02 Oct 2020

Being a segregase, it’s the job of valosin-containing protein (VCP) to straighten out a myriad of poorly folded proteins. We can now add tau to that list. In a paper published in Science on October 1, researchers led by Edward Lee at the University of Pennsylvania, Philadelphia, report that a rare mutation in the VCP gene causes an autosomal-dominant form of frontotemporal dementia. Autopsy samples—and even a brain biopsy—from carriers of the D395G mutation revealed neurons riddled with vacuoles, and also with neurofibrillary tangles of tau that appear to be much like those found in Alzheimer’s disease. The researchers christened this neuropathology “vacuolar tauopathy.” Biochemical studies, as well as findings from mutant knock-in mice, indicated that the mutation derails VCP’s “disaggregase” activity, allowing a build-up of tangles.

The pAsp395Gly mutation in VCP causes autosomal-dominant FTD.
The mutation disrupts its ability to break up tau aggregates.
Carriers accumulate endocytic vacuoles and tangles in their brains.

“The identification of a new mutation in a chaperone protein that causes a novel tauopathy is very exciting,” commented Laura Blair of the University of South Florida in Tampa. “Links between VCP and tau have been made over the last 10 years by a number of groups, and these data significantly strengthen this connection.”

Manuela Neumann of the German Center for Neurodegenerative Diseases in Tübingen thinks that while this VCP mutation might be very rare, the realization that VCP is a tau disaggregase could have implications for other tauopathies.

As an ATPase Associated with diverse cellular Activities (AAA+), VCP harnesses the power of ATP hydrolysis to perform diverse functions. It yanks proteins out of membranes, sorts them for interaction with other proteins, and unfolds them to hasten their destruction by proteases (for review, see van den Boom and Meyer, 2018). When VCP malfunctions, numerous biological pathways get disrupted (Abisambra et al., 2013; Mar 2013 news).

Mutations in VCP have been implicated in a range of diseases, including multisystem proteinopathy (MSP), which afflicts bone, muscle, and neurons; amyotrophic lateral sclerosis; and frontotemporal dementia (Watts et al., 2004; Dec 2010 news; Mar 2013 news). In ALS/FTD, inclusions of TDP-43 and ubiquitin are the predominant underlying pathology associated with VCP variants (Neumann et al., 2007).

That was what Lee, a neuropathologist, expected to find when he autopsied the brain of a VCP mutation carrier. Instead, he found something entirely different: neurofibrillary tangles of tau, and neurons overloaded with vacuoles. Nary a TDP-43 aggregate.

It turned out that this particular mutation meddles with VCP in a different way than do its other known variants. Co-first authors Nabil Darwich and Jessica Phan and colleagues found the D395G variant in three siblings who had behavioral-variant FTD. One unaffected parent did not carry the mutation, and the other, for whom genetic data was not available, had died around the age of 65 with a diagnosis of FTD. The aspartate-to-glycine swap appeared to be inherited in an autosomal-dominant fashion; all carriers developed FTD even though only two generations are available in this pedigree. The D395G mutation was absent from several large genomic databases, including gnomAD, the Alzheimer’s Disease Genetics Consortium, and the 1000 Genomes Project, suggesting that it is exceedingly rare. The researchers had been following the family for several years when one sibling died at age 55.

Alzheimer-Like Tau. Antibodies to multiple forms of tau present in AD bound inclusions in the brain of a DG-VCP mutation carrier with FTD. Electron microscopy (bottom right) revealed a tau fibril. [Courtesy of Darwich et al., Science, 2020.]

Upon autopsy, the researchers found tau tangles that resembled those seen in AD. Immunohistochemistry, biochemical studies, and electron microscopy indicated that the tangles consisted of paired helical filaments of phosphorylated 3R and 4R isoforms of tau (see image above).

However, unlike in AD, and much like in FTD, the carrier’s tangle burden was highest in the frontal cortex. Some also appeared in the temporal and parietal cortices. Their distribution correlated strongly with neuronal loss, which was consistent with the typical pattern for frontotemporal lobar degeneration. Gliosis also tracked with the distribution of tangles and neurodegeneration. Subsequently, a tau-PET scan of one of the living affected siblings revealed tau tangles that co-localized with atrophy detected by MRI.

