Objective Since the first report of CHCHD10 gene mutations in amyotrophiclateral sclerosis (ALS)/frontotemporaldementia (FTD) patients, genetic variation in CHCHD10 has been inconsistently linked to disease. A pathological assessment of the CHCHD10 protein in patient neuronal tissue also remains to be reported. We sought to characterise the genetic and pathological contribution of CHCHD10 to ALS/FTD in Australia.
Methods Whole-exome and whole-genome sequencing data from 81 familial and 635 sporadic ALS, and 108 sporadic FTD cases, were assessed for genetic variation in CHCHD10. CHCHD10 protein expression was characterised by immunohistochemistry, immunofluorescence and western blotting in control, ALS and/or FTD postmortem tissues and further in a transgenic mouse model of TAR DNA-binding protein 43 (TDP-43) pathology.
Results No causal, novel or disease-associated variants in CHCHD10 were identified in Australian ALS and/or FTD patients. In human brain and spinal cord tissues, CHCHD10 was specifically expressed in neurons. A significant decrease in CHCHD10 protein level was observed in ALS patient spinal cord and FTD patient frontal cortex. In a TDP-43 mouse model with a regulatable nuclear localisation signal (rNLS TDP-43 mouse), CHCHD10 protein levels were unaltered at disease onset and early in disease, but were significantly decreased in cortex in mid-stage disease.
Conclusions Genetic variation in CHCHD10 is not a common cause of ALS/FTD in Australia. However, we showed that in humans, CHCHD10 may play a neuron-specific role and a loss of CHCHD10 function may be linked to ALS and/or FTD. Our data from the rNLS TDP-43 transgenic mice suggest that a decrease in CHCHD10 levels is a late event in aberrant TDP-43-induced ALS/FTD pathogenesis.
- amyotrophic lateral sclerosis
- coiled-coil-helix-coiled-coil-helix domain containing 10 protein
Statistics from Altmetric.com
- amyotrophic lateral sclerosis
- coiled-coil-helix-coiled-coil-helix domain containing 10 protein
Amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD) are late-onset neurodegenerative diseases that show significant clinical, genetic and pathological overlap.1 Up to 50% of ALS patients develop some form of cognitive impairment, with 15% being diagnosed with comorbid FTD (ALS/FTD).2 Further, ALS and FTD share genetic aetiology, and are both characterised by the presence of protein aggregates in affected neurons.2 Such protein aggregates are positive for the TAR DNA-binding protein 43 (TDP-43) protein in over 90% of ALS cases and around 50% of FTD cases.3 As such, these two conditions are considered to represent a spectrum of neurodegenerative disease.2
Approximately 10% of ALS patients and 33% of FTD patients exhibit familial inheritance of disease, while the remaining cases appear to occur sporadically.4 5 In 2014, Bannwarth et al6 reported a novel mutation in the gene encoding the coiled-coil-helix-coiled-coil-helix domain containing 10 protein (CHCHD10, MIM#615903) in a French family with a mitochondrial DNA instability disorder. The affected family members exhibited a range of accompanying phenotypes, including symptoms suggestive of ALS and FTD. Subsequently, a familial ALS (FALS)/FTD patient of Spanish descent was found to carry an identical mutation.6 Numerous non-synonymous CHCHD10 variants have since been reported in ALS/FTD cohorts.7–16 However, CHCHD10 mutations have also been found to be absent from multiple pure ALS cohorts,17 18 and at lower frequencies in ALS patients compared with ethnically matched FTD cases,8 18–21 suggesting CHCHD10 mutations may be more closely linked to FTD dominant phenotypes than pure ALS.
The function of CHCHD10 is largely unknown, though it has been implicated in mitochondrial organisation by interaction with the mitochondrial contact site and cristae organising system (MICOS) protein complex.22 It has also been reported to form complexes with TDP-43 and play a role in retaining the nuclear localisation of TDP-43.23 However, no nucleocytoplasmic translocation of TDP-43 was observed in ALS patient-derived fibroblasts carrying a CHCHD10 p.Gly66Val mutation.24 In the original CHCHD10 mutation positive family with mitochondrial DNA instability disorder, mitochondrial fragmentation, crystalloid inclusions and structural alterations were observed in patient muscle and skin fibroblasts.6 To date, no pathological assessment of CHCHD10 has been reported in brain and spinal cord tissues from ALS and/or FTD patients, with or without a CHCHD10 mutation.
