Congenital Myopathies in the Adult Neuromuscular Clinic: Diagnostic Challenges and Pitfalls E341

Home , DNM2, MYH2, Nemaline myopathy, RNF5, SLFN12L, Tau protein

Volume 5, Number 4, August 2019 Neurology.org/NG

A peer-reviewed clinical and translational neurology open access journal

ARTICLE Congenital myopathies in the adult neuromuscular clinic: Diagnostic challenges and pitfalls e341

ARTICLE Systematic review and meta-analysis of cardiac involvement in mitochondrial myopathy e339

ARTICLE Genetic risk of Parkinson disease and progression: An analysis of 13 longitudinal cohorts e348

ARTICLE Genome-wide brain DNA methylation analysis suggests epigenetic reprogramming in Parkinson disease e342 TABLE OF CONTENTS Volume 5, Number 4, August 2019 Neurology.org/NG

e344 Altered CSF levels of monoamines in hereditary spastic paraparesis 10: A case series M. Andr´easson, K. Lagerstedt-Robinson, K. Samuelsson, G. Solders, K. Blennow, M. Paucar, and P. Svenningsson Open Access

e347 MAPT p.V363I mutation: A rare cause of corticobasal degeneration S. Ahmed, M.D. Fairen, M.S. Sabir, P. Pastor, J. Ding, L. Ispierto, A. Butala, C.M. Morris, C. Schulte, T. Gasser, E. Jabbari, O. Pletnikova, H.R. Morris, J. Troncoso, E. Gelpi, A. Pantelyat, and S.W. Scholz Open Access

e345 Novel mutation in HTRA1 in a family with diﬀuse white matter lesions and inﬂammatory features A. Ziaei, X. Xu, L. Dehghani, C. Bonnard, A. Zellner, A.Y. Jin Ng, S. Tohari, B. Venkatesh, C. Haffner, B. Reversade, V. Shaygannejad, and M.A. Pouladi Open Access

e348 Genetic risk of Parkinson disease and progression: An analysis of 13 longitudinal cohorts H. Iwaki, C. Blauwendraat, H.L. Leonard, G. Liu, J. Maple-Grødem, J.-C. Corvol, L. Pihlstrøm, M. van Nimwegen, S.J. Hutten, K.-D.H. Nguyen, J. Rick, S. Eberly, F. Faghri, P. Auinger, K.M. Scott, R. Wijeyekoon, V.M. Van Deerlin, D.G. Hernandez, A.G. Day-Williams, A. Brice, G. Alves, A.J. Noyce, O.-B. Tysnes, J.R. Evans, D.P. Breen, K. Estrada, C.E. Wegel, F. Danjou, D.K. Simon, B. Ravina, M. Toft, Articles P. Heutink, B.R. Bloem, D. Weintraub, R.A. Barker, C.H. Williams-Gray, B.P. van de Warrenburg, J.J. Van Hilten, C.R. Scherzer, A.B. Singleton, and e339 Systematic review and meta-analysis of cardiac M.A. Nalls involvement in mitochondrial myopathy Open Access A. Quadir, C.S. Pontifex, H. Lee Robertson, C. Labos, and G. Pfeffer Open Access e349 New family with HSPB8-associated autosomal dominant rimmed vacuolar myopathy e340 Human GABRG2 generalized epilepsy: Increased S. Al-Tahan, L. Weiss, H. Yu, S. Tang, M. Saporta, A. Vihola, somatosensory and striatothalamic connectivity T. Mozaffar, B. Udd, and V. Kimonis M. Pedersen, M. Kowalczyk, A. Omidvarnia, P. Perucca, S. Gooley, Open Access S. Petrou, I.E. Scheffer, S.F. Berkovic, and G.D. Jackson Open Access Clinical/Scientific Notes e341 Congenital myopathies in the adult neuromuscular e343 Missense mutations in DYT-TOR1A clinic: Diagnostic challenges and pitfalls dystonia S. Nicolau, T. Liewluck, J.A. Tracy, R.S. Laughlin, and M. Milone Z. Iqbal, J. Koht, L. Pihlstrøm, S.P. Henriksen, C. Cappelletti, M.B. Russel, O. Norberto de Souza, I.M. Skogseid, and M. Toft Open Access Open Access e342 Genome-wide brain DNA methylation analysis suggests epigenetic reprogramming in Parkinson disease e346 Adult-onset variant ataxia-telangiectasia J.I. Young, S.K. Sivasankaran, L. Wang, A. Ali, A. Mehta, D.A. Davis, diagnosed by exome and cDNA sequencing D.M. Dykxhoorn, C.K. Petito, G.W. Beecham, E.R. Martin, D.C. Mash, M. Krenn, I. Milenkovic, G. Eckstein, F. Zimprich, T. Meitinger, M. Pericak-Vance, W.K. Scott, T.J. Montine, and J.M. Vance T. Foki, and M. Wagner Open Access Open Access Video TABLE OF CONTENTS Volume 5, Number 4, August 2019 Neurology.org/NG

Corrections e350 Missense mutations in DYT-TOR1A dystonia e354 Genetic risk of Parkinson disease and progression: An analysis of 13 longitudinal cohorts e355 Genome-wide Brain DNA methylation analysis suggests epigenetic reprogramming in Parkinson Cover image disease Protein sequencing of the HSPB8 gene. The mRNA that results from this mutation has five splice variants with size ranging from 27 to 244 aa. Only the first three contain an α-crystallin domain and are coding. See e349 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 © 2019 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 Executive Office, 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 Catherine M. Rydell, 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, Kathleen M. Pieper, Senior Managing Editor, Neurology New York, NY 10001. Contacts: Eileen Henry, tel: 732-778-2261, fax: 973-215- 2485, [email protected] and in Europe: Craig Silver, tel: +44 Lee Ann Kleffman, Managing Editor, Neurology® Genetics 7855 062 550 or e-mail: [email protected]. Sharon L. Quimby, Managing Editor, Neurology® Clinical Practice Careers & Events: Monique McLaughlin, Wolters Kluwer, Two Commerce Morgan S. Sorenson, Managing Editor, Neurology® Neuroimmunology & Neuroinflammation Square, 2001 Market Street, Philadelphia, PA 19103, tel: 215-521-8468, fax: 215- Neurology 521-8801; [email protected]. Andrea Rahkola, Production Editor, Reprints: Meredith Edelman, Commercial Reprint Sales, Wolters Kluwer, Two Robert J. Witherow, Senior Editorial Associate Commerce Square, 2001 Market Street, Philadelphia, PA 19103, tel: 215-356-2721; Karen Skaja, Senior Editorial Associate [email protected]; [email protected]. Special projects: US & Canada: Alan Moore, Wolters Kluwer, Two Kaitlyn Aman Ramm, Editorial Assistant Commerce Square, 2001 Market Street, Philadelphia, PA 19103, tel: Kristen Swendsrud, Editorial Assistant 215-521-8638, [email protected]. International: Andrew Wible, Senior Manager, Rights, Licensing, and Partnerships, Wolters Kluwer; Justin Daugherty, Editorial Assistant [email protected]. Madeleine Sendek, MPH, Editorial Assistant

Publisher Wolters Kluwer Baltimore, MD

Publishing Staff Kim Jansen, Executive Publisher Jessica Heise, Production Team Leader, Neurology Journals Megen Miller, Production Editor Steve Rose, Editorial Assistant Stacy Drossner, Production Associate A peer-reviewed clinical and translational neurology open access journal Neurology.org/NG

Neurology® Genetics

Editor Neurology® Genetics Stefan M. Pulst, MD, Dr med, FAAN Vision : 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, Margherita Milone, MD, PhD editorials, and reviews to elucidate the role Raymond P. Roos, MD, FAAN of genetic and epigenetic variations in Jeﬀery M. Vance, MD, PhD 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

Neurology® Journals

Editor-in-Chief Classification of Evidence Robert A. Gross, MD, PhD, FAAN Review Team

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

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

Ombudsman David S. Knopman, MD, FAAN

Scientiﬁc Integrity Advisor Robert B. Daroﬀ, MD, FAAN ARTICLE OPEN ACCESS Systematic review and meta-analysis of cardiac involvement in mitochondrial myopathy

Asfia Quadir, Carly Sabine Pontifex, BSc, Helen Lee Robertson, MLIS, Christopher Labos, MD, MSc, and Correspondence Gerald Pfeffer, MD, PhD Dr. Pfeffer [email protected] Neurol Genet 2019;5:e339. doi:10.1212/NXG.0000000000000339 Abstract Objective Our goal was to perform a systematic review of the literature to demonstrate the prevalence of cardiac abnormalities identiﬁed using cardiac investigations in patients with mitochondrial myopathy (MM).