**Vacuolar Tauopathy.** In a DG-VCP mutation carrier, the back of the brain suffered vacuolization (top), while the frontal, temporal, and parietal cortices developed tauopathy. *[Courtesy of Darwich et al., Science, 2020.]*

Where were the vacuoles? Oddly, these seemingly empty spaces had opened in brain regions mostly spared from tangles, for example in the occipital neocortex. The vacuoles crowded near neuronal membranes and expressed the marker EEA-1 on their membranes, suggesting they were endocytic. Unlike the notorious vacuoles detected in prion disease, these vacuoles did not coincide with regions ravaged by neurodegeneration.

Lee told Alzforum that he does not know what causes this vacuolization. He suspects it could arise from a malfunction in sorting endosomes, since VCP is known to play a role in this process via its interaction with EEA-1. In the brain of the mutation carrier, VCP did not associate with vacuoles. The findings suggested that the VCP mutation could lead to vacuolization and tauopathy via two distinct mechanisms.

As Lee’s group was writing their paper, a case of FTD with the same VCP mutation emerged from a genetic study in Greece co-led by Sokratis Papageorgiou (Ramos et al., 2019). Lee reached out to the Greek researchers, who had the patient’s family medical history. There, too, the mutation was associated with what appeared to be an autosomal-dominant pattern of inheritance, over three generations. It was absent in the proband’s two unaffected siblings, and though genetic data was unavailable on the parents or grandparents, one parent and one grandparent had died with dementia with FTD-like behavioral symptoms around age 50.

An MRI scan revealed a pattern of hyperintensity outlining cortical gyri, called cortical ribboning. Because this is a known feature of a vacuolized cortex and can flag prion disease, the Greek patient underwent a frontal biopsy. The researchers shared some of the sample with Lee, who confirmed the tissue was laden with tau tangles.

How might this mutation affect VCP function? D395 lies within the D1 ATPase domain (see below), and in silico analysis indicated the aspartate-to-glycine swap would destabilize it. Indeed, in vitro, recombinant D395G VCP had one-third less ATPase activity than normal VCP. In contrast, another VCP mutation known to cause MSP and ALS/FTD had the opposite effect, raising ATPase activity. This gain of function has been documented for several VCP mutations that cause MSP (Manno et al., 2010; Blythe et al., 2019).

**Toggling ATPase.** The pAsp395Gly mutation (red) resides in the D1 ATPase domain of VCP and inhibits ATPase activity. VCP mutations that cause MSP (blue) lie closer to the N-terminus and enhance ATPase activity. *[Courtesy of Darwich et al., Science, 2020.]*

How might this loss of a VCP function goad tau pathology? A clue emerged from postmortem study. In brain samples from a person with AD, their wild-type VCP mingled with aggregated tau in dystrophic neurites around plaques. In samples from an FTD patient with the DG-VCP mutation, mutated VCP steered clear of aggregated tau. Perhaps wild-type VCP serves as a disaggregase, unfolding tangles? In AD, this activity may ultimately be insufficient to suppress tangles, the scientists believe.

Subsequent biochemical experiments supported this hunch. The researchers mixed insoluble tau aggregates extracted from an AD brain with recombinant wild-type VCP, and its co-factors UFD1L and NPLOC4, and measured thioflavin S fluorescence to gauge subsequent changes to tau fibrillization. When ATP was added to the reaction, tau fibrillization dropped by a third compared to reactions that included a non-hydrolyzable form of ATP. This suggested that active, wild-type VCP detangled tau fibrils. Even in the presence of ATP, D395G-VCP was a poor tau detangler, leading to about half the drop in tau aggregation as wild-type VCP. In line with previous reports that VCP recognizes poly-ubiquitin-tagged substrates, ubiquitin was required for wild-type VCP’s disaggregase activity. Biosensor cell lines using FRET to report tau aggregation told a similar story.