In this study, we aimed to determine the prevalence of CHCHD10 mutations, or disease associated variants, in Australian ALS and FTD patient cohorts. We also sought to characterise CHCHD10 protein expression in ALS and/or FTD patient postmortem neuronal tissues as well as in a TDP-43 transgenic mouse model.25 Given that the protein products encoded by ALS genes, such as TARDBP and UBQLN2, show abnormal pathology in ALS patients regardless of mutation status,26–28 we hypothesised that CHCHD10 may also contribute to wider patient pathology. While causal CHCHD10 mutations were not found within the Australian patient cohort, we report neuronal expression of CHCHD10 in human brain and spinal cord postmortem tissues. Further, we show a significant decrease in CHCHD10 protein levels in ALS and FTD patients and in a TDP-43 transgenic mouse model with disease progression. This confirmed that CHCHD10 is a neuronal protein within brain and spinal cord tissues, and implicates its role in ALS/FTD pathogenesis via a loss-of-function mechanism.
Eighty-one FALS patients (including 61 probands), and 635 sporadic ALS (SALS) patients were ascertained from the Macquarie University Neurodegenerative Disease Biobank, Molecular Medicine Laboratory (Concord Hospital), Australian Motor Neuron Disease DNA bank (Royal Prince Alfred Hospital) and Brain and Mind Centre (University of Sydney). An additional 108 FTD patients were also recruited from the Brain and Mind Centre. All participants were recruited under informed written consent as approved by the human research ethics committees of the Sydney South West Area Health Service, Macquarie University, Sydney South East and Illawarra Area Health Service or University of Sydney. Most participants were of European descent. Patients were clinically diagnosed with ALS based on El Escorial criteria29 or behavioural variant FTD according to international consensus criteria.30 Genomic DNA was extracted from peripheral blood using standard protocols.
DNA from FALS patients who were negative for mutations in SOD1 and the pathogenic hexanucleotide repeat expansions in C9orf72 underwent whole-exome sequencing (WES, n=81). Briefly, WES was performed at Macrogen (Seoul, Korea) using an Illumina HiSeq2000 platform and the TruSeq Exome Enrichment kit (Illumina, California, USA) or SureSelectXT Human All Exon V5 + UTR kit (Agilent, California, USA). Full details of the cohort are described by McCann et al.31 Whole-genome sequencing (WGS) was performed on the Illumina HiSeq X Ten platform using the TruSeq PCR-free library preparation (V.2.5) (Kinghorn Centre for Clinical Genomics, Sydney) for SALS patients (n=635), including 16 C9orf72 positive patients, and FTD patients (n=108) negative for C9orf72 and MAPT, including two patients carrying a GRN mutation. WES and WGS raw data were processed using the Genome Analysis ToolKit (GATK, Broad Institute, Massachusetts, USA) and the corresponding best practices,32 33 with resultant variant call format (VCFs) files annotated using ANNOVAR,34 including in silico protein predictions from dbNSFP (V.3.3a).35 Read depth of the CHCHD10 coding region was calculated using Qualimap V.2.2.1,36 using WES BAM files (n=81) and WGS BAM files (n=152, representative samples).
CHCHD10 variant analysis
Using UNIX and the R statistical environment, custom bioinformatics scripts were applied to annotated VCFs to identify all genetic variants present in the coding region of CHCHD10 (NM_213720) in all subjects, regardless of mutation status. First, the presence of CHCHD10 variants previously reported in the literature as being linked to ALS and/or FTD was determined (c.34C>T; p.Pro12Ser, c.44G>T; p.Arg15Leu, c.64C>T, p.His22Tyr, c.68C>T; p.Pro23Leu, c.95C>A; p.Ala32Asp, c.100C>T; p.Pro34Ser, c.104C>A; p.Ala35Asp, c.170T>A; p.Val57Glu, c.176C>T; p.Ser59Leu, c.197G>T; p.Gly66Val, c.239C>T; p.Pro80Leu, c.244C>T; p.Gln82Ter and c.286C>A; p.Pro96Thr). To identify novel variants in CHCHD10 that contribute to disease, all CHCHD10 variants were filtered for exonic, non-synonymous variants.
