healthy general population, with just 2 repeats being the most common form,
C9orf72 ALS/FTD cases carry 800–4,400 repeats [
5].
C9orf72 positive FTD (c9FTD) cases may show clinically typical FTD features and have been described to most commonly present with behavioural variant frontotemporal dementia, often with prominent psychiatric and amnestic symptoms [
19].
Pathologically, c9FTD patients have unique characteristics, including p62-positive neuronal cytoplasmic inclusions (NCIs) in cerebellar and dentate fascia granule cells and pyramidal neurons of the hippocampus [
19,
23,
26]. The pathogenic mechanisms by which the hexanucleotide repeat expansion causes disease are unclear and both gain- and loss-of-function mechanisms have been proposed to play a role [
3,
9,
10,
22]. Here, we present a case of FTD with a homozygous
C9orf72 hexanucleotide repeat expansion and compare with heterozygous cases. Clinical features, neuropathology and expression data that we describe below carry important implications for disease pathogenesis and genetic counselling.
Materials and methods
DNA extraction and genotyping
Genomic DNA was extracted from peripheral blood using the Nucleon BACC2 DNA extraction kit (RPN8502) following the supplied protocol. DNA concentrations were determined using a Nanodrop ND-1000 spectrophotometer, and adjusted to a working concentration of 20 ng/μl TE buffer.
Rs3849942 genotyping: The surrogate marker rs3849942, defining the haplotypes at risk of expansion, was genotyped by allelic discrimination using the 5′ nuclease assay in conjunction with Minor Groove Binding (MGB) probes. The custom-designed assay was performed on the SDS7500 Fast Real Time PCR system (ABI) and genotyping calls were made using software v2.0.6.
Hexanucleotide repeat number assessment
Hexanucleotide repeat number was assessed by repeat primed PCR and carried out as previously described [
19]. Fragment length analysis was undertaken on an ABI 3730xl automated sequencer. Analysis of repeat primed PCR (rpPCR) electrophoretograms was performed using Peak Scanner v1.0 (ABI). In addition, repeat number was assessed by fluorescent labelled PCR followed by fragment length analysis on an Applied Biosystems (ABI) 3730xl automated sequencer as previously described [
19].
Microsatellite analysis
Microsatellite analysis was performed using ten markers spanning approximately 13.1 Mb of genomic DNA centred on the C9orf72 gene. PCR amplicons were generated using fluorescently end-labelled primers for microsatellite markers D9S1814(VIC), D9S976(FAM), D9S171(NED), D9S1121(VIC), D9S169(FAM), D9S263(HEX), D9S270(FAM), D9S104(FAM), D9S147E(NED) and D9S761(FAM). DNA products were electrophoresed on an ABI 3730xl automated sequencer. Data were analysed using ABI GeneMapper software v4.0 [Applied Biosystems (ABI)].
Southern blotting
Adaptation of standard blotting methods included the probing of AluI/DdeI digested genomic DNA with an oligonucleotide hybridisation probe from Eurofins MWG Operon (Germany) that comprised five hexanucleotide repeats (GGGGCC)
5 labelled 3′ and 5′ with digoxigenin (DIG). Further methods followed the DIG Application Manual [Roche Applied Science (RAS)], except for the supplementation of DIG Easy Hyb buffer with 100 μg/ml denatured fragmented salmon sperm DNA. Detection was carried out as recommended in the DIG Application Manual using CSPD ready-to-use (RAS) as chemiluminescent substrate visualised on fluorescent detection film (RAS). Hexanucleotide repeat number was estimated by visual interpolation using a plot of log10 base pair number against migration distance which was created in Excel (Microsoft) and subtraction of the wild-type allele fragment size (199 bp). Full methods have been previously reported [
5].
For single probe Southern hybridisation, 5 μg genomic DNA were digested with EcoRI. The Roche DIG labelling and detection system was used with a 1 kb C9orf72 genomic DNA specific probe.
Pathological analysis
Analysis was undertaken according to the Queen Square Brain Bank protocol [
16]. p62 (1:200, BD Transduction Laboratories), TDP-43 (amino acids 1-261, Abnova, 1:800), C9RANT [
3] and ubiquilin-2 [
6] were analysed. Using a semi-quantitative approach, the frequency of p62-positive neuronal cytoplasmic inclusions was compared with TDP-43-positive inclusions in the reported regions of the frontal cortex, temporal cortex, hippocampus and cerebellum.
Semi-quantitative analysis of p62 and TDP-43-positive lesions was performed using a five-tiered semi-quantitative grading scale. The scoring of the pathological lesions was as follows: ‘0’ for the absence of p62-positive neuronal cytoplasmic inclusions (NCIs) and neuronal intranuclear inclusions (NIIs); ‘+’ for 1–5 inclusions present in an average of at least five microscopic fields using a 20× objective; ‘++’ for 6–10; ‘+++’ for between 11 and 20 inclusions; ‘++++’ for >20 lesions.
