23 September 2011. Meet C9ORF72. Researchers have found that über-long repeat expansions in this uncharacterized gene cause amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD). The finding genetically links the two conditions, provides researchers with a fresh pathway to untangle, and solves one of the mysteries in the field. Two independent teams from the Mayo Clinic in Jacksonville, Florida, and the National Institute on Aging (NIA) in Bethesda, Maryland, report their results online in Neuron September 21. The authors discovered that a non-coding hexamer, GGGGCC, repeats up to 1,600 times near the beginning of the C9ORF72 gene in ALS cases. People with no hint of the disease have between two and 23 of the repeats. This expansion explains many cases of ALS and FTD, and a spectrum of conditions that mix symptoms of both. The finding is “undoubtedly one of the most singular genetic discoveries in the field,” wrote Simon Cronin of Beaumont Hospital in Dublin, Ireland, in an e-mail to ARF. Cronin was not involved in either study.
This discovery solves a five-year-long mystery in ALS and FTD genetics. Numerous linkage analyses (e.g., ARF related news story on Gijeselinck et al., 2010; see also Related Papers) and genomewide association studies (e.g., ARF related news story on Laaksovirta et al., 2010 and Shatunov et al., 2010; see also Related Papers) showed that the p arm of the ninth chromosome harbored a common genetic factor related to familial and sporadic ALS and FTD—but no one could figure out what the variant was. Standard sequencing approaches failed.
The identification of the hexamer repeat explains why. GC-rich repeats present challenges because the nucleotides bind opposing strands tightly and curl into secondary structures that preclude the usual sequencing and polymerase chain reaction (PCR) techniques. In addition, these repeats were hidden in an intron—not the first place one would look for a disease-causing mutation. The very nature of the variant stymied attempts to identify it. “When you consider that we knew [the location] within 32,000 base pairs—which is like a street in a city—it is amazing that it took us so long to find it,” said Bryan Traynor, co-senior author of the NIA study. Co-first authors Alan Renton, Elisa Majounie, Raphael Gibbs, and Jennifer Schymick also work at the NIA. “The other problem is that the expansion is so enormous,” added Ian Mackenzie, who works at the University of British Columbia in Vancouver, Canada. Mackenzie joined senior author Rosa Rademakers and first author Mariely DeJesus-Hernandez of the Mayo Clinic on their paper.
According to the papers, the repeat expansion accounts for 24-46 percent of familial ALS cases and 12-29 percent of inherited FTD instances in European populations. That makes it the most influential gene in ALS—dethroning superoxide dismutase 1, which is mutated in 10-20 percent of familial cases (Rosen et al., 1993)—and puts it on par with progranulin mutations, which are found in 25 percent of inherited FTD cases (see ARF related news story on Cruts et al., 2006 and Baker et al., 2006). Scientists have long thought the two diseases, which coincide in some individuals, represent opposite ends of a continuous spectrum (Geser et al., 2010), and this genetic evidence clinches it, Mackenzie said.
Rademakers and colleagues previously described a family they named VSM-20, for collaborators in Vancouver, University of California San Francisco, and the Mayo Clinic in Rochester, Minnesota (Boxer et al., 2011). Affected members of the kindred exhibit symptoms of ALS or FTD and inherited disease in an autosomal-dominant manner. The team had already sequenced exon and exon/intron boundaries across the chromosome 9p region from this family to no avail.
The researchers turned to the non-coding areas, where they discovered the GGGGCC hexanucleotide repeated three to 23 times between C9ORF72 exons 1a and 1b (alternative splicing determines which exon is translated) in healthy people. To examine the expansion further, they designed PCR primers to amplify the region for restriction fragment-length analysis, but something was decidedly odd about their results. As expected, non-affected family members exhibited two different alleles—for example, one with three repeats and one with nine. But every person affected by disease appeared to be homozygous, with only one repeat length present. Moreover, that repeat length always seemed to match an allele from the unaffected parent—as if nothing was inherited from the affected parent.
The key here is that affected people only appeared homozygous—the researchers knew that could not be correct. “The discovery in Rosa Rademakers’ laboratory really hinged on following up on apparently non-Mendelian inheritance…something that many people would dismiss as lab error,” noted Ammar Al-Chalabi of Kings College London, U.K., in an e-mail to ARF (see full comment below). Al-Chalabi did not participate in either Neuron report.
Rademakers' team realized they must have been missing out on the disease-linked allele entirely. They concluded that the missing variant was not showing up because it was not amenable to PCR—it must be either a deletion or some sort of amplification-resistant sequence. Thinking the secondary structure of the repeats might be interfering with the amplification, the researchers designed a PCR protocol around a primer that bound within the repeat sequence. Since the hexamer-binding primer could attach anywhere within the repeats, the technique did not amplify a single, full-length expansion, but a range of fragments up to full length. From this swath of amplified DNA, it was clear that people with ALS or FTD possessed a vastly expanded repeat section. Based on a handful of Southern blots for C9ORF72 in lymphoblasts and cortical samples, the authors estimated that 700-1,600 repeats were present in family members with ALS or FTD.