Variant allele counts were compared between patients and unrelated control individuals using Fisher’s exact testing. Intergenic, upstream and downstream variants were not analysed. The p value significance threshold (alpha=0.05) was corrected for multiple-testing using Bonferroni corrections based on the number of analysed variants. Each of FALS, SALS and FTD patient allele counts were compared with three control datasets; the non-Finnish European (NFE) subset from the Genome Aggregation database (gnomAD, n=63 369),37 and two ethnically matched control cohorts, the Medical Genome Reference Bank (MGRB, n=1144) and the Diamantina Australian Control Collection (DACC, University of Queensland, n=967). The MGRB and DACC cohorts consisted of neurologically healthy Australian individuals of predominantly Western European descent and all individuals from the MGRB cohort were over 70 years of age. Fisher’s exact testing was not completed for variants absent from the MGRB and DACC datasets if flanking variants had low sequence coverage. An average of 15 350 alleles was used to calculate p values for variants absent in the NFE gnomAD control dataset.
Subjects and tissues
Patient postmortem formalin-fixed paraffin-embedded human cervical spinal cord, motor cortex and frontal cortex sections (5 µm), and fresh-frozen motor cortex tissue were obtained from the New South Wales Brain Bank Network. The frontal cortex tissues were collected from the superior frontal cortex region, corresponding to the Brodmann area 9. A summary of the case details is listed in table 1. No significant difference in sex distribution, age at death or postmortem interval was evident within the spinal cord, motor cortex and frontal cortex cohorts, as determined using Welch two Sample t-test in R (V.3.5.2). P values are listed in online supplementary table 1. All patients were characterised by TDP-43 inclusions. Motor cortex tissues match the spinal cord cohort except for four additional C9orf72 positive cases in the spinal cord cohort.
All mouse experiments were approved by and performed according to the requirements of the Animal Ethics Committee of Macquarie University (approval #2015–042) with reference to the Australian code for the care and use of animals for scientific purposes. Mice breeding and tissue collection methods are detailed in the online supplementary methods.
Immunohistochemistry and immunofluorescence
Immunohistochemistry (IHC) analysis of CHCHD10 for visualisation of staining patterns and semi-quantification of protein levels in motor cortex and frontal cortex tissues was performed on spinal cord, motor cortex and frontal cortex tissue sections as previously described using rabbit polyclonal anti-CHCHD10 (1:400; Sigma-Aldrich, Missouri, USA).38 To assess colocalisation between CHCHD10 and phosphorylated TDP-43 or a mitochondrial marker VDAC1, dual immunofluorescence (IF) was performed on spinal cord, motor cortex and frontal cortex tissues using primary anti-CHCHD10 and either mouse monoclonal anti-VDAC1 (1:500, BioLegend, California, USA), or mouse monoclonal anti-TDP-43 phosphorylated Ser409/410 (1:5000; Cosmo Bio, Japan) antibodies, followed by relevant Alexa Fluor conjugated secondary antibodies.
Visualisation and analysis of tissue sections
IHC sections were captured using the Virtual Microscope ScanScope Unit and ScanScope Consol programme before being visualised using the Image Scope programme (Leica Biosystems, Germany). IHC sections were also visualised using the ZEISS Axio Imager 2 microscope. IF sections were imaged with a ZEISS LSM 880 inverted confocal laser-scanning microscope and analysed using FIJI.39
To quantify CHCHD10 protein expression in spinal cord tissues, the mean CHCHD10 fluorescence intensity from at least four motor neurons were obtained using methods described by Fifita et al.38 For frontal cortex tissues, the semi-quantification was performed by a researcher and each section was categorised into strong, moderate or weak based on staining level blindly.
