Scientists are hearing the pitter-patter of C9ORF72 mouse feet, a potential new model for familial forms of amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD). In the May 14 Science, Leonard Petrucelli and colleagues at the Mayo Clinic in Jacksonville, Florida, describe a mouse that expresses 66 of the GGGGCC C9ORF72 repeats associated with disease. The animals exhibit a triad of key pathological features: dense foci containing transcribed repeats, aggregates of dipeptides translated from those RNAs, and phosphorylated TDP-43 in cytoplasmic inclusions. Moreover, the mice are hyperactive and antisocial, behavior reminiscent of FTD.

All the right markers.

C9ORF72 mice accumulated puncta of repeat RNA (red, top), globs of dipeptide repeats (red, bottom), and cytoplasmic TDP-43 inclusions (green). [Image courtesy of Science/AAAS.]

“This is a very strong start toward a complete model,” commented Brian Kraemer of the University of Washington in Seattle, who was not involved in the study. “The mice seem to have the major pathology in spades.”

Hundreds or thousands of repeats, located in the first intron of the C9ORF72 gene, are the most common genetic cause of ALS and FTD. The repeat RNAs accumulate and are translated into repetitive dipeptides, which also aggregate (see Jan 2013 news; Feb 2013 newsnews).

Petrucelli and first author Jeannie Chew used an adeno-associated virus to deliver the repeats into the brain of wild-type mice. Chew packed the virus heads with a DNA construct containing 66 copies of the repeat, not the full C9ORF72 gene itself. She drove expression of the repeat DNA using a β-actin promoter. As a control, she made another version with just two repeats, a number found in healthy people. She injected the viruses into the brain ventricles of mouse pups just after they were born. The procedure did not disrupt the pups’ overall development, and reliably achieved broad transgene expression across neurons of the brain and spinal cord, Petrucelli said.

Chew and colleagues reported their analysis of the infected mice at 6 months of age. Fluorescent in situ hybridization revealed GGGGCC foci throughout the brain and spinal cord. Globular inclusions of the translated dipeptides were also present.

Inclusions containing phosphorylated TDP-43 are another hallmark of human C9ORF72 disease. The mice displayed these aggregates, too, but only in about 8 percent of the cells in the cortex and hippocampus. That might not sound like a large proportion, but scientists who spoke with Alzforum said it was reasonable. “I do not know that there is any mouse model out there that has this much phosphorylated TDP-43—even the TDP-43 [transgenic] mice,” Kraemer commented. The mice confirm that the C9ORF72 repeats are sufficient to induce phosphorylation and aggregation of endogenous TDP-43, Petrucelli said, and suggest that anti-C9ORF72 treatments might mitigate TDP-43 proteinopathy in people who carry the repeat gene.

Compared with their counterparts with just two repeats, the 66-repeat mice possessed 17 percent fewer neurons in the cortex, and 11 percent fewer Purkinje cells in the cerebellum. The neurons of the hippocampus, thalamus, and spinal cord were unaffected at this age.

Next, Chew tested the mice for FTD-like behavior. When placed in an open area, the animals seemed anxious, hugging the dim walls rather than exploring the well-lit center field, as did two-repeat mice. However, they traveled faster and further overall, suggesting they were hyperactive. The authors note that hyperactivity can follow behavioral disinhibition in patients with dementia. The mice were also antisocial. When given the choice between interacting with an unknown mouse and an empty chamber, they were more likely to hang out alone. “The FTD readouts are exquisite, a real jackpot,” said Cat Lutz of the Jackson Laboratory in Bar Harbor, Maine.

What about ALS-like symptoms? Chew tested balance and grip on a rotating rod, a typical test for motor-neuron deficiency. On the first day, the two-repeat and 66-repeat mice both fell off after about two minutes. The two-repeat mice rapidly got the hang of it, balancing for about four minutes by the fourth day of testing. The 66-repeat mice, in contrast, hardly improved. Their disability could be due to motor problems, learning problems, or both, Petrucelli said. Perhaps the animals will develop further characteristics of ALS as they age, Lutz speculated.

Several laboratories are wrestling the C9ORF72 repeats into mice, mostly via a repeat-heavy transgene cloned onto a bacterial artificial chromosome, said Lutz (see Nov 2013 news). Lutz and collaborator Robert Baloh of the Cedars-Sinai Medical Center in Los Angeles are working on this, but were not involved in Petrucelli’s project. Most scientists in the race have managed to obtain mice with the RNA foci and dipeptide inclusions, said Lutz, but these animals have little evidence of neurodegeneration or disease symptoms. Lutz told Alzforum her C9ORF72 mice have lived to 15 months so far with no sign of behavioral or movement problems.

Lutz suggested that Petrucelli and Chew achieved stronger phenotypes than other laboratories because their approach was more severe than the single-copy transgene many other scientists have tried. In addition, mice engineered with germline transgenes possess the extra repeats from conception, and that may give them time to adapt, she theorized. In contrast, Chew’s mice received the transgene at birth, under a strong promoter. “You are hitting it hard, and you are hitting it fast,” she said.

Experts who spoke with Alzforum agreed that Petrucelli’s mice have potential as preclinical models of human disease. Lutz noted that many mouse labs are already equipped for AAV injection, and could quickly obtain large groups of C9ORF72 mice by starting with wild-type pups.

