EDITORIAL DNA slows effects of C9orf72 mutations An epigenetic brake on genetic inheritance

Jeremy J. Day, PhD Hexanucleotide expansion mutations in the C9orf72 verbal recall decline on neuropsychological testing.6 Erik D. Roberson, MD, gene are the most common autosomal dominant cause These findings provide evidence that the DNA meth- PhD of amyotrophic lateral sclerosis (ALS) and frontotem- ylation of C9orf72 is protective, not detrimental, in poral dementia (FTD). It remains unclear how patients with ALS or FTD due to C9orf72 C9orf72 expansions cause disease, with data to support expansions. Correspondence to 3 different hypotheses1: (1) loss of function due to The study is consistent with other recent findings Dr. Roberson: [email protected] decreased C9orf72 expression, (2) gain of function from the same group that higher levels of C9orf72 due to RNA foci that sequester critical proteins, and DNA methylation are associated with fewer RNA foci ® 2015;84:1616–1617 (3) gain of function due to expression of dipeptide and dipeptide repeat aggregates,3 plus longer disease repeat proteins encoded by the expansion. course and later death.4 Together, these findings indi- The loss-of-function hypothesis is based on ob- cate that epigenetic mechanisms counteract servations of reduced C9orf72 levels in expansion at C9orf72, with DNA methylation suppressing ef- carriers, mediated at least in part through epige- fects of the repeat expansion. Combined with reports netic mechanisms, including increases in DNA that reducing C9orf72 in mice using antisense oligo- methylation within the C9orf72 .2–5 For nucleotides does not have apparent toxicity7 and most genes, DNA methylation reduces gene expres- other similar findings, these data may spell the begin- sion, and this is true for C9orf72.2,3 While ning of the end for the loss-of-function hypothesis of increased DNA methylation is observed in C9orf72 pathogenesis. C9orf72 carriers, it has been unclear whether it is While the study of McMillan et al.6 has several protective (by reducing gain-of-function effects) or strengths, including the combination of neuroimag- detrimental (by causing loss of function). In this ing and neuropathologic analyses and a longitudinal issue of Neurology®, McMillan et al.6 present data component, it also has limitations. This is a rela- suggesting a protective effect of C9orf72 DNA tively small study, and the participants were an methylation. admixture of different clinical phenotypes, including The authors studied 20 individuals with C9orf72 ALS, behavioral variant FTD, FTD-ALS, and the expansions. Prior studies showed that in controls and nonfluent/agrammatic variant of primary progress- patients with FTD without C9orf72 expansions, ive aphasia. While an exploratory analysis did not there is little DNA methylation in this region of find differences in methylation rates among clinical C9orf72.2,5 By contrast, in C9orf72 expansion car- subtypes, the sample sizes are too small to be defin- riers, the DNA methylation rate for this locus ranged itive, and it remains possible that the observed dif- up to almost 40%. McMillan et al.6 first examined the ferences in regional volumes and cell counts could be relationship between C9orf72 methylation and driven by differential segregation of certain clinical regional gray matter volumes by structural MRI. phenotypes toward higher (or lower) methylation Those with more methylation had less atrophy in rates. several regions. Next they examined the rela- Another concern regarding this study is that the tionship between DNA methylation and neuronal neuroimaging and longitudinal analyses were based loss. Greater C9orf72 methylation was associated with on analysis of C9orf72 methylation in blood, rather less neuron loss in both hippocampus and frontal than in brain tissue. However, for many genes, levels cortex.6 Finally, and perhaps most interestingly, the of DNA methylation in blood correlate with levels in authors examined longitudinal imaging and cognitive brain, and this holds true for C9orf72.4 The fact that data over about 1 year. Consistent with the cross- DNA methylation measurements from blood corre- sectional data, more C9orf72 methylation was associ- late with multiple disease measures poses an oppor- ated with slower rates of volume loss by MRI and tunity for an easily accessible biomarker.

See page 1622 From the Center for Neurodegeneration and Experimental Therapeutics (E.D.R.), Evelyn F. McKnight Brain Institute (J.J.D., E.D.R.), and Departments of Neurology (E.D.R.) and Neurobiology (J.J.D., E.D.R.), University of Alabama at Birmingham. Go to Neurology.org for full disclosures. Funding information and disclosures deemed relevant by the authors, if any, are provided at the end of the editorial.