Western blotting of human motor cortex, and mouse brain or spinal cord samples
Methods for the preparation of human motor cortex and mouse brain or spinal cord protein lysates, and western blotting conditions are detailed in the online supplementary methods. For human samples, membranes were blocked in Odyssey Blocking Buffer in TBS (LI-COR Biosciences, Nebraska, USA) for 1 hour at room temperature followed by overnight incubation at 4°C with primary antibodies: CHCHD10 (as above, 1:250), Neuronal Nuclei Antigen (NeuN) mouse monoclonal 1:1000 (Merck, Germany) and GAPDH mouse monoclonal 1:5000 (Proteintech). For mouse samples, membranes were blocked in 3% bovine serum albumin in tris-buffered saline for 1 hour at room temperature followed by overnight incubation at 4°C with primary antibodies: rabbit polyclonal anti-CHCHD10 (1:1000, #25 671–1-AP, ProteinTech), rabbit polyclonal anti-TDP-43 (1:2000, #10 782–2-AP, ProteinTech), mouse monoclonal NeuN 1:1000 (Merck, Germany) and mouse monoclonal GAPDH (1:10 000, #60 004–1-Ig, ProteinTech). Membranes were then incubated for 1 hour at room temperature with IRDye 680LT donkey anti-rabbit IgG and 800CW donkey anti-mouse IgG, 1:20 000 (LI-COR Biosciences). Membranes were visualised using the Odyssey CLx imaging system and protein bands quantified with the Image Studio Lite software (LI-COR Biosciences).
Statistical analysis of CHCHD10 expression
A linear mixed effect model was fitted to assess whether there was an association between CHCHD10 protein expression level and whether an individual was classified as a control, an ALS patient positive for the C9orf72 repeat expansion or an ALS patient negative for the C9orf72 repeat expansion in spinal cord motor neurons. Sample identification numbers were included as a random effect in the model to account for multiple observations from the same individual. R V.3.5.1 and the ‘lmerTest’ package were used for this analysis. Prism 8 build-in one-way analysis of variance was used when analysing CHCHD10 protein levels in human motor cortex tissues, respectively. Post hoc Dunnett’s t-test was applied to compare values between controls and patients in these samples. Paired Student’s t-test (two-tailed) was performed when comparing CHCHD10 expression in mouse tissues. A probability of p<0.05 was considered significant. When analysing motor cortex and frontal cortex IHC staining, Fisher’s exact testing was performed, and significance was set to 0.0166667 after Bonferroni correction.
Genetic variation in CHCHD10 among ALS and FTD patients
WES and WGS data were interrogated for the presence of CHCHD10 genetic variants in FALS, SALS and FTD patients. The mean read depths for the coding region of CHCHD10 in WES and WGS data were 10.89 and 42.64, respectively. Three of the 13 variants previously reported as linked to ALS/FTD (p.Pro34Ser, p.Pro80Leu and p.Pro96Thr) were identified in the Australian patient cohort (table 2). Six SALS, two SALS and one FTD patient carried each of these variants, respectively. Each of these three variants were also present in control individuals from the NFE gnomAD dataset (table 2). Interestingly, p.Pro80Leu (observed in two SALS cases) was absent from Australian controls, and trended towards overrepresentation in SALS compared with NFE gnomAD controls (p=0.03), though this was not significant after Bonferroni correction (described in Association analysis of population-based genetic variants). Additionally, p.Pro34Ser was overrepresented in SALS when compared with Australian DACC controls (p=0.0038); however this was not replicated using NFE gnomAD or Australian MGRB control cohorts (table 2). One known rare non-synonymous variant (p.Tyr135His; rs145649831) was also identified in one SALS and one FALS case (table 2). This variant was predicted to be a benign polymorphism by 15/18 in silico protein prediction programme. No novel CHCHD10 missense or nonsense variants were identified. All subjects that were found to carry a genetic variant in CHCHD10 were negative for C9orf72 and GRN mutations.
Association analysis of population-based genetic variants
Within the FALS WES data and SALS/FTD WGS data, 8 and 27 variants were identified for analysis, respectively. Therefore, the Bonferroni corrected significance thresholds were p<0.00625 (FALS analysis) and p<0.00185 (SALS and FTD analyses). Association analysis of population-based CHCHD10 SNPs using the Fisher’s exact test found no variants significantly associated with FALS or FTD (table 2, online supplementary results). One intronic SNP (rs62241575) was significantly associated with SALS compared with NFE gnomAD controls. However, analysis using the Australian control cohorts MGRB and DACC failed to replicate this association.
