Neurologic Outcomes in Friedreich Ataxia: Study of a Single-Site Cohort E415

Home , Basilar invagination, Cerebellar tonsil, Etiracetam, IGFALS, Lentiform nucleus, OPA3

Volume 6, Number 3, June 2020 Neurology.org/NG

A peer-reviewed clinical and translational neurology open access journal

ARTICLE Neurologic outcomes in Friedreich ataxia: Study of a single-site cohort e415

ARTICLE Prevalence of RFC1-mediated spinocerebellar ataxia in a North American ataxia cohort e440

ARTICLE Mutations in the m-AAA proteases AFG3L2 and SPG7 are causing isolated dominant optic atrophy e428

ARTICLE Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy revisited: Genotype-phenotype correlations of all published cases e434 Academy Officers Neurology® is a registered trademark of the American Academy of Neurology (registration valid in the United States). James C. Stevens, MD, FAAN, President Neurology® Genetics (eISSN 2376-7839) is an open access journal published Orly Avitzur, MD, MBA, FAAN, President Elect online for the American Academy of Neurology, 201 Chicago Avenue, Ann H. Tilton, MD, FAAN, Vice President Minneapolis, MN 55415, by Wolters Kluwer Health, Inc. at 14700 Citicorp Drive, Bldg. 3, Hagerstown, MD 21742. Business offices are located at Two Carlayne E. Jackson, MD, FAAN, Secretary Commerce Square, 2001 Market Street, Philadelphia, PA 19103. Production offices are located at 351 West Camden Street, Baltimore, MD 21201-2436. Janis M. Miyasaki, MD, MEd, FRCPC, FAAN, Treasurer © 2020 American Academy of Neurology. Ralph L. Sacco, MD, MS, FAAN, Past President Neurology® Genetics is an official journal of the American Academy of Neurology. Journal website: Neurology.org/ng, AAN website: AAN.com CEO, American Academy of Neurology Copyright and Permission Information: Please go to the journal website (www.neurology.org/ng) and click the Permissions tab for the relevant Mary E. Post, MBA, CAE article. Alternatively, send an email to [email protected]. Chief Executive Officer General information about permissions can be found here: https://shop.lww.com/ journal-permission. 20l Chicago Ave Disclaimer: Opinions expressed by the authors and advertisers are not Minneapolis, MN 55415 necessarily those of the American Academy of Neurology, its affiliates, or of the Publisher. The American Academy of Neurology, its affiliates, and the Tel: 612-928-6000 Publisher disclaim any liability to any party for the accuracy, completeness, efficacy, or availability of the material contained in this publication (including drug dosages) or for any damages arising out of the use Editorial Office or non-use of any of the material contained in this publication. Patricia K. Baskin, MS, Executive Editor Advertising Sales Representatives: Wolters Kluwer, 333 Seventh Avenue, Rachel A. Anderson, Administrative Assistant New York, NY 10001. Contacts: Eileen Henry, tel: 732-778-2261, fax: 973-215- 2485, [email protected] and in Europe: Craig Silver, tel: +44 Morgan S. Sorenson, Managing Editor 7855 062 550 or e-mail: [email protected]. fl Careers & Events: Monique McLaughlin, Wolters Kluwer, Two Commerce Neurology® Neuroimmunology & Neuroin ammation Square, 2001 Market Street, Philadelphia, PA 19103, tel: 215-521-8468, fax: 215- Neurology® Genetics 521-8801; [email protected]. Reprints: Meredith Edelman, Commercial Reprint Sales, Wolters Kluwer, Two Kathleen M. Pieper, Senior Managing Editor, Neurology® Commerce Square, 2001 Market Street, Philadelphia, PA 19103, tel: 215-356-2721; Karen Skaja, Senior Editorial Coordinator [email protected]; [email protected]. Skyler M. Kane, Editorial Coordinator Special projects: US & Canada: Alan Moore, Wolters Kluwer, Two Commerce Square, 2001 Market Street, Philadelphia, PA 19103, tel: Margaret A. Rei, Editorial Coordinator 215-521-8638, [email protected]. International: Andrew Wible, Senior Manager, Rights, Licensing, and Partnerships, Wolters Kluwer; Lee Ann Kleffman, Managing Editor, Neurology® Clinical Practice [email protected]. Andrea Rahkola, Production Editor Kristen Swendsrud, Production Coordinator Sharon L. Quimby, Digital Managing Editor Kaitlyn Aman Ramm, Digital Multimedia/Graphics Coordinator Justin Daugherty, Digital Multimedia/Podcast Coordinator Madeleine Sendek, MPH, Digital Marketing Communications Coordinator

Publisher Wolters Kluwer Baltimore, MD

Publishing Staff Sarah Carrera, MBA, Executive Publisher Jessica Heise, Production Team Leader, Neurology Journals Megen Miller, Production Editor Steve Rose, Editorial Assistant Stacy Drossner, Production Associate

Copyright ª 2020 American Academy of Neurology. Unauthorized reproduction of this article is prohibited. A peer-reviewed clinical and translational neurology open access journal Neurology.org/NG

Neurology® Genetics

Editor Stefan M. Pulst, MD, Dr med, FAAN Vision Neurology®: Genetics will be the premier peer- reviewed journal in the ﬁeld of neurogenetics. Deputy Editor Massimo Pandolfo, MD, FAAN Mission Neurology: Genetics will provide neurologists Associate Editors and clinical research scientists with Alexandra Durr, MD, PhD outstanding peer-reviewed articles, Suman Jayadev, MD editorials, and reviews to elucidate the role Margherita Milone, MD, PhD of genetic and epigenetic variations in Raymond P. Roos, MD, FAAN diseases and biological traits of the central Editorial Board and peripheral nervous systems. Hilary Coon, PhD Giovanni Coppola, MD ChantalDepondt, MD, PhD Editorial Tel: 612-928-6400 Brent L. Fogel, MD, PhD, FAAN Inquiries Toll-free: 800-957-3182 (US) AnthonyJ. Griswold, PhD Fax: 612-454-2748 Orhun H. Kantarci, MD [email protected] Julie R. Korenberg, PhD, MD Davide Pareyson, MD Shoji Tsuji, MD,PhD DinekeS. Verbeek,PhD Stay facebook.com/NeurologyGenetics Connected David Viskochil, MD,PhD twitter.com/greenjournal JulianeWinkelmann, MD Juan I. Young, PhD youtube.com/user/NeurologyJournal

instagram.com/aanbrain

Neurology® Journals

Editor-in-Chief Jos´eG.Merino,MD,MPhil,FAAN

Deputy Editor Classification of Evidence Olga Ciccarelli, MD, PhD, FRCP Review Team Melissa J. Armstrong, MD Section Editors Richard L. Barbano, MD,PhD, FAAN Biostatistics RichardM.Dubinsky,MD,MPH,FAAN Richard J. Kryscio, PhD Jeffrey J. Fletcher, MD, MSc Sue Leurgans, PhD Gary M. Franklin, MD, MPH, FAAN V. Shane Pankratz, PhD David S. Gloss II, MD,MPH&TM John J. Halperin, MD,FAAN fi Classi cation of Evidence Evaluations Jason Lazarou, MSc, MD Gary S. Gronseth, MD, FAAN Steven R. Messé, MD, FAAN Equity, Diversity, & Inclusion (EDI) Pushpa Narayanaswami, MBBS, DM, RoyH.Hamilton,MD,MS,FAAN FAAN Holly E. Hinson, MD, MCR, FAAN Alex Rae-Grant, MD

Podcasts Stacey L. Clardy, MD, PhD, FAAN Jeﬀrey B. Ratliﬀ, MD, Deputy Podcast Editor

Ombudsman Jonathan W. Mink, MD, PhD, FAAN

Scientiﬁc Integrity Advisor David S. Knopman, MD, FAAN

Copyright ª 2020 American Academy of Neurology. Unauthorized reproduction of this article is prohibited. TABLE OF CONTENTS Volume 6, Number 3, June, 2020 Neurology.org/NG

The Helix e424 TGM6 L517W is not a pathogenic variant for spinocerebellar ataxia type 35 e438 The Helix: Editorial Changes Y. Chen, D. Wu, B. Luo, G. Zhao, and K. Wang S.M. Pulst Open Access Open Access e425 Expanding the phenotypic and molecular spectrum of Editorial RNA polymerase III–related leukodystrophy e436 Intronic pentanucleotide expansion in the replication S. Perrier, L. Gauquelin, C. Fallet-Bianco, M.K. Dishop, RFC1 M.A. Michell-Robinson, L.T. Tran, K. Guerrero, L. Darbelli, M. Srour, factor 1 gene ( ) is a major cause of adult-onset K. Petrecca, D.L. Renaud, M. Saito, S. Cohen, S. Leiz, B. Alhaddad, ataxia T.B. Haack, I. Tejera-Martin, F.I. Monton, N. Rodriguez-Espinosa, D. Pohl, S. Nageswaran, A. Grefe, E. Glamuzina, and G. Bernard S.M. Boesch and M.A. Nance Open Access Open Access e426 Phenotypic variability in chorea-acanthocytosis Articles associated with novel VPS13A mutations e414 Neuraxial dysraphism in EPAS1-associated syndrome V. Niemelä, A. Salih, D. Solea, B. Lindvall, J. Weinberg, G. Miltenberger, T. Granberg, A. Tzovla, L. Nordin, T. Danfors, due to improper mesenchymal transition I. Savitcheva, N. Dahl, and M. Paucar J.S. Rosenblum, A.J. Cappadona, D.P. Argersinger, Y. Pang, H. Wang, Open Access M.A. Nazari, J.P. Munasinghe, D.R. Donahue, A. Jha, J.G. Smirniotopoulos, M.M. Miettinen, R.H. Knutsen, B.A. Kozel, Z. Zhuang, K. Pacak, and J.D. Heiss e428 Mutations in the m-AAA proteases AFG3L2 Open Access and SPG7 are causing isolated dominant optic atrophy e415 Neurologic outcomes in Friedreich ataxia: Study of M. Charif, A. Chevrollier, N. Gueguen, C. Bris, D. Goudenège, a single-site cohort V. Desquiret-Dumas, S. Leruez, E.Colin,A.Meunier,C.Vignal, V. Smirnov, S. Defoort-Dhellemmes, I. Drumare Bouvet, M. Pandolfo C. Goizet, M. Votruba, N. Jurkute, P. Yu-Wai-Man, F. Tagliavini, Open Access L. Caporali, C. La Morgia, V. Carelli, V. Procaccio, X. Zanlonghi, I. Meunier, P. Reynier, D. Bonneau, P. Amati-Bonneau, and e416 Polygenic risk scores of several subtypes of epilepsies G. Lenaers in a founder population Open Access C. Moreau, R.-M. Rébillard, S. Wolking, J. Michaud, F. Tremblay, e430 A. Girard, J. Bouchard, B. Minassian, C. Laprise, P. Cossette, and Genotyping single nucleotide polymorphisms for S.L. Girard allele-selective therapy in Huntington disease Open Access D.O. Claassen, J. Corey-Bloom, E.R. Dorsey, M. Edmondson, S.K. Kostyk, M.S. LeDoux, R. Reilmann, H.D. Rosas, F. Walker, e417 Clinical and pathologic phenotype of a large family V. Wheelock, N. Svrzikapa, K.A. Longo, J. Goyal, S. Hung, and M.A. Panzara with heterozygous STUB1 mutation Open Access M.O. Mol, J.G.J. van Rooij, E. Brusse, A.J.M.H. Verkerk, S. Melhem, W.F.A. den Dunnen, P. Rizzu, C. Cupidi, J.C. van Swieten, and e440 RFC1 L. Donker Kaat Prevalence of -mediated spinocerebellar ataxia Open Access in a North American ataxia cohort D. Aboud Syriani, D. Wong, S. Andani, C.M. De Gusmao, Y. Mao, e418 Acute encephalopathy after head trauma in a patient M. Sanyoura, G. Glotzer, P.J. Lockhart, S. Hassin-Baer, V. Khurana, C.M. Gomez, S. Perlman, S. Das, and B.L. Fogel with a RHOBTB2 mutation Open Access A.C.S. Knijnenburg, J. Nicolai, L.A. Bok, A. Bay, A.P.A. Stegmann, M. Sinnema, and M. Vreeburg Clinical/Scientific Notes Open Access e423 Biallelic LINE insertion mutation in HACD1 causing e420 Cerebellar ataxia, neuropathy, hearing loss, and congenital myopathy intellectual disability due to AIFM1 mutation F. Al Amrani, C. Gorodetsky, L.-N. Hazrati, K. Amburgey, M. Pandolfo, M. Rai, G. Remiche, L. Desmyter, and I. Vandernoot H.D. Gonorazky, and J.J. Dowling Open Access Open Access

Copyright ª 2020 American Academy of Neurology. Unauthorized reproduction of this article is prohibited. TABLE OF CONTENTS Volume 6, Number 3, June, 2020 Neurology.org/NG

e432 Expanding the phenotype of MTOR-related disorders and the Smith-Kingsmore syndrome A. Elizondo-Plazas, M. Ibarra-Ram´ırez, A. Garza-B´aez, and L.E. Mart´ınez-de-Villarreal Open Access

Views and Reviews e434 Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy revisited: Genotype-phenotype correlations of all Cover image published cases Ex vivo Micro-CT of EPAS1-gain-of-function Transgenic Mouse Model demonstrating faulty ossiﬁcation of the posterior elements of G. Xiromerisiou, C. Marogianni, K. Dadouli, C. Zompola, ﬁ D. Georgouli, A. Provatas, A. Theodorou, P. Zervas, C. Nikolaidou, the cervical and thoracic spine, speci cally the spinous process; the S. Stergiou, P. Ntellas, M. Sokratous, P. Stathis, G.P. Paraskevas, transverse processes are similarly hypodense and there is A. Bonakis, K. Voumvourakis, C. Hadjichristodoulou, a dysraphism of T1. G.M. Hadjigeorgiou, and G. Tsivgoulis See e414 Open Access

Copyright ª 2020 American Academy of Neurology. Unauthorized reproduction of this article is prohibited. Volume 6, Number 3, June 2020 The Helix Neurology.org/NG

Stefan M. Pulst, MD, Dr med, Editor, Neurology®: Genetics The Helix: Editorial Changes

Neurol Genet June 2020 vol. 6 no. 3 e438. doi:10.1212/NXG.0000000000000438

It gives me great pleasure to welcome Dr. Suman Jayadev to the group of Associate Editors of Neurology® Genetics. She is a board-certiﬁed neurologist and Assistant Professor of Neurology at the University of Washington. She runs an adult neurogenetics clinic at the University of Washington and is also the Clinical Core Leader of the University of Washington Alzheimer Disease Research Center. Her neurogenetics laboratory focuses on inﬂammatory mechanisms related to neurodegeneration.

Dr. Jayadev has taken over from Dr. Jeffery Vance who has provided his expertise on the genetics of neurodegenerative diseases for the first 5 years of the journal. I want to thank Jeff for his tireless efforts, insights, collegiality, and friendship.

Study funding No targeted funding reported.

Disclosure S.M. Pulst serves on the editorial boards of the Journal of Cerebellum, NeuroMolecular Medicine, Experimental Neurology, Neurogenetics, and Nature Clinical Practice Neurology and as Editor-in- Chief of Current Genomics; receives research support from the NIH and the National Ataxia Foundation; has served on the speakers’ bureau of Athena Diagnostics; receives publishing royalties from Churchill Livingston, AAN Press, Academic Press, and Oxford University Press; has consulted for Ataxion Therapeutics; has received license fee payments from Cedars-Sinai Medical Center; holds multiple patents; and receives an honorarium from the AAN as the Editor of Neurology: Genetics. Go to Neurology.org/NG for full disclosures.

From the Department of Neurology University of Utah, Salt Lake City. This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloading and sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Copyright © 2020 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. e1 EDITORIAL OPEN ACCESS Intronic pentanucleotide expansion in the replication factor 1 gene (RFC1) is a major cause of adult-onset ataxia

Sylvia M. Boesch, MD, and Martha A. Nance, MD Correspondence Ass. Prof. Dr. Boesch Neurol Genet 2020;6:e436. doi:10.1212/NXG.0000000000000436 [email protected]

The ataxias comprise diseases of both genetic and nongenetic origin with extreme clinical and RELATED ARTICLE genetic heterogeneity. They may present as a pure cerebellar form or as part of a more complex neurologic syndrome. Progressive, neurodegenerative sporadic adult-onset ataxias (SAOAs) Prevalence of RFC1- mediated spinocerebellar without a known cause have a prevalence rate of 2.2–12.4 per 100,000. In several ataxia cohorts, ataxia in a North American repetitive genetic screening using high-coverage ataxia-specific gene panels in combination with – ataxia cohort next-generation sequencing (NGS) failed to identify a causative gene in 50%–90% of SAOAs.1 3 Cerebellar ataxia, neuropathy, vestibular areflexia syndrome (CANVAS), first described by Page e440 Brownstein et al.,4 is a slowly progressive neurodegenerative disorder with adult onset, affecting the cerebellum, sensory neurons, and the vestibular system. CANVAS is usually sporadic, but occasionally occurs in siblings. Two research groups recently identified large biallelic intronic AAGGG expansions in replication factor C subunit 1 (RFC1) resulting in CANVAS, an adult- – onset neurodegenerative ataxia.5 7 RFC1 normally loads proliferating cell nuclear antigen onto DNA and activates DNA polymerases δ and e to promote the coordinated synthesis of both strands during replication or after DNA damage.8

In this issue of Neurology® Genetics, Syriani et al.9 investigated the prevalence of intronic AAGGG expansions in the RFC1 gene in a North American cohort of 911 predominantly adult-onset patients with undiagnosed familial or sporadic ataxia. Testing in this cohort revealed 29 patients with biallelic expansions (3.2%), one-third of whom had the full CANVAS syndrome. The remaining had late-onset ataxia frequently accompanied by neuropathy (60%). All RFC1 expansion carriers were Caucasian. The rate of heterozygosity was as high as 6.8%, which may be caused by overrepresented alleles with repeat lengths below 400 repeats—the pathogenic threshold in RFC1 anticipated in 2 previous studies.

Nucleotide repeat disorder reloaded Repetitive DNA sequences constitute approximately one-third of the genome. There is evidence that they may contribute to diversity within and between species. They display considerable variability in length between individuals, which is presumed to have no detrimental consequences unless the repeat number is expanded beyond a gene-speciﬁc threshold. Pathologic unstable repeat expansions are classiﬁed according to their length, repeat sequence, gene location, and underlying pathologic mechanisms. Large (hundreds-thousands of copies) pathogenic repeat expansions are typically located in noncoding regions including promoters, introns, and untranslated regions of genes and can show somatic instability. Repeat expansions in introns are thought to produce aberrant repeat-bearing RNAs that interact with and sequester a wide variety of essential proteins, resulting in cellular toxicity.

Targeted non–sequence-based testing is still the method of choice to detect nucleotide repeat expansions in the human genome. Commonly used NGS techniques such as whole genome sequencing (WGS) and whole exome sequencing (WES) fail to detect repetitive regions.

From the Medical University Innsbruck (S.M.B.); and Struthers Parkinson’s Center (M.A.N.).

Go to Neurology.org/NG for full disclosures. Funding information is provided at the end of the article. This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloading and sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Copyright © 2020 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 Indeed, in the initial study in inherited CANVAS, WGS did of the pathogenic pentanucleotide and its function remains to not detect the causative mutation in RFC1.5 The use of tar- be elucidated by additional studies.5,6,10 geted non–sequence-based techniques and Southern blot finally led to the detection of 4 distinct intronic repeat The fact that this highly prevalent ataxia gene was unknown conformations in RFC1: AAAAG11 (the wild-type sequence) until now, highlights both the importance of precise pheno- and longer expansions of AAAAGn, AAAGGn, and AAGGGn. typing and sampling, as well as the use of analytic techniques Of these, the AAGGG pentanucleotide expanded up to beyond currently available panels and NGS. Intronic repeat 400–2,000 repeats was the only disease-causing condition. expansions, in particular, are difficult to identify but may be The expansion occurs in the poly(A) tail of an AluSx3 ele- common causes of neurodegenerative disease. ment, and differs in both size and nucleotide sequence from the reference (AAAAG)11 allele. Study funding No targeted funding reported.

Haplotype and allele carrier frequency Disclosure in RCF1 The authors report no disclosures relevant to the manuscript. Go to Neurology.org/NG for full disclosures. The same ancestral haplotype is shared by the majority of familial and positive RFC1 cases, as well as some healthy car- Publication history riers of 2 (AAAGG)exp alleles. It is likely that the nucleotide Received by Neurology: Genetics April 15, 2020. Accepted in change from AAAAG to AAAGG or AAGGG represents an final form April 20, 2020. ancestral founder event, followed by the pathologic expansion of the repeated unit, whose size seems to be related to its References 1. Fogel BL, Lee H, Deignan JL, et al. Exome sequencing in the clinical diagnosis of guanine-cytosine content. Up to now, analyses of the core sporadic or familial cerebellar ataxia. JAMA Neurol 2014;71:1237–1246. Erratum in: haplotype in the mixed ethnic cohort confirmed the European JAMA Neurol 2015;72:128. core haplotype estimated to have arisen more than 25,000 years 2. Giordano I, Harmuth F, Jacobi H, et al. Clinical and genetic characteristics of sporadic adult-onset degenerative ataxia. Neurology 2017;89:1043–1049. ago. Although the AAGGG repeat expansion has been identi- 3. Sun M, Johnson AK, Nelakuditi V, et al. Targeted exome analysis identifies the genetic fied in non-European individuals (Native American, Arabic, basis of disease in over 50% of patients with a wide range of ataxia-related phenotypes. Genet Med 2018;21:195–206. and Japanese), it remains highly overrepresented in pop- 4. Bronstein AM, Mossman S, Luxon LM. The neck-eye reflex in patients with reduced ulations of European descent, with frequencies of 4%–6.8% vestibular and optokinetic function. Brain 1991;114:1–11. 7,9–11 5. Cortese A, Simone R, Sullivan R, et al. Author Correction: biallelic expansion of an (White and Hispanic). intronic repeat in RFC1 is a common cause of late-onset ataxia. Nat Genet 2019;51:920. 6. Cortese A, Tozza S, Yau WY, et al. Cerebellar ataxia, neuropathy, vestibular areflexia 9 syndrome due to RFC1 repeat expansion. Brain 2020;143:480–490. In summary, the study by Syriani et al. supports the notion 7. Rafehi H, Szmulewicz DJ, Bennett MF, et al. Bioinformatics-based identification of that the newly discovered RCF1 gene is a major cause of expanded repeats: a non-reference intronic pentamer expansion in RFC1 causes CANVAS. Analysis of RCF1 should be included in clinical CANVAS. Am J Hum Genet 2019;105:151–165. 8. Overmeer RM, Gourdin AM, Giglia-Mari A, et al. Replication factor C recruits DNA diagnostic testing of adult-onset neurodegenerative ataxia, polymerase delta to sites of nucleotide excision repair but is not required for PCNA especially when neuropathy is present. There are multiple recruitment. Mol Cell Biol 2010;30:4828–4839. 9. Syriani DA, Wong D, De Gusmao CM, et al. Prevalence of RFC1-mediated spino- areas for future work, including deep phenotyping in sporadic cerebellar ataxia in a North American ataxia cohort. Neurol Genet 2020;6:e440. doi: adult-onset ataxias, analyses for a correct determination of 10.1212/NXG.0000000000000440. 10. Akçimen F, Ross JP, Bourassa CV, et al. Investigation of the RFC1 repeat expansion in pathogenic repeat lengths, and the stability of this pentanu- a Canadian and a Brazilian ataxia cohort: identification of novel conformations. Front cleotide repeat sequence across siblings and generations Genet 2019;10:1219. fi 11. Nakamura H, Doi H, Mitsuhashi S, et al. Long-read sequencing identifies the path- within families. Moreover, an understanding of the ne ogenic nucleotide repeat expansion in RFC1 in a Japanese case of CANVAS. J Hum structure of RFC1 as it relates to the final repeat composition Genet 2020;65:475–480.

2 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG ARTICLE OPEN ACCESS Neuraxial dysraphism in EPAS1-associated syndrome due to improper mesenchymal transition

Jared S. Rosenblum, MD,* Anthony J. Cappadona, BS,* Davis P. Argersinger, BS, Ying Pang, MD, PhD, Correspondence Herui Wang, PhD, Matthew A. Nazari, MD, Jeeva P. Munasinghe, PhD, Danielle R. Donahue, BS, Dr. Heiss [email protected] Abhishek Jha, MD, James G. Smirniotopoulos, MD, Markku M. Miettinen, MD, PhD, Russell H. Knutsen, BA, Beth A. Kozel, MD, PhD, Zhengping Zhuang, MD, PhD, Karel Pacak, MD, PhD, DSc,* and John D. Heiss, MD*

Neurol Genet 2020;6:e414. doi:10.1212/NXG.0000000000000414 Abstract Objective To investigate the eﬀect of somatic, postzygotic, gain-of-function mutation of Endothelial Per- Arnt-Sim (PAS) domain protein 1 (EPAS1) encoding hypoxia-inducible factor-2α (HIF-2α)on posterior fossa development and spinal dysraphism in EPAS1 gain-of-function syndrome, which consists of multiple paragangliomas, somatostatinoma, and polycythemia.

Methods Patients referred to our institution for evaluation of new, recurrent, and/or metastatic paragangliomas/pheochromocytoma were confirmed for EPAS1 gain-of-function syndrome by identification of the EPAS1 gain-of-function mutation in resected tumors and/or circulating leukocytes. The posterior fossa, its contents, and the spine were evaluated retrospectively on available MRI and CT images of the head and neck performed for tumor staging and restaging. The transgenic mouse model underwent Microfil vascular perfusion and subsequent intact ex vivo 14T MRI and micro-CT as well as gross dissection, histology, and immunohistochemistry to assess the role of EPAS1 in identified malformations.

Results All 8 patients with EPAS1 gain-of-function syndrome demonstrated incidental posterior fossa malformations—one Dandy-Walker variant and 7 Chiari malformations without syringomyelia. These ﬁndings were not associated with a small posterior fossa; rather, the posterior fossa volume exceeded that of its neural contents. Seven of 8 patients demonstrated spinal dysraphism; 4 of 8 demonstrated abnormal vertebral segmentation. The mouse model similarly demonstrated features of neuraxial dysraphism, including cervical myelomeningocele and spinal dysraphism, and cerebellar tonsil displacement through the foramen magnum. Histology and immunohistochemistry demonstrated incomplete mesenchymal transition in the mutant but not the control mouse.

Conclusions This study characterized posterior fossa and spinal malformations seen in EPAS1 gain-of- function syndrome and suggests that gain-of-function mutation in HIF-2α results in improper mesenchymal transition.

*These authors contributed equally to the manuscript.

From the National Institutes of Health (J.S.R., A.J.C., H.W., Z.Z.), National Cancer Institute Neuro-Oncology Branch; National Institutes of Health (D.P.A., J.D.H.), National Institute of Neurological Disorders and Stroke, Surgical Neurology Branch; National Institutes of Health (Y.P., A.J., K.P.), Eunice Kennedy Shriver National Institute of Child Health and Human Development, Section on Medical Neuroendocrinology; Georgetown Hospital (M.A.N.), Internal Medicine and Pediatrics, Washington DC; National Institutes of Health (J.P.M., D.R.D.), National Institute of Neurological Disorders and Stroke, Mouse Imaging Facility, Bethesda, MD; George Washington University (J.G.S.), Radiology, Washington DC; National Library of Medicine (J.G.S.), MedPix®; National Institutes of Health (M.M.M.), Center for Cancer Research, National Cancer Institute, Laboratory of Pathology; and National Institutes of Health (R.H.K., B.A.K.), National Heart Lung and Blood Institute, Translational Vascular Medicine Branch, Bethesda, MD.

Go to Neurology.org/NG for full disclosures. Funding information is provided at the end of the article.

The Article Processing Charge was funded by NIH. This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloading and sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Copyright © 2020 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 Glossary CMI = Chiari malformation; NICHD = National Institute of Child Health and Development.

Chiari malformations (CMIs) encompass a spectrum of included MRI and CT of the head and neck and CT of the neuraxial failures of dorsal induction depending on the extent lower spine. We evaluated the available MRI scans to charac- of longitudinal neuraxial involvement.1 Asymptomatic CM1 terize the anatomy of the brain, including the posterior fossa may be incidentally discovered on MRI in adults.2 CMI has and cervical spine. We evaluated the CT reconstructed images been attributed to the internal volume of the posterior fossa to describe the bony anatomy of the calvarium, skull base, and being too small to contain its neural contents, producing spine. downward displacement of the cerebellar tonsils, which may – obstruct CSF flow and cause syringomyelia.3 5 CMI is Measurements sometimes associated with syndromes that impair normal We used the previously described methods to measure pos- – 9,19–25 bone development, e.g., achondroplasia.3,6 9 terior fossa structures. Details in e-methods (links.lww. com/NXG/A252). We previously identified CMI and occult spinal dysraphism in 2 patients with EPAS1 gain-of-function syndrome.10 Early Bone density calculations (Hounsfield Units) from CT somatic gain-of-function mutations in EPAS1 encoding the abdomen/pelvis protein hypoxia-inducible factor-2α (HIF-2α) causes a syn- We evaluated the bone density at the L5 vertebral body and sacrum on available CT scans using the same settings,26,30 drome characterized by multiple paragangliomas, somatosta- 26,27 tinoma, and polycythemia.11 These EPAS1 mutations affect using OsiriX Imaging Software. Details in e-methods the oxygen degradation domain of HIF-2α and impair hy- (links.lww.com/NXG/A252). droxylation by prolyl hydroxylase domain-containing protein 2 and subsequent association of HIF-2α with the von Hippel- Laboratory studies 12,13 α Lindau protein. This impairs the degradation of HIF-2 Patient EPAS1 mutation analysis 14,15 and the response to normal or increasing oxygen tension. Genomic DNA was extracted from patient tumor tissue and α Reduced HIF-2 activity promotes endochondral and intra- leukocytes via the NucleoSpin Tissue Kit (Macherey-Nagel, fi 16,17 membranous ossi cation and bone repair. We thus hy- Bethlehem, PA). PCR amplified EPAS1 exons. Primer sets for α pothesized that HIF-2 gain-of-function in our patients led to exon amplification were previously described.11 Sanger se- persistent hypoxic signaling, incomplete mesenchymal de- quencing determined the DNA sequence of each exon in velopment, reduced bony development of the spine and skull 11 10 tumors and leukocytes. Peptide nucleic acid sequencing in base, and CM1 with associated sacral abnormalities. blood leukocytes later confirmed mosaicism of the mutations in the 2 index patients.18 The present study evaluates 8 patients with EPAS1 gain-of- function syndrome for the presence of posterior fossa and Transgenic mouse model spinal malformations. We also investigated the pathogenesis All animal studies were reviewed and approved by the Animal of the identified malformations in a transgenic EPAS1 gain-of- Care and Use Committee of NICHD. Transcription activator- function syndrome mouse model. like effector nucleases-mediated homologous recombination Epas1A529V established mutant mice via a previous study.28 This introduces the point mutation, A529V (guanine, cytosine, ade- Methods nine>guanine, thymine, adenine) in exon 12 of Epas,distaltothe Standard protocol approvals, registrations, insertion of a reverse oriented neo cassette flanked by loxP sites and patient consents in intron 11. In its native state, the neo cassette inhibits tran- The institutional review board of the Eunice Kennedy Shriver scription from the altered allele, resulting in a haploinsufficiency. National Institute of Child Health and Development This haploinsufficient mouse does not demonstrate poly- 28 (NICHD, ClinicalTrials.gov Identifier: NCT00004847) ap- cythemia or noradrenergic phenotype. When bred to the en- proved the study protocol, and written informed consent was zyme 2a cyclization recombinase (E2a-Cre)mouse,Cre obtained from all participants. The research study followed all expressed in the early postzygotic embryo leads to excision of the 29 applicable institutional and governmental regulations con- loxP flanked neo cassette on a sporadic basis, allowing the cerning the ethical use of animals. successful transcription of the A529V gain-of-function mutant allele in a subset of the cells. Therefore, 4 categories of mice Patient selection and evaluation result from each litter: (1) the gain-of-function mutant (A529V Patients met the criteria of the syndrome including para- variant positive, Cre positive, henceforth referred to as “mu- ganglioma, polycythemia, and confirmed EPAS1 somatic mutant”), (2) the Cre-only control (variant negative, Cre positive), tation.18 Anatomic imaging for tumor staging and restaging (3) the haploinsufficient mouse (variant positive, Cre negative),

2 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG Figure 1 Posterior fossa and spine malformations in patients with EPAS1 gain-of-function syndrome

Panels (A–C): Representative images of cerebellar tonsillar displacement and posterior fossa morphology in the patients with incidental Chiari malformation; the same patient is shown. Panel (A): MRI of the brain, sagittal postcontrast T1-weighted sequence shows the left cerebellar tonsil at the lowest tonsillar position, 8 mm below the foramen magnum. The clivus is narrow and underdeveloped; the supraoccipital bone and the uncalcified synchondroses of the odontoid are also shown (arrowheads). Panel (B): Coronal T1-weighted sequence demonstrating the location of the measurement in (A). Panel (C): Axial postcontrast T1-weighted sequence showing crowding of the brainstem by the cerebellar tonsils within the foramen magnum (arrows). Panel (D): MRI of the brain, sagittal postcontrast T1-weighted sequence shows a Dandy-Walker variant malformation with communication between the mega cisterna magna (arrow) and the fourth ventricle. There is improper cerebellar rotation (dashed arrow). The same morphology of the clivus, odontoid, and supraoccipital bone was seen in the patients with Chiari malformation (arrowheads). Panel (E): Axial postcontrast T1-weighted sequence of the same patient as in panel (D) shows the connection between the mega cisterna magna (arrow) and the fourth ventricular space (dashed arrow). Panel (F): Representative lumbosacral spineCT volumetric reconstruction of patient 3 shows dysraphism of the posterior elements of the entire sacrum. Panel (G): Volumetric reconstruction of CT ofthe lumbosacral spine of patient 7 shows abnormal sacral segmentation. and (4) a wild type (WT) control (variant negative, Cre nega- before bilateral thoracotomy. Mice were then positioned su- tive). Previous quantification of this allele using digital droplet pine and incised in the midline from the xiphoid to the pubis. PCR in reverse transcribed complementary DNA in multiple Once exposed, a hemostat secured the xiphoid. The di- tissues showed that the excision occurred in 30%–100% of cells aphragm was removed. The rib cage was cut bilaterally just in gain-of-function mutants, depending on the tissue type.28 short of the internal thoracic arteries and then elevated to reveal the thoracic cavity. The inferior vena cava was trans- Wild type or Cre-only mice were used as controls for all ected and the descending thoracic aorta exposed. Two sutures experiments; haploinsufficient mice from the litter were not were placed under the aorta using 7-0 silk (Teleflex, Coventry, studied. Experiments were performed on 3- to 5-month-old CT) and a small opening made in the vessel. A perfusion mice. Five mutant mice (4 males, 1 female) and 4 control mice catheter, consisting of 15 cm of PE-10 tubing (Instech, Ply- (2 males, 2 females) underwent Microfil vascular perfusion mouth Meeting, PA) threaded onto a 30G needle (Beckton and subsequent ex vivo 14T MRI, CT scans, volumetric re- Dickinson, Franklin Lakes, NJ) and connected to a syringe, construction, and gross dissection. For bone mineral density was placed in the opening, advanced beyond the sutures, and calculation in the mouse model, ex vivo micro-CT was re- secured in place. Five milliliters of 10-4 M sodium nitroprus- peated on 8 nonperfused mice to avoid the density incurred side (Sigma, St. Louis, MO) in 1× phosphate-buffered saline by Microfil in the bone vascular spaces. Two mutant and 2 was then perfused through the vasculature to remove excess control mouse nonperfused calvaria were isolated with dura blood and assure maximal dilation. Microfil (Flow Tech Inc, and venous sinuses intact for decalcification, histology, he- Carver, MA), mixed in a 2:7:1 (Microfil:diluent:polymerizing matoxylin & eosin staining, and immunohistochemistry. agent) ratio, optimized to allow arterial-venous transit, was slowly perfused until visibly exiting the inferior vena cava. At Vascular perfusion of mouse model this point, the inferior vena cava and right auricle were ligated, Mice were heparinized, allowed to ambulate for 2 minutes, the animal was placed in the prone position, and the skull was and subsequently euthanized via carbon dioxide narcosis exposed to visualize the dural venous sinuses and diploic

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 3 veins. Microfil injection was continued to fill vascular net- ) 3 works (figure e-1, links.lww.com/NXG/A252). PFV (cm

Ex vivo imaging of the mouse model head and spine ) 3 Mouse samples were scanned via 14T MRI and micro-CT.

PFCV (cm OsiriX Imaging Software and Bruker software suite for MRI and CT imaging facilitated image reconstruction and analysis. respectively. Details in e-methods (links.lww.com/NXG/ KI (mm) A252).

Histology and immunohistochemistry

PFH (mm) Immunohistochemistry detected cells stained by HIF-2α primary antibody (Abcam [ab109616], 1:500 ab:PBST) within formalin-fixed, decalcified, paraffin-embedded sec-

PMH (mm) tions. Details in e-methods (links.lww.com/NXG/A252).

Data archiving The published article and supplementary information ﬁles tentorium-twining line angle; VCSFS = ventral CSF space.

Tonsil shape include data generated or analyzed during this study. osition; PFCV = posterior fossa contents volume; PFV = posterior fossa volume;

LTP (mm) Results Posterior fossa characteristics in patients with EPAS1 gain-of-function syndrome LCT (mm) All 8 patients had conﬁrmed EPAS1 somatic mutation. Of these, all had malformation of the posterior fossa and 7 demonstrated spinal dysraphism. Figure 1 shows the repre- RCT (mm) sentative images of the posterior fossa malformations, and section e-4 (links.lww.com/NXG/A252) shows the spinal malformations. One patient demonstrated a Dandy-Walker SOL (mm) variant malformation characterized by mild improper closure of the cerebellar vermis and a wide communication between the fourth ventricle and cisterna magna (ﬁgure 1, D and E). CL (mm) Seven of the 8 patients demonstrated downward position of the cerebellar tonsils. Table 1 contains the measurements of the posterior fossa and contents for each patient; section e-3

DCSFS (mm) includes additional patient sagittal MRI measurements. The cerebellar tonsils extended by more than 5 mm below the foramen magnum in 2 patients. The distance of the cerebellar tonsils below the foramen magnum (excluding both the pe- VCSFS (mm) diatric patient and the patient with the Dandy-Walker variant) was 4 ± 3 mm (mean ± SD) (table 2). No patient had pla-

TA (°) tybasia (range 115.36°–132.99°; mean Boogaard Angle 123.2° ± 6.4°). The posterior fossa was larger (205 ± 24 cm3) than its gain-of-function syndrome. Measurements characterizing the posterior fossa malformations in the 8 syndrome patients. 3

TTA (°) contained neural elements (174 ± 20 cm ). Table 2 shows the

EPAS1 normal values for these structures. No patient demonstrated syringomyelia or obstruction of subarachnoid spaces at the

BA (°) craniovertebral junction.

Bone mineral density in patients with EPAS1 gain-of-function syndrome In 7 of 8 of patients, bone density, evaluated at the L5 vertebral body and sacrum on available CT scans,26,30 was in the c.1595A>Gp.Y532C 132.99c.1589C>Tp.A530V 35.98 126.39c.1588G>Ap.A530T 39.53 38.19 115.36 6.7c.1615G>AP.D539N 48.73 43.93 124.61 11.1c.1591C>Tp.P531S 51.17 53.62 118.2 11 18.9c.1589C>Tp.A530V 51.58 15.8 6.3 43.67 129.23c.1615G>Ap.D539N 60.59 36.31 120.12 39.3c.1615G>Ap.D539N 16.1 13.4 38.27 34.43 40.8 120.35 16.8 45 11.3 53.33 40.51 38.19 8.8 46.9 19.6 55.81 35.4 6.3 15.5 7.8 4.1 47.7 48.7 18.9 38.3 7.3 0 38.8 0 20.1 49.4 3.4 45.5 48.7 7.3 0 2.7 4.1 48.7 37.7 2.2 2.2 Round 45 2.7 Round 3.9 0 3.4 0 16 6 15 Round 2.6 Round 0 2.2 33.1 17.9 3.9 30.2 8.2 13.2 Round 48.5 Round 35.6 41.6 Round 27.1 8.2 193.2 19.6 8.2 198.5 50.1 15.4 36.1 Pegged 161.2 208.9 39.5 158.2 36.3 166.8 12.8 29.8 175.2 49.6 47.9 148.8 44.1 35 246.8 205.3 167.9 199.8 48.9 184.0 147.4 212.9 194.8

Syndrome patient measurements on MRI of the brain normal range compared with available literature for sex- and age-matched averages (table 3). Patient 7, who displayed severe abnormal segmentation of the sacrum (ﬁgure 1G), Table 1 Patient Mutation status 1 2 3 4 5 6 7 8 Abbreviations: BA = Boogaard angle; CL =PFH clival = length; DCSFS posterior = dorsal fossa CSF height; space; KI PMH = = Klaus index; pontomedullary LCT height; = left RCT cerebellar tonsil; = LTP right = lowest cerebellar tonsillar tonsil; p SOL = supraoccipital bone length; TA = tentorial angle; TTA = Posterior fossa measurements in patients exhibited a low density, one SD below the mean normal value.

4 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG Table 2 EPAS1 gain-of-function syndrome patient posterior fossa characteristics

Characteristic Mean SD

Boogaard angle (133.8 ± 6.5°)21 123.2 6.4

Tuberculum sellae-IOP-straight sinus tentorial angle (26.5°–56.4°)20 39.0 3.8

Nasion-tuberculum sellae-straight sinus tentorial angle (27°–52°)9 49.6 8.2

Ventral CSF space (12 ± 2.3 mm)21 10 2

Dorsal CSF space (19 ± 2.3 mm)21 18 2

Clival length (43.4 ± 4.4 mm)21 44 4

Supraoccipital bone length (41 ± 5 mm)21 45 5

Right cerebellar tonsil (mm) 33

Left cerebellar tonsil (mm) 33

Lowest tonsillar ectopia (mm) 43

Patients with >5 mm tonsillar ectopia (n) 2 —

Pontomedullary junction to foramen magnum (19 ± 3mm)21 15 4

Posterior fossa height (32 ± 3mm)21 34 3

Klaus index (38.0 ± 5mm)21 47 3

Posterior fossa volumetric measurement (159.58 ± 29.73 cm3)35 205 24

Posterior fossa contents volume (196.1 ± 10.8 cm3)36 174 20

Abbreviation: IOP = internal occipital protuberance. EPAS1 gain-of-function syndrome patient posterior fossa characteristics. Mean and SD of measurements of the posterior fossa in the 6 adult EPAS1 gain-of-function syndrome patients with incidentally discovered cerebellar tonsil displacement through the foramen magnum are shown.

Neuraxial dysraphism in syndrome the posterior fossa and spine similar to the patient findings, mouse model although of greater severity (figure 2). The displacement of MRI (14T) of the head and neck of the intact mouse after the cerebellar vermis through the foramen magnum was vascular Microfil perfusion, fixation, and gadolinium contrast identified in 5 of 7 mutant mice (figure 2, A and B), and infusion demonstrated a spectrum of dysraphic processes of cervical myelomeningocele was found in 5 of 7 evaluated

Table 3 Patient bone density (HU)

Male Female

L5 Sacrum Literature control lumbar L5 Sacrum Literature control lumbar Patient Age density density density (HU) by age26 Patient Age density density density (HU) by age26

1 46 252.6 191.6 192.8 ± 15.5 5 20 296.0 230.0 240.2 ± 43.2

2 27 329.7 255.6 248.1 ± 52.1 8 29 299.0 — 240.2 ± 43.2

3 38 296.9 282.5 188.9 ± 35.3

4 13 289.7 251.3 253.5 ± 29.6

6 43 297.7 242.7 192.8 ± 15.5

7 58 142.3 98.2 186.7 ± 40.7

Mean 37.5 268.1 24.5 297.5

SD 15.7 66.4 6.4 2.1

Abbreviation: HU = Hounsfield unit. Bone density (HU) from CT scans in EPAS1 gain-of-function syndrome patients. Bone density of the L5 vertebral body and sacrum was evaluated from available CT scans of the abdomen/pelvis. The pelvic scan in patient 8 did not extend to include the entire sacrum. Patient age is at the time of the scan. Literature control age ranges separated by decade and reported in HU.

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 5 Figure 2 Dysraphisms in EPAS1 gain-of-function transgenic mouse model

Representative images of posterior fossa and spinal cord neural tube defects in the transgenic mouse model of EPAS1 gain-of-function syndrome are shown. Panel (A): Ex vivo 14T MRI axial plane at the craniocervical junction shows cerebellar parenchyma (arrow) extending through the foramen magnum with an attachment to the upper cervical spinal cord (arrowhead). Panel (B): Ex vivo 14T MRI sagittal plane performed on a Microfil-perfused mouse demonstrates a cervical myelomeningocele (arrowheads); large leptomeningeal veins surrounding the spinal cord are shown (arrows). Panel (C): Ex vivo micro-CT ofthe intact mouse shows abnormal vertebral segmentation of left L5 transverse process and S1 ala (arrows). Panel (D): Ex vivo micro-CT of the intact mouse demonstrates faulty ossification of the posterior elements of the cervical and thoracic spine (double-lined arrows), specifically the spinous process (arrow); the transverse processes are similarly hypodense (asterisk). There is a dysraphism of T1 (dashed arrow). Panel (E): Axial gross dissection of the cervical myelomeningocele shows a large leptomeningeal vein (arrowhead) below the parenchymal tissue marked by the dashed arrow in panel (B). Parenchymal spinal cord tissue is seen outside the normal border (arrow). Lack of ossification of the posterior elements of the spinal column (asterisk) is appreciated lateral to the primary ossification center (dashed arrow).

mutant mice, including 3 of the mice with cerebellar dis- temporal bone, petrous temporal bone, and occipital bone placement (figure 2B). All mutant mice had enlarged lep- (figure 3, A and B). Furthermore, the tonsillar depression tomeningeal veins throughout the head and neck, and animals on the right, which is posterior to this abnormal bone, is with myelomeningocele had a large vein coursing through smaller than the left and is open to the foramen magnum. that structure (figure 2E). This finding recapitulates the finding of aberrant venous anatomy in 4 of the 8 patients. In these patients, the venous Micro-CT of the head and spine in the same Microfil-perfused drainage of the skull base around the foramen magnum is mutant mice demonstrated several consistent features altered (figure 3, C–E). First, the right jugular bulb is (figure 2D). Cervical spinous processes appeared absent at consistently developmentally enlarged, with a consistently some levels. The spinous processes and lamina that were diminutive left jugular bulb. Second, on the left, a large present, such as at C2, were hypodense, suggesting incomplete posterior condylar emissary vein is seen exiting its re- ossification. All mutant mice had occult dysraphism of the spective foramen with disorganized bony borders; this posterior elements at the T1 vertebra. Gross dissection of the emissary vein is not seen on the right side with the large samples confirmed cervical myelomeningocele shown by jugular bulb. Finally, this large emissary vein is seen micro-CT (figure 2D). Mouse lumbosacral micro-CT scans are draining to a large cervical venous plexus not appreciated shown in e-supplement section 5 (links.lww.com/NXG/A252). on the contralateral side.

The mutant mice also demonstrated abnormal development Decalciﬁed histologic sections of the calvarium at the level of of the bones of the posterior fossa related to the course of the torcular Herophili and petrous temporal bone, removed the major dural veins. In a representative image shown in intact, were evaluated by hematoxylin and eosin staining on ﬁgure 3A, the right sigmoid sinus can be seen coursing light microscopy. Histology showed varying patterns of size through an anomalous canal that is not present on the left. and the presence of bone marrow cavities, i.e., regions both This canal is formed by a bony ridge connected to an ab- lacking and abundant and transitional cartilage with poor normally segmented portion of bone joining the squamous separation from periosteum compared with the littermate

6 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG Figure 3 Anomalous venous patterns affecting clival-occipital development in patients and mouse model

Panel (A): Volumetric reconstruction of the ex vivo CT of the Microfil-perfused mouse model demonstrates an abnormal development of the right petrous temporal bone (asterisk), with a bony ridge from an aberrant segment of bone overlying the sigmoid sinus; this is not seen on the left side. Panel (B): The same reconstruction seen from 180° view compared with panel A. The jugular vein from the sigmoid sinus on the right (arrow) is seen on the other side of the abnormal bony ridge (asterisk). The tonsillar depressions in the occipital bone are appreciated (dashed arrows); the right, which is open to the foramen magnum, is smaller than the left because of the anomalous course of the venous system through the abnormal petrous temporal bone. Panels (C–E): Representative images of anomalous pattern of the posterior condylar emissary vein and occipital bone in the EPAS1 gain-of-function patients; the same patient (5) is shown. Panel (C): MRI of the brain, coronal postcontrast T1-weighted MR shows the posterior condylar emissary vein (PCEV) (arrow) arising from the sigmoid sinus and exiting through the occipital bone on the left; this is not seen on the right (arrowhead). Panel (D): Sagittal T1-weighted shows the course of the PCEV identified in panel (A) (arrow). Panel (E): Axial postcontrast T1-weighted sequence shows the PCEV coursing through the occipital bone with poor definition of a true foramen (arrow); the right jugular bulb is also seen to be developmentally larger than the left (arrowheads). controls (figure e-2, A and B, links.lww.com/NXG/A252). Discussion Immunohistochemistry with an antibody against HIF-2α showed strongly positive staining in the bone marrow in This report further characterizes the posterior fossa and spinal the mutant compared with the littermate control (figure e-2, malformations in EPAS1 gain-of-function syndrome. CMI and C and D). The transitional cartilage in the mutant did not sacral dysraphism were the most common malformations seen in stain with the HIF-2α antibody. our patient cohort. The co-occurrence of CM1, Dandy-Walker variant malformation, sacral dysraphism, and abnormal vertebral Bone mineral density in the mouse model segmentation suggest that they share a common mechanism in Bone mineral density calculations of S1 using reconstructed this syndrome, i.e., failure of mesenchymal transition mediated CT scans of 4 mutant mice and 4 WT mice demonstrated no by HIF-2α.Ourfindings align with proposals of previous significant difference between the mutant and control samples investigators that posterior fossa and dorsal spine malformations (figure e-3, links.lww.com/NXG/A252). For WT, the sample arise through the mechanisms preventing normal development mean was 0.36 with 95% confidence limits for the true pop- of brain, spinal cord, and CSF circulation pathways.1,4 ulation mean of 0.18 and 0.54. For the mutant, the sample mean was 0.37 with 95% confidence limits for the true pop- The larger than normal posterior fossa volume seen in ulation mean of 0.20 and 0.53. patients with EPAS1 gain-of-function syndrome with CMI

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 7 differs from the smaller than normal posterior fossa volume in The histology and immunohistochemistry of the calvarium in most patients with idiopathic CM1. The larger posterior fossa the disease mouse model further elucidates the pathogenesis in EPAS1 gain-of-function syndrome may prevent the symp- of the dysraphic processes in the syndrome. There appears to toms, obstruction of CSF flow, and syringomyelia that fre- be varying patterns of bone marrow architecture throughout quently occur in idiopathic patients with CM1. In EPAS1 the calvarium of the mutant mouse model compared with the gain-of-function syndrome, the posterior fossa volume can littermate control. Furthermore, the bone appeared in contain the entire cerebellum, eliminating disproportion be- a transitional stage of ossification with poor distinction between the posterior fossa container and contents as a mecha- tween the layers of the surrounding mesenchymal tissue. This nism of tonsillar descent. The cause of the cerebellar ectopia suggested that the decreased presence and ossification of in patients with EPAS1 gain-of-function syndrome likely ari- bone is because of failure of mesenchymal transition. ses from congenital malposition of the cerebellar tonsils. The congenital malposition theory is supported by the finding of Bone mineral density calculations in the mice confirmed no Dandy-Walker variant malformation, a known failure of cer- difference between the mutant mouse and littermate controls. ebellar rotation, in one of our patients. This was consistent with the patient bone density calculations, which were found to be elevated for 7 patients but within 2 The large posterior fossa in EPAS1 gain-of-function syndrome SDs of the mean for sex- and age-matched literature reference appears because of longer bones developing in the posterior values for all patients.26, Patient 7 had severely abnormal fossa. The basioccipital portion of the clivus, the supra- segmentation of the sacrum, which likely contributed to the occipital bone, and the dens have synchondroses.10 Raybaud low calculated bone density from CT, which is based on implicates stunted development of these structures in the a volumetric reconstruction. It appears based on these data, pathogenesis of idiopathic CMI.31 In morphologic studies of gross dissection, and histology that the bone that is present large series of patients with CMI,32 the posterior fossa is and ossified is normal, whereas there are some areas of bone smaller than its contents. We propose that in CM1 associated which have failed to transition from the developmental stages with EPAS1 gain-of-function syndrome, the duration of the of mesenchyme. For example, in the spine, gross dissection longitudinal cartilaginous development of the endochondral revealed a remnant primary ossification center with in- bones was prolonged, ossification was disrupted, and longer completely ossified posterior elements of the spine sur- and narrower posterior fossa bones resulted. rounding the myelomeningocele. Furthermore, it appears that this failure of mesenchymal transition occurs in regions with In our mouse model and patients, we observed that de- large developmentally remnant venous structures. velopmental anomalies of skull base veins, including the sigmoid sinus/jugular bulb and the posterior condylar emissary This study further identified and characterized posterior fossa veins, disrupted bony development around the foramen malformations and spinal dysraphism in patients with EPAS1 magnum. In embryo, veins organize before other vascular gain-of-function syndrome. We investigated the pathogenesis structures after passive diffusion is exceeded.33 Later, a hyp- of malformation and dysraphism in our disease mouse model oxic gradient stimulates the ingrowth of arteries into the and found evidence of persistent hypoxic signaling, persistent metaphysis toward the epiphysis, leading to cartilage forma- venous elements, and failure of mesenchymal transition. This tion and ossification.16 The mechanism of development of study further supports a common mechanism for posterior anomalous posterior fossa veins and enlarged posterior fossa fossa and spinal malformations in EPAS1 gain-of-function volume appears to be (1) reduced hypoxic gradient in the syndrome and establishes the role of HIF-2⍺ in their developing bone, (2) abnormally prolonged longitudinal de- pathogenesis. velopment of the bone, and (3) failure of mesenchymal transition. Acknowledgment The authors would like to acknowledge the Mouse Imaging The transgenic mouse model allowed investigation of the Facility of NINDS at NIH for the MRI and micro-CT studies. pathogenesis of these developmental disorders of the neuraxis. This study was supported, in part, by the Intramural Research The malformations were evaluated without disturbing their Programs of the Eunice Kennedy Shriver NICHD and NCI, anatomy using ex vivo MRI and CT scans after vascular perfu- NIH. The authors would also like to thank the patients and sion. The mouse model demonstrated a wide spectrum of participating health care professionals. malformations, including CMI, spinal dysraphism, and myelomeningocele associated with failure of dorsal induction, im- Study funding proper closure of the neural tube, and incomplete bone Supported by the intramural programs of the Eunice Kennedy development.34 Of note, the case of CMI shown in figure 2A Shriver National Institute of Child Health and Human De- demonstrated white matter connections between the cerebellum velopment; the National Cancer Institute; and the National and spinal cord, similar to the connections present in the cervical Institute of Neurologic Disorders and Stroke, NIH. myelomeningocele. Although this was not seen in the patient imaging, it does support malposition of the cerebellum as part of Disclosure the common mechanism behind CMI-IV in this syndrome. Disclosures available: Neurology.org/NG.

