Background
Frontotemporal lobar degeneration (FTLD) is the third most common neurodegenerative cause of dementia after Alzheimer's disease (AD) and dementia with Lewy bodies (DLB). [
1,
2] It stems from the degeneration of neurons in the superficial frontal cortex and anterior temporal lobes. Typically, this results in several distinct clinical presentations characterised by changes in personality and behaviour, including a decline in manners and social skills representative of frontotemporal dementia, as well as language disorders of expression and comprehension, known as progressive aphasia and semantic dementia, respectively. [
3] Contributing to the spectrum of clinical phenotypes seen in FTLD is the co-occurrence of FTLD with motor neurone disease (MND). [
4] MND, also referred to as amyotrophic lateral sclerosis (ALS) is characterised by degeneration of upper and lower motor neurons, leading to progressive muscle wasting, weakness and spasticity which ultimately results in profound global paralysis and death, usually due to respiratory failure.
FTLD is also a pathologically heterogeneous disorder and can be categorised into cases without detectable inclusions known as dementia lacking distinctive histopathology (DLDH), cases with tau-positive pathology known as tauopathies, and the most frequently recognised cases have ubiquitin-positive, tau-negative inclusions (FTLD-U). [
5] The TAR DNA binding protein (TDP-43) is a nuclear protein implicated in exon splicing and transcription regulation, [
6] and was recently identified as a major protein component of the ubiquitin-immunoreactive inclusions characteristic of sporadic and familial FTLD-U, with and without MND, as well as in sporadic cases of MND [
7‐
9]. Recently, mutations in the TDP-43 (TARDBP) gene have recently been reported in familial and sporadic forms of MND. [
10‐
14]
There is increasing evidence that FTLD and MND may represent two phenotypic variants resulting from a common underlying genetic cause. This is supported by both the presence of ubiquitin/TDP-43 pathology and also by genetic loci on chromosome 9 in families with FTLD and MND. Hosler et al. [
15] identified a region on chromosome 9q21-22 from linkage data from 5 American FTLD-MND families. Subsequently, both Vance et al. [
16] and Morita et al. [
17] reported linkage to chromosome 9p13.2-21.3 in large FTLD-MND kindreds from Holland and Scandinavia, respectively. Finally, three other families were identified by Valdmanis et al. [
18] with linkage to the chromosome 9p locus. Yan et al. [
19] have also provided a preliminary abstract report of significant linkage in 15 FTLD-MND families. To date, only one gene, IFT74 has been postulated to be the causative gene of chromosome 9p-linked FTLD-MND. [
20] However, only a single family has been identified with a mutation in the IFT74 gene, suggesting genetic heterogeneity in this region. Here, we report a large FTLD-MND family from Australia with linkage to chromosome 9p21.1-q21.3 and TDP-43 positive pathology, further supporting the evidence for a novel gene associated with this type of neurodegenerative disorder.
Conclusion
Frontotemporal lobar degeneration (FTLD) is a clinically, pathologically and genetically heterogeneous disorder. To date, at least 22 families with FTLD and/or MND have now been reported with genetic linkage to chromosome 9p [
13‐
16] providing strong evidence that an additional FTLD gene exists. In this study we describe a large Australian FTLD-MND family that shows linkage to the chromosome 9p21.1-21.2 locus. With a significant two-point LOD score of 3.24 and a multi-point LOD score of 3.41, this is the only study that have provided statistically significant evidence for linkage from a single pedigree, the other pedigrees having two-point LOD scores of 2.41 [
16], 2.81 [
18] and 2.33 [
17]. This means that we can rely on our haplotype analysis with greater statistical certainty. In addition, the family shows considerable clinical heterogeneity, compared to some other families that have been linked to the same locus.
Our genome-wide linkage analysis led to the identification of a genetic locus on chromosome 9p21.1-9q21.3. The resulting 57 Mb disease haplotype region overlaps with three other FTLD-MND loci identified by Morita et al., [
17] Vance et al., [
16] and Valdmanis et al. [
18] (Figure
5) providing further evidence for this region as the disease locus. In combination, the four linkage studies collectively define a likely disease haplotype of 7.0 Mb between D9S169 and D9S1805 (Figure
5). We note that this haplotype does not overlap with the most recent preliminary abstract report of a 7.4 cM haplotype on the 9p region by Yan et al. [
31] although, it does overlap with a region that was originally reported in abstract form by Yan et al. [
19] Given that Yan's latest reported region is probably based on recombination events drawn from multiple families it is possible that one of the defining break points may be a false positive due to the low statistical power of individual pedigrees, [
31] or that a disease haplotype boundary was defined by a phenocopy as observed in our pedigree. The region defined by D9S169 and D9S1118 (Figure
4) harbours the five transcripts that have been thoroughly screened in this study and by Momeni et al. [
20] with no plausible mutations having been detected.