RNA extraction and transcription analysis
RNA extraction was performed using the miRNeasy Kit (Qiagen) and RNA quality was evaluated using the Agilent 2100 Bioanalyzer. Reverse transcription was performed using the QuantiTect kit (Qiagen) and real-time quantitative RT-PCR (qPCR) was performed using Taqman probes according to manufacturer’s instructions (Applied Biosytems Primers and probes were designed in order to specifically detect the V1, V2 and V3 C9orf72 isoforms (Supplementary table 1). Dilution curve experiments using the V1 and V3 C9orf72 primers vs beta-actin showed comparable efficiencies. Reactions were performed in triplicate and gene expression values were normalised using the housekeeping genes ACTB and GAPDH. Relative quantification of gene expression was calculated via the comparative threshold cycle (ddCt) method. Regression analysis was performed by plotting the expression values vs the number of mutated alleles using Graphpad Prism v5.03.
Discussion
We report the first FTD patient carrying a homozygous C9orf72 hexanucleotide repeat expansion. Our finding that the clinicopathological features were severe, but within the range of reported heterozygous cases, suggests that the condition is truly dominant; therefore, we provide evidence in favour of a gain-of-function mechanism. Whilst the affected sister with an age of onset of 21 could also be potentially homozygous for the C9orf72 expansion, it is highly unlikely that the patient’s mother and also the father’s son from his second (non-consanguineous) marriage, both with an onset in their 20s, were homozygous. These family members were not available for assessment.
We report Southern blotting for detection of
C9orf72 expansions using two different approaches with the probe directed towards either the hexanucleotide repeat itself (Fig.
2c) or the adjacent genomic region (Fig.
2d). The former has the advantage of being more sensitive and is able to detect a significant amount of somatic mosaicism, whilst the latter has the advantage of detecting also the normal allele and therefore its absence in the setting of homozygosity. A size shift, possibly due to somatic instability, is documented between blood and brain DNA samples, as previously reported [
5].
Three major
C9orf72 transcripts have been described: V1, in which the hexanucleotide expansion is in the promoter region, and V2 and V3, in which the expansion lies in the first intron and is therefore transcribed [
9]. Previous analyses in
C9orf72 heterozygous cases showed a reduction in V1 and V2 [
9,
10]. Consistent with these findings, our results indicate a reduction of all three transcripts in heterozygous cases, with a further reduction in the homozygous case; but importantly, we unequivocally show that transcripts derived from expansion-containing alleles are present and clearly detectable in the homozygous case. One study [
9] observed a greater reduction in the V1 isoform than V2 and V3. Although our results show that all variants are reduced, we nonetheless observe a greater reduction in V1 compared to V2 and V3 in the homozygous case.
The pathogenic mechanism of the
C9orf72 repeat expansion is unknown and both loss of function (LOF) [
10] and gain of function [
22] have been proposed to play a role. Although the two mechanisms are not mutually exclusive, the clinicopathological data along with the expression profile of the case presented here are consistent with a predominant gain-of-function mechanism. A pure LOF mutation would be expected to cause a much more severe clinical presentation and pathology in the homozygous state, which was not observed in this case. Indeed heterozygous LOF mutations in other neurodegenerative diseases present with radically different phenotypes when homozygous, whereas those associated with a toxic gain of function tend to be present with the same clinical syndrome, although in many cases with an earlier onset. Examples include heterozygous LOF mutations in the progranulin gene (
PGRN) causing FTD but homozygous mutations causing a lysosomal storage disorder [
4,
29]; mutations in
GBA causing Parkinson’s disease in heterozygosity and Gaucher’s disease in homozygosity [
1]; and mutations in
TREM2 causing Alzheimer’s disease and Nasu–Hakola disease [
11,
14]; however, homozygous mutations in gain-of-function neurodegenerative diseases such as inherited prion disease (
PRNP), ALS (
SOD1) and Huntington disease (
HTT), manifest with severe phenotypes, that fall within the range of disease [
13,
15,
28,
30]. Further, a severe form of disease, but with no additional clinical features, such as presented here, was described for homozygous cases of myotonic dystrophy types 1 and 2 (DM1 and DM2) [
2,
27]. Importantly, these disorders are similar to
c9FTD in that they are also caused by very long expansions of non-coding repeats, and functional evidence suggests they occur via a gain-of-function mechanism [
7]. The severe pathology here reported in the C9orf72 homozygous case could be compatible with a dosage dependence of the GOF mechanism. Gain and loss of function have been shown to coexist in other neurodegenerative disorders [
8,
18,
31,
32]. Given the decreased levels of all
C9orf72 transcripts, further work is needed to assess the possible role of loss of function in
C9orf72 ALS/FTD, but our results argue against a pure loss-of-function mechanism.
Lastly,
C9orf72 is estimated to have a frequency of 1:700 in the UK population [
5], thus making the predicted frequency of homozygous cases to be approximately 1:2 × 10
6. Although we present a single case and clinical variability is likely to occur, our results show that homozygous
C9orf72 expansion is viable and that it can present as typical FTD. Genetic diagnostic tests are already available for
C9orf72 and are carried out by performing both repeat-primed PCR and Southern blotting. Our results underline the importance of performing Southern blots in the diagnostic test, in order to detect potential homozygous cases. Although homozygosity may or may not have prognostic implications, it carries huge implications for genetic counselling. In summary, we provide the first report of a homozygous
C9orf72 repeat expansion FTD case, which provides new insight into disease pathogenesis and important implications for genetic testing.