8 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG Publication history Received by Neurology: Genetics October 4, 2019. Accepted in ﬁnal form Appendix (continued)

February 6, 2020. Name Location Role Contribution

Markku M. Laboratory Author Analysis of histology and Miettinen, MD, of Pathology, immunohistochemistry; PhD NCI, CCR, revision of the Appendix Authors NIH, manuscript Bethesda, Name Location Role Contribution MD

Jared S. NOB, NCI, Author Conceived the study; Russel H. TVMB, Author Performed the Rosenblum, MD NIH, performed the Knutsen, BA NHLBI, NIH, experiments; data Bethesda, experiments and patient Bethesda, analysis; revision of the MD measurements; data MD manuscript analysis; identification of malformations in the Beth A. Kozel, MD, TVMB, Author Data analysis of the patients and mice; PhD NHLBI, NIH, transgenic mouse drafted the original Bethesda, model; revision of the manuscript MD manuscript

Anthony J. NOB, NCI, Author Performed the Zhenping NOB, NCI; Author Oversight; drafting the Cappadona, BS NIH, experiments and patient Zhuang, MD, PhD SNB, NINDS, manuscript Bethesda, measurements; data NIH, MD analysis; identification of Bethesda, the malformations in the MD patients and mice; revised the manuscript Karel Pacak MD, SMN, NICHD, Author Clinical patient care; PhD, DSc NIH, oversight; drafting the Davis P. SMB, NINDS, Author Patient measurements; Bethesda, manuscript Argersinger, BS NIH, data analysis; MD Bethesda, manuscript the revisions MD John D. Heiss, MD SNB, NINDS, Author Conception of the NIH, project; identification of Ying Pang, MD, SMN, NICHD, Author Performed the Bethesda, the malformations in the PhD NIH, experiments; patient MD patients; drafting the Bethesda data analysis; confirmed manuscript mutation status in the patients

Herui Wang, PhD NOB, NCI, Author Generation of the NIH, transgenic mouse Bethesda, References MD 1. Padget DH. Development of so-called dysraphism; with embryologic evidence of clinical Arnold-Chiari and Dandy-Walker malformations. Johns Hopkins Med J 1972; Matthew A. Internal Author Patient data analysis; 130:127–165. Nazari, MD Medicine manuscript the revision 2. Bano S, Chaudhary S, Yadav S. Congenital Malformation of the Brain, and Neuroimaging—Clinical Applications; Bright P, editor. London: InTech; 2012. Pediatrics, Available at: intechopen.com/books/neuroimaging-clinical-applications/congenital- Georgetown malformations-of-the-brain Accessed January 23, 2019. Hospital, 3. Oldfield EH. Pathogenesis of Chiari I-pathophysiology of syringomyelia: implications Washington for therapy: a summary of 3 decades of clinical research. Neurosurgery 2017;64(CN_ DC suppl_1):66–77. 4. Mar´ın-Padilla M. Cephalic axial skeletal-neural dysraphic disorders: embryology and Jeeva P. MIF, NINDS, Author Mouse model imaging; pathology. Can J Neurol Sci 1991;18:153–169. Munasinghe, PhD NIH, data analysis; revision of 5. Buell TJ, Heiss JD, Oldfield EH. Pathogenesis and cerebrospinal fluid hydrodynamics Bethesda, the manuscript of the Chiari I malformation. Neurosurg Clin N Am 2015;26:495–499. MD 6. Pan KS, Heiss JD, Brown SM, Collins MT, Boyce AM. Chiari I malformation and basilar invagination in fibrous dysplasia: prevalence, mechanisms, and clinical impli- Danielle R. MIF, NINDS, Author Mouse model imaging; cations. J Bone Miner Res 2018;33:1990–1998. Donahue, BS NIH, data analysis; revision of 7. Kniffin CL; McKusick VA. OMIM-online Mendelian Inheritance in Man: Chiari Bethesda, the manuscript Malformation Type, I. 1992. Available at: omim.org/entry/118420#title. Accessed MD January 13, 2019; Updated July 10, 2016. 8. Urbizu A, Khan TN, Ashley-Koch AE. Genetic dissection of Chiari malformation type Abhishek Jha, MD SMN, NICHD, Author Data collection and I using endophenotypes and stratification. J Rare Dis Res Treat 2017;2:35–42. NIH, analysis; manuscript 9. Kao SC, Waziri MH, Smith WL, Sato Y, Yuh WT, Franken EA Jr. MR imaging of the Bethesda, revision craniovertebral junction, cranium, and brain in children with achondroplasia. Am J MD Roentgenol 1989;153:565–569. 10. Rosenblum JS, Maggio D, Pang Y, et al. Chiari malformation type I in EPAS1- James G. Radiology, Author Confirmation of the associated syndrome. Int J Mol Sci 2019;20:E2819. doi: 10.3390/ijms20112819. Smirniotopoulos, George patient malformations; 11. Zhuang Z, Yang C, Lorenzo F, et al. Somatic HIF-2A gain-of-function mutations in MD Washington patent data analysis; paraganglioma with polycythemia. N Engl J Med 2012;367:922–930. University, revision of the 12. Yang C, Sun MG, Matro J, et al. Novel HIF-2A mutations disrupt oxygen sensing, Washington manuscript leading to polycythemia, paragangliomas, and somatostatinomas. Blood 2013;121: DC; National 2563–2566. Library of 13. Semenza GL. Hypoxia-inducible factors in physiology and medicine. Cell 2012;148: Medicine, 399–408. MedPix, 14. Chowdhury R, Leung IKH, Tian YM, et al. Structural basis for oxygen degradation Bethesda, domain selectivity of the HIF prolyl hydroxylases. Nat Commun 2016;7:1–10. MD 15. Kaelin WG Jr, Ratcliffe PJ. Oxygen sensing by metazoans: the central role of the HIF hydroxlase pathway. Mol Cell 2008;30:393–402.

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 9 16. Dunwoodie SL. The role of hypoxia in development of the mammalian embryo. Dev 27. Szpinda M, Baumgart M, Szpinda A, et al. Morphometric study of the T6 vertebra Cell 2009;17:755–773. and its three ossification centers in the human fetus. Surg Radiol Anat 2013;35: 17. Lee SY, Park KH, Yu HG, et al. Controlling hypoxia-inducible factor-2α is critical for 901–916. maintaining bone homeostasis in mice. Bone Res 2019;13:7–14. 28. Wang H, Cui J, Yang C, et al. A transgenic mouse model of Pacak-Zhuang syndrome 18. Därr R, Nambuba J, Del Rivero J, et al. Novel insights into the polycythemia- with Epas1 gain-of-function mutation. Cancers 2019;11:667. paraganglioma-somatostatinoma syndrome. Endocr Relat Cancer 2016;23:899–908. 29. Lakso M, Pichel JG, Gorman JR, et al. Efficient in vivo manipulation of mouse 19. Guvenc G, Sarp AF, Kizmazoglu C, et al. Craniometric analysis of skullbase with genomic sequences at the zygote stage. Proc Natl Acad Sci 1996;93:5860–5865. magnetic resonance imaging in patients with Chiari malformation. J Craniofac Surg 30. Hendrickson NR, Pickhardt PJ, del Rio AM, Rosas HG, Anderson PA. Bone mineral 2019;30:818–822. density T-scores derived from CT attenuation numbers (Hounsfield units): clinical 20. Sayed HR, Jean WC. A novel method to measure the tentorial angle and the impli- utility and correlation with dual-energy X-ray absorptiometry. Iowa Orthop J 2018;38: cations on surgeries of the pineal gland. World Neurosurg 2018;111:e213–e220. 25–31. 21. Bogdanov EI, Heiss JD, Mendelevich EG, Mikhaylov IM, Haass A. Clinical and 31. Raybaud C, Jallo GI. Chapter 2: Chiari I deformity in children: etiopathogenesis and neuroimaging features of “idiopathic” syringomyelia. Neurology 2004;62:791–794. radiologic diagnosis. In: Manto M, Huisman TAGM, editors. Handbook of Clinical 22. Pinter NK, McVige J, Mechtler L. Basilar invagination, basilar impression, and pla- Neurology. Vol 155 (3rd series). Edinburgh: Elsevier; 2018:25–48. tybasia: clinical and imaging aspects. Curr Pain Headache Rep 2016;20:49. 32. Nishikawa M, Sakamoto H, Hakuba A, Nakanishi N, Inoue Y. Pathogenesis of Chiari 23. Halvorson KG, Kellogg RT, Keachie KN, Grant GA, Muh CR, Waldau B. Morpho- malformation: a morphometric study of the posterior cranial fossa. J Neurosurg 1997; metric analysis of predictors of cervical syrinx formation in the setting of Chiari I 86:40–47. malformation. Pediatr Neurosurg 2016;51:137–141. 33. Marin-Padilla M. Cerebral microvessels. In: Pfaff DW, Volkow ND, editors. Neuro- 24. Hardway FA, Holste K, Ozturk G, et al. Sex-dependent posterior fossa anatomical science in the 21st Century. New York: Springer Science; 2016:2–23. differences in trigeminal neuralgia patients with and without neurovascular compression: 34. Altman NR, Naidich TP, Braffman BH. Posterior fossa malformations. AJNR 1992; a volumetric MRI age- and sex-matched case-control study. J Neurosurg 2019;1:1–8. 13:691–724. 25. Yamauchi T, Yamazaki M, Okawa A, et al. Efficacy and reliability of highly functional 35. Kanodia G, Parihar V, Yadav YR, Boatel PR, Sharma D. Morphometric analysis of the open source DICOM software (OsiriX) in spine surgery. J Clin Neurosci 2010;17: posterior fossa and foramen magnum. J Neurosci Rural Pract 2012;3:261–266. 756–759. 36. Bagci AM, Lee SH, Nagornaya N, Green BA, Algerian N. Automated posterior cranial 26. Schreiber JJ, Anderson PA, Wellington KH. Use of computed tomography for fossa volumetry by MRI: applications to Chiari malformation type I. AJNR Am J assessing bone mineral density. Neurosurg Focus 2014;37:E4. Neuroradiol 2013;34:1758–1763.

10 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG ARTICLE OPEN ACCESS Neurologic outcomes in Friedreich ataxia Study of a single-site cohort

Massimo Pandolfo, MD, FAAN Correspondence Dr. Pandolfo Neurol Genet 2020;6:e415. doi:10.1212/NXG.0000000000000415 [email protected] Abstract Objective To investigate the pattern of progression of neurologic impairment in Friedreich ataxia (FRDA) and identify patients with fast disease progression as detected by clinical rating scales.

Methods Clinical, demographic, and genetic data were analyzed from 54 patients with FRDA included at the Brussels site of the European Friedreich’s Ataxia Consortium for Translational Studies, with an average prospective follow-up of 4 years.

Results Afferent ataxia predated other features of FRDA, followed by cerebellar ataxia and pyramidal weakness. The Scale for the Assessment and Rating of Ataxia (SARA) best detected progression in ambulatory patients and in the first 20 years of disease duration but did not effectively capture progression in advanced disease. Dysarthria, sitting, and upper limb coordination items kept worsening after loss of ambulation. Eighty percent of patients needing support to walk lost ambulation within 2 years. Age at onset had a strong influence on progression of neurologic and functional deficits, which was maximal in patients with symptom onset before age 8 years. All these patients became unable to walk by 15 years after onset, significantly earlier than patients with later onset. Progression in the previous 1 or 2 years was not predictive of progression in the subsequent year.

Conclusions The SARA is a sensitive outcome measure in ambulatory patients with FRDA and has an excellent correlation with functional capabilities. Ambulatory patients with onset before age 8 years showed the fastest measurable worsening. Loss of ambulation in high-risk patients is a disease milestone that should be considered as an end point in clinical trials.

From the Service of Neurology, Hopitalˆ Erasme, and Laboratory of Experimental Neurology, Universit´e Libre de Bruxelles (ULB), Brussels, Belgium.

Go to Neurology.org/NG for full disclosures. Funding information is provided at the end of the article.

The Article Processing Charge was funded by the author. This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloading and sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Copyright © 2020 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 Glossary ADL = activities of daily living; ALS = amyotrophic lateral sclerosis; CI = conﬁdence interval; EFACTS = European Friedreich’s Ataxia Consortium for Translational Studies; FARS = Friedreich Ataxia Rating Scale; FRDA = Friedreich ataxia; INAS = Inventory of Non-Ataxic Signs; mFARS = modiﬁed FARS; SARA = Scale for the Assessment and Rating of Ataxia; PRO = patient-reported outcome.

Friedreich ataxia (FRDA) is an autosomal recessive multisystem quantify ataxia in a variety of clinical situations, without disorder characterized by neurologic impairment, hypertrophic attempting to capture the complexity of a specific condition.13 cardiomyopathy, skeletal abnormalities, and carbohydrate in- It has a lower number of items than the mFARS (8 vs 18), so its tolerance. Most patients are homozygous for the hyperexpansion administration is faster, and patients are less fatigued. Training of a guanosine-adenosine-adenosine (GAA) repeat in the first is also easier, a plus in multicentric studies. The Inventory of intron of the Frataxin (FXN) gene,1 which triggers the formation Non-Ataxic Signs (INAS) was developed along with the SARA of repressive chromatin inhibiting FXN messenger RNA tran- to provide a reliable descriptive of neurologic comorbidities in scription.2 Longer repeats lead to more severe repression of fra- patients with ataxia.14 taxin expression, such that most residual FXN in patients with FRDA derives from the allele with the shorter GAA repeat Of the 2 ongoing collaborative prospective natural history (GAA1). Patients with earlier onset and more severe disease studies in FRDA, the European Friedreich’s Ataxia Consortium usually have longer GAA1.3,4 FXN is needed for synthesis of iron- for Translational Studies (EFACTS) uses the SARA, whereas sulfur clusters in mitochondria.5 Its deficiency leads to mito- the Friedreich’s Ataxia–Clinical Outcome Measures Study chondrial dysfunction, oxidative stress, and altered iron metabo- (FA-COMS) in the United States, Canada, and Australia uses lism,6 all potential therapeutic targets. New therapeutic strategies the FARS. The FA-COMS published prospective data on a co- aim to restore FXN levels by upregulating the endogenous gene hort of 812 patients with FRDA, 234 with up to year-5 follow- or by protein or gene replacement therapy. As these approaches up.15 The average annual increase in the mFARS score was ;2 move to clinical development, trial design becomes a critical issue. points. Lower baseline FARS scores predicted faster progression, FRDA is a rare disease, so clinical trials can only enroll a limited whereas the repeat length of GAA1 only showed a marginal number of patients. The need for efficient design is compounded trendwithFARS-basedmeasures.Individualswhowereaged by the number of treatments entering the clinical arena, inevitably <16 years at baseline has the fastest FARS deterioration. The competing for these patients. Two aspects are critical: (1) iden- EFACTS published a cross-sectional analysis of its core cohort tifying the patient population in which disease progression can of 600 patients in 201516 and the 2-year prospective follow-up of best be detected and (2) selecting the most sensitive and robust 471 patients from the same cohort in 2016.17 In the prospective outcome measures, including clinical assessments, patient- study, younger age at onset was associated with faster SARA reported outcomes (PROs), and biomarkers. deterioration, but the effect was minor (−0.02 points per year of age at onset), and the average rate was not different in patients FRDA neuropathology is characterized by marked differences in with typical onset before age 25 years (0.75 points per year, 95% the vulnerability of neuronal systems and in the timing when they confidence interval [CI] 0.62–0.88) and those with later onset become affected.7 Clinically, this translates into different timing and (0.86 points per year, 95% CI, 0.57–1.16). Deterioration in progression rate of proprioceptive,8 cerebellar,9 and pyramidal10 SARA slowed after 24 years of disease duration and was faster signs and symptoms, affecting the sensitivity of rating scales and the with lower SARA score at baseline. choice of appropriate clinical outcomes at different disease stages. Although both the FA-COMS and the EFACTS found that The International Cooperative Ataxia Rating Scale11 was ini- earlier age at onset results in faster worsening, the detected tially used to quantify the severity of FRDA neurologic symp- effects were of relatively limited size. Neither study compared toms in natural history studies and clinical trials, but current progression of different age at onset groups vs disease duration. studies use the Friedreich Ataxia Rating Scale (FARS)8 or Power calculations from both the FA-COMS and the EFACTS the Scale for the Assessment and Rating of Ataxia (SARA).12 were based on average progression rates and concluded that The FARS was conceived as a FRDA-specific scale capturing a 2-year study is needed to detect 50% slowing of disease pro- the various neurologic features of the disease, including sensory gression with a manageable samplesizeofaround100patients. loss, weakness, and amyotrophy, in addition to ataxia. It also includes an assessment of activities of daily living (ADL), The published FA-COMS and EFACTS, while offering a general staging of disease progression, and several quantita- a highly valuable overview of FRDA progression, have 2 im- tive performance measures. More recently, the modified FARS portant limitations. The first is the lack of detail in defining the (mFARS), which dropped nonataxia items, has been accepted pattern of neurologic deterioration, and the second is the lack by the US Food and Drugs Administration as outcome measure of the clear definition of a rapidly progressing patient group in an FRDA clinical trial. The SARA was instead developed to that may provide maximum power for a clinical trial.

2 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG The present study analyzes a single EFACTS site cohort of 54 66.8% of variance), SARA progression is clearly driven by dif- patients with FRDA with the aim of characterizing the pattern ferent items at different stages of FRDA. The gait and the of disease progression and identifying the most rapidly pro- stance items, whose maximum scores account for about 1/3 of gressing subset of patients. Its findings are expected to orient the maximum total SARA score (14/40), are responsible for a more detailed analysis in the overall EFACTS and FA- half of the SARA score progression until patients remain am- COMS cohorts, eventually to be used to improve the design bulatory, i.e., to an overall score of 23–27, when they reach their of future clinical trials in FRDA. maximum (figure 1, A and B). Although all recently diagnosed patients show some gait abnormality, some patients with a total SARA score <10 have a normal stance score, i.e., they can stand Methods in tandem position with eyes open for 10 seconds. The SARA does not test stance with eyes closed, but data from the FA- Between 2010 and 2019, the Brussels EFACTS site has in- 18 COMS and personal observation show that almost all cluded 54 patients in the natural history study. The study was patients even shortly after diagnosis are unable to stand with approved and is monitored by the institution’s ethics com- feet close together and eyes closed, i.e., they have a positive mittee. Patients provided written informed consent for annual Romberg sign. The only SARA item performed without visual clinical assessments, repeated genetic testing, and blood and control, the heel-to-shin slide, is also uniformly abnormal even urine sampling for biomarker studies. Clinical data were in recently diagnosed patients with SARA <10 (figure 1C) and recorded in the EFACTS clinical database; genetic testing was rapidly progresses reaching its maximal score when ambulation performed at the Brussels site, as for the entire EFACTS is lost. The sitting item (figure 1D) is often initially normal, in cohort, confirming the molecular diagnosis of FRDA and some cases up to a total SARA score of 20, and then it pro- determining GAA repeat sizes. Data are available from 263 gresses linearly. It is one of the items contributing to SARA annual visits, corresponding to an average follow-up of almost progression in advanced disease after loss of ambulation, 4 years (range: baseline only to 8 years). Almost all visits were reflecting worsening cerebellar ataxia and truncal weakness in performed by the same examiner (M.P.). advanced disease. Dysarthria and upper limb cerebellar ataxia (speech, finger chase, and nose-to-finger items, figure 1, E–G) Data from the Brussels cohort were extracted from the may also be initially absent and show most progression after EFACTS database as of October 25, 2019. The open source loss of ambulation. The SARA upper limb coordination items software Jamovi (jamovi.org) was used for most statistical assess spatial irregularities, i.e., dysmetria and tremor, which are analyses; the G*Power package was used for power calcu- 19 not prominent in FRDA, so they rapidly move from low to lations. Graphs were drawn using Microsoft Excel or Jamovi. p maximum score when the movement becomes impossible Values are corrected for multiple comparisons when needed. because of very severe ataxia and weakness. The alternating CIs are indicated when appropriate. upper limb movement (diadochokinesia) item becomes se- fi Data availability verely abnormal relatively early ( gure 1H), due to slowing fi more than irregularity of movement. Slowing of repetitive Data tables not allowing patient identi cation are available on fi request. movements is a common nding in patients with FRDA and is considered to be a consequence of pyramidal degeneration.20 Results Effect of disease duration on SARA progression As in the overall EFACTS cohort, disease progression showed Patients considerable slowing after 20–25 years17 (figure 2A), with an The 54 patients were equally distributed by sex (27 men and 27 average yearly SARA progression of 0.92 points in the first 25 women). Age at baseline visit ranged from 7 to 69 years (mean = years and of 0.30 points afterward. For this reason, further 24.9, SD = 13.3). Age at symptom onset ranged from 3 to 60 analyses focused on the first 20 years of disease progression. years (mean = 13.6, SD = 9.7). Seventeen patients had onset before age 8 years, 20 patients between age 8 and 14 years, and 17 Effect of age at symptom onset and of GAA1 on patients at age 15 years or after. Disease duration ranged from 3 to SARA progression 42 years (mean = 16.9, SD = 9.0). Data from the first 20 years of Patients with earlier onset have faster progression. The limited disease progression were available from 15 patients with onset size of the Brussels cohort did not allow a detailed analysis with before age 8 years, 17 patients with onset between age 8 and 14 a small age window, but the effect is very clear just by stratifying years, and 17 patients with onset at age 15 years or after. GAA patients in 3 groups based on age at onset. In the first 20 years repeat sizes were available for 49 patients (mean GAA1 = 649, SD after onset, patients with onset age <8 years (n = 15, 84 yearly = 226), of these 29 had GAA1 > 600 and 20 had GAA1 ≤ 600. visits) progressed most rapidly at an estimated linear rate of 1.90 ± 0.186 (slope ± standard error) SARA points per year vs Items driving SARA progression 1.33 ± 0.157 points for the group with onset between age 8 and Although all SARA items are tightly correlated, as previously 14 years (n = 17, 75 yearly visits) and 0.71 ± 0.168 points for reported13 and confirmed in this study sample by principal those with onset at age ≥15 years (n = 17, 92 yearly visits), all component analysis (single major component explaining differences being statistically significant (figure 2, B and C).

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 3 Figure 1 SARA items vs total SARA score

Scatterplots of scores of each SARA item vs total SARA scores (n = 251), showing fitted Loess regression lines with confidence bands. (A) Gait; (B) stance; (C) heel-to- shin slide; (D) sitting; (E) finger chase; (F) finger-to-nose; (G) speech; (H) alternate hand movements (diadochokinesis). SARA = Scale for the Assessment and Rating of Ataxia.

The GAA1 repeat length also correlated with the SARA triplets (123 yearly visits for GAA1 > 600; 71 yearly visits for progression rate (figure 3A). However, as previously ob- GAA1 ≤ 600). served, GAA1 only accounted for less than half of the variability in age at onset (R2 = 0.345, p < 0.001, figure 3B), so Loss of ambulation its effect was less marked and could only be detected by More rapid progression in patients with earlier onset stratifying patients in 2 groups using a cutoff of 600 GAA translated into shorter time to loss of ambulation (p <

4 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG Figure 2 SARA scores vs disease duration in patients with different onset age

(A) Scatterplot of SARA scores (n = 252) vs disease duration in the Brussels European Friedreich Ataxia Consortium for Translational Study patients (n = 54). Linear regression lines are shown for disease duration <25 years and ≥25 years. (B) Scatterplots of SARA scores vs disease duration in patients with age at onset at <8 years (SARA scores n = 84), between 8 and 14 years (SARA scores n = 75), and ≥15 years (SARA scores n = 92), showing fitted linear regression lines with confidence bands. (C) p Values of comparisons of linear regression slopes between pairs of age at onset groups. AO = age at onset; SARA = Scale for the Assessment and Rating of Ataxia.

0.001). This occurred at an average total SARA score of 25. constant (SARA gait score 6) support was a strong predictor Survival analysis showed that patients with onset age <8 of loss of ambulation within 2 years, occurring in 80% of years reached a SARA score of 25 after a median time of 14 patients. Conversely, no patient with a SARA gait score ≤4 years, and all of them reached this score by 15 years lost the ability to walk in the following 2 years. Of notice, (ﬁgure 4). The median time to SARA 25 was 17 years in a score of 5 was recorded in few patients, indicating that patients with onset age between 8 and 14 years and >20 years occasional support is only needed for a short time before the in those with onset age ≥15 years, with no overlap of CIs need for constant support. The stance item (p < 0.001) and (ﬁgure 4). The need for occasional (SARA gait score 5) or the total SARA score (p = 0.002) showed the best correlation

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 5 have good sensitivity to disease progression in the overall Figure 3 SARA scores vs disease duration in patients with EFACTS.16,17 In the EFACTS, the ADL scale is administered different GAA1 through a structured interview conducted by the investigator, involving the patient and caregiver(s), rather than as a purely PRO as in the FA-COMS, which modifies its psychometric properties. In the Brussels cohort, the ADL and SARA scores were tightly correlated (R2 = 0.82, p < 0.001, figure 5A). Thus, a similar effect of age at onset was observed for ADL progression (figure 5B) as for SARA, although it was less marked, being only detected when comparing patients with onset before age 8 years with those with onset at age 8 years or later, and only marginally significant (p = 0.049).

Effect of previous scores on SARA progression As reported for the overall EFACTS cohort,17 a lower SARA score in the previous visit predicted a faster progression rate. A sharp slowing of progression rate occurred at loss of ambulation (SARA gait item ≥7), with a signiﬁcant diﬀerence in SARA progression between ambulatory and nonambulatory patients when all 1-year changes are compared and regardless of the age at onset group (1.62 points/y vs 0.34 points/y, p = 0.004). As presented above, SARA progression in nonambulatory patients can only be driven by a limited subset of items, limiting the sensitivity of the scale and reducing the signal-to-noise ratio.

Of interest, in the present study, SARA progression in the previous year or 2 years was not predictive of similar progression at the following visit. This ﬁnding reﬂects the expected regression to the mean occurring in repeated assessments of a noisy measure. Therefore, selection of rapidly progressing patients cannot be based on this criterion.

Figure 4 Survival analysis of loss of ambulation in the first 20 years after onset

(A) Scatterplots of SARA scores vs disease duration in patients with GAA1 repeat length >600 (SARA scores n = 123) and ≤600 (SARA scores n = 80), showing fitted linear regression lines with confidence bands. (B) Scatterplot of GAA1 repeat lengths vs age at onset in Brussels European Friedreich Ataxia Consortium for Translational Study patients (n = 51), showing fitted Loess regression line with confidence band. SARA = Scale for the Assessment and Rating of Ataxia.

with the gait item, so they were also good predictors of loss of ambulation.

Loss of ambulation was primarily due to ataxia rather than weakness. Despite the presence of extensor plantar responses at baseline in 50/54 patients (92.5%), weakness of the lower limbs was absent or minimal in ambulatory patients, whereas it rapidly became prominent 1 year after loss of ambulation, with an increase in the total lower limb weakness INAS score from 0.41 ± 0.90 to 3.97 ± 1.90 (p < 0.001).

Kaplan-Meier plot with confidence intervalsoflossofambulationoverthefirst20 ADL score progression years of disease duration in patients with age at onset at <8 years, between 8 and The FARS ADL score is a relevant measure of patients’ 14 years, and ≥15 years. SARA = Scale for the Assessment and Rating of Ataxia. functional capacities in everyday life, previously shown to

6 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG Figure 5 ADL vs SARA scores and disease duration

(A) Scatterplot of ADL vs SARA scores (n = 255) in Brussels European Friedreich Ataxia Consortium for Translational Study patients, showing fitted linear regression line with confidence band. (B) Scatterplots of ADL scores vs disease duration in patients with age at onset at <8 years (ADL scores n = 72) and ≥8 years (ADL scores n = 137), showing fitted linear regression lines with confidence bands. ADL = activities of daily living; SARA = Scale for the Assessment and Rating of Ataxia.

Power calculations Discussion Including ambulatory patients with onset before age 8 years would need a sample size of 97 to detect 50% slowing in SARA Taken together, the results of the present study delineate a pat- progression over 1 year at 90% power by the t test, assuming tern of progression of neurologic impairment in FRDA. They fi ff a 2-sided criterion for detection that allows for a maximum type con rm that a erent ataxia predates cerebellar ataxia and pyra- I error rate of α = 0.05. Just selecting for ambulatory patients midal weakness. Symptom onset corresponds to the appearance (SARA < 25) with a disease duration of <25 years, who proof cerebellar ataxia, affecting gait before stance, speech, and limb gressed an average of 1.60 SARA points/year (CI, 1.19–2.01), coordination. Pyramidal involvement is initially manifest only as would increase the sample size to 193. In comparison, a 1-year extensor plantar responses, then as slowing of rapid alternating trial with unselected patients as calculated in the overall movements, affecting the diadochokinesia SARA item. Pyrami- EFACTS cohort would require 548 patients.17 These calcu- dal weakness eventually becomes prominent, mostly after loss of lations are based on the assumption of identical progression in ambulation. It affects the trunk and limbs, contributing to the a clinical trial vs natural history study and do not keep into progression of the sitting, limb coordination, and possibly speech account the placebo effect shown to occur in FRDA trials,21 so items in patients who are dependent on a wheelchair. they have to be taken as indicative and likely overestimating the needed sample size, particularly if more powerful statistics that The SARA efficiently captures progression up to a score of make use of repeated measures are used. about 25 when ambulation is lost. More sensitive measures of

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 7 neurologic functions that keep worsening in wheelchair- Therefore, there were very few ambulatory patients with very bound patients are needed. These include speech,22 swal- early onset (age <8 years), the vast majority of whom have <13 lowing,23 vision, hearing, and upper limb coordination. The years of disease duration, as indicated by survival analysis of loss Composite Cerebellar Functional Severity score,24 assessing of ambulation (figure 5). upper limb coordination under visual control in 2 timed functional tests, may be more sensitive to progression in Combining all observations, maximal sensitivity to pro- nonambulatory patients. Serious Games, another upper limb gression is found in ambulatory patients with onset before age coordination test reflecting cerebellar function, is also more 8 years, all of whom have <20 years of disease duration. In sensitive in this group.25 Additional upper limb function tests addition to SARA and ADL progression, loss of ambulation using computer-assisted analyses, particularly in the context of can be a relevant outcome for clinical trials in this high-risk ADL, are the object of ongoing studies. patient group. Loss of ambulation is a major milestone in FRDA natural history, whose impact on patients’ lives cannot The tight correlation between the SARA and ADL scores, al- be overstated. It can be predicted to occur within 2 years with ready detected in the overall EFACTS cohort,16,17 was not 80% probability in patients who need even intermittent sup- surprising, as several items measure the same function in both port for walking (SARA gait item score ≥5). Even more rapid scales. However, the ADL scale directly assesses how everyday progression can be expected in the age at onset at <8 years, as life activities are affected, providing a sort of validation of the shown by survival analysis. clinical relevance of the corresponding SARA items. Further- more, the ADL scale includes some functions not scored in the All the observations made in this small patient series need to be SARA, such as swallowing and urinary disturbances. These corroborated and refined in the overall EFACTS and possibly observations support the use of the ADL scale as a coprimary or FA-COMS cohorts. As no site effect was detected in the overall secondary outcome in clinical trials in FRDA, as it is clinically analysis of the EFACTS cohort,16,17 the findings of the present meaningful and sufficiently sensitive to progression. study are likely to predict results in these larger groups. Fur- thermore, although the EFACTS and the FA-COMS use dif- One of the goals of the present study was to tentatively identify ferent rating scales, these are tightly correlated, so overall a subgroup of patients with faster progression that may be analyses can be performed. In this regard, a recent analysis of selected for inclusion in initial therapeutic trials to maximize the the FA-COMS cohort reached very similar conclusions to the sensitivity and minimize the sample size. Unfortunately, the present study, showing that patients with earlier symptom simple assessment of the SARA progression rate in the previous onset also had faster progression and earlier loss of ambulation, 1 or 2 years could not be used for this purpose. Although this which was heralded by inability to stand without support.26 approach is effective in more rapidly progressing diseases such as amyotrophic lateral sclerosis, in FRDA, the noise of the The gained knowledge about FRDA natural history would measure masks any real difference in the progression rate so allow us to greatly improve the follow-up of these patients and that regression to the mean is what mostly predicts the sub- trial design, at a time when multiple highly promising treat- sequent SARA score. The average yearly increase in the SARA ments are likely to come into the clinical arena. score was higher in ambulatory patients, possibly because a ceiling effect is reached for the SARA when gait and stance Author contributions scores, which account for >1/3 of the total score, are maximal. M. Pandolfo: design or conceptualization of the study, clinical Significant slowing of SARA progression also occurred 20–25 evaluation, analysis and interpretation of the data, and writing years after symptom onset, as in the overall EFACTS the manuscript. cohort.16,17 This may be the consequence of the fact that at this time, essentially all patients are wheelchair bound, except those Study funding with particularly slow progression. Age at onset turned out to The EFACTS study has been funded by the European Union be the major determinant of the SARA progression rate in the 7th Framework Programme from 2010 until 2015, by the first 20 years of disease, with those with onset at <8 years charity Euroataxia from 2015 until 2018, and by Voyager showing significantly more rapid worsening. As expected, the Therapeutics from 2015 to present. ADL score also worsened more rapidly in the same group. Of interest, the GAA1 repeat length was less a good predictor of Disclosure the progression rate than age at onset, possibly because the M. Pandolfo has received consulting fees from ApoPharma, latter reflects the combination of all factors that modify disease BioMarin, Minoryx, and Voyager Therapeutics and royalties severity, of which GAA1 is only one. Such a strong effect of from Athena Diagnostics. He serves as Deputy Editor of onset age was not clearly observed in the published overall Neurology: Genetics. Go to Neurology.org/NG for full 16,17 analyses of the EFACTS natural history study. Apossible disclosures. explanation is in the age composition of the core EFACTS cohort, which included almost only adults. In that cohort, the Publication history early-onset (age ≤14 years) group at baseline had an average Received by Neurology: Genetics November 25, 2019. Accepted in final disease duration of 17 years, with a minimum of 10 years.16 form February 5, 2020.

8 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG References 14. Schmitz-Hübsch T, Coudert M, Bauer P, et al. Spinocerebellar ataxia types 1, 2, 3, and – 1. Campuzano V, Montermini L, Moltò MD, et al. Friedreich’s ataxia: autosomal re- 6: disease severity and nonataxia symptoms.. Neurology 2008;71:982 989. cessive disease caused by an intronic GAA triplet repeat expansion. Science 1996;271: 15. Patel M, Isaacs CJ, Seyer L, et al. Progression of Friedreich ataxia: quantitative – 1423–1427. characterization over 5 years. Ann Clin Transl Neur 2016;3:684 694. 2. Gottesfeld JM. Molecular mechanisms and therapeutics for the GAA·TTC expansion 16. Reetz K, Dogan I, Costa AS, et al. Biological and clinical characteristics of the Eu- ’ disease Friedreich ataxia. Neurotherapeutics 2019;16:1–18. ropean Friedreich s Ataxia Consortium for Translational Studies (EFACTS) cohort: – 3. Montermini L, Richter A, Morgan K, et al. Phenotypic variability in Friedreich ataxia: a cross-sectional analysis of baseline data. Lancet Neurol 2015;14:174 182. role of the associated GAA triplet repeat expansion. Ann Neurol 1997;41:675–682. 17. Reetz K, Dogan I, Hilgers RD, et al. Progression characteristics of the European ’ 4. Dürr A, Cossée M, Agid Y, et al. Clinical and genetic abnormalities in patients with Friedreich s Ataxia Consortium for Translational Studies (EFACTS): a 2 year cohort – Friedreich’s ataxia. N Engl J Med 1996;335:1169–1175. study. Lancet Neurol 2016;15:1346 1354. 5. Patra S, Barondeau DP. Mechanism of activation of the human cysteine desulfurase 18. Rummey C, Corben LA, Delatycki MB, et al. Psychometric properties of the Frie- complex by frataxin. Proc Natl Acad Sci U S A 2019;116:19421–19430. dreich Ataxia Rating Scale. Neurol Genet 2019;5:371. 6. Martelli A, Puccio H. Dysregulation of cellular iron metabolism in Friedreich ataxia: 19. Corti M, Casamento-Moran A, Delmas S, et al. Temporal but not spatial dysmetria from primary iron-sulfur cluster deficit to mitochondrial iron accumulation. Front relates to disease severity in FA. J Neurophysiol 2020;123:718–725. Pharmacol 2014;5:130. 20. Marty B, Naeije G, Bourguignon M, et al. Evidence for genetically determined de- 7. Koeppen AH, Mazurkiewicz JE. Friedreich ataxia: neuropathology revised. generation of proprioceptive tracts in Friedreich ataxia. Neurology 2019;93: J Neuropathol Exp Neurol 2013;72:78–90. e116–e124. 8. Subramony S, May W, Lynch D, et al. Measuring Friedreich ataxia: interrater re- 21. Pandolfo M, Arpa J, Delatycki MB, et al. Deferiprone in Friedreich ataxia: a 6-month liability of a neurologic rating scale. Neurology 2005;64:1261–1262. randomized controlled trial. Ann Neurol 2014;76:509–521. 9. Koeppen AH, Davis AN, Morral JA. The cerebellar component of Friedreich’s ataxia. 22. Vogel AP, Wardrop MI, Folker JE, et al. Voice in Friedreich ataxia. J Voice 2017;31: Acta Neuropathol 2011;122:323–330. 243.e9–243.e19. 10. Storey E, Tuck K, Hester R, Hughes A, Churchyard A. Inter-rater reliability of the In- 23. Keage M, Delatycki MB, Dyer J, Corben LA, Vogel AP. Changes detected in swal- ternational Cooperative Ataxia Rating Scale (ICARS). Movement Disord 2004;19:190–192. lowing function in Friedreich ataxia over 12 months. Neuromuscular Disord 2019;29: 11. Trouillas P, Takayanagi T, Hallett M, et al. International Cooperative Ataxia Rating 786–793. Scale for pharmacological assessment of the cerebellar syndrome. The Ataxia Neu- 24. Melac A, Mariotti C, Pierucci A, et al. Friedreich and dominant ataxias: quantitative ropharmacology Committee of the World Federation of Neurology. J Neurol Sci differences in cerebellar dysfunction measurements. J Neurol Neurosurg Psychiatry 1997;145:205–211. 2018;89:559. 12. Vasco G, Gazzellini S, Petrarca M, et al. Functional and gait assessment in children and 25. Bonnechère B, Jansen B, Haack I, et al. Automated functional upper limb evaluation of adolescents affected by Friedreich’s ataxia: a one-year longitudinal study. PLoS One patients with Friedreich ataxia using serious games rehabilitation exercises. 2016;11:e0162463. J Neuroeng Rehabil 2018;15:589. 13. Schmitz-Hübsch T, du Montcel TS, Baliko L, et al. Scale for the assessment and rating 26. Rummey C, Farmer JM, Lynch DR. Predictors of loss of ambulation in Friedreich’s of ataxia: development of a new clinical scale. Neurology 2006;66:1717–1720. ataxia. EClinicalMedicine 2020;18:100213.

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 9 ARTICLE OPEN ACCESS Polygenic risk scores of several subtypes of epilepsies in a founder population

Claudia Moreau, MSc, Rose-Marie Rébillard, MD, Stefan Wolking, MD, Jacques Michaud, MD, Correspondence Frédérique Tremblay, BSc, Alexandre Girard, BSc, Joanie Bouchard, BSc, Berge Minassian, MD, Dr. Girard [email protected] Catherine Laprise, PhD, Patrick Cossette, MD, PhD, and Simon L. Girard, PhD

Neurol Genet 2020;6:e416. doi:10.1212/NXG.0000000000000416 Abstract Objective Polygenic risk scores (PRSs) are used to quantify the cumulative eﬀects of a number of genetic variants, which may individually have a very small eﬀect on susceptibility to a disease; we used PRSs to better understand the genetic contribution to common epilepsy and its subtypes.

Methods We ﬁrst replicated previous single associations using 373 unrelated patients. We then calculated PRSs in the same French Canadian patients with epilepsy divided into 7 epilepsy subtypes and population-based controls. We ﬁtted a logistic mixed model to calculate the variance explained by the PRS using pseudo-R2 statistics.

Results We show that the PRS explains more of the variance in idiopathic generalized epilepsy than in patients with nonacquired focal epilepsy. We also demonstrate that the variance explained is diﬀerent within each epilepsy subtype.

Conclusions Globally, we support the notion that PRSs provide a reliable measure to rightfully estimate the contribution of genetic factors to the pathophysiologic mechanism of epilepsies, but further studies are needed on PRSs before they can be used clinically.

From the Centre Intersectoriel en Santé Durable (C.M., F.T., A.G., J.B., C.L., S.L.G.), UniversitéduQuébec à Chicoutimi, Saguenay; Axe Neurosciences (R.-M.R., S.W., P.C.), Centre de recherche de l’Université de Montréal, Université de Montréal; Centre de recherche du CHU Ste-Justine (J.M.), Université de Montréal, Canada; and Department of Pediatrics and Neurology and Neurotherapeutics (B.M.), UT Southwestern Medical Center, TX.

Go to Neurology.org/NG for full disclosures. Funding information is provided at the end of the article.

The Article Processing Charge was funded by the authors. This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloading and sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Copyright © 2020 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 Glossary CAE = childhood absence epilepsy; FC = French Canadian; IGE = idiopathic generalized epilepsy; ILAE = International League Against Epilepsy; MAF = minor allele frequency; NAFE = nonacquired focal epilepsy; PC = principal component; PRS = polygenic risk score.

In the last decade, many groups have been working on dif- clear clinical focal semiology in the 6 months before starting ferent genetic techniques and statistics to better understand treatment AND (2) an MRI scan of the brain that did not the complex genetic mechanisms underlying epilepsy.1 Last demonstrate any potentially epileptogenic lesion (no lesion) year, a large genome-wide association study on epilepsy OR (3) documented hippocampal sclerosis and lesion other identified 16 loci associated with the disease, and many of than mesial temporal sclerosis (other lesion). these were already known or suspected.2 Despite these efforts, there is still a substantial missing heritability component in For IGE, patients were at least 4 years of age at the diagnosis and epilepsy genetics.3 the IGE subtype (childhood absence epilepsy, juvenile absence epilepsy, juvenile myoclonic epilepsy or IGE not otherwise It is likely that a wide spectrum of genetic factors is in play, specified) was determined according to the 1989 ILAE syn- ranging from very rare mutations with large effects to relatively drome definitions using clinical and EEG characteristics. In IGE, rare variants with medium effect sizes and finally to common we also included patients with epilepsy with eyelid myoclonia variants with smaller risk effects. Polygenic risk scores (PRSs) (Jeavons), which is an idiopathic generalized form of reflex aim to quantify the cumulative effects of a number of variants, epilepsy characterized by childhood-onset, unique seizure which may individually have a very small effect on susceptibility. manifestations, striking light sensitivity, and possible occurrence They have been used previously in many common traits and of generalized tonic-clonic seizures alone. – – diseases such as heart disease4 6 and in neurologic disorders.7 10 Our epilepsy cohort consisted of 643 patients diagnosed with In this study, we aim to use PRSs to see whether this method familial epilepsy. We used 1 patient per family for the analyses can explain more of the epilepsy genetics than the classical for a total of 373 not closely related FC patients, 192 with IGE methods did. We take advantage of the recent meta-analysis and 132 with NAFE (and 49 unclassified epilepsies because of GWAS metrics2 to calculate PRSs in 373 unrelated French a lack of information). We validated that the remaining patients Canadian (FC) patients with epilepsy divided into 7 subtypes. The population is known for its well-documented recent (400 ff years) founder e ect and its particular genetic background, Table 1 Basic statistics for different epilepsy subtypes which makes it an ideal population for genetic studies. French Canadians are also closely related to the European population, Epilepsy subtype n Women, n Mean age 2 which is predominant in the GWAS. All epilepsiesa 373 213 46

IGEa 192 114 43 Methods CAE 34 18 35 JAE 18 10 44 Standard protocol approvals, registrations, and patient consents JME 81 52 48 This study was approved by the Centre de Recherche du GTCS 24 13 41

Centre Hospitalier Universitaire de Montr´eal ethics commit- NAFEa 132 70 48 tee, and written informed consent was obtained for all patients. NAFE HS 22 10 55

Phenotyping of patients NAFE no lesion 71 40 46 The epilepsy cohort was composed of families with at least 3 aﬀected individuals with idiopathic generalized epilepsy (IGE) or NAFE other lesion 13 10 51

nonacquired focal epilepsy (NAFE) previously collected and di- Abbreviations: CAE = childhood absence epilepsy; GTCS = generalized tonic- agnosed by neurologists. The clinical epilepsy phenotype is defined clonic seizures alone; HS = documented hippocampal sclerosis; IGE = idiopathic fi generalized epilepsy; JAE = juvenile absence epilepsy; JME = juvenile myoclonic based on the Classi cation of the Epilepsy Syndromes established epilepsy, NAFE = nonacquired focal epilepsy; NAFE HS = nonacquired focal by the International League Against Epilepsy (ILAE).11 epilepsy documented hippocampal sclerosis; NAFE no lesion = nonacquired focal epilepsy no documented epileptogenic lesion; NAFE other lesion = nonacquired focal epilepsy lesion other than mesial temporal sclerosis. fi fi a The difference between the total numbers of patients and of IGE and NAFE More speci cally, the operational de nitions of the epilepsy broad subtypes and the sum of their respective subtypes comes from the phenotypes studied in the project for NAFE were as follows: patients who could not be classified into any of the narrow subtypes because of a lack of clinical information (49 unclassified patients, 35 un- (1) patients were aged at least 5 years and had experienced at classified IGE, and 26 unclassified NAFE). least 2 unprovoked seizures with focal EEG abnormalities or

2 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG were not closely related (first cousins or more related) using software.13 We then removed 121 individuals of non-FC PLINK. FC ancestry was assessed by self-declared ethnicity and descent using the first 2 principal components in addition to principal component (PC) analysis (figure e-1, links.lww.com/ self-identification of patients whenever this information was NXG/A253). Table 1 shows the different subtypes of epilepsy available. Principal component analysis (figure e-1, links. that are represented in our cohort. In addition, we selected 954 lww.com/NXG/A253) was performed using Eigensoft14 on FC individuals from a reference population data set.12 pruned SNPs (pairwise r2 < 0.2 in sliding windows of size 50 shifting every 5 SNPs) at 5% minor allele frequency (MAF). Data availability We finally aligned the data set to the GRCh37 genome build The patients’ genotype data used in the present study will be for further imputation following the method described available on request. here.15

Genotyping and imputation The Sanger Imputation Service was used to conduct For this study, we used whole-genome genotyping data for whole-genome imputation of SNPs.16 We selected the 12 the patient and the French Canadian control cohorts. All Human Reference Consortium data set as the reference samples were processed on either the Illumina Omni Express panel. Postimputation quality control filters were applied (number of single nucleotide polymorphisms = 710,000) or to remove SNPs within imputed data with an imputation − the Illumina Omni 2.5 (number of single nucleotide poly- info score of <0.9 or HWE p value of <1 e 6, and only morphisms = 2,500,000 including the Omni Express core) biallelic SNPs at MAF 1% or higher were kept for further depending on the availability of the chip regardless if they analyses. were controls or patients. Genotypes of all samples were merged, and only positions present on both chips were kept. Association analysis We performed cleaning steps to remove individuals having We used PLINK software for the logistic association analysis more than 2% missing genotypes among all SNPs, SNPs with the first 10 PCs and sex as covariates. Associations were with more than 2% missing SNPs over all individuals, and only tested for the 20 SNPs found significant in the ILAE study SNPs with Hardy-Weinberg p value <0.001 using PLINK and only in the epilepsy subtypes in which they were originally

Figure 1 PRS density for broad epilepsy types

PRS density plots for French Canadian controls and (A) all patients with epilepsy, (B) patients with IGE, and (C) patients with NAFE. IGE = idiopathic generalized epilepsy; NAFE = nonacquired focal epilepsy; PRS = polygenic risk score.

