I think this is nice, convincing work. However, lots of new questions occur and remain to be answered.

Rosen et al. clearly show (with an overwhelming set of data) for the first time that reduced progranulin (GRN) levels in shRNA-expressing human neuronal progenitor cells, in brains from GRN knockout mice, or in brains of FTLD patients with GRN loss-of-function mutations, result in upregulation of activating components of the Wnt signaling pathway, whereas inhibitors of the Wnt pathway are downregulated. The enhanced Wnt signaling due to GRN deficiency was also observed in mature differentiated cells, and did not depend on cell proliferation. This is of particular interest, since in neurodegenerative diseases, adult differentiated neurons are affected. How GRN expression mechanistically affects the expression of components of the Wnt pathway is not addressed by the authors.

It would be interesting if overexpression of GRN has the opposite effect. In schizophrenia, increased NRG1-, BDNF- and TGF-β signaling and decreased Wnt signaling has been reported (  Read more

I think this is nice, convincing work. However, lots of new questions occur and remain to be answered.

Rosen et al. clearly show (with an overwhelming set of data) for the first time that reduced progranulin (GRN) levels in shRNA-expressing human neuronal progenitor cells, in brains from GRN knockout mice, or in brains of FTLD patients with GRN loss-of-function mutations, result in upregulation of activating components of the Wnt signaling pathway, whereas inhibitors of the Wnt pathway are downregulated. The enhanced Wnt signaling due to GRN deficiency was also observed in mature differentiated cells, and did not depend on cell proliferation. This is of particular interest, since in neurodegenerative diseases, adult differentiated neurons are affected. How GRN expression mechanistically affects the expression of components of the Wnt pathway is not addressed by the authors.

It would be interesting if overexpression of GRN has the opposite effect. In schizophrenia, increased NRG1-, BDNF- and TGF-β signaling and decreased Wnt signaling has been reported (Kalkman, 2009). Since Rosen et al. elegantly show that Wnt signaling is beneficial for neuronal survival upon GRN deficiency, it would be interesting to analyze if growth factor withdrawal in general affects Wnt signaling.

Moreover, Wnt signaling is only increased in GRN mutation carriers, not in FTLD-TDP cases without GRN mutation; therefore, Wnt seems to be no general hallmark of FTLD-TDP, and whether increased Wnt signaling is beneficial for all FTLD-TDP cases remains to be shown. In other neurodegenerative diseases, such as AD and PD, altered Wnt signaling has been reported. Also, Wnt plays a role in maintenance of survival of neurons as a kind of synaptotropic factor. Therefore, I think it is unlikely that the Wnt signaling is specifically involved in FTLD-TDP.

Influencing Wnt signaling pathways might provide therapeutic benefits, though it needs to take into account that Wnt also occurs in oncogenic processes. The Wnt pathway would be a negligible diagnostic tool for GRN mutation carriers because GRN expression in any body fluid is the best diagnostic tool. However, variations in Wnt signaling might contribute to the variable age of disease onset within patients carrying the identical GRN mutation. To investigate whether protective Wnt signaling plays a role for the low disease penetrance for GRN mutation carriers might be quite exciting.

Since we have shown that lysosomal alkalization enhances GRN expression (ARF related news story on Capell et al., 2011), I find it very interesting that the study confirmed that GRN plays an important role in lysosomal function, and that lysosomal genes represent the most significant expression changes in nine-month-old GRN-/- mice.

View all comments by Anja Capell

The study by Dan Geschwind and colleagues using WGCNA is novel and highly interesting, not only providing a general view of transcriptional alterations associated with reduced granulin (GRN) expression, but also uncovering a previously unknown link between GRN and Wnt pathways. Consistent findings of changes in expression of apoptosis and ubiquitination pathway genes in GRN-knocked down neurons and frontotemporal dementia (FTD) brain tissues suggest the clinical relevance of the results. This elegant work represents one of the first systematic studies of neural transcriptome changes in GRN-deficient FTD cases, and will likely stimulate further research in both mechanistic understanding of FTDs and new therapeutic development.

View all comments by Jane Wu

These are very interesting data, adding to the many cases of potential pleiotropy that we are starting to see more often by applying next-generation sequencing to the study of neurodegenerative diseases.

Before the availability of these technologies, sequencing would be performed only in the genes known to be associated with a specific clinical picture. Now what we are seeing is that some of these cases have mutations in genes that haven't been directly associated with that specific clinical entity, but rather with a similar disease, in this case, another form of dementia.

The early age at onset for these cases (especially case 2) makes it less probable that AD is co-occurring with FTD just by chance, and suggests the existence of common molecular pathways between FTD and AD. We can easily establish a parallel with TREM2 or C9ORF72. In the TREM2 gene, homozygous mutations were initially identified to cause Nasu-Hakola disease; later, the same types of mutations were found to cause FTD without any bone symptoms and, more recently, heterozygous variants were shown to...  Read more

These are very interesting data, adding to the many cases of potential pleiotropy that we are starting to see more often by applying next-generation sequencing to the study of neurodegenerative diseases.

Before the availability of these technologies, sequencing would be performed only in the genes known to be associated with a specific clinical picture. Now what we are seeing is that some of these cases have mutations in genes that haven't been directly associated with that specific clinical entity, but rather with a similar disease, in this case, another form of dementia.

The early age at onset for these cases (especially case 2) makes it less probable that AD is co-occurring with FTD just by chance, and suggests the existence of common molecular pathways between FTD and AD. We can easily establish a parallel with TREM2 or C9ORF72. In the TREM2 gene, homozygous mutations were initially identified to cause Nasu-Hakola disease; later, the same types of mutations were found to cause FTD without any bone symptoms and, more recently, heterozygous variants were shown to increase the risk for AD. Similarly the expansion on C9ORF72 was initially found in cases diagnosed with ALS, FTD, or ALS/FTD, providing genetic evidence that these diseases are most likely part of a pathological spectrum.

Interestingly, TREM2 is also involved in inflammatory processes, and the roles of microglia and inflammatory cellular status are becoming more central as molecular pathological pathways for different dementias.

These results clearly stress the importance of taking a non-restrictive approach to the screening of genetic variants, which can now be achieved by using targeted sequencing panels or exome/genome sequencing. The identification of more cases like these will allow the study of previously unrecognized associations between genetic variability in the different genes involved in these disorders.

View all comments by Rita Guerreiro

Perry et al. have identified two patients with Alzheimer’s disease (AD) pathology who also have a mutation in the progranulin (GRN) gene. This is an extremely important finding, considering that the GRN mutations are known to cause the autosomal-dominant form of frontotemporal dementia (FTD) through the haploinsufficiency of functional progranulin. This observation is consistent with previous work, which has suggested that certain variants of GRN, such as SNP rs5848, increase the risk of AD in different AD cohorts. Thus, it is possible that the mutations/variants in GRN may also play a role as risk factors in AD. However, that both patients also harbored the ApoE4 allele, which is the most prominent genetic risk factor for AD, emphasizes the importance of conducting further studies before making firm conclusions about GRN in AD.

View all comments by Mikko Hiltunen