To learn how the D395G mutation would affect tau aggregation in the brain, the scientists used CRISPR gene editing to generate knock-in mice expressing one or two copies of D395G-VCP in place of their normal gene. VCP and tau levels in the knock-ins were on par with wild-type controls, and they had no overt neurodevelopmental problems or neurodegeneration. They also had no tau tangles, in keeping with the idea that VCP does not initiate tau aggregation, but untangles it. Since mouse tau does not fibrillize spontaneously, the researchers had to introduce human pathological tau to test the effect of the mutation on detangling. They injected tau aggregates extracted from AD brain into the dorsal hippocampi and overlying cortices of both wild-type and knock-ins. Three months later, hyperphosphorylated, aggregated tau had crept into neuroanatomically connected brain regions. Mice expressing one or two copies of DG-VCP had more severe pathology in several of these regions, including the retromammillary nucleus, entorhinal cortex, and ventral hippocampus.

Together, the data suggested that VCP acts as a tau disaggregase and that D395G thwarts this, promoting tau pathology.

Lee thinks there are other possible consequences. For example, altered VCP trafficking and substrate binding could explain why wild-type VCP associated with tangles in the AD brain, but D395G-VCP did not in the two FTD patients who carry the variant.

The role of VCP in other tauopathies, including AD, is also unclear, said Lee. A previous study found slightly less VCP in AD brain than in controls, as well as a build-up of phosphorylated tau in VCP-deficient cells (Dolan et al., 2011).

“It’s hard to wrap your head around how many things this protein does,” said Lee, complicating it as a potential therapeutic target, he said. If VCP activity could be enhanced to specifically turn over poly-ubiquitinated substrates, it is possible that a workable therapeutic range might be found, he said. If D395G causes tau pathology in humans, why do other VCP mutations that increase its ATPase activity lead to neurodegeneration with TDP-43 pathology? Lee said these mutations could interfere with other VCP functions and/or disrupt specific interactions between VCP and TDP-43.

“I find it particularly surprising that other proteins known to aggregate in a poly-ubiquitinated form, such as α-synuclein or TDP-43, seem not to be affected by this VCP impairment,” wrote Neumann. “While it would have been informative to include preparations of ubiquitinated forms of α-synuclein and TDP-43 in the assays, dissection of the potentially specific mode of action of VCP on tau aggregates will be very interesting in the future,” she wrote.

Rita Guerreiro of the Van Andel Research Institute in Grand Rapids, Michigan, noted that it is rare to see such an in-depth study of one mutation. “This finding clearly indicates the need to fully study different mutations in the same gene and not assume all have the same effects on disease” she wrote (full comment below).—Jessica Shugart

Comments

- Marc Diamond
  University of Texas, Southwestern Medical Center
- Sushobhna Batra
  UT Southwestern Medical Center
- Posted: 02 Oct 2020
This study from Dr. Edward Lee’s group provides an exciting glimpse into the potential role of valosin-containing protein (VCP) in ameliorating tau aggregation. The work highlights a novel autosomal-dominant mutation (D395G) in the D1ATPase domain of VCP which results in early onset neurodegeneration in families. The associated neuropathology shows an interesting inverse relationship between the location of vacuoles and Alzheimer’s-like tau tangles in the brain. The authors have termed this frontotemporal degeneration “vacuolar tauopathy.”

Through several in vitro experiments, the authors present evidence that the identified mutation inhibits the ATPase activity of VCP, decreasing its disaggregase activity on tau fibrils in vitro. They have further determined that a polyubiquitin signature on the substrate is essential for VCP’s disaggregation effect. Finally, the observation of increased tau pathology and spread in the knock-in mice expressing the mutant form of VCP recapitulates an essential function that it might play in curbing tau pathology, adding an important physiological test to their study.

This finding is a major contribution to the field. It promises to be an important lead into further dissecting the mechanism(s) of tau aggregation or disaggregation. It lays a solid foundation of what could be an important role of VCP in tauopathies, and it raises several questions regarding our understanding of the tau assembly formation/breakdown process. To begin, it will be interesting to determine the stage(s) of tau aggregation (initiation/maintenance/clearance) and species of tau (monomer/oligomers/ aggregates) affected by VCP.

Does tau ubiquitination precede VCP interactions, and, if so, what are the precise site(s) of ubiquitination, are they unique for distinct tauopathies, and how might ubiquitinated tau be routed for downstream processing? It is also intriguing, in light of other disaggregases such as HSP104 in yeast, to consider whether VCP might actually promote seeding in certain contexts. Given VCP’s links to the proteasome system, autophagy, and even stress granules, to name a few, it will be important to parse the pathways (or combinations of them) specific to clearance of tau fibrils. It will also be very interesting to determine the characteristics of the fragments generated by VCP’s disaggregase activity. For example, does the breakdown by VCP create benign tau species incapable of aggregating or does it release pathogenic monomers/oligomers that can continue to function as seeds?