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 3 Figure 2 PRS density for IGE subtypes

PRS density plots for French Canadian controls and patients with IGE syndrome (A) CAE, (B) GTCS, (C) JME, and (D) JAE. CAE = childhood absence epilepsy; GTCS = generalized tonic-clonic seizures alone; IGE = idiopathic generalized epilepsy; JAE = juvenile absence epilepsy; JME = juvenile myoclonic epilepsy; PRS = polygenic risk score.

reported. We used a p value threshold of 0.0025 to account for Statistical analyses multiple testing (n = 20). We fitted a logistic regression mixed model using R. We then calculated the Nagelkerke pseudo-R2 (using the PRS at the p PRS calculation value threshold that best predicts the phenotype) with and 17 PRSs were calculated with PRSice software using ILAE meta- without the PRS as the full and null model. Note that pseudo- 2 analysis on epilepsy summary statistics. Because the BETA R2 is reported on the observed scale to avoid overfitting. was not provided for the METAL analyses (all, generalized, and focal epilepsies’ analyses), we used the formula from reference 18 to calculate it. We used the first 10 PCs in addition to sex as Results covariates, recalculating eigenvectors for each patient subset including controls using SNPs at MAF 0.05 pruned (as de- We used whole-genome genotyping on 373 unrelated patients scribed above). PRSs were standardized for graphs. having epilepsy and 954 population controls. All individuals

4 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG Figure 3 PRS density for NAFE subtypes

PRS density plots for French Canadian controls and patients with NAFE with (A) HS, (B) no documented lesion, and (C) lesions other than HS. HS = documented hippocampal sclerosis; NAFE = nonacquired focal epilepsy; PRS = polygenic risk score. were confirmed with French Canadian ancestry. FC control and 3 show the same analysis for the IGE and NAFE subtypes. individuals used in this study have already been demon- The best-fits are shown in figures e-3 and e-4 (links.lww.com/ strated to cluster with Western Europeans.19 First, we NXG/A253). wanted to assess whether the associations found by the ILAE study2 were valid for our cohort. Table e-1, links.lww.com/ The next logical question was to investigate whether the NXG/A253, presents the statistics of the association analy- PRS could be used to discriminate between a patient with sis. One locus was found to be significant, and 3 were close to epilepsy and a control. Table 2 presents the logistic mixed significance (p value threshold = 0.0025). These results show model statistics and the variance explained by the PRS that our founder FC population shares a portion of the ep- calculated using the Nagelkerke pseudo-R2 (on the ob- ilepsy genetic risks with the Western European populations served scale) for patients and controls for all epilepsy sub- studied by the ILAE. types. The variance explained by the PRS varies among epilepsy subtypes, but is generally higher for IGE types than Next, to assess whether SNPs taken together could explain forNAFEtypes.Thisisreflected by the higher odds ratio a portion the epilepsy phenotype, we used the basic statistics and variance explained by the PRS in IGE broad subtype of the ILAE study2 to construct the PRS. Figure 1 shows the compared with NAFE and corroborates what was found in density plots of standardized PRS values of patients compared arecentstudy.20 with controls for the 3 broad epilepsy types; best-fit p values are shown in figure e-2 (links.lww.com/NXG/A253). Our Discussion first observation was that the PRS distribution is more shifted to the right in IGE than in NAFE, which is consistent with the The strongest association in our FC cohort was observed with heritability estimates reported in the ILAE study.2 Figures 2 the SNP rs1402398. This SNP is located in the noncoding

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 5 Table 2 PRS and pseudo-R2 statistics based on the logistic mixed model

Epilepsy subtype PRS p value threshold No. of SNPs OR (95% CI) p Value (Wald) Pseudo-R2

All epilepsies 0.1606 45,634 1.12 (1.07–1.17) <0.001 0.031

IGE 0.06265 24,293 1.24 (1.17–1.32) <0.001 0.087

CAE 0.079 25,616 4.85 (2.31–10.19) <0.001 0.094

JAE 0.0044 2,705 74.88 (2.82–1,990.77) 0.01 0.042

JME 0.0041 2,941 8.47 (3.67–19.58) <0.001 0.057

GTCS 0.046 15,624 4.41 (0.69–28.31) 0.118 0.012

NAFE 0.09695 33,200 1.12 (1.05–1.19) 0.001 0.019

NAFE HS 1 95,225 1.82 (0.92–3.58) 0.084 0.016

NAFE no lesion 0.00065 686 2.56 (0.68–9.74) 0.167 0.0047

NAFE other lesion 0.0007 756 43.36 (2.86–658.04) 0.007 0.058

Abbreviations: CAE = childhood absence epilepsy; GTCS = generalized tonic-clonic seizures alone; HS = documented hippocampal sclerosis; IGE = idiopathic generalized epilepsy; JAE = juvenile absence epilepsy; JME = juvenile myoclonic epilepsy; NAFE = nonacquired focal epilepsy; NAFE HS = nonacquired focal epilepsy documented hippocampal sclerosis; NAFE no lesion = nonacquired focal epilepsy no documented epileptogenic lesion; NAFE other lesion = nonacquired focal epilepsy lesion other than mesial temporal sclerosis; PRS = polygenic risk score.

region surrounding genes FANCL and BCL11A. These genes This study was conducted on a documented founder pop- have been linked with epilepsies through association studies,21 ulation. The FC population is well known for its high prevalence but no other functional or clinical evidence highlights their of specific disease-causing mutations.22,23 For epilepsy, although roles in the disease. we cannot exclude that some of the associations found were driven by rare haplotypes, we show here that the genetic eti- Although we successfully replicated associations, we believe ology of the disease is consistent with that of the general Eu- that the biggest contribution of our study lies in the PRSs ropean population. In future work, we will try to assess whether that were established for each epilepsy type. This is, to our the strong PRS found in some epilepsy subtypes could be knowledge, one of the first documented examples of how explained by rarer haplotypes, as we would expect in a founder PRSs can be used for epilepsy genetic studies for different population. subtypes of epilepsy, although a recent study has shown that for broad epilepsy subtypes.20 Although this measure cannot Globally, we support the notion that PRSs provide a reli- yet be translated into clinical use, our analysis shows that the able measure to rightfully estimate the contribution of ge- additive value of common variants can be used to better netic factors to the pathophysiologic mechanism of understand the disease. epilepsies.

One definite pitfall of our study is the small size of our Acknowledgment cohort. The initial GWAS was performed on more than The authors are thankful to Compute Canada/Calcul 15,000 patients with epilepsy. Our study only included 373 Québec for the access to storage and computing resources. patients with epilepsy and thus cannot have the same out- They thank Alexandre Bureau for his useful expertise in reach as the initial one. This is why we did not report biostatistics. They also thank Editage (editage.com) for genome-wide association statistics and focused only on the English language editing. They are extremely grateful to all replication of associated SNPs. We believe that the small size patients and their families for participating in this research. of our cohort also affects the PRS calculations, but to They thank Damian Labuda and Hélène Vézina for their a smaller degree. work on the QRS control cohort.

For these reasons, we have to take the variance explained by the Study funding PRS with caution. However, for the broad phenotypes, we This work was supported by funding from Genome Quebec/ explain 4 times more of the variance for patients with IGE than Genome Canada and from the CIHR (#420021). what we explain for patients with NAFE, as shown elsewhere.20 This also supports the fact that epilepsy should be divided into Disclosure subtypes when studying the genetic mechanism underlying the C. Moreau, R.-M. R´ebillard, S. Wolking, J. Michaud, disease, as some epilepsy types were reasonably well explained F. Tremblay, A. Girard, J. Bouchard, B. Minassian, C. Laprise, by the PRS (i.e., childhood absence epilepsy). P. Cossette, and S.L. Girard report no disclosure. This study

6 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG was not industry sponsored. Go to Neurology.org/NG for full disclosures. Appendix (continued)

Name Location Contribution Publication history The manuscript was previously published in bioRxiv (doi.org/ Simon L. Centre Intersectoriel en Designed and conceptualized Girard, PhD Santé Durable, Université the study, supervised the 10.1101/728816). Received by Neurology: Genetics du Québec à Chicoutimi, analysis of the data, and November 7, 2019. Accepted in final form February 13, 2020. Saguenay finalized the manuscript

References 1. Koeleman BPC. What do genetic studies tell us about the heritable basis of common Appendix Authors epilepsy? Polygenic or complex epilepsy? Neurosci Lett 2018;667:10–16. 2. The International League Against Epilepsy Consortium on Complex Epilepsies. Name Location Contribution Genome-wide mega-analysis identifies 16 loci and highlights diverse biological mechanisms in the common epilepsies. Nat Commun 2018;9:5269. Claudia Centre Intersectoriel en Designed and 3. Thomas RH, Berkovic SF. The hidden genetics of epilepsy-a clinically important new Moreau, Santé Durable, Université conceptualized the study, paradigm. Nat Rev Neurol 2014;10:283–292. MSc du Québec à Chicoutimi, analyzed the data, and 4. Khera AV, Chaffin M, Aragam KG, et al. Genome-wide polygenic scores for common Saguenay drafted the manuscript for diseases identify individuals with risk equivalent to monogenic mutations. Nat Genet intellectual 2018;50:1219–1224. content 5. Wünnemann F, Lo KS, Langford-Avelar A, et al. Validation of genome-wide polygenic risk scores for coronary artery disease in French Canadians. Circ Genomic Precis Med ’ Rose-Marie Axe Neurosciences, Centre Determined patients 2019;12:243–248. ’ Rébillard, de recherche de l Université epilepsy subtype 6. Inouye M, Abraham G, Nelson CP, et al. Genomic risk prediction of coronary artery MD de Montréal, Universitéde disease in 480,000 adults: implications for primary prevention. J Am Coll Cardiol Montréal, Canada 2018;72:1883–1893. 7. Fullerton JM, Koller DL, Edenberg HJ, et al. Assessment of first and second degree relatives of Stefan Axe Neurosciences, Centre Determined patients’ individuals with bipolar disorder shows increased genetic risk scores in both affected relatives Wolking, de recherche de l’Université epilepsy subtype and young at-risk Individuals. Am J Med Genet B Neuropsychiatr Genet 2015;168:617–629. MD de Montréal, Universitéde 8. Dima D, de Jong S, Breen G, Frangou S. The polygenic risk for bipolar disorder Montréal, Canada influences brain regional function relating to visual and default state processing of emotional information. Neuroimage Clin 2016;12:838–844. Jacques Centre de recherche du Collected and genotyped 9. Ahn K, An SS, Shugart YY, Rapoport JL. Common polygenic variation and risk for Michaud, CHU Ste-Justine, Université patients with epilepsy childhood-onset schizophrenia. Mol Psychiatry 2016;21:94–96. MD de Montréal, Canada 10. Boies S, Mérette C, Paccalet T, Maziade M, Bureau A. Polygenic risk scores distin- guish patients from non-affected adult relatives and from normal controls in schizo- Frédérique Centre Intersectoriel en Interpreted the data and phrenia and bipolar disorder multi-affected kindreds. Am J Med Genet Part B Tremblay Santé Durable, Université revised the manuscript for Neuropsychiatr Genet 2018;177:329–336. du Québec à Chicoutimi, intellectual content 11. Berg AT, Berkovic SF, Brodie MJ, et al. Revised terminology and concepts for or- Saguenay ganization of seizures and epilepsies: report of the ILAE Commission on Classifica- tion and Terminology, 2005-2009. Epilepsia 2010;51:676–685. Alexandre Centre Intersectoriel en Interpreted the data and 12. Quebec Reference Sample. Website [online]. Available at: www.quebecgenpop.ca/ Girard Santé Durable, Université revised the manuscript for home.html. Accessed August 6, 2019. du Québec à Chicoutimi, intellectual content 13. Purcell S, Neale B, Todd-Brown K, et al. PLINK: a tool set for whole-genome association Saguenay and population-based linkage analyses. Am J Hum Genet 2007;81:559–575. 14. Price AL, Patterson NJ, Plenge RM, Weinblatt ME, Shadick NA, Reich D. Principal Joanie Centre Intersectoriel en Interpreted the data and components analysis corrects for stratification in genome-wide association studies. Bouchard Santé Durable, Université revised the manuscript for Nat Genet 2006;38:304–309. du Québec à Chicoutimi, intellectual content 15. McCarthy Group Tools. Website [online]. Available at: www.well.ox.ac.uk/;wrayner/ Saguenay tools/. Accessed June 14, 2019. 16. Sanger Imputation Service. Website [online]. Available at: imputation.sanger.ac.uk/). Berge Department of Pediatrics Collected and genotyped Accessed March 18, 2019. Minassian, and Neurology and patients with epilepsy 17. Euesden J, Lewis CM, O’Reilly PF. PRSice: polygenic risk score software. Bio- MD Neurotherapeutics, UT informatics 2015;31:1466–1468. Southwestern Medical 18. Zhu Z, Zhang F, Hu H, et al. Integration of summary data from GWAS and eQTL Center, TX studies predicts complex trait gene targets. Nat Genet 2016;48:481–487. 19. Roy-Gagnon MH, Moreau C, Bherer C, et al. Genomic and genealogical investigation of Catherine Centre Intersectoriel en Contributed in the control the French Canadian founder population structure. Hum Genet 2011;129:521–531. Laprise, Santé Durable, Université group 20. Leu C, Stevelink R, Smith AW, et al. Polygenic burden in focal and generalized PhD du Québec à Chicoutimi, epilepsies. Brain 2019;142:3473–3481. Saguenay 21. The International League Against Epilepsy Consortium on Complex Epilepsies. Genetic determinants of common epilepsies: a meta-analysis of genome-wide asso- Patrick Axe Neurosciences, Centre Collected and genotyped ciation studies. Lancet Neurol 2014;13:893–903. Cossette, de recherche de l’Université patients with epilepsy 22. Scriver CR. Human Genetics : lessons from Quebec populations. Annu Rev MD, PhD de Montréal, Universitéde Genomics Hum Genet 2001;2:69–101. Montréal, Canada 23. Laberge AM, Michaud J, Richter A, et al. Population history and its impact on medical genetics in Quebec. Clin Genet 2005;58:287–301.

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 7 ARTICLE OPEN ACCESS Clinical and pathologic phenotype of a large family with heterozygous STUB1 mutation

Merel O. Mol, MD, Jeroen G.J. van Rooij, MSc, Esther Brusse, MD, PhD, Annemieke J.M.H. Verkerk, PhD, Correspondence Shamiram Melhem, BSc, Wilfred F.A. den Dunnen, MD, PhD, Patrizia Rizzu, MD, PhD, Chiara Cupidi, MD, PhD, Dr. Mol [email protected] John C. van Swieten, MD, PhD, and Laura Donker Kaat, MD, PhD

Neurol Genet 2020;6:e417. doi:10.1212/NXG.0000000000000417 Abstract Objective To describe the clinical and pathologic features of a novel pedigree with heterozygous STUB1 mutation causing SCA48.

Methods We report a large pedigree of Dutch decent. Clinical and pathologic data were reviewed, and genetic analyses (whole-exome sequencing, whole-genome sequencing, and linkage analysis) were performed on multiple family members.

Results Patients presented with adult-onset gait disturbance (ataxia or parkinsonism), combined with prominent cognitive decline and behavioral changes. Whole-exome sequencing identiﬁed a novel heterozygous frameshift variant c.731_732delGC (p.C244Yfs*24) in STUB1 segregating with the disease. This variant was present in a linkage peak on chromosome 16p13.3. Neuropathologic examination of 3 cases revealed a consistent pattern of ubiquitin/p62-positive neuronal inclusions in the cerebellum, neocortex, and brainstem. In addition, tau pathology was present in 1 case.

Conclusions This study conﬁrms previous ﬁndings of heterozygous STUB1 mutations as the cause of SCA48 and highlights its prominent cognitive involvement, besides cerebellar ataxia and movement disorders as cardinal features. The presence of intranuclear inclusions is a pathologic hallmark of the disease. Future studies will provide more insight into its pathologic heterogeneity.

From the Department of Neurology (M.O.M., J.G.J.v.R., E.B., S.M., J.C.v.S., L.D.K.); Department of Internal Medicine (J.G.J.v.R., A.J.M.H.V.), Erasmus Medical Center, Rotterdam; Department of Pathology and Medical Biology (W.F.A.d.D.), University Medical Centre Groningen, Groningen, the Netherlands; German Center for Neurodegenerative Diseases (DZNE) (P.R.), Tuebingen, Germany; IRCCS Centro Neurolesi “Bonino Pulejo” (C.C), Messina, Italy; and Department of Clinical Genetics (L.D.K.), Erasmus Medical Center, Rotterdam, the Netherlands.

Go to Neurology.org/NG for full disclosures. Funding information is provided at the end of the article.

Copyright © 2020 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 Glossary AD = Alzheimer disease; CHIP = C-terminus of Hsp70-interacting protein; CI = cytoplasmic inclusion; DNS =diﬀuse nuclear staining; FTD = frontotemporal dementia; GATK = Genome Analysis Toolkit; NII = neuronal intranuclear inclusion; PSP = progressive supranuclear palsy; SCA = spinocerebellar ataxia; STR = short tandem repeat; WES = whole-exome sequencing; WGS = whole-genome sequencing.

Spinocerebellar ataxias (SCAs) comprise a heterogeneous Standard protocol approvals, registrations, group of disorders characterized by progressive cerebellar and patient consents ataxia in combination with noncerebellar signs, including The study was approved by the Medical Ethical Committee of (extra) pyramidal features, peripheral neuropathy, and cog- the Erasmus Medical Center Rotterdam. Brain autopsy was nitive impairment.1 In most cases, prominent loss of cere- conducted by the Netherlands Brain Bank according to their bellar Purkinje neurons is observed, yet neuropathologic Legal and Ethical Code of Conduct. Written informed con- changes can be diverse with degeneration extending to all sent was obtained from all participants and/or their legal brain areas. Frequently, intranuclear neuronal inclusions serve representatives. as a morphologic marker.2 Genetic studies Currently, over 40 autosomal dominant SCAs have been identified.1 Polyglutamine SCAs caused by trinucleotide CAG Linkage analysis repeat expansion represent the most common form. Other Single nucleotide polymorphism (SNP) array genotyping genetic causes include repeat expansion in noncoding regions (Illumina Human OmniExpress-24 v1.0 BeadChip) was and conventional mutations.3 A large percentage of patients performed on 7 patients (III-8, III-9, III-10, III-11, III-12, III- with familial SCA (30%–48%) remain without an identified 13, and IV-1). SNP call data were adapted by GenomeStudio genetic defect.4 In the past decade, an increasing number of (Illumina) for linkage analyses using Allegro, implemented in 13,14 ff rare causal variants have been recognized in both dominant easyLINKAGE Plus. An a ected-only linkage analysis was 5 performed. Mendelian inheritance check was performed with and recessive forms of ataxia. Among these are variants in 15 STUB1 (STIP1 homologous and U-box-containing protein 1; the program PedCheck. SNPs showing mendelian incon- OMIM#607207), the gene encoding the protein C-terminus sistencies were excluded from the calculations. Multipoint of Hsp70-interacting protein (CHIP). This molecular co- linkage analysis was performed with an SNP spacing of 0.3 chaperone plays a prominent role in protein quality control cM. Logarithm of the odds (LOD) scores were calculated processes and the cellular stress response.6 Recessive STUB1 under the assumption of an autosomal dominant disorder. mutations were identified in families with SCAR16, showing Whole-exome and genome sequencing a wide variability of symptoms, including hypogonadism, Whole-exome sequencing (WES) was performed on the epilepsy, autonomic dysfunction, and dementia.7,8 More re- same 7 patients as those included in the linkage analysis. cently, heterozygous variants in STUB1 have been implicated Whole-genome sequencing (WGS) was performed on 3 of in autosomal dominant SCA (SCA48; OMIM#618093), these patients (III-12, III-13, and IV-1). Genomic DNA was characterized by a complex phenotype including cognitive/ – fragmented to 150–200 base pairs (bp), end paired, ade- affective symptoms and movement disorders.9 12 nylated, and ligated to adapters. The SeqCap capturing kit for Illumina Paired-End Sequencing Library (version 2.0.1; In this study, we present a large Dutch SCA48 family asso- NimbleGen) was used. The captured fragments were puri- ciated with a novel heterozygous frameshift mutation in the fied and sequenced on an Illumina Hiseq2000 platform U-box domain of STUB1 and describe the neuropathologic using 100 bp paired-end reads. WGS was performed by features of 3 patients in this family. Novogene on an Illumina Hiseq2000 platform using 150 bp paired-end reads. Bioinformatic analysis was performed Methods using an in-house pipeline based on published tools. Briefly, sequence reads were aligned to the human reference ge- Clinical data nome (hg19) using Burrows-Wheeler Aligner (version Medical records and neuroimaging of 9 affected individuals 0.7.3a).16 Subsequently, alignment data were processed and from 1 large family of Dutch ancestry were collected and recalibrated using Picard (version 1.90) and the Genome reviewed (figure 1 and table 1). Seven of them underwent Analysis Toolkit (GATK version 2.5.2). SNPs and small a single clinical assessment by neurologists from the re- indels were called using GATK’s HaplotypeCaller and search group (E.B., J.C.v.S., and L.D.K.) during a local visit. Variant recalibration following best practices.17 The WES Blood samples were collected from 7 affected and 5 un- and WGS data sets were generated separately. The lists of affected family members; DNA was extracted using standard variants from the WES data were annotated with Annovar18 procedures. and filtered to include (1) exonic and splice site variants, (2)

2 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG Figure 1 Pedigree of the family

Filled black symbols represent patients who were personally examined by the researchers. Filled gray symbols are affected individuals based on medical records (III-2 and III-3) or on reports from family members (II-2, III-1, III-5, and III-6). Individuals III-7, III-14, III-15, III-16, and III-17 were clinically unaffected and all deceased >75 years of age. Individual II-3 is an obligate carrier but did not have neurologic symptoms according to the family. Individual III-4 was diagnosed with Alzheimer disease without motor symptoms and without cerebellar atrophy and not considered to have the same phenotype. The numbers inside the symbols represent the number of individuals. Sex is not specified to protect anonymity. Transmission was independent of sex (including male to male transmission). + = mutated allele; – = wild type. nonsynonymous, (3) with rare occurrence in publicly Clone, Santa Cruz, 1:100), 1C2 (mouse 5TF1-1C2-172 available databases (1000G, Exome Aggregation Consor- Clone, Chemicon, 1:3,200), and STUB1 (rabbit anti- tium and Exome Sequencing Project; minor allele frequency STUB1 Abcam ab2917, 1:100). <0.1%), and (4) QD score >5 (quality score normalized by allele depth). Functional predictions by Sorting Intolerant Data availability From Tolerant and/or PolyPhen-2 were taken into account. The deidentified data generated and analyzed in the current In addition, to detect the possible presence of expanded study will be available and shared by request to the corre- polyglutamine repeats, short tandem repeats (STRs) were sponding author (M.O.M.) from any qualified investigator. called using WGS data in the chr16p13.3 linkage peak using These data include deidentified clinical data, raw exome/ LobSTR (v3.0.3) according to best practices.19 STRs were genome files, microsatellite genotyping, and additional pic- filtered for heterozygosity in all 3 carriers. The repeat motif tures of stainings from the studied material. and number of copies per carrier were extracted and analyzed manually. RefSeq gene annotation was added using Annovar (ref; 20601685). The candidate variant was Results checked by Sanger sequencing (details provided in sup- Clinical characteristics plementary file, links.lww.com/NXG/A251). The pedigree structure is presented in figure 1, and the main clinical features of 9 affected individuals are summarized in Neuropathology table 1. Individuals II-2, II-3, III-1, III-2, III-3, III-5, and III-6 Three patients died during follow-up (III-10, III-12, and III- were deceased before the start of the study, and DNA was not 13); brain autopsy was performed by the Netherlands Brain available. Patient III-4 was not included in the genetic analyses Bank within 4 hours after death. Formalin-fixed (10%) and (linkage and WES) because the clinical signs did not fully paraffin-embedded tissue blocks were available for exami- match with the other family members (i.e., isolated cognitive nation. Eight-micrometer sections of all cortical regions, dysfunction suggestive of Alzheimer disease [AD] without subcortical nuclei, brainstem, and cerebellum underwent ataxia or movement disorders). Mean age at onset of disease routine staining. Immunohistochemistry was performed in the 9 patients was 65.3 ± 6.6 years. These patients deceased using the following antibodies: hyperphosphorylated tau after a mean disease duration of 13.2 ± 2.4 years. The pre- (AT8, Innogenetics, Ghent, Belgium; 1:40), beta-amyloid senting features consisted of slowly progressive gait distur- (anti-amyloid, DAKO, Glostrup, Denmark; 1:100, following bance (mainly ataxic and frequently described as a waddled formic acid pretreatment), alpha-synuclein (anti-synuclein, gait) combined with prominent cognitive dysfunction and Zymed Laboratories, San Francisco, CA; undiluted, follow- behavioral changes with impaired insight, disinhibition, and ing formic acid pretreatment), TDP-43 (anti-phospho TDP- perseverations. The cognitive deficits were among the first 43, Cosmo Bio, 1:100 and Proteintech Group, 1:100), symptoms in some patients, showing a gradual decline in ubiquitin (anti-ubiquitin, DAKO, Glostrup, Denmark; 1: memory performance and executive functioning. Imaging in 500, following 80°C antigen retrieval), p62 (mouse D3 6 of 9 patients showed cerebellar atrophy; 2 had moderate

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 3 Table Summary of demographic and clinical data of affected individuals

Clinical features Initial Patient AAO DD AAD CI D G UL SP GP P Ch Imaging diagnosis Cognitive/behavioral symptoms

III-2 71 13 84 ± ? + ? ? ? − + NA Huntington Social withdrawal, perseverance

III-3 66 14 80 + + + + ? ? ? + CAa Huntington Intellectual disability since birth

III-8 72 14 86 + + ± − + −−−CA and AD Memory deficits, impaired language GAb comprehension, disorientation, perseverance, and irritability

III-9 69 14 83 + + + − + −−−CAa OPCA Memory deficits, apraxia, disorientation, perseverance, and mental speed decline

III-10c 65 18 83 + + −− ++ +− CA and PSP Memory deficits, impulsiveness, and GAb disinhibition; MMSE 24/30, FAB 7/18

III-11 67 11 78 + + ± ± −− −− GAa Dementia Perseverance, apraxia, executive NOS dysfunction, disinhibition, and lack of insight

III-12c 61 10 71 + + ± −−+±±CAa SCA, Memory deficits, passivity, impaired dementia language comprehension, and lack of insight

III-13c 50 14 64 + + + −−+ − ±CAb FTD Childish behavior, inappropriate laughing, disinhibition, perseverance, and aggressiveness

IV-1 67 11 78 + ± + ± + + ± ± GAa Dementia Memory deficits, impulsiveness, NOS executive dysfunction, and lack of insight; MMSE 20/30, FAB 5/18

Abbreviations: AAO = age at first symptom onset; AAD = age at death; DD = disease duration; FAB = frontal assessment battery; FTD = frontotemporal dementia; MMSE = Mini-Mental State Examination. Clinical symptoms: CI = cognitive impairment; Ch = chorea/motoric restlessness; D = dysarthria; G = gait ataxia; GP = gaze palsy; P = parkinsonism; SP = saccadic pursuit; UL = upper limb ataxia; − = absent; ± = subtle; + = present; ? = unknown. Imaging: CA = cerebellar atrophy; GA = generalized atrophy; NA = not available. Diagnosis: AD = Alzheimer disease; Dementia NOS = dementia not otherwise specified; OPCA = olivopontocerebellar atrophy; PSP = progressive supranuclear palsy; SCA = spinocerebellar ataxia. a CT imaging. b MR imaging. c Pathologic examination.

to severe generalized atrophy (for 1 patient, no scan was pronounced frontal symptoms and semantic deficits, but with available). Initial clinical diagnoses by reviewing medical preservation of episodic memory. On MRI, severe cerebellar records were olivopontocerebellar atrophy, Huntington dis- and bilateral parietal atrophy was observed. The hippocam- ease (without genetic confirmation), unspecified chorea, pus, frontal, and temporal lobes showed no apparent de- progressive supranuclear palsy (PSP), AD, frontotemporal generation. This patient died after a disease duration of 14 dementia (FTD), and unspecified dementia. years.

Two patients are further described in more detail: patient III- Genetic analyses 10 presented at age 65 years with a parkinsonian gait, frequent Diagnostic testing excluded mutations in MAPT, C9orf72, falls, and bradykinesia unresponsive to levodopa. Sub- PSEN1, PRNP, HTT, DRPLA, FMR1, SCA1, SCA2, SCA3, sequently, memory problems developed together with be- SCA6, SCA7,andSCA17. Genome-wide linkage analysis havioral changes and impulsiveness. On neurologic using SNP array data did not reveal a significance linkage examination, dysarthria, an upward gaze palsy, bilateral bra- peak (LOD score >3). However, a few regions showed dykinesia, a shuffling gait with freezing, and abnormal postural suggestive linkage, including the region on chromosome reflexes were found. Neuroimaging with MRI showed gen- 16p13.3 with an LOD score of 2.35. After filtering, WES data eralized atrophy, including cerebellar volume loss. This pa- of 7 patients revealed 5 candidate variants, all of them in the tient died 18 years after disease onset at age 83 years. heterozygous state and located on chromosome 16p13.3. These included 4 nonsynonymous missense variants Patient III-13 presented with progressive behavioral changes (AXIN1, IGFALS, PDK1, and THOC6)and1frameshift from age 50 years, including inappropriate laughing, aggres- variant (STUB1). Analysis of WGS data of 3 patients using siveness, and perseverations. Neurologic examination the same filtering steps confirmed these 5 variants; no ad- revealed dysarthria, diffuse motor restlessness, and stereotypic ditional variants were found. In these patients, LobSTR was movements. Neuropsychological evaluation indicated used on the chromosome 16p13.3 linkage area to identify

4 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG possible expanded polyglutamine repeats, but none were and III-13 showed few abnormalities, except for a few positive detected. neurons in the temporal neocortex (III-12) and a moderate number of pretangles (III-13). Of interest, AT8 staining in Of these 5 candidates, the frameshift variant in STUB1 is the III-10 showed many tufted astrocytes, glial staining, and most likely candidate because this gene has recently been tangles in the thalamus and subthalamic nucleus (figure 2I). linked to both recessive (SCAR16) and dominant ataxia Tau pathology was also present in many other regions (SCA48). The 2-bp deletion c.731_732delGC (i.e., temporal cortex (figure 2J), hippocampus, caudate, (p.C244Yfs*24) is located in exon 6, which encodes the U-box putamen, substantia nigra, locus coeruleus, cerebellar dentate domain together with the last exon. This variant is absent in nucleus, and spinal motor neurons). Immunohistochemistry publicly available databases, and Sorting Intolerant From for alpha-synuclein and TDP-43 antibodies in several selected Tolerant predicts a damaging effect (confidence score 0.86). brain regions was negative in all 3 cases. The mutation was confirmed by Sanger sequencing (figure e-1, links.lww.com/NXG/A251). Subsequently, 5 unaffected relatives were tested (III-7, III-14, III-15, III-16, and III-17), Discussion and in 4 of 5, the variant was absent. Individual III-4, who was not considered to have the same phenotype, did not have the We present a multigenerational family with late-onset ataxia fi STUB1 variant. There was no report of neurologic symptoms associated with a novel heterozygous STUB1 mutation, tting in individual II-3, who is an obligate carrier of the STUB1 the diagnosis of SCA48. Although the various initial diagnoses variant, but died at age 64 years due to a myocardial infarction. of the patients illustrate the clinical heterogeneity of this new SCA subtype, there are some strongly overlapping features. Neuropathology The clinical presentation is dominated by cognitive decline Neuropathologic examination was performed in 3 patients and profound behavioral changes, combined with cerebellar (III-10, III-12, and III-13). Brain weight varied slightly (1,304, ataxia and movement disorders. 1,223, and 1,090 g, respectively). Macroscopic inspection showed cerebellar atrophy of vermis and hemispheres in all 3 Cognitive impairment has been found in other SCA subtypes, cases, although less profound in III-10. Routine staining on but is usually preceded by cerebellar symptoms.20 Here, it was the 3 brains showed no abnormalities in cortical areas, except present in all affected family members and occasionally as the for focal neuron loss in the parietal and occipital cortices in first manifestation. One case was initially diagnosed as FTD, III-13. Cases III-12 and III-13 contained no apparent hippo- although frontal and temporal atrophy were lacking. Other campal plaques or tangles (Braak stage 1), whereas a few published SCA48 pedigrees with mutations in close proximity tangles and neuropil threads were seen in the Ammon horn in to the current mutation showed a broad range of cognitive – III-10 (Braak stage 2). All 3 cases showed nearly complete loss symptoms.9 12 Another interesting observation is the pres- of Purkinje cells with Bergmann gliosis (figure 2A) and severe ence of other movement disorders besides gait ataxia. Chorea neuronal loss in the mesencephalon and medulla oblongata. or uncontrolled motor activity was frequently reported, In III-10, the subthalamic nucleus also showed atrophy with showing resemblance with the chorea described in several – severe gliosis. Staining of the cerebellum of III-13 with other families.10 12 Parkinsonism was also present in 4 of 11 ubiquitin showed several neuronal intranuclear inclusions described Italian patients with SCA48.11 Atypical parkinson- (NIIs) in cerebellar granular cells (figure 2C). In all 3 cases, ism with features resembling PSP was observed in 1 pa- immunohistochemistry with p62 showed similar NII in the tient. The combination of cognitive and movement disorders cerebellum (figure 2D) and in all cortical areas, most preva- in SCA48 also resembles SCA17 and dentatorubral- lent in posterior regions (figure 2E). NIIs were also present in pallidoluysian atrophy. In both disorders, dementia and the substantia nigra and deep pontine nuclei. In the hippo- chorea are part of the clinical features.1,21 campus, many p62-positive inclusions were found, whereas the dentate gyrus showed punctuate cytoplasmic inclusions The clinical presentation of SCA48 shows overlap with (CIs) and occasional diffuse nuclear staining (DNS). Staining SCAR16. The latter is more complex with a myriad of phe- with 1C2 antibody in III-13 showed DNS in granular and notypes and seems to constitute a multisystemic disorder stellate neurons of the cerebellum and in cortical areas (figure usually presenting at an earlier age.7,22 Endocrine abnormal- 2F). In the substantia nigra, CIs were observed, and the gyrus ities have been described as a major feature of SCAR16, and cinguli showed CI, DNS, and dystrophic neurites. 1C2 hypogonadism was present in patients with SCA48.11 Al- staining also exposed DNS in the inferior olives (figure 2G) though not formally tested, endocrine abnormalities were not and CI in the hypoglossal nucleus with a granular pattern reported in the medical records of the current pedigree. (figure 2H). In III-10, faint 1C2 reactivity was seen in the cerebellum, less prominent than with p62. In cortical regions, The homozygous and compound heterozygous STUB1 similar DNS was seen as in III-13. 1C2 staining was not mutations in SCAR16, and the recent identification of het- performed in III-12. Staining with the antibody against CHIP erozygous STUB1 mutations in adult-onset familial ataxia, in III-13 showed a diffuse staining of neurons, not different support STUB1 as the most likely candidate gene in this from controls (data not shown). AT8 staining in cases III-12 family. The identified frameshift variant is predicted to result

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 5 Figure 2 Histopathologic features of affected cases III-13 (A–H) and III-10 (I and J)

(A and B) Nearly complete loss of Purkinje cells, loss of neurons, and spongiosis in the cerebellar granular layer of the patient (A) compared with the control brain (B) in H/E staining; (C and D) NII in the cerebellar granular layer with ubiquitin (C) and p62 staining (D); (E) NII in the occipital cortex with p62 staining; (F) DNS in the occipital cortex with 1C2 staining; (G) DNS in the inferior olive with 1C2 staining; (H) granular CI in the hypoglossal nucleus with 1C2 staining; (I) AT8-positive neurons, neurites, and glial staining in the thalamic nuclei; (J) AT8-positive tufted astrocyte in the putamen. The arrowheads indicate NII, which are magnified in figure 2C-F. CI = cytoplasmic inclusion; DNS = diffuse nuclear staining; H/E = hematoxylin and eosin staining; NII = neuronal intranuclear inclusion.

in a premature stop codon and is located in the highly con- region (U-box) is responsible for the ubiquitination of mis- served U-box domain, similar to previously reported variants folded proteins destined for elimination.24 Missense muta- causing SCAR16 and SCA48.8,9,11,23 The protein encoded by tions in the U-box domain of CHIP impair the intracellular STUB1, CHIP, regulates several proteins involved in neuro- degradation of ataxin-3 microaggregates by preventing their degenerative disorders.6 Through its tetratricopeptide repeat ubiquitination, thereby accelerating disease progression.25 We domain, CHIP interacts with chaperones (Hsp70 and Hsp90) hypothesize that a similar pathophysiologic mechanism is and a broad range of co-chaperones. The ubiquitin ligase likely in SCA48 due to impaired function of CHIP.

6 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG Earlier work has shown that mutations in the U-box domain studies of other STUB1 cases are scarce, with a single neu- causing SCAR16 destabilize CHIP, supporting a loss-of- ropathologic report of an SCAR16 patient.32 This case function mechanism.26 CHIP null mice and animal models showed similar severe Purkinje cell loss and p62-positive NIIs, with homozygous missense mutation (p.T246M) exhibited and CHIP staining revealed nuclear staining in neurons of the striking similarities with characteristics seen in SCAR16 deep layers of the frontal cortex. We also observed diffuse patients, including neuronal degeneration and ataxic motor staining of neurons with CHIP antibody, but not different disturbances.7,27 Patients with recessive STUB1 mutations had from controls (data not shown). decreased CHIP protein levels in isolated fibroblasts in varying degrees, dependent on specific mutations.27 The U-box do- NIIs have been considered the hallmark of disorders with repeat main is especially important, as mutations in this domain have expansions.33 However, several studies have demonstrated the astrongereffect on the overall loss of CHIP function and presence of NIIs in the absence of pathologically expanded strongly associate with cognitive dysfunction.28 Indeed, most polyglutamine repeats.34 Moreover, NIIs in the present cases did frameshift mutations identified so far in SCA48 are located in not show 1C2 positivity, which is typically present in SCA caused the U-box domain (table e-1, links.lww.com/NXG/A251).9,11 by polyglutamine repeat expansion.2 Instead, we did observe Whether this results in nonsense mediated RNA decay or a pattern of DNS with 1C2. Granular cytoplasmic and DNS of a truncated protein (with a possible dominant negative effect) 1C2 as an earlier stage in aggregate formation have been de- remains to be elucidated. Additional functional studies on dif- scribedinbothSCA3andHuntingtondisease.35 Other studies ferent mutations in STUB1 are essential to provide more in- have demonstrated that 1C2 recognizes nonexpanded ataxin-3 sight why some mutations cause SCA48 in heterozygous state present in NIIs.36,37 It is possible that recruitment of ataxin-3 into and others SCAR16 in a recessive trait. The co-occurrence of the nucleus is involved in the pathogenesis of STUB1 disease.38 recessive and dominant mutations in the same gene has been Finally, 1C2 immunoreactivity might be less specific, as it has also described before. For example, mutations in HTRA1 are re- been detected in nonrepeat disorder, such as a SCA6 case with sponsible for cerebral autosomal recessive arteriopathy with 22–24 polyglutamine stretches39 and in healthy controls.40 Fu- subcortical infarcts and leukoencephalopathy (CARASIL), ture investigations are needed to clarify the exact meaning of the while heterozygous mutations causing a late-onset form of different types of staining patterns on the cellular (dys)function. small vessel disease have been reported.29 Likewise, truncating heterozygous progranulin mutations cause FTD, whereas ho- Intriguingly, extensive tau pathology was found in 1 patient mozygous mutations lead to neuronal ceroid lipofuscinosis.30 with clinical similarities to PSP (III-10). This has up till now only been seen in SCA11, a pure progressive cerebellar In the current study, 2 carriers of the STUB1 mutation did not ataxia.2 Several studies highlight the role of CHIP in tau pa- develop symptoms: 1 unaffected carrier died over age 75 thology. It attenuates the pathologic changes associated with years, and another (obligate) carrier died at age 64 years tau, playing a primary role in tau ubiquitination and degra- without neurologic disease. This could indicate incomplete dation.41 CHIP levels were found to be elevated in AD, sug- penetrance, or it could reflect the diversity in phenotypical gesting a role in tangle maturation,42 and CHIP depletion in expression, given the large variety in ages at onset in the mice was associated with accumulation of hyper- present family (50–72 years). Therefore, it cannot be ex- phosphorylated tau.43 The discovery of different pathologies cluded that parents of some patients with SCAR16 are at risk within 1 kindred is known to occur also in other (genetic) of developing SCA48. neurodegenerative diseases.44

It would also be of interest to investigate other genetic factors The present pedigree along with the recently described fam- that could influence disease penetrance and progression. ilies underscores the association between heterozygous Previous studies have shown that expanded ataxin-2 repeats STUB1 mutations and SCA48. The combination of familial predispose to other neurodegenerative disorders besides late-onset gait ataxia, cognitive decline with behavioral SCA2.31 In line with this, it would be interesting to explore changes, and other movement disorders (parkinsonism, whether repeat length variation in ataxin-3, a known inter- chorea, and stereotypic movements) is indicative of SCA48. actor of CHIP, might interfere with the function of CHIP and We show that the pathology of SCA48 is dominated by influence the clinical variability in SCA48. Although we have Purkinje cell loss, p62-positive intranuclear inclusions, and not found toxic exposures in patients (data not shown), it is may also include tau pathology. Additional functional studies worthwhile to study more systematically whether environ- and reports of newly identified SCA48 and SCAR16 families mental factors can trigger the unfolded protein response and will provide more insight into the heterogeneous spectrum of may therefore have an effect on disease onset or progression STUB1-related disorders. in STUB1-related disorders. Acknowledgment Neuropathologic changes were present in cerebellar but also The authors are indebted to all the patients who made this neocortical areas, suggesting diffuse involvement of the brain. study possible. They also thank Prof. A. Rozemuller from the The almost complete loss of Purkinje cells in the present 3 Netherlands Brain Bank for the neuropathologic examination cases resembles that of most other SCAs.2 Neuropathologic of the cases.

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 7 Study funding 4. Ruano L, Melo C, Silva MC, Coutinho P. The global epidemiology of hereditary ataxia fi and spastic paraplegia: a systematic review of prevalence studies. Neuroepidemiology This work was nancially supported by Prinses Beatrix Fonds 2014;42:174–183. (grant number 01-0128) and by the Internationaal Parkinson 5. Synofzik M, Nemeth AH. Recessive ataxias. Handb Clin Neurol 2018;155:73–89. 6. Dickey CA, Patterson C, Dickson D, Petrucelli L. Brain CHIP: removing the culprits Fonds (IPF). in neurodegenerative disease. Trends Mol Med 2007;13:32–38. 7. Shi CH, Schisler JC, Rubel CE, et al. Ataxia and hypogonadism caused by the loss of ubiquitin ligase activity of the U box protein CHIP. Hum Mol Genet 2014;23: Disclosure 1013–1024. Several authors of this publication are members of the Eu- 8. Shi Y, Wang J, Li JD, et al. Identification of CHIP as a novel causative gene for autosomal recessive cerebellar ataxia. PLoS One 2013;8:e81884. ropean Reference Network for Rare Neurological Diseases - 9. Genis D, Ortega-Cubero S, San Nicolas H, et al. Heterozygous STUB1 mutation Project ID No 739510. Disclosures available: Neurology. causes familial ataxia with cognitive affective syndrome (SCA48). Neurology 2018;91: e1988–e1998. org/NG. 10. De Michele G, Lieto M, Galatolo D, et al. Spinocerebellar ataxia 48 presenting with ataxia associated with cognitive, psychiatric, and extrapyramidal features: a report of two Italian families. Parkinsonism Relat Disord 2019;65:91–96. Publication history 11. Lieto M, Riso V, Galatolo D, et al. The complex phenotype of spinocerebellar ataxia Received by Neurology: Genetics November 15, 2019. Accepted in final type 48 in eight unrelated Italian families. Eur J Neurol 2020;27:498–505. ff form February 19, 2020. 12. Palvadeau R, Kaya-Gulec ZE, Simsir G, et al. Cerebellar cognitive-a ective syndrome preceding ataxia associated with complex extrapyramidal features in a Turkish SCA48 family. Neurogenetics 2020;21:51–58. 13. Hoffmann K, Lindner TH. easyLINKAGE-plus—automated linkage analyses using large-scale SNP data. Bioinformatics 2005;21:3565–3567. Appendix Authors 14. Gudbjartsson DF, Jonasson K, Frigge ML, Kong A. Allegro, a new computer program for multipoint linkage analysis. Nat Genet 2000;25:12–13. Name Location Contribution 15. O’Connell JR, Weeks DE. PedCheck: a program for identification of genotype in- compatibilities in linkage analysis. Am J Hum Genet 1998;63:259–266. Merel O. Erasmus Medical Center, Analyzed and interpreted 16. Li H, Durbin R. Fast and accurate short read alignment with Burrows–Wheeler Mol, MD Rotterdam, the the data and drafted the transform. Bioinformatics 2009;25:1754–1760. Netherlands manuscript 17. McKenna A, Hanna M, Banks E, et al. The genome analysis toolkit: a mapreduce framework for analyzing next-generation DNA sequencing data. Genome Res 2010; Jeroen G.J. Erasmus Medical Center, Analyzed and interpreted 20:1297–1303. van Rooij, Rotterdam, the the data and revised the 18. Wang K, Li M, Hakonarson H. ANNOVAR: functional annotation of genetic variants MSc Netherlands manuscript from high-throughput sequencing data. Nucleic Acids Res 2010;38:e164. 19. Gymrek M, Golan D, Rosset S, Erlich Y. lobSTR: a short tandem repeat profiler for Esther Erasmus Medical Center, Role in the acquisition of personal genomes. Genome Res 2012;22:1154–1162. Brusse, MD, Rotterdam, the clinical data and revised the 20. Burk K. Cognition in hereditary ataxia. Cerebellum 2007;6:280–286. PhD Netherlands manuscript 21. Tsuji S. Dentatorubral-pallidoluysian atrophy. Handb Clin Neurol 2012;103: 587–594. Annemieke Erasmus Medical Center, Analyzed and interpreted 22. Hayer SN, Deconinck T, Bender B, et al. STUB1/CHIP mutations cause Gordon J.M.H. Rotterdam, the the data and revised the Holmes syndrome as part of a widespread multisystemic neurodegeneration: evidence Verkerk, Netherlands manuscript from four novel mutations. Orphanet J Rare Dis 2017;12:31. PhD 23. Synofzik M, Schule R, Schulze M, et al. Phenotype and frequency of STUB1 mutations: next-generation screenings in caucasian ataxia and spastic paraplegia cohorts. Shamiram Erasmus Medical Center, Performed laboratory work Orphanet J Rare Dis 2014;9:57. Melhem, BSc Rotterdam, the 24. Murata S, Chiba T, Tanaka K. CHIP: a quality-control E3 ligase collaborating with Netherlands molecular chaperones. Int J Biochem Cell Biol 2003;35:572–578. 25. Williams AJ, Knutson TM, Colomer Gould VF, Paulson HL. In vivo suppression of Wilfred F.A. University Medical Centre Interpreted the data and polyglutamine neurotoxicity by C-terminus of Hsp70-interacting protein (CHIP) den Dunnen, Groningen, Groningen, revised the manuscript supports an aggregation model of pathogenesis. Neurobiol Dis 2009;33:342–353. MD, PhD the Netherlands 26. Kanack AJ, Newsom OJ, Scaglione KM. Most mutations that cause spinocerebellar ataxia autosomal recessive type 16 (SCAR16) destabilize the protein quality-control Patrizia German Center for Analyzed and interpreted E3 ligase CHIP. J Biol Chem 2018;293:2735–2743. Rizzu, MD, Neurodegenerative the data and revised the 27. Shi CH, Rubel C, Soss SE, et al. Disrupted structure and aberrant function of CHIP PhD Diseases (DZNE), manuscript mediates the loss of motor and cognitive function in preclinical models of SCAR16. Tuebingen, Germany PLoS Genet 2018;14:e1007664. 28. Madrigal SC, McNeil Z, Sanchez-Hodge R, et al. Changes in protein function underlie Chiara IRCCS Centro Neurolesi Analyzed and interpreted the disease spectrum in patients with CHIP mutations. J Biol Chem 2019;294: Cupidi, MD, “Bonino Pulejo,” Messina, the data 19236–19245. PhD Italy 29. Favaretto S, Margoni M, Salviati L, Pianese L, Manara R, Baracchini C. A new Italian family with HTRA1 mutation associated with autosomal-dominant variant of CAR- – John C. van Erasmus Medical Center, Designed and ASIL: are we pointing towards a disease spectrum? J Neurol Sci 2019;396:108 111. ffi Swieten, Rotterdam, the conceptualized the study 30. Ward ME, Chen R, Huang HY, et al. Individuals with progranulin haploinsu ciency MD, PhD Netherlands and revised the manuscript exhibit features of neuronal ceroid lipofuscinosis. Sci Transl Med 2017;9:eaah5642. 31. Ross OA, Rutherford NJ, Baker M, et al. Ataxin-2 repeat-length variation and neu- – Laura Erasmus Medical Center, Designed and rodegeneration. Hum Mol Genet 2011;20:3207 3212. Donker Rotterdam, the conceptualized the study, 32. Bettencourt C, de Yebenes JG, Lopez-Sendon JL, et al. Clinical and neuropathological Kaat, MD, Netherlands analyzed and interpreted features of spastic ataxia in a Spanish family with novel compound heterozygous – PhD the data, and revised the mutations in STUB1. Cerebellum 2015;14:378 381. manuscript 33. Yamada M, Sato T, Tsuji S, Takahashi H. CAG repeat disorder models and human neuropathology: similarities and differences. Acta Neuropathol 2008;115:71–86. 34. Fujigasaki H, Uchihara T, Koyano S, et al. Ataxin-3 is translocated into the nucleus for the formation of intranuclear inclusions in normal and Machado-Joseph disease References brains. Exp Neurol 2000;165:248–256. 1. Klockgether T, Mariotti C, Paulson HL. Spinocerebellar ataxia. Nat Rev Dis Primers 35. Seidel K, Siswanto S, Fredrich M, et al. Polyglutamine aggregation in Huntington’s 2019;5:24. disease and spinocerebellar ataxia type 3: similar mechanisms in aggregate formation. 2. Seidel K, Siswanto S, Brunt ER, den Dunnen W, Korf HW, Rub U. Brain pathology of Neuropathol Appl Neurobiol 2016;42:153–166. spinocerebellar ataxias. Acta Neuropathol 2012;124:1–21. 36. Perez MK, Paulson HL, Pittman RN. Ataxin-3 with an altered conformation that 3. Durr A. Autosomal dominant cerebellar ataxias: polyglutamine expansions and be- exposes the polyglutamine domain is associated with the nuclear matrix. Hum Mol yond. Lancet Neurol 2010;9:885–894. Genet 1999;8:2377–2385.

8 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG 37. Takahashi J, Tanaka J, Arai K, et al. Recruitment of nonexpanded polyglutamine 41. Saidi LJ, Polydoro M, Kay KR, et al. Carboxy terminus heat shock protein 70 inter- proteins to intranuclear aggregates in neuronal intranuclear hyaline inclusion disease. acting protein reduces tau-associated degenerative changes. J Alzheimers Dis 2015;44: J Neuropathol Exp Neurol 2001;60:369–376. 937–947. 38. Scaglione KM, Zavodszky E, Todi SV, et al. Ube2w and ataxin-3 coordinately regulate 42. Sahara N, Murayama M, Mizoroki T, et al. In vivo evidence of CHIP up-regulation the ubiquitin ligase CHIP. Mol Cell 2011;43:599–612. attenuating tau aggregation. J Neurochem 2005;94:1254–1263. 39. Ishikawa K, Owada K, Ishida K, et al. Cytoplasmic and nuclear polyglutamine 43. Dickey CA, Yue M, Lin WL, et al. Deletion of the ubiquitin ligase CHIP leads to the aggregates in SCA6 Purkinje cells. Neurology 2001;56:1753–1756. accumulation, but not the aggregation, of both endogenous phospho- and caspase-3- 40. Herndon ES, Hladik CL, Shang P, Burns DK, Raisanen J, White CL III. Neuroan- cleaved tau species. J Neurosci 2006;26:6985–6996. atomic proﬁle of polyglutamine immunoreactivity in Huntington disease brains. 44. Santpere G, Ferrer I. LRRK2 and neurodegeneration. Acta Neuropathol 2009;117: J Neuropathol Exp Neurol 2009;68:250–261. 227–246.

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 9 ARTICLE OPEN ACCESS Acute encephalopathy after head trauma in a patient with a RHOBTB2 mutation

Annemarie C.S. Knijnenburg, MD,* Joost Nicolai, MD, PhD,* Levinus A. Bok, MD, PhD, Akin Bay, BSc, Correspondence Alexander P.A. Stegmann, PhD, Margje Sinnema, MD, PhD, and Maaike Vreeburg, MD, PhD Drs. Knijnenburg [email protected] Neurol Genet 2020;6:e418. doi:10.1212/NXG.0000000000000418 Abstract Objective De novo missense mutations in the RHOBTB2 gene have been described as causative for developmental and epileptic encephalopathy.

Methods The clinical phenotype of this disorder includes early-onset epilepsy, severe intellectual disability, postnatal microcephaly, and movement disorder. Three RHOBTB2 patients have been described with acute encephalopathy and febrile epileptic status. All showed severe EEG abnormalities during this episode and abnormal MRI with hemisphere swelling or reduced diﬀusion in various brain regions.