Methods This systematic review surveys the available evidence for cardiac investigations in MM from a total of 21 studies including 825 participants. Data were stratiﬁed by genetic mutation and clinical syndrome.

Results We identified echocardiogram and ECG as the principal screening modalities that identify cardiac structural (29%) and conduction abnormalities (39%) in various MM syndromes. ECG abnormalities were more prevalent in patients with m.3243A>G mutations than other gene defects, and patients with mitochondrial encephalopathy, lactic acidosis, and stroke-like episodes (MELAS) had a higher prevalence of ECG abnormalities than patients with other clinical syndromes. Echocardiogram abnormalities were significantly more prevalent in patients with m.3243A>G or m.8344A>G mutations compared with other genetic mutations. Similarly, MELAS and MERRF had a higher prevalence compared with other syndromes. We observed a descriptive finding of an increased prevalence of ECG abnormalities in pediatric patients compared with adults.

Conclusions This analysis supports the presence of a more severe cardiac phenotype in MELAS and myoclonic epilepsy with ragged red ﬁbres syndromes and with their commonly associated genetic mutations (m.3243A>G and m.8344A>G). This provides the ﬁrst evidence basis on which to provide more intensive cardiac screening for patients with certain clinical syndromes and genetic mutations. However, the data are based on a small number of studies. We rec- ommend further studies of natural history, therapeutic response, pediatric participants, and cardiac MRI as areas for future investigation.

From the Hotchkiss Brain Institute (A.Q., C.S.P., G.P.), University of Calgary; Health Sciences Library (H.L.R.), University of Calgary, Alberta; Queen Elizabeth Health Complex (C.L.), Montreal, Quebec; and Department of Clinical Neurosciences (G.P.), Cumming School of Medicine, University of Calgary, Alberta, Canada.

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.

Study sponsorship: Dr. Pfeffer receives institutional support from the Department of Clinical Neurosciences, the Cumming School of Medicine, and Hotchkiss Brain Institute (University of Calgary). 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 © 2019 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 Glossary BBB = bundle branch block; CI = confidence interval; CMR = cardiac MRI; CPEO = chronic progressive external ophthalmoplegia; KSS = Kearns-Sayre syndrome; LVH = left ventricular hypertrophy; MELAS = mitochondrial encephalopathy, lactic acidosis, and stroke-like episodes; MERRF = myoclonic epilepsy with ragged red fibres; MM = mitochondrial myopathy; mtDNA = mitochondrial DNA; WPW = Wolff-Parkinson-White syndrome.

Mitochondrial myopathies (MMs) are clinically heteroge- We also performed a search through the bibliographies of all neous disorders resulting from defects in the respiratory included articles to ensure that no other relevant articles were chain, preferentially affecting organs with high-energy missed, although no additional publications for inclusion were requirements. Cardiac dysfunction is common as part of the identified by these means. phenotype of MM and is manifested by cardiomyopathy, cardiac conduction defects, and/or heart failure.1 The iden- Study selection tification of cardiac dysfunction is particularly important be- We included all studies in which patients had a confirmed di- cause of its impact on quality of life, morbidity, mortality, and agnosis of MM using genetic testing and/or muscle pathology. especially because it is amenable to treatment.2 Some mito- We did not include patients with KSS because this syndrome chondrial syndromes, such as Kearns-Sayre syndrome (KSS) includes cardiac dysfunction as a core feature of its diagnostic include cardiac dysfunction as a core feature.2 However, most criteria. Our objective was to identify all published studies of the common mitochondrial syndromes may or may not documenting cardiac complications of MM and using at least include cardiac dysfunction, which can occur at any point in one modality of cardiac investigation. Studies that did not dis- the disease course, and as such screening investigations are close the investigative modality or presented focused or in- recommended for early detection. At present, the optimal complete data were not included. Cross-sectional studies, cohort investigations and time intervals for screening are not estab- studies, or clinical trials were included in our review. Case lished, and studies describing the natural history of cardiac reports (i.e., small studies of 4 or fewer cases) were excluded. dysfunction in MM have been limited.3 Data extraction This systematic review serves to summarize the available ev- Data extraction was independently performed by 2 of the idence regarding screening investigations to: (1) provide data authors (A.Q. and C.S.P.), using a custom-designed data ex- indicating the diagnostic yield of different cardiac inves- traction form, which was piloted on 5 randomly selected eli- fi tigations, (2) compare the frequency of cardiac abnormalities gible studies before being nalized. In cases of disagreement, 3 in differing mitochondrial syndromes and genotypes, (3) of the authors (A.Q., C.S.P., and G.P.) discussed the dis- compare the frequency of diagnostic abnormalities between crepancies to reach a consensus. pediatric and adult patients, and (4) identify limitations in the evidence and areas for future study. The overall goal of this The information extracted included year of study, duration of study is to systematically demonstrate the prevalence of car- follow-up, number of participants, age, the diagnostic tests diac abnormalities across various MM syndromes. used, the number of subjects with abnormal results with ECG, echocardiography (echo), Holter monitor, cardiac MRI (CMRI), and nuclear medicine studies. We also collected data Methods regarding treatment and clinical outcomes where available. When data were incomplete, we attempted to contact authors Two investigators (A.Q. and G.P.) created a preliminary search of the relevant articles to obtain more detailed data and re- strategy that was subsequently refined by a medical librarian ceived such data for 2 publications.4,5 (H.L.R.). The search was conducted on June 4, 2018, on the following databases: Epub Ahead of Print, In-Process & Other For purposes of our analysis, data from child (aged <18 years) Non-Indexed Citations, Ovid MEDLINE Daily and Ovid and adult (aged ≥18 years) patients were recorded and ana- MEDLINE, EMBASE, Cochrane Central Register of Con- lyzed separately. Adult patients who had disease onset in trolled Trials, Scopus, and Web of Science. Combinations of childhood were still included in the adult group. Data were subject headings, keywords, and synonyms used included MMs, stratified using 2 separate approaches: based on genetic sub- mitochondrial encephalomyopathies, Kearns-Sayre Syndrome, types of MM (m.3243A>G mutation, m.8344A>G mutation, cardiomyopathy, cardiac arrhythmia, and tachycardia. Full other mitochondrial DNA (mtDNA) point mutations, single search terms are available in supplemental data (links.lww.com/ large-scale mtDNA deletions, and nuclear gene mutations) NXG/A159). The authors (A.Q. and G.P.) independently and based on clinical syndromes (mitochondrial encepha- screened all titles and abstracts identified by the initial search. lopathy, lactic acidosis, and stroke-like episodes [MELAS], We obtained full-text versions of studies identified as being chronic progressive external ophthalmoplegia [CPEO] or potentially relevant, which were then independently assessed. other phenotypes of MM). Clinical syndromes were defined Only English-language publications were considered. based on how they had been identified within individual

2 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG articles. In cases in which the specific syndrome was not de- [CI: 0.00–0.49]). Given the heterogeneity in outcome fined but the genotype was available, if sufficient in- reporting, it was not possible to perform analysis of specific dividualized data were presented, we would infer the clinical ECG abnormalities with the exception of bundle branch block syndrome for individual patients (e.g., a patient with the (BBB), which is presented in figure e-1 (links.lww.com/NXG/ m.3243A>G mutation with stroke-like episodes, encepha- A159). This abnormality was seldom reported in MM pop- lopathy, and lactic acidosis would be interpreted as a MELAS ulations, and the overall prevalence was zero in many studies. syndrome case). Patients with unspecified syndromes were included in a category labeled “other phenotypes.” A total of 9 articles included ECG findings in pediatric patients (figure 3). We cannot conclude that there is any Based on this approach, we allowed cases to be counted in both significant difference between syndromes because of the small the genetic group and the clinical syndrome group if in- number of reported participants. The overall prevalence of dividualized data were reported for patients having both a de- ECG abnormalities was 0.55 [CI: 0.38–0.71]. We observed fined clinical syndrome and genetic lesion. The rationale for this descriptively that abnormalities in m.3243A>G patients (0.79 approach was that there would be clinical value to providing [CI: 0.48–1.00]) were higher than detected in other genetic a systematic analysis of data according to both genetic and or phenotypic categories. In comparison to adults, pediatric clinical classifications. However, for overall prevalence estimates, patients with MM had a higher prevalence of ECG abnor- each patient was counted only once. We did not perform a risk of malities (0.55 [CI: 0.38–0.71] compared with 0.39 [CI: bias assessment because the included studies were uncontrolled 0.28–0.50]), as a descriptive finding, which did not achieve cross-sectional studies without an identified intervention. statistical significance (p = 0.07).