1616 © 2015 American Academy of Neurology ª 2015 American Academy of Neurology. Unauthorized reproduction of this article is prohibited. How might come into play in clinical TALE (transcription activator-like effector) systems neurology? biology, where understanding to direct epigenetic modifiers to targeted sites in epigenetic dysregulation of tumor suppressor genes DNA, and are currently being applied to enable cus- and oncogenes has facilitated both biomarker dis- tom reorganization of epigenetic marks in animal covery and development of novel therapeutics, models.11 Although these approaches are at best many may provide a useful analogy. Epigenetic biomarkers years from clinical applications, such revolutionary have been identified in several malignancies and help advances in technology highlight the enticing poten- predict cancer incidence and survival rates in vulner- tial for epigenetics-based therapeutics in neurologic able populations.8 Likewise, epigenetics-based ther- disease. apeutics have been Food and Drug Administration– approved for at least 2 , refractory cutaneous STUDY FUNDING T-cell lymphoma and myelodysplastic syndrome. No targeted funding reported. These therapies prolong survival and improve qual- DISCLOSURE ity of life, and other promising new drugs are in the The authors report no disclosures relevant to the manuscript. Go to 9 pipeline. Neurology.org for full disclosures. Neuroepigenetics is a younger field and its clinical potential is less well-defined. However, several studies REFERENCES have linked specific epigenetic alterations to Alz- 1. Gendron TF, Belzil VV, Zhang YJ, Petrucelli L. Mecha- nisms of toxicity in C9FTLD/ALS. Acta Neuropathol heimer disease, Parkinson disease, ALS, and epilepsy. 2014;127:359–376. In cases where these differences can be detected in 2. Xi Z, Zinman L, Moreno D, et al. Hypermethylation of

blood, epigenetic biomarkers could be used to identify the CpG island near the G4C2 repeat in ALS with a vulnerable populations before disease onset or consid- C9orf72 expansion. Am J Hum Genet 2013;92:981–989. ered alongside other risk factors and incorporated into 3. Liu EY, Russ J, Wu K, et al. C9orf72 hypermethylation treatment decisions. Preclinically, considerable evi- protects against repeat expansion-associated pathology in ALS/FTD. Acta Neuropathol 2014;128:525–541. dence also suggests potential for epigenetics-based ther- 4. Russ J, Liu EY, Wu K, et al. Hypermethylation of repeat apeutics in neurodegenerative disorders, including expanded C9orf72 is a clinical and molecular disease mod- improved cognitive function following treatment with ifier. Acta Neuropathol 2015;129:39–52. drugs that modulate the epigenome.10 Although bio- 5. Belzil VV, Bauer PO, Gendron TF, Murray ME, marker applications for neuroepigenetics are probably Dickson D, Petrucelli L. Characterization of DNA hyper- closer to fruition than treatments are, discoveries of an methylation in the cerebellum of c9FTD/ALS patients. – epigenetic role in the etiology of these disorders present Brain Res 2014;1584:15 21. 6. McMillan CT, Russ J, Wood EM, et al. C9orf72 promoter a new avenue for therapeutics. hypermethylation is neuroprotective: neuroimaging and If C9orf72 hypermethylation in FTD and ALS is neuropathologic evidence. Neurology 2015;84:1622–1630. neuroprotective, an obvious therapeutic strategy for 7. Lagier-Tourenne C, Baughn M, Rigo F, et al. Targeted mutation carriers would be to boost DNA methyla- degradation of sense and antisense C9orf72 RNA foci as tion at C9orf72. Supplementation with the methyl therapy for ALS and frontotemporal degeneration. Proc – donor, S-adenosyl methionine, is an established but Natl Acad Sci USA 2013;110:E4530 E4539. 8. Anjum S, Fourkala EO, Zikan M, et al. A BRCA1-muta- nonspecific approach to increasing DNA methyla- tion associated DNA methylation signature in blood cells tion. However, it could have unwanted off-target ef- predicts sporadic breast cancer incidence and survival. fects. Such lack of genetic specificity demonstrates a Genome Med 2014;6:47. key pitfall of current epigenetically targeted treat- 9. Dawson MA, Kouzarides T. Cancer epigenetics: from ments, but a rapidly emerging set of epigenetic engi- mechanism to therapy. Cell 2012;150:12–27. neering techniques is beginning to allow the necessary 10. Gräff J, Tsai LH. The potential of HDAC inhibitors as cognitive enhancers. Annu Rev Pharmacol Toxicol 2013; genetic specificity, at least in experimental settings. 53:311–330. These tools utilize the inherent DNA sequence rec- 11. Konermann S, Brigham MD, Trevino AE, et al. Optical ognition capacity of bacterial CRISPR (clustered reg- control of mammalian endogenous transcription and epi- ularly interspaced short palindromic repeats) or genetic states. Nature 2013;500:472–476.

Neurology 84 April 21, 2015 1617 ª 2015 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.