17 February 2010. The genetics of frontotemporal dementia (FTD) just got more interesting. On February 14, Nature Genetics posted a paper—authored by much of the FTD research community—describing a genomewide association study that identified new SNPs linked to frontotemporal lobal dementia with TDP-43 pathology (FTLD-TDP). The February Archives of Neurology brought multiple snippets of news as well. Researchers from the University of Pennsylvania in Philadelphia and the University of Washington in Seattle spearheaded a broad study of the FTD disease spectrum, and came up with a large list of new mutations in progranulin (PGRN) that may be relevant to disease. In the same journal, researchers from the Mayo Clinic in Rochester, Minnesota, report on yet another PGRN mutation with a varying phenotype, and scientists from the Netherlands contribute to the list of possible variants in the gene FUS. “This is an exhilarating time to be clinicians and scientists in the FTLD field,” wrote Bradley Boeve of the Mayo Clinic in Rochester, in an Archives of Neurology editorial accompanying the disease spectrum paper.
Frontotemporal dementia has multiple causes. Approximately half of familial cases are due to mutations in tau; the other half exhibit TDP-43 proteinopathy, due to mutations in TDP-43 itself or in PGRN. A smaller proportion, caused by none of those proteins, is associated with mutations in FUS.
Introducing TMEM106B (You May Not Have Heard of It...)
The plan to go fishing, GWAS-style, for genetic risk factors for FTLD was hatched by Virginia Lee and John Trojanowski of the University of Pennsylvania in Philadelphia. They recruited joint senior author Hakon Hakonarson at Children’s Hospital in Philadelphia and joint first authors Vivianna Van Deerlin, Maria Martinez-Lage, and Alice Chen-Plotkin at the University of Pennsylvania and Patrick Sleiman at Children’s Hospital. Lee invited scientists in the FTLD field to submit samples in 2007, and—a few thousand e-mails later—researchers from 45 sites in 11 countries had anted up. “I would challenge you to find other fields that play together so well,” Trojanowski said of his 99 coauthors. “I think it is an important paper, despite the fact that much of it is names and addresses of authors,” joked Allen Roses of Duke University in Durham, North Carolina, who was not part of the study group. He expects knowledge of the new locus will come in handy in his own work, helping him to identify and separate people with FTD from other subjects in an Alzheimer disease study.
Researchers’ high hopes for GWASs have not always been borne out (e.g., see ARF related news story on Chiò et al., 2009). “This is the kind of GWAS that one might have been skeptical of at the outset, because the numbers were small,” Trojanowski said. “ The FTLD cohort contained 515 cases. The researchers relied on private funding to start the project. The trick was to carefully select a “squeaky-clean” subject pool, Trojanowski said, limited to only TDP-43-positive cases. The researchers confirmed TDP-43 pathology with autopsy tissue in most cases; a handful were people with PGRN mutations. Controls were 2,509 DNA samples, statistically matched to the cases by ethnic background and other characteristics, Van Deerlin said.
Once the analysis was complete, the authors had identified three SNPs, all in the same region of chromosome 7, associated with FTLD status. Each had a p-value in the -10 to the -11 range—impressive significance, Roses noted. In each case, FTLD was associated with the more common allele; having an FTLD SNP increased odds of disease by approximately 1.6. These SNPs were linked to disease both among cases with PGRN mutations and cases without. Of course, a SNP gives only a nearby address, and does not represent the actual disease-relevant genetic lesion. But in this case, Van Deerlin said, there is only one gene in the area, making that gene, transmembrane protein 106B (TMEM106B), a very likely.
If you have never heard of TMEM106B, you are hardly alone. “Nothing is published, nothing is known, no antibodies, no nothing,” Van Deerlin said. To dig into TMEM106B’s potential mechanism, the group evaluated the gene’s expression in the brains of people who had FTLD-TDP. They used quantitative PCR to assay TMEM106B mRNA levels. Those carrying the risk-associated SNPs had higher levels of TMEM106B mRNA in the cerebral cortex, suggesting the SNPs are linked to higher expression of the gene. TMEM106B expression was 2.5-fold higher in tissues from people with FTLD-TDP, compared to unaffected controls.
There is always the possibility that some other gene, in linkage disequilibrium with the identified SNPs, is the true problem. But the expression data make a “compelling” case,” said Kirk Wilhelmsen of the University of North Carolina in Chapel Hill. “Having sort of a smoking gun, that it changes the expression level of a gene, is relatively satisfying,” he said. This potential mechanism is also encouraging when one considers treatment strategies, he noted; dampening a gene’s overexpression is theoretically easier than repairing a protein that is damaged or missing.
The next step, already underway, according to Trojanowski, is to delve into the normal function of TMEM106B and sequence the gene directly in human samples. Also, like any GWAS, replication is necessary. Roses suggested that one further approach is to perform phylogenetic analysis, as he did in a 2009 study of the Alzheimer disease genetic risk factor ApoE (Roses et al., 2009), to better delineate the specific gene sequence that leads to disease. For doctors and people at risk for FTLD, he noted, “You do not want something that is associated—you want to know what the genetic lesion is.”