Statistical analysis Echocardiography The proportion of patients with abnormal tests was summa- Fourteen of the included studies provided data on echocardi- ff 6 rized with a random e ects meta-analysis, with the estimate ography findings in adult patients. The overall prevalence of of heterogeneity being taken from the inverse-variance. To abnormalities was calculated as 0.29 (CI: 0.17–0.42) (figure 4). account for studies with no outcomes of interest, i.e., zero 7 The best-studied group was patients with the m.3243A>G events, an arcsine transformation was applied. Heterogeneity mutation, in which data from 246 participants were reported. was assessed using the I-squared statistic. Here, it was expected that m.3243A>G and m.8344A>G would Data availability be associated with a higher prevalence of abnormalities than All data pertaining to this work (list of abstracts, articles other genetic defects. This was supported by the data when m.3243A>G and m.8344A>G (0.41 (CI: 0.21–0.62) and 0.44 reviewed, data entry spreadsheet, and statistical analysis) will – be made available upon request by any qualified investigator. (CI: 0.23 0.66)) were compared with single mtDNA deletions (0.05 [CI: 0.01–0.12]), other mtDNA point mutations (0.08 [CI: 0.01–0.20]), and nuclear gene mutations (0.00 [CI: Results 0.00–0.07]). When analyzed by clinical syndrome, MELAS We identified 8,601 articles in our literature search for title and (0.80 [CI: 0.57–0.97]) had a higher prevalence compared with abstract review for relevance. After the primary screen, we CPEO and other phenotypes. It was possible to perform an retained 171 articles for full-text review. After assessing these analysis for left ventricular hypertrophy (LVH), which had studies for eligibility, 21 studies were included in our meta- a prevalence of 0.53 in patients with MELAS (CI: 0.25–0.80) analysis. The Preferred Reporting Items for Systematic Reviews and a prevalence of 0.18 (CI: 0.10–0.27) in patients with MM and Meta-Analyses flowchart is presented in figure 1.8 The basic overall (figure e-2, links.lww.com/NXG/A159). characteristics of the included studies are presented in table 1. Echocardiogram findings in child patients were reported in ECG a total of 9 studies. The overall prevalence of abnormalities In our review, 14 studies provided data on ECG findings in was 0.35 (CI: 0.23–0.47) (figure 5). There were no significant adult patients. The overall prevalence of abnormalities was 0.39 differences in prevalence between the syndromes or geno- (95% confidence interval [CI]: 0.28–0.50) (figure 2). The data types, although the data consisted of a small number of indicate a significantly higher prevalence of ECG abnormalities patients and originated from a small number of studies. Be- in patients with the m.3243A>G mutation (0.65 [CI: tween pediatric and adult patients, there was no difference in 0.33–0.92]) compared with the m.8344A>G mutation (0.26 the overall prevalence of abnormalities. [CI: 0.10–0.46]), single mtDNA deletions (CI: 0.21 [0.13–0.305]), other mtDNA point mutations (0.21 [CI: Cardiac MRI 0.09–0.35]), and nuclear gene mutations (0.13 [CI: CMRI was performed in adult patients in 5 studies, and there 0.02–0.29]). However, the data for several of these groups was substantial heterogeneity in the results. Meta-analyzing all originate predominantly from a single study.9 When analyzed 5 studies revealed an overall prevalence of abnormalities of by clinical syndrome, patients with MELAS had significantly 0.35 (CI: 0.09–0.66), but given the small number of partic- more ECG abnormalities (0.88 [CI: 0.71–0.99]) than patients ipants and studies, it is difficult to draw any further con- with CPEO (0.32 [CI: 0.17–0.49]) or other phenotypes (0.17 clusions (figure e-3, links.lww.com/NXG/A159). Only one

Neurology.org/NG Neurology: Genetics | Volume 5, Number 4 | August 2019 3 Figure 1 PRISMA flowchart

PRISMA = Preferred Reporting Items for Systematic Reviews and Meta- Analyses.

study reported CMRI results in children,10 with data from such as LVH and ventricular pre-excitation abnormalities, only a single participant. were most prevalent in m.3243A>G or MELAS patients. Holter Discussion Holter monitoring was assessed in adults in only 4 – studies5,10 12 with very low diagnostic yield and with an overall Overall, patients with MM have a high prevalence of cardiac prevalence estimate for abnormalities of 0.03 (CI: 0.00–0.14) abnormalities, although the data indicate that MELAS has (figure e-4, links.lww.com/NXG/A159). This suggests that a higher prevalence of ECG abnormalities compared with other Holter monitor is a low-yield investigation for adult patients. clinical syndromes, and the m.3243A>G mutation is associated Only one study included Holter monitor data from children.10 with a higher prevalence of ECG abnormalities than other genetic defects. Abnormalities detected by echo are more preva- Common ECG and echo abnormalities were found in several lent in MELAS and myoclonic epilepsy with ragged red fibres studies of patients with various MM mutations and clinical (MERRF) compared with other syndromes, and when analyzed syndromes, but they were unsuitable for meta-analysis, given by genetic category, the m.3243A>G and m.8344A>G muta- the sparsity of the data. They are presented additively in table tions were associated with a higher prevalence than other e-1 (links.lww.com/NXG/A159). These data provide an mutations. Comparing children to adult patients, we report the overall view of the diversity of findings in these common MM descriptive finding of a higher prevalence of ECG abnormalities syndromes. Overall, the most severe cardiac abnormalities, in children compared with adults with MM, supporting the

4 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG Table 1 Characteristics of included studies

Cohort Cardiac modalities Total participants Mean age of Age Sex Author, year PMID type studied included participants (y) range (y) (M:F)

Akaike et al. 199722 9068909 Adult ECG, echo 5 48.0 20–64 2:3

Anan et al. 199535 7850981 Child, ECG, echo 15 41.0 12–54 9:6 adult

Baik et al. 201019 21189931 Child ECG 57 4.8 NA 26:31

Baik et al. 201220 23038991 Child ECG, echo 27 5.0 NA 11:16

Catteruicca et al. 25559684 Child, ECG, echo, Holter, cardiac 15 46.0 8–71 5:10 201510 adult MRI (CMRI)

Cordeiro et al. 24209401 Child Echo 63 0.77 0.17–1.37 30:33 200918

Florian et al. 201517 26001801 Adult ECG, CMRI 64 46.7 35–65 28:36

Galetta et al. 201436 25139213 Adult ECG, echo 20 55.3 NA 5:15

Hollingsworth et al. 22513320 Adult ECG, echo, CMRI 10 42.5 30–55 5:5 201215

Ikawa et al. 200721 17280875 Child, ECG, echo, MIBI 5 39.8 16–51 2:3 adult

Limongelli et al. 20083621 Child, ECG, echo, Holter 30 39.2 16–62 13:17 201011 adult

