Mutations in the progranulin gene cause a good many cases of frontotemporal dementia (FTD), the most common form of early onset dementia. But not everyone with a progranulin mutation gets FTD—some lucky souls are protected by variants in a second gene, TMEM106B. While the genetic interaction is clear, just what TMEM106B does in the cell, and how it overrides destructive progranulin mutations, has been a mystery. Now, a TMEM106B knockout mouse, generated in the lab of Stephen Strittmatter at Yale University, New Haven, Connecticut, shows that the protection takes place in lysosomes. These organelles, the work suggests, are the battlefields where both genes regulate protein degradation in opposite directions. Their respective effects also diverge: Knocking out progranulin caused neuronal death and behavioral deficits, while removing TMEM106B from the progranulin knockouts prevented both these phenotypes. Published in the July 19 Neuron, the work shines a spotlight on lysosomes not only in FTD, but also in a growing number of related conditions, including Alzheimer’s, where TMEM106B affects risk.

“The important finding in this paper is the involvement of lysosomal function in the pathway to neurodegeneration,” said Daniela Galimberti, University of Milan in Italy. “A number of studies in humans implicate TMEM106B or progranulin polymorphisms in a list of different diseases, and nobody has had an explanation for these associations. So this information can be useful in many fields of neurodegeneration,” she told Alzforum.

The paper’s limitation is that genetic knockout poorly reflects how genetic polymorphisms lead to disease in humans, said Galimberti. “We start from the mice to get to very basic mechanisms, but we have to think carefully about how to translate these findings to humans because the model is artificial,” she said. 

Yin and Yang.

Proteomic analysis (LFQ-LCMS) of TMEM106B or progranulin knockout mice reveals reciprocal regulation of multiple lysosomal enzymes. [Courtesy of Neuron, Klein et al.]

TMEM106B does not cause FTD, but it affects if and when people with progranulin mutations, or other causal mutations, get the disease (see Feb 2010 newsCruchaga et al., 2011; Sep 2016 conference news). Pathogenic progranulin mutations cause a loss of progranulin protein production, and protective alleles of TMEM106B boost plasma progranulin levels. This led to the idea that TMEM106B’s protective effect lay in reconstituting progranulin function. TMEM106B is a transmembrane protein that resides in lysosomes and regulates their function (Brady et al., 2013); however, the actual mechanism by which it modifies neurodegeneration is unclear. 

In the new study, first authors Zoe Klein and Hideyuki Takahashi set out to explore how TMEM106B functions and how it interacts with progranulin. They first examined an existing mouse strain lacking Grn, the murine equivalent of the human progranulin gene. The knockouts develop microgliosis, a feature of FTD, and retinal neurodegeneration, a feature of the granulin-related disease neuronal ceroid lipofuscinosis (Hafler et al, 2014). The researchers exhaustively catalogued changes in gene expression and protein levels, using whole-transcriptome sequencing of mRNA from cerebral cortices of two-month-old mice and proteomic analysis of membrane organelles from their forebrains. Both exercises pointed to an early upregulation of lysosome function. In a gene ontogeny analysis of mRNA and protein data, lysosome function topped both lists as most affected in the knockouts compared to wild-type mice. Among 364 mouse lysosome genes, 29 changed expression and 19 altered their protein levels. Of those, the majority went up. That upregulation was accompanied by lifelong increases in activity of lysosome enzymes in the brain. Overall, lysosomal proteolysis in cultured primary cortical neurons jumped by one-quarter in the knockouts compared to wild-type.

In contrast to previous studies (Oct 2012 news), the analysis showed no effect on microglial markers of inflammation in these mice but did reveal upregulation of innate immune proteins of the complement system.

For TMEM106B, the authors created a knockout and performed a similar transcriptome and proteome analysis. Surprisingly, gene transcription in these mice was minimally affected—in total only 54 genes were differentially expressed between the null mice and wild-type littermates. The only differentially expressed lysosomal gene was TMEM106B itself. However, at the protein level, things looked different. Several of the same proteases that went down in the progranulin knockout went up when TMEM106B was absent.