It was a rather curious lab result that clued in Traynor and colleagues to the repeats. They focused on one Welsh and one Dutch ALS-FTD family (Pearson et al., 2011; Mok et al., 2011). Collaborators included co-senior authors Nigel Williams of Cardiff University and Stuart Pickering-Brown of the nearby Royal Gwent Hospital in Wales, U.K. and co-first author Adrian Waite, also in Cardiff. Sara Rollinson at the University of Manchester, U.K., was also a co-first author. In The Netherlands, Peter Heutink of VU University in Amsterdam was a co-senior author, and Javier Simón-Sánchez and John van Swieten of the University Medical Center Rotterdam were co-first authors.
When plain old sequencing failed to identify anything unusual in the affected family members, “we realized, we have to get a little bit extreme,” Traynor said. The team collected DNA samples from a few members of FTD-ALS kindreds and isolated the ninth chromosome’s DNA so they would not have to search the entire genome. They sequenced deep, obtaining 300 independent sequences of the entire chromosome except for an odd p-arm region where the coverage, according to computer analysis, dropped to just twofold. Looking at the sequences themselves, the researchers concluded that the computer had misaligned the base pairs. Manually matching the nucleotides, they realized that the sticking point was a whole slew of GGGGCC repeats in C9ORF72. “It really was a 'eureka' moment,” Traynor recalled. The group also designed a PCR assay to identify the repeats.
As of yet, neither group has come across a person with more than 23 yet fewer than 700 C9ORF72 repeats. That mid-range might cause disease or be a risk factor, Rademakers speculated.
What is C9ORF72? Nobody knows. It possesses no common, identifiable protein domains to hint at its function. It is fairly well conserved across species, though, so researchers think it must be important. The gene is expressed in the brain as well as other tissues. At this point, researchers have no idea how the C9ORF72 expansion can cause ALS, FTD, or a combination. The manifestation could depend on genetic or environmental influence, random chance, or some combination of those factors, Traynor said.
The repeat expansion is likely to be an unstable region, Traynor said. Sequences packed with cytosines and guanines are difficult for DNA polymerase to cross, Traynor explained, so a few repeats could easily expand. Once a couple of hundred or so of the repeats are present, they could fold up, making it easy for a duplication to occur during DNA replication, he added. Folding of repeat expansions has been shown to compromise replication of polyglutamine-expanded huntingtin genes previously.
The discovery fits current trends in neurodegeneration research, noted Michael van Es of the University Medical Center Utrecht in The Netherlands in an e-mail to ARF. Van Es was not involved in the current papers. He noted that the lines between Mendelian, sporadic, and complex-cause disorders are becoming increasingly blurred. It is also becoming harder to delineate distinct diseases. C9ORF72 is one in a growing set of genes, van Es wrote, linked to one or more conditions, for example, TAR DNA binding protein 43 (TDP-43) in ALS, FTD, and PD; ataxin 2 in ALS, spinocerebellar ataxia, and PD (see ARF related news story on Elden et al., 2010); and tau in FTD and PD.
Expansions in coding DNA, such as polyglutamines in Huntington’s and other neurodegenerative diseases, are well known. Rademakers said that finding the repeats felt like harking back to the beginning of neurogenetics, where polyQ repeats in Huntingtin provided the first big break in the field (see Snell et al., 1993). Van Es noted that the current work provides further evidence that examining only the genome’s coding sequences is a form of tunnel vision, with at least some disease variants occurring elsewhere. Noncoding repeats have also been found in myotonic dystrophy (Mahadevan et al., 1992), fragile-X-associated tremor/ataxia syndrome (Tassone et al., 2004), and spinocerebellar ataxia (Kobayashi et al., 2011; Moseley et al., 2006; Sato et al., 2009).
People with the C9ORF72 GGGGCC expansions exhibit TDP-43 inclusions, a common feature in many kinds of ALS and FTD. There are two likely ways in which these intron repeats could trigger disease. For one, the expansions might affect C9ORF72 directly. Instability of the mRNA might prevent C9ORF72 translation, leading to a loss of function, suggested Guy Rouleau of the University of Montréal, who did not participate in the studies. Alternatively, the GC-rich mRNA could cause trouble by recruiting and sequestering RNA binding proteins and other RNAs into aggregates that prevent proper splicing of RNAs beyond just C9ORF72, as proposed for fragile X and spinocerebellar ataxia (Galloway and Nelson, 2009; Daughters et al., 2009; reviewed in Todd and Paulson, 2010). In that case, the repeat-rich mRNA—not the protein—would be responsible for disease. Given the expansion size, “the number of proteins that are bound to this are going to be astronomical, potentially,” said Stanley Appel of The Methodist Hospital Research Institute in Houston, Texas, who was not involved in either study.