Lindroos et al. 26112752 Adult CMRI 14 46.5 36.2–57 4:10 201616

Nesbitt et al. 201324 23355809 Adult Echo 129 NA 0.91–74 50:79

Okajima et al. 9875091 Child, ECG, echo 11 16.5 6–23 8:3 199827 adult

Pfeffer and Mezei 22987704 Adult ECG, echo 15 56.9 18–83 8:7 20123

Ueno and Shiotani 10598894 Adult Echo 10 48.6 NA 6:4 199923

Vydt et al. 200712 17223431 Adult ECG, echo, Holter 12 35.0 18–57 5:7

Wahbi et al. 20105 20177121 Child, ECG, echo, Holter 18 42.6 12–71 8:10 adult

Wahbi et al. 20159 26224072 Adult ECG, echo 272 44.0 28–58 115: 157

Wortmann et al. 17407476 Child ECG, echo 5 5.6 0–14 3:2 200728

Yilmaz et al. 20124 22143423 Adult ECG, CMRI 37 NA 45–61 14:23

Abbreviation: NA = not available. concept that MM has a more severe cardiac phenotype in data are very heterogeneous and from a limited number of childhood compared with adult patients. However, this diﬀer- patients. In studies including m.3243A>G and/or MELAS ence was not signiﬁcant (p = 0.07), perhaps relating to the small patients, 2 studies15,16 reported no CMRI abnormalities, and 1 number of participants in pediatric studies. study17 reported CMRI abnormalities in nearly all patients. This discrepancy can hopefully be addressed in future research. One The majority of the available data are from ECG and echo and study of CMRI in 14 participants showed no structural abnor- both modalities demonstrated a high prevalence of detected malities, but an overall reduction of myocardial glucose uptake, abnormalities. We expected that a higher prevalence of abnor- which may be a biomarker of interest for future study.16 Only malities would be detected with CMRI, given that CMRI is a single study included CMRI data from a pediatric participant.10 recognized as a structural imaging test with higher sensitivity and Overall, CMRI has thus far had very limited study in MM, and reproducibility.13,14 The available evidence shows similar di- further investigation will be required to understand the advan- agnostic yield of CMRI in comparison with echo; however, the tages and limitations of this modality in patients with MM.

Neurology.org/NG Neurology: Genetics | Volume 5, Number 4 | August 2019 5 Figure 2 Prevalence of ECG abnormalities with ES (95% CI) in adult participants

Note that patients are reported according to genetic defect (upper portion of the figure) and according to their clinical syndrome (lower portion of figure), and some patients may be duplicated if both a genetic defect and clinical syndrome was provided. This equally applies to figures 3–5. ES = effect size; mtDNA = mitochondrial DNA.

Pediatric patients may characteristically be considered to have This suggests that pediatric patients with MM should be more a more severe phenotype than adult patients with MM. Based closely monitored for cardiac conduction abnormalities. on this, it was expected that pediatric patients would have However, we cannot exclude the possibility that the prevalence a higher prevalence of abnormalities, which was suggested by of abnormalities for pediatric MM could have been inﬂated by our data comparing ECG abnormalities in children and adults. the incidental inclusion of patients with KSS.

6 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG Figure 3 Prevalence of ECG abnormalities with ES (95% CI) in pediatric participants

ES = effect size.

We observed that most studies of children included patients might be areas for future study, and in the course of this – with nonspeciﬁc MM syndromes.18 20 Only 2 studies repor- review, some of the included articles contained data regarding ted m.8344A>G patients,5,10 in which all patients had no symptoms, sudden cardiac death, and natural history. We abnormalities on ECG, but 1 patient had abnormalities found 8 articles reporting symptoms of cardiac dysfunction in detected by echo.10 Further investigation into pediatric MM patients before diagnosis, including chest pain,4,11,17 various patients, with better described clinical syndromes and mo- degrees of dyspnea,4,10,11,17,22 syncope,11 palpitation10,11 and lecular defects, should be conducted to determine the severity exercise limitations.11 Although 1 article found all patients to of the cardiac phenotype in these patients. In our analysis, be asymptomatic,3 5 articles reported cardiac symptoms oc- Holter monitoring was a very low-yield investigation in adult curring before investigation for MM.9,11,17,21,25,26 One article9 and child patients with MM, and its use as a routine screening reported 7 patients with a history of major adverse cardiac procedure should perhaps be reassessed. events including heart failure and third-degree atrioventricular block, whereas another study25 reported a patient who had In our study, 10 articles described patients with MM with previously underwent cardiac transplantation. Two – – diabetes.9,11,12,15 17,21 24 Of the 122 patients with diabetes, 47 studies21,26 presented patients with cardiomyopathy pre- of them were diagnosed with MELAS. Given that diabetes is ceding their diagnosis of MM. Of interest, 1 study concluded a major risk factor for cardiac dysfunction, it might be that 31% of their patients with MM were previously diagnosed expected that patients with MELAS have greater cardiac risk with cardiac conditions, particularly arrhythmia, impaired left as a consequence. This elevated risk was borne out by the ventricular systolic function, LVH, coronary artery disease, data, and the interaction of diabetes on cardiac dysfunction in and left bundle branch block.17 However, it is unclear in 1 patients with MM is worthy of future study. study whether indicators of cardiac abnormalities were present before or after the diagnosis of MM.22 The goal of this study was to systematically review the prevalence of cardiac abnormalities in MM syndromes; it was not An important limitation encountered in the literature is that a goal of this work to characterize the natural history, symp- few studies presented long-term follow-up data. Only 6 of the toms, sudden-death occurrence, or therapies. However, these 21 included articles reported follow-up data regarding the

Neurology.org/NG Neurology: Genetics | Volume 5, Number 4 | August 2019 7 Figure 4 Prevalence of echocardiography abnormalities with ES (95% CI) in adult participants

ES = effect size; mtDNA = mitochondrial DNA.

8 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG Figure 5 Prevalence of echocardiography abnormalities with ES (95% CI) in pediatric participants

ES = effect size. evolution of cardiac dysfunction over time.3,5,9,11,27,28 Five of with MM showed LVH development, whereas 1 patient with these reported detailed follow-up data3,9,11,27,28 are as follows: MELAS developed systolic impairment.

One study examined the occurrence of major cardiac events in A study reported 6 pediatric m.3243A>G patients, of which 2 a population of 260 patients, retrospectively followed for died at an early stage of disease.28 One patient had died be- a median of 7 years.9 Twenty-seven of the patients died of cause of multiorgan failure, whereas the other had died be- cardiac events during the follow-up period, and the likelihood cause of cardiorespiratory failure. One patient had developed of death was associated with the presence of abnormalities on mild LVH and tricuspid regurgitation without pulmonary cardiac screening tests (42% in patients with 2 or more ab- hypertension in the ﬁrst year of follow-up. One patient had normalities on ECG/echo compared with 10% in the total developed stroke-like episodes in adulthood. The remaining study population). Patients with large-scale single mtDNA patients did not develop clinical symptoms of cardiac dys- deletions and the m.3243A>G mutation had the most severe function over 3 years of follow-up. cardiac phenotype (although this former category may have incidentally included patients with KSS). Last, one study examined the development of cardiac dysfunction in adult and pediatric patients with MELAS (10 of 11 Another publication presented follow-up data on 15 patients harbored the m.3243A>G mutation) during subsequent fol- diagnosed with adult-onset CPEO for a mean of 6.5 years with low-up.27 They observed 11 patients, 4 adults, and 7 children, ECG and echo.3 A total of 5 patients developed new abnor- over an average time span of 6.9 years, whereas 6 patients were malities, but in 4 of these cases, the abnormalities were un- followed up for more than 5 years. Three patients showed related to MM; only one patient developed a mild a worsening ejection fraction, whereas 2 patients died, a child cardiomyopathy attributed to MM. and an adult, one from causes unrelated to MM and the other due to congestive heart failure, respectively. A previous investigation followed 32 patients for an average of 4.1 years, undergoing serial echocardiograms and clinical as- There are other works that did not meet our criteria for in- sessment.11 Two patients with initially normal ECG de- clusion in this systematic analysis that have also provided veloped hypertrophic cardiomyopathy during follow-up. important contributions to our understanding of the natural Serial echocardiography in a patient with CPEO and a patient history of cardiac dysfunction in MM. A previous study of 228

Neurology.org/NG Neurology: Genetics | Volume 5, Number 4 | August 2019 9 Table 2 Summary of findings from systematic review of the evidence

Recommendation

Principal screening tests ECG Echocardiography

Other screening tests requiring further study Cardiac MRI (CMRI)

Recommended screening Available evidence did not allow for a specific recommendation, although we provide 2 general recommendations: In adults, patients with MELAS and MERRF should be screened more frequently than patients with other syndromes In patients with mitochondrial myopathy (MM), ECG screening should be more frequent for children than adults

Common abnormalities in CPEO Bundle branch block (BBB) Left ventricular dysfunction