Given the opposite effects of the knockouts, and the human genetic interaction, the researchers crossed the two strains. In the double-null mice, lysosome function normalized: Protease levels and overall proteolysis returned to wild-type conditions when both genes were gone. Curiously, the TMEM106B knockout selectively affected lysosomes, not other phenotypes of the granulin knockout. For example, loss of TMEM106B did not alter lipofuscin accumulation, or expression of the microglial markers CD68 or complement C1q.

How might TMEM106B affect lysosome function? The scientists discovered that several subunits of the vacuolar-ATPase were downregulated. V-ATPase is an enzyme that makes lysosomes acidic, and cortical neurons cultured from the knockouts contained less acidic lysosomes. Co-immunoprecipitation indicated that TMEM106B directly associates with the enzyme’s AP1 subunit. Knocking out progranulin had no effect on V-ATPase.

The results suggest that, in the double knockouts, reduced V-ATPase activity somehow normalizes lysosome activity. In support of this idea, the V-ATPase inhibitor bafilomycin A mimicked the effect of knocking out TMEM106B, decreasing lysosomal enzyme levels and proteolysis in cultured progranulin-null neurons.

What about the behavior phenotype? That, too, returned to normal in the double knockout. Progranulin-null mice are hyperactive; this hyperactivity in an open-field test came down to wild-type levels when TMEM106B was absent. Also, the progranulin knockouts venture out boldly in the elevated plus maze, which may be a model for the disinhibition FTD patients exhibit. This behavior, too, abated in the double knockouts. So did neurodegeneration: The double knockouts preserve retinal neurons that die in progranulin-null animals.

In people, the twin findings that progranulin mutations are loss-of-function and that TMEM106B variants boost serum progranulin have suggested restoring progranulin levels as a therapeutic approach to FTD (Finch et al., 2011). The new work suggests that modifying lysosome activity might be another avenue.

Perhaps the two are related? Several years ago, Anja Capell, Christian Haass and colleagues at Ludwig-Maximilians-University in Munich reported that bafilomycin A, the same V-ATPase inhibitor used in the new work, attenuated lysosome acidification and proteolysis, and boosted progranulin production in neurons from patients with progranulin loss-of-function mutations (Feb 2011 news). On the other hand, the double knockout mice in the current study suggest that holding neurodegeneration at bay does not absolutely require that progranulin be present.

Curiously, in vitro studies reported that overexpression of TMEM106B hampers lysosome acidification (Aug 2012 news; Busch et al., 2016), while the current study observed this same effect by knocking out TMEM106B. The authors speculate in the paper that when it comes to TMEM106B, perhaps too much is as bad as too little, and cells need to get it just right to function properly. Strittmatter declined to be interviewed on the record for this story.

What’s next? Galimberti would like to see additional analysis of the wealth of gene expression and proteomics data generated for the knockouts. “We have a lot of information here in the list of genes described as over- or underexpressed. Now we have to look into the pathways, and go in depth to study each candidate to try to understand the story of neurodegeneration in these mice.” With luck, that may reveal new therapeutic targets, she said.—Pat McCaffrey 


  1. Zoe A. Klein and colleagues in Stephen Strittmatter's lab show that deficiency of TMEM106B partially rescues the lysosomal effects of knocking out progranulin (Grn) in mice.

    Klein et al. show that in Grn knockout mice, increased levels of lysosomal hydrolases, which have been reported previously by several research groups, result in enhanced enzyme activity and consequently in enhanced protein degradation via the lysosomal pathway.

    Interestingly, they convincingly showed that loss of TMEM106B has an opposite effect on activity of lysosomal enzymes.  However, only two out of three analyzed hydrolases show a reduced activity and thereby provide a kind of rescue phenomenon of this particular effect in Grn knockout mice.

    That the loss of TMEM106B rescues the enhanced lysosomal activity in the Grn KO mouse due to less acidification in lysosomes is a quite exciting finding. However, the data that TMEM106B stabilizes the Vo subunit of the vATPase appear somewhat weak, as neither the endogenous V-ATPase Vo or AP1 subunits could be shown on immunoblot. Less acidic lysosomes are expected to change the maturation of many lysosomal hydrolases, which is unfortunately not shown.