CHCHD10 protein is specifically expressed in neurons in human spinal cord, motor cortex and frontal cortex
Following the genetic analysis, we next sought to examine whether changes in CHCHD10 protein localisation are evident in ALS and/or FTD patients with TDP-43 pathologies, even in the absence of CHCHD10 mutations. We first performed IHC staining in a cohort of ALS and/or FTD patient postmortem tissues. CHCHD10 showed primarily neuronal expression at all three locations. In ALS patient and control spinal cord, CHCHD10 staining was specifically observed in anterior horn motor neurons and neuropils in the grey matter region but was generally absent from other cell types (figure 1A). In the motor cortex and frontal cortex of ALS, ALS/FTD or pure FTD patients and controls, CHCHD10 was predominantly observed in the medium and large pyramidal neurons, and the Betz cells, in cortical layers (figure 1C,E). At all three locations, CHCHD10 showed cytoplasmic staining in controls and patients with or without known ALS gene mutations (figure 1B,D,F). Consistent with previous studies, we confirmed that CHCHD10 colocalises with a mitochondrial marker VDAC1 in spinal cord neurons in both controls and patients (online supplementary figure 1).
No CHCHD10 inclusions were observed in the majority of spinal cord, motor cortex or frontal cortex neurons, nor was CHCHD10 seen to colocalise with pTDP-43 inclusions (figure 2). However, CHCHD10 inclusion-like structures were occasionally seen in spinal cord motor neurons from three SALS cases, and motor cortex tissues from one SALS case (online supplementary figure 2).
CHCHD10 protein levels are decreased in ALS spinal cord and FTD frontal cortex tissues
Quantification of CHCHD10 protein level in spinal cord was performed by CHCHD10 immunofluorescent intensity in control or ALS motor neurons (figure 3A). By fitting a linear mixed effect model to the CHCHD10 protein expression data, we found CHCHD10 expression to be significantly reduced for ALS patients who carried the C9orf72 repeat expansion (p=0.00056) and ALS patients without the C9orf72 repeat expansion (p<0.0001), compared with controls (figure 3B and online supplementary table 2). In motor cortex, CHCHD10 was quantified via western blotting. As CHCHD10 is primarily expressed in neurons, we normalised CHCHD10 to a neuronal marker NeuN and a loading control GAPDH. No significant difference was observed in CHCHD10 protein level between control and patients with different genotypes or between control and ALS patients as a whole (figure 3C,D,E). This is possibly due to the variable CHCHD10 protein expression and variable neuronal content in each sample. To overcome this, we also assessed CHCHD10 level in Betz cells via IHC scoring. Motor cortex tissues from cases where tissue sections were available were immunostained with CHCHD10 antibody and the level of CHCHD10 was blindly scored as strong, moderate or weak (online supplementary figure 3 and online supplementary table 3). Similar to motor cortex western blotting results (figure 3C,E), both control and patient Betz cells demonstrated variable levels of staining and no significant difference between the number of cases with strong or moderate and weak CHCHD10 staining (p=0.4909).
No tissue lysates were available from frontal cortex tissues for western blot analysis, and quantification via IF was not suitable due to the high lipofuscin content that caused high background staining in cortex tissues.40 As such, CHCHD10 expression was semi-quantified whereby IHC staining was blindly scored as strong, moderate or weak (figure 4). Frontal cortex tissues from ALS/FTD patients and controls demonstrated similar CHCHD10 neuronal staining intensities, with strong to moderate expression observed in a combined 6/6 (100%) controls and 5/6 (83%) ALS/FTD cases (table 3, figure 4). In contrast, predominantly weak CHCHD10 neuronal staining was observed in the frontal cortex of 5/6 (83%) pure FTD patients (table 3, figure 4). The Fisher’s exact test (Bonferroni corrected threshold of p<0.016667) indicated a significant decrease in CHCHD10 protein levels between FTD patients and controls (p=0.0079); however, no significant changes were apparent between strong and weak ALS/FTD cases and controls (p>0.99) or ALS/FTD cases and pure FTD cases (p=0.0476).
