Genetic analysis of VCP variants in a Turkish dementia cohort

Introduction

Valosin-containing protein (VCP) belongs to the AAA+ ATPase family and is involved in many molecular pathways, particularly concerning protein degradation such as the ubiquitin-proteasome and lysosomal systems (Meyer et al., 2012). VCP contains an N-domain and a C-terminus surrounding two ATPase domains (D1 and D2). Heterozygous mutations in VCP are causative for multisystem proteinopathy (MSP) which can present with a variety of phenotypes including inclusion body myopathy, Paget’s disease of bone, and frontotemporal dementia (IBMPFD) (Figure 1A). Such mutations, mostly located within the N-domain and ATPase domains, have been reported to increase ATPase activity (Niwa et al., 2012). Recent work further characterizing the effects of VCP variants in MSP, reported that they result in faster substrate binding and unfolding when specifically studying the interaction between VCP and the VCP substrates NPLOC4 and UFD1L (Blythe et al., 2017, 2019).

• Download figure
• Open in new tab

Figure 1.

A: A schematic of the phenotypes presenting in multisystem proteinopathy. Phenotypes encompass inclusion body myopathy, Paget’s disease of bone, and frontotemporal dementia. Arrows point to body regions that are commonly affected by each phenotype. B: Partial pedigree tree for the Turkish pDLB patient with VCP p.Ile216Met variant. Square = male; circle = female; diamond = unknown sex; black = individuals affected with dementia; white = individuals with no reported dementia phenotype; diagonal line = deceased individual; double bar = consanguineous marriage; arrow = the proband. Where tested, “+/-” indicates a heterozygous variant carrier and “-/-” indicates an individual not carrying the variant of interest. An affected sibling (IV:1) previously diagnosed with AD had an age of onset in their mid-70s, and a 5-year disease duration. They are currently entirely dependent. An unaffected sibling (IV:2) died from non-dementia reasons. Another unaffected sibling (IV:3) in their early 70s had DNA available and Sanger sequencing was performed to test for this variant. The proband (IV:4) had an age of onset in their early 60s. An affected sibling (IV:5) previously diagnosed with AD had an age of onset in their mid-50s, and 5-year disease duration. They are now fully dependent. It was not possible to obtain a sample from this sibling. A sibling with Down’s syndrome (IV:6) died when they were in their late teens. The unaffected sibling (IV:7) died in their late teens because of an accident. Another unaffected sibling (IV:8) died in their 50s due to non-dementia related reasons. C: Sanger traces for proband and two siblings. These show the heterozygous variant present in both the proband (IV:4) and affected sibling (IV:1), with no variant in the unaffected sibling (IV:3). D: VCP protein domains. Domain positions are as reported by DeLaBarre and Brunger, 2003. VCP p.Ile216Met is marked in red and previously reported variants are marked in yellow (see Supplementary Table 4 for genomic positions and a description of the phenotypes reported for variant carriers). Panels A and D were created with BioRender.com.

In addition to MSP, VCP mutations have also been reported to cause autosomal dominant Charcot-Marie Tooth disease type 2Y (Gonzalez et al., 2014; Jerath et al., 2015) and amyotrophic lateral sclerosis with or without frontotemporal dementia (FTD) (Johnson et al., 2010). VCP has also been investigated in Parkinson’s disease (PD) (Majounie et al., 2012; Regensburger et al., 2017) with decreased VCP expression suggested as a biomarker for early-onset PD (Alieva et al., 2020). VCP variants have also been reported in Alzheimer’s disease (AD); one study specifically screened VCP following linkage near this locus on chromosome 9 and identified VCP:p.Arg92His in 2% of the AD siblings contributing to the linkage peak, and rs10972300-T which demonstrated an association with AD in the US cases but not those from the UK with no further reports of replication (Kaleem et al., 2007; Mehta et al., 2013). Moreover, in one family with VCP:p.Arg155His, amyotrophic lateral sclerosis was reported in two individuals, and neuropathological-confirmed AD was reported in one individual (Kaleem et al., 2007; Mehta et al., 2013). Most recently, VCP p.Asp395Gly was reported in 2 families from the US and Greece with vacuolar tauopathy, a form of FTD characterized by neuronal vacuoles and tau neurofibrillary tangles; the authors suggested that VCP functions as a tau disaggregase (Darwich et al., 2020).

As the number of diseases and populations studied for VCP variants increases, we set out to specifically investigate VCP variants in a Turkish dementia cohort.

Materials and Methods

Results

We identified a VCP coding variant in one individual within the studied Turkish dementia cohort. This was a heterozygous missense VCP variant NM_007126.5:c.648A>G in an individual diagnosed with prodromal dementia with Lewy bodies (pDLB) according to the McKeith criteria (McKeith et al., 2020). This equates to VCP p.Ile216Met at the protein level.

This patient in their mid-60s was first assessed in their early 60s for cognitive decline. They were reported to have cognitive problems two years before initial assessment. The main complaints from both the patient and their family regarded cognitive slowness and perception problems. They had no psychotic or other behavioral symptoms. In cognitive testing their mini-mental state examination score was 29/30, their main cognitive deficit was reduced verbal fluency. They reported multiple autonomic symptoms including urinary incontinence, constipation, orthostatic hypotension, and rapid eye movement sleep behavior disorder. The clinical diagnosis was pDLB in accordance with diagnostic criteria for this disorder (McKeith et al., 2020). Based on whole-exome sequencing data, this proband is homozygous for the APOE e3 allele.

Their family history was remarkable for dementia (Figure 1B). The proband’s relative, now in their early 90s, (individual III:3 in the pedigree) had dementia for three years before death (age of onset in their mid-80s, they died of cancer after five years). The patient had consanguineous parents (aunt-uncle children). The proband’s mother died when she was in her late 80s and was reported to be unaffected at the time of death. The proband’s father died when he was in his mid-70s without dementia. The mother also had consanguineous parents.