Results We describe the episode of acute encephalopathy after head trauma in a 5-year-old RHOBTB2 patient. At admission, Glasgow coma scale score was E4M4V1. EEG was severely abnormal showing a noncontinuous pattern with slow activity without epileptic activity indicating severe encephalopathy. A second EEG on day 8 was still severely slowed and showed focal delta activity frontotemporal in both hemispheres. Gradually, he recovered, and on day 11, he had regained his normal reactivity, behavior, and mood. Two months after discharge, EEG showed further decrease in slow activity and increase in normal electroen- cephalographic activity. After discharge, parents noted that he showed more hyperkinetic movements compared to before this period of encephalopathy. Follow-up MRI showed an increment of hippocampal atrophy. In addition, we summarize the clinical characteristics of a second RHOBTB2 patient with increase of focal periventricular atrophy and development of hemiparesis after epileptic status.

Conclusions Acute encephalopathy in RHOBTB2 patients can also be triggered by head trauma.

*These authors contributed equally to this work.

From the Department of Neurology (A.C.S.K., J.N.), Maastricht University Medical Center, Maastricht; Department of Pediatrics (L.A.B.), M`axima Medical Center, Veldhoven; and Department of Clinical Genetics (A.B., A.P.A.S., M.S., M.V.), Maastricht University Medical Center, Maastricht, the Netherlands.

Funding information and disclosures are provided at the end of the article. Full disclosure form information provided by the authors is available with the full text of this article at Neurology.org/NG.

The Article Processing charge was funded by the authors. This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloading and sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Copyright © 2020 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 Early 2018, de novo missense mutations in the RHOBTB2 phenobarbital and later valproic acid and levetiracetam. From gene were described as causative for a developmental and the age of 1 year 2 months, new attacks were noted characterized epileptic encephalopathy.1 All these mutations are located in by agitation and dystonic posturing. EEG registration proved the BTB-encoding region of the RHOBTB2 gene. The clinical that those attacks were nonepileptic. Despite genetic and met- phenotype of this disorder includes early-onset epilepsy, se- abolic analyses, no etiologic diagnosis could be made. Because vere intellectual disability, postnatal microcephaly, and periodic kinesigenic dyskinesia is known to react to low-dose a movement disorder (dystonic, paroxysmal, or chorea-like). carbamazepine, carbamazepine monotherapy was started. At the Recently, 3 RHOBTB2 patients with acute encephalopathy age of 2.5 years, 2 doses of carbamazepine were missed because 2 and febrile epileptic status were described. All showed severe of gastroenteritis. This resulted in frequent epileptic seizures, EEG abnormalities during this episode and abnormal MRI necessitating midazolam IV continuously. Follow-up EEG at the ff with hemisphere swelling or reduced di usion in various brain age of 2 years 7 months is shown in figure 1A. regions. Gradually, severe global psychomotor delay was noted. At the We describe the episode of acute encephalopathy of 1 child in age of 2 years 8 months, trio whole exome sequencing showed detail and summarize the characteristics of acute encepha- a de novo mutation in de RHOBTB2 gene, c.1448 G>A lopathy in another patient because of a similar episode. The (p.Arg483His), then of unknown significance. Later, the index case adds new information not described till now; be- details of this boy were published (individual 2), emphasizing sides fever, acute encephalopathy in RHOBTB2 patients can the need for international collaboration in the case of de novo also be triggered by other mechanisms as head trauma. mutations of unknown significance. The mutation was proved to be pathogenic by the in vitro analysis showing impaired degradation of the mutant RHOBTB2 by the proteasome and Case 1 neurologic defects in mutant Drosophila.1 The index patient was born at term in 2013 as the first child from healthy nonrelated parents. He had a good start at birth. For almost 4 years, he was quite stable showing hyperkinetic Already at the age of 4 days, seizures were noticed. He showed dyskinesia and vivid startle reactions that were proven by simul- opisthotonus, head rotation, and did not react. On the first taneous EEG recording to be epileptic. He had no long-lasting examination at the age of 3.5 months, poor visual fixation episodes of painful dystonic posturing and no other kind of and mild axial hypotonia were noted. He was treated with seizures.

Figure 1 Follow-up EEG

(A) Three years before admission, (B) 1 day after admission, (C) 8 days after admission, and (D) 2 months after discharge.

2 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG In September 2018, he had a mild trauma when his head hit of stools showed enterovirus. However, CSF analysis for en- his bicycle. According to his parents, he was severely fright- terovirus was negative. ened but did not lose consciousness. A little later, he did not respond to his parents, was staring, and vomited several times. He suffered from many seizures with unilateral clonic jerks, eye A CT scan of the brain in a local hospital did not show any deviation, and desaturation. These were treated with midazolam traumatic injury. Because anisocoria was noted, a non- nasally, increase of dose of carbamazepine, and initiation of lev- convulsive epileptic status was suspected and treated with etiracetam. On the fourth day after the accident, he reacted a little a single dose of midazolam IV. Several hours after the onset of to his parents. Because of the negative results, acyclovir, ceftriax- symptoms, he developed fever up to 102.7°F, and the next one, and dexamethasone were stopped. A second EEG on day 8 day, he was transported to the department of child neurology was still severely slowed and showed focal delta activity fronto- because of unexplained persistent encephalopathy. temporal in both hemispheres (figure 1C). MRI showed subtle reduced diffusion in the hippocampal area of the left hemisphere On admission, he showed a E4M4V1 Glasgow coma scale and the right-sided hippocampal atrophy (figure 2, A and B). score. His pupils showed a normal light reaction. He showed an Gradually, he recovered, and on day 11, he regained his normal alternating straight-forward or right-sided lateral deviation of reactivity, behavior, and mood. Two months after discharge, both eyes. EEG registration was made to exclude a non- a new EEG showed further decrease in slow activity (figure 1D). convulsive epileptic status. EEG was severely abnormal showing a noncontinuous pattern with slow activity without epileptic After discharge, parents noted an increase in hyperkinetic activity indicating a severe encephalopathy (figure 1B). movements. Follow-up MRI showed increment of hippocampal brain atrophy (figure 2, C and D). Because of the fever and unexplained encephalopathy, viral encephalitis and bacterial meningitis were considered, and Another patient with a de novo ROHBTB2 mutation, obtained treatment with acyclovir, ceftriaxone, and dexamethasone was by trio whole exome sequencing (also with 22q11 duplication) started. A lumbar puncture showed a mild pleocytosis (14 × with an episode of encephalopathy after only mild hyperther- 106/L leucocytes [normal <5 × 106/L], 86% granulocytes). mia 100.4°F and similar EEG and MRI findings was identified. Cultures of CSF, blood, and urine were negative. PCR viral We include summarized findings of both patients in the table. analyses for HSV type I and type II were negative. PCR analysis The parents of both patients gave consent for publication.

Figure 2 Follow-up MRI of the cerebrum

(A) DWI during admission and (B) ADC during admission, showing subtle diffusion restriction in the hippocampal area. (C) IR 3 years before admission and (D) IR 3 months after admission showing atrophy in the hippocampal area. ADC = apparent diffusion coefficient; DWI = diffusion- weighted imaging; IR = inversion recovery.

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 3 Discussion Acute encephalopathy is a generic term for a subtype of encephalopathy mainly occurring in young children with acute brain dysfunction preceded by an infection but without evi- Clinical follow-up Increased hyperkinetic movements Decreased motor skills dence of a brain infection.3 Acute encephalopathy is more frequent in Japan compared with other countries in the world. This disorder can be triggered by diﬀerent infections, including inﬂuenza, herpes virus type 6, gastrointestinal infec-

Follow-up MRI Hippocampal atrophy White matter damage tions, and RS virus. Besides, the use of NSAIDs has been proven to be a risk factor.3

Some genetic risk factors for the occurrence of postviral acute encephalopathy are known. RANBP2 mutations are found in most children with familial or recurrent acute (necrotizing) encephalopathy.4 Acute encephalopathy has also been de- left hippocampal area Cortical diffusion restriction left hemisphere scribed in children with SCN1A5 and SCN2A mutations6 and in children with certain carnitine palmitoyl transferase II polymorphism7 and ADORA2A polymorphisms.8 Acute encephalopathy is not described in children with 22q11 duplications. Severely slowed Diffusion restriction Severely slowed left > right Children with de novo RHOBTB2 mutations are rarely found. Three children with a RHOBTB2 mutation were identiﬁed in a cohort of 1,230 children with infantile or early-childhood epilepsy.2 It has been found that the altered protein RHOBTB2 underlying cause Yes 100.4°F, no underlying cause is relatively resistant to degradation, leading to overexpression of the altered protein. In Drosophila, overexpression of the ortholog RhoBTB is associated with seizure susceptibility and severe locomotor defects.1 epilepticus The children with RHOBTB2 mutations described with acute encephalopathy have some symptoms and signs in common. First of all, it is triggered by fever or, in this case, by a mild trauma with severe frightening, with fever developing several Precipitating event Seizures Hyperthermia EEG MRI-cerebrum hours after the onset of encephalopathy. The episodes are long lasting. EEG recordings are severely slowed and gradually recover. Brain MRI reveals swelling or reduced diﬀusion of various brain regions and can show focal atrophy over time. Length of episode, d In addition, long-term symptoms can worsen after the recovery of an episode of acute encephalopathy. Age at onset, y 14 12 None Status It seems that children with a RHOBTB2 mutation do not show stable intellectual impairment, and several show regression- or stagnation-associated with the occurrence of seizures or acute encephalopathy.1,2,9 When the seizures occur at young age, this regression or stagnation is possibly not

227-19,791,607) duplication recognized.

Boy NoneGirl 22q11.2 (17,275, 5 11 Trauma MultipleBoth Yes, 102.7°F, no our patients had an increase in clinical symptoms after acute encephalopathy and focal brain atrophy, as described in other patients with RHOBTB2 mutations. It has been described after febrile epileptic status but can also occur after

c.1448 G>A (p.Arg483His) c. 1532 G>A (p.Arg551Gln) head trauma or with mild fever. Patient characteristics in 2 patients with pathogenic RHOBTB2 variants and episode of acute encephalopathy Study funding Table Patient Mutation Sex Comorbity 1 2 No targeted funding reported.

4 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG Disclosure Disclosures available: Neurology.org/NG. Appendix (continued)

Name Location Contribution Publication history Received by Neurology: Genetics August 1, 2019. Accepted in ﬁnal form Margje Maastricht University Medical Revision of the Sinnema, MD, Center, Maastricht, the manuscript February 26, 2020. PhD Netherlands

Maaike Maastricht University Medical Revision of the Vreeburg, MD, Center, Maastricht, the manuscript PhD Netherlands Appendix Authors

Name Location Contribution References Annemarie Maastricht University Medical Writing first draft of 1. Straub J, Konrad EDH, Grüner J, et al. Missense variants in RHOBTB2 cause a de- C.S. Center, Maastricht, the the manuscript velopmental and epileptic encephalopathy in humans, and altered levels cause neu- Knijnenburg, Netherlands rological defects in Drosophila. Am J Hum Genet 2018;102:44–57. MD 2. Belal H, Nakashima M, Matsumoto H, et al. De novo variants in RHOBTB2, an atypical Rho GTPase gene, cause epileptic encephalopathy. Hum Mutat 2018;39:1070–1075. Joost Nicolai, Maastricht University Medical Writing first draft of 3. Mizuguchi M, Yamanouchi H, Ichiyama T, Shiomi M. Acute encephalopathy associated MD, PhD Center, Maastricht, the the manuscript with influenza and other viral infections. Acta Neurol Scand Suppl 2007;186:45–56. Netherlands 4. Gika AD, Rich P, Gupta S, Neilson DE, Clarke A. Recurrent acute necrotizing encephalopathy following influenza A in a genetically predisposed family. Dev Med Levinus A. Màxima Medical Center, Acquisition of patient Child Neurol 2010;52:99–102. Bok, MD, PhD Veldhoven, the Netherlands data, revision of the 5. Saitoh M, Shinohara M, Hoshino H, et al. Mutations of the SCN1A gene in acute manuscript encephalopathy. Epilepsia 2012;53:558–564. 6. Fukasawa T, Kubota T, Negoro T, et al. A case of recurrent encephalopathy with Akin Bay, BSc Maastricht University Medical Revision of the SCN2A missense mutation. Brain Dev 2015;37:631–634. Center, Maastricht, the manuscript 7. Shinohara M, Saitoh M, Takanashi J, et al. Carnitine palmitoyl transferase II poly- Netherlands morphism is associated with multiple syndromes of acute encephalopathy with various infectious diseases. Brain Dev 2011;33:512–517. Alexander P.A. Maastricht University Medical Revision of the 8. Shinohara M, Saitoh M, Nishizawa D, et al. ADORA2A polymorphism predisposes Stegman, PhD Center, Maastricht, the manuscript children to encephalopathy with febrile status epilepticus. Neurology 2013;80:1571–1576. Netherlands 9. Lopes F, Barbosa M, Ameur A, et al. Identification of novel genetic causes of Rett syndrome-like phenotypes. J Med Genet 2016;53:190–199.

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 5 ARTICLE OPEN ACCESS Cerebellar ataxia, neuropathy, hearing loss, and intellectual disability due to AIFM1 mutation

Massimo Pandolfo, MD, Myriam Rai, PhD, Gauthier Remiche, MD, PhD, Laurence Desmyter, PhD, and Correspondence Isabelle Vandernoot, MD Dr. Pandolfo [email protected] Neurol Genet 2020;6:e420. doi:10.1212/NXG.0000000000000420 Abstract Objective To describe the clinical and molecular genetic ﬁndings in a family segregating a novel mutation in the AIFM1 gene on the X chromosome.

Methods We studied the clinical features and performed brain MRI scans, nerve conduction studies, audiometry, cognitive testing, and clinical exome sequencing (CES) in the proband, his mother, and maternal uncle. We used in silico tools, X chromosome inactivation assessment, and Western blot analysis to predict the consequences of an AIFM1 variant identiﬁed by CES and demonstrate its pathogenicity.

Results The proband and his maternal uncle presented with childhood-onset nonprogressive cerebellar ataxia, hearing loss, intellectual disability (ID), peripheral neuropathy, and mood and behavioral disorder. The proband’s mother had mild cerebellar ataxia, ID, and mood and behavior disorder, but no neuropathy or hearing loss. The 3 subjects shared a variant (c.1195G>A; p.Gly399Ser) in exon 12 of the AIFM1 gene, which is not reported in the exome/genome sequence databases, aﬀecting a critical amino acid for protein function involved in NAD(H) binding and predicted to be pathogenic with very high probability by variant analysis programs. X chromosome inactivation was highly skewed in the proband’s mother. The mutation did not cause quantitative changes in protein abundance.

Conclusions Our report extends the molecular and phenotypic spectrum of AIFM1 mutations. Specific findings include limited progression of neurologic abnormalities after the first decade and the coexistence of mood and behavior disorder. This family also shows the confounding effect on the phenotype of nongenetic factors, such as alcohol and drug use and side effects of medication.

From the Neurology Service (M.P., G.R.), Hopitalˆ Erasme; Laboratory of Experimental Neurology (M.P., M.R.); and Medical Genetics Service (L.D., I.V.), Hopitalˆ Erasme, Universit´e Libre de Bruxelles, Belgium.

Go to Neurology.org/NG for full disclosures. Funding information is provided at the end of the article.

Copyright © 2020 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 Glossary CES = clinical exome sequencing; CR = creatine; ID = intellectual disability; MRS = magnetic resonance spectroscopy; PBMC = peripheral blood mononuclear cell; SCA = spinocerebellar ataxia.

We describe a family segregating a complex syndrome be- Standard protocol approvals, registrations, cause of a previously unreported mutation in the AIFM1 and patient consents (apoptosis-inducing factor, mitochondrion-associated, 1) Patients provided written informed consent for genetic gene, encoding a homodimeric flavoprotein tethered to the analysis and for the use of their coded data for research pur- mitochondrial inner membrane that is required to maintain poses, as approved by the Ethics Committee of the Hopitalˆ the mitochondrial respiratory complex I. AIFM1 has a rote- Erasme, Brussels, Belgium. none-sensitive reduced nicotinamide adenine dinucleotide (NADH):Ubiquinone oxidoreductase activity, whose func- Data availability tional role remains unclear.1 In addition, under conditions of Coded clinical and genetic data not allowing patient identi- mitochondrial stress, a cleaved fragment of AIFM1 is also fication and details on pipeline and filtering of CES results are implicated in caspase-independent programmed cell death available on request. induction.2 Mutations may affect one or both of these functions.3 Cerebellar atrophy is prominent in the Harlequin (Hq) mouse, a spontaneous Aifm1 mutant,4 but is not a common Results feature of the multiple phenotypes associated with AIFM1 Clinical presentation mutations in humans. These include a severe neonatal mito- The family tree is shown in figure 1. No information is available chondrial encephalomyopathy,5 a more slowly progressive on the maternal grandparents of the proband. In the proband, encephalopathy,6 the association of sensorineural hearing loss instability and clumsiness were first noticed when he was 2 years and axonal neuropathy called Cowchock syndrome,7 other old. At age 10, he had severe gait and limb ataxia, dysarthria, and “mitochondrial” phenotypes of variable severity and symp- abnormal eye movements with jerky pursuit and slow saccades. toms, and other unique presentations as infantile motor BrainMRIwasperformedatage2,4,7,9,and13years.Al- neuron disease,8 distal motor neuropathy,9 ventriculomegaly, though the first 2 investigations were reported to be normal, mild and myopathy.10 Cerebellar ataxia, mostly of childhood onset, cerebellar atrophy was seen at age 7, which progressed in the has been occasionally reported in association with other following years, affecting both vermis and hemispheres (figure 2, phenotypes of variable severity.6,8,10,11 In this article, we ex- A and B). Knee and ankle reflexes were absent at age 2, and there tend the spectrum of AIFM1-related phenotypes and report was mild leg amyotrophy, suggesting that polyneuropathy was a novel AIFM1 mutation affecting the NAD(H) binding site. initially responsible for motor symptoms. Cognitive development was considered normal until age 2, and then, it Methods We collected DNA samples on 3 family members. Initial ge- Figure 1 Family pedigree netic screening involved testing the proband for spinocerebellar ataxias (SCAs) caused by cytosine-adenosine- guanosine repeat expansions (SCA1, 2, 3, 6, 7, 17). Because this was negative, next generation sequencing of 3,638 genes associated with pathologic human phenotypes (clinical exome sequencing [CES]) was performed on all 3 individuals’ DNA.

X chromosome inactivation in the proband’s mother peripheral blood mononuclear cells (PBMCs) was assessed by ampliﬁcation of the cytosine-adenosine-guanosine repeat in the androgen receptor gene before and after cleavage with the methylation-sensitive enzymes HpaII and CfoI.

For Western blot analysis, PBMCs from the proband and his mother were isolated by centrifugation on Ficoll, mechan- ically lysed and separated into a nuclear and cytoplasmic fraction by centrifugation. AIFM1 was detected using a goat The proband, individual III-1, is indicated with an arrowhead. AIFM1 geno- polyclonal primary antibody (Novus Biologicals, Abingdon, types are indicated as wt (G399) or M (S399). UK) and chemiluminescence detection.

2 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG Figure 2 Patients MRI showing cerebellar atrophy

Coronal (A, C, and E) and sagittal (B, D, and F) T1- (A-F) and T2-weighted (C) brain MRI images from the proband at age 13 (A and B), his maternal uncle at age 42 (C and D), and proband’s mother at age 42 (E and F). Notice the mild cerebellar atrophy in the proband’s mother and the moderate cerebellar atrophy in the proband and his uncle. stagnated. At age 10, he had a total IQ of 50, with no significant was attributed to chronic treatment with valproate as mood difference between verbal and perceptual reasoning scores. stabilizers. Her cognitive function was in the mild intellectual Hearing was still normal at age 8, but hearing loss was present at disability (ID) range. She had no hearing loss. Brain MRI at age 10. The clinical picture remained stable in the second decade, age 43 showed mild cerebellar atrophy (figure 2, C and D). with some improvement with intense rehabilitation. Magnetic resonance spectroscopy (MRS) revealed decreased N-acetyl-aspartate/creatine (CR) and choline/CR ratios in The proband’s mother had mild truncal and limb ataxia, the cerebellum and brainstem, with a detectable lactate peak. dysarthria, mild loss of vibration sense at external malleoli, and a bilateral extensor plantar response. Tendon reflexes in her The proband’smaternaluncle,twinbrotheroftheproband’s lower limbs were brisk at age 42, then weakened and dis- mother, when examined in his 40s had similar clinical features of appeared by age 46. Gaze-evoked nystagmus, present at age childhood-onset ataxia that stopped progressing in the second 42, also disappeared by age 46, whereas saccades become slow decade, ID with prominent executive dysfunction, peripheral and hypometric. She had had mood instability and behavioral neuropathy, and deafness. Mood instability and behavioral problems since her adolescence, leading to repeated referrals problems, for which he received antidepressants, neuroleptics, to psychiatric services. She had received a diagnosis of bipolar and mood stabilizers, had been prominent since childhood. disorder and was treated with antidepressants, neuroleptics, When seen at age 42, he had moderate generalized bradykinesia and mood stabilizers. A rest and action tremor of her hands and rigidity, which resolved after stopping neuroleptics and

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 3 valproate. At age 42, MRI showed diffuse cerebellar atrophy, databases (PM2), it is a missense variant in a gene that has affecting both vermis and hemispheres (figure 2, E and F). a low rate of benign missense variation and in which missense Blood lactate was mildly increased on several occasions. variants are a common mechanism of disease (PP2), and multiple lines of computational evidence support a deleteri- Genetic analyses ous effect on the gene or gene product (PP3). In addition, the CES revealed a variant (c.1195G>A; p.Gly399Ser) in exon 12 patients’ phenotype has highly specific features associated of the AIFM1 gene (NM_004208) on the X chromosome with AIFM1 mutations, in particular axonal polyneuropathy shared by the proband, his maternal uncle, and his mother. and deafness (PP4). Highly skewed X inactivation in the This variant is not reported in the literature and is not present heterozygous proband’s mother (91:9 ratio) further supports in ExAC and gnomAD (supplemental material, links.lww. the pathogenicity of this variant. com/NXG/A254). Western blot analysis of AIFM1 (not shown) revealed no AIFM1 has 3 domains: flavin adenine dinucleotide-binding difference in size or abundance between the proband, his (residues 128–262 and 401–480), NADH-binding (residues mother, and normal male and female controls, indicating that 263–400), and C-terminal (residues 481–608) (atlas- the mutation did not affect the synthesis, maturation, or geneticsoncology.org//Genes/AIFM1ID44053chXq25. degradation of the protein. html). The variant affects a highly conserved glycine that is located in the NADH-binding domain, where it directly interacts with NADH (figure 3). It is predicted to be deleterious by Polyphen-2, SIFT, and MutationTaster, likely pre- Discussion venting or destabilizing NADH binding to AIFM1, which is The phenotype of our male patients partially overlaps with 12 needed for its oxidoreductase activity and dimerization. Cowchock syndrome because they both have axonal polyneuropathy and deafness in addition to cerebellar ataxia and ID. Following the American College of Medical Genetics guide- However, cerebellar atrophy, which is a prominent feature in this lines, the variant, even if found in a single family, is classified as family has rarely been reported in human patients with AIFM1 likely pathogenic because it is located in a critical and well- mutation. Interesting, cerebellar ataxia progression essentially established functional domain without benign variation occurred in the first decade, followed by stabilization and, in the (PM1), it is absent from controls in exome and genome proband, even improvement with intensive rehabilitation. Treatment with riboflavin,reportedtobebeneficial in 2 patients with ataxia and AIFM1 mutations, might have provided further 11 Figure 3 Amino acid change affecting the AIFM1 NAD(H)A improvement. Whether the mood and behavior disorder also binding site prominent in this family is coincidental, possibly because of an unfavorable environment combined to other genetic risk factors, or a consequence of the mutation remains speculative. However, the very similar psychiatric and cognitive profiles of the proband and his uncle suggest that the mutation may indeed have a predisposing role, if not a causative role.

Of interest all previously reported AIFM1 mutations are recessive, with only hemizygous men being clinically affected, whereas in our family, a heterozygous woman seems to be symptomatic, although in a much milder form than her brother and son. However, in her case, the causative role of the AIFM1 mutation may be questioned. First, highly skewed X inactivation makes it unlikely that a mutation present in a very small proportion of active X chromosomes can be disease causing, even if, by affecting NADH binding, it may prevent dimerization of the protein. In addition, although brain lactate was detected by MRS in this patient, this finding is not specific and cannot prove a role of the mutation in causing her neurologic features. In addition, this patient has a history of alcohol and drug abuse and use of neuroleptics Schematic of the AIFM1 structure around the NAD(H)A and FAD binding sites. and mood stabilizers, which are likely to have clouded her The mutated G399 residue (boxed) is directly involved in NAD(H) binding. Adapted with permission from American Chemical Society from Ferreira P, clinical picture. In this regard, her tremor was clearly induced et al.12 Copyright 2014 American Chemical Society, Washington, DC. All by valproate, whereas polyneuropathy and mild cerebellar permission requests for this image should be made to the copyright holder. FAD = flavin adenine dinucleotide. atrophy might have been secondary to alcohol abuse, although we cannot exclude that the AIFM1 mutation may have

4 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG enhanced her vulnerability to these factors. Notably, iatro- genic complications due to neuroleptic treatment also oc- Appendix (continued) curred in the proband and in his uncle; a strong reminder of Name Location Contribution how a variety of factors, including nongenetic ones, may aﬀect the phenotype of a genetic disorder and be potentially mis- Gauthier Universit´eLibrede Drafting/Revising the manuscript; Remiche, MD, Bruxelles, Belgium and data acquisition leading in the diagnostic process. PhD

Laurence UniversitéLibrede Analysis or interpretation of data Study funding Desmyter, Bruxelles, Belgium This study was supported by internal funding from the Service PhD ˆ of Neurology and of Medical Genetics, Hopital Erasme, Isabelle UniversitéLibrede Drafting/revising the manuscript; Université Libre de Bruxelles, Brussels, Belgium. Vandernoot, Bruxelles, Belgium data acquisition; and analysis or MD interpretation of data Disclosure M. Pandolfo is Deputy Editor for Neurology: Genetics. M. Rai, References G. Remiche, L. Desmyter, and I. Vandernoot have nothing to 1. Elguindy MM, Nakamaru-Ogiso E. Apoptosis-inducing factor (AIF) and its family disclose. Go to Neurology.org/NG for full disclosure. member protein, AMID, are rotenone-sensitive NADH:ubiquinone oxidoreductases (NDH-2). J Biol Chem 2015;290:20815–20826. 2. Susin SA, Lorenzo HK, Zamzami N, et al. Molecular characterization of mitochondrial Publication history apoptosis-inducing factor. Nature 1999;397:441–446. fi 3. Sevrioukova IF. Structure/function relations in AIFM1 variants associated with Received by Neurology: Genetics January 7, 2020. Accepted in nal form neurodegenerative disorders. J Mol Biol 2016;428:3650–3665. March 7, 2020. 4. Klein JA, Longo-Guess CM, Rossmann MP, et al. The harlequin mouse mutation downregulates apoptosis-inducing factor. Nature 2002;419:367–374. 5. Ghezzi D, Sevrioukova I, Invernizzi F, et al. Severe X-linked mitochondrial encephalomyopathy associated with a mutation in apoptosis-inducing factor. Am J Hum Genet 2010;86:639–649. Appendix Authors 6. Ardissone A, Piscosquito G, Legati A, et al. A slowly progressive mitochondrial encephalomyopathy widens the spectrum of AIFM1 disorders. Neurology 2015;84:2193–2195. Name Location Contribution 7. Rinaldi C, Grunseich C, Sevrioukova IF, et al. Cowchock syndrome is associated with a mutation in apoptosis-inducing factor. Am J Hum Genet 2012;91:1095–1102. Massimo Université Libre de Data acquisition; study concept 8. Diodato D, Tasca G, Verrigni D, et al. A novel AIFM1 mutation expands the phe- Pandolfo, MD Bruxelles, Belgium or design; analysis or notype to an infantile motor neuron disease. Eur J Hum Genet 2016;24:463–466. interpretation of data; 9. Sancho P, Sánchez-Monteagudo A, Collado A, et al. A newly distal hereditary motor contribution of vital reagents/ neuropathy caused by a rare AIFM1 mutation. Neurogenetics 2017;18:245–250. tools/patients; acquisition of 10. Kettwig M, Schubach M, Zimmermann FA, et al. From ventriculomegaly to severe data; and study supervision muscular atrophy: expansion of the clinical spectrum related to mutations in AIFM1. Mitochondrion 2015;21:12–18. Myriam Rai, Université Libre de Analysis or interpretation of data; 11. Heimer G, Eyal E, Zhu X, et al. Mutations in AIFM1 cause an X-linked childhood PhD Bruxelles, Belgium contribution of vital reagents/ cerebellar ataxia partially responsive to riboflavin. Eur J Paediatr Neurol 2018;22:93–101. tools/patients; and acquisition of 12. Ferreira P, Villanueva R, Mart´ınez-Júlvez M, et al. Structural insights into the co- data enzyme mediated monomer-dimer transition of the pro-apoptotic apoptosis inducing factor. Biochemistry 2014;53:4204–4215.

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 5 ARTICLE OPEN ACCESS TGM6 L517W is not a pathogenic variant for spinocerebellar ataxia type 35

Yanxing Chen, MD, PhD, Dengchang Wu, MD, PhD, Benyan Luo, MD, PhD, Guohua Zhao, MD, PhD, and Correspondence Kang Wang, MD, PhD Dr. Zhao [email protected] Neurol Genet 2020;6:e424. doi:10.1212/NXG.0000000000000424 or Dr. Wang [email protected] Abstract Objective To investigate the pathogenicity of the TGM6 variant for spinocerebellar ataxia 35 (SCA35), which was previously reported to be caused by pathogenic mutations in the gene TGM6.

Methods Neurologic assessment and brain MRI were performed to provide detailed description of the phenotype. Whole-exome sequencing and dynamic mutation analysis were performed to identify the genotype.

Results The proband, presenting with myoclonic epilepsy, cognitive decline, and ataxia, harbored both the TGM6 p.L517W variant and expanded CAG repeats in gene ATN1. Further analysis of the other living family members in this pedigree revealed that the CAG repeat number was expanded in all the patients and within normal range in all the unaﬀected family members. However, the TGM6 p.L517W variant was absent in 2 aﬀected family members, but present in 3 healthy individuals.

Conclusions The nonsegregation of the TGM6 variant with phenotype does not support this variant as the disease-causing gene in this pedigree, questioning the pathogenicity of TGM6 in SCA35.

From the Department of Neurology (Y.C., G.Z.), the Second Affiliated Hospital, School of Medicine, Zhejiang University; and Department of Neurology (D.W., B.L., K.W.), the First Affiliated Hospital, School of Medicine, Zhejiang University, Hangzhou, China.

Go to Neurology.org/NG for full disclosures. Funding information is provided at the end of the article.

Copyright © 2020 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 Glossary AD = autosomal dominant; DRPLA = dentatorubral-pallidoluysian atrophy; GTCS = generalized tonic-clonic seizure; SCA = spinocerebellar ataxia.

Spinocerebellar ataxias (SCAs) are a group of autosomal members. Briefly, 22-FAM-labeled primer sets were used for dominant (AD) disorders that mainly affect the cerebellar triplet repeat primed PCR, followed by PCR product processed function. SCAs are genetically heterogeneous, and there are by capillary electrophoresis with the ABI3730xl (Applied Bio- more than 40 subtypes of SCAs and more than 30 genes systems, Foster, CA). The primer sequences of all SCAs genes identified responsible for them (neuromuscular.wustl.edu/). were designed using reference sequences from GenBank. Data SCA35 was reported to be associated with mutation in trans- were analyzed with GeneMapper v4.0 (Applied Biosystems). glutaminase 6 gene (TGM6), which codes for transglutaminase 6 protein (TG6) in 2010.1 A 4-generation Chinese AD-SCA Data availability family was identified to have 9 affected members showing The data that support the findings of this study are available progressive gait instability, scanning speech, and poor dexterity from the corresponding author on reasonable request. in hands. Linkage analysis suggested that the disease-causing gene was located on chromosome 20p13-12.2, and exome sequencing showed that the c.1550T>G (p.L517W) variant Results cosegregated with the phenotypes. Another variant c.980A>G The proband (IV:1, figure, A) is a 28-year-old woman, who fi 2 (D327G) was identi ed in another AD-SCA family. In the presented to the neurology outpatient clinic with a 13-year ff following years, another 13 variants were reported in di erent history of unprovoked seizures. She developed generalized ethnic groups including Chinese, Asian, European, and tonic-clonic seizures (GTCSs) at age 15 years, which mainly 2–9 Hispanic, mainly in Asia. Here, we report an AD family with occurred in the morning, on awakening. She was then pre- 6 patients presenting with ataxia. Nonsegregation of the TGM6 scribed with several antiepileptic drugs, including valproate, p.L517W variant with phenotype questions the pathogenicity lamotrigine, and levetiracetam, as monotherapy or polytherapy of this variant. with different combinations, but still experiencing 3–5GTCSs each year. At age 25 years, she started to develop myoclonic jerks, which manifested with a brief, shock-like, involuntary Methods movement of the upper limbs, more severe distally and aggra- fi Participants vated by motion, and clonazepam was added with signi cant The 4-generation pedigree was recruited from the First Af- improvement. One year ago, the patient gradually exhibited filiated Hospital of Zhejiang University School of Medicine. unsteadiness, speech disturbance, and memory decline. Neu- Neurologic examination was performed by at least 2 senior rologic examination revealed mild cerebellar dysarthria, dys- fi neurologists. metria performing nger-to-nose test, and positive Romberg sign. Ophthalmoplegia, nystagmus, and saccadic movements Standard protocol approvals, registrations, were not observed. She had normal muscle strength without and patient consents rigidity, normal deep tendon reflexes, and a Mini-Mental State This study was approved by the Medical Ethics Committee of Examination score of 20. Brain MRI revealed mild atrophy of the First Affiliated Hospital of Zhejiang University School of the cerebellum and cerebral cortex (figure). Video- Medicine (No. 2017-326). Written informed consent was electroencephalography (video-EEG) showed normal back- obtained from each participant or from a legal representative. ground activities and intermittent spikes involving the posterior head region. Photic sensitivity was absent. Considering the Genetic analysis progressiveness and the involvement of multiple neurologic Genomic DNA was extracted from the peripheral blood of systems of this patient, inherited neurodegenerative diseases family members. Exon-enriched DNA sequencing and bio- were suspected. Whole-exome sequencing identified a reported informatic analysis were performed on the Illumina HiSeq variant c.1550T>G (p.L517W) in the TGM6 gene. Given that X-ten (Illumina, CA) in high-output mode with 150 bp paired- epilepsy has not been reported in patients with TGM6 muta- end reads following the manufacturer’s instructions (Illumina) tions, we further studied ataxia caused by trinucleotide (or in the proband. The suspected pathogenic variants were vali- pentanucleotide) repeat expansion. Surprisingly, genetic testing dated by Sanger sequencing. The results of Sanger sequencing using capillary electrophoresis revealed 14/61 CAG repeats in performed in other living family members were used for the atrophin-1 (ATN1) gene on chromosome 12p13.31, which cosegregation analysis. A dedicated panel for screening SCAs is a well-established gene for DRPLA. However, this raises the including SCA1, 2, 3, 6, 7, 8, 10, 12, 17, and dentatorubral- question whether this patient happened to harbor double pallidoluysian atrophy (DRPLA) was used to test repeat mutations, which might have additive effects on the phenotype, expansions of related mutations in all the available family or the TGM6 p.L517W variant is not actually pathogenic.

2 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG Figure Family pedigree and brain MRI of the proband

(A) Family pedigree of DRPLA disease. (B) Brain MRI of the proband showing mild cerebellar atrophy. (C) Normal control of the TGM6 gene. (D) The TGM6 variant in the patient. Arrows indicate the variant site. DRPLA = dentatorubral-pallidoluysian atrophy.

We subsequently studied the whole family and found that unstable CAG trinucleotide repeat expansion in ATN1. Nor- family members I:2, II:1, II:3, III:1, and III:5 developed cere- mally, the CAG tract bears 6–35 repeats, which expands to over bellar ataxia and cognitive decline/dementia in their fifties or 49 in patients with DRPLA.10 The clinical features of DRPLA sixties (table). The TGM6 p.L517W variant was absent in are strikingly heterogeneous, depending on the age at disease patients III:5 and II:3, but present in unaffected individuals (II: onset and the prominent genetic anticipation.11 Patients with 4, III:3, and III:7). The nonsegregation of the TGM6 variant juvenile-onset (onset before age 20 years) frequently exhibit with phenotype does not support this variant as the disease- progressive myoclonic epilepsy, intellectual disabilities, and causing gene in this family. On the other hand, the CAG repeat ataxia. Various forms of seizures are common features in all number was expanded in all the patients (II:3, III:1, III:5, and patients with onset before age 20 years. Patients with onset IV:1) and within normal range in all the unaffected family after age 20 years typically present symptoms including cere- members with available DNA, perfectly cosegregating with bellar ataxia, choreoathetosis, and dementia. Therefore, the disease phenotype. No DNA was available from the deceased clinical manifestations of the proband comply with juvenile- patients (I:1, I:2, and II:1). Patient II:1 died of gastric cancer at onset DRPLA, while the other affected family members also fit age 53 years. The presence of the ATN1 variant in her children well the clinical spectrum of DRPLA. indicates that patient II:1 should be a carrier. This observation strongly suggests that the disease is caused by the CAG repeat The identification of TGM6 as a causative gene for SCA35 was expansion in ATN1 rather than the p.L517W variant in TMG6. first reported in 2010.1 The TGM6 p.L517W variant cosegregated with the phenotype in a Chinese 4-generation SCA family. Cosegregation of another variant of TGM6 with the Discussion phenotype was also identified in another 2-generation family. More variants in the TGM6 gene were later discovered in In the current study, we present a 4-generation pedigree with patients with ataxia manifestations by other groups, supporting ataxia and cognitive decline/dementia. One well-established the role of TG6 in ataxia syndrome.2,3,8 However, with the wide CAG expansion in ATN1 and another reported pathogenic application of the next-generation sequencing in clinical prac- TGM6 p.L517W variant were identified. Cosegregation anal- tice in recent years, increasing genetic results are available, ysis confirmed the ATN1 mutation as the disease-causing gene, which challenges the previously reported pathogenic variants casting doubts on the pathogenicity of the TGM6 p.L517W that have not been extensively validated. Besides, the emerging variant. public database-based variant analysis provides an accessible and efficient approach to test the pathogenicity probability of ATN1 is a well-established gene for DRPLA, which is a rare, suspected variants. Recently, in a Chinese exome sequencing inherited AD neurodegenerative disorder resulting from an cohort, 8 families were found to harbor the reported TGM6

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 3 Table Clinical information and genetic results in the family members

Family Age at c.1550T>G ATN1 CAG member Sex Age (y) onset (y) Clinical symptoms variant repeats

I:2 Female Deceased NA Ataxia / /

II:1 Female Deceased Fifties Ataxia / /

II:2 Male 83 / None Negative 14/17

II:3 Male 75 Sixties Ataxia/dementia Negative 9/54

II:4 Male 72 / None Positive 17/18

III:1 Male 56 52 Ataxia/cognitive Positive 17/55 decline

III:2 Female 50 / None Negative 14/18

III:3 Male 55 / None Negative 14/17

III:5 Female 53 48 Ataxia/cognitive Negative 14/56 decline

III:7 Male 51 / None Positive 17/17

IV:1 Female 28 15 Seizure/myoclonus/ Positive 14/61 ataxia/cognitive decline

Abbreviation: NA = not available.

variants but share no features of SCA35. They further reviewed provides direct and compelling evidence that the TMG6 the public database genomAD and found that the reported p.L517W is not pathogenic. pathogenic variants, including L517W, are significantly more common in East Asians than in other ethnic groups. Gene A few functional studies of the TMG6 variants indicate a bi- constraint metrics showed that both missense and loss-of- ologically possible link with neurodegeneration. The sub- function variants in TGM6 are unlikely to be disease causing. cellular distribution, expression, and in vitro activity of the 2 Inflation analysis of the reported pathogenic TGM6 variants variants of TGM6 (D327G and L517W) have been in- showed that there is an at least 111-fold inflation over disease vestigated. In 1 study, it was found that neither of the mutants prevalence of all AD SCAs. This level of inflation is far beyond changed the subcellular localization of TG6, but exhibited de- the threshold of 10-fold, which indicates a high chance of creased transglutaminase activity and exerted the cells more misdiagnosis/misclassification of the variant or very low pen- vulnerable to staurosporine-induced apoptosis.13 On the other etrance. They raised the concern that misclassification of be- hand, another study found that 5 previously reported TGM6 nign or low penetrant variants as pathogenic is a significant mutations, including L517W, were associated with nuclear problem, which would often result in genetic misdiagnosis.12 depletion of TG6 and loss of the transglutaminase activity, and However, this study still fails to reach a definite conclusion thus leading to the activation of the unfolded protein response regarding the causality between the genotype and phenotype. and neuronal death.8 These studies showed an increased vul- First, the inability to perform cosegregation analysis in this nerability of cells transfected with TGM6 variants, suggesting study cannot rule out the likelihood of incomplete penetrance a role of TG6 in neuronal viability. However, variants associ- of the variants. Second, the size of database and the ethnic ated biological function changes might not be responsible for recruiting bias should be taken into consideration when inter- the clinical manifestations observed in patients, which are also preting the results. Last, the majority of index individuals in commonly seen in risk genes. Even healthy individuals carry their cohort were infants or children, and half of their asymp- many rare protein-disrupting variants.14 Therefore, further tomatic variant–carrying parents were under middle age. Thus, functional studies that can replicate the disease-relevant phe- there exists possibility that they might become symptomatic at notypes are needed to establish the causal relationship between older age, given that the mean age at onset for SCA35 was 43.7 TGM6 variants and the disease phenotype. years.1 Pedigree cosegregation analysis could provide additional evidence on the pathogenicity of the variants, especially We report a pedigree with ataxia, in which a well-established that lack of segregation of a variant is strong evidence against ATN1 mutation and a TGM6 variant were identified. Non- pathogenicity. Therefore, the pedigree we report here is of segregation of the TGM6 L517W variant with disease phe- special value. The nonsegregation of the variant in this pedigree notype indicates that this variant is not disease causing. We do

4 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG not rule out the possibility that certain genuine pathogenic TGM6 variants might exist. Appendix (continued)

Name Location Contributions Acknowledgment The authors thank the patients and their families for Guohua Department of Neurology, Designed and Zhao, MD, the Second Affiliated conceptualized the study; participating in this study. PhD Hospital, School of major role in acquisition Medicine, Zhejiang and interpretation of data; Study funding University, Hangzhou, and critical review China This study was supported by the National Key R&D Program of China (2017YFC0907700), National Natural Science Kang Wang, Department of Neurology, Designed and MD, PhD the First Affiliated Hospital, conceptualized the study; Foundation of China (81870826), and Zhejiang Provincial School of Medicine, major role in acquisition Natural Science Foundation of China (LY19H180006, Hangzhou, China and interpretation of data; and critical review LY18H090004).

Disclosure References Y. Chen, D. Wu, B. Luo, G. Zhao, and K. Wang report no 1. Wang JL, Yang X, Xia K, et al. TGM6 identified as a novel causative gene of spino- disclosures. Go to Neurology.org/NG for full disclosures. cerebellar ataxias using exome sequencing. Brain 2010;133:3510–3518. 2. Li M, Pang SY, Song Y, Kung MH, Ho SL, Sham PC. Whole exome sequencing identifies a novel mutation in the transglutaminase 6 gene for spinocerebellar ataxia in Publication history a Chinese family. Clin Genet 2013;83:269–273. fi 3. Pan LL, Huang YM, Wang M, et al. Positional cloning and next-generation se- Received by Neurology: Genetics November 16, 2019. Accepted in nal quencing identified a TGM6 mutation in a large Chinese pedigree with acute myeloid form March 13, 2020. leukaemia. Eur J Hum Genet 2015;23:218–223. 4. Guo YC, Lin JJ, Liao YC, Tsai PC, Lee YC, Soong BW. Spinocerebellar ataxia 35: novel mutations in TGM6 with clinical and genetic characterization. Neurology 2014; 83:1554–1561. 5. Fasano A, Hodaie M, Munhoz RP, Rohani M. SCA 35 presenting as isolated treatment- resistant dystonic hand tremor. Parkinsonism Relat Disord 2017;37:118–119. Appendix Authors 6. Yang ZH, Shi MM, Liu YT, et al. TGM6 gene mutations in undiagnosed cerebellar ataxia patients. Parkinsonism Relat Disord 2018;46:84–86. Name Location Contributions 7. Choi KD, Kim JS, Kim HJ, et al. Genetic variants associated with episodic ataxia in Korea. Sci Rep 2017;7:13855. Yanxing Department of Neurology, Drafted the manuscript; 8. Tripathy D, Vignoli B, Ramesh N, et al. Mutations in TGM6 induce the unfolded Chen, MD, the Second Affiliated major role in acquisition protein response in SCA35. Hum Mol Genet 2017;26:3749–3762. PhD Hospital, School of and interpretation of data; 9. Lin CC, Gan SR, Gupta D, Alaedini A, Green PH, Kuo SH. Hispanic spinocerebellar Medicine, Zhejiang and critical review ataxia type 35 (SCA35) with a novel frameshift mutation. Cerebellum 2019;18:291–294. University, Hangzhou, 10. Wardle M, Morris HR, Robertson NP. Clinical and genetic characteristics of non- China Asian dentatorubral-pallidoluysian atrophy: a systematic review. Mov Disord 2009; 24:1636–1640. Dengchang Department of Neurology, Acquisition and 11. Tsuji S. Dentatorubral-pallidoluysian atrophy. Handb Clin Neurol 2012;103: Wu, MD, the First Affiliated Hospital, interpretation of data and 587–594. PhD School of Medicine, critical review 12. Fung JLF, Tsang MHY, Leung GKC, et al. A significant inflation in TGM6 genetic risk Hangzhou, China casts doubt in its causation in spinocerebellar ataxia type 35. Parkinsonism Relat Disord 2019;63:42–45. Benyan Department of Neurology, Acquisition and 13. Guan WJ, Wang JL, Liu YT, et al. Spinocerebellar ataxia type 35 (SCA35)-associated Luo, MD, the First Affiliated Hospital, interpretation of data and transglutaminase 6 mutants sensitize cells to apoptosis. Biochem Biophys Res PhD School of Medicine, critical review Commun 2013;430:780–786. Hangzhou, China 14. Tennessen JA, Bigham AW, O’Connor TD, et al. Evolution and functional impact of rare coding variation from deep sequencing of human exomes. Science 2012;337:64–69.

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 5 ARTICLE OPEN ACCESS Expanding the phenotypic and molecular spectrum of RNA polymerase III–related leukodystrophy

Stefanie Perrier, HBSc,* Laurence Gauquelin, MD, FRCPC,* Catherine Fallet-Bianco, MD, Correspondence Megan K. Dishop, MD, Mackenzie A. Michell-Robinson, MSc, Luan T. Tran, MSc, Kether Guerrero, MSc, Dr. Bernard [email protected] Lama Darbelli, PhD, Myriam Srour, MDCM, PhD, Kevin Petrecca, MD, PhD, FRCSC, Deborah L. Renaud, MD, Michael Saito, MD, Seth Cohen, MD, Steffen Leiz, MD, Bader Alhaddad, MD, Tobias B. Haack, MD, Ingrid Tejera-Martin, MD, Fernando I. Monton, MD, PhD, Norberto Rodriguez-Espinosa, MD, Daniela Pohl, MD, PhD, Savithri Nageswaran, MBBS, MPH, Annette Grefe, MD, Emma Glamuzina, MD, and Genevi`eve Bernard, MD, MSc, FRCPC

Neurol Genet 2020;6:e425. doi:10.1212/NXG.0000000000000425 Abstract Objective To expand the phenotypic spectrum of severity of POLR3-related leukodystrophy and identify genotype- phenotype correlations through study of patients with extremely severe phenotypes.

Methods We performed an international cross-sectional study on patients with genetically proven POLR3-related leukodystrophy and atypical phenotypes to identify 6 children, 3 males and 3 females, with an extremely severe phenotype compared with that typically reported. Clinical, radiologic, and molecular features were evaluated for all patients, and functional and neuropathologic studies were performed on 1 patient.

Results Each patient presented between 1 and 3 months of age with failure to thrive, severe dysphagia, and developmental delay. Four of the 6 children died before age 3 years. MRI of all patients revealed a novel pattern with atypical characteristics, including progressive basal ganglia and thalami abnormalities. Neuropathologic studies revealed patchy areas of decreased myelin in the cerebral hemispheres, cerebellum, brainstem, and spinal cord, with astrocytic gliosis in the white matter and microglial activation. Cellular vacuolization was observed in the thalamus and basal ganglia, and neuronal loss was evident in the putamen and caudate. Genotypic similarities were also present between all 6 patients, with one allele containing a POLR3A variant causing a premature stop codon and the other containing a speciﬁc intronic splicing variant (c.1771-7C>G), which produces 2 aberrant transcripts along with some wild-type transcript.

Conclusions We describe genotype-phenotype correlations at the extreme end of severity of the POLR3-related leukodystrophy spectrum and shed light on the complex disease pathophysiology.

*These authors contributed equally to the manuscript.

From the Department of Neurology and Neurosurgery (S.P., L.G., M.A.M.-R., L.T.T., K.G., L.D., M. Srour, K.P., G.B.), McGill University; Child Health and Human Development Program (S.P., M.A.M.-R., L.T.T., K.G., L.D., M. Srour, G.B.), Research Institute of the McGill University Health Centre; Department of Pediatrics (L.G., L.T.T., K.G., L.D., M. Srour, G.B.), McGill University, Montreal, Quebec, Canada; Division of Clinical and Metabolic Genetics (L.G.), Division of Neurology, the Hospital for Sick Children, University of Toronto, Ontario, Canada; Department of Pathology (C.F.-B.), CHU Sainte-Justine, Universit´e de Montreal, Quebec, Canada; Division of Pathology and Laboratory Medicine (M.K.D.), Phoenix Children’s Hospital, AZ; Department of Human Genetics (L.T.T., K.G., L.D., G.B.), McGill University, Montreal, Quebec, Canada; McGill University (K.P.), Brain Tumour Research Center Montreal Neurological Institute and Hospital, Quebec, Canada; Department of Neurology (D.L.R.), Department of Clinical Genomics, Department of Pediatrics, Mayo Clinic, Rochester, MN; Department of Pediatrics (M. Saito), University of California Riverside School of Medicine, Riverside Medical Clinic, CA; Department of Pediatrics (S.C.), Beaver Medical Group, Redlands, CA; Division of Pediatric Neurology (S.L.), Department of Pediatrics, Klinikum Dritter Orden, Munich, Germany; Institute of Human Genetics (B.A., T.B.H.), Technische Universit¨at Munchen,¨ Munich, Germany; Institute of Medical Genetics and Applied Genomics (T.B.H.), University of Tubingen,¨ Germany; Department of Neurology (I.T.-M., F.I.M., N.R.-E.), Hospital Universitario Nuestra Señora de Candelaria, Santa Cruz de Tenerife, Canary Islands, Spain; Department of Neurology (D.P.), Children’s Hospital of Eastern Ontario, University of Ottawa, Ontario, Canada; Department of Pediatrics (S.N.) and Department of Neurology (A.G.), Wake Forest School of Medicine, Winston-Salem, NC; Adult and Paediatric National Metabolic Service (E.G.), Starship Children’s Hospital, Auckland, New Zealand; and Division of Medical Genetics (G.B.), Department of Specialized Medicine, Montreal Children’s Hospital and McGill University Health Centre, Quebec, Canada.

Go to Neurology.org/NN for full disclosures. Funding information is provided at the end of the article.

The Article Processing Charge was funded by the Foundation of Stars. This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloading and sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Copyright © 2020 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 Glossary nc-RNA = noncoding RNA; NMD = nonsense mediated decay; POLR3-HLD = RNA polymerase III-related hypomyelinating leukodystrophy; tRNA = transfer RNA.