“The fantastic thing about this model is we see everything—it has the foci, it has the dipeptide repeats, it has the TDP-43 inclusions,” commented Johnathan Cooper-Knock of the University of Sheffield, U.K. Even so, he raised some caveats about how well the new model parallels the human condition. Chew’s mice make only 66 hexanucleotide repeats, compared with the hundreds or thousands often found in patients. In addition, they only make RNAs and dipeptides from the sense direction, but C9ORF72 repeats are known to be transcribed and translated in the antisense direction, as well (see Nov 2013 news). Some recent studies finger the poly-proline-arginine dipeptides made by antisense translation as potentially the most toxic species, Cooper-Knock pointed out, in which case this model would be missing a key feature (see Dec 2014 news). Only as researchers flesh out the details of human C9ORF72 disease will they be able to rank Chew’s mice, and other models, on how well they match patient physiology, Cooper-Knock said.

Petrucelli intends to address the antisense question by packing an antisense construct into the viral vector, he told Alzforum. He also plans to investigate different lengths of repeat, as well as different ages.—Amber Dance


  1. This is an exciting and potentially very important new model system for C9 forms of ALS and FTLD. The paper underscores the utility of somatic cell gene transfer with AAV vectors for disease modeling. The vector approach is rapid and therefore expedites hypothesis testing.

  2. This paper by Chew et al. describes a mouse model that recapitulates the neuropathological and clinical phenotypes of the G4C2 hexanucleotide repeat expansion in C9ORF72, the major genetic cause of FTD and ALS. The mice were generated by intracerebroventricular AAV-injection of two or 66 G4C2 repeats at postnatal day zero. The neuropathological and behavioral phenotypes were examined at six months post-injection.

    Similarly to the C9FTD/ALS patients, the (G4C2)66 mice showed the presence of nuclear RNA foci and ubiquitin-positive C9RAN inclusions, as well as expression of poly(GP) dipeptides. The mice also developed phospho-TDP43-containing inclusions, but these, even when they resided in the same cells as poly(GA) inclusions, did not contain poly(GA). These findings are in accordance with previous data from human patients, which have indicated that the dipeptide repeat (DPR)-containing inclusions are separate from those harboring phospho-TDP-43 (Mori et al., 2013). Here, Chew et al. also present an intriguing idea that the extended G4C2 repeats might function as initiators of TDP-43 pathology. On the other hand, because TDP-43 inclusions are present also in ALS or FTD patients not carrying the G4C2 repeat expansion, the extended repeats may not represent the sole trigger of the TDP-43 pathology. Nevertheless, in G4C2 expansion cases, DPRs might function as initiators or modulators and thus result in the development of enhanced TDP-43 pathology. Characterization of the potential underlying mechanism(s) and further testing of this interesting idea by using other approaches, for example, antisense oligonucleotides to block the generation of the extended repeats in these mice, will provide more information on this plausible relationship.

    In addition to the typical neuropathological changes, behavioral changes were detected in (G4C2)66 mice compared to (G4C2)2 mice. (G4C2)66 mice exhibited hyperactivity and anxiety-like and asocial behaviors, which are typical symptoms in patients with the C9ORF72-associated behavioral variant FTD. Also, motor skill deficits were noticed in (G4C2)66 mice, but a decrease in body weight was detected only in females. Thus, it seems that the (G4C2)66 mice exhibit more of the clinical features of behavioral-variant FTD than of ALS.

    Taken together, the (G4C2)66 mice represent the first mammalian model widely recapitulating both neuropathological and behavioral alterations associated with the C9ORF72 G4C2-repeat expansion in human FTD and ALS patients. The model provides excellent possibilities for further characterizations and understanding of the mechanisms of FTD/ALS pathogenesis as well as for testing emerging therapeutic approaches. 


    . The C9orf72 GGGGCC repeat is translated into aggregating dipeptide-repeat proteins in FTLD/ALS. Science. 2013 Mar 15;339(6125):1335-8. Epub 2013 Feb 7 PubMed.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Chicago—RNA Inclusions Offer Therapeutic Target in ALS
  2. RNA Twist: C9ORF72 Intron Expansion Makes Aggregating Protein
  3. Second Study Sees Intron in FTLD Gene Translated
  4. Researchers Revel in C9ORF72 Advances at RNA Symposium
  5. Sense, Antisense: C9ORF72 Makes Both Forms of RNA, Peptides
  6. Live-Cell Studies Blame Arginine Peptides for C9ORF72’s Crimes

Further Reading


  1. . Antisense RNA foci in the motor neurons of C9ORF72-ALS patients are associated with TDP-43 proteinopathy. Acta Neuropathol. 2015 Jul;130(1):63-75. Epub 2015 May 6 PubMed.
  2. . The Spectrum of C9orf72-mediated Neurodegeneration and Amyotrophic Lateral Sclerosis. Neurotherapeutics. 2015 Apr;12(2):326-39. PubMed.
  3. . Characterization of the dipeptide repeat protein in the molecular pathogenesis of c9FTD/ALS. Hum Mol Genet. 2015 Mar 15;24(6):1630-45. Epub 2014 Nov 14 PubMed.
  4. . C9orf72 expansions in frontotemporal dementia and amyotrophic lateral sclerosis. Lancet Neurol. 2015 Mar;14(3):291-301. Epub 2015 Jan 29 PubMed.

Primary Papers

  1. . Neurodegeneration. C9ORF72 repeat expansions in mice cause TDP-43 pathology, neuronal loss, and behavioral deficits. Science. 2015 Jun 5;348(6239):1151-4. Epub 2015 May 14 PubMed.