Supporting the toxic RNA hypothesis, the Mayo team observed nuclear RNA aggregates containing C9ORF72 mRNA in spinal cord and frontal cortex postmortem sections from people who had the expansion. If the RNA is indeed causing problems, Appel said, then it suggests a potential future therapeutic: antisense oligonucleotides to block or destroy the repeat sequences. This kind of therapy is already in the works for myotonic dystrophy (Furling et al., 2003).
John Fink of the University of Michigan in Ann Arbor, who was not part of either study team, said he was initially skeptical that the C9ORF72 expansion could explain all of chromosome 9p-linked neurodegeneration. But when he saw how many sporadic cases carried the repeats—more than 20 percent in the NIA study—he was sold, he told ARF.
Traynor’s team focused on a Finnish population because the country has unusually high rates of ALS (Murros and Fugelholm, 1983). His collaborators in that region included co-senior author Pentti Tienari and co-first author Hannu Laaksovirta at the University of Helsinki. There, the expansion accounted for 46 percent of the familial disease and 21 percent of sporadic cases in a cohort of 402 people with ALS. Along with a SOD1 D90A mutation common in Scandinavia, the C9ORF72 variant explains the majority of ALS in Finland, Traynor said. The team also examined 75 Finns with FTD, among whom 29 percent carried the C9ORF72 expansion.
Traynor did not expect that the long expansion would affect many people beyond Northern Europe. But when the team screened 238 people with inherited ALS from North America, Germany, and Italy, “my jaw was dropping because literally every second sample had this kind of expansion,” he recalled. “I never thought I would see the day when one mutation would explain so many of familial ALS cases.”
Similarly, in a series of 696 North Americans with FTD and/or ALS, the Mayo group determined that the repeat expansion was responsible for 24 percent of familial ALS and 4 percent of the sporadic disease. SOD1 mutations, in comparison, only accounted for 12 percent of the familial and none of the sporadic cases in their analysis. In the case of FTD, the expansion was linked to 12 percent of familial and 3 percent of sporadic cases. For comparison, progranulin mutations caused 8 percent of the familial and 3 percent of the sporadic FTD cases.
With C9ORF72 such a black box, there are plenty of questions to ask and experiments to perform about its biology and pathological role. “We have a whole new area of research that we can focus on now,” Rademakers said. “It is very likely that this discovery will eventually help identify new targets for therapies.”
Moreover, “it is also going to be important for clinical testing because it is present at such a high frequency,” said Vivianna Van Deerlin of the University of Pennsylvania in Philadelphia. “I am going to be working as quickly as possible to develop a clinical test in my lab.” Simply knowing the root of disease can be a great relief to a family (see BBC News story on the Welsh family in the NIA study).—Amber Dance.
Dejesus-Hernandez M, Mackenzie IR, Boeve B, Boxer AL, Baker M, Rutherford NJ, Nicholson AM, Finch NA, Gilmer HF, Adamson J, Kouri N, Wojtas A, Sengdy P, Hsiun GY, Karydas A, Seeley WW, Josephs KA, Coppola G, Geschwind DH, Wszolek ZK, Feldman H, Knopman D, Petersen R, Miller BL, Dickson D, Boylan D, Graff-Radford N, Rademakers R. Expanded GGGGCC hexanucleotide repeat in non-coding region of C9ORF72 causes chromosome 9p-linked frontotemporal demential and amyotrophic lateral sclerosis. Neuron. 2011 Sep 21. Abstract
Renton AE, Majounie E, Waite A, Simón-Sánchez J, Rollinson S, Gibbs JR, Schymick JC, Laaksovirta H, van Swieten JC, Myllykangas L, Kalimo H, Paetau A, Abramzon Y, Remes AM, Kaganovich A, Scholz SW, Duckworth J, Ding J, Harmer DW, Hernandez DG, Johnson JO, Mok K, Ryten M, Trabzuni D, Guerrerio RT, Orrell RW, Neal J, Murray A, Pearson J, Jansen IE, Sondervan D, Seelaar H, Blake D, Young K, Halliwell N, CAllister J, Toulson G, Richardson A, Gerhard A, Snowden J, Mann D, Neary D, Nalls MA, Peuralinna T, Jansson L, Isoviita VM, Kaivorinne AL, Hölttä-Vuori M, Ikonen E, Sulkava R, Benatar M, Wuu J, Chìo A, Restagno G, Borghero G, Sabatelli M, The ITALSGEN Consortium, Heckerman D, Roaeva E, Zinman L, Rothstein J, Sendtner M, Drepper C, Eichler EE, Alkan C, Abdullaev Z, Pack SD, Dutra A, Pak E, Hardy J, Singleton A, Williams NM, Heutink P, Pickering-Brown S, Morris HR, Tienari PF, Traynor BJ. A hexanucleotide repeat expansion in C9ORF72 is the cause of chromosome 9p21-linked ALS-FTD. Neuron. 2011 Sep 21. Abstract