RNA polymerase III-related hypomyelinating leukodystrophy Neuroradiology (POLR3-HLD; MIM: 607694, 614381, 616494), or 4H leu- Brain MRI review was performed on latest available scans kodystrophy, is one of the most common hypomyelinating by L.G. and G.B. based on previously published criteria leukodystrophies, typically associated with the cardinal clini- for hypomyelination and POLR3-HLD imaging – cal features of hypogonadotropic hypogonadism and characteristics.1,5,6,17 19 The earliest studies were also ana- – hypodontia.1 3 POLR3-HLD commonly presents in child- lyzed when available. Only one study was available for P5 hood, with motor delay or regression, prominent cerebellar and P6. features, mild pyramidal signs, and variable cognitive involvement.1 Typical brain MRI pattern includes diffuse Neuropathology hypomyelination with relative preservation (T2 hypointen- Neuropathologic investigations were performed on post- sity) of the anterolateral nucleus of the thalamus, globus mortem brain tissue from P2; details are provided in sup- pallidus, dentate nucleus, optic radiations, and pyramidal plemental methods (links.lww.com/NXG/A257). tracts in the posterior limb of the internal capsule, along with – cerebellar atrophy and thinning of the corpus callosum.4 6 Genetic analysis Variants in POLR3A were identified by exome sequencing POLR3-HLD is caused by biallelic pathogenic variants in using genomic DNA extracted from blood samples, according POLR3A, POLR3B, POLR1C,orPOLR3K, which encode to standard protocols. Variants were validated by Sanger se- subunits of RNA polymerase III (POLR3), an enzyme re- quencing and analyzed for familial segregation when DNA sponsible for transcription of several noncoding RNAs (nc- was available. RNAs), including transfer RNAs (tRNAs), 5S ribosomal RNA, U6 small nuclear RNA, 7S RNAs, and other small nu- Cell culture and cycloheximide treatment – cleolar RNAs.7 15 The precise mechanism underlying the To evaluate the presence of nonsense mediated decay (NMD), pathogenesis of hypomyelination remains to be fully eluci- fibroblasts derived from P2 were subjected to treatment with dated; 2 main mechanistic hypotheses include (1) defects in cycloheximide. Experimental details are described in supple- transcription capability of POLR3 causing disruptions in mental methods (links.lww.com/NXG/A257). tRNA levels, thereby altering global translation during myelination, which require large production of essential myelin Western blot proteins, or (2) impairments in specific POLR3-transcribed Immunoblots were performed using brain tissue protein nc-RNAs required for myelin development.7,10,16 extracts of P2 and an age/sex-matched control. Detailed protocols are outlined in supplemental methods (links.lww. Here, we expand the phenotypic spectrum of POLR3-HLD com/NXG/A257). through description of clinical, radiologic, and molecular features of six patients with an extremely severe phenotype and present Data availability functional and neuropathologic investigations on one patient. Data supporting this study’s findings are available on reasonable request. Raw data from participants (i.e., raw genetic data and MRI data sets) are not made publicly available to Methods protect patient privacy. Patients and study design An international cross-sectional study was performed be- Results tween 2016 and 2019, including a retrospective chart review of 6 patients (P1-6) from 5 families with atypical phenotypes Clinical characteristics identified from a repository of genetically proven POLR3- Patients 1–6 (P1-6) presented during infancy, between ages 1 HLD patients. and 3 months, with prominent feeding difficulties and failure to thrive. They exhibited severe developmental delay and Standard protocol approvals, registrations, motor regression before age 1 year. None achieved in- and patient consents dependent walking. Clinical characteristics are summarized in This research was approved by the Montreal Children’s table 1 and table e-1 (links.lww.com/NXG/A257). Hospital and McGill University Health Center Research Ethics Boards (11-105-PED; 2019-4972). Informed consent Of the 6 patients, 3 (3/6, 50%) had laryngomalacia and 2 was obtained from all patients or legal guardians. underwent supraglottoplasty. All had dysphagia and required

2 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG enteral tube feeding, with 5 (5/6, 83%) requiring a gastro- was not seen in any studies; however, mild to severe supra- stomy or gastrojejunostomy tube placement between ages 5 tentorial atrophy was present in all cases (figure 1, G–H). No and 15 months. Four patients (4/6, 67%) developed severe signs of pituitary involvement were noted. respiratory insufficiency, and 3 required supplemental oxygen and/or noninvasive respiratory support between ages 5 and Neuropathology 15 months, with 1 later having a tracheostomy at age 13 Preserved brain tissue of P2, who died at age 13 months from months. In addition, 2 patients (2/6, 33%) had suspected respiratory complications, was subjected to neuropathologic paroxysmal episodes of dysautonomia, with excessive sweat- study (figure 2). The brain weighed 777 g, below expected ing and retching. brain weight and comparable to typical weight at age 8 months.20 Macroscopic examination revealed normal sym- Non-neurologic features typical of POLR3-HLD included metry with well-formed cerebral hemispheres and cerebellum delayed dentition (3/6, 50%) and ophthalmologic abnor- (figure 2A). On gross examination, white matter was slightly malities, including hyperopia and cortical visual impairment reduced, but demonstrated normal appearance without gray (4/6, 67%). All patients were too young for hypogonado- discoloration or cavitation. The lateral ventricles and cere- tropic hypogonadism to be appreciated. bellum had normal size (figure 2B), and corpus callosum thickness was normal for age. Neurologic examination revealed acquired microcephaly in 4 patients (4/6, 67%). Five (5/6, 83%) had a combination of Histologic analysis of the neocortex and hippocampus axial hypotonia and upper motor neuron signs (spasticity revealed some ischemic neurons because of the final hypoxic- and/or hyperreflexia) in the limbs. Generalized dystonia and/ ischemic injury preceding death. No mineralization of cortical or chorea was seen in all patients. Restricted upgaze and ab- neurons or evidence of inflammatory infiltrate, necrosis, or normal saccades were occasionally noted. Two patients microglial nodules was present. exhibited hypomimia. White matter demonstrated patchy areas of rarefaction with Progressive decline and respiratory complications led to the mild myelin pallor. Oligodendrocytes showed normal mor- death of P1, P2, and P3 before age 2 years and P4 at age 3 phology and density in all studied areas, including pale areas, years. P5 and P6 are alive and currently aged 5 and 3 years, and features of demyelination were absent. White matter respectively. also exhibited diffuse astrocytic gliosis, both chronic (fibrillary) and subacute (protoplasmic), with activation of Radiologic characteristics microglia but without macrophagic changes associated with Brain MRI characteristics of P1-6 are summarized in table 2 phagocytic activity (figure 2, C–F). Changes in white matter and figure 1, which compares a typical POLR3-HLD MRI to appeared more severe in the parietal lobes (figure 2, C–D). P3. All 10 studies available for the 6 patients showed evidence No Rosenthal fibers or axonal spheroids were seen. Immu- of insufficient myelin deposition, but criteria for diffuse nohistochemistry did not reveal any axonal lesions. The hypomyelination were not met (figure 1, E–K).6,17 Overall, corpus callosum and corticospinal tracts demonstrated there was more myelin than usually seen in POLR3-HLD and normal myelination. additional distinctive MRI characteristics. T2 hyperintensity of the hilus of the dentate nucleus, associated with T2 Cellular vacuolization was seen in the thalamus and basal hypointensity (preserved myelination) of the dentate nucleus ganglia. Atrophy of the putamen was evident with enlarged itself and peridentate region, was seen in all studies (figure Virchow-Robin spaces and severe neuronal loss, associated 1F). In 9/10 studies (90%), the posterior brainstem exhibited with both chronic and subacute diffuse gliosis, along with rare similar features, with T2 hyperintensity (decreased myelin calcifications and considerable activation of microglia (figure content) of the posterior medulla, posterior-inferior pons, and 2, G–I). Discrete neuronal loss was evident in the caudate. posterior aspect of the middle cerebellar peduncles, in a pat- Within the pallidum, numerous pale nuclei of Alzheimer type tern suggestive of axonal degeneration (figure 1, F, I, and J). II glia were present due to the terminal anoxia, and no ap- The latest imaging studies of 2 patients (2/6, 33%), obtained preciable neuronal loss was evident. The adenohypophysis at ages 10 and 11 months, also revealed T2 hyperintensity of did not demonstrate pathologic abnormalities. the red nucleus (figure 1K). In addition, 8/10 studies (80%) revealed abnormal signal of the lentiform nuclei, which Hemisections of the brainstem demonstrated mild to mod- appeared hyperintense on T2 sequences compared with gray erate pyknosis in the pons and olivary nuclei of the medulla, matter and isointense to unmyelinated white matter. The consistent with acute ischemic changes. Patchy areas of re- same 8 studies also showed atrophy of the thalami (figure duced myelin were seen in the brainstem. The cerebellum 1G). The 2 scans without these findings were the 2 earliest demonstrated severe lesions of poorly myelinated white studies (P2, age 2 months; P3, age 3 months); however, matter, with diffuse and mainly chronic (fibrillary) gliosis, but follow-up MRIs showed that these changes developed over without notable morphological changes in oligodendrocytes time. Basal ganglia atrophy was seen only in the 5 latest scans (figure 2, J–L). The cerebellar cortex and dentate nucleus obtained between ages 10 and 15 months. Cerebellar atrophy appeared normal, and Bielschowsky staining did not reveal

Table 1 Clinical, MRI, molecular, and pathologic features associated with the typical and severe POLR3-related leukodystrophy phenotypes

Feature Typical phenotype Severe phenotype

Clinical characteristics

Age at onset 3–4y 1–3mo

Age of death Adulthood 1–3 y (2/6 patients still alive)

Symptoms at onset Developmental delay and motor regression Failure to thrive and developmental delay

Developmental delay Mild to moderate Severe

Dysphagia Late Early and severe

Respiratory End of disease course Early and severe insufficiency

Severe myopia Very common Too young

Dental abnormalities Common Delayed dentition seen in 3/6

Hypogonadotropic Common Too young hypogonadism

Brain MRI Hypomyelination with preservation of specific structures, thinning of the corpus Very atypical: more myelin than typical phenotype, supratentorial atrophy, and additional features callosum, and cerebellar atrophy including progressive abnormalities of the basal ganglia and thalami

Genetics POLR3A, POLR3B, POLR1C,orPOLR3K biallelic pathogenic variants POLR3A (NM_007055.3) >200 variants Compound heterozygous Allele 1: P1: c.2119C>T, p.Q707* P2: c.1681C>T, p.R561* P3: c.1051C>T, p.R351* P4: c.1051C>T, p.R351* P5: c.601delA, p.I201Lfs*18 P6: c.3583delG, p.D1195Ifs*47 Allele 2: P1-6: c.1771-7C>G

Pathology Prominent and diffuse decreased myelin, secondary axonal loss, and relative Patchy areas of decreased myelin, neuronal loss in the putamen and caudate, and vacuolization in the preservation of myelin in perivascular regions26 thalamus and basal ganglia erlg.r/GNuooy eeis|Vlm ,Nme ue2020 June | 3 Number 6, Volume | Genetics Neurology: Neurology.org/NG

Table 2 MRI features of patients with the severe POLR3-related leukodystrophy phenotype

Typical POLR3-HLD characteristics Additional atypical characteristics

Classic T2 hyperintense Age at T2 hypointensity Thin More myelin than T2 hyperintense posterior- Basal MRI Diffuse of specific corpus Cerebellar typically seen in hilus of the T2 hyperintense inferior T2 hyperintense Thalami ganglia Supratentorial ID (mo) hypomyelination structures callosum atrophy POLR3-HLD dentate nuclei lentiform nuclei brainstem red nuclei atrophy atrophy atrophy

Patient 6 −−+/mild − ++++− + − +/mild 1

15 −−+/mild − ++++− + + +/moderate- severe

Patient 2 −−+ − ++−− −−−+/mild 2

7 −−+ − ++++− + − +/mild- moderate

Patient 3 −−+/mild − ++− + −−−+/mild 3

10 −−+/mild − + + + + + + + +/mild- moderate

Patient 8 −−+/mild − ++++− + − +/mild- 4 moderate

11 −−+/mild − + + + + + + + +/mild- moderate

Patient 14 −−+ − ++++− + + +/moderate- 5 severe

Patient 10 −−+/thin − ++++− + + +/moderate 6 isthmus 5 Figure 1 MRI characteristics

Sagittal T1-weighted (A, E) and axial T2-weighted (B–D, F–K) images. (A–D) Typical POLR3-HLD; MRI obtained at age 6 years. Hypomyelination with relative preservation (T2 hypointensity) of the dentate nucleus (red arrow; B), anterolateral nucleus of the thalamus (double-lined arrow; C), optic radiations (arrowhead; C), globus pallidus, and corticospinal tracts in the posterior limb of the internal capsule (not shown). Thinning of the corpus callosum and cerebellar atrophy are also seen. (E–K) Severe phenotype; MRI of patient 3 obtained at age 10 months. Mild insufficient myelin deposition, not meeting the criteria for diffuse hypomyelination. Loss of myelin (T2 hyperintensity) in the posterior brainstem (red arrows; F, I, J), red nucleus (red dashed arrow; K), and hilus of the dentate nucleus (double-lined arrow; F). Abnormal signal of the lentiform nucleus (arrowhead; G). Supratentorial atrophy (G–H) and diffuse atrophy of the basal ganglia and thalami (G) are also seen.

clear evidence of decreased axons. Moderate pyknosis was revealed the presence of 2 aberrant transcripts resulting from seen in Purkinje cells; however, there was no appreciable loss abnormal splicing, including one lacking exon 14 causing of neurons. In the spinal cord, patchy areas of reduced myelin a frameshift and premature stop codon (p.P591Mfs*9) and the were noted. other lacking exons 13–14 causing loss of amino acids 548–637 (p.G548_Y637del). In addition, sequencing of the band corre- Genetic findings sponding to complementary DNA of wild-type length revealed Each patient harbored a specific combination of com- the presence of both the nonsense transcript (c.1681C>T/ pound heterozygous variants, including a variant causing a pre- p.R561*) and wild-type transcript, confirming the splice site mature stop codon on one allele (P1: c.2119C>T/p.Q707*, variant is leaky (figure 3B). Thus, 4 transcripts were detected, P2: c.1681C>T/p.R561*, P3&4: c.1051C>T/p.R351*, with sequences corresponding to (1) wild-type, (2) the non- P5: c.601delA/p.I201Lfs*18, P6: c.3583delG/p.D1195Ifs*47) sense variant, and those resulting from aberrant splicing events and a specific intronic splicing variant on the other (P1-6: including (3) lack of exon 14, and (4) lack of both exons 13–14 c.1771-7C>G). We hypothesized that this splicing variant was (figure 3C). leaky as complete absence of POLR3A is incompatible with life. PCR amplification using complementary DNA from fibroblasts As transcripts containing nonsense variants are typically of P2 revealed 2 additional bands compared with controls (figure targeted for NMD, we hypothesized that the c.1681C>T/ 3, figure e-1, links.lww.com/NXG/A257). Sequencing of bands p.R561* variant transcript was subjected to degradation. We

6 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG Figure 2 Neuropathology of the POLR3-HLD severe phenotype (patient 2)

(A–B) Macroscopic appearance of the (A) right cerebral hemisphere, and (B) coronal sections showing a slight decrease of the volume of the white matter without appreciable ventricular enlargement. (C) Luxol fast blue–cresyl violet (Kluver-Barrera)¨ staining demonstrating areas of poor myelination in the parietal white matter, but a normally myelinated corpus callosum. (D) Higher magnification of poorly myelinated white matter (10×), and (E) GFAP IHC revealing astrocytic gliosis (20×). (F) IBA1 IHC revealing activated microglia of the occipital white matter (10×). (G–I) Hemalun-phloxin staining revealing abnormalities of the putamen including (G) enlarged Virchow-Robin spaces (10×), (H) neuronal loss and gliosis with few calcifications (20×), and (I) neuronal death (60×). (J) Luxol fast blue–cresyl violet–stained section of the cerebellum revealing hypomyelination of the cerebellar white matter surrounding the dentate nucleus. (K) GFAP IHC demonstrating gliosis in the cerebellar white matter (20×), and (L) IBA1 IHC revealing activated microglia (10×). GFAP = glial fibrillary acidic protein; IHC = immunohistochemistry. evaluated the presence of NMD in P2 ﬁbroblasts compared Because complete lack of POLR3A is incompatible with life, we with a control using cycloheximide, a compound that sought to determine whether the detected residual wild-type inhibits transcriptional elongation and consequently NMD. transcript would lead to wild-type protein expression. We Following cycloheximide treatment, an increase in band 1 performed immunoblot analysis on protein extracts from fro- (corresponding to the wild-type transcript and nonsense zen brain tissue of P2 and an age/sex-matched control. To variant transcript) was observed by semiquantitative PCR ensure detection of only wild-type full-length protein, we chose (ﬁgure 3B), indicating that the nonsense transcript is sub- a POLR3A antibody with an epitope spanning amino acid jected to NMD under normal conditions. residues 607–698. In P2, this antibody cannot bind to the

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 7 abnormal protein products as the epitope binds to residues aspecific MRI pattern (i.e., more myelin than the typical located in the truncated POLR3A region, i.e., after the pre- phenotype, with progressive basal ganglia involvement). mature stop codon (p.R561*) and contained/semicontained in the deleted residues resulting from the splicing variant Neuropathologic examination revealed areas of reduced my- (p.P591Mfs*9, p.G548_Y637del). Thus, this antibody only elin in the brainstem. On MRI, all studies but one showed allows detection of wild-type POLR3A (figure 3C, figure e-1, evidence of decreased myelin in specific posterior-inferior links.lww.com/NXG/A257). We observed reductions in aver- brainstem structures. Wallerian degeneration affecting spe- age normalized POLR3A levels both in brain gray matter cific tracts could at least partly explain these findings, although (84.7% reduction, 95% CI = 69.3%–100%, d = 1.28) and white no clear axonal loss was documented on postmortem studies. matter (54.8% reduction, 95% CI = 20.1%–89.5%, d = 1.34) of The dentate nuclei appeared normal on neuropathologic P2 compared with control (figure 3, D and E). Gray matter analyses, consistent with the MRI pattern of preservation of displayed a greater reduction in POLR3A compared with white the dentate nuclei and peridentate region. On MRI, reduced matter (average difference 29.9%; 95% CI = 0.7%–59.0%; myelin was restricted to the hilus. d=1.77). Although it is well known that hypomyelination is not obligate in POLR3-HLD,18,25 the discrepancy between the relatively Discussion mild insufficient myelin deposition and the diffuse supratentorial atrophy was highly unusual and consistent across all Here, we present an expanded spectrum of POLR3-HLD MRIs. Although previous studies have revealed that oligoden- through description of 6 patients with a very severe phenotype drocytes are primarily affected in the typical form of POLR3- and similar genotype. The dramatic clinical presentation, in- HLD,1,26 our patients’ MRI and pathologic findings support the cluding prominent feeding and breathing difficulties and early hypothesis that the severe form is primarily neuronal, with death in 4 patients, is strikingly different from the typical POLR3- associated myelination deficits. We hypothesize that the path- HLD phenotype. A large phenotypic study of POLR3-HLD ophysiology associated with the severe phenotype varies sub- revealed typical onset at age 3–4 years with mild to moderate stantially from typical POLR3-HLD and involves several neural motor delay and/or regression.1 Dysphagia and respiratory in- cell types. As myelination is a complex process involving sufficiency were late findings. Death typically occurred in adult- a multitude of signaling events between neurons and glia, it is hood, where the youngest to die was aged 8 years.1 possible that an increased disruption of POLR3 activity, or the production of aberrant transcripts, could manifest adversely in fi The MRI pattern associated with this phenotype is distinct; more cell types than in a milder de cit. It is known that dys- despite very severe clinical manifestations, all patients had regulation of transcription and translation-related genes is often notably more myelin with different imaging features than associated with neurologic involvement, highlighting the im- typical POLR3-HLD. An evolving change in signal pattern portance of precise protein expression regulation during neural 27–30 was seen in the lentiform nuclei, with thalami atrophy, development. For example, defects in genes encoding ff aminoacyl-tRNA synthetases cause a variety of phenotypes, progressing to more di use basal ganglia atrophy. This 31–38 correlated with the prominent basal ganglia and thalami ranging from hypomyelination to brain malformations. pathologic abnormalities, including atrophy, calcifications, Given the broad clinical spectrum of phenotypes associated and severe neuronal loss in the putamina. Two patients also with POLR3 deficiency, it is clear that pathogenic variants in had red nuclei signal abnormalities. Recently, a similar MRI POLR3 genes have distinct effects on various cellular phenotype was described in patients with a c.1771-7C>G or processes.25,39 Variants in POLR3A have been associated with c.1771-6C>G variant, in trans with a missense, nonsense, splice phenotypes ranging from spastic ataxia–related disorders to site, or synonymous variant.21,22 Clinical severity varied neonatal progeroid syndrome, whereas variants in POLR3B according to the trans POLR3A variant; patients homozygous have been associated with isolated hypogonadotropic hypo- for the splicing variant typically displayed a milder phenotype, gonadism, without hypomyelination or hypodontia, and a dis- whereas those harboring a trans loss of function variant dis- – tinct phenotype of cerebellar hypoplasia with endosteal 21 24 – played severe features with early onset. Of interest, sclerosis.25,39 41 patients homozygous or compound heterozygous for the c.1771-7C>G and/or c.1771-6C>G variants did not display In contrast to this extremely severe clinical presentation, white matter involvement and were described as only having we also identified 3 adults with a very mild phenotype and the the neuronal MRI features, including striatal involvement with same homozygous pathogenic POLR3B variant (c.1568T>A/ caudate nucleus and putamen atrophy, and occasional red p.V523E) in our patient cohort. These patients were all di- 21–24 nuclei signal abnormalities. We hypothesize that these agnosed incidentally in adolescence/adulthood, based on specific splicing variants cause a cell-specificeffect (i.e., basal brain MRI performed for unrelated reasons, or through ge- ganglia neurons) compared with other POLR3-HLD variants. netic investigation of typical POLR3-HLD affected relatives. This could explain why, when this variant is combined with They had minimal findings on neurologic examination, if any, a loss of function allele, patients with a severe phenotype have and MRI revealed milder findings than usually seen in

8 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG Figure 3 Molecular and protein level implications of pathogenic variants in patient 2

(A) Sanger sequencing results of RT-PCR products generated from patient 2 fibroblasts, as visualized by agarose gel electrophoresis in (B), in which 3separate bands were excised and sequenced. In band 1, the presence of the POLR3A wild-type transcript is detected, as well as the transcript containing the paternally inherited nonsense variant (c.1681C > T; p.R561*), confirming that the splice site variant is leaky. Sequencing of the 2 additional bands confirms that the maternally inherited splice site variant (c.1771-7C>G) causes production of 2 additional transcripts, including 1 transcript with a deletion of exon 14, which produces a new open reading frame that results in a premature stop codon (p.P591Mfs*9), and the other containing a deletion of exons 13–14, which leads to the loss of amino acids 548–637 (p.G548_Y637del). (B) RT-PCR products with primers in POLR3A exons 11 and 15 revealing 2 additional bands in patient 2 fibroblasts compared with control fibroblasts. Cycloheximide treatment shows a stabilization of the mRNA containing the nonsense variant (band 1), confirming that it is targeted by NMD. β-Actin is shown as a loading control. (C) Schematic summary of each transcript detected following mRNA splicing. The starred region in the wild-type transcript denotes the POLR3A antibody epitope spanning from amino acids 607–698 for the immunoblots depicted in (D). (D) Immunoblots of protein lysates from frozen brain tissue of patient 2 (age 13 months) compared with that of an age/sex-matched control (age 14 months). Samples were collected from the subcorticalwhite matter (left) and the cortical gray matter (right). (E) Normalized expression of POLR3A in the brain of patient 2 compared to that in the control. Chemiluminescent intensity of the POLR3A signal at 164 kDa was normalized to the intensity of the β-tubulin signal at 51 kDa for each blot. Average values of normalized protein expression are derived from 4 Western blot replicates, and error bars represent standard error of the mean. Full-length POLR3A is detected in both control and patient 2 white and gray matter, with decreases seen in patient samples compared with the control. bp = base pairs; CHX = cycloheximide; Del = deletion; Ex = exon; mRNA = messenger RNA; NMD = nonsense medicated decay; P2 = patient 2; RT-PCR = reverse transcription PCR; WT = wild type.

POLR3-HLD. Two were previously described as having the of unknown etiology. She is also independent for all activities mildest phenotype in a past large cohort study of POLR3- of daily living and maintained an active role in the care of her HLD,1 and the third, who has not been reported, is an adult oﬀspring. These cases highlight the extreme variability in woman in her late 70s for whom limited information is disease severity of POLR3-HLD, which can range from very available. She is currently still ambulatory, and was able to mild to exceptionally severe. reproduce, making it unlikely she had fertility concerns due to hypogonadotropic hypogonadism. She is described as having Each patient with a severe clinical presentation had a similar mild intellectual challenges and hearing loss from childhood genotype, including a premature stop codon on one allele and

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 9 aspecific splicing variant (c.1771-7C>G) on the other. In our scholarships from Can-GARD (Canadian Gene Cure Advanced patient, we demonstrated that this variant is leaky and causes Therapies for Rare Disease) and R.S. McLaughlin and Teva alternative splicing events producing 2 aberrant transcripts, Canada Innovation Funds from the Faculty of Medicine, Uni- corresponding to results in a past study that investigated this versité Laval. M.A. Michell-Robinson is supported by the McGill variant in homozygous form.25 It is thought that this variant Faculty of Medicine (MD-PhD Program). T.B. Haack was creates a new enhancer binding site, and competition for en- supported by the German Bundesministerium für Bildung und hancer binding at either the native acceptor splice site (SRp40 Forschung through the Juniorverbund in der Systemmedizin enhancer protein) or aberrant binding site (SC35 enhancer “mitOmics” (FKZ01ZX1405C), the intramural fortüne program protein) is likely the cause of incomplete inactivation of the (#2435-0-0), and the Deutsche Forschungsgemeinschaft (Ger- native acceptor splice site and leaky production of the wild-type man Research Foundation) Projektnummer (418081722). transcript.23,25 We also confirmed that the transcript containing the nonsense variant was degraded by NMD. Moreover, as Disclosure POLR3 is a housekeeping gene, complete loss of its function is S. Perrier reports no disclosures. L. Gauquelin has received incompatible with life, which is further supported by the em- scholarships from Can-GARD (Canadian Gene Cure Ad- bryonic lethal Polr3a knock-out mouse.42 Thus, leaky expres- vanced Therapies for Rare Disease) and from the R.S. sion of some wild-type protein is not unexpected as all patients McLaughlin and Teva Canada Innovation Funds from the with a severe phenotype survived until early childhood. Al- Faculty of Medicine, Université Laval. C. Fallet-Bianco, M.K. though we were able to detect the production of some wild- Dishop, M.A. Michell-Robinson, L.T. Tran, K. Guerrero, L. type POLR3A protein in brain tissue of P2, protein levels were Darbelli, M. Srour, K. Petrecca, and D.L. Renaud report no significantly decreased, supporting the hypothesis that minimal disclosures. M. Saito has received compensation from Shire production of POLR3A is insufficient for proper neuro- Pharmaceutical (2017) for an unrelated study. S. Cohen development and growth. Less POLR3A protein was detected reports no disclosures. S. Leiz has received financial compen- in gray matter compared with white matter, lending further sation for lectures from Desitin, Hamburg. B. Alhaddad, T.B. support to our hypothesis that the severe phenotype is a pri- Haack, I. Tejera-Martin, and F.I. Monton report no disclosures. marily neuronal disorder. N. Rodriguez-Espinosa has received research grants from Gri- fols S.A. D. Pohl has received compensation for consulting, These findings illustrate an expanded phenotypic spectrum of presentations, and support for travel and conference atten- POLR3-HLD through presentation of patients with biallelic path- dance from Bayer-Schering, Biogen, Forward Pharma, Merck ogenic variants in POLR3A and an extremely severe phenotype. Serono, Sanofi, Teva, and Novartis. S. Nageswaran, A. Grefe, Identifying genotype-phenotype relationships advances our un- and E. Glamuzina report no disclosures. G. Bernard has re- derstanding of the disease course, providing valuable information ceived compensation for traveling to meetings and advisory for clinicians and allowing patients and families to have proper boards from Ionis, Shire/Takeda, Children’s Hospital of Phil- genetic counseling. Our functional and pathologic studies shed light adelphia, and Actelion Pharmaceuticals. She served on the on the pathogenesis of the severe form of POLR3-HLD, opening scientific advisory board for Ionis (2019) and has received the door for the development of targeted disease interventions. research grants from Shire/Takeda and Bluebird Bio. Go to Neurology.org/NN for full disclosures. Acknowledgment The authors thank the patients and their families for participating Publication history in this study. This work would not have been possible without Received by Neurology: Genetics November 6, 2019. Accepted in final support from The Yaya Foundation for 4H Leukodystrophy. The form March 25, 2020. authors acknowledge the McGill University and Genome Quebec Innovation Center. Appendix Authors

Study funding Name Location Contribution This study was supported by grants from the Foundation of Stefanie McGill University; Designed and Stars, Canadian Institutes of Health Research (CIHR; Perrier, HBSc Research Institute of the conceptualized the study; 201610PJT-377869), Fondation Les Amis d’Elliot, Fondation McGill University Health acquisition of data; ’ Centre, Montreal, QC, analysis and Lueur d Espoir pour Ayden, Fondation le Tout pour Loo, Canada interpretation of data; and Leuco-Action, and Réseau de Médecine Génétique Appliquée of drafted and revised the manuscript for intellectual the Fonds de Recherche en SantéduQuébec (FRQS). This content research was enabled in part by support provided by Compute Laurence McGill University, Designed and Canada (computecanada.ca). G. Bernard has received the CIHR Gauquelin, Montreal, QC, Canada; conceptualized the study; New Investigator Salary Award (2017–2022). S. Perrier is sup- MD, FRCPC The Hospital for Sick acquisition of data; Children, University of analysis and ported by the FRQS Doctoral Scholarship, Fondation du Grand Toronto, ON, Canada interpretation of data; and défi Pierre Lavoie Doctoral Scholarship, McGill Faculty of drafted and revised the Medicine F.S.B. Miller Fellowship, and RI-MUHC Desjardins manuscript for intellectual content Studentship in Child Health Research. L. Gauquelin has received

10 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG Appendix (continued) Appendix (continued)

Name Location Contribution Name Location Contribution

Catherine CHU Sainte-Justine, Acquisition of data; Ingrid Tejera- Hospital Universitario Acquisition of data and Fallet-Bianco, Universit´e de Montreal, analysis and Mart´ın, MD Nuestra Señora de revised the manuscript for MD QC, Canada interpretation of data; and Candelaria, Santa intellectual content drafted and revised the Cruz de Tenerife, manuscript for intellectual Canary Islands, content Spain

Megan K. Phoenix Children’s Acquisition of data; Fernando I. Hospital Universitario Acquisition of data and Dishop, MD Hospital, AZ analysis and Monton, MD, Nuestra Señora de revised the manuscript for interpretation of data; and PhD Candelaria, Santa Cruz de intellectual content revised the manuscript for Tenerife, Canary Islands, intellectual content Spain

Mackenzie A. McGill University; Acquisition of data; analysis Norberto Hospital Universitario Acquisition of data and Michell- Research Institute of the and interpretation of data; Rodr´ıguez- Nuestra Señora de revised the manuscript for Robinson, McGill University Health and drafted and revised the Espinosa, MD Candelaria, Santa Cruz de intellectual content MSc Centre, Montreal, QC, manuscript for intellectual Tenerife, Canary Islands, Canada content Spain

Luan T. Tran, McGill University; Research Acquisition of data and Daniela Pohl, Children’s Hospital of Acquisition of data and MSc Institute of the McGill revised the manuscript for MD, PhD Eastern Ontario, revised the manuscript for University Health Centre, intellectual content University of Ottawa, intellectual content Montreal, QC, Canada Canada

Kether McGill University; Research Acquisition of data; Savithri Wake Forest School of Acquisition of data and Guerrero, Institute of the McGill analysis and Nageswaran, Medicine, Winston-Salem, revised the manuscript for MSc University Health Centre, interpretation of data; and MBBS, MPH NC intellectual content Montreal, QC, Canada revised the manuscript for intellectual content Annette Wake Forest School of Acquisition of data and Grefe, MD Medicine, Winston-Salem, revised the manuscript for Lama McGill University; Acquisition of data; NC intellectual content Darbelli, PhD Research Institute of the analysis and McGill University Health interpretation of data; and Emma Starship Children’s Acquisition of data and Centre, Montreal, QC, drafted and revised the Glamuzina, Hospital, Auckland, New revised the manuscript for Canada manuscript for intellectual MD Zealand intellectual content content Geneviève McGill University; Designed and Myriam McGill University; Research Acquisition of data and Bernard, MD, Research Institute of the conceptualized the study; Srour, MDCM, Institute of the McGill revised the manuscript for MSc FRCPC McGill University Health acquisition of data; analysis PhD University Health Centre, intellectual content Centre; Montreal and interpretation of data; Montreal, QC, Canada Children’s Hospital and revised the manuscript for McGill University Health intellectual content; and Kevin McGill University; Brain Acquisition of data and Centre, QC, Canada study supervision Petrecca, MD, Tumour Research Center revised the manuscript for PhD, FRCSC Montreal Neurological intellectual content Institute and Hospital, QC, Canada References Deborah L. Mayo Clinic, Rochester, Acquisition of data and 1. Wolf NI, Vanderver A, van Spaendonk RM, et al. Clinical spectrum of 4H leuko- Renaud, MD MN revised the manuscript for dystrophy caused by POLR3A and POLR3B mutations. Neurology 2014;83: intellectual content 1898–1905. 2. Timmons M, Tsokos M, Asab MA, et al. Peripheral and central hypomyelination with Michael University of California Acquisition of data and hypogonadotropic hypogonadism and hypodontia. Neurology 2006;67:2066–2069. Saito, MD Riverside School of revised the manuscript for 3. Bernard G, Vanderver A. POLR3-Related leukodystrophy. In: Adam MP, Ardinger Medicine; Riverside Medical intellectual content HH, Pagon RA, et al, editors. GeneReviews. Seattle, WA: University of Washington, Clinic, CA Seattle; 2017. 4. Vrij-van den Bos S, Hol JA, La Piana R, et al. 4H Leukodystrophy: a brain magnetic Seth Cohen, Beaver Medical Group, Acquisition of data and resonance imaging scoring system. Neuropediatrics 2017;48:152–160. MD Redlands, CA revised the manuscript for 5. La Piana R, Tonduti D, Gordish Dressman H, et al.. Brain magnetic resonance intellectual content imaging (MRI) pattern recognition in Pol III-related leukodystrophies. J Child Neurol 2014;29:214–220. Steffen Leiz, Klinikum Dritter Orden, Acquisition of data and 6. Steenweg ME, Vanderver A, Blaser S, et al.. Magnetic resonance imaging pattern MD Munich, Germany revised the manuscript for recognition in hypomyelinating disorders. Brain 2010;133:2971–2982. intellectual content 7. Bernard G, Chouery E, Putorti ML, et al. Mutations of POLR3A encoding a catalytic subunit of RNA polymerase Pol III cause a recessive hypomyelinating leukodystro- Bader Technische Universität Acquisition of data and phy. Am J Hum Genet 2011;89:415–423. Alhaddad, Munchen,¨ Munich, revised the manuscript for 8. Saitsu H, Osaka H, Sasaki M, et al. Mutations in POLR3A and POLR3B encoding MD Germany intellectual content RNA Polymerase III subunits cause an autosomal-recessive hypomyelinating leukoencephalopathy. Am J Hum Genet 2011;89:644–651. Tobias B. Technische Universität Acquisition of data and 9. Daoud H, Tétreault M, Gibson W, et al. Mutations in POLR3A and POLR3B are Haack, MD Munchen,¨ Munich, revised the manuscript for a major cause of hypomyelinating leukodystrophies with or without dental abnor- Germany; Institute of intellectual content malities and/or hypogonadotropic hypogonadism. J Med Genet 2013;50:194–197. Medical Genetics and 10. Thiffault I, Wolf NI, Forget D, et al. Recessive mutations in POLR1C cause a leuko- Applied Genomics, dystrophy by impairing biogenesis of RNA polymerase III. Nat Commun 2015;6:7623. University of Tubingen,¨ 11. Tétreault M, Choquet K, Orcesi S, et al. Recessive mutations in POLR3B, encoding Germany the second largest subunit of Pol III, cause a rare hypomyelinating leukodystrophy. Am J Hum Genet 2011;89:652–655.

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 11 12. Gutierrez M, Thiffault I, Guerrero K, et al. Large exonic deletions in POLR3B gene 29. Lee TI, Young RA. Transcriptional regulation and its misregulation in disease. Cell cause POLR3-related leukodystrophy. Orphanet J Rare Dis 2015;10:69. 2013;152:1237–1251. 13. Dorboz I, Dumay-Odelot H, Boussaid K, et al. Mutation in POLR3K causes hypomyelinating 30. Tahmasebi S, Khoutorsky A, Mathews MB, Sonenberg N. Translation deregulation in leukodystrophy and abnormal ribosomal RNA regulation. Neurol Genet 2018;4:e289. human disease. Nat Rev Mol Cell Biol 2018;19:791–807. 14. Jasiak AJ, Armache KJ, Martens B, Jansen RP, Cramer P. Structural biology of RNA 31. Mendes MI, Gutierrez Salazar M, Guerrero K, et al. Bi-allelic mutations in EPRS, polymerase III: subcomplex C17/25 X-ray structure and 11 subunit enzyme model. encoding the glutamyl-prolyl-aminoacyl-tRNA synthetase, cause a hypomyelinating Mol Cell 2006;23:71–81. leukodystrophy. Am J Hum Genet 2018;102:676–684. 15. Dieci G, Fiorino G, Castelnuovo M, Teichmann M, Pagano A. The expanding RNA 32. Taft RJ, Vanderver A, Leventer RJ, et al. Mutations in DARS cause hypomyelination polymerase III transcriptome. Trends Genet 2007;23:614–622. with brain stem and spinal cord involvement and leg spasticity. Am J Hum Genet 16. Choquet K, Forget D, Meloche E, et al. Leukodystrophy-associated POLR3A 2013;92:774–780. mutations down-regulate the RNA polymerase III transcript and important regulatory 33. Wolf NI, Salomons GS, Rodenburg RJ, et al. Mutations in RARS cause hypomyeli- RNA BC200. J Biol Chem 2019;294:7445–7459. nation. Ann Neurol 2014;76:134–139. 17. Schiffmann R, van der Knaap MS. Invited article: an MRI-based approach to the 34. Feinstein M, Markus B, Noyman I, et al. Pelizaeus-Merzbacher-like disease diagnosis of white matter disorders. Neurology 2009;72:750–759. caused by AIMP1/p43 homozygous mutation. Am J Hum Genet 2010;87: 18. La Piana R, Cayami FK, Tran LT, et al. Diffuse hypomyelination is not obligate for 820–828. POLR3-related disorders. Neurology 2016;86:1622–1626. 35. Zhang X, Ling J, Barcia G, et al. Mutations in QARS, encoding glutaminyl-tRNA 19. Cayami FK, Bugiani M, Pouwels PJW, Bernard G, van der Knaap MS, Wolf NI. 4H synthetase, cause progressive microcephaly, cerebral-cerebellar atrophy, and in- leukodystrophy: lessons from 3T imaging. Neuropediatrics 2018;49:112–117. tractable seizures. Am J Hum Genet 2014;94:547–558. 20. Schulz DM, Giordano DA, Schulz DH. Weights of organs of fetuses and infants. Arch 36. Accogli A, Russell L, Sebire G, et al. Pathogenic variants in AIMP1 cause pontocer- Pathol 1962;74:244–250. ebellar hypoplasia. Neurogenetics 2019;20:103–108. 21. Hiraide T, Kubota K, Kono Y, et al. POLR3A variants in striatal involvement without 37. Ognjenovic J, Simonovic M. Human aminoacyl-tRNA synthetases in diseases of the diffuse hypomyelination. Brain Dev 2020;42:363–368. nervous system. RNA Biol 2018;15:623–634. 22. Harting I, Al-Saady M, Krageloh-Mann I, et al. POLR3A variants with striatal in- 38. Friedman J, Smith DE, Issa MY, et al. Biallelic mutations in valyl-tRNA synthetase volvement and extrapyramidal movement disorder. Neurogenetics 2020;21:121–133. gene VARS are associated with a progressive neurodevelopmental epileptic enceph- 23. Azmanov DN, Siira SJ, Chamova T, et al. Transcriptome-wide effects of a POLR3A alopathy. Nat Commun 2019;10:707. gene mutation in patients with an unusual phenotype of striatal involvement. Hum 39. Jay AM, Conway RL, Thiffault I, et al. Neonatal progeriod syndrome associated Mol Genet 2016;25:4302–4314. with biallelic truncating variants in POLR3A. Am J Med Genet A 2016;170: 24. Wu S, Bai Z, Dong X, et al. Novel mutations of the POLR3A gene caused POLR3-related 3343–3346. leukodystrophy in a Chinese family: a case report. BMC Pediatr 2019;19:289. 40. Wambach JA, Wegner DJ, Patni N, et al. Bi-allelic POLR3A loss-of-function variants 25. Minnerop M, Kurzwelly D, Wagner H, et al. Hypomorphic mutations in POLR3A are cause autosomal-recessive wiedemann-rautenstrauch syndrome. Am J Hum Genet a frequent cause of sporadic and recessive spastic ataxia. Brain 2017;140:1561–1578. 2018;103:968–975. 26. Vanderver A, Tonduti D, Bernard G, et al. More than hypomyelination in Pol-III 41. Richards MR, Plummer L, Chan YM, et al. Phenotypic spectrum of POLR3B disorder. J Neuropathol Exp Neurol 2013;72:67–75. mutations: isolated hypogonadotropic hypogonadism without neurological or dental 27. Abbott JA, Francklyn CS, Robey-Bond SM. Transfer RNA and human disease. Front anomalies. J Med Genet 2017;54:19–25. Genet 2014;5:158. 42. Choquet K, Yang S, Moir RD, et al. Absence of neurological abnormalities in mice 28. Kapur M, Monaghan CE, Ackerman SL. Regulation of mRNA translation in neurons: homozygous for the Polr3a G672E hypomyelinating leukodystrophy mutation. Mol a matter of life and death. Neuron 2017;96:616–637. Brain 2017;10:13.

12 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG ARTICLE OPEN ACCESS Phenotypic variability in chorea-acanthocytosis associated with novel VPS13A mutations

Valter Niemel¨a, MD, PhD, Ammar Salih, MD, Daniela Solea, MD, Bjorn¨ Lindvall, MD, Jan Weinberg, MD, PhD, Correspondence Gabriel Miltenberger, MD, Tobias Granberg, MD, PhD, Aikaterini Tzovla, MD, Love Nordin, PhD, Dr. Paucar [email protected] Torsten Danfors, MD, PhD, Irina Savitcheva, MD, PhD, Niklas Dahl, MD, PhD, and Martin Paucar, MD, PhD

Neurol Genet 2020;6:e426. doi:10.1212/NXG.0000000000000426

Abstract MORE ONLINE Video Objective To perform a comprehensive characterization of a cohort of patients with chorea- acanthocytosis (ChAc) in Sweden.

Methods Clinical assessments, targeted genetic studies, neuroimaging with MRI, [18F]-ﬂuorodeoxyglucose (FDG) PET, and dopamine transporter with 123I FP-CIT (DaTscan) SPECT. One patient underwent magnetic resonance spectroscopy (MRS).

Results Four patients living in Sweden but with different ethnical backgrounds were included. Their clinical features were variable. Biallelic VPS13A mutations were confirmed in all patients, including 3 novel mutations. All tested patients had either low or absent chorein levels. One patient had progressive caudate atrophy. Investigation using FDG-PET revealed severe bilateral striatal hypometabolism, and DaTscan SPECT displayed presynaptic dopaminergic deficiency in 3 patients. MRS demonstrated reduced N-acetylaspartate/creatine (Cr) ratio and mild elevation of both choline/Cr and combined glutamate and glutamine/Cr in the striatum in 1 case. One patient died during sleep, and another was treated with deep brain stimulation, which transiently attenuated feeding dystonia but not his gait disorder or chorea.

Conclusions Larger longitudinal neuroimaging studies with diﬀerent modalities, particularly MRS, are needed to determine their potential role as biomarkers for ChAc.

From the Department of Neurology (V.N.), Uppsala University Hospital; Department of Neurology (A.S.), Västerås Hospital, Sweden; Department of Neurology (D.S.), Gävle Hospital; Department of Neurology (B.L.), University Hospital in Orebro;¨ Department of Neurology (J.W., M.P.), Karolinska University Hospital, Stockholm, Sweden; Department of Neurology (G.M.), Ludwig-Maximilians-Universität Munchen,¨ Munich, Germany; Department of Clinical Neuroscience (T.G., M.P.), Karolinska Institutet, Stockholm; Department of Radiology (T.G., A.T.), Karolinska University Hospital, Stockholm; Department of Diagnostic Medical Physics (L.N.), Karolinska University Hospital Solna, Stockholm; Division of Clinical Geriatrics (L.N.), Department of Neurobiology, Care Sciences and Society, Karolinska Institutet, Stockholm; Department of Surgical Sciences (T.D.), Section for Nuclear Medicine and PET, Uppsala University Hospital; Department of Medical Radiation Physics and Nuclear Medicine (I.S.), Karolinska University Hospital, Stockholm; and Department of Immunology, Genetics and Pathology (N.D.), Science for Life Laboratory, Uppsala University, Sweden.

Go to Neurology.org/NG for full disclosures. Funding information is provided at the end of the article.

Copyright © 2020 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 Glossary BG = basal ganglia; CC/IT = intercaudate distance to inner table width ratio; ChAc = chorea-acanthocytosis; Cho = choline; Cr = creatine; DBS = deep brain stimulation; FDG = ﬂuorodeoxyglucose; FH/CC = frontal horn width to intercaudate distance ratio; Glx = glutamine; HD = Huntington disease; MLS = McLeod syndrome; MRS = magnetic resonance spectroscopy; NAA = N-acetylaspartate.

Chorea-acanthocytosis (ChAc) and McLeod syndrome (MLS) Data availability are the main forms of neuroacanthocytosis disorders, charac- Data not provided in this article are available in a trusted re- terized by progressive and incurable hyperkinesias, neuro- pository (doi.org/10.5061/dryad.7h44j0zr9). muscular abnormalities, and a reduced life span. However, epilepsy is common in ChAc, whereas dilated cardiomyopathy is a distinctive feature in 2/3 of patients with MLS.1 Caudate Results atrophy is progressive in MLS but has not been described in ChAc2 caused by biallelic mutations in the vacuolar protein Phenotype and genetics sorting 13 homolog A (VPS13A) gene, encoding chorein.1 The clinical features and mutations are summarized in table 1. Briefly, the mean age at onset was 34 (range 30–38) years, and Currently, almost 1,000 cases with ChAc have been reported in – various populations. Here, we characterize 4 patients affected the disease duration was 9.5 (range 2 17) years. All patients by ChAc and 3 new truncating mutations in VPS13A. presented with cognitive decline of variable degree, and laboratory investigations revealed elevated creatine kinase levels and acanthocytes in blood smears. Feeding dystonia was Methods prominent in patient 1 and therefore treated with botulinum injections in the tongue base3 and later with bilateral deep The patients (2 males and 2 females) were evaluated at 4 dif- brain stimulation (DBS) of the internal globus pallidus that ferent centers in Sweden (table 1). Patients or their next-of-kin resulted in a transient attenuation of feeding dystonia. provided oral and written consent for this study, approved by the Western blot analysis revealed absent chorein in patient 1 and Regional Ethics Committee of Stockholm. Patients underwent low levels in patients 3 and 4. Chorea, dystonia, and vocal- clinical evaluation of motor features with Unified Huntington’s izations were initial features in patients 1 and 3, and their Disease Rating Scale, cognitive screening with Montreal Cog- severe gait disorder was overcome when walking backward. nitive Assessment, psychometric testing, laboratory tests, blood Both patients were compound heterozygous for VPS13A smears, and/or ancillary tests such as neurophysiologic tests and/ mutations. In patient 2, seizures were the first symptom, fol- or muscle biopsy. Analysis of the VPS13A gene was performed lowed by hyperkinesias and myopathy at later stages. Genetic following exclusion of a trinucleotide expansion in the huntingtin investigation revealed homozygosity for a VPS13A mutation. (HTT) gene. The ClinVar database (ncbi.nlm.nih.gov/clinvar) The patient died during sleep at age 42 years despite good wasusedtoassessnoveltyofVPS13A mutations, and Muta- seizure control, and autopsy revealed incidental sarcoidosis tionTaster (mutationtaster.org) was used to predict the effects of affecting the heart. Patient 3 was diagnosed with tourettism mutations.Westernblotforchoreininbloodwasperformedin during adolescence, followed by progressive hyperkinesias patients 1, 3, and 4 (Western blot analysis for chorein were and a gait disorder that became evident at age 30 years. performed with the financial support of the Advocacy for Neu- Symptoms progressed with generalized dystonia, pre- roacanthocytosis Patients in the laboratories of Drs. Bettina dominantly of the torso, and marked blepharospasm. Genetic Schmid (Biochemistry/DZNE) and Adrian Danek (Neurology) investigation revealed compound heterozygosity for VPS13A at Ludwig-Maximilians-Universität Munich, Germany). Neuro- mutations. A younger sibling of patient 3, not investigated for imaging included MRI, [18F]-fluorodeoxyglucose (FDG)PET/ VPS13A mutations, had seizures without motor abnormalities CT, and assessment of presynaptic dopamine transporter with and died during sleep. Patient 4 had chorea and depression; 123I-labeled N-(3-fluoropropyl)-2β-carbomethoxy-3β-(4-iodo- her genetic investigation revealed homozygosity for a VPS13A phenyl)nortropane (FP-CIT) DaTscan SPECT. The parame- mutation. Three of the 6 VPS13A mutations were new and ters evaluated in Huntington disease (HD), frontal horn width to pathogenic (table 1 and supplementary material, links.lww. intercaudate distance ratio (FH/CC) and intercaudate distance com/NXG/A255). The course of disease for patient 1 is to inner table width ratio (CC/IT), were measured. Magnetic shown in video 1; patients 3 and 4 declined video recordings resonance spectroscopy (MRS) was performed only in patient 1. of their examination. Briefly, the concentrations of N-acetylaspartate (NAA), choline (Cho), combined glutamate and glutamine (Glx), creatine (Cr), Neuroimaging and phosphocreatine were measured in the putamen and globus MRI displayed variable degrees of striatal atrophy in all pallidus. These results were compared with normative data from patients that was progressive in patient 3. Patient 1 had FH/ 20 healthy controls (supplementary material, links.lww.com/ CC = 1.6 (normal range: 2.2–2.6) and CC/IT = 0.18 (normal NXG/A255). range: 0.09–0.12) at age 37 years, with unchanged ratios 5

2 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG Table 1 Acanthocytes were present in all 3 patients, and chorein was absent in blood in patient 1

Case no. 1 2 3 4

Ethnicity Latin American Swedish Swedish Syrian Features and genetics Sex M F M F

Clinical course Age at onset 33 36 30 37

Current age or age at deathb 46 42b 47c 40

Symptom at onset Personality change and Epilepsy Gait disorder, facial Chorea and chorea chorea, and fatigue vocalizations

Motor features Predominant motor symptoms Chorea, dystonia, Chorea, postural Generalized dystonia Chorea, postural instability, and instability, oral dystonia, and severe including oral bradykinesiaa and vocal tics blepharospasm tics

Chorea + + + +

UHDRS (latest) 43 NA 55 25

Feeding dystonia + − (+) −

Cognitive and Cognitive impairment + + (+) + behavioral function

MoCA 28 NA NA 13

Psychiatric symptoms + Frontal disinhibition − + Tourettism + Depression

Neuromuscular Muscle weakness/atrophy + + ND − assessment

Reduced deep tendon reflexes + + + +

Neuropathy (confirmed with − NA −− ENeG)

Myopathy + + NA −

Other Epilepsy − + −− neurologic features

Laboratory CK (x upper reference value) + (15x) + (19x) + (4x) + (3x) testing (Ref. men 0.8–6.7 and women 0.6–3.5 μkat/L)

Elevated AST or ALT + + −−

Acanthocytes + + + +

Chorein Absent NA Low Low

Imaging studies MRI caudate atrophy + + + Progressive +

Reduced striatal uptake +NA++ (DaTscan)

Striatal hypometabolism (FDG- +NA++ PET)

Genetics Mutations in VPS13A c.266dupT; exon 51–59 Homozygous c.4162_ c.2428-2A>G; Homozygous del (p.Ile90Tyrfs*; p?) 4166 delins C c.1595+4_ c.7867C>T (p.Leu1388Glu-fsx6) 1595+7delAGTA (p?; (p.Arg2623X) p?)