CHCHD10 protein expression is decreased in the cortex but not the spinal cord in a transgenic TDP-43 mouse model with disease progression
Analysis of human tissues suggested that CHCHD10 protein level changes may contribute to ALS and FTD even in the absence of a CHCHD10 mutation. However, studies of human autopsy tissue are unable to indicate whether these changes begin early in disease. We therefore investigated the timeline of changes in CHCHD10 protein levels in cortex and spinal cord of a TDP-43 mouse model of disease characterised by accumulation of pathological TDP-43 in neurons.25 CHCHD10 expression was examined in rNLS TDP-43 transgenic mice at 2, 4 and 6 weeks posttransgene induction (corresponding to disease onset, early and mid-stages of disease, respectively) and littermate controls. Western blot analysis demonstrated no significant difference in CHCHD10 protein levels in cortex tissue in rNLS TDP-43 mice at 2 and 4 weeks posttransgene induction compared with littermate controls (figure 5A,B,C,D). However, in mice at 6 weeks posttransgene induction, a significant decrease in CHCHD10 protein level was observed in comparison to littermate controls (p=0.0416) (figure 5E,F). Neuronal content as analysed by immunoblotting did not differ significantly between control and rNLS TDP-43 mice at any of these time points despite a trend towards decrease at 6 weeks (online supplementary figure 4), suggesting that CHCHD10 loss may be independent of neuron loss in these animals. In the spinal cord, there was no statistically significant difference in CHCHD10 protein levels between rNLS TDP-43 mice and littermate controls at any of the 2, 4 or 6 week timepoints (online supplementary figure 5). These data suggest that changes in CHCHD10 protein level are a relatively late feature of TDP-43-linked disease and are a more prominent feature of pathology in the cortex rather than the spinal cord in this model.
Mitochondrial dysfunction has long been recognised in ALS and FTD patients; however, its role in disease onset and progression remains unclear. Following the initial report of ALS/FTD causal mutations in the mitochondrial protein encoding gene CHCHD10,6 and varied findings from subsequent genetic and functional studies, we sought to further assess the contribution of both genetic and pathological variation in CHCHD10 to ALS/FTD pathogenesis.
Three previously reported ALS and/or FTD-linked genetic variants in CHCHD10 (p.Pro34Ser, p.Pro80Leu and p.Pro96Thr) were present in the Australian ALS and FTD patient cohorts. Notably, none of the patients who carried these variants had characteristics of mitochondrial disorders. All three variants were also present at similar frequencies in control databases, thus their role in disease remains to be confirmed. Since its initial report in ALS/FTD,19 41 the p.Pro34Ser variant has been shown to be a population-based SNP underrepresented in publicly available control cohorts.20 21 42 One population-based CHCHD10 intronic SNP (rs62241575) was implicated as a potential SALS risk allele when compared with NFE gnomAD controls. However, this was not replicated using Australian control cohorts, suggesting that the putative disease association may instead reflect Australian population structure. Interestingly, we also found that the p.Pro80Leu variant reported as pathogenic by Ronchi et al,41 was overrepresented in SALS patients when compared with NFE gnomAD controls. However, this too was not supported when using Australian control cohorts. Notably, in their initial report, Ronchi et al41 identified p.Pro80Leu in two SALS patients and found it to be absent from the 1000 Genomes and Exome Variant Server control databases, and an additional 286 Italian controls (totalling ~7500 control individuals). Here, we utilised data from over 65 000 healthy individuals across three control cohorts, providing greater power to identify novel or rare variants, and disease association. These data suggest that genetic variation in CHCHD10 is not a common cause of ALS or FTD in Australia.
Our results add to the mixed evidence surrounding the genetic contribution of CHCHD10 to ALS/FTD. As is the case for many other ALS genes, it is likely that there are population differences in the frequency of CHCHD10 variation. This may explain the absence of CHCHD10 from our Australian cohort in contrast to their repeated identification in French and other European cohorts.6 7 9–12 16 Though putative causal ALS/FTD mutations have been reported in cohorts of various ethinicities,7–16 as reported here, other studies have found CHCHD10 mutations to be absent from ALS/FTD cohorts of similar ethnicities, particularly pure ALS cohorts.17 18 Together, this suggests that genetic variation in CHCHD10 is a rare cause of ALS/FTD. The pathogenicity of the reported CHCHD10 mutations remain to be confirmed.20 21 42 43 Interestingly, Tazelaar et al43 have also suggested that mutations in CHCHD10 are most likely associated with more complex phenotypes, where ALS/FTD is accompanied by clinical features indicative of mitochondrial defects, such as myopia or deafness.