Changes in personality and behaviour, motor dysfunction as well as Ub/TDP-43 positive pathology represent the core clinical and neuropathological features characteristic of FTLD-MND families linked to 9p. In this study we describe an FTLD-MND family with additional clinical and pathological findings, not previously described in the chromosome 9p-linked families. Early and severe memory impairment is generally held to be an exclusion criterion for the clinical diagnosis of FTLD. [
1] None of the other chromosome 9-linked pedigrees have reported major memory impairment as their primary diagnoses, although Morita et al. [
17] mentioned that memory deficits were detected in three affected individuals in their pedigree. However this aspect was not reported as their primary diagnosis as their memory deficits were detected during neuropsychological tests three years before death. Moreover, Momeni et al. [
20] reported that one of their patients had additional AD-like pathology, namely diffuse β-amyloid (Aβ) positive plaques in the absence of neuritic plaques and tangles. We too describe a patient (III:2) who presented with clinical symptoms typical of AD and at autopsy not only had indisputable TDP-43 positive neuronal cytoplasmic inclusions but also had amyloid-plaques and neurofibrillary tangles characteristic of AD (Figure
2). It has been postulated that Apolipoprotein E (APOE) may play a role in the development of Aβ deposition in FTLD cases [
32]. There is no apparent association of APOE status with the presence of Aβ deposition in family 14 as the individual who was homozygous e4/e4 (III:12) had less Aβ deposition than the two individuals who were heterozygous e3/e4 (III:2 and III:3).
The first reported linkage of a novel FTLD-MND locus to chromosome 9p was in 2006, although no convincing candidate genes have yet been identified. The issue of phenocopies and the error of reliance on a single meiotic recombination events to define minimal disease regions could be a crucial factor in the failure to identify the disease gene. The recombination breakpoints reported in the literature by Morita
et al. [
17] and Valdmanis
et al. [
18] are based on a single recombination event in a single pedigree. Moreover, both of these pedigrees have two-point LOD scores less than 3. Vance et al. [
16], using a pedigree with a LOD score of 2.4, showed recombination in multiple individuals. However, several of the individuals with the disease haplotype do not have FTLD-MND, [
16] calling into question the relevance of this recombination breakpoint. Our reported increase in the minimal disease region should inform the other groups that the chromosome 9 locus may be more significantly more telomeric than predicted by the existing recombination breakpoints. Moreover, we report the existence of a case with clinical Alzheimer's disease, and FTLD-U neuropathology, who shares the disease haplotype. This result highlights the possibility that the classification of late-onset AD patients in the other linked pedigrees as sporadic dementia cases or unaffected may be erroneous, thereby reducing statistical power, or possibly even excluding pedigrees from linkage analysis. In summary, multiple families with FTLD-MND, without mutations in the known dementia genes, have been linked to chromosome 9p. This strongly suggests that the locus on chromosome 9 play a major role in pathogenic pathways that lead to FTLD-MND, making it imperative to identify the causative gene(s).
Acknowledgements
Australian Postgraduate Award (AAL), the National Health & Medical Research Council (Australia) Project Grants 276407 and 510217, RD Wright Fellowship 230862 (JBJK), Medical Postgraduate Scholarship 325640 (CTL) and Research Fellowships 350827 (GMH) and 157209 (PRS), and the Rebecca Cooper Medical Research Foundation Ltd. Blundy family donation. We thank all patients and family members who participated in this study. We also thank Jim McBride and the staff of the Peter Wills Bioinformatic Centre, Garvan Institute of Medical Research for IT assistance, Heather McCann for assistance with immunohistochemistry and Robyn Flook of the South Australian Brain Bank.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
JBJK, PRS conceived this study. AAL, CDS acquired the data. EMT, JH, GAN, WSB, PKP, CTL collected blood and clinical data from family. PB, GMH performed the neuropathological analyses. JBJK, AAL, PP, GMH and PRS participated in the mamangement, analysis, interpretation of data and drafting of manuscript. All have critically revised the manuscript for important intellectual content and seen and approved the final version.