Abbreviations: + = present; − = absent; (+) = mild feature; ALT = alanine transaminase; AST = aspartate aminotransferase; CK = creatine kinase; FDG = fluorodeoxyglucose; MoCA = Montreal Cognitive Assessment; NA = no assessed; ND = no determined; UHDRS = Unified Huntington’s Disease Rating Scale. All patients also had areflexia and hyperCKemia. Mutations for patients 2 and 3 are new. a Treated with bilateral deep brain stimulation of the internal globus pallidus. b Sudden unexpected death in epilepsy (SUDEP). Age at death was 42. c A younger brother to patient 3 had epilepsy at age 24 years and died during sleep at age 28 years; however, he did not show obvious movement abnormalities.

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 3 years later. MRS for patient 1 demonstrated a mild reduction dystonia and blepharospasm. The only previous MRS study of NAA/Cr ratio and mild elevation of Cho/Cr and Glx/Cr in in genetically confirmed ChAc demonstrated decreased the striatum (table e-1 and figure e-1, links.lww.com/NXG/ NAA/Cr ratio in the atrophic striatum of 2 brothers.8 This A255). Patient 3 had FH/CC = 1.5 and CC/IT = 0.20 at age abnormality reflects neuronal loss and gliosis. In contrast, 39 years. Six years later, the FH/CC was 1.2 and CC/IT 0.26, the myo-inositol/Cr ratio was increased in this part of the indicating progressive caudate atrophy (figure e-4). Patients 2 BG. MRS abnormalities have previously been described in and 4 had iron deposition in the globus pallidus. In patients 1, cortical areas of 5 patients with MLS, but abnormalities in 3, and 4, we observed profound reduction of glucose metab- thestriatumwerenotdetermined.11 In HD, MRS abnor- olism in the dorsal striatum (caudate nucleus and putamen) malities are widespread and depend on disease stage. Our and reduced dopamine transporter binding (figure 1 and results showing reduced NAA, reflecting neuronal loss and supplementary document). Patients 1 and 3 also had an in- gliosis, are in line with previous observations. The elevations creased glucose metabolism in the thalamus, whereas patient of Cho, reflecting increased cellular membrane turnover and 4 had a normal thalamic metabolism. An extended description glutamate/Glx, suggest a possible role of glutamate excito- on the clinical course and neuroimaging findings are found in toxicity in ChA as a potential biomarker in the pathophysi- the supplementary files (figures e-1 to e-5). ology of ChAc.

A common clinical feature is the progression into bradyki- Discussion nesia, observed in 2 of 4 patients. The low binding of DaTscan to dopamine transporter in the striatum is a rea- Here, we add to the number of VPS13A mutations associated with variable phenotypic expression in 4 cases of ChAc. sonable correlate to parkinsonism in patients 1 and 3; neuronal loss has been described in patients with ChAc with Similar to HD, structural neuroimaging for ChAc demon- 1 strated both caudate atrophy with dilatation of the anterior parkinsonism. On the other hand, reduced binding in patient horns and variable cortical atrophy.1,4 Other MRI findings 4 is likely a premonitory sign of hypokinesia and no a side ff include increased T2-weighted signal in the caudate and e ect of neuroleptic since DaTscan is a presynaptic tracer. 1 To date, there are only 5 reports on genetically confirmed putamen and hippocampal sclerosis and atrophy, in contrast 3,e2,e3,e8,e9 to some patients with ChAc with normal findings in the basal ChAc in Scandinavia. All 6 mutations reported ganglia (BG).5 In addition, iron accumulation in the BG in herein are truncating, and 3 of them are novel. This is con- patient 2 is in line with previous findings in ChAc1. Caudate sistent with previously reported VSP13A gene mutations, of atrophy is progressive in MLS,2 but this is, to our knowledge, which the majority predict a complete or partial loss of e3 the first report on progressive caudate atrophy in ChAc. protein function. Low or absent chorein levels in 3 of our patients strengthen the notion that the VPS13A variants are Consistent with previous studies on ChAc, we identified indeed pathogenic. The reasons for the wide clinical vari- 1 severe hypometabolism in the BG, loss of dopaminergic ability in ChAc remain elusive. Patient 3 presented with nigrostriatal projections, and dopamine transporter bind- tourettism during adolescence and was diagnosed with ChAc – ing sites in all assessed patients.6 9 Importantly, hypo- at age 30 years, followed by a disease course spanning 17 metabolism in the BG is not specificforChAc,asitoccurs years. This is remarkable considering that the mean disease also in MLS and HD. Furthermore, patients 1 and 3 had duration in ChAc is 11 years.e4 Furthermore, sudden un- increased metabolism in the thalamus. Whereas thalamic expected death in epilepsy (SUDEP) appears to be common hypermetabolism has been reported in patients with in ChAc and was recently reported in 6 of 52 patients.e4 blepharospasm,10 we noted that only patient 3 has generalized Patient 2 died during sleep at age 42 years, raising the

Figure 1. Neuroimaging findings in chorea-achantocytosis

Axial T2-weighted FLAIR MRI of patient 3 (at age 46 years) shows a pronounced atrophy of the caudate nuclei and putamina and increased signal intensity from the putamina (arrows) (A) DaTscan SPECT and fluorodeoxyglucose PET in the same patient at age 43 years demonstrate very low DAT density in both putamen (B, arrows) respectively severe hypometabolism in the striata (C, arrows). DAT = dopamine transporter; FLAIR = fluid-attenuated inversion recovery.

4 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG suspicion of SUDEP. However, it cannot be excluded that the presence of sarcoidosis in her heart could have predisposed to Appendix (continued) a lethal arrhythmia. Name Location Contribution

Although our report brings additional information to the Gabriel Department of Western blot analyses and Miltenberger, Neurology, Ludwig- revising the manuscript genetic and clinical variability in ChAc, the retrospective na- MD Maximilians-Universit¨at ture of our study precluded volumetric MRI analyses. Finally, Munchen¨ our study calls for further longitudinal neuroimaging studies Tobias Karolinska University Drafting and revising the using diﬀerent modalities to assess their potential role as Granberg, Hospital and Karolinska manuscript and analysis MD, PhD Institutet and interpretation of biomarkers in ChAc. neuroimaging data

Acknowledgment Aikaterini Karolinska University Drafting and revising the Tzovla, MD Hospital manuscript and analysis The authors are truly grateful to the patients and their next-of- and interpretation of kin for consenting for this report. The authors also thank neuroimaging data psychologist Å. Bergendal and Dr. G. Schechtmann for Love Nordin, Karolinska University Drafting and revising the cognitive evaluations respectively performing DBS surgery in PhD Hospital and Karolinska manuscript and analysis ’ Institutet and interpretation of MRS patient 1. M. Paucar s research is supported by Region data Stockholm. Torsten Uppsala University Drafting and revising the Danfors, MD, Hospital manuscript and analysis Study funding PhD and interpretation of PET M. Paucar is supported by Region Stockholm and N. Dahl by data the Swedish Research Council (2015-02424). Irina Karolinska University Drafting and revising the Savitcheva, Hospital manuscript and analysis MD, PhD and interpretation of PET Disclosure data V. Niemel¨a, A. Salih, D. Solea, B. Lindvall, J. Weinberg, G. Niklas Dahl, Uppsala University Revising the manuscript; Miltenberger, T. Granberg, A. Tzovla, L. Nordin, T. Danfors, MD, PhD Hospital study concept and design; I. Savitcheva, N. Dahl, and M. Paucar report no disclosures. analysis and interpretation of data; and study Go to Neurology.org/NG for full disclosure. supervision and coordination

Publication history Martin Karolinska University Revising the manuscript; Received by Neurology: Genetics October 16, 2019. Accepted in ﬁnal Paucar, MD, Hospital and Karolinska study concept and design; form March 27, 2020. PhD Institutet analysis and interpretation of data; and study supervision and coordination

Appendix Authors References Name Location Contribution 1. Velayos Baeza A, Dobson-Stone C, Rampoldi L, et al. Chorea-acanthocytosis. 2002. In: Adam MP, Ardinger HH, Pagon RA, et al, editors. GeneReviews® [Internet]. Valter Uppsala University Drafting and revising the Seattle, WA: University of Washington, Seattle; 1993-2019. Niemelä, MD, Hospital manuscript; study 2. Valko PO, Hänggi J, Meyer M, Jung HH. Evolution of striatal degeneration in PhD coordination; and analysis McLeod syndrome. Eur J Neurol 2010;17:612–618. and interpretation of data 3. Paucar M, Lindestad PÅ, Walker RH, et al. Teaching video NeuroImages: feeding dystonia in chorea-acanthocytosis. Neurology 2015;85:e143–e144. Ammar Salih, Västerås Hospital Drafting and revising the 4. Walterfang M, Looi JC, Styner M, et al. Shape alterations in the striatum in chorea- MD manuscript; patient care; acanthocytosis. Psychiatry Res 2011;192:29–36. and analysis and 5. Estévez-Fraga C, López-Sendón Moreno JL, Mart´ınez-Castrillo JC; Spanish Collab- interpretation of data orative Neuroacanthocytosis Group. Phenomenology and disease progression of chorea-acanthocytosis patients in Spain. Parkinsonism Relat Disord 2018;49:17–21. Daniela Solea, Gävle Hospital Drafting and revising the 6. Brooks DJ, Ibanez V, Playford ED, et al. Presynaptic and postsynaptic striatal dopa- MD manuscript; patient care; minergic function in neuroacanthocytosis: a positron emission tomographic study. and analysis and Ann Neurol 1991;30:166–171. interpretation of data 7. Müller-Vahl KR, Berding G, Emrich HM, et al. Chorea-acanthocytosis in monozygotic twins: clinical findings and neuropathological changes as detected by diffusion tensor Bjorn¨ University Hospital in Drafting and revising the imaging, FDG-PET and (123)I-beta-CIT-SPECT. J Neurol 2007;254:1081–1088. Lindvall, MD Orebro¨ manuscript; patient care; 8. Ismailogullari S, Caglayan AO, Bader B, et al. Magnetic resonance spectroscopy in two and analysis and siblings with chorea-acanthocytosis. Mov Disord 2010;25:2894–2897. interpretation of data 9. Cui R, You H, Niu N, et al. FDG PET brain scan demonstrated glucose hypometabolism of bilateral caudate nuclei and putamina in a patient with chorea-acan- Jan Weinberg, Karolinska University Drafting and revising the thocytosis. Clin Nucl Med 2015;40:979–980. MD, PhD Hospital manuscript; patient care; 10. Suzuki Y, Mizoguchi S, Kiyosawa M, et al. Glucose hypermetabolism in the thalamus and analysis and of patients with essential blepharospasm. J Neurol 2007;254:890–896. interpretation of data 11. Dydak U, Mueller S, Sándor PS, Meier D, Boesiger P, Jung HH. Cerebral metabolic alterations in McLeod syndrome. Eur Neurol 2006;56:17–23.

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 5 ARTICLE OPEN ACCESS Mutations in the m-AAA proteases AFG3L2 and SPG7 are causing isolated dominant optic atrophy

Majida Charif, PhD, Arnaud Chevrollier, PhD, Na¨ıg Gueguen, PhD, Céline Bris, PhD, David Goudenège, PhD, Correspondence Valérie Desquiret-Dumas, PhD, Stéphanie Leruez, MD, Estelle Colin, MD, PhD, Audrey Meunier, MD, Dr. Lenaers [email protected] Catherine Vignal, MD, PhD, Vasily Smirnov, MD, Sabine Defoort-Dhellemmes, MD, Isabelle Drumare Bouvet, MD, Cyril Goizet, MD, PhD, Marcela Votruba, MD, PhD, Neringa Jurkute, MD, PhD, Patrick Yu-Wai-Man, MD, PhD, Francesca Tagliavini, PhD, Leonardo Caporali, PhD, Chiara La Morgia, MD, PhD, Valerio Carelli, MD, PhD, Vincent Procaccio, MD, PhD, Xavier Zanlonghi, MD, Isabelle Meunier, MD, PhD, Pascal Reynier, MD, PhD, Dominique Bonneau, MD, PhD, Patrizia Amati-Bonneau, MD, PhD, and Guy Lenaers, PhD

Neurol Genet 2020;6:e428. doi:10.1212/NXG.0000000000000428 Abstract Objective To improve the genetic diagnosis of dominant optic atrophy (DOA), the most frequently inherited optic nerve disease, and infer genotype-phenotype correlations.

Methods Exonic sequences of 22 genes were screened by new-generation sequencing in patients with DOA who were investigated for ophthalmology, neurology, and brain MRI.

Results We identified 7 and 8 new heterozygous pathogenic variants in SPG7 and AFG3L2. Both genes encode for mitochondrial matricial AAA (m-AAA) proteases, initially involved in recessive hereditary spastic paraplegia type 7 (HSP7) and dominant spinocerebellar ataxia 28 (SCA28), respectively. Notably, variants in AFG3L2 that result in DOA are located in different domains to those reported in SCA28, which likely explains the lack of clinical overlap between these 2 phenotypic manifestations. In comparison, the SPG7 variants identified in DOA are interspersed among those responsible for HSP7 in which optic neuropathy has previously been reported.

Conclusions Our results position SPG7 and AFG3L2 as candidate genes to be screened in DOA and indicate that regulation of mitochondrial protein homeostasis and maturation by m-AAA proteases are crucial for the maintenance of optic nerve physiology.

From the MitoLab Team (M.C., A.C., C.B., D.G., V.D.-D., S.L., V.P., P.R., D.B., P.A.-B., G.L.), UMR CNRS 6015—INSERM U1083, Institut MitoVasc, Angers University and Hospital; Genetics and immuno-cell therapy Team (M.C.), Mohammed First University, Oujda, Morocco; Departments of Biochemistry and Genetics (C.B., D.G., V.D.-D., E.C., V.P., P.R., D.B., P.A.-B.), University Hospital Angers; Department of Ophthalmology (A.M.), Centre Hospitalier Universitaire Saint-Pierre, Brussels, Belgium; Neuroophthalmology Department (C.V.), Rothschild Ophthalmologic Foundation, Paris; Exploration of Visual Function and Neuro-Ophthalmology Department (V.S., S.D.-D., I.D.B.), Lille University Hospital, Rue Emilie Laine, Lille Cedex; CHU Bordeaux (C.G.), Service de Génétique Médicale, Centre de Référence « Neurogénétique » and Université de Bordeaux, INSERM U 1211, Laboratoire Maladies Rares, Génétique et Métabolisme (MRGM) Bordeaux; School of Optometry and Vision Sciences (M.V.), Cardiff University and Cardiff Eye Unit, University Hospital of Wales; NIHR Biomedical Research Centre at Moorfields Eye Hospital and UCL Institute of Ophthalmology (N.J., P.Y.-W.-M.), London; Department of Clinical Neurosciences (P.Y.-W.-M.), Cambridge Centre for Brain Repair and MRC Mitochondrial Biology Unit, University of Cambridge; Cambridge Eye Unit (P.Y.-W.-M.), Addenbrooke’s Hospital, Cambridge University Hospitals, UK; IRCCS Istituto Delle Scienze Neurologiche di Bologna (F.T., L.C., C.L.M., V.C.), Bellaria Hospital; Unit of Neurology (C.L.M., V.C.), Department of Biomedical and NeuroMotor Sciences (DIBINEM), University of Bologna, Italy; Centre de Compétence Maladies Rares (X.Z.), Clinique Pluridisciplinaire Jules Verne, Nantes; and National Centre in Rare Diseases (I.M.), Genetics of Sensory Diseases, University Hospital, Montpellier, France.

Go to Neurology.org/NG for full disclosures. Funding information is provided at the end of the article.

Copyright © 2020 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 Glossary DOA = dominant optic atrophy; HSP7 = hereditary spastic paraplegia; OCT = optical coherence tomography; RGC = retinal ganglion cell; RNFL = retinal nerve ﬁber layer; SCA28 = spinocerebellar ataxia.

Dominant optic atrophy (DOA, MIM*605290) is the most optic atrophy, initially screened for OPA1, OPA3,andWFS1 commonly inherited optic neuropathy, leading to irreversible exonic sequences and the 3 primary Leber hereditary optic loss of retinal ganglion cells (RGCs), optic nerve degeneration, neuropathy mutations. Cases negative after this primary and central visual loss.1,2 More than 400 OPA1 variants were screening were analyzed by resequencing a panel of 22 genes – identified in DOA individuals,3 6 resulting in excess of mito- (table e-1, links.lww.com/NXG/A260) involved in inherited chondrial fission.7,8 Surprisingly, a similar clinical presentation optic neuropathies or in mitochondrial dynamics. Library was also reported in individuals with dominant DNM1L preparation for each sample was performed using an Ion mutations9 (MIM603850) and mitochondrial network hyper- AmpliSeq Library Kit 2.0 (Cat. no. 4480441) according to fusion, thus providing evidence that alterations of both fusion the manufacturer’s protocol (Thermo Fisher Scientific, and fission compromise RGC survival. This hypothesis was MAN0006735). Sample emulsion PCR, emulsion breaking, further supported by the identification in syndromic DOA and enrichment were performed using the Ion 540 Kit–Chef families of dominant mutations in MFN2 (MIM608507)10 (Cat. no. A27759) according to the manufacturer’s and OPA3 (MIM606580),11 2 additional genes acting on mi- instructions (Thermo Fisher Scientific, MAN0010851). tochondrial dynamics. More recently, a single SPG7 Sequencingwasperformedusinga540ChIPsonanIonS5 (MIM602783) mutation and a single AFG3L2 (MIM604581) Sequencer using the barcoded samples. Sequencing data – mutation were reported in DOA families,12 14 although were processed using our own dedicated bioinformatics mutations in these genes are commonly known to be re- pipeline, as described elsewhere.20 Candidate pathogenic sponsible for the recessive hereditary spastic paraplegia type 7 variants were validated by Sanger sequencing, and their (HSP7)15 and dominant spinocerebellar ataxia 28 (SCA28),16 segregation was assessed in DNAs from other members of respectively. In addition, the occurrence of heterozygous var- the families, when available. iants in SPG7 and AFG3L2 was identified in a patient affected with DOA and parkinsonism,17 a clinical presentation found in Cell studies few patients with OPA1.18 SPG7 and AFG3L2 are paralogue Fibroblasts from AFG3L2 individuals P1: III:2 and P2: II:1 from genes encoding mitochondrial matricial AAA (m-AAA)- family 9 and 15, respectively, were generated from skin biopsies proteases involved in protein homeostasis and the cleavage of andculturedin2/3Dulbecco’s Minimum Essential Medium the OMA1 and YME1L mitochondrial proteases, which con- (DMEM, Gibco) supplemented with 1/3 AmnioMAX (Gibco), trol the shift between profusion long and profission short 10% fetal calf serum (Lonza), and 1% Penicillin-Streptomycin- OPA1 isoforms.19 Amphotericin B (Lonza). Mitochondrial network analysis, respiratory chain enzymatic activities, and mtDNA copy number 9 This prompted us to screen SPG7 and AFG3L2 exonic were assessed as described. sequences in patients with DOA without molecular diagnosis. We report the identification of pathogenic variants in these 2 Data availability genes in nonsyndromic patients with DOA. All data relevant to this study are contained within the article.

Methods Results Standard protocol approvals, registrations, Identification of SPG7 and AFG3L2 pathogenic and patient consents variants in individuals with DOA or isolated Written informed consent to perform genetic analyses was optic atrophy using a targeted obtained from each subject involved in this study or from the sequencing panel parents of individuals younger than 18 years of age, according to Six hundred cases without positive result after screening the protocols approved by the ethical committees of the dif- OPA1, OPA3, and WFS1 exonic sequences and the 3 primary ferent institutes involved in this study and in agreement with the Leber hereditary optic neuropathy mutations were included Declaration of Helsinki (Institutional Review Board Committee in a resequencing program focused on 22 genes, among of the University Hospital of Angers, Authorization number: which were those already ﬁrmly established for DOA and AC-2012-1507). recessive optic atrophy and candidate genes encoding actors of the mitochondrial dynamics. After eliminating frequent Genetic analysis (>1/10.000) and nonpathogenic variants, according to the Genomic DNA was extracted from peripheral blood cells SIFT, PolyPhen, MutationTaster, and LRT prediction tools, from multinational cohorts of DOA and sporadic cases of we identiﬁed 7 and 8 individuals harboring a SPG7 or an

2 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG Figure 1 SPG7 and AFG3L2 pedigrees

Description of the pedigrees with SPG7 (top) and AFG3L2 (middle and bottom) mutations and their segregation among the DOA families. DOA = dominant optic atrophy.

AFG3L2 pathogenic heterozygous variant, respectively. Conversely, all AFG3L2 mutated individuals disclosed a severe These variants were confirmed by Sanger sequencing and optic atrophy with visual acuities ranging from 0.2/10 to 2/10, analyzed for segregation whenever possible in the respective except for the 3 members of family 7 who had visual acuity families. Segregation of the 15 variants fitted with the clinical scores above 4/10 (table 1). First ophthalmologic examination features of affected individuals for whom DNA samples were occurred in a broad range of age, with some individuals being available (Figure 1). In family 13, the c.1126G > A variant in affected early during the first 2 decades, as reported for patients AFG3L2 was not found in both parents, suggesting that it with OPA1.1 All patients disclosed optic nerve pallor and highly occurred de novo. reduced RNFL at OCT scanning (figure 2B). Similar to SPG7, no brain MRI abnormality was reported in investigated Phenotypic manifestations of SPG7 and AFG3L2 patients, except for patient II.1 from family 14, who had a pi- mutation carriers tuitary adenoma without cerebellar atrophy. This individual All individuals included were referred to ophthalmology had a hearing impairment in addition to a very low visual acuity departments for visual acuity impairment. At inclusion, none of (table 1). them complained of spastic paraplegia or SCA, and in all cases but one, the brain MRI was normal. All patients with SPG7 Functional consequences of SPG7 and presented optic disk pallor and accordingly, reduced retinal AFG3L2 mutations nerve fiber layer (RNFL) at optical coherence tomography SPG7 and AFG3L2 encode highly similar proteins with 5 (OCT) scanning, mainly on the temporal side (figure 2A). conserved domains (figure 3). Four of the DOA mutations in Nevertheless, their visual acuity alterations were mild, with SPG7 are referenced with a frequency close to 1e-05 in the scores ranging from 3/10 to 10/10 (table 1) and occurring GnomAD database, whereas the 3 others were not referenced. during midlife for all individuals except one (family 3, II-1). All SPG7 variants responsible for DOA are interspersed with Patient II.1 from family 4 with the best visual acuity only pre- the recessive variants responsible for HSP7, and 2 of them are sented a significant reduction of the RNFL at the OCT exam- deletions leading to a frameshift at positions 258 and 654. ination. No additional symptom was observed for the index These latter data suggest that SPG7 haploinsufficiency might cases and their relatives when the clinical data were available, be the primary causal pathologic process in DOA. Surprisingly, except for a hearing impairment in family 5 (table 1). other heterozygote composite frameshift mutations were

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 3 Figure 2 Ophthalmologic exploration of SPG7 and an AFG3L2 affected individuals

Left: Eye fundus pictures of individuals with SPG7 (A, family 5, II.1 and family 6 II.1) and individuals with AFG3L2 (B, family 9, III:2 and family 12, II.1) revealing the temporal pallor of the optic discs in both REs and LEs. Right: RNFL by optic coherence tomography in individuals, disclosing the mild reduction of RNFL thickness in the individuals with SPG7 (A) and the severe one in the individuals with AFG3L2 (B). The green area defines the 5th to 95th, the yellow area the 1st to 5th, and the red area below the 1st percentiles. INF = inferior quadrants; LE = left eye; NAS = nasal; RE = right eye; RNFL = retinal nerve fiber layer assessment; SUP = superior; TEMP = temporal.

reported to cause HSP7, even at an earlier position toward the the SCA28 phenotype (figure 3B). Two fibroblast cell lines N-end of the protein (figure 3A). Unfortunately, no individual were established from individuals III:2 from family 9 and II:1 with a SPG7 variant accepted to provide a skin biopsy to infer from family 15 to assess their mitochondrial shape and fibroblasts for functional validation. physiology. A tendency toward mitochondrial fragmentation was observed in AFG3L2 fibroblasts (figure e-1, links.lww. Seven out of the 8 AFG3L2 variants identified in this study com/NXG/A261), together with a significant reduction of were not referenced in any database. They result in missense CI, CIII, and CIV enzymatic activities, which correlated with amino acid changes (table 1), and all but one (p.Thr644Ser asignificant citrate synthase decreased activity, suggesting from family 14) are located in domains different from the a reduction of the mitochondrial mass. This prompted the one involved in SCA28 individuals, suggesting that they af- analysis of mitochondrial DNA copy number (figure e-1), fect another AFG3L2 function than the one responsible for which showed a significant 50% reduction in AFG3L2 cells.

4 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG Table 1 Clinical data of the patients with SPG7 and AFG3L2

Other ORF GnomAD Family Patient Sex Age VA symptoms Brain MRI Gene mutation Protein change rs # Freq.

1 I.1 F 42 4/10 — Normal SPG7 c.1960_ p.Val654Profs*7 Unknown 1963delGTCA

2 II.1 M 43 3/10 Liver ND SPG7 c.773- p.Val258Glyfs*30 rs768136171 8.03e-06 cirrhosis 774delTG

3 II.1 F 8 5/10 — Normal SPG7 c.934G>C p.Ala312Pro Unknown

4 II.1 F 66 10/ — Normal SPG7 c.392G>A p.Arg131His rs985921704 7.99e-06 10

5 II.1 F 31 9/10 Hearing ND SPG7 c.1048C>T p.Pro350Ser rs199789849 2 e-05 impairment

6 II.1 F 64 5/10 — Normal SPG7 c.209C>A p.Pro70His Unknown

7 II.1 F 62 2&9/ — Normal SPG7 c.2191G>A p.Ala731Thr rs747521455 3.19e-5 10

8 III.2 M 36 5/10 — Normal AFG3L2 c.1248A>T p.Arg416Ser Unknown

IV.1 M 7 5/10 — Normal AFG3L2 c.1248A>T p.Arg416Ser

IV.4 M 8 4/10 — Normal AFG3L2 c.1248A>T p.Arg416Ser

9 II.2 F 47 2/10 — Normal AFG3L2 c.1541C>T p.Pro514Leu Unknown

III.1 F 26 1/10 — Normal AFG3L2 c.1541C>T p.Pro514Leu

III.2 M 22 1/10 — Normal AFG3L2 c.1541C>T p.Pro514Leu

III.3 F 15 1/10 — Normal AFG3L2 c.1541C>T p.Pro514Leu

10 I.2 M 80 1/10 — ND AFG3L2 c. 1010 G>A p.Gly337Glu Unknown

II.1 M 61 1.6/ — ND AFG3L2 c. 1010 G>A p.Gly337Glu 10

II.2 M 60 1.6/ — ND AFG3L2 c. 1010 G>A p.Gly337Glu 10

II.3 M 58 1.6/ — ND AFG3L2 c. 1010 G>A p.Gly337Glu 10

11 I.1 F 46 0.5/ — Normal AFG3L2 c.1009G>A p.Gly337Arg Unknown 10

II.1 F 13 0.5/ — Normal AFG3L2 c.1009G>A p.Gly337Arg 10

II.2 M 7 0.8/ — Normal AFG3L2 c.1009G>A p.Gly337Arg 10

12 II.1 F 25 0.2/ — Normal AFG3L2 c.1541C>T p.Pro514Leu Unknown 10

13 II.1 F 9 1.3/ — Normal AFG3L2 c.1126G>A p.Glu376Lys Unknown 10

14 II.1 M 59 1.6/ — Normal AFG3L2 c.1931C>G p.Thr644Ser rs1226952405 3.98e-06 10

15 II.1 F 19 0.5/ Hearing Pituitary AFG3L2 c.221A>C p.Glu74Ala Unknown 10 impairment adenomas

Abbreviations: GnomAD Freq. = frequency in the genome aggregation database; ND = not done; ORF = open reading frame; rs # = reference sequence number; VA = visual acuity.

Discussion indistinguishable from that seen in patients with DOA harboring dominant OPA1 mutations. As in OPA1-positive DOA, Dominant SPG7 and AFG3L2 mutations can result in isolat- SPG7 and AFG3L2 mutation carriers can present with a broad ed optic nerve involvement with a clinical phenotype spectrum of visual impairment ranging from asymptomatic

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 5 Figure 3 Structural representation of SPG7 and AFG3L2 amino acid changes related to mutations in individuals with DOA, HSP7, and SCA28/SPAX5

(A): Structure and domains of the SPG7 protein with the amino acid changes associated with DOA (top) and HSP7 (bottom); red, mutations identified in this study; purple, a DOA mutation previously reported; and black: HSP7 published mutations. (B): Structure and domains of the AFG3L2 protein with the amino acid changes associated with DOA (top) and to other diseases (bottom); red, mutations identified in this study; purple, a previously reported DOA mutation; black, published mutations responsible for SCA28; blue, published mutations responsible for recessive spastic ataxia SPAX5; in green, myoclonus and pyramidal signs; and in orange, microcephaly, early onset seizures, spasticity, and basal ganglia atrophy. DOA = dominant optic atrophy; SCA28 = spinocerebellar ataxia; SPAX5 = spastic ataxia-neuropathy syndrome.

mutation carriers to legal blindness and, in some of them, to Dominant mutations in AFG3L2 were initially found in a more severe syndromic manifestation with sensorineural individuals affectedwithSCA28anddolocalizeinexons15 deafness.21 and 16, in addition to one in exon 10.22 Recessive consanguineous AFG3L2 mutations in exon 15 were also identified It is intriguing that SPG7 mutations can behave both dominantly in a spastic ataxia-neuropathy syndrome23 (SPAX5), and recessively with variable tissue specificity. Of interest, in whereas 2 additional dominant AFG3L2 variants were a previous work reporting novel SPG7 mutations, 10 affected identified in an individual affectedwithmyoclonusandpy- individuals underwent an ophthalmologic examination and all of ramidal signs24 and a recessive mutation in another family them had evidence of a mild optic neuropathy with bilateral optic affected with microcephaly, early onset seizures, spasticity, disc pallor and thinning of the peripapillary RNFL on OCT and basal ganglia atrophy.25 Ofinterest,the8missense imaging.12 Conversely, the neurologic examination by the same AFG3L2 variants that we identified in individuals with DOA team of the first-identified SPG7-related DOA family did not are involving other domains than those identified earlier, evidence any gait or walking difficulties in mutation carriers. In thus explaining the absence of clinical overlap between all addition, the SPG7 variants that we identified in individuals with the symptoms previously described for AFG3L2 mutations DOA are interspersed with the recessive variants resulting in and the optic atrophy found in this study. This is reinforced HSP7, and in both diseases frameshift variants are contributing by the normality of the brain MRI and the absence of ataxia to the pathophysiologic mechanism. Altogether, these data in AFG3L2-related patients with DOA. Nevertheless, we suggest a clinical overlap related to SPG7 mutations between suggest that the ophthalmologic follow-up of these individ- DOA on one end and HSP7 on the other end and all possible uals should be accompanied by a neurologic examination to mixed phenotype inbetween. These observations should prompt eventually diagnose early manifestations of cerebellar ataxia to perform a systematic neuroophthalmological examination of or any other clinical symptoms. individuals with HSP7 and their heterozygous parents. Similarly, DOA individuals with a SPG7 variant should have a neurologic The 2 m-AAA proteases encoded by SPG7 and AFG3L2 follow-up, particularly older than the age of 40 years, to evaluate genes have been described as promoting OPA1 cleavage to the eventual occurrence of spasticity and also cognitive impair- its short profission isoforms.19 This activity is under the ment, epilepsy, and cerebellar atrophy. control of OMA1 and YME1L proteases that directly process

6 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG the cleavage of OPA1.26 Our data suggest that this mecha- 02361449 and VC by the “Ricerca Corrente” funding, both nism might be affected by the dominant variants in both from the Italian Ministry of Health. The views expressed are genes, although at different levels, depending on the mutated those of the author(s) and not necessarily those of the NHS, gene. Indeed, AFG3L2 forms both homopolymers and the NIHR, or the Department of Health. M. Charif, A. heteropolymers with SPG7; therefore, pathogenic variants Chevrollier, N. Gueguen, C. Bris, D. Goudenège, V. should affect the activity of both types of polymers. Con- Desquiret-Dumas, S. Leruez, E. Colin, A. Meunier, C. Vignal, versely, SPG7 can only form heteropolymers with AFG3L2, V. Smirnov, S. Defoort-Dhellemmes, I. Drumare Bouvet, C. but not homopolymers, implying that pathogenic variants Goizet,M.Votruba,N.Jurkute,P.Yu-Wai-Man,F.Taglia- should only affect the activity of the AFG3L2-SPG7 heter- vini, L. Caporali, C. La Morgia, V. Carelli, V. Procaccio, X. opolymers. This might explain the relatively mild visual Zanlonghi, I. Meunier, P. Reynier, D. Bonneau, P. Amati- deficits in individuals with SPG7 variants, contrasting with Bonneau, and G. Lenaers report no disclosures relevant to the more severe visual loss observed in individuals with the manuscript. Go to Neurology.org/NG for full AFG3L2 variants and the syndromic DOA plus phenotype disclosures. found in a patient harboring concurrent mutations in both genes. It further questions the possible specificity of these Publication history mutations for the regulation of OPA1 processing. In this Received by Neurology: Genetics November 22, 2019. Accepted in final respect, mitochondrial dynamic might be affected by the form April 6, 2020. other SPG7 mutations but overwhelmed by the severe HSP7 phenotype, whereas it is apparently not affected by the AFG3L2 variants involved in SCA28. This reflect the fact that no optic atrophy has been yet reported in the mouse Appendix Authors models harboring Spg7 or Afg3l2 mutations. Name Location Contribution

Majida Charif, University of Acquisition and analysis of Thus, we provide compelling evidence that heterozygous PhD Angers, France the data and drafting SPG7 and AFG3L2 mutations should be considered in the a significant portion of the manuscript and casesofisolatedDOA,moresowhenalreadyfoundtobe figures OPA1-negative. Our ﬁndings stress the central role mediated Arnaud University of Acquisition and analysis of by m-AAA proteases in the regulation of mitochondrial dy- Chevrollier, Angers, France the data namics and how dysfunction of these pathways compro- PhD mise the RGC integrity and survival, resulting in optic Na¨ıg Gueguen, University and Acquisition and analysis of neuropathy. PhD Hospital of Angers, the data France

Acknowledgment C´eline Bris, PhD University and Acquisition and analysis of The authors are indebted to Dr. Vittoria Petruzzella, Silvana Hospital of Angers, the data France Guerriero, Anna Maria De Negri, and Michele Carbonelli for clinical investigations and thank Dr. Menetou for stimulating David University and Acquisition and analysis of Gouden`ege, PhD Hospital of the data critical discussions. Angers, France

Study funding Valérie University and Acquisition and analysis of The authors are indebted for the financial support to the Desquiret- Hospital of Angers, the data Universitéd’Angers, CHU d’Angers, the Région Pays de la Dumas, PhD France Loire, Angers Loire Métropole, the Fondation Maladies Rares, Stéphanie University and Clinical investigation and Leruez, MD Hospital of Angers, phenotyping the Fondation VISIO, Kjer-France, Ouvrir Les Yeux, Retina France France, UNADEV, Fondation de France, and Association Estelle Colin, MD, University and Clinical investigation and Française contre les Myopathies. PhD Hospital of Angers, phenotyping France

Disclosure Audrey Meunier, University and Clinical investigation and P. Yu-Wai-Man is supported by a Clinician Scientist Fel- MD Hospital of Brussels, phenotyping Belgium lowship Award (G1002570) from the Medical Research Council (UK) and also receives funding from Fight for Sight Catherine Vignal, Rothschild Clinical investigation and MD, PhD Ophthalmologic phenotyping (UK), the Isaac Newton Trust (UK), the UK National In- Foundation, Paris, stitute of Health Research (NIHR) as part of the Rare Dis- France eases Translational Research Collaboration, and the NIHR Vasily Smirnov, University and Clinical investigation and Biomedical Research Centre based at Moorﬁelds Eye Hos- MD Hospital of Lille, phenotyping pital NHS Foundation Trust and UCL Institute of Oph- France thalmology. L. Caporali is supported by the Grant GR-2016- Continued

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 7 References Appendix (continued) 1. Lenaers G, Hamel C, Delettre C, et al. Dominant optic atrophy. Orphanet J Rare Dis 2012;7:46. Name Location Contribution 2. Yu-Wai-Man P, Chinnery PF. Dominant optic atrophy: novel OPA1 mutations and revised prevalence estimates. Ophthalmology 2013;120:1712. 3. Alexander C, Votruba M, Pesch UE, et al. OPA1, encoding a dynamin-related Sabine Defoort- University and Clinical investigation and GTPase, is mutated in autosomal dominant optic atrophy linked to chromosome Dhellemmes, MD Hospital of Lille, phenotyping 3q28. Nat Genet 2000;26:211–215. France 4. Delettre C, Lenaers G, Griffoin JM, et al. Nuclear gene OPA1, encoding a mitochondrial dynamin-related protein, is mutated in dominant optic atrophy. Nat Genet Isabelle Drumare University and Clinical investigation and 2000;26:207–210. Bouvet, MD Hospital of Lille, phenotyping 5. Ferré M, Caignard A, Milea D, et al. Improved locus-specific database for OPA1 France mutations allows inclusion of advanced clinical data. Hum Mutat 2015;36:20–25. 6. Le Roux B, Lenaers G, Zanlonghi X, et al. OPA1: 516 unique variants and 831 patients Cyril Goizet, MD, University and Clinical investigation and registered in an updated centralized Variome database. Orphanet J Rare Dis 2019;14: PhD, Hospital of phenotyping 214. Bordeaux, France 7. Olichon A, Landes T, Arnauné-Pelloquin L, et al. Effects of OPA1 mutations on mitochondrial morphology and apoptosis: relevance to ADOA pathogenesis. J Cell Marcela Votruba, University and Clinical investigation and Physiol 2007;211:423–430. MD, PhD Hospital of Cardiff, phenotyping 8. Bertholet AM, Delerue T, Millet AM, et al. Mitochondrial fusion/fission dynamics in UK neurodegeneration and neuronal plasticity. Neurobiol Dis 2016;90:3–19. 9. Gerber S, Charif M, Chevrollier A, et al. Mutations in DNM1L, as in OPA1, result in Neringa Jurkute, University College of Clinical investigation and dominant optic atrophy despite opposite effects on mitochondrial fusion and fission. MD, PhD London, UK phenotyping Brain 2017;140:2586–2596. 10. Rouzier C, Bannwarth S, Chaussenot A, et al. The MFN2 gene is responsible for Patrick Yu-Wai- University College of Clinical investigation and mitochondrial DNA instability and optic atrophy ’plus’ phenotype. Brain 2012;135: Man, MD, PhD London, UK phenotyping 23–34. 11. Reynier P, Amati-Bonneau P, Verny C, et al. OPA3 gene mutations responsible Francesca University of Acquisition and analysis of for autosomal dominant optic atrophy and cataract. J Med Genet 2004;41:e110. Tagliavini, PhD Bologna, Italy the data 12. Klebe S, Depienne C, Gerber S, et al. Spastic paraplegia gene 7 in patients with spasticity and/or optic neuropathy. Brain 2012;135:2980–2993. Leonardo University of Acquisition and analysis of 13. Charif M, Roubertie A, Salime S, et al. A novel mutation of AFG3L2 might cause Caporali, PhD Bologna, Italy the data dominant optic atrophy in patients with mild intellectual disability. Front Genet 2015; 6:311. Chiara La University and Clinical investigation and 14. Colavito D, Maritan V, Suppiej A, et al. Non-syndromic isolated dominant optic Morgia, MD, PhD Hospital of Bologna, phenotyping atrophy caused by the p.R468C mutation in the AFG3 like matrix AAA peptidase Italy subunit 2 gene. Biomed Rep 2017;7:451–454. 15. Casari G, De Fusco M, Ciarmatori S, et al. Spastic paraplegia and OXPHOS im- Valerio Carelli, University and Clinical investigation and pairment caused by mutations in paraplegin, a nuclear-encoded mitochondrial met- MD, PhD Hospital of Bologna, phenotyping alloprotease. Cell 1998;93:973–983. Italy 16. Di Bella D, Lazzaro F, Brusco A, et al. Mutations in the mitochondrial protease gene AFG3L2 cause dominant hereditary ataxia SCA28. Nat Genet 2010;42:313–321. Vincent University and Clinical investigation and 17. Magri S, Fracasso V, Plumari M, et al. Concurrent AFG3L2 and SPG7 mutations Procaccio, MD, Hospital of Angers, phenotyping associated with syndromic parkinsonism and optic atrophy with aberrant OPA1 PhD France processing and mitochondrial network fragmentation. Hum Mutat 2018;39: 2060–2071. Xavier Clinique Jules Verne, Clinical investigation and 18. Carelli V, Musumeci O, Caporali L, et al. Syndromic parkinsonism and dementia Zanlonghi, MD, Nantes, France phenotyping associated with OPA1 missense mutations. Ann Neurol 2015; 78: 21–38. 19. Ehses S, Raschke I, Mancuso G, et al. Regulation of OPA1 processing and mito- Isabelle Meunier, Hospital of Clinical investigation and chondrial fusion by m-AAA protease isoenzymes and OMA1. J Cell Biol 2009; 187: MD, PhD Montpellier, France phenotyping 1023–1036. 20. Felhi R, Sfaihi L, Charif M, et al. Next generation sequencing in family with MNGIE Pascal Reynier, University and Clinical investigation and syndrome associated to optic atrophy: novel homozygous POLG mutation in the MD, PhD Hospital of Angers, phenotyping; Acquisition C-terminal sub-domain leading to mtDNA depletion. Clin Chim Acta 2019;488: France and analysis of the data 104–110. 21. Yu-Wai-Man P, Griffiths PG, Gorman GS, et al. Multi-system neurological disease is Dominique University and Clinical investigation and common in patients with OPA1 mutations. Brain 2010; 133: 771–786. Bonneau, MD, Hospital of Angers, phenotyping and revised the 22. Cagnoli C, Stevanin G, Brussino A, et al. Missense mutations in the AFG3L2 pro- PhD, France manuscript for intellectual teolytic domain account for ;1.5% of European autosomal dominant cerebellar content ataxias. Hum Mutat 2010;31:1117–1124. 23. Pierson TM, Adams D, Bonn F, et al. Whole-exome sequencing identifies ho- Patrizia Amati- University and Conception and design of mozygous AFG3L2 mutations in a spastic ataxia-neuropathy syndrome linked to Bonneau, MD, Hospital of Angers, the study and acquisition mitochondrial m-AAA proteases. PLoS Genet 2011;7:e1002325. PhD France and analysis of the data 24. Mancini C, Orsi L, Guo Y, et al. An atypical form of AOA2 with myoclonus associated with mutations in SETX and AFG3L2. BMC Med Genet 2015;16:16. Guy Lenaers, University and Conception and design of 25. Eskandrani A, AlHashem A, Ali ES, et al. Recessive AFG3L2 mutation causes pro- PhD Hospital of Angers, the study and drafting gressive microcephaly, early onset seizures, spasticity, and basal ganglia involvement. France a significant portion of the Pediatr Neurol 2017;71:24–28. manuscript and figures 26. Anand R, Wai T, Baker MJ, et al. The i-AAA protease YME1L and OMA1 cleave OPA1 to balance mitochondrial fusion and fission. J Cell Biol 2014;204:919–929.

8 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG ARTICLE OPEN ACCESS Genotyping single nucleotide polymorphisms for allele-selective therapy in Huntington disease

Daniel O. Claassen, MD, Jody Corey-Bloom, MD, PhD, E. Ray Dorsey, MD, Mary Edmondson, MD, Correspondence Sandra K. Kostyk, MD, PhD, Mark S. LeDoux, MD, PhD, Ralf Reilmann, MD, H. Diana Rosas, MD, Dr. Claassen [email protected] Francis Walker, MD, Vicki Wheelock, MD, Nenad Svrzikapa, MS, Kenneth A. Longo, PhD, Jaya Goyal, PhD, Serena Hung, MD, and Michael A. Panzara, MD, MPH

Neurol Genet 2020;6:e430. doi:10.1212/NXG.0000000000000430 Abstract Background The huntingtin gene (HTT) pathogenic cytosine-adenine-guanine (CAG) repeat expansion responsible for Huntington disease (HD) is phased with single nucleotide polymorphisms (SNPs), providing targets for allele-selective treatments.

Objective This prospective observational study deﬁned the frequency at which rs362307 (SNP1) or rs362331 (SNP2) was found on the same allele with pathogenic CAG expansions.

Methods Across 7 US sites, 202 individuals with HD provided blood samples that were processed centrally to determine the number and size of CAG repeats, presence and heterozygosity of SNPs, and whether SNPs were present on the mutant HTT allele using long-read sequencing and phasing.

Results Heterozygosity of SNP1 and/or SNP2 was identiﬁed in 146 (72%) individuals. The 2 polymorphisms were associated only with the mHTT allele in 61% (95% high density interval: 55%, 67%) of individuals.

Conclusions These results are consistent with previous reports and demonstrate the feasibility of genotyping, phasing, and targeting of HTT SNPs for personalized treatment of HD.

From the Vanderbilt University Medical Center (D.O.C.), Nashville, TN; University of California San Diego (J.C.-B.), La Jolla; University of Rochester Medical Center (E.R.D.), NY; HD Reach (M.E.), Raleigh, NC; Ohio State University (S.K.K.), Columbus; University of Memphis and Veracity Neuroscience, LLC (M.S.L.), TN; George-Huntingon-Institute & Department of Clinical Radiology University of Muenster (R.R.), Department of Neurodegeneration, Hertie Institute for Clinical Brain Research, University of Tuebingen, Germany; Havard Medical School (H.D.R.), Massachusetts General Hospital, Boston; Wake Forest University School of Medicine (F.W.), Winston Salem, NC; University of California Davis Health (V.W.), Sacra- mento, CA; Wave Life Sciences USA, Inc. (N.S., K.A.L., J.G., S.H., M.A.P.), Cambridge, MA; and Department of Paediatrics (N.S.), Medical Sciences Division, University of Oxford, UK.

Go to Neurology.org/NG for full disclosures. Funding information is provided at the end of the article.

The Article Processing Charge was funded by Wave Life Sciences. This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloading and sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Copyright © 2020 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 Glossary CAG = cytosine-adenine-guanine; HD = Huntington disease; mHTT = mutant HTT; SNP = single nucleotide polymorphism; wtHTT = wild-type HTT; UHDRS = Uniﬁed Huntington Disease Rating Scale.

Most individuals with Huntington disease (HD) are heterozy- Methods gous for the cytosine-adenine-guanine (CAG) repeat, having one wild-type (wtHTT) and one abnormally expanded mutant Ambulatory men and women aged 25–65 years with diagnostic huntingtin gene (mHTT) allele.1 Allele-selective targeting of the motor features of HD (Unified Huntington Disease Rating 7 mHTT transcript offers a personalized approach to HD treat- Scale [UHDRS] Diagnostic Confidence Score of 4 and stage I ment1 and has the potential advantage of keeping wtHTT or II HD with UHDRS Total Functional Capacity scores ≥7) protein relatively intact. It has been suggested that wtHTT were eligible. At one clinic visit, blood samples were collected in protein is required for normal neurologic function and may be PAXgene Blood DNA and RNA tubes (PreAnalytiX, Switzer- – neuroprotective in the adult brain.2 4 Oneapproachistotarget land) as per manufacturer’s instructions and shipped frozen for specific single nucleotide polymorphisms (SNPs) found on the processing. mHTT allele. Multiple SNPs have an increased frequency in HD but do not affect diagnosis or disease course.5,6 According to Blood samples were processed at a central laboratory using 3 previous reports, 65%–70% of individuals with HD of European steps. First, the number of CAG repeats was confirmed by PCR ancestry carry SNP rs362307 (SNP1), SNP rs362331 (SNP2), and the size was determined using a Bioanalyzer (3500 Genetic or both SNPs.1 According to the Genome Aggregation Data- Analyzer, Applied Biosystems, San Francisco, CA). Second, base (gnomAD.broadinstitute.org), SNP1 and SNP2 frequen- zygosity at the targeted SNP(s) was determined by Sanger se- cies vary by population and are higher in Latinos and Africans, quencing. Finally, for samples with confirmed normal and ex- respectively, whereas both are lower in Asian populations. For panded CAG repeats and SNP heterozygosity at either SNP1 or selective treatment to be feasible, an individual must be het- SNP2, a PacBio (Menlo Park, CA) long-read sequencing in- erozygous for a target SNP and it must be colocated on the same vestigational assay determined the haplotype phase of the SNP allele or haplotype phased, with the expanded CAG repeat. In with the CAG expansion. this observational study, individuals with HD were recruited from HD clinics in the United States (US), genotyped, and Demographic information was descriptively summarized. experimentally phased to evaluate the prevalence of SNP1 and The frequency of SNP1 or SNP2 T variant on mHTT was SNP2 on the same allele as the expanded CAG repeat. determined as the posterior chain product probability of

2 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG the frequency of heterozygosity at either SNP and the with CAG expansion were available for 128 individuals, frequency of the U variant on the mHTT allele. The prob- of which 108 (84%) had SNP1 only (n = 41), SNP2 only ability was calculated using a beta distribution model, which (n = 28), or both SNPs (n = 39) present on the mHTT provides the mode and 95% highest density intervals. allele (figure 1). For the other 20 heterozygous individ- SNP prevalence was compared across sites and by sex and uals, target SNPs were found on the wtHTT allele. Two ethnicity. individuals had expanded CAG repeats on both alleles. Sixteen individuals had inconclusive phasing results be- Standard protocol approvals and cause of failed quality controlonmultiplesamplepro- data availability cessing attempts (up to 4) and interpretable data could not This observational study was conducted in accordance with be provided. the Declaration of Helsinki with ethics committees’ approval from all 7 participating US centers. Anonymized data will be The most frequently occurring genotype for HTT was shared on request from any qualified investigator. a normal allele with 18 CAG repeats and mutant allele with 43 CAG repeats (figure 2). The prevalence of SNP1 and/or SNP2 was consistent across study sites and inde- Results pendent of the sex of individuals. Because of the small ff Participants numbers of individuals with di erent ethnicities, no con- From February 2017 to September 2018, 203 individuals clusions regarding ethnicity and SNP prevalence could be with HD were enrolled (table); 1 individual was excluded for made. older age. Overall, the frequency at which SNP1 and/or SNP2 were as- SNP heterozygosity and phasing sociated only with the mHTT allele in this observational cohort Nearly three-quarters of individuals (146/202; 72%) were was 0.61 (95% high-density interval: 0.55, 0.67), based on the heterozygous for SNP1 only (n = 52), SNP2 only (n = 46), probability of both SNP heterozygosity and phasing on the or both SNPs (n = 48) (figure 1). Thus, approximately mHTT allele ([146/202] × [108/128]). one-quarter of individuals (56/202; 28%) were found to have the same SNP on both alleles or did not have a SNP on either allele. Among the heterozygotes, the se- Discussion quencing results to determine SNP haplotype phasing This is the first study to demonstrate the feasibility of rapid assessment of SNP prevalence and haplotype phasing in a relatively large number of individuals with HD (>200) using next- Table Characteristics of study participants generation sequencing of HD transcript. Heterozygosity for Individuals with SNP1, SNP2, or both SNPs was established in most individuals Characteristics HD (N = 203) with HD (61%). These results suggest that clinical trials in this Age, y, mean (SD) 49.7 (10.0) population are feasible and future SNP1/2 selective treatments could potentially address a significant portion of the Sex, n (%) HD population. Male 104 (51)

Female 99 (49) This study directly phased patient samples, and the results are consistent with SNP frequencies reported in previous Race, n (%) studies using computational methods.5,8,9 Pﬁster et al.5 White 193 (95) sequenced 24 SNPs using genomic DNA from 109 German

Black or African American 6(3) and US individuals with HD and found an increased frequency of SNP1 (>48%) vs other SNPs. The addition of 2 Native American or Alaska Native and Whitea 2(1) SNP sites was calculated to incorporate approximately 75% 5 Asian 1 (0.5) of the individuals with HD tested. In another study, unrelated Italian (European Caucasian) individuals with HD Native Hawaiian or other Pacific Islander 1 (0.5) heterozygous for the CAG repeat (N = 327) were geno- Ethnicity, n (%) typed at 26 SNP sites, including SNP1 and SNP2.8 Of

Hispanic or Latino 12 (6) these, 86% of individuals were heterozygous at one or more SNP loci and may be amenable to allele-selective therapy. Not Hispanic or Latino 191 (94) SNP2 heterozygosity was most prevalent in this HD pop- CAG repeats, median (range) 43 (38–62) ulation (46.2%), increasing the estimated probability of heterozygosity at either SNP1 or SNP2 to 65%.8 Using the Abbreviations: CAG = cytosine-adenine-guanine; HD = Huntington disease. a Two participants indicated American Indian and White. University of British Columbia and Tissue Bank for HD Research database, direct sequencing was performed for

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 3 Figure 1 Phasing results

HDI = high-density interval.