Common abnormalities in m.3243A>G/MELAS and Left ventricular hypertrophy (LVH) m.8344A>G/MERRF Wolff-Parkinson-White syndrome (WPW)

Areas for future study Natural history studies Studies of treatment outcomes Investigations of new modalities (CMRI and nuclear medicine), in comparison with established modalities (ECG/echo) Investigation into diabetes as a risk factor for cardiac dysfunction associated with MELAS and possibly other mitochondrial diseases Additional studies in pediatric patients

patients with single mtDNA deletion syndromes presented MELAS.32 In contrast, adult-onset CPEO is generally con- data from initial presentation and after an average of 18.7 sidered to have a milder phenotype, and a previous study has years of follow-up.29 Over this time interval, the prevalence of suggested a screening interval as long as 3–5 years.3 The cardiac conduction abnormalities increased from 1.3% to results of our analysis appear to be consistent with these 5.3%, and the prevalence of cardiomyopathy increased from recommendations, but it should be emphasized that the data 0% to 2.6%. This demonstrates a lower prevalence of cardiac are based on a small number of studies, and the prevalence of dysfunction than seen in our systematic analysis, which is abnormalities is high for most syndromes. a surprising finding given the inclusion of numerous KSS patients (which prevented it from being included in our sys- Some cardiac findings that have been associated with MM tematic analysis). In contrast, another study of patients with were rarely or not identified in this systematic analysis. Left MELAS and the m.3243A>G mutation had a mean follow-up ventricular noncompaction has been associated with several of 3.8 years and identified death due to cardiac causes in 4 of mtDNA point mutations,31,33 and we identified a single case 55 patients, which appears to be a more severe phenotype with a mtDNA point mutation at position m.11232.11 Re- than expected for this short duration of follow-up.30 This high strictive or histiocytoid cardiomyopathies have also been as- level of variability between publications emphasizes the need sociated with MM,34 but these were also not identified as part for further study of natural history after extended follow-up. of this analysis. This was likely due to the inclusion criteria for this review, which excluded individual case reports and small Based on the above data, the following observations are case series. possible: (1) cardiac abnormalities are very common across various MM syndromes, (2) analysis of clinical syndromes Limitations of this review relate to the heterogeneity in study indicates that MELAS and MERRF appear to have the most design, ascertainment methods, cardiac investigations, and severe cardiac phenotype in adult patients, and (3) analysis of outcome reporting. The majority of included publications genetic subtypes reveals a more severe cardiac phenotype for were case series focused on cardiologic issues, which may have the m.3243A>G and m.8344A>G mutations compared with biased the reported cases toward a higher prevalence of car- other MM genetic defects. Cardiac screening investigations diac involvement. The majority of the data pertain to patients should be made on a case-by-case basis, but the following with mtDNA-related diseases, and patients with nuclear mi- suggestions based on the evidence may be helpful and are tochondrial disorders were not well represented. Details re- summarized in table 2. ECG and echo are the best-established garding the genetic investigations, which were performed screening tests for MM; the role for CMRI requires further (specific assays, their sensitivity, and other negative genetic investigation. Generally, patients with MELAS and MERRF investigations), were not generally reported. Details regarding are considered to be at a higher risk of cardiac complications, muscle pathology interpretation were not generally reported, and annual screening has been recommended.31 A common but our assumption was that a statement regarding diagnostic and well-described cardiac complication is Wolff-Parkinson- muscle pathology would be reliable in the reported litera- White syndrome (WPW), which is overrepresented in ture. In 2 studies,4,17 data from patients with CPEO were

10 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG aggregated with data from a few patients with Kearns-Sayre also thank Dr Ali Yilmaz and Dr. Karim Wahbi who provided Syndrome (KSS), so it is possible that the prevalence of additional information regarding their published data for this cardiac abnormalities was inflated by the incidental inclusion review. of some patients with KSS to the CPEO group (given that patients with KSS necessarily have cardiac conduction defects Study funding as part of their syndrome). In some of the studies including Asfia Quadir is the recipient of O’Brien Centre Summer patients with the m.3243A>G mutation, data from mutation Student Awards in 2017 and 2018. carriers who were asymptomatic or paucisymptomatic were ff 9,12,15,24 included, which could also have a ected the results. Disclosure This illustrates another limitation of our study that patients Disclosures available: Neurology.org/NG. with the m.3243A>G mutation included aggregate data from patients with MELAS syndrome or other phenotypes, which Publication history does not provide an accurate picture of the cardiac findings in Received by Neurology: Genetics November 2, 2018. Accepted in final phenotypic subgroups with the m.3243A>G mutation. A form April 30, 2019. similar limitation is present for patients with the m.8344A>G mutation, in which only a subset would have had MERRF syndrome. There may also have been some overlap in the groups (for example, CPEO could overlap with single large- scale mtDNA deletions, mtDNA point mutations, and nuclear Appendix Authors gene mutation groups); based on the available information, Name Location Role Contribution the groups were assigned based on how they had been des- ignated in their respective publications. Based on this classi- Asfia Quadir University of First author Data collection, fi ff fi Calgary, data cation, it is also possible that di erences in the de nition of Calgary, interpretation, and each syndrome existed between the included publications, Alberta, drafting and Canada editing of the which is also a limitation of this study. In 1 article, it was manuscript for necessary to assume which abnormalities were from ECG intellectual (BBB and WPW) compared with Holter monitor (ventricular content 10 dysrhythmias). In 1 article studying CMRI, only aggregate Carly Sabine University of Author Data collection, Pontifex, Calgary, data data were presented, but as this was in the normal range, we BSc Calgary, interpretation, and 15 assumed a normal result for all patients. Finally, our analysis Alberta, editing of the Canada manuscript for could not assess the severity of individual reported results, so intellectual even if the prevalence of results was similar between syn- content dromes, we cannot rule out the possibility that some of the Helen Lee University of Author Development of abnormalities may have been more severe in some syndromes Robertson, Calgary, search strategy than others. MLIS Calgary, and editing of the Alberta, manuscript for Canada intellectual This systematic analysis provides evidence to support a more content aggressive cardiac screening for subsets of patients with MM, Christopher Queen Author Statistical analysis, based on a more severe cardiac phenotype for MELAS and Labos, MD, Elizabeth data MSc Health interpretation, and MERRF among clinical syndromes and a more severe phe- Complex, editing of the notype for m.3243A>G and m.8344A>G mutations among Montreal, manuscript for fi Quebec, intellectual genetic defects. We also report a descriptive nding of more Canada content prevalent ECG abnormalities in pediatric MM compared with Gerald University of Corresponding Design and adult MM. The prevalence of abnormalities with Holter Pfeffer, MD, Calgary, author conceptualization monitoring is low and suggests that it may not be necessary PhD Calgary, of the study, data Alberta, collection, data for routine screening. More research is needed to determine Canada interpretation, and how CMRI compares with echo as a modality for structural drafting and cardiac imaging. There are overall limited data on natural editing of the manuscript for history and treatment response, and further longitudinal study intellectual of these conditions will be of benefit to the field. Further study content including pediatric participants would also be of benefit, given the small numbers of studies and participants identified in this review. References 1. Meyers DE, Basha HI, Koenig MK. Mitochondrial cardiomyopathy: pathophysiology, diagnosis, and management. Texas Heart Inst J 2013;40:385–394. Acknowledgment 2. Pfeffer G, Chinnery PF. Diagnosis and treatment of mitochondrial myopathies. Ann Med 2013;45:4–16. The authors specially thank University of Calgary libraries and 3. Pfeffer G, Mezei MM. Cardiac screening investigations in adult-onset progressive Theresa Connolly for their assistance with this review. They external ophthalmoplegia patients. Muscle Nerve 2012;46:593–596.