Overall, Lee and colleagues should be congratulated on their rigorous work, which has linked disparate diseases through a common factor, and has set the stage for many further exciting studies.

View all comments by Marc Diamond View all comments by Sushobhna Batra

- Laura Blair
  University of South Florida
- Posted: 02 Oct 2020
This is an impressive new paper. The identification of a new mutation in a chaperone protein that causes a novel tauopathy is very exciting. Links between VCP and tau have been made over the last 10 years by a number of groups. These data significantly strengthen this connection and provide strong evidence of a newly identified functional role for VCP in disaggregating AD-tau PHFs in an ATP- and polyubiquitin-dependent manner.

Future studies are needed to determine how this VCP mutation affects autophagy, proteasome degradation, and ER stress and if these factors contributed to the tau accumulation in vivo.

View all comments by Laura Blair

- Rita Guerreiro
  Van Andel Institute
- Posted: 02 Oct 2020
This is a really interesting study. These are small families but, for the first family, even though segregation is not perfect, this is most probably the causative mutation. The fact that the same mutation is seen in another non-related individual with the same phenotype also adds evidence to the causative role.

This mutation is not present in HEX or TOPMed, and, as the authors mention, also not in gnomAD.

It's not common to have this in-depth study of one mutation. The results show how valuable these can be: Different phenotypes can arise from mutations in the same gene and, at the same time, different mechanisms can lead to a same general outcome (neurodegeneration). This finding clearly indicates the need to fully study different mutations in the same gene and not assume all have the same effects on disease.

View all comments by Rita Guerreiro

- James Shorter
  University of Pennsylvania School of Medicine
- Posted: 02 Oct 2020
This paper is a very interesting set of studies, which begin to reveal how a neurodegenerative tauopathy might emerge from a specific mutation in the AAA+ ATPase, VCP. VCP appears to antagonize tau aggregation, and the disease-linked VCP mutation appears to yield a VCP hypomorph with reduced activity.

It remains uncertain why tau is specifically affected and not other neurodegenerative disease proteins, such as TDP-43. It will be of great interest to determine the precise mechanism by which VCP antagonizes tau aggregation and whether this activity might be harnessed therapeutically in other settings.

View all comments by James Shorter

- Manuela Neumann
  University of Tübingen and DZNE AG Neumann
- Posted: 02 Oct 2020
This is a very well-performed study with very interesting findings, namely the identification of two families with a novel VCP mutation associated with a new type of vacuolar tauopathy.

While such families might be very rare (at least I am not aware that I have seen such cases in our brain banks in the past), the finding with altered disaggregase activity of this VCP mutation leading to tau pathology is very exciting with impact for future drug development approaches in tauopathies in general.

I find it particularly surprising that other proteins known to aggregate in a poly-ubiquitinated form such as α-synuclein or TDP-43 seem not to be affected by this VCP impairment (at least not resulting in aggregates by neuropathological analysis). While it would have been informative to include preparations of ubiquitinated forms of α-synuclein and TDP-43 in the assays, dissection of the potentially specific mode of action of VCP on tau aggregates will be very interesting in the future.

View all comments by Manuela Neumann

- Carmen Nussbaum
  Ludwig-Maximilians-University of Munich
- Eliana Nachman
  VIB-KU Leuven Center for Brain & Disease Research
- Posted: 06 Oct 2020
In this exciting work, Darwich et al. characterize an autosomal-dominant D395G mutation in the AAA+ ATPase chaperone VCP, which is associated with "vacuolar tauopathy" (VT), a novel form of tauopathy. Dr. Edward Lee's group provides compelling evidence that recombinant wild-type VCP and its cofactors UFD1L and NPLOC4 are capable of disaggregating Tau fibrils and that the VT-linked mutation exhibits decreased disaggregation activity. This correlated with a more severe Tau pathology in a mouse model carrying the D395G mutation, suggesting that reduced resolubilization of aggregated Tau causes VT.