Immunohistochemical analysis in postmortem tissues showed that CHCHD10 is primarily expressed in neurons in spinal cord, motor cortex and frontal cortex and colocalises with a mitochondrial marker in spinal cord neurons. This, together with data from other studies,6 suggests that CHCHD10 may play a specific role in neuronal mitochondria. Occasional CHCHD10 inclusion-like structures were also observed in this study. Future work in an extended cohort is required to determine the prevalence and biological consequence of these structures.
Previous studies of mutant CHCHD10 protein have implicated both loss-of-function and gain-of function mechanisms. Several studies have reported decreased CHCHD10 protein levels in fibroblasts and lymphoblasts derived from CHCHD10 mutation positive patients.22 24 44 45 However, knockout of CHCHD10 does not cause overt phenotypes in mice, leading others to argue that a gain of toxic function, rather than depletion of CHCHD10 is associated with disease.46 47 Here, we showed that CHCHD10 is significantly reduced in ALS spinal cord tissues, suggesting that there is a loss of CHCHD10 function at least at the end stage of the disease. This reduction of CHCHD10 may lead to MICOS reduction, abnormal mitochondrial cristae structure and synaptic damage leading to motor neuron death.23 48 It is also worth noting that while CHCHD10 protein expression levels were significantly higher for control samples than ALS cases, there was considerable variation, particularly among control samples (figure 3B and supplementary table 2). A larger study is required to further investigate changes to CHCHD10 expression. Interestingly, we did not see CHCHD10 expression changes in motor cortex tissues, despite these two tissue types having been derived from the same patient cohort. However, CHCHD10 protein levels were analysed using two different methods in these tissues: IF quantification specifically in spinal cord motor neurons and western in motor cortex total lysate. As CHCHD10 is expressed primarily in neurons, analysis in a total tissue lysate may mask changes only seen in neurons. To overcome this, we have also semi-quantified CHCHD10 level in Betz cells in a small cohort where motor cortex tissue sections were available and found no significant difference between control and ALS patients. It is therefore possible that there is a cell-type specific difference, which results in differential CHCHD10 protein levels between these two central nervous system regions. We also found a significant reduction of CHCHD10 in pure FTD patient frontal cortex tissues, but not ALS/FTD patient. Interestingly, previous genetic reports have also indicated that CHCHD10 mutations were seen more frequently in FTD patients than pure ALS cohorts.8 17–21 Currently, it is unknown what molecules or pathways influence whether a patient develops ALS or FTD, or comorbid disease. Although a larger study is warranted, our data suggest that CHCHD10 may be more relevant to FTD pathology, and that this protein and its related pathways may help delineate the pathogenic mechanisms distinguishing these two related diseases.
We used a rNLS TDP-43 mouse cortex and spinal cord total tissue lysates, rather than a cell-type specific analysis, to assess whether CHCHD10 reduction was an early event or only occurred during the end stage of disease, which is otherwise not possible to assess with human tissues. We found a significant decrease in CHCHD10 protein levels in rNLS TDP-43 transgenic mice motor cortex tissue after extended disease progression, but not in transgenic mice at disease onset or early disease stages, suggesting CHCHD10 reduction may be a late, downstream result of TDP-43 abnormalities. In contrast, mouse spinal cord tissues did not show significant CHCHD10 reduction at any of the disease stages, suggesting that the mouse cortex may be more susceptible to changes in CHCHD10 protein levels than spinal cord. The loss of CHCHD10 seen in mouse tissues could be partially due to the loss of neurons in the rNLS TDP-43 mouse as cortical atrophy can be observed from 4 weeks off-Dox.25 However, we did not find a significant decrease in neuronal content in motor cortex tissues at any tested time point, suggesting that the observed CHCHD10 loss was independent of neuronal loss. Combining results from human autopsy and the TDP-43 transgenic mouse, we propose that prolonged exposure of pathological TDP-43 is likely to be a trigger of CHCHD10 protein level reduction, which may explain why knockout of CHCHD10 alone does not produce ALS/FTD phenotypes in mice that do not have TDP-43 pathology.46 Our data also provide an additional link between CHCHD10 and TDP-43, which complements the findings of Woo et al,23 who showed that knockdown, or the expression of mutant CHCHD10 increases the accumulation of cytoplasmic TDP-43. Our study further suggests that the accumulation of cytoplasmic TDP-43 can trigger pathological changes in wild-type CHCHD10.