Neurology.org/NG Neurology: Genetics | Volume 5, Number 4 | August 2019 11 4. Yilmaz A, Gdynia H-J, Ponfick M, et al. Cardiovascular magnetic resonance imaging 20. Baik R, Yu R, Lee YM, Kang HC, Lee JS, Kim HD. Early cardiac evaluation in children (CMR) reveals characteristic pattern of myocardial damage in patients with mito- with non-specific mitochondrial disease with isolated mitochondrial respiratory chain chondrial myopathy. Clin Res Cardiol 2012;101:255–261. complex I defect. J Paediatr Child Health 2012;48:1016–1020. 5. Wahbi K, Larue S, Jardel C, et al. Cardiac involvement is frequent in patients with the 21. Ikawa M, Kawai Y, Arakawa K, et al. Evaluation of respiratory chain failure in mito- m.8344A>G mutation of mitochondrial DNA. Neurology 2010;74:674–677. chondrial cardiomyopathy by assessments of 99mTc-MIBI washout and 123I- 6. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986;7: BMIPP/99mTc-MIBI mismatch. Mitochondrion 2007;7:164–170. 177–188. 22. Akaike M, Kawai H, Yokoi K, et al. Cardiac dysfunction in patients with chronic 7. Rucker G, Schwarzer G, Carpenter J, Olkin I. Why add anything to nothing? The progressive external ophthalmoplegia. Clin Cardiol 1997;20:239–243. arcsine difference as a measure of treatment effect in meta-analysis with zero cells. Stat 23. Ueno H, Shiotani H. Cardiac abnormalities in diabetic patients with mutation in the Med 2009;28:721–738. mitochondrial tRNA(Leu(UUR)) gene. Jpn Circ J 1999;63:877–880. 8. Moher D, Liberati A, Tetzlaff J, Altman DG; PRISMA Group. Preferred reporting 24. Nesbitt V, Pitceathly RDS, Turnbull DM, et al. The UK MRC Mitochondrial Disease items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Patient Cohort Study: clinical phenotypes associated with the m.3243A>G Med 2009;151:264–269. mutation—implications for diagnosis and management. J Neurol Neurosurg Psy- 9. Wahbi K, Bougouin W, Behin A, et al. Long-term cardiac prognosis and risk strati- chiatry 2013;84:936–938. fication in 260 adults presenting with mitochondrial diseases. Eur Heart J 2015;36: 25. Malfatti E, Laforet P, Jardel C, et al. High risk of severe cardiac adverse events in 2886–2893. patients with mitochondrial m.3243A>G mutation. Neurology 2013;80: 10. Catteruccia M, Sauchelli D, Della Marca G, et al. “Myo-cardiomyopathy” is commonly 100–105. associated with the A8344G “MERRF” mutation. J Neurol 2015;262:701–710. 26. Holmgren D, Wahlander H, Eriksson BO, Oldfors A, Holme E, Tulinius M. Car- 11. Limongelli G, Tome-Esteban M, Dejthevaporn C, Rahman S, Hanna MG, Elliott PM. diomyopathy in children with mitochondrial disease; clinical course and cardiological Prevalence and natural history of heart disease in adults with primary mitochondrial findings. Eur Heart J 2003;24:280–288. respiratory chain disease. Eur J Heart Fail 2010;12:114–121. 27. Okajima Y, Tanabe Y, Takayanagi M, Aotsuka H. A follow up study of myocardial 12. Vydt TCG, de Coo RFM, Soliman OII, et al. Cardiac involvement in adults with involvement in patients with mitochondrial encephalomyopathy, lactic acidosis, and m.3243A>G MELAS gene mutation. Am J Cardiol 2007;99:264–269. stroke-like episodes (MELAS). Heart 1998;80:292–295. 13. Grothues F, Smith GC, Moon JC, et al. Comparison of interstudy reproducibility of 28. Wortmann SB, Rodenburg RJ, Backx AP, Schmitt E, Smeitink JA, Morava E. Early cardiovascular magnetic resonance with two-dimensional echocardiography in normal cardiac involvement in children carrying the A3243G mtDNA mutation. Acta Paediatr subjects and in patients with heart failure or left ventricular hypertrophy. Am J Cardiol 2007;96:450–451. 2002;90:29–34. 29. Mancuso M, Orsucci D, Angelini C, et al. Redefining phenotypes associated with 14. Gardner BI, Bingham SE, Allen MR, Blatter DD, Anderson JL. Cardiac magnetic mitochondrial DNA single deletion. J Neurol 2015;262:1301–1309. resonance versus transthoracic echocardiography for the assessment of cardiac vol- 30. Kaufmann P, Engelstad K, Wei Y, et al. Natural history of MELAS associated with umes and regional function after myocardial infarction: an intrasubject comparison mitochondrial DNA m.3243A>G genotype. Neurology 2011;77:1965–1971. using simultaneous intrasubject recordings. Cardiovasc Ultrasound 2009;7:38. 31. Bates MG, Bourke JP, Giordano C, d’Amati G, Turnbull DM, Taylor RW. Cardiac 15. Hollingsworth KG, Gorman GS, Trenell MI, et al. Cardiomyopathy is common in involvement in mitochondrial DNA disease: clinical spectrum, diagnosis, and man- patients with the mitochondrial DNA m.3243A>G mutation and correlates with agement. Eur Heart J 2012;33:3023–3033. mutation load. Neuromuscul Disord 2012;22:592–596. 32. Sproule DM, Kaufmann P, Engelstad K, Starc TJ, Hordof AJ, De Vivo DC. Wolff- 16. Lindroos MM, Parkka JP, Taittonen MT, et al. Myocardial glucose uptake in patients with Parkinson-White syndrome in patients with MELAS. Arch Neurol 2007;64: the m.3243A>G mutation in mitochondrial DNA. J Inherit Metab Dis 2016;39:67–74. 1625–1627. 17. Florian A, Ludwig A, Stubbe-Drager B, et al. Characteristic cardiac phenotypes are 33. Arbustini E, Favalli V, Narula N, Serio A, Grasso M. Left ventricular noncompaction: detected by cardiovascular magnetic resonance in patients with different clinical a distinct genetic cardiomyopathy? J Am Coll Cardiol 2016;68:949–966. phenotypes and genotypes of mitochondrial myopathy. J Cardiovasc Magn Reson 34. El-Hattab AW, Scaglia F. Mitochondrial cardiomyopathies. Front Cardiovasc Med 2015;17:40. 2016;3:25. 18. Cordeiro M, Scaglia F, Lopes Da Silva S, et al. The brain-heart connection in mito- 35. Anan R, Nakagawa M, Miyata M, et al. Cardiac involvement in mitochondrial diseases. chondrial respiratory chain diseases. Neuroradiol J 2009;22:558–563. A study on 17 patients with documented mitochondrial DNA defects. Circulation 19. Baik R, Chae JH, Lee YM, Kang HC, Lee JS, Kim HD. Electrocardiography as an early 1995;91:955–961. cardiac screening test in children with mitochondrial disease. Korean J Pediatr 2010; 36. Galetta F, Franzoni F, Mancuso M, et al. Cardiac involvement in chronic progressive 53:644–647. external ophthalmoplegia. J Neurol Sci 2014;345:189–192.

12 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG ARTICLE OPEN ACCESS Human GABRG2 generalized epilepsy Increased somatosensory and striatothalamic connectivity

Mangor Pedersen, PhD, Magdalena Kowalczyk, MSc, Amir Omidvarnia, PhD, Piero Perucca, MD, PhD, Correspondence Samuel Gooley, MBBS, Steven Petrou, PhD, Ingrid E. Scheffer, MBBS, PhD, Samuel F. Berkovic, MD, and Dr. Pedersen [email protected] Graeme D. Jackson, MD

Neurol Genet 2019;5:e340. doi:10.1212/NXG.0000000000000340 Abstract Objective To map functional MRI (fMRI) connectivity within and between the somatosensory cortex, putamen, and ventral thalamus in individuals from a family with a GABAergic deﬁcit segregating with febrile seizures and genetic generalized epilepsy.

Methods We studied 5 adults from a family with early-onset absence epilepsy and/or febrile seizures and GABRG2[R43Q] a GABAA receptor subunit gamma2 pathogenic variant ( ) vs 5 age-matched controls. We infer diﬀerences between participants with the GABRG2 pathogenic variant and controls in resting-state fMRI connectivity within and between the somatosensory cortex, putamen, and ventral thalamus.

Results We observed increased fMRI connectivity within the somatosensory cortex and between the putamen and ventral thalamus in all individuals with the GABRG2 pathogenic variant compared with controls. Post hoc analysis showed less pronounced changes in fMRI connectivity within and between the primary visual cortex and precuneus.