This is an intriguing finding, especially with respect to other cellular amyloid disaggregation systems. The cytosolic Hsp70 disaggregation machinery, consisting of the constitutively expressed Hsc70 and its co-chaperones, a class B J- domain protein and an Hsp110-type nucleotide exchange factor (NEF) (Gao et al., 2015; Nachman et al., 2020), or the proteasome (Cliffe et al., 2019), are also able to disentangle Tau fibrils, but at the cost of generating smaller fragments.

For example, we have recently shown that amyloid Tau disaggregation by the Hsp70 chaperone system liberated both small oligomeric and monomeric species, which were able to induce self-propagating Tau species in a biosensor cell line (Nachman et al., 2020). This Hsp70 disaggregase could therefore have a similar function as the Hsp104-Hsp70 bichaperone system in yeast prion propagation and contribute to the accumulation of toxic amyloid conformers. In support of this idea, we were able to demonstrate that reduction of the essential co-chaperone Hsp-110 reduced the aggregation, toxicity and intercellular spreading of α-synuclein in a C. elegans model (Tittelmeier et al., 2020). Therefore, it would be very interesting to characterize the products of VCP-mediated Tau disaggregation regarding their seeding and spreading competence.

The Hsp70 disaggregation machinery targets a broad spectrum of amyloid structures and showed activity against α-synuclein, all six Tau variants, and HTTExon1Q48 (Gao et al., 2015; Nachman et al., 2020; Scior et al., 2018). In contrast, VCP seems to have an extremely narrow substrate spectrum and to specifically target ubiquitinated Tau. Further investigation to determine the substrate spectrum of VCP, e.g., its effect on different Tau conformers, would be interesting. How is VCP embedded in the highly interconnected proteostasis network? Could other co-factors extend its substrate specificity? Is VCP-mediated disaggregation directly coupled to substrate degradation, and therefore more beneficial in vivo? And with regard to the severe vacuolization observed in these patients: Does VCP play a role in the intra- and/or intercellular transport of Tau?

This excellent paper highlights the role of chaperone action in the etiology of neurodegenerative diseases. Future work will shed light on how VCP and other chaperone amyloid disaggregases contribute to disease progression in various proteinopathies, and it will be exciting to explore how such systems can be exploited as therapeutic targets.

References:

Cliffe R, Sang JC, Kundel F, Finley D, Klenerman D, Ye Y. Filamentous Aggregates Are Fragmented by the Proteasome Holoenzyme. Cell Rep. 2019 Feb 19;26(8):2140-2149.e3. PubMed.

Gao X, Carroni M, Nussbaum-Krammer C, Mogk A, Nillegoda NB, Szlachcic A, Guilbride DL, Saibil HR, Mayer MP, Bukau B. Human Hsp70 Disaggregase Reverses Parkinson's-Linked α-Synuclein Amyloid Fibrils. Mol Cell. 2015 Sep 3;59(5):781-93. Epub 2015 Aug 20 PubMed.

Nachman E, Wentink AS, Madiona K, Bousset L, Katsinelos T, Allinson K, Kampinga H, McEwan WA, Jahn TR, Melki R, Mogk A, Bukau B, Nussbaum-Krammer C. Disassembly of Tau fibrils by the human Hsp70 disaggregation machinery generates small seeding-competent species. J Biol Chem. 2020 Jul 10;295(28):9676-9690. Epub 2020 May 28 PubMed.

Scior A, Buntru A, Arnsburg K, Ast A, Iburg M, Juenemann K, Pigazzini ML, Mlody B, Puchkov D, Priller J, Wanker EE, Prigione A, Kirstein J. Complete suppression of Htt fibrilization and disaggregation of Htt fibrils by a trimeric chaperone complex. EMBO J. 2018 Jan 17;37(2):282-299. Epub 2017 Dec 6 PubMed.

Tittelmeier J, Sandhof CA, Ries HM, Druffel-Augustin S, Mogk A, Bukau B, Nussbaum-Krammer C. The HSP110/HSP70 disaggregation system generates spreading-competent toxic α-synuclein species. EMBO J. 2020 Jul 1;39(13):e103954. Epub 2020 May 25 PubMed.

View all comments by Carmen Nussbaum View all comments by Eliana Nachman