The reduced expression of CHCHD10 in post-mortem patient neuronal tissue and during late disease stages in the rNLS TDP-43 mouse suggests that CHCHD10 plays a role in disease progression. Previous reports, particularly familial studies, have implicated mutant CHCHD10 in disease onset. Other reported disease associated CHCHD10 variants show weaker evidence and may be false positives, particularly as CHCHD10 has been shown to be relatively tolerant to variation.49 Nevertheless, there is combined evidence that support the role of abnormal CHCHD10 in both onset and progression of disease.
In summary, we have demonstrated that genetic variation in CHCHD10 is not a common cause of ALS/FTD among Australian patients. However, loss of CHCHD10 is evident in both ALS and FTD patient neuronal tissue and likely results from TDP-43 abnormalities. Further study is required to fully characterise the molecular changes linking pathological TDP-43 and CHCHD10.
The authors thank L Adams, A Crook, C Cecere and J O’Connor for their assistance in sample collection and compiling family information, the Genome Aggregation Database (gnomAD) and the groups that provided exome and genome variant data to this resource (a full list of contributing groups can be found at http://gnomad.broadinstitute.org/about), Paul Leo, Emma Duncan and Matthew Brown for access to whole exome data from the Diamantina Australian Control Collection 1.0, the MGRB Collaborative (http://sgc.garvan.org.au/mgrb/initiatives) for use of Australian WGS data, the New South Wales Brain Bank and Sydney Brain Bank for providing tissues, and CMRI for access to the Virtual Microscope ScanScope Unit.
Twitter @NeuroBritt, @Dr_KLWilliams, @jnnp_bmj, @WalkerNeuroLab, @@BlairMNDgroup
EPM, JAF, IPB and SY contributed equally.
Correction notice This article has been corrected since it was published Online First. The funding statement has been amended, and ORCIDs added.
Contributors Study concept and design: EPM, JAF and SY with input from JA, AKW and IPB. Acquisition of data: major contribution from EPM, JAF, SY and NG and contributions from JG, SCMF, KLW, ALH, PM, SEF, KYZ, SSLW, CJJ and BAB. Analysis (including statistical) and interpretation of the data: EPM, JAF, NG, JG and SY. Data processing: NAT. Statistical analysis only: LH. Collection of clinical information and samples: OP, JH, JBJK, GMH, MCK, DBR and GAN. Study supervision: SY, DB, JA and IPB. Manuscript preparation: EPM, JAF, NG and SY. Critical revision of the manuscript: EPM, JAF, NG, SY, AKW, JA and IPB. All authors read and approved the final manuscript.
Funding This work was funded by Macquarie University (PhD scholarship to EPM), the Motor Neuron Disease Research Institute of Australia (Bill Gole Postdoctoral Research Fellowship to JAF, PhD scholarship top-up to EPM and BAB), MND Australia (Leadership Grant to IPB), the National Health and Medical Research Council of Australia (NHMRC)(grants 1095215, 1092023, 1124005 and RD Wright Career Development Fellowship 1140386 to AW), an Australian Government Research Training Program (scholarship to BAB), the Ross Maclean Fellowship and the Brazil Family Program for Neurology. The Brain and Mind Centre ALS and FTD cohorts were collected through Forefront, a collaborative research group dedicated to the study of frontotemporal dementia and motor neuron disease, funded by the NHMRC (grants 1037746, 1095127 and 1132524). The Diamantina Control Cohort includes data obtained from projects funded by NHMRC Project Grants 1032571 and 511132. GMH is supported by a NHMRC Senior Principal Research Fellowship (grant 1079679). The postmortem brain tissue cohorts were collected by the New South Wales Brain Banks Network with the Sydney Brain Bank supported by the University of New South Wales and Neuroscience Research Australia and the New South Wales Brain Tissue Resource Centre supported by the National Institute on Alcohol Abuse and Alcoholism, NIHR28AA012725.
Competing interests None declared.
Patient consent for publication Not required.
Ethics approval Animal Ethics Committee of Macquarie University (approval #2015-042).
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement Data are available upon reasonable request.
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.