Conclusions Although our sample size was small, this preliminary study suggests that individuals with a GABRG2 pathogenic variant, raising risk of febrile seizures and generalized epilepsy, display underlying increased functional connectivity both within the somatosensory cortex and in striatothalamic networks. This human network model aligns with rodent research and should be further validated in larger cohorts, including other individuals with generalized epilepsy with and without known GABA pathogenic variants.

From the The Florey Institute of Neuroscience and Mental Health (M.P., M.K., A.O., S.P., I.E.S., G.D.J.), Parkville; Department of Neurology (I.E.S.), Royal Children’s Hospital, Parkville; Department of Neuroscience (P.P.), Central Clinical School, Monash University; Department of Neurology (P.P.), The Royal Melbourne Hospital, Parkville; Department of Neurology (P.P.), Alfred Health, Melbourne; Department of Medicine (P.P., S.P.), The Royal Melbourne Hospital, The University of Melbourne, Parkville; Epilepsy Research Centre (S.G., I.E.S., S.F.B., G.D.J.), Department of Medicine, The University of Melbourne, Austin Health, Heidelberg; and Department of Pediatrics (I.E.S.), The University of Melbourne, Parkville, VIC, Australia.

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.

We previously reported an Australian family with a GABA Methods type A receptor subunit gamma2 pathogenic variant (GABRG2[R43Q]) presenting with febrile seizures, febrile Participants and clinical information seizures plus, and absence epilepsy.1 This family has decreased We recruited 5 adults from a previously reported family with 1 cortical inhibition, demonstrated by transcranial magnetic the GABRG2 [R43Q] pathogenic variant (mean age [SD]: stimulation,2 and decreased cortical benzodiazepine receptor 36.4 ± 4.2 years; clinical information in the table). They were binding on PET.3 Aligning with human data, mice with the compared with 5 age- and sex-matched controls (mean age same Gabrg2[R43Q] pathogenic variant have reduced cortical [SD] 36.8 ± 4.1 years). inhibition shown by deficits in GABA-mediated synaptic Standard protocol, approvals, and currents in the somatosensory cortex and anatomic changes in patient consents cortical interneuron positioning.4,5 This study was approved by the Austin Human Research Ethics Committee, and participants gave written informed Although these investigations provide clues about the mecha- consent to participate. nisms underlying the GABRG2 pathogenic variant, it is un- ff known whether brain connectivity is a ected in humans with fMRI preprocessing fi the R43Q pathogenic variant. To study this issue, we quanti ed We acquired 10 minutes of resting-state fMRI data on a Sie- changes in functional MRI (fMRI) connectivity in individuals mens 3T Skyra scanner with a voxel size of 3 × 3 × 3 mm and GABRG2 with this pathogenic variant. We focused on regions repetition time of 3 seconds. The fMRI data were slice-timing Gabrg2 previously studied in mice: the somatosensory cor- corrected, realigned, coregistered to T1 images, tissue seg- tex, the ventral thalamus, and the putamen (as a recent fMRI mented, spatially normalized, and filtered between 0.01 and study in rats demonstrated that GABAA antagonists enhance 0.08 Hz. See reference 8 for further details about our fMRI 6 connectivity of the somatosensory cortex and striatum ). methods. Together, these 3 brain regions comprise a well-known pathway where the cortex sends excitatory signals to the fMRI connectivity analysis putamen, which in turn exerts strong inhibitory control over We used 2 different estimates of fMRI connectivity. (1) Regional the ventral thalamus via the globus pallidus and substantia homogeneity: to calculate voxel-averaged local connectivity nigra pars reticulata before relaying excitatory information within each node of our brain network model (somatosensory back to the cortex. cortex, ventral thalamus, and putamen).ThismeasuresKendall W correlations between a single voxel and its 26 nearby voxels in Based on previous research in humans with the GABRG2 three-dimensional space. Its values range between 0 (minimal [R43Q] pathogenic variant2,3 and mice with the Gabrg2[R43Q] local connectivity) and 1 (maximal local connectivity). (2) Par- pathogenic variant,4,5,7 we hypothesize that fMRI connectivity tial correlation: to calculate voxel-averaged Pearson correlations within the somatosensory cortex and between subcortical of fMRI time series between each node pair, while regressing out regions is altered in people with a GABRG2 pathogenic variant. indirect correlation effects from all other connection pairs. Partial

Table Clinical, EEG and imaging phenotype of individuals with the R43Q GABRG2 pathogenic variant1

Patient Seizure GABRG2 pathogenic no. Sex Age (y) Onset Offset type EEG Syndrome MRI AEDs variant

1 F 41 Infant Early FS, AS GSW CAE Normal 0 Positive teens

2 F 33 Infant Ongoing FS, AS GSW CAE Normal 0 Positive

3 M 31 Infant Toddler FS Normal FS only Normal 0 Positive

4 M 39 Infant 36 y FS, AS, GTCS, GSW, L- EOAE (childhood), L-TLE Normal 1 Positive FIAS TIRDA (adulthood) (CBZ)

5 M 38 Infant Toddler FS Normal FS only Normal 0 Positive

Abbreviations: AED = current antiepileptic drug; AS = absence seizure; CAE = childhood absence epilepsy; CBZ = carbamazepine; EOAE = early-onset absence epilepsy; FIAS = focal impaired awareness seizure; FS = febrile seizure; GTCS = generalized tonic-clonic seizure; GSW = generalized spike-wave discharge; L-TLE = left temporal lobe epilepsy; L-TIRDA = left temporal intermittent rhythmic delta activity.

2 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG correlation values range between −1 (maximal negative corre- variant compared with controls (Hedges’ g = 1.46; 95% CI = lation) and 1 (maximal positive correlation). 0.43–2.41—figure, A). Hedges’ g effect sizes were lower between GABRG2 and healthy controls for fMRI connec- Effect size analysis tivity within the putamen (Hedges’ g = 0.55; CI 95 = ’ g ff We used Hedges standardized e ect sizes and 95th per- −0.85–1.84) and ventral thalamus (Hedges’ g = 0.40; CI 95 = fi centile con dence intervals (95% CIs) to quantify mean dif- −0.96 to 1.70). ferences in brain connectivity between individuals with the GABRG2 pathogenic variant and controls (as Hedges’ g is recommended over Cohen’s d in studies with a limited sample GABRG2: fMRI connectivity is increased sizes). Hedges’ g effect sizes = 0.2 (small); 0.5 (moderate); 0.8 between the thalamus and putamen (large); and 1.2 (very large). We observed stronger fMRI connectivity between the putamen and ventral thalamus in all individuals with the GABRG2 Data availability pathogenic variant compared with controls (Hedges’ g = 1.98; Anonymized original data will be shared by request from any – —fi fi 95% CI = 0.11 3.45 gure, F). It is worth noting that the quali ed investigator. GABRG2 participant with strongest connectivity also experienced seizures into adulthood (see participant 2 in the table Results for more clinical information). There were no large effect sizes between GABRG2 and controls for fMRI connectivity be- GABRG2: fMRI connectivity is increased within tween the putamen and somatosensory cortex (Hedges’ g = the somatosensory cortex 0.68; 95% CI = −0.75–1.98) or the ventral thalamus and We observed stronger fMRI connectivity within the somato- somatosensory cortex (Hedges’ g = 0.26; 95% CI = −1.09 sensory cortex in all individuals with the GABRG2 pathogenic to 1.55).

Figure Functional connectivity within and between brain nodes in GABRG2 participants displayed by purple magenta color bars and healthy controls displayed by blue cyan color bars

Peak regional homogeneity across all voxels within the somatosensory cortex (A), putamen (B), and ventral thalamus (C), and partial correlations averaged across all voxels between the somatosensory cortex-putamen (D), somatosensory cortex-ventral thalamus (E), and ventral thalamus-putamen (F). Error bars denote the standard error across participants. There was little difference in head movement between groups (mean movement [SD] in GABRG2 =0.12mm± 0.05; mean movement [SD] in controls = 0.16 mm ± 0.04). Our 3 brain regions of interest—(1) bilateral somatosensory cortex (yellow); (2) putamen (orange); and (3) ventral thalamus (red) (G)—were delineated using the NeuroSynth database (neurosynth.org) incorporating findings from a multitude of previous fMRI studies describing these brain regions and their behavioral correlates, as well as their functional connectivity patterns. *A and F hadlarge effect size.