    Furthermore, impaired lysosomal acidification most likely results in impaired lysosomal and autophagic protein degradation and finally in a lysosomal storage disease. For example, this has been suggested for CLN3 mutations (Gachet et al., 2005), which are associated with juvenile Batten disease.

    Thus, it may not be a good option to treat a neuronal ceroid lipofuscinosis (NCL) caused by the total loss of Grn with inhibition of lysosomal acidification either with alkalizing drugs or by altering TMEM106B expression, thereby further impairing protein degradation. The most critical question which needs to be answered before deciding on a treatment is how enhanced degradation caused by Grn deficiency results in a neuronal lysosomal storage phenotype. 

    No lysosomal storage disease has been described yet as being the result of enhanced lysosomal function.​

  2. Since the advent of the genome-wide association study, we've found many genetic risk factors for diseases, including TMEM106B as a common variant genetic risk factor for frontotemporal dementia. However, it has been hard to understand what these risk factors do, and how our finding them can ever help a person with that disease.

    An interesting aspect of this paper is that it looks at TMEM106B as a genetic modifier for a Mendelian gene in which haploinsufficiency mutations cause FTD with high penetrance, the GRN gene. The authors show through the use of mouse models that the interaction between GRN and TMEM106B, suggested by cell culture work from our group and others, also occurs in vivo, through specific effects on lysosomes. To me, as a practicing neurologist as well as a scientist, this suggests that developing TMEM106B as a therapeutic target in the thousands of people with FTD due to GRN mutations might be a reasonable approach.

    Further, the paper pinpoints a potential interaction between TMEM106B and components of the vacuolar ATPase. While I would have liked to see a little more data in this area, this finding suggests a specific molecular mechanism for the altered lysosomal acidification seen by the authors and others (including our group) when TMEM106B expression levels are manipulated.

    It is intriguing that only some of the phenotypes in the GRN null mouse are rescued by deletion of TMEM106B—and notably, the accumulation of lipofuscin and the microglial abnormalities seen in this animal model are not rescued. Despite this, the authors find rescue of retinal degeneration, which is promising from the standpoint of thinking about targeting TMEM106B to rescue neurodegeneration in people with FTD due to GRN mutations.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Genetics of FTD: New Gene, PGRN Variety, and a Bit of FUS
  2. FTD Gene Hunt Turns Up CYLD and Modifying Factors
  3. Microglial Progranulin Douses Neural Inflammation
  4. Back to Basics? Boosting pH Puts Cells in Progranulin-Pumping Mode
  5. FTD Risk Factor Confirmed, Alters Progranulin Pathways

Paper Citations

  1. . Association of TMEM106B gene polymorphism with age at onset in granulin mutation carriers and plasma granulin protein levels. Arch Neurol. 2011 May;68(5):581-6. PubMed.
  2. . The frontotemporal lobar degeneration risk factor, TMEM106B, regulates lysosomal morphology and function. Hum Mol Genet. 2013 Feb 15;22(4):685-95. PubMed.
  3. . Progressive retinal degeneration and accumulation of autofluorescent lipopigments in Progranulin deficient mice. Brain Res. 2014 Nov 7;1588:168-74. Epub 2014 Sep 16 PubMed.
  4. . TMEM106B regulates progranulin levels and the penetrance of FTLD in GRN mutation carriers. Neurology. 2011 Feb 1;76(5):467-74. PubMed.
  5. . Increased expression of the frontotemporal dementia risk factor TMEM106B causes C9orf72-dependent alterations in lysosomes. Hum Mol Genet. 2016 Apr 28; PubMed.

Further Reading

No Available Further Reading

Primary Papers

  1. . Loss of TMEM106B Ameliorates Lysosomal and Frontotemporal Dementia-Related Phenotypes in Progranulin-Deficient Mice. Neuron. 2017 Jul 19;95(2):281-296.e6. PubMed.