To test for segregation, we Sanger-sequenced DNA available from the proband and two siblings. We confirmed the presence of the variant in the proband and observed the same variant in the affected sibling (individual IV:1 in the pedigree) and no variant in the unaffected sibling (individual IV:3 in the pedigree) (Figure 1C).

VCP p.Ile216Met is located within the D1 ATPase domain (Figure 1D) and VCP p.Ile216 is well-conserved across different species, suggesting functional importance of variants at this position (see Supplementary Figure 1 for multiple sequence alignment).. The variant was considered damaging and deleterious by in silico prediction tools including PolyPhen-2, SIFT, and CADD (Kumar et al., 2009; Adzhubei et al., 2010; Kircher et al., 2014). Moreover, it was not reported in either gnomAD, the GME Variome, or the TR Variome. There were also no additional candidate variants (and specifically no homozygous variants) identified in this case when exonic variants in genes known to cause, or be associated with risk of neurodegenerative diseases were screened (see Supplementary Table 2 for this gene list and Supplementary Table 3 for the variants identified in our proband). This suggests that this variant is both rare and is not a common population-specific variant. This proband also does not carry previously published definite or potentially pathogenic variants as summarized in Supplementary Table 1. Furthermore, we did not identify other VCP Ile216Met carriers when we consulted additional datasets (see Methods) or any VCP variant carriers with similar phenotypes when we consulted GeneMatcher (Sobreira et al., 2015).

Discussion

Here we report a heterozygous VCP variant (NM_007126.5:c.648A>G; p.Ile216Met) in a pDLB patient. This missense variant has been previously reported on ClinVar with uncertain significance for IBMPFD (submitted by Invitae).

Although the absence of dementia in the proband’s parents (Individuals III:1 and III:2 in the pedigree) and the multiple loops of consanguinity in the family would suggest an autosomal recessive mode of disease inheritance, there are several reasons why a heterozygous variant could be the cause of disease in this family. The variant might present with incomplete penetrance, or either parent might have developed the disease at a later age. In particular, the father can be considered as a potential carrier based on the observed consanguinity and age at death identical to age at disease onset of sibling IV:1. Moreover, it should be noted that the maternal relative was diagnosed with dementia when in their mid-80s. The clinical picture described in this family carrying the VCP p.Ile216Met variant does not represent a known clinical phenotype for pathogenic VCP mutations (Supplementary Table 4); however, some similarities between the proband described here and previous reports of VCP variant carriers with PD include REM sleep behavior disorder, urinary incontinence, and absence of psychosis (Chan et al., 2012; Spina et al., 2013; Regensburger et al., 2017). Sphincter disturbances were also reported in a large MSP pedigree (Miller et al., 2009). To date, no correlations between mutations and the incidence of specific clinical phenotypes have been described, with the exception of the association of p.R159C with a later age of onset of myopathy and absence of Paget’s disease of bone when compared to other mutations (Al-Obeidi et al., 2018).

A limitation of this current study is the lack of certainty regarding the pathogenicity of VCP p.Ile216Met. Based on the data available, it is not possible to definitively conclude that this is the Mendelian cause of disease. Still, the association of this variant with the disease is supported by several lines of evidence: 1) the limited (due to sample availability) segregation of the variant with disease in the family; 2) the very low population frequency of the variant, being absent from population databases and only reported once in ClinVar; and 3) by similarities with clinical features described for other VCP variant carriers.

If this VCP variant is indeed the cause of disease in the studied family, it may have an effect on disease independently or through the development of Lewy body pathology. Given that VCP immunoreactivity has been shown in Lewy bodies of cases with DLB (Mori et al., 2013), one would expect the latter to be more probable, however, we cannot conclude that this patient has Lewy pathology without autopsy confirmation.

Considering the previous genotype-phenotype correlations for VCP mutations, one might have expected for this VCP variant to present as FTD. In this respect, it is interesting to note that a clinical overlap between FTD and DLB has been reported with some patients described as presenting clinical features suggestive of both FTD and DLB (Claassen et al., 2008). The clinical picture for GRN mutation carriers is also broad, with reports including, among others, DLB and AD (Orme et al., 2020); this is intriguing since GRN variants are a known Mendelian cause of FTD.

DLB is a relatively understudied clinical phenotype accounting for at least 7.5% of all dementias in clinic-based studies (Walker et al., 2015). The potential under- and misdiagnosis of DLB is reinforced by the knowledge that the two affected siblings (one of which had a sample available and was determined to also carry the same VCP variant) had both previously received a diagnosis of Alzheimer’s disease. On the other hand, perhaps this also suggests a pathological mechanism, induced by the VCP variant, that causes protein aggregation and leads to dementia, independently of Lewy body formation.

To the best of our knowledge, this is the first report of a VCP variant in pDLB, and the first report of a potentially pathogenic VCP variant in a Turkish family. Although we have not been able to identify a second DLB case carrying this same variant, it is still important to describe these results to help corroborate potentially similar future findings.

Conflict of Interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Author Contributions

ND, RG and JB designed the study and wrote the manuscript. ND, KW, RG and JB performed data analyses and interpreted the data. ZT, GG, HH, EL, BS, HG, BB and ME collected and characterized samples for inclusion in the study. All authors provided critical feedback and helped shape the research, analysis and manuscript. All authors read and approved the final manuscript.

Funding

There was no specific funding for this study.

Data Availability Statement

Data generated during this study is included in this published article and supplementary information files. We have submitted this VCP variant (NM_007126.5:c.648A>G) to ClinVar (accession number: SCV002524135).