A Panoply of PGRN Problems
One frontotemporal dementia is not like another. “FTD has varied clinical phenotypes, and there are several different pathological phenotypes,” Van Deerlin said. She and coauthors at the University of Pennsylvania and the University of Washington in Seattle set out to document the spectrum of PGRN-associated FTLDs with another large-scale study, published in the Archives of Neurology. Joint first authors were Chang-En Yu and Thomas Bird, with Van Deerlin and Gerald Schellenberg, formerly in Washington and now at the University of Pennsylvania, as senior authors.
The researchers collected data from eight centers on 434 people with FTD and 111 controls with TDP-43 pathology but no FTD. The FTD cases included many forms of the disease, such as FTLD with ubiquitin-positive inclusions (FTLD-U), FTD with motor neuron disease, corticobasal degeneration, and Pick disease. The controls included subjects with amyotrophic lateral sclerosis (ALS), Alzheimer disease, and parkinsonism, among other conditions. The scientists sequenced the PGRN gene from each sample and used that data and in vitro splicing assays to predict the pathogenic potential of the variants they discovered.
PGRN mutations were limited to cases with symptoms of FTD or corticobasal syndrome and proteinopathy with ubiquitin-tagged deposits. PGRN mutations were not present in other conditions such as AD and ALS. However, some diseases were not fully represented in the sample, noted Marc Cruts of VIB-University of Antwerp, who was not part of the study team. For example, he wrote in an e-mail to ARF, some people with an AD diagnosis have been found to have PGRN mutations (Brouwers et al., 2007). Therefore, Cruts wrote, these diseases cannot be definitively excluded from the PGRN spectrum.
The sequencing revealed 58 PGRN variants, 26 not previously published. The researchers discounted variants in the 3’-UTR and in controls as unlikely to be pathogenic. Twenty-two likely pathogenic mutations caused a premature stop codon in the sequence. Five of them altered splice sites, and likely caused exon skipping, leading to a premature stop codon later on. In vitro splicing assays confirmed this prediction. Sequence analysis indicated that most of these mutations would alter protein function.
Overall, PGRN mutations were found in 6.9 percent of all clinical FTD cases, 16 percent of familial cases, and 21.4 percent of cases with confirmed ubiquitin-tagged inclusions. Many of the mutations cause premature stop codons, and the corresponding mRNA is likely degraded before it can make a protein. Thus, people carrying such a mutation on one chromosome would only make half as much PGRN protein as they should. This haploinsufficiency appears to be the most common cause of PGRN-associated disease, but it is not yet known how reduced PGRN leads to TDP-43 proteinopathy, which is found in all FTD cases with PGRN mutations, or disease symptoms. Perhaps, Van Deerlin and Lee speculated, TMEM106B will turn out to be the “missing link” between PGRN and TDP-43.
“I think we have gone a long distance to nailing down what may be the nearly complete array of mutations,” said Trojanowski, who also participated in this study. They likely did not hit all possible PGRN mutations, though. In fact, in the same issue of the Archives of Neurology, another group at the Mayo Clinic in Rochester, Minnesota, describes yet another PGRN mutation with rather unusual clinical features. This work was led by first author Brendan Kelley, now at the University of Cincinnati in Ohio, and joint senior authors Boeve and Ronald Petersen at the Mayo Clinic.
Kelley and colleagues studied 10 affected members of the same local family with a familial neurodegenerative disorder linked to progranulin. The disorder looked a lot like Alzheimer disease. In fact, some members of the family lived with an AD diagnosis until they died. The first generation was not seen at the Mayo Clinic, but when Petersen and colleagues saw the second generation, AD seemed like the best conclusion. The patients had forgetfulness and mild cognitive impairment. But the diagnosis began to make less sense when members of the third generation started seeing Boeve and other doctors at the Mayo Clinic. “Generation three presented at an earlier age and they had more variable symptoms,” Boeve said. Memory loss, as in AD, was not reliably an early symptom, but nor were the personality changes and executive dysfunction characteristic of FTD. One person had aphasia. Some family members started out with an AD diagnosis that was later changed to FTD. One even had both AD and FTD pathology, although the FTD was considered to be the primary problem.
Five of the second- and third-generation family members underwent MRI, and several were examined at autopsy. The pathology was suspicious; some subjects evinced degeneration and ubiquitin inclusions in the frontotemporal lobe. But there was no genetic explanation until 2006. “Once the progranulin mutation was found, then the light bulb went on,” Boeve said (see ARF related news story on Cruts et al., 2006 and Baker et al., 2006). Of those family members able to offer a DNA sample, all had the same single base pair deletion, causing a frameshift mutation and likely a nonfunctional gene, as in the other PGRN mutations described above.