Canadians of European origin (n = 65) and confirmed in investigational compounds are stereopure antisense oli- a replication group (n = 203).9 Of 190 SNPs identified, 23% gonucleotides (ASOs) synthesized by precisely controlling were common (minor allele frequency >0.20). The maxi- the chirality of the phosphorothioate linkages to enable mum coverage of a single SNP was 52%, whereas targeting selective targeting of the SNPs of interest. In general, between 1 and 4 SNPs was theorized to cover 89% of the stereopure ASOs have increased lipophilicity and stability HD population. and enhanced RNase H1 activity than comparable stereo- random ASOs.10 The ability to prospectively define an individual’sspecific HTT SNP haplotype permits consideration of personalized This study confirms the feasibility of rapidly detecting allele-selective gene-silencing methods. In the PRECISION- SNP1 and/or SNP2 in the HD population in the United HD trials (NCT03225833, NCT03225846), the presence States and opens the possibility of selectively targeting of SNP1 and/or SNP2 with the CAG expansion is de- mHTT transcript in eligible patients. It is important that the termined using a similar process. Based on these results, proof of concept of this approach may lead to the identifi- participants receive targeted therapy with WVE-120101 or cation and targeting of other SNPs in the HD population, WVE-120102 for SNP1 and SNP2, respectively, aiming to allowing others to potentially benefit from allele-selective selectively lower mHTT without affecting wtHTT.These treatment.

4 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG Figure 2 Length of CAG repeats among population with HD

The normal allele with 18 CAG repeats and the mutant allele with 43 CAG repeats were shown to be the frequently occurring genotype for HTT in the patients tested. The data did not pass the normal distribution using the Shapiro-Wilk normality test.

Acknowledgment Michael J. Fox Foundation, NIH/NINDS, National Science The authors thank Shawn Levy, Nripesh Prasad, and Dan Foundation, Nuredis Pharmaceuticals, Patient-Centered Out- Dorset at HudsonAlpha for performing Sanger sequencing comes Research Institute, Pfizer, Prana Biotechnology, Raptor and phasing assay, Giulia Malferrari at Biorep for performing Pharmaceuticals, Roche, Safra Foundation, Teva Pharmaceut- CAG testing, and Ramakrishna Boyanapalli from Wave Life icals, and the University of California Irvine; provides editorial Sciences for his input and oversight on the assay and data. services for Karger Publications; and has ownership interests with Blackfynn (data integration company) and Grand Rounds Study funding (second opinion service). M. Edmondson reports no dis- No targeted funding reported. closures. S.K. Kostyk has received clinical trial research support from Wave Life Sciences; grants from Huntington’sDisease Disclosure Society of America, CHDI Foundation, Roger A Vaughn The study was sponsored by Wave Life Sciences Ltd., Cam- Fund/Columbus Medical Foundation, Azevan/NIH/NINDS bridge, MA. D.O. Claassen has served as a consultant/advisory NeuroNext STAIR trial, Voyager Therapeutics, Vaccinex, board member for Lundbeck, Teva Neuroscience, Acadia, Pfizer, and Auspex/Teva; personal fees for travel/consulting AbbVie, and Wave Life Sciences. J. Corey-Bloom has received from Lundbeck; and honoraria for speaking at Huntington research support from CHDI, Roche/Genentech, Teva Phar- Study Group. M.S. LeDoux has received a grant from Wave Life maceuticals, and Vaccinex Inc.; she is on the speakers bureau Sciences. R. Reilmann reports payments to George- for Roche/Genentech and Teva Pharmaceuticals. W.R. Dorsey Huntington-Institute from Teva, Pfizer, uniQure, Wave Life has received honoraria for speaking at the American Academy Sciences, Roche, Ipsen, Vaccinex, Mitoconix, Prilenia, Raptor, of Neurology courses, the American Neurologic Association, Omeros, Prana Biotechnology, Desitin, and AOP Orphan. and the University of Michigan; received compensation for H.D. Rosas has received grant support from NINDS and is consulting services from 23andMe, AbbVie, American Well, a consultant for Wave Life Sciences. F. Walker has received Biogen, CLINTREX, DeciBio, Denali Therapeutics, Glax- grants from Vaccinex, Pfizer, and Teva; and consulting fees oSmithKline, Grand Rounds, Karger, Lundbeck, MC10, from AskBio. V. Wheelock has received compensation as MedAvante, Medical-legal services, Mednick Associates, a consultant for Roche Pharmaceuticals. N. Svrzikapa, K.A. NINDS, Olson Research Group, Optio, Prilenia, Putnam Longo, J. Goyal, S. Hung, and M.A. Panzara are employees of Associates, Roche, Sanofi, Shire, Sunovion Pharma, Teva, Wave Life Sciences USA, Inc. Go to Neurology.org/NG for full UCB, and Voyager Therapeutics; received grants from Abbvie, disclosures. Acadia Pharmaceuticals, AMC Health, BioSensics, Burroughs Wellcome Fund, Davis Phinney Foundation, Duke University, Publication history Food and Drug Administration, GlaxoSmithKline, Greater Received by Neurology: Genetics September 17, 2019. Accepted in final Rochester Health Foundation, Huntington Study Group, form February 24, 2020.

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 5 Appendix Authors Appendix (continued)

Name Location Role Contribution Name Location Role Contribution

Daniel Vanderbilt Lead and Study design; H. Diana Harvard Medical Author Study design; Claassen, University, corresponding subject Rosas, MD School, subject MD Nashville, TN author recruitment/ Massachusetts recruitment/ acquisition of General Hospital, acquisition of data; analyzed Boston data; analyzed and interpreted and interpreted the data; drafted the data; the manuscript revised the for intellectual manuscript for content intellectual content Jody Corey- University of Author Subject Bloom, MD California San recruitment/ Diego acquisition of Francis Wake Forest Author Analyzed and data; analyzed Walker, MD University School interpreted and interpreted of Medicine, the data and the data; Winston Salem, revised revised the NC the manuscript manuscript for for intellectual intellectual content content Vicki University of Author Subject E. Ray University of Author Study design; Wheelock, California Davis recruitment/ Dorsey, MD Rochester, NY analyzed and MD Health, acquisition of interpreted Sacramento data; analyzed the data; and interpreted revised the the data; manuscript for revised the intellectual manuscript for content intellectual content Mary HD Reach, Author Study design; Edmondson, Raleigh, NC analyzed and Nenad Wave Life Author Study design; MD interpreted Svrzikapa, Sciences USA, analyzed and the data; MS Inc., Cambridge, interpreted revised the MA; University of the data; manuscript for Oxford, United revised intellectual Kingdom the manuscript content for intellectual content Sandra K. Ohio State Author Subject Kostyk, MD, University, recruitment/ Kenneth A. Wave Life Author Study design; PhD Columbus acquisition of Longo, PhD Sciences USA, analyzed and data; analyzed Inc., Cambridge, interpreted and interpreted MA the data; the data; revised the revised the manuscript manuscript for for intellectual intellectual content content Jaya Goyal, Wave Life Author Study design; Mark S. University of Author Study design; PhD Sciences USA, analyzed and LeDoux, MD, Memphis and subject Inc., Cambridge, interpreted PhD Veracity recruitment/ MA the data; revised Neuroscience, acquisition of the manuscript LLC, Memphis, data; analyzed for intellectual TN and interpreted content the data; revised the Serena Wave Life Author Study design; manuscript Hung, MD Sciences USA, subject for intellectual Inc., Cambridge, recruitment/ content MA acquisition of data; analyzed Ralf George- Author Study design; and interpreted Reilmann, Huntingon- analyzed and the data; revised MD Institute and interpreted the manuscript Dept. of Clinical the data; for intellectual Radiology, revised the content University of manuscript for Muenster, Hertie intellectual Michael A. Wave Life Author Study design; Institute for content Panzara, Sciences USA, analyzed and Clinical MD, MPH Inc., Cambridge, interpreted the Brain Research, MA data; revised the University manuscript for of Tuebingen, intellectual Germany content

6 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG References 6. van Bilsen PH, Jaspers L, Lombardi MS, Odekerken JC, Burright EN, Kaemmerer WF. Identification and allele-specific silencing of the mutant huntingtin allele in 1. Kay C, Skotte NH, Southwell AL, Hayden MR. Personalized gene silencing thera- Huntington’s disease patient-derived fibroblasts. Hum Gene Ther 2008;19: – peutics for Huntington disease. Clin Genet 2014;86:29 36. 710–719. 2. Leavitt BR, van Raamsdonk JM, Shehadeh J, et al. Wild-type huntingtin protects 7. Huntington Study Group. Unified Hungtington’s Disease Rating Scale: reliability and neurons from excitotoxicity. J Neurochem 2006;96:1121–1129. consistency. Mov Disord 1996;11:136–142. 3. Dietrich P, Johnson IM, Alli S, Dragatsis I. Elimination of huntingtin in the adult 8. Lombardi MS, Jaspers L, Spronkmans C, et al. A majority of Huntington’s disease mouse leads to progressive behavioral deficits, bilateral thalamic calcification, and patients may be treatable by individualized allele-specific RNA interference. Exp altered brain iron homeostasis. PLoS Genet 2017;13:e1006846. Neurol 2009;217:312–319. 4. Gauthier LR, Charrin BC, Borrell-Pages M, et al. Huntingtin controls neurotrophic 9. Warby SC, Montpetit A, Hayden AR, et al. CAG expansion in the Huntington disease support and survival of neurons by enhancing BDNF vesicular transport along gene is associated with a specific and targetable predisposing haplogroup. Am J Hum microtubules. Cell 2004;118:127–138. Genet 2009;84:351–366. 5. Pfister EL, Kennington L, Straubhaar J, et al. Five siRNAs targeting three SNPs may 10. Iwamoto N, Butler DC, Nenad S, et al. Control of phosphorothioate stereochemistry provide therapy for three-quarters of Huntington’s disease patients. Curr Biol 2009; substantially increases the efficacy of antisense oligonucleotides. Nat Biotechnol 2017; 19:774–778. 35:845–851.

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 7 ARTICLE OPEN ACCESS Prevalence of RFC1-mediated spinocerebellar ataxia in a North American ataxia cohort

Dona Aboud Syriani, Darice Wong, PhD, Sameer Andani, BS, Claudio M. De Gusmao, MD, Yuanming Mao, BS, Correspondence May Sanyoura, PhD, Giacomo Glotzer, Paul J. Lockhart, PhD, Sharon Hassin-Baer, MD, Dr. Fogel [email protected] Vikram Khurana, MD, PhD, Christopher M. Gomez, MD, PhD, Susan Perlman, MD, Soma Das, PhD, and or Dr. Das Brent L. Fogel, MD, PhD [email protected] Neurol Genet 2020;6:e440. doi:10.1212/NXG.0000000000000440

Abstract RELATED ARTICLE Objective Editorial We evaluated the prevalence of pathogenic repeat expansions in replication factor C subunit 1 Intronic pentanucleotide (RFC1) and disabled adaptor protein 1 (DAB1) in an undiagnosed ataxia cohort from North expansion in the replication America. factor 1 gene (RFC1)is a major cause of adult-onset ataxia Methods A cohort of 596 predominantly adult-onset patients with undiagnosed familial or sporadic cer- Page e436 ebellar ataxia was evaluated at a tertiary referral ataxia center and excluded for common genetic causes of cerebellar ataxia. Patients were then screened for the presence of pathogenic repeat expansions in RFC1 (AAGGG) and DAB1 (ATTTC) using ﬂuorescent repeat-primed PCR (RP- PCR). Two additional undiagnosed ataxia cohorts from diﬀerent centers, totaling 302 and 13 patients, respectively, were subsequently screened for RFC1, resulting in a combined 911 subjects tested.

Results In the initial cohort, 41 samples were identified with 1 expanded allele in the RFC1 gene (6.9%), and 9 had 2 expanded alleles (1.5%). For the additional cohorts, we found 20 heterozygous samples (6.6%) and 17 biallelic samples (5.6%) in the larger cohort and 1 heterozygous sample (7.7%) and 3 biallelic samples (23%) in the second. In total, 29 patients were identified with biallelic repeat expansions in RFC1 (3.2%). Of these 29 patients, 8 (28%) had a clinical diagnosis of cerebellar ataxia, neuropathy, and vestibular areflexia syndrome (CANVAS), 14 had cerebellar ataxia with neuropathy (48%), 4 had pure cerebellar ataxia (14%), and 3 had spinocerebellar ataxia (10%). No patients were identified with expansions in the DAB1 gene (spinocerebellar ataxia type 37).

Conclusions In a large undiagnosed ataxia cohort from North America, biallelic pathogenic repeat expansion in RFC1 was observed in 3.2%. Testing should be strongly considered in patients with ataxia, especially those with CANVAS or neuropathy.

From the Department of Neurology (D.A.S., D.W., Y.M., S.P., B.L.F.), Program in Neurogenetics, David Geffen School of Medicine, University of California, Los Angeles; Department of Neurology (D.W., B.L.F.), Clinical Neurogenomics Research Center, David Geffen School of Medicine, University of California, Los Angeles; Department of Human Genetics (S.A., M.S., S.D.), University of Chicago, IL; Department of Neurology (C.M.D.G., V.K.), Brigham and Women’s Hospital and Harvard Medical School, Boston, MA; Department of Neurology (G.G., C.M.G.), University of Chicago, IL; Bruce Lefroy Centre (P.J.L.), Murdoch Children’s Research Institute; Department of Paediatrics (P.J.L.), University of Melbourne, Parkville, Australia; Sackler Faculty of Medicine (S.H.-B.), Tel-Aviv University, Tel-Aviv, Israel; and Department of Human Genetics (B.L.F.), David Geffen School of Medicine, University of California, Los Angeles.

Go to Neurology.org/NG for full disclosures. Funding information is provided at the end of the article.

Copyright © 2020 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 Glossary CANVAS = cerebellar ataxia, neuropathy, and vestibular areﬂexia syndrome; DAB1 = Disabled Adaptor Protein 1; RFC1 = replication factor C subunit 1; SCA37 = spinocerebellar ataxia type 37.

Cerebellar ataxia is a heterogeneous genetic disorder charac- enrolled at the University of Chicago and 13 patients enrolled at terized by inability to control balance and coordination. the Brigham and Women’s Hospital under institutional review Roughly 50% of patients remain undiagnosed despite ad- board-approved procedures with identical inclusion criteria – vanced genomic testing.1 4The most common genetic ataxias, were subsequently tested as well. For the purpose of assessing as well as several rarer forms, are caused by nucleotide repeat disease prevalence, only probands from affected families were expansions, which typically require targeted non–sequence- included in this study, and all pathogenic repeat expansions – based testing to identify.5 8 were confirmed in additional affected family members when available. Recent studies identified a recessive intronic (AAGGG) repeat expansion in replication factor C subunit 1 (RFC1) re- Repeat expansion testing lated to cerebellar ataxia, neuropathy, and vestibular areflexia RFC1 gene repeat expansion analysis syndrome (CANVAS) in Australia and the United 9,10 Fluorescent repeat-primed PCR (RP-PCR) was performed to Kingdom. In addition, this expansion may be responsible detect RFC1 pathogenic (AAGGG)n alleles using previously for up to 22% (33/150) of sporadic cerebellar ataxia and 63% 9,10 9 published primers with an optimized touchdown PCR pro- (32/51) of ataxia associated with sensory neuropathy. Sim- tocol and Qiagen HotStarTaq. One primer set included the ilarly, a dominant pathologic pentanucleotide (ATTTC) re- forward 59FAM-ACTGACAGTGTTTTTGCCTGT-39 primer, peat insertion was identified within a normal (ATTTT) the anchor 59-CAGGAAACAGCTATGACC-39 primer, and the tandem repeat element in the intronic 59 untranslated region repeat 59-CAGGAAACAGCTATGACCAAGGGAAGGGA of the disabled adaptor protein 1 (DAB1) gene causing spi- AGGGAAGGGAAGGG-39 primer that identifies the (AAGGG) nocerebellar ataxia type 37 (SCA37) in patients from the 11 repeats. A second primer set included the forward 59FAM- southern Iberian Peninsula. TCAAGTGATACTCCAGCTACACCGT-39 primer, the anchor 59-CAGGAAACAGCTATGACC-39 primer, and 3 repeat To address the frequency of these repeat expansion disorders in primers that identify the (AAGGG) repeats 59-CAGGAAA- North America, we assessed a large cohort of 596 patients from CAGCTATGACCAACAGAGCAAGACTCTGTTTCA the United States with unidentified cerebellar ataxia. We AAAAAGGGAAGGGAAGGGAAGGGAA-39,59-CAGGAAA- identified biallelic RFC1 expansion in 1.5% (n = 9) and found CAGCTATGACCAACAGAGCAAGACTCTGTTTCA no patients with a pathogenic DAB1 expansion. We further AAAAGGGAAGGGAAGGGAAGGGAA-39,and59-CAG- tested 2 additional cohorts from different centers (the larger of GAAACAGCTATGACCAACAGAGCAAGACTCTGTT which consisted of approximately one-third samples from TCAAAAGGGAAGGGAAGGGAAGGGAA-39.Fragment patients in Canada, with the remainder from the United States) length analysis was performed using an Applied Biosystems and identified RFC1-mediated ataxia cases in 5.6% (17/302) 3730xl DNA Analyzer with Peak Scanner software (v. 2.0). To and 23% (3/13), respectively, for a total prevalence of 3.2% determine whether the genotype of samples with positive RP- (29/911). PCR results was heterozygous or biallelic, standard PCR was performed using published primers,9 forward 59-TCAAGTGA- Methods TACTCCAGCTACACCGTTGC-39 primer and the reverse 59 GTGGGAGACAGGCCAATCACTTCAG-39 primer. Obser- Standard protocol approvals, registrations, vation of a band at or near 348 bp (wild-type size) corresponding and patient consents to repeat sizes of less than approximately 60 repeats (approxi- Patients were enrolled at the University of California, Los mately 650 bp) identified patients as heterozygotes. As an internal Angeles (UCLA) Ataxia Center, clinically assessed for acquired control to prevent false positives, standard PCR was also per- causes of ataxia, and then considered for genetic causes.2 Only formed simultaneously on the same sample with primers patients with negative testing for the common genetic ataxias designed to amplify a 282-bp band from the SPG11 gene (figure (SCA1, SCA2, SCA3, SCA6, SCA7, and Friedreich ataxia) were e-1, links.lww.com/NXG/A266). included in this study. The majority (; two-thirds) were adult and sporadic onset. All patients consented for DNA collection For a proportion of samples at the University of Chicago, 3 for genetic analysis. Peripheral blood was collected from separate reactions each using 100 ng of genomic DNA were patients, and DNA was then isolated and purified using the performed to confirm the existence of a true biallelic Gentra Puregene Blood Kit (Qiagen) for genetic testing. The (AAGGG) repeat expansion (figure e-2, links.lww.com/NXG/ study methods used were approved by the UCLA Institutional A267). First, a flanking PCR was performed using primers that Review Board. Additional DNA samples from 302 patients surrounded the RFC1 region of interest. The flanking primers

2 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG Table 1 Patient demographics

UCLA University of Chicago Harvard Combined

A. Age and sex

Total 596 302 13 911

Femalea 300 (50.3%) 148 (49.0%) 8 (61.5%) 456 (50.1%)

Average age at onsetb (y) 55 (±17) 48 (±22) 56 (±11) 53 (±19)

B. Clinical presentation

CANVAS 3 (0.5%) 7 (2.3%) 1 (7.7%) 11 (1.2%)

Cerebellar ataxia neuropathyb 41 (6.9%) 18 (6.0%) 4 (30.8%) 63 (6.9%)

Episodic ataxia 19 (3.2%) 22 (7.3%) 0 41 (4.5%)

MSA 104 (17.4%) 6 (2.0%) 1 (7.7%) 111 (12.2%)

Pure cerebellar ataxia 129 (21.6%) 62 (20.5%) 2 (15.4%) 193 (21.2%)

Spastic ataxia 38 (6.4%) 24 (7.9%) 1 (7.7%) 63 (6.9%)

Spastic paraplegia 35 (5.9%) 4 (1.3%) 0 39 (4.3%)

Spinocerebellar ataxia 158 (26.5%) 144 (47.9%) 4 (30.8%) 306 (33.6%)

Other 69 (11.6%) 15 (5.0%) 0 84 (9.2%)

C. Race/ethnicityc

Asian 54 (9.1%) 12 (4.0%) 0 66 (7.2%)

Native Hawaiian or Pacific Islander 1 (0.2%) 0 0 1 (0.1%)

Native America or Alaska Native 6 (1.0%) 0 0 6 (0.7%)

Black 15 (2.5%) 10 (3.3%) 0 25 (2.7%)

White, Hispanic, or Latino 58 (9.7%) 5 (1.7%) 0 63 (6.9%)

White, non-Hispanic 393 (65.9%) 192 (63.6%) 12 (92.3%) 597 (65.5%)

Unspecified 76 (12.8%) 84 (27.8%) 1 (7.7%) 161 (17.7%)

Abbreviations: CANVAS = cerebellar ataxia, neuropathy, and vestibular areflexia syndrome; MSA = multiple system atrophy. A) Age and sex of the enrolled subjects are shown. B) The major clinical presentations of the patients enrolled in this study are shown. The presence of peripheral neuropathy was determined either clinically or electrophysiologically. Spastic ataxia and spastic paraplegia are distinguished based on which symptom was clinically estimated to be predominant. Spinocerebellar ataxia is used to indicate patients with cerebellar ataxia as the primary symptom but with notable features other than spasticity or neuropathy (e.g., dementia, epilepsy, or extrapyramidal signs). Although all patients exhibited ataxia, patients who did not clearly fit the major diagnostic categories listed were labeled as other (e.g., primary extrapyramidal conditions such as parkinsonism). MSA includes both possible and probable cases as defined by current diagnostic criteria. C) Race and ethnicity of the enrolled subjects are shown. Patients who did not choose to disclose this information are listed as unspecified. a Ten subjects in the UCLA cohort did not report sex. b Sixteen subjects in the University of Chicago cohort did not have a reported age at onset. c Individuals identifying membership in more than 1 race were counted separately for each race. included a labeled forward primer9 (59-/56-FAM/ACTGA- performed using the same forward primer as the flanking, the CAGTGTTTTTGCCTGT-39)(10μm) and a reverse primer9 M13 anchor primer, and an (AAGGG) specific primer9 (59- (59-GGCTGAGGCAGGAGATTCAC-39)(10μm). Second, CAGGAAACAGCTATGACC_AAGGGAAGGGAAGG- RP‐PCR was performed to detect individual nonpathogenic GAAGGGAAGGG-39). To obtain full amplification of the (AAAAG) motifs. The primers for the (AAAAG) RP-PCR expanded (AAGGG) motif, the Qiagen HotStarTaq chemistry included the same forward primer as the flanking reaction, an was used with 400 μm of deoxyribonucleotide triphosphates M13 anchor primer (59-CAGGAAACAGCTATGACC-39) using a standard PCR. All 3 products were loaded on an (10 μm) and an (AAAAG) specific primer (59-CAGGAAA- ABI3730xl DNA analyzer after denaturing with a 5% GS500 CAGCTATGACC_AAAAGAAAAGAAAAGAAAA- Rox/formamide mixture and subsequently analyzed using GAAAAG-39)(1μm). Both the flanking and (AAAAG) RP- GeneMarker v2.6.0 (SoftGenetics Inc.). PCR products were amplified using the Takara LA PCR kit in combination with a touchdown PCR. Third, to detect in- To validate the assays performed at the different centers, dividual pathogenic (AAGGG) motifs, a PCR reaction was a proportion of samples (>50%) determined to have

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 3 heterozygous or biallelic AAGGG expansions were assessed at cohort are described in table 1. Average age was 55 years, 50% least 2 times by both expansion RP-PCR testing and standard of the patients were female, and 66% were white, non- PCR genotyping in 2 separate laboratories. Hispanic. The most common phenotypes were spinocerebellar ataxia (26.5%), pure cerebellar ataxia (21.6%), and DAB1 gene repeat expansion analysis multiple system atrophy (17.4%). To assess for the presence RP-PCR was performed to detect expansion of the normal of pathogenic (AAGGG) repeat expansions in RFC1, fluo- (ATTTT) DAB1 repeat region and for detection of the path- rescent repeat-primed fragment analysis was performed and 11,12 ogenic (ATTTC)n insertion using published primers. identified at least 1 expansion in 50 of 596 patients (8.4%, Fragment length analysis was performed as described above. figure 1 and figure e-1, links.lww.com/NXG/A266). Standard PCR was used to genotype subjects for the presence of Data availability a heterozygous or a pathogenic biallelic expansion. Nine The data sets generated and/or analyzed during the current subjects (1.5%) were found with biallelic expansions. Of these study are available from the corresponding author on rea- patients, 3 subjects presented clinically with CANVAS (33%, sonable request. 100% of phenotype), 5 had cerebellar ataxia with neuropathy (56%, 12% of phenotype), and 1 had spinocerebellar ataxia Results (11%, 0.6% of phenotype). We examined the prevalence of repeat expansions in RFC1 To validate these findings, we tested 2 additional cohorts from and DAB1 in a large cohort from the tertiary referral Ataxia centers in different regions of the United States, the University Center at UCLA. The demographics of this 596 subject of Chicago (UC) and Brigham and Women’s Hospital affiliated

Figure 1 RFC1 expansion analysis

Representative RP-PCR results from a patient with disease due to (A) biallelic expanded (AAGGG) pathogenic alleles or a control individual (B) with wild-type alleles. Samples with RP-PCR evidence of an expanded RFC1 allele were genotyped by standard PCR for biallelic expansion (C). Standard PCR allows categorization of individuals as biallelic with pathogenic expansions (no band, lanes 1-4), heterozygous wild-type (348 bp band, arrow, lanes 5-8), heterozygous with a non-pathogenic polymorphic expansion (variable sized bands, lanes 9-11), or wild-type with one or more non-pathogenic polymorphic expansion(s) (variable sized band(s), lanes 12-14). + = biallelic control; − = wild-type control; M = marker; RFC1 = replication factor C subunit 1.

4 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG Table 2 Demographics of patients with heterozygous expansions in the RFC1 gene

Sample Phenotype Sex Age at onset (y) Ethnicity Inheritance

H1 CANVAS Female 52 White, non-Hispanic Familial

H2 Cerebellar ataxia + neuropathy Male 72 White, Hispanic, or Latino Sporadic

H3 Cerebellar ataxia + neuropathy Male 20 White, non-Hispanic Familial

H4 Cerebellar ataxia + neuropathy Female 75 White, non-Hispanic Sporadic

H5 Episodic ataxia Male 36 White, Hispanic, or Latino Sporadic

H6 Episodic ataxia Male 30 White, non-Hispanic Familial

H7 Episodic ataxia Female 22 White, non-Hispanic Familial

H8 Episodic ataxia Male 10 White, non-Hispanic Sporadic

H9 MSA Female 66 White, non-Hispanic Sporadic

H10 MSA Male 63 White, non-Hispanic Sporadic

H11 MSA Female 56 Asian Sporadic

H12 MSA Male Unknown Unspecified Unspecified

H13 MSA Male 55 White, non-Hispanic Sporadic

H14 MSA Female 77 White, non-Hispanic Sporadic

H15 MSA Male 61 Unspecified Sporadic

H16 MSA Male 63 White, non-Hispanic Sporadic

H17 MSA Male 54 White, non-Hispanic Sporadic

H18 Pure cerebellar ataxia Male 72 White, non-Hispanic Sporadic

H19 Pure cerebellar ataxia Female 61 White, non-Hispanic Sporadic

H20 Pure cerebellar ataxia Female 66 Unspecified Sporadic

H21 Pure cerebellar ataxia Female 62 White, non-Hispanic Sporadic

H22 Pure cerebellar ataxia Female 76 Unspecified Sporadic

H23 Pure cerebellar ataxia Unknown 66 White, non-Hispanic, Native Sporadic American, or Alaska Native

H24 Pure cerebellar ataxia Female 37 White, non-Hispanic Sporadic

H25 Pure cerebellar ataxia Male 68 White, non-Hispanic Sporadic

H26 Pure cerebellar ataxia Female 70 White, non-Hispanic Sporadic

H27 Pure cerebellar ataxia Male 21 White, Hispanic, or Latino Sporadic

H28 Pure cerebellar ataxia Male 57 White, non-Hispanic Familial

H29 Spastic ataxia Male 54 White, non-Hispanic Sporadic

H30 Spastic ataxia Female 63 White, Hispanic, or Latino Sporadic

H31 Spastic ataxia Male 78 White, non-Hispanic Sporadic

H32 Spastic ataxia Female 57 White, non-Hispanic Sporadic

H33 Spastic ataxia Male 68 White, non-Hispanic Familial

H34 Spastic ataxia Female 38 Unspecified Sporadic

H35 Spastic ataxia Male 71 Unspecified Familial

H36 Spastic paraplegia Male 61 White, non-Hispanic Unspecified

H37 Spastic paraplegia Female 32 White, non-Hispanic Sporadic

H38 Spinocerebellar ataxia Male 84 Unspecified Sporadic

Continued Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 5 Table 2 Demographics of patients with heterozygous expansions in the RFC1 gene (continued)

Sample Phenotype Sex Age at onset (y) Ethnicity Inheritance

H39 Spinocerebellar ataxia Male 48 White, non-Hispanic Sporadic

H40 Spinocerebellar ataxia Female 66 White, non-Hispanic Sporadic

H41 Spinocerebellar ataxia Female 44 White, non-Hispanic Sporadic

H42 Spinocerebellar ataxia Male Unknown Unspecified Unspecified

H43 Spinocerebellar ataxia Male Unknown Unspecified Unspecified

H44 Spinocerebellar ataxia Female 69 White, Hispanic, or Latino Sporadic

H45 Spinocerebellar ataxia Female 76 White, Hispanic, or Latino Familial

H46 Spinocerebellar ataxia Female 62 White, non-Hispanic Familial

H47 Spinocerebellar ataxia Male 36 White, non-Hispanic Sporadic

H48 Spinocerebellar ataxia Female 35 White, non-Hispanic Familial

H49 Spinocerebellar ataxia Female White, non-Hispanic Familial

H50 Spinocerebellar ataxia Female 42 White, non-Hispanic Sporadic

H51 Spinocerebellar ataxia Female 35 Asian Sporadic

H52 Spinocerebellar ataxia Female 47 Unspecified Sporadic

H53 Spinocerebellar ataxia Male 44 White, non-Hispanic Sporadic

H54 Spinocerebellar ataxia Male 41 White, non-Hispanic Familial

H55 Other Unknown 24 Unspecified Unspecified

H56 Other Female 69 White, non-Hispanic Sporadic

H57 Other Female 35 White, Hispanic, or Latino Familial

H58 Other Male 78 White, non-Hispanic Sporadic

H59 Other Unknown 65 Unspecified Unspecified

H60 Other Male 53 Unspecified Sporadic

H61 Other Female 55 White, non-Hispanic Familial

H62 Other Female 24 White, non-Hispanic Sporadic

Abbreviations: CANVAS = cerebellar ataxia, neuropathy, and vestibular areflexia syndrome; MSA = multiple system atrophy; RFC1 = replication factor C subunit 1.

with Harvard Medical School. Demographics were similar to the table 2), and 29 of 911 subjects showed pathogenic biallelic UCLA cohort (table 1). The larger UC cohort, consisting of expansions across all cohorts for a total prevalence of 3.2% (29/ both sporadic and familial cases with roughly one-third of 911, table 3). Of the biallelic cases, the majority of the patients subjects originating from Canada, showed heterozygous were white (24/29, 83%), 1 patient was Hispanic (1/29, 3.4%), expansions in 20 individuals (6.6%) and pathogenic biallelic and the race/ethnicity of 5 patients was not reported (17%). ﬁ expansion in 17 patients (5.6%) ( gure e-2, links.lww.com/ Twelve of the cases showed sporadic onset (41%). In total, 8 NXG/A267). Four subjects had CANVAS (24%, 57% of phe- subjects had CANVAS (28%, 73% of phenotype), 14 had cer- notype), 7 had cerebellar ataxia with neuropathy (41%, 39% of ebellar ataxia with neuropathy (48%, 22% of phenotype), 4 had phenotype), 4 had pure cerebellar ataxia (23.5%, 6.5% of phe- pure cerebellar ataxia (14%, 2.1% of phenotype), and 3 had notype), and 2 had spinocerebellar ataxia (12%, 1.4% of phe- spinocerebellar ataxia (10%, 1.0% of phenotype). notype). In the smaller Harvard cohort, 1 individual showed heterozygous expansion (7.7%), and 3 of 13 patients (23%) had For DAB1 repeat expansion analysis, 83/596 (13.9%) subjects pathogenic biallelic expansions. Two of these had cerebellar showed an expanded ATTTT allele by RP-PCR analysis; ataxia with neuropathy (67%, 50% of phenotype), while 1 had however, none of these patients possessed the pathogenic CANVAS (33%, 100% of phenotype). Collectively, 62 indi- ATTTC insertion, indicating that no patients within the cohort viduals were found with heterozygous expansion (6.8%, 62/911, had SCA37.

6 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG Table 3 Demographics of patients with biallelic expansions in the RFC1 gene

Sample Phenotype Sex Age at onset (y) Ethnicity Inheritance

B1 CANVAS Male 78 White, non-Hispanic Sporadic

B2 CANVAS Female 69 White, non-Hispanic Familial

B3 CANVAS Male 63 White, non-Hispanic Familial

B4 CANVAS Female 33 White, non-Hispanic Familial

B5 CANVAS Male 40 White, non-Hispanic Sporadic

B6 CANVAS Female 61 Unspecified Unknown

B7 CANVAS Female 56 White, non-Hispanic Sporadic

B8 CANVAS Male 60 White, non-Hispanic Sporadic

B9 Cerebellar ataxia + neuropathy Female 81 White, non-Hispanic Familial

B10 Cerebellar ataxia + neuropathy Female 66 White, Hispanic, or Latino Familial

B11 Cerebellar ataxia + neuropathy Female 71 White, non-Hispanic Sporadic

B12 Cerebellar ataxia + neuropathy Male 69 White, non-Hispanic Sporadic

B13 Cerebellar ataxia + neuropathy Female 71 White, non-Hispanic Sporadic

B14 Cerebellar ataxia + neuropathy Female 42 White, non-Hispanic Familial

B15 Cerebellar ataxia + neuropathy Female 39 White, non-Hispanic Familial

B16 Cerebellar ataxia + neuropathy Male 38 White, non-Hispanic Familial

B17 Cerebellar ataxia + neuropathy Female 64 Unspecified Familial

B18 Cerebellar ataxia + neuropathy Female 55 Unspecified Familial

B19 Cerebellar ataxia + neuropathy Female 45 White, non-Hispanic Sporadic

B20 Cerebellar ataxia + neuropathy Male 37 White, non-Hispanic Familial

B21 Cerebellar ataxia + neuropathy Female 55 White, non-Hispanic Sporadic

B22 Cerebellar ataxia + neuropathy Female 64 White, non-Hispanic Sporadic

B23 Pure cerebellar ataxia Male 45 Unspecified Familial

B24 Pure cerebellar ataxia Female 76 White, non-Hispanic Familial

B25 Pure cerebellar ataxia Female 62 White, non-Hispanic Familial

B26 Pure cerebellar ataxia Male 55 White, non-Hispanic Familial

B27 Spinocerebellar ataxia Female Unknown Unspecified Unknown

B28 Spinocerebellar ataxia Male 45 White, non-Hispanic Sporadic

B29 Spinocerebellar ataxia Female 38 White, non-Hispanic Sporadic

Abbreviations: CANVAS = cerebellar ataxia, neuropathy, and vestibular areflexia syndrome; RFC1 = replication factor C subunit 1.

Discussion European ancestry is consistent with previous reports.9,10 Of the heterozygotes with race and ethnicity data available, 96% Overall, in a large undiagnosed ataxia cohort of 911 patients (47/49, table 2) were of white ancestry, and overall, single or from 3 tertiary referral centers in the United States, biallelic biallelic expansions were detected in 7.1% and 3.6%, re- pathogenic (AAGGG) repeat expansions in RFC1 were spectively, of the total white population surveyed in this observed in 3.2% (n = 29, 95% CI 2.0%–4.3%) of patients study. The high rate of heterozygosity (6.8%) in our cohort from the United States (25/29, 86%) and Canada (4/29, is notable but is similar to 1 previous study, which calculated 14%). The observation that the majority of cases (24/24, an allele frequency of 4% in control populations of 69 and 100%, table 3), where ethnicity was known, were white of 133 individuals based on haplotype in next-generation

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 7 sequencing data sets,10 although another study found a fre- RFC1 expansions anticipated in 1 of every 5.0 (95% CI quency of 0.7% in a cohort of 304 healthy controls using RP- 3.3–10.3) and 1 of every 1.4 (95% CI 1.0–2.3) patients PCR.9 Although our methods cannot accurately size larger tested, respectively. Taken together, these results suggest repeats, standard PCR analysis indicated that under the that RFC1 expansion testing is high yield in cases of CAN- conditions of our AAGGG RP-PCR assay, we were able to VAS and cerebellar ataxia with neuropathy but should also detect small expansions up to 60 repeats above wild type be considered in the genetic workup of patients with un- (;650 bp, figure 1 and figure e-1, links.lww.com/NXG/ diagnosed pure cerebellar and spinocerebellar ataxia. A266), and therefore, we suspect that our high rate of detection may be due, in part, to the detection of confounding Acknowledgment mildly expanded alleles below the estimated 400 repeat The authors thank all the patients and their families for their margin of pathogenicity9 (figure 1 and figure e-1, links.lww. participation in this study. com/NXG/A266). It is interesting to note that small AAGGG expansions have not previously been reported in Study funding patients or controls,9,10 and because all subjects tested pre- This work was supported by the National Institute for Neu- sented with some form of cerebellar ataxia, we cannot ex- rological Disorders and Stroke [R01NS082094 to BLF]. BLF clude the possibility that expanded AAGGG repeats may acknowledges support through donations to the University of contribute to the development of cerebellar ataxia in some California by the Rochester Ataxia Foundation. PJL was heterozygous individuals. We also cannot rule out a contri- supported by the Vincent Chiodo Foundation. CMG and VK bution of false-positive detection of other small polymorphic acknowledge support from the National Ataxia Foundation nonpathogenic non-AAGGG repeats9 in some heterozygous and the Brigham Research Institute. individuals. We did confirm all biallelic subjects with RP- PCR and standard PCR in at least 2 experiments each across Disclosure 2 separate laboratories and further determined that none of All authors declare that there is no conflict of interest. Go to these individuals harbored the most common non- Neurology.org/NG for full disclosures. pathogenic expanded repeat, AAAAG9 (data not shown). Publication history In addition, the pathogenic (ATTTC) repeat in the DAB1 This manuscript was previously posted on bioRxiv: doi: gene, causative for SCA37, was not observed in our large un- 10.1101/790006. Received by Neurology: Genetics diagnosed ataxia cohort of 596 individuals of mostly white, November 7, 2019. Accepted in final form March 15, 2020. non-Hispanic ancestry, consistent with the observation of afoundereffect in the Iberian Peninsula11,13 and suggesting that although this disorder should be considered in that pop- Appendix Authors ulation, it is likely extremely rare in the United States. Although Name Location Contribution no pathogenic ATTTC insertions were found, it is possible that fl Dona Aboud University of California, Acquisition of data, extremely large repeats of ATTTT anking a pathogenic Syriani Los Angeles analysis of data, and ATTTC insertion might prevent amplification of products drafting of the manuscript

from the RP-PCR, so false negatives cannot be ruled out in this Darice Wong, University of California, Analysis of data, study study, although this has not been commonly observed in PhD Los Angeles design, and drafting of the published reports.11,12 manuscript Claudio M. De Brigham and Women’ Acquisition of data and Consistent with previous reports, our study identified bial- Gusmao, MD Hospital and Harvard revision of the manuscript Medical School for intellectual content lelic RFC1 expansions in a high percentage of patients with CANVAS (n = 8, 73%) and cerebellar ataxia with neuropa- Sameer University of Chicago Acquisition of data, 9 Andani, BS analysis of data, and thy (n = 14, 22%). Although we do not have electrophysi- revision of the manuscript ologic data on all subjects, of the biallelic patients identified, for intellectual content all appeared to have a large fiber neuropathy (data not shown), which would be an important focus for further Yuanming University of California, Acquisition of data, Mao, BS Los Angeles analysis of data, and clinical investigation. In addition, we also observed biallelic drafting of the manuscript expansion in a notable percentage of patients with pure for intellectual content

cerebellar ataxia (n = 4, 2.1%) and generalized spinocer- May University of Chicago Acquisition of data, ebellar ataxia (n = 3, 1.0%), a frequency on par with the Sanyoura, analysis of data, and ﬁ PhD revision of the manuscript majority of rare ataxic disorders identi able by clinical se- for intellectual 2 quencing. Based on this study, the presence of neuropathy content confers a 20.1% absolute beneﬁt to testing (95% CI Giacomo University of Chicago Acquisition of data and 9.7%–30.6%) over pure cerebellar ataxia alone, and the Glotzer revision of the manuscript presence of both neuropathy and vestibulopathy further for intellectual content increases this to 70.6% (95% CI 44.2%–97.0%), with biallelic

8 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG Appendix (continued) Appendix (continued)

Name Location Contribution Name Location Contribution

Paul J. Murdoch Children’ Study design and revision Brent L. University of California, Study design and Lockhart, Research Institute of the manuscript for Fogel, MD, Los Angeles supervision, acquisition of PhD intellectual PhD data, analysis of data, and content drafting and revision of the manuscript for Sharon Sackler Faculty of Acquisition of intellectual content Hassin-Baer, Medicine, Tel-Aviv data and revision MD University, Tel-Aviv, Israel of the manuscript for intellectual content References Vikram Brigham and Women’s Acquisition of 1. Sun M, Johnson AK, Nelakuditi V, et al. Targeted exome analysis identiﬁes the genetic Khurana, MD, Hospital and Harvard data, study basis of disease in over 50% of patients with a wide range of ataxia-related phenotypes. PhD Medical School design, and Genet Med 2019;21:195–206. revision of the manuscript 2. Fogel BL, Lee H, Deignan JL, et al. Exome sequencing in the clinical diagnosis of for intellectual sporadic or familial cerebellar ataxia. JAMA Neurol 2014;71:1237–1246. content 3. Rexach J, Lee H, Martinez-Agosto JA, Nemeth AH, Fogel BL. Clinical application of next- generation sequencing to the practice of neurology. Lancet Neurol 2019;18:492–503. Christopher University of Chicago, Acquisition of 4. Ngo KJ, Rexach JE, Lee H, et al. A diagnostic ceiling for exome sequencing in M. Gomez, Chicago data, study design, cerebellar ataxia and related neurological disorders. Hum Mutat 2020;41:487–501. MD, PhD analysis of data, 5. Paulson H. Repeat expansion diseases. Handb Clin Neurol 2018;147:105–123. and revision 6. Mundwiler A, Shakkottai VG. Autosomal-dominant cerebellar ataxias. Handb Clin of the manuscript Neurol 2018;147:173–185. for intellectual 7. Fogel BL. Autosomal-recessive cerebellar ataxias. Handb Clin Neurol 2018;147: content 187–209. 8. Wallace SE, Bird TD. Molecular genetic testing for hereditary ataxia: what every Susan University of California, Acquisition of neurologist should know. Neurol Clin Pract 2018;8:27–32. Perlman, MD Los Angeles data and 9. Cortese A, Simone R, Sullivan R, et al. Biallelic expansion of an intronic repeat in revision of the manuscript RFC1 is a common cause of late-onset ataxia. Nat Genet 2019;51:649–658. for intellectual 10. Rafehi H, Szmulewicz DJ, Bennett MF, et al. Bioinformatics-based identiﬁcation of content expanded repeats: a non-reference intronic pentamer expansion in RFC1 causes CANVAS. Am J Hum Genet 2019;105:151–165. Soma Das, University of Chicago Acquisition of 11. Seixas AI, Loureiro JR, Costa C, et al. A pentanucleotide ATTTC repeat insertion in PhD data, data the non-coding region of DAB1, mapping to SCA37, causes spinocerebellar ataxia. analysis, study Am J Hum Genet 2017;101:87–103. design, and 12. Loureiro JR, Oliveira CL, Sequeiros J, Silveira I. A repeat-primed PCR assay for penta- revision of the manuscript nucleotide repeat alleles in spinocerebellar ataxia type 37. J Hum Genet 2018;63:981–987. for intellectual 13. Corral-Juan M, Serrano-Munuera C, Rabano A, et al. Clinical, genetic and neuro- content pathological characterization of spinocerebellar ataxia type 37. Brain 2018;141: 1981–1997.

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 9 CLINICAL/SCIENTIFIC NOTES OPEN ACCESS Biallelic LINE insertion mutation in HACD1 causing congenital myopathy

Fatema Al Amrani, MD, Carolina Gorodetsky, MD, Lili-Naz Hazrati, MD, Kimberly Amburgey, MSc, Correspondence Hernan D. Gonorazky, MD, and James J. Dowling, MD, PhD Dr. Dowling [email protected] Neurol Genet 2020;6:e423. doi:10.1212/NXG.0000000000000423

Congenital myopathies are clinically and genetically heterogeneous, resulting from mutations in at least 30 diﬀerent genes.1 The classical presentation is neonatal hypotonia and nonprogressive weakness with normal creatine phosphokinase, although there is a broad range in terms of age at onset and clinical presentation. Historically, congenital myopathies have been deﬁned and diagnosed based on muscle biopsy. However, with advances in genomics, genetics have taken primacy in the diagnostic pathway.2

HACD1 encodes 3-hydroxyacyl-CoA dehydratase 1, which participates in the biosynthesis of very-long-chain fatty acids (VLCFAs). A single family with 8 aﬀected members was previously described with homozygous nonsense mutation in HACD1. The clinical presentation included severe neonatal hypotonia and weakness, with impressive improvement through childhood and early adulthood. Muscle biopsy was consistent with congenital ﬁber-type disproportion.3

Case report We present the second case of HACD1 myopathy. The affected individual underwent initial evaluation for hypotonia as a neonate. She is the second child of consanguineous (first cousin) parents of Sri Lankan descent. She was born by C-section at 35+2/7 weeks and noted to have hypotonia at birth. She had poor oral intake requiring nasogastric tube supplementation but no respiratory difficulties. Physical examination at age 19 days showed no dysmorphic features, normal cranial nerve examination, severe axial and appendicular hypotonia, and a near- complete absence of antigravity movements. Workup included brain MRI, chromosomal microarray, Prader-Willi testing, and metabolic screening laboratories and did not provide a diagnosis. Management was supportive, with occupational and physiotherapy and feeding assistance.

Neuromuscular consultation was solicited at age 4 years. At that time, she had proximal muscle weakness and mild gross motor delay. She had achieved ambulation, could do stairs without difficulty, and could ride a bike with training wheels. Her feeding difficulties had resolved. Physical examination revealed a long, myopathic facies but no clear facial weakness (figure, A), with normal eyelid strength and no restriction of eye movements. She had reduced muscle tone, hyperextensibility at the elbows, but no joint contractures. She had mild proximal extremity muscle weakness (Medical Research Council grade 4/5 at deltoid and hip flexors/extensors), a negative Gowers sign, and mild Trendelenburg gait and was unable to run.

Creatine phosphokinase was within normal limits. Nerve conductions were normal, and EMG revealed myopathic units. Muscle biopsy, obtained from quadriceps, showed rare central nuclei, type 1 predominance, and relative type 1 hypotrophy (ﬁgure, B). Muscle ultrastructure, as determined by electron microscopy, was not obviously disturbed. Genetic testing via a congenital myopathy

From the Division of Neurology (F.A.A., C.G., K.A., H.D.G., J.J.D.), Hospital for Sick Children; Department of Pathology and Laboratory Medicine (L.-N.H.), Hospital for Sick Children; and Department of Pediatrics and Molecular Genetics (J.J.D.), University of Toronto, Ontario, Canada.

Go to Neurology.org/NG for full disclosures. Funding information is provided at the end of the article.

(A) Patient at age 4 years. Note the long and slightly myopathic facies but no clear facial weakness and no restrictions of eye movements. (B–E) Biopsy of the right quadriceps muscle at age 3 years. (B) Hematoxylin and eosin staining showed variability in muscle fiber size and shape and rare fibers with central nuclei (arrow). (C) NADH staining. (D–E) ATPase pH 4.2 (D) and 9.4 (E) staining showed type I predominance and relative type I fiber hypotrophy. Scale bars = 30 μm. NADH = nicotinamide adenine dinucleotide.

multigene panel (28 genes, sequencing plus copy number variant membrane phospholipid composition.5,6 This may explain the testing, University of Chicago) identiﬁed a homozygous long observation of myoﬁber hypotrophy seen on muscle biopsy. interspersed nuclear element (LINE) insertion of approximately However, neither our case nor the previously reported one 1,250 bps in exon 6 of HACD1 (c.739_740delins1250). Both exhibits obvious alterations of the tubule-reticular system parents were heterozygous for the mutation. (i.e., no features of centronuclear myopathy).

The previous HACD1 pedigree described was a consanguin- Discussion eous family with 8 affected individuals.3 Each affected individual had a homozygous nonsense mutation in HACD1 Here, we present a patient with homozygous mutation in (c.744C>A; p.Tyr248Ter). Functional studies confirmed that HACD1, representing only the second case report of this mutation led to loss of HACD1 protein expression. All HACD1-related congenital myopathy. The clinical pheno- affected patients presented with a severe phenotype at birth type and muscle biopsy findings of our case match the initial that gradually improved. All patients had reduced muscle tone report and suggest a condition characterized by initially severe and proximal weakness, although the oldest patient (aged 35 weakness that markedly improves over time and nonspecific years) worked as a truck driver with daily heavy lifting without muscle pathology most in keeping with congenital fiber-type limitation. EMG showed myopathic findings, and muscle biopsy disproportion. showed decreased type 1 fiber size, increased type 1 fiber number, and rare internally located nuclei. These findings are strikingly HACD1 encodes 3-hydroxyacyl-CoA dehydratase, a protein similar to our patient and together create a consistent clinical and that catalyzes the third step in VLCFA synthesis. VLCFAs pathologic phenotype associated with HACD1 mutation. are fatty acids with a chain length of >20 carbons that function in numerous biological processes, including skin Of interest, the mutation mechanism in our patient is a LINE barrier formation, fetal growth, and brain development. insertion in exon 6. LINE insertion in HACD1 has been de- There are 4 HACD paralog enzymes responsible for the scribed to cause a form of centronuclear myopathy in Lab- third step of VLCFA synthesis,4 each with differential ex- rador retrievers.6,7 Clinical features in these dogs include pression. HACD1 is expressed mainly in the skeletal muscle hypotonia, generalized muscle weakness, abnormal posture, and heart,5 which likely accounts for the muscle-specific stiff hoping gate, and exercise intolerance. Unlike the patients presentation in our patient, although cardiac abnormalities were described with HACD1 mutation, however, the affected dogs not observed in our case or in the previously reported family. show progressive clinical features and muscle biopsies with HACD1 has been implicated in myoblast fusion and mainte- findings of centronuclear myopathy. The explanation for nance of the tubule-reticular system via its ability to regulate these differences is not known, but most likely relates to

2 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG function variability of HACD1 between species and/or genetic modiﬁcation. Appendix (continued)

Name Location Contribution Study funding No targeted funding reported. Lili-Naz Hospital for Acquisition of data; generation of the Hazrati, MD Sick Children figure; and editorial input on the manuscript Disclosure Kimberly Hospital for Acquisition of data; generation of the This case report has been contributed to, seen, and approved Amburgey, Sick Children figure; and editorial input on the by all the authors. All the authors fulﬁll the authorship credit MSc manuscript requirements. No honorarium grant or other form of payment Hernan D. Hospital for Acquisition and interpretation of data was received for the preparation of this case report. F. Al Gonorazky, Sick Children and editorial input on the manuscript Amrani, C. Gorodetsky, L.-N. Hazrati, K. Amburgey, H.D. MD Gonorazky, and J.J. Dowling report no disclosures relevant to James J. Hospital for Conceived on the study; data the manuscript. There is no ﬁnancial disclosure relevant to Dowling, MD, Sick Children interpretation; and codraft and edit of PhD the manuscript this project. Go to Neurology.org/NG for full disclosures.

Publication history References fi 1. Gonorazky HD, Bönnemann CG, Dowling JJ. The genetics of congenital myopathies. Received by Neurology: Genetics October 9, 2019. Accepted in nal form Handb Clin Neurol 2018;148:549–564. March 20, 2020. 2. Gonorazky HD, Dowling JJ, Volpatti JR, Vajsar J. Signs and symptoms in congenital myopathies. Semin Pediatr Neurol 2019;29:3–11. 3. Muhammad E, Reish O, Ohno Y, et al. Congenital myopathy is caused by mutation of HACD1. Hum Mol Genet 2013;22:5229–5236. Appendix Authors 4. Ikeda M, Kanao Y, Yamanaka M, et al. Characterization of four mammalian 3-hydroxyacyl-CoA dehydratases involved in very long-chain fatty acid synthesis. FEBS Lett 2008;582:2435–2440. Name Location Contribution 5. Blondelle J, Ohno Y, Gache V, et al. HACD1, a regulator of membrane composition and fluidity, promotes myoblast fusion and skeletal muscle growth. J Mol Cell Biol Fatema Al Hospital for Major role in data acquisition and 2015;7:429–440. Amrani, MD Sick Children codrafted the manuscript 6. Walmsley GL, Blot S, Venner K, et al. Progressive structural defects in canine centronuclear myopathy indicate a role for HACD1 in maintaining skeletal muscle Carolina Hospital for Acquisition of data and codrafted the membrane systems. Am J Pathol 2017;187:441–456. Gorodetsky, Sick Children manuscript 7. Pele M, Tiret L, Kessler JL, Blot S, Panthier JJ. SINE exonic insertion in the PTPLA MD gene leads to multiple splicing defects and segregates with the autosomal recessive centronuclear myopathy in dogs. Hum Mol Genet 2005;14:1417–1427.

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 3 CLINICAL/SCIENTIFIC NOTES OPEN ACCESS Expanding the phenotype of MTOR-related disorders and the Smith-Kingsmore syndrome

Anasofia Elizondo-Plazas, MD, Marisol Ibarra-Ram´ırez, MD, Azalea Garza-B´aez, MD, and Correspondence ´ı Laura Elia Mart´ınez-de-Villarreal, MD Dr. Mart nez-de-Villarreal [email protected] Neurol Genet 2020;6:e432. doi:10.1212/NXG.0000000000000432

Heterozygous germline mutations in mammalian target of rapamycin (MTOR)(OMIM 601231) are known to underlie Smith-Kingsmore syndrome (SKS; OMIM 616638), an infrequent entity with autosomal dominant inheritance, also known as macrocephaly- intellectual disability-neurodevelopmental disorder-small thorax syndrome (ORPHA 457485).1 Among the clinical features of SKS, the most common features include intellectual disability, macrocephaly, epilepsy, and facial dysmorphism. The aim of this case is to raise awareness of a distinct phenotypical presentation of SKS manifesting with bilateral cataracts and no history of seizures.

Case presentation A 5-year-old boy with macrocephaly and a history of developmental delay that included lag in motor and language milestones presented to our clinic. He is the firstborn of nonconsanguineous parents with unremarkable medical and family histories. His mother underwent adequate prenatal control and preeclampsia. During the third trimester, macrocephaly was diagnosed on ultrasound suspicious of hydrocephalus. The baby was born by elective cesarean section at 38 weeks because of preeclampsia and oligohydramnios. Birth weight and length were 3,960 g (P90) and 53 cm (P80), respectively. Head circumference referred to be normal although it was not recorded. Diagnosis of cryptorchidism prompted bilateral orchidopexy at 10 months of age. Brain MRI was performed showing left temporal lobe cyst (figure, A and B). At 5 years of age, physical examination was remarkable for macrocephaly and prominent and downward oblique eyes with tele- canthus, strabismus, and bilateral cataracts (figure, C and D). In addition, nevus flammeus on nasal bridge and tip, anteverted nostrils, and a long philtrum were also observed. No viscer- omegalies were palpable, and no other notable features were found. Somatometry showed a weight of 20.5 kg (P50-75), height of 113.5 cm (P75), head circumference of 56 cm (+3 SD), inner intercanthal distance of 4 cm (>2 SD), philtrum of 2 cm in length (+2 SD), palpebral fissures of 3.2 cm in length bilaterally (+1 SD), auricular pavilions of 6 × 4 cm (P50), hand of 13 cm (P75-97), and middle finger of 6 cm (P75-97). On cranial X-rays, increased anteroposterior diameter (dolichocephaly) was observed with no other evident abnormalities. With clinical suspicion of SKS, an exome sequencing was performed.

Results The clinical exome was made, resulting in the diagnosis of a monogenic disorder, clinically indistinctive from Smith-Kingsmore, Weaver, and Marshall syndromes. According to the American College of Medical Genetics and Genomics (ACMG) classiﬁcation, a likely pathogenic variant was identiﬁed in heterozygosis in the MTOR gene, c.5663T>G (p.Phe1888Cys). The reported variant was not found in gnomAD exomes nor genomes. However, in silico analysis (Functional Analysis through Hidden Markov Models, MutationAssessor, MutationTaster, and

From the Department of Medical Genetics (A.E.-P, M.I.-R, L.E.M.-V) and Department of Radiology (A.G.-B.), Hospital Universitario “Dr. Jose Eleuterio Gonzalez”, Universidad Autonoma de Nuevo Leon, Monterrey, Mexico.

Go to Neurology.org/NG for full disclosures. Funding information is provided at the end of the article.

The Article Processing Charge was funded by the Authors. This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloading and sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

(A) Brain MRI in T1 sagittal sequence identifying megalencephaly. (B) Brain MRI in T2 sequence, (B.a) coronal and (B.b) axial both showing an arachnoid cyst in the left temporal pole (Galassi classification type I) (arrows). (C) Macrocephaly: low implantation and dysplastic ear pavilions with prominent antihelix. (D) Oblique downward palpebral fissures of 3.2 cm each (>1 SD) with hypertelorism. Strabismus and bilateral cataracts: flammeus nevus on bridge and nasal tip, anteverted nasal wings, and long philtrum (Likert III). (E) Pectus excavatum and nipples of low implantation with teletelia.

Sorting Intolerant from Tolerant) predicts this variant as polymicrogyria-polydactyly-hydrocephalus and megalencephaly pathogenic. The result obtained is compatible with the genetic capillary malformation syndrome.5 In 2017, 4 patients with SKS diagnosis of the Smith-Kingsmore syndrome. Neither of the were described, claiming that it belongs to the group of parents were tested for this variant because they presented with “mTORopathies,” a term introduced to describe neurologic dis- no clinical suspicion for SKS. orders characterized by the altered cerebrocortical architecture, abnormal neuronal morphology, and intractable epilepsy as a consequence of mTOR signaling hyperactivation, suggesting a histo- Discussion pathologic substrate for epileptogenesis.6 A recently published article reported a de novo MTOR gain of function variant in a pa- ﬁ The rst SKS case reported was a girl with megalencephaly and tient with SKS and antiphospholipid syndrome, expanding both the intractable seizures, in which an exome sequencing showed genetic and phenotypic spectra of MTOR-associated diseases.7 a phenotypically relevant heterozygous de novo variant, c.4448G>T (p.Cys1483Phe) in MTOR.2 Except for the seizures, our patient did show common characteristics of SKS, Conclusion although cataracts have never been described. The detected To date, only 10 MTOR gene variants have been described in variant in the MTOR gene has been previously reported in 28 families with SKS. We hereby describe an unusual pre- a pair of twins, who presented with seizures, cognitive delay, sentation of the spectrum of mTORopathies. In this particular intellectual disability, behavioral disorders, hypotonia, and case, the patient presented with bilateral cataracts and re- 3 ﬁ macrocephaly. It is classi ed as likely pathogenic (Class 2) markably no history of seizures in a 5-year lapse. according to the recommendations of the ACMG.

It is well known that mammalian target of the rapamycin Study funding (mTOR) pathway integrates both intracellular and extracellular No targeted funding reported. signals. As such, it serves as a regulator of cell metabolism, growth, proliferation, and survival. The mTOR protein is a 289-kDa Disclosure serine/threonine kinase and belongs to the phosphoinositide A. Elizondo-Plazas, M. Ibarra-Ram´ırez, A. Garza-B´aez, and 3-kinase (PI3K)–related kinase family. This kinase positively L.E. Mart´ınez-de-Villarreal report no disclosures relevant to regulates cell growth and proliferation by promoting many ana- the manuscript. Go to Neurology.org/NG for full disclosures. bolic processes, including biosynthesis of proteins, lipids, and organelles.4 Mutations in genes in the PI3K/Akt/mTOR Publication history pathway have also been described in multiple (hemi)mega- Received by Neurology: Genetics December 25, 2019. Accepted in ﬁnal lencephaly-associated syndromes, including megalencephaly- form April 9, 2020.

2 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG References Appendix Authors 1. Moosa S, Böhrer-Rabel H, Altmüller J, et al. Smith-Kingsmore syndrome: a third family with the MTOR mutation C.5395G>A p.(Glu1799Lys) and evidence for pa- Name Location Contribution ternal gonadal mosaicism. Am J Med Genet A 2016;173:264–267. 2. Smith L, Saunders CJ, Dinwiddie DL, et al. Exome sequencing reveals de novo Anasofia University Hospital “Dr. Clinical approach on the germline mutation of the mammalian target of rapamycin (MTOR) in a patient with Elizondo-Plazas, José E. Gonzalez,” UANL patient and drafted the megalencephaly and intractable seizures. J Genomes Exomes 2013;2:63–72. MD manuscript 3. Møller RS, Weckhuysen S, Chipaux M, et al. Germline and somatic mutations in the MTOR gene in focal cortical dysplasia and epilepsy. Neurol Genet 2016;2:e118. Marisol Ibarra- University Hospital “Dr. Data collection and 4. Laplante M, Sabatini DM. MTOR signaling at a glance. J Cell Sci 2009;122: Ram´ırez, MD José E. Gonzalez,” UANL revision of the manuscript 3589–3594. 5. Baynam G, Overkov A, Davis M, et al. A germline MTOR mutation in aboriginal Azalea Garza- University Hospital “Dr. Imaging interpretation Australian siblings with intellectual disability, dysmorphism, macrocephaly, and small Báez, MD José E. Gonzalez,” UANL and data analysis thoraces. Am J Med Genet A 2015;167:1659–1667. 6. Gordo G, Tenorio J, Arias P, et al. MTOR mutations in Smith-Kingsmore Laura Elia University Hospital “Dr. Revision for intellectual syndrome: four additional patients and a review. Clin Genet 2018;93; Mart´ınez-de- José E. Gonzalez,” UANL content 762–775. Villarreal, 7. Rodr´ıguez-Garc´ıa ME, Cotrina-Vinagre FJ, Bellusci M, et al. A novel de novo MD MTOR gain-of-function variant in a patient with smith-kingsmore syndrome and antiphospholipid syndrome. Eur J Hum Genet 2019;27:1369–1378.

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 3 VIEWS AND REVIEWS OPEN ACCESS Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy revisited Genotype-phenotype correlations of all published cases

Georgia Xiromerisiou, MD, PhD, Chrysoula Marogianni, MD, MSc, Katerina Dadouli, MSc, Correspondence Christina Zompola, MD, Despoina Georgouli, MD, MSc, Antonios Provatas, MD, PhD, Dr. Xiromerisiou [email protected] Aikaterini Theodorou, MD, Paschalis Zervas, MD, Christina Nikolaidou, MD, Stergios Stergiou, MD, Panagiotis Ntellas, MD, Maria Sokratous, MD, MSc, Pantelis Stathis, MD, PhD, Georgios P. Paraskevas, MD, PhD, Anastasios Bonakis, MD, PhD, Konstantinos Voumvourakis, MD, PhD, Christos Hadjichristodoulou, MD, PhD, Georgios M. Hadjigeorgiou, MD, PhD, and Georgios Tsivgoulis, MD, PhD

Neurol Genet 2020;6:e434. doi:10.1212/NXG.0000000000000434 Abstract Objective The aim of this study was to evaluate the correlation between the various NOTCH3 mutations and their clinical and genetic proﬁle, along with the presentation of a novel mutation in a patient.

Methods Here, we describe the phenotype of a patient with cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL) harboring a novel mutation. We also performed an extensive literature research for NOTCH3 mutations published since the identiﬁcation of the gene and performed a systematic review of all published cases with NOTCH3 mutations. We evaluated the mutation pathogenicity in a great number of patients with detailed clinical and genetic evaluation and investigated the possible phenotype-genotype correlations.

Results Our patient harbored a novel mutation in the NOTCH3 gene, the c.3084 G > C, corresponding to the aminoacidic substitution p.Trp1028Cys, presenting with seizures as the first neurologic manifestation. We managed to find a correlation between the pathogenicity of mutations, severity of the phenotype, and age at onset of CADASIL. Significant differences were also identified between men and women regarding the phenotype severity.

Conclusions The collection and analysis of these scarce data published since the identification of NOTCH3 qualitatively by means of a systematic review and quantitatively regarding genetic profile and pathogenicity scores, highlight the significance of the ongoing trend of investigating phenotypic genotypic correlations.

From the Department of Neurology (G.X., C.M., D.G., A.P., M.S., G.M.H.), University Hospital of Larissa, Faculty of Medicine, School of Health Sciences, University of Thessaly, Larissa, Greece; Second Department of Neurology (C.Z., A.T., P.Z., A.B., K.V., G.T.), “Attikon” University Hospital, School of Medicine, National and Kapodistrian University of Athens, 12462 Athens, Greece; Department of Neurology (G.M.H.), Medical School, University of Cyprus, Nicosia, Cyprus; Department of Hygiene and Epidemiology (K.D., C.H.), Faculty of Medicine, University of Thessaly, Larissa, Greece; Department of Medical Oncology (P.N.), University Hospital of Ioannina, Ioannina, Greece; Department of Neurology (P.S.), Mediterraneo Hospital, Glyfada, Athens, Greece; Histopathological Department (C.N., S.S.), Hippokration General Hospital Thessaloniki; and Department of Neurology (G.P.P.), School of Medicine, National and Kapodistrian University of Athens, Eginition Hospital, Athens, Greece.

Go to Neurology.org/NG for full disclosures. Funding informaton is provided at the end of the article.

Copyright © 2020 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 Glossary CADASIL = cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy; CADD = combined annotation-dependent depletion; CI = conﬁdence interval; CMB = cerebral microbleed; ECD = extracellular domain; EGFr = epidermal growth factor-like repeat; GOM = granular osmiophilic material; TIA = transient ischemic attack.

Cerebral autosomal dominant arteriopathy with subcortical EGFR. Most of these mutations are missense mutations, al- infarcts and leukoencephalopathy (CADASIL) is the most though there are only a few in-frame deletions or splice-site common heritable cause of stroke in adults younger than 65 mutations.11 The gene mutation analysis of NOTCH3 is the years old.1 Although the description of the first case was made gold standard to diagnose this genetically inherited disease, around 1955,2 the official characterization of the disorder was and there are more than 230 different mutations located in 20 defined in 1993 after the discovery of the responsible gene, different exons reported in patients with CADASIL.12 NOTCH3 on chromosome 19.3 Although the clinical manifestations of the disease are directly linked to the brain lesions, In this review, we report a patient with CADASIL with a novel there is a systemic arteriopathy that affects the skin, the spleen, heterozygous NOTCH3 mutation and epileptic seizures as the the liver, the kidneys, and the aorta apart from the brain.4 very first manifestation of the disorder. In silico analysis revealed the pathogenicity of the mutation. However, we CADASIL consists of 4 basic clinical characteristics, which are proceeded to the skin biopsy to detect deposits of GOM. migraine with aura, relapsing episodes of transient ischemic attacks (TIAs) and ischemic strokes, psychiatric symptoms as In addition, we performed a systematic review of all published apathy and severe mood swings, and gradual cognitive im- cases with NOTCH3 mutations. We evaluated the mutation pairment, which eventually lead to severe dementia.3 pathogenicity in all previously reported cases to investigate possible phenotype-genotype correlations. Another fundamental element of this disorder is the leukoencephalopathy and the subcortical infracts identified on the brain MRI, especially in the external capsules and anterior Methods pole of temporal lobes.5 Case description: clinical findings A 62-year-old woman visited the outpatient stroke clinic of a ter- Pathologic findings have confirmed the extensive changes in tiary care stroke center in Athens, Greece. The patient experienced the brain parenchyma, compatible with chronic small artery episodes of loss of consciousness, some of them accompanied by disease, mainly affecting the white matter in the periventricular tonic contraction of the upper arms and bite of her tongue, for the areas and the region of basal ganglia. It is worth mentioning that past 40 years. She has been receiving an antiepileptic drug (leve- the cortex, which appeared unaffected in neuroimaging, in tiracetam 1250 mg daily) for the past 2 years because of repeat a macroscopic examination displays extended neuronal apo- episodes of tonic-clonic seizures. She had also experienced an ptosis. Furthermore, in microscopic testing of the brain lesions, fi ischemic lacunar stroke in the distribution of the left middle ce- aspecic arteriopathy has been revealed in which there is rebral artery 4 months before her visit to the outpatient clinic of a thickening of the arterial wall of small penetrating cerebral ’ 6 our department that resulted in right hemiparesis. The patient s and leptomeningeal arteries, leading to lumen stenosis. At the husband and children reported changes in her personality and same time, with an electronic microscope in a specimen from cognitive decline, gradually worsening over the past 5 years. Her a pathologic skin biopsy, deposits of granular osmiophilic ma- family history revealed a brother with epilepsy and a father with terial (GOM) located in the basement membrane of smooth posttraumatic epilepsy who died at the age of 60 years old. muscle cells can be identified.7 CADASIL is the only disorder fi whose GOM has been identi ed. However, some reports on The patient had no history of hypertension or diabetes or other the sensitivity of detecting GOM in skin biopsy of patients with known cardiovascular risk factors. She was receiving escitalo- 8,9 genetically proven CADASIL have been contradictory. pram (20 mg) for depression, levetiracetam (1250 mg) for seizures, clopidogrel (75 mg), and atorvastatin (20 mg) for CADASIL is an autosomal dominant inherited arteriopathy secondary stroke prevention. caused by mutations in the NOTCH3. The NOTCH3 gene encodes a single pass transmembrane protein, with receptor The patient was admitted in our department for further di- properties. This receptor is mainly expressed in the smooth agnostic workup. We repeated brain MRI that revealed extensive muscle cells of blood vessels and pericytes.10 After ligand bilateral white matter lesions located mainly in the frontal lobes, binding, the intracellular part translocates to the nucleus and the anterior temporal lobe, the external capsules, the centrum activates transcription factors. NOTCH3 has 33 exons, but all semiovale, and the basal ganglia (figure 1, A–C). There was no the mutations found are located in exons 2–24, which are evidence of acute infarction. Furthermore, multiple cerebral responsible for encoding 34 epidermal growth factor repeats, microbleeds (CMBs) located predominantly in the thalami were

2 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG Figure 1 Brain MRI scan (A–C) of the patient showing extensive leukoencephalopathy, mainly in the frontal lobes, the anterior temporal lobe, the external capsules, and the basal ganglia in the FLAIR sequences (D–F) of the daughter showing multiple hyperintensities in the FLAIR sequences

documented on susceptibility-weighted imaging. The patient available family members was extracted from the peripheral white underwent an extensive workup for stroke (including trans- blood cells according to the standard protocols. esophageal echocardiography, neck, and brain MR angiography, full coagulation disorder panel, repeat 24-hour holter ECG The NOTCH3 analysis led to the identification of a heterozy- recordings, molecular genetic screening for Fabry disease, im- gous mutation in the exon 19 NOTCH3 gene, c.3084 G > C, munologic screening for autoimmune disorders) that was un- corresponding with the amino acid substitution p.Trp1028Cys. remarkable. We also found no evidence for demyelinating CNS This is a novel mutation, compatible with CADASIL syn- disorder on CSF analysis (normal results with absent oligoclonal drome, leading to the gain of a cysteine residue in one of the 34 bands) and cervical as well as thoracic spinal cord MRI. epidermal growth factor-like repeat (EGFr) domains of the protein encoded by NOTCH3 (1,312). This results in an un- Bedside neuropsychologic examination showed moderate even number of cysteine residues in the given EGFr domain, cognitive decline (Mini-Mental State Examination 20/30) with most likely modifying the tertiary structure of the protein. defects, mainly in the attention and recall ability and a Frontal Assessment Battery (6 of 18) that revealed a serious disorder in According to the American College of Medical Genetics and conceptualization and mental flexibility. Genomics and the Association for Molecular Pathology 2015 guidelines, the pathogenicity potential of the p.Trp1028Cys The patient’s daughter complained of chronic tension-type variant is “pathogenic” based on the following criteria: (1) null headaches and was further investigated with brain MRI that variant (PVS1)—amissensemutationleadingtoagainof revealed multiple hypertense subcortical white matter lesions acysteineresidue,(2)theabsenceofthevariantfromthecon- (figure 1, D–F). The patient’s son was asymptomatic and re- trols in the Exome Sequencing Project, 1,000 Genomes Project, fused to undergo brain MRI despite our suggestions. or Exome Aggregation Consortium (PM2), (3) in silico bioinformatics tools (Homologene, GEPR, Varsome) predicted Case description: genetic analysis that the variant causes a deleterious effect on the gene (PP3) Given the aforementioned extended workup, the relevant clinical because it occurs in a highly conserved area across multiple course, the symptoms, and the compatible neuroimaging, the species accordingly (ncbi.nlm.nih.gov/homologene), and (4) suspicion of CADASIL was raised, so we proceeded to NOTCH3 the patient’s phenotype is highly specific for the disease (PP4). genetic analysis. Based on the mode of inheritance, the age at onset, and the phenotype, we suspected a genetic disorder. All Case description: histopathologic staining study participants provided written informed consent for further For immunohistochemistry, we used 10-lm tissue specimens. genetic analysis. Our study was approved by the Ethics Com- Endogenous peroxidase activity was eliminated by treatment mittee of the University Hospital of Attikon. Genomic DNA of all with 0.5% periodic acid solution for 10 minutes. The sections

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 3 Figure 2 Flow chart depicting our literature search and study selection for the systematic review

were incubated in a blocking buffer (1% bovine serum albu- Data collection and eligibility criteria min and 5% rabbit serum in phosphate-buffered saline). A Studies were selected when they met the following criteria: (1) monoclonal antibody against the extracellular domain of diagnosis of CADASIL was confirmed by a mutation analysis, NOTCH3 diluted 1:100 in blocking buffer, served as the (2) clinical findings of CADASIL were described, (3) the primary antibody. The sections were counterstained with identified mutation was described, (4) when the same patient Victoria blue. To observe the internal elastic lamina of vas- population was presented in more than one publication, the cular walls, we used Victoria blue–hematoxylin and eosin one with the largest sample size or most recent publication date staining. No specific immunoreactive signal was detected in was used as the primary study for data extraction, (5) languages the vessels; therefore, the specimen was scored as negative. being limited to English, and (6) Full study.

Review of published CADASIL data Evaluation of mutations For the evaluation of the pathogenicity of the published muta- Systematic search tions, computational prediction analysis was used. A combined We conducted a systematic review of the literature, in- annotation-dependent depletion (CADD) algorithm was used vestigating all known mutations of NOTCH3 and their phe- for scoring the deleteriousness of single nucleotide variants and notypic characteristics. We performed searches on PubMed, insertion/deletions on the function of NOTCH3.Predictionis Cochrane Library, and MEDLINE (via PubMed) to identify based on empirical rules applied to the sequence, phylogenetic, all published studies before August 2019. A combination of and structural information characterizing the amino acid sub- the terms“CADASIL,”“Cerebral autosomal dominant arterio- stitution.13,14 A qualitative characterization of the mutations was pathy with subcortical infarcts and leukoencephalopathy,” based on the score: the CADD score is above >30, was highly “NOTCH3,” and “NOTCH3 MUTATIONS” were used. All pathogenic, above >20 was pathogenic, between 15 and 20 was studies presenting original data that reported the clinical, likely pathogenic, below <15 was likely benign, and finally below genetic, and radiologic characteristics of patients with <10 was benign. The Human Splicing Finder tool37 was used to CADASIL were included for further review. English language evaluate mutations that potentially affect splicing.15 was the only filter used initially. References of selected articles and reviews were also searched for additional records. Statistical analysis We used descriptive statistics to present demographic, clinical, and Data extracted from each study were title; author; year of other characteristics of the patients overall. Data were checked for publication; the exon, the mutation and the exact amino acid deviation from normal distribution (Shapiro-Wilk normality test). change, the total cases screened carrying the mutant allele and Categorical data were analyzed with the use of a χ2 test; Student t sociodemographic characteristics (origin, age, and sex), age at test, Mann-Whitney U test, analysis of variance, Kruskal-Wallis onset, family history, clinical features of the disease (migraine, test and were performed for continuousdataasappropriate. stroke, psychiatric disorders, cognitive decline, acute en- Spearman correlation analyses were used to estimate the corre- cephalopathy, and atypical findings), and findings from di- lations between quantitative variables. A multivariate analysis was agnostic procedures (MRI findings and skin biopsy). performed in the form of multiple linear regression and

4 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG Figure 3 A world map showing the frequencies of CADASIL mutations across various countries according to our reviewed cases

multinomial regression and multinomial (polytomous) logistic China, and Italy. The frequencies of CADASIL mutations across regression. The cutoff value of age at onset was calculated using various countries according to our reviewed cases are presented the receiver operating characteristic curve. For all the analyses, in a world map shown in figure 3. However, the largest case a 5% significance level was set. The Bonferroni method was used series with CADASIL that were excluded from our analysis to correct the significance level where it was necessary.16,17 The because of the lack of detailed clinical and genetic information analysis was carried out with SPSS version 25.0. originated from the Netherlands, Germany, UK, and Japan.

Detailed clinical information was available for 752 patients. The Results overall mean age of the reported cases was 52 ± 14 years, whereas the mean age at onset was 43 ± 14 years. Apprimately Literature review 64% of the patients that have been included in the study had We screened a total of 695 articles yielded by our literature a positive family history, 10% had no family history, and no search; we excluded 476 that evaluated animal models, lacked information was available for the remaining 25%. clinical, and genetic data and that were duplicates. After full- text retrieval and review, we included the 224 remaining Migraine was reported in 23% of patients. Most of all reported articles because they met the predefined eligibility criteria and cases presented with stroke (52%), followed by cognitive de- we proceeded in the qualitative analysis (Supplementary 1, cline (46%) while psychiatric disorders had a prevalence of links.lww.com/NXG/A262). The selected studies consisted 24%. Psychiatric disorders included predominantly depressive of noncontrolled case series and case reports (figure 2). symptoms, apart from a minority of cases with bipolar disorder (2 cases) and psychosis (1 case). Epileptic disorders were Demographics: clinical phenotype reported in 4% of the patients. A total of 752 patients (50% men) were included in our systematic review. Most of the patients were Caucasians of Euro- There are certain clinical characteristics mentioned in the in- pean origin (45%) and Asians followed with 43%. Only a few cluded studies, which are atypical or in some cases, could be patients were included originating from North and South explained from the recurrent stroke episodes that the patients America, Africa, and Australia/New Zealand. Most case reports suffered. For example, we have found sensory deficits of the upper and small case series report patients originated from Japan, arms and findings of polyneuropathy in 10 cases. Furthermore,

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 5 Figure 4 The distribution of the most common mutations across the NOTCH3 gene

several cases (approximately 7%) with intracerebral hemorrhage not find any information in 25% of the patients regarding the have been described. An interesting fact is the presence of presence of white matter lesions. Furthermore, CMBs have movement disorders such as ataxia and parkinsonism in ap- been reported in 7% of cases and were absent in 28% of the proximately 4% of our included patients. Symptoms such as brain MRI scans. However, in most cases (65%), the presence dizziness, vertigo, gait disturbance, and dysarthria could probably of microbleeds was not reported. be consequences of the recurrent strokes. At some point in the course of their disease, approximately 2% of patients presented Skin biopsy data, a sensitive diagnostic tool in the diagnosis of a confusional state and/or encephalopathy (CADASIL coma) CADASIL, were not available for more than half of the after severe headache that lasted several days.18,19 reported cases. Approximately 39% of patients had been tested with a skin biopsy. Approximately 14% of the patients We divided the patients according to the presentation of one had compatible findings with the diagnosis of CADASIL, or a combination of 2 or all of the above major clinical whereas 24% were tested negative for the characteristic characteristics [phenotype 1 = patients with migraine, phe- granular deposits in the basal lamina of blood vessels. In most notype 2 = patients with migraine and stroke or stroke only, studies, GOM was confirmed with electron microscopy, and phenotype 3 = (1) patients with migraine and stroke and whereas in the rest of them with immunohistochemistry. cognitive disorder and/or psychiatric disorder and (2) patients with stroke and cognitive decline and/or psychiatric Mutations disorder]. We noticed that most patients, 34%, presented with Most mutations described were located in the exon 4 (29%), phenotype 3 followed by phenotype 2 (32%). Phenotype 1 followed by exon 3 (14%), exon 11 (8%), and exon 19 (6%) was documented in 7% of the patients included in this sys- (figure 4). Most of the mutations are predicted to result in tematic review. a gain or loss of at least one of a cysteine residue; 14% of the mutations described did not involve a cysteine residue. Most Neuroimaging and skin biopsy mutations were missense mutations (74%), whereas a few The predominant radiologic manifestations of CADASIL on splicing mutations and deletions were also reported. In total, the brain MRI include hyperintense located in the white 85% of these mutations were pathogenic and 12% were found matter of the anterior temporal poles, centrum semiovale, to be highly pathogenic, whereas approximately 1% corre- external capsule, basal ganglia, and pons. Most of the patients sponded to benign mutations. Regarding quantitative patho- had radiologic signs of leukoencephalopathy in their brain genicity scores, the mean score was 26 (max = 44, min = 0.6, MRI scans (72%), where the precise distribution of the white SD = 4). The most common mutation was p.R169C in exon 4, matter lesions was not mentioned in most of them. We did followed by p.R182C in the same exon.

6 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG Pathogenicity score appears to be solely aﬀected by the lo- Figure 5 Univariate analysis showing the effect of patho- cation of the mutation (p < 0.001) and is not associated with genicity score of the mutations on the age at on- phenotype severity (ﬁgure 8). All 3 phenotypes described set of the disorder harbor the same pathogenic mutations in qualitative and quantitative analysis.

The clinical manifestations and phenotype severity did not vary significantly between patients harboring cysteine and cysteine-sparing mutations. However, the age at onset was significantly earlier (mean difference = 7.8 years 95% CI = 2.2, 13.5) in patients with cysteine-sparing mutations (p = 0.009). The mean age at onset in patients with cysteine and cysteine- sparing mutations was 51 and 43 years old, respectively.

Regarding the clinical profile and the phenotypic characteristics among Asian and Caucasian patients, there are certain significant differences. In particular, phenotypes 1 (patients with migraine) and 2 (patients with migraine and stroke or stroke only) of CADASIL were more prevalent in Caucasians than in Asians (figure 3). Regarding distribution of mutations, there are several differences between Asians and Caucasians because cysteine mutations are more prevalent in Asian populations (figure 4).

Genotype-phenotype correlation Initially, we performed several univariate analyses concerning Discussion the associations of all clinical characteristics and phenotypes The present review summarizes the clinical and neuroimaging with genotypic characteristics and pathogenicity scores. We findings of a novel mutation in exon 19 of the NOTCH3 gene, found that the age at onset of CADASIL was significantly as- the c.3084 G > C, corresponding to the aminoacidic sub- sociated with the pathogenicity score of the mutation (rho = stitution p.Trp1028Cys. We have also documented that type −0.165, p < 0.001). Highly pathogenic and pathogenic mutations appear to induce an earlier age at onset compared with likely pathogenic, benign, and likely benign variants. The higher Figure 6 Bar chart showing the differences between men the pathogenicity score, the earlier the onset of the disease and women regarding phenotype severity (figure 5).

Multivariable analyses, taking into account several clinical and genetic factors, further indicate that the age of disease onset was independently (B = −0.4, 95% conﬁdence interval [CI] = −0.7 to 0.1; p < 0.004) associated with pathogenicity score (table e-1, links.lww.com/NXG/A263).

Regarding phenotype severity, women appear to have a diﬀer- ent phenotype compared with men (p = 0.003), as presented in ﬁgure 6. Migraine is more common by 44% in women than men, while women usually present a less severe phenotype.

Phenotype severity is also highly correlated with age (p < 0.001) (table e-2, links.lww.com/NXG/A264). Migraine is prevalent at the age of 40 years, whereas a more severe phenotype that contains stroke or stroke and migraines is more prevalent in older patients (50–60 years). Cognitive decline or other psychiatric disorders manifest at the same age range (50–60 years) as stroke (figure 7). This indicates a cumulative effect of neurologic deficits with the progression of the disorder. More specifically, the mean age at onset in patients presenting with migraine, stroke and cognitive decline, or other Phenotype severity-sex, p value = 0.003. psychiatric disorders is 40, 52 and 55 years, respectively.

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 7 and pathogenicity of the mutation are independently associated with the age of disease onset. Finally, phenotypic and genotypic Figure 7 Box plot depicting the distribution of phenotypes characteristics diﬀer signiﬁcantly between Asians and of CADASIL across age Caucasians.

Most mutations in the NOTCH3 gene are located within the large extracellular domain (ECD) of the transmembrane receptor. This domain consists of 34 epidermal growth factor (EGF)-like repeats that contain cysteines. The pathogenic mutations are associated with changes in the number of cysteines, leading to a misfolding of the receptor. This misfolding contributes signiﬁcantly to the formation of oligomers and ECD aggregation, which is considered to be the pathogenic mechanism of the disease.20,21

A few mutations have been found that do not affect the number of cysteines. The interesting fact is that these mutations result in an uneven number of cysteine residues. The pathogenic role of these mutations is controversial.22 Several of these atypical mutations appear to be associated with conformational changes Phenotype severity-age, p < 0.001. in the protein similar to the changes observed for typical cysteine involving mutations. However, these mutations seem to cause a similar phenotype without any significant differences regarding Many technical factors, related to the actual biopsy or the severity the age at onset or phenotype severity.21 Our findings clearly of the disorder, may be the main reasons for the discrepancy across demonstrate this concentration of mutations in the ECD domain, studies regarding the sensitivity of the method. According to our especially in exon 4 and 3.23 Furthermore, they highlight the study, it seems that the implementation of this tool is not widely significant association of the location of the mutations with recognized. A small percentage of studies reported electron mi- pathogenicity score and phenotype severity. We have not found croscopy studies on skin biopsies or immunohistochemistry any differences in the age at onset and the clinical phenotype analysis using a NOTCH3 monoclonal antibody.28 The studies overall regarding the cysteine and cysteine-sparing mutations. that did not show the typical depositions presented similar phe- These results are consistent with most previous studies in Cau- notype with the most of all the other described cases and involved casian or Asian populations. However, there are small regional patients with typical and cysteine sparing mutations as well. In studies in Taiwan and Korea that presented different results, most centers investigating and examining patients with CADASIL indicating that the most common mutations were in exon 11 and phenotype, it seems that skin biopsy is reserved for patients with 18. These differences may simply reflect a founder effect.24,25 unclassified variants. Genetic testing is the core diagnostic tool for the disorder. Our described case failed to manifest these charac- Our study investigated several possible associations between teristic findings with immunochemistry. Electron microscopy the genetic profile and clinical manifestations in all previously studies are more sensitive than immunochemistry.29 Technical published cases within the parameters stated earlier. Several reasons, the location of the biopsy, and the severity of the disorder studies have thoroughly described the phenotypic manifes- could be possible explanations for the lack of typical findings.30 tations and the natural history of the disorder, evaluating the frequency and age distribution of several characteristics.26 Furthermore, we noticed in our study that leukoencephalopathy There are also other studies, with a limited number of in MRI scans is reported in most patients with genetically patients, investigating possible phenotypic-genotypic corre- confirmed CADASIL and CMBs in a minority of them. How- lation by comparing patients carrying mutations involving ever, a detailed description of MRI findings is not universally cysteine residues with patients with cysteine-sparing muta- followed and so certain clinicoradiological associations cannot tions without showing any significant differences.27 be drawn. There are several other radiologic focused studies showing that the MRI lesion load and pattern can vary quite It has also been found that genetic variants located in the significantly across patients. A consistent finding is the presence EGFr domain 7–34 are identified in the general population, of symmetrical white matter hyperintensities on T2-weighted whereas in patients with CADASIL the variants are pre- and fuid-attenuated inversion recovery images.31 Anterior dominately located in the EGFR 1. This must be further temporal pole changes have been associated with high sensi- investigated with wider population studies to elucidate the tivity and specificity for the disorder. External capsule changes role of many variants in small vessel disease phenotype. also have a high sensitivity but low specificity. The occurrence of CMBs is disproportionate across published cases.32 Dilated Skin biopsy is considered an important diagnostic tool for perivascular spaces and subcortical and cortical atrophy have CADASIL diagnosis, highly specific but with variable sensitivity. also been detected.33 The only significant finding that was

8 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG Figure 8 Bar chart showing the distribution of mutations across the exons of the gene according to their pathogenicity score

Εxon-pathogenicity score, p value< 0.001. highlighted during our study, with the limited data available, was Regarding cognitive decline, we should also mention that the fact that MRI CMBs seem to be an age-related phenomenon cognitive performance was either not measured with the same associated with the progression of the disorder. assessments across all studies or it was not performed at all.37 Therefore, the presence of impairment is possibly under- Regarding clinical presentation, migraine is often the earliest estimated. Longitudinal studies are also needed to investigate feature of the disease. According to the previous studies, it changes related to disease progression and possible association seems that migraine is the first clinical symptom in 41% of with MRI changes. This will shed light to the pathogenic symptomatic patients and an isolated symptom in 12%.34,35 processes underlying these symptoms.38 Migraine is also reported in approximately 55–75% of Cauca- sian cases, although it is less frequent in Asian populations. Other atypical clinical characteristics have also been reported in TIAs and stroke are reported in approximately 85% of symp- patients with CADASIL.39 A very important manifestation tomatic individuals.35 Several studies have shown that the total seems to be acute encephalopathy or coma.40 This is a rather lacunar lesion load is strongly associated with the development misleading manifestation that has been reproduced several times of disability.4 Cognitive impairment in CADASIL involves in- in the literature but is extremely rare. Acute encephalopathy has formation processing speed and executive functions mostly. It been reported in 2 cases as the initial manifestation of the dis- is also often associated with apathy and depression. Our study order in our reviewed cases. However, an acute encephalopathic presents findings consistent with previous research. However, presentation of the disease was previously described in 10% of we managed to investigate the effect of several factors on patients who were part of a the British CADASIL prevalence phenotype severity according to the presence of several clinical study.41 All these patients had a history of migraine. The severe symptoms. At some point, the results clearly indicated that symptoms that they presented were episodes of migraine or phenotype severity is an age-related phenomenon so the dis- epileptic seizures that lasted longer than usual, resulting in order seems progressive with the accumulation of new lacunar confusion and disorientation and were self-limited. infarcts that affect cognition. Pathogenicity score does not significantly affect the combination of symptoms that a patient Several small cohort studies have previously described specific presents, but it appears to be related to the age at onset. We phenotypic and genotypic characteristics of patients with cannot make assumptions based on the pathogenicity of the CADASIL in detail (Supplementary data, links.lww.com/ mutation for the severity of the disorder, so genetic advice NXG/A262). They have also focused on specific findings, should be limited to the presence of the disorder and not to the such as ophthalmologic manifestations, cognitive profile, and appearance of specific phenotypic characteristics. Another im- MRI features or the effect of other factors such as cardio- portant clinical conclusion that we drew was the fact that vascular risk factors on phenotype.42,43 Some of these studies phenotype severity is also affected by sex, with men manifesting include a large number of patients >200. However, the clinical more severe symptoms.36 and genetic information for every patient separately is not

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 9 publicly available. This fact prevented us from including these patients in our analysis and posed a bias concerning the fre- Appendix (continued)

quency of mutations across several countries and more de- Name Location Contribution tailed genotypic phenotypic correlations. Panagiotis Ntellas, University Hospital Statistical analysis MD of Ioannina, However, we strongly believe that the collection and analysis of all Ioannina, Greece. these scarce data published since the identification of NOTCH3 Maria Sokratous University of Data extraction qualitatively by means of a systematic review and quantitatively in MD, MSc Thessaly, Larissa, terms of genetic profile and pathogenicity scores, highlight the Greece fi signi cance of the ongoing trend of investigating phenotypic ge- Pantelis Stathis, MD, Mediterraneo Revised the manuscript notypic correlations. PhD Hospital, Athens, for intellectual content Greece

Study funding Georgios P. National and Revised the manuscript No targeted funding. Paraskevas, MD, Kapodistrian for intellectual content PhD University of Athens, Greece Disclosure Anastasios Bonakis, National and Revised the manuscript All authors declare that they have no conﬂicted interests. Go MD,PhD Kapodistrian for intellectual content to Neurology.org/NG for full disclosures. University of Athens, Greece

Publication history Konstantinos National and Revised the manuscript ﬁ Voumvourakis, MD, Kapodistrian for intellectual content Received by Neurology: Genetics January 7, 2020. Accepted in nal form PhD University of April 2, 2020. Athens, Greece

Christos University of Statistical analysis Hadjichristodoulou, Thessaly, Larissa, MD, PhD Greece

Appendix Authors Georgios M. University of Interpreted the data; Hadjigeorgiou, MD, Cyprus, Nicosia, revised the manuscript Name Location Contribution PhD, Cyprus for intellectual content

Georgia University of Design and Georgios Tsivgoulis, National and Interpreted the data; Xiromerisiou MD, Thessaly, Larissa, conceptualized the study; MD, PhD Kapodistrian revised the manuscript PhD Greece analyzed the data; drafted University of for intellectual content; the manuscript for Athens, Greece case description intellectual content

Chrysoula University Hospital literature search Marogianni, MD, of Larissa, Greece References MSc 1. Tournier-Lasserve E, Joutel A, Melki J, et al. Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy maps to chromosome Katerina Dadouli, University of Statistical analysis 19q12. Nat Genet 1993;3:256–259. MSc Thessaly, Larissa, 2. Van Bogaert L. Encéphalopathie sous corticale progressive (Binswanger) à Greece evolutionrapide´ chez des soeurs. Méd Hellen 1955;24:961–972. 3. Federico A, Bianchi S, Dotti MT. The spectrum of mutations for CADASIL diagnosis. “ ” Christina Zompola, Attikon University Case description Neurol Sci 2005;26:117–124. MD Hospital, Athens, 4. Tan RYY, Markus HS. CADASIL: migraine, encephalopathy, stroke and their inter Greece relationships. PLoS One 2016;11:e0157613. 5. O’Sullivan M, Jarosz JM, Martin RJ, et al. MRI hyperintensities of the temporal lobe Georgouli Despoina, University Hospital literature search and external capsule in patients with CADASIL. Neurology 2001;56:628–634. MD, MSc of Larissa, Greece 6. Craggs L, Yamamoto Y, Ihara M, et al. White matter pathology and disconnection in the frontal lobe in cerebral autosomal dominant arteriopathy with subcortical infarcts and Antonios Provatas, University Hospital Data extraction leukoencephalopathy (CADASIL). Neuropathol Appl Neurobiol 2014;40:591–602. MD, PhD of Larissa, Greece 7. Goebel HH, Meyermann R, Rosin R, Schlote W. Characteristic morphologic manifestation of CADASIL, cerebral autosomal-dominant arteriopathy with subcortical “ ” Aikaterini Attikon University Case description infarcts and leukoencephalopathy, in skeletal muscle and skin. Muscle Nerve 1997;20: Theodorou, MD Hospital, Athens, 625–627. Greece 8. Markus HS, Martin RJ, Simpson MA, et al. Diagnostic strategies in CADASIL. Neurology 2002;59:1134–1138. “ ” Paschalis Zervas, Attikon University Case description 9. Tikka S, Mykkänen K, Ruchoux M-M, et al. Congruence between NOTCH3 muta- MD Hospital, Athens, tions and GOM in 131 CADASIL patients. Brain 2009;132:933–939. Greece 10. Joutel A, Corpechot C, Ducros A, et al. Notch3 mutations in CADASIL, a hereditary adult-onset condition causing stroke and dementia. Nature 1996;383:707–710. Christina Hippokration Histopathologic 11. Monet-Lepretreˆ M, Bardot B, Lemaire B, et al. Distinct phenotypic and functional Nikolaidou, MD General Hospital examination features of CADASIL mutations in the Notch3 ligand binding domain. Brain A J Thessaloniki, Neurol 2009;132(pt 6):1601–1612. Greece 12. Muiño E, Gallego-Fabrega C, Cullell N, et al. Systematic review of cysteine-sparing NOTCH3 missense mutations in patients with clinical suspicion of CADASIL. Int J Stergios Stergiou, Hippokration Histopathologic Mol Sci 2017;18:1964. MD General Hospital examination 13. Hack R, Rutten J, Lesnik Oberstein SA (1993). CADASIL. 2000 Mar 15 [Updated Thessaloniki, 2019 Mar 14]. In: Adam MP, Ardinger HH, Pagon RA, et al. (editors), GeneReviews® Greece [Internet]. Seattle (WA): University of Washington, Seattle; 1993-2020. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1500/

10 Neurology: Genetics | Volume 6, Number 3 | June 2020 Neurology.org/NG 14. Rentzsch P, Witten D, Cooper GM, Shendure J, Kircher M. CADD: Predicting the 29. Gustavsen WR, Reinholt FP, Schlosser A. Skin biopsy findings and results of neu- deleteriousness of variants throughout the human genome. Nucleic Acids Research ropsychological testing in the first confirmed cases of CADASIL in Norway. Eur J 2019;47:D886–D894. Neurol 2006;13:359–362. 15. Human Splicing Finder—Version 3.1. (n.d.). Available at: umd.be/HSF/. Accessed 30. Morroni M, Marzioni D, Ragno M, et al. Role of electron microscopy in the diagnosis October 13, 2019. of cadasil syndrome: a study of 32 patients. PLoS One 2013;8:e65482. 16. Berkeley’s stats website, Statistics for Bioinformatics, Available at: stat.berkeley.edu/ 31. Dziewulska D, Sulejczak D, We¸˙zyk M. What factors determine phenotype of cerebral users/mgoldman. Accessed September 2019. autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy 17. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and pow- (CADASIL)? Considerations in the context of a novel pathogenic R110C mutation in erful approach to multiple testing. J R Stat Soc Ser B (Methodological) 1995;57: the NOTCH3 gene. Folia Neuropathologica 2017;55:295–300. 289–300. 32. Shi Y, Li S, Li W, et al. MRI lesion load of cerebral small vessel disease and cognitive 18. Eswaradass VP, Ramasamy B, Kalidoss R, Gnanagurusamy G. Cadasil coma: unusual impairment in patients with CADASIL. Front Neurol 2018;9:862. cause for acute encephalopathy. Ann Indian Acad Neurol 2015;18:483–484. 33. Puy L, De Guio F, Godin O, et al. Cerebral microbleeds and the risk of incident 19. Ragno M, Pianese L, Morroni M, et al. “CADASIL coma” in an Italian homozygous ischemic stroke in CADASIL (cerebral autosomal dominant arteriopathy with sub- CADASIL patient: comparison with clinical and MRI findings in age-matched het- cortical infarcts and leukoencephalopathy). Stroke 2017;48:2699–2703. erozygous patients with the same G528C NOTCH3 mutation. Neurol Sci 2013;34: 34. Wang Z, Yuan Y, Zhang W, et al. NOTCH3 mutations and clinical features in 33 1947–1953. mainland Chinese families with CADASIL. J Neurol Neurosurg Psychiatry 2011;82: 20. Lackovic V, Bajcetic M, Lackovic M, et al. Skin and sural nerve biopsies: ultrastructural 534–539. findings in the first genetically confirmed cases of CADASIL in Serbia. Ultrastructural 35. Guey S, Mawet J, Hervé D, et al. Prevalence and characteristics of migraine in Pathol 2012;36:325–335. CADASIL. Cephalalgia 2016;36:1038–1047. 21. Santa Y, Uyama E, Chui DH, et al. Genetic, clinical and pathological studies of 36. Burkett JG, Dougherty C. Recognizing CADASIL: a secondary cause of migraine with CADASIL in Japan: a partial contribution of Notch3 mutations and implications of aura. Curr Pain Headache Rep 2017;21:21. smooth muscle cell degeneration for the pathogenesis. J Neurol Sci 2003;212:79–84. 37. Gunda B, Hervé D, Godin O, et al. Effects of gender on the phenotype of CADASIL. 22. Wollenweber FA, Hanecker P, Bayer-Karpinska A, et al. Cysteine-sparing CADASIL Stroke 2012;43:137–141. mutations in NOTCH3 show proaggregatory properties in vitro. Stroke 2015;46: 38. Yoon CW, Kim YE, Seo SW, et al. NOTCH3 variants in patients with subcortical 786–792. vascular cognitive impairment: a comparison with typical CADASIL patients. Neu- 23. Matsushima T, Conedera S, Tanaka R, et al. Genotype–phenotype correlations of robiol Aging 2015;36:2443–2447. cysteine replacement in CADASIL. Neurobiol Aging 2017;50:169.e7–169.e14. 39. Brookes RL, Hollocks MJ, Tan RY, Morris RG, Markus HS. Brief screening of vascular 24. Ueda A, Ueda M, Nagatoshi A, et al. Genotypic and phenotypic spectrum of cognitive impairment in patients with cerebral autosomal-dominant arteriopathy with CADASIL in Japan: the experience at a referral center in Kumamoto University from subcortical infarcts and leukoencephalopathy without dementia. Stroke 2016;47: 1997 to 2014. J Neurol 2015;262:1828–1836. 2482–2487. 25. Lee YC, Liu CS, Chang MH, et al. Population-specific spectrum of NOTCH3 40. Spinicci G, Conti M, Cherchi MV, Mancosu C, Murru R, Carboni N Unusual clinical mutations, MRI features and founder effect of CADASIL in Chinese. J Neurol 2009; presentations in subjects carrying novel NOTCH3 gene mutations. J Stroke Cere- 256:249–255. brovasc Dis 2013;22:539–544. 26. Rutten JW, Dauwerse HG, Gravesteijn G, et al. Archetypal NOTCH3 mutations 41. Feuerhake F, Volk B, Ostertag C, et al. Reversible coma with raised intracranial frequent in public exome: implications for CADASIL. Ann Clin Translational Neurol pressure: an unusual clinical manifestation of CADASIL. Acta Neuropathologica 2016;3:844–853, 2002;103:188–192. 27. Liu X, Zuo Y, Sun W, et al. The genetic spectrum and the evaluation of CADASIL 42. Schon F, Martin RJ, Prevett M, Clough C, Enevoldson TP, Markus HS. “CADASIL screening scale in Chinese patients with NOTCH3 mutations. J Neurol Sci 2015;354: coma”: an underdiagnosed acute encephalopathy. J Neurol Neurosurg Psychiatry 63–69. 2003;74:249–252. 28. Dichgans M, Mayer M, Uttner I, et al. The phenotypic spectrum of CADASIL: clinical 43. Cumurciuc R, Massin P, Paquesˆ M, et al. Retinal abnormalities in CADASIL: a ret- findings in 102 cases. Ann Neurol 1998;44:731–739. rospective study of 18 patients. J Neurol Neurosurg Psychiatry 2004;75:1058–1060.

Neurology.org/NG Neurology: Genetics | Volume 6, Number 3 | June 2020 11