Ultimately, scientists still do not understand how mutations in the same gene can cause such a panoply of presentations, making it difficult to predict disease course even if genetic information is available. “These two papers are complementary because the point of both is that neuropathology is important,” Bird said. Without pathology data, he said, it is much harder to predict what the genetics of a person with FTD will be. Genetics do not make much difference to treatment options at this point. But understanding the underlying genetic cause makes a big difference to family members concerned about their own level of risk. “One has to consider a PGRN mutation, even in late-onset Alzheimer disease with a strong family history,” Boeve said. “How many late-onset Alzheimer’s families are not having progranulin tested, and would additional insights be provided into pathogenesis?”
And Don’t Forget FUS
FTD and ALS could be viewed as opposite ends of a TDP-43 proteinopathy spectrum, with FTD affecting cortical neurons and ALS affecting motor neurons, and they can co-occur. Recently, FUS was linked to these diseases (see ARF related news story on Kwiatkowski et al., 2009 and Vance et al., 2009; and ARF related news story and Neumann et al., 2009). A group of researchers at the University Medical Center Utrecht in the Netherlands analyzed FUS variants in 52 Dutch people with familial ALS and 970 healthy controls. This report, led by joint first authors Ewout Grown and Michael van Es and senior author Leonard van den Berg, is also printed in February’s Archives of Neurology.
The researchers identified three mutations, including a novel one, in FUS. However, some control cases also had FUS variants. “Caution is warranted when interpreting results in a clinical setting,” the authors write. Doctors and genetic counselors should be aware, they write, that not all FUS variants cause disease and that some have incomplete penetrance.—Amber Dance.
Van Deerlin VM, Sleiman PM, Martinez-Lage M, Chen-Plotkin A, Wang LS, Graff-Radford NR, Dickson DW, Rademakers R, Boeve BF, Grossman M, Arnold SE, Mann DM, Pickering-Brown SM, Seelaar H, Heutink P, van Swieten JC, Murrell JR, Ghetti B, Spina S, Grafman J, Hodges J, Spillantini MG, Gilman S, Lieberman AP, Kaye JA, Woltjer RL, Bigio EH, Mesulam M, Al-Sarraj S, Troakes C, Rosenberg RN, White CL 3rd, Ferrer I, Lladó A, Neumann M, Kretzschmar HA, Hulette CM, Welsh-Bohmer KA, Miller BL, Alzualde A, de Munain AL, McKee AC, Gearing M, Levey AI, Lah JJ, Hardy J, Rohrer JD, Lashley T, Mackenzie IR, Feldman HH, Hamilton RL, Dekosky ST, van der Zee J, Kumar-Singh S, Van Broeckhoven C, Mayeux R, Vonsattel JP, Troncoso JC, Kril JJ, Kwok JB, Halliday GM, Bird TD, Ince PG, Shaw PJ, Cairns NJ, Morris JC, McLean CA, Decarli C, Ellis WG, Freeman SH, Frosch MP, Growdon JH, Perl DP, Sano M, Bennett DA, Schneider JA, Beach TG, Reiman EM, Woodruff BK, Cummings J, Vinters HV, Miller CA, Chui HC, Alafuzoff I, Hartikainen P, Seilhean D, Galasko D, Masliah E, Cotman CW, Tuñón MT, Martínez MC, Munoz DG, Carroll SL, Marson D, Riederer PF, Bogdanovic N, Schellenberg GD, Hakonarson H, Trojanowski JQ, Lee VM. Common variants at 7p21 are associated with frontotemporal lobar degeneration with TDP-43 inclusions. Nat Genet. 2010. Epub ahead of print 14 Feb. Abstract
Yu CE, Bird TD, Bekris LM, Montine TJ, Leverenz JB, Steinbart E, Galloway NM, Feldman H, Woltjer R, Miller CA, Wood EM, Grossman M, McCluskey L, Clark CM, Neumann M, Danek A, Gelasko DR, Arnold SE, Chen-Plotkin A, Karydas A, Miller BL, Trojanowski JQ, Lee VM, Schellenberg GD, Van Deerlin VM. The spectrum of mutations in progranulin: a collaborative study screeing 545 cases of neurodegeneration. Arch Neurol. 2010 Feb;67(2):161-70. Abstract
Boeve BF. Progress on progranulin. Arch Neurol. 2010 Feb;67(2):145-7. Abstract
Kelley BJ, Haldar W, Boeve BF, Baker M, Shiung M, Knopman DS, Rademakers R, Hutton M, Adamson J, Kuntz KM, Dickson DW, Parisi JE, Smith GE, Petersen RC. Alzheimer disease-like phenotype associated with the c.154delA mutation in progranulin. Arch Neurol. 2010 Feb;67(2):171-7. Abstract
Groen EJ, van Es MA, van Vught PW, Spliet WG, van Engelen-Lee J, de Visser M, Wokke JH, Schelhaas JH, Ophoff RA, Fumoto K, Pasterkamp RJ, Dooijes D, Cuppen E, Veldink JH, van den Berg LH. FUS mutations in familial amyotrophic lateral sclerosis in the Netherlands. Arch Neurol. 2010 Feb;67(2):224-30. Abstract