Neurology.org/NG Neurology: Genetics | Volume 5, Number 4 | August 2019 3 GABRG2: post hoc analysis shows less fMRI disorder to the demands of an imaging study that requires connectivity changes of the visual cortex travel and attendance at a single site. Despite this limitation, and precuneus our preliminary findings give insight into network changes To test whether our primary result was specific to the so- that may underlie human genetic epilepsy caused by matosensory cortex, thalamus, and putamen or it reflected a GABRG2 pathogenic variant, and they align with results a “whole-brain process”, we analyzed 2 additional brain nodes: from the Gabrg2 animal model, which also shows increased (1) primary visual cortex and (2) precuneus. We observed no activity of the somatosensory cortex.4 Nevertheless, our large effect size in fMRI connectivity for these 2 brain regions human network model based on this family with a GABRG2 between GABRG2 and control participants. fMRI connec- variant should be further validated in larger generalized ep- tivity within the visual cortex: Hedges’ g = 0.45; 95% CI = ilepsy cohorts with and without known GABA pathogenic −0.93 to 1.74. fMRI connectivity within the precuneus: variants. Hedges’ g = 0.25; 95% CI = −1.12 to 1.56. fMRI connectivity between the visual cortex and precuneus: Hedges’ g = 0.43; Acknowledgment 95% CI = −0.95–1.72. The authors thank Susannah Bellows for recruiting the participants with the GABRG2 pathogenic variant. Discussion Study funding The Florey Institute of Neuroscience and Mental Health We found that 5 individuals with a GABRG2 pathogenic acknowledges the strong support from the Victorian variant and a history of seizures (5/5 febrile seizures; 3/5 Government and in particular the funding from the Op- absence seizures) have increased fMRI connectivity within the erational Infrastructure Support Grant. They also ac- somatosensory cortex and between the putamen and ventral knowledge the facilities and the scientific and technical thalamus compared with 5 healthy controls. This finding assistance of the National Imaging Facility (NIF) at the suggests a “network-specific” rather than a “whole-brain” ef- Florey node and The Victorian Biomedical Imaging Ca- fect, as our post hoc analysis revealed no difference in con- pability (VBIC). This study was supported by the National nectivity between GABRG2 and control participants of either Health and Medical Research Council (NHMRC) of the primary visual cortex or the precuneus. Australia, grant numbers 628952 and 1060312 (GJ Prac- titioner Fellowship) and program grant number 1091593 Emerging animal research has highlighted the important role (S. Petrou, I. E. Scheffer,S.F.Berkovic,andG.D. of specific brain areas in generalized epilepsy, in particular Jackson). absence epilepsy where seizures are thought to originate in 9 the somatosensory cortex and are modulated by thala- Disclosure 10 mostriatal circuits. Although it is not straightforward to Disclosures available: Neurology.org/NG. compare animal and human studies, our findings provide preliminary evidence that somatosensory cortex and sub- Publication history cortical structures are hyperconnected in people with a ge- Received by Neurology: Genetics January 28, 2019. Accepted in final form netic predisposition to develop febrile and also absence April 29, 2019. seizures. This finding is further supported by our previous GABRG2 transcranial magnetic stimulation study—in the same family—showing neuronal hyperexcitability of the 2 perimotor cortex. Appendix Authors

It is nontrivial to quantify whether microscale neuronal dys- Name Location Role Contribution function and macroscale fMRI connectivity are related be- Mangor The Florey Author Conception and design cause of their vast difference in spatial scales. However, this Pedersen, Institute of of the study, acquisition PhD Neuroscience and and analysis of data, and study presented us with an opportunity to (indirectly) assess Mental Health writing the first draft of whether inhibitory neuronal dysfunction is reflected in fMRI the manuscript GABRG2 connectivity, as we know that people with a path- Magdalena The Florey Author Conception and design of ogenic variant have abnormal inhibitory GABAergic neuronal Kowalczyk, Institute of the study and acquisition fi MSc Neuroscience and and analysis of data function. We postulate that our fMRI connectivity ndings Mental Health are related to inhibitory neuronal abnormalities, especially Amir The Florey Author Conception and design hyperconnectivity between the thalamus and the putamen, as Omidvarnia, Institute of of the study and analysis the latter brain region consists almost exclusively of medium PhD Neuroscience and of data spiny inhibitory GABAergic neurons. Mental Health Piero The University of Author Conception and design of Our small sample size is a consequence of the difficulty of Perucca, MD, Melbourne the study and acquisition PhD and analysis of data recruiting a single family with a genetically homogenous

4 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG References Appendix (continued) 1. Marini C, Harkin LA, Wallace RH, Mulley JC, Scheﬀer IE, Berkovic SF. Childhood absence epilepsy and febrile seizures: a family with a GABAA receptor mutation. Brain – Name Location Role Contribution 2003;126:230 240. 2. Fedi M, Berkovic SF, Macdonell RAL, Curatolo JM, Marini C, Reutens DC. Intra- Samuel The University of Author Conception and design of cortical hyperexcitability in humans with a GABAA receptor mutation. Cereb Cortex – Gooley, Melbourne the study and acquisition 2008;18:664 669. MBBS and analysis of data 3. Fedi M, Berkovic SF, Marini C, Mulligan R, Tochon-Danguy H, Reutens DC. A GABAA receptor mutation causing generalized epilepsy reduces benzodiazepine re- Steven The Florey Author Conception and design ceptor binding. Neuroimage 2006;32:995–1000. Petrou, PhD Institute of of the study and 4. Tan HO, Reid CA, Single FN, et al. Reduced cortical inhibition in a mouse model of Neuroscience and codesigned and drafted familial childhood absence epilepsy. Proc Natl Acad Sci U S A 2007;104: Mental Health a significant portion of 17536–17541. the manuscript 5. Wimmer VC, Li MYS, Berkovic SF, Petrou S. Cortical microarchitecture changes in genetic epilepsy. Neurology 2015;84:1308–1316. Ingrid E. The University of Author Conception and design 6. Nasrallah FA, Singh KKDR, Yeow LY, Chuang KH. GABAergic eﬀect on resting-state Scheffer, Melbourne of the study and functional connectivity: dynamics under pharmacological antagonism. Neuroimage MBBS, PhD codesigned and drafted 2017;149:53–62. a significant portion of 7. Currie SP, Luz LL, Booker SA, Wyllie DJA, Kind PC, Daw MI. Reduced local input to the manuscript fast-spiking interneurons in the somatosensory cortex in the GABAA γ2 R43Q mouse – Samuel F. The University of Author Conception and design model of absence epilepsy. Epilepsia 2017;58:597 607. Berkovic, Melbourne of the study and 8. Pedersen M, Curwood EK, Vaughan DN, Omidvarnia AH, Jackson GD. Abnormal MD codesigned and drafted brain areas common to the focal epilepsies: multivariate pattern analysis of fMRI. a significant portion of Brain Connect 2016;6:208–215. the manuscript 9. Meeren HK, Pijn JP, Van Luijtelaar EL, Coenen AM, Lopes da Silva FH. Cortical focus drives widespread corticothalamic networks during spontaneous absence seiz- Graeme D. The Florey Author Conception and design ures in rats. J Neurosci 2002;22:1480–1495. Jackson, MD Institute of of the study and drafted 10. Slaght SJ, Paz T, Chavez M, Deniau JM, Mahon S, Charpier S. On the activity of the Neuroscience and a significant portion of corticostriatal networks during spike-and-wave discharges in a genetic model of ab- Mental Health the manuscript sence epilepsy. J Neurosci 2004;24:6816–6825.

Neurology.org/NG Neurology: Genetics | Volume 5, Number 4 | August 2019 5 ARTICLE OPEN ACCESS Congenital myopathies in the adult neuromuscular clinic Diagnostic challenges and pitfalls

Stefan Nicolau, MD, Teerin Liewluck, MD, Jennifer A. Tracy, MD, Ruple S. Laughlin, MD, and Correspondence Margherita Milone, MD, PhD Dr. Milone [email protected] Neurol Genet 2019;5:e341. doi:10.1212/NXG.0000000000000341 Abstract Objective To investigate the spectrum of undiagnosed congenital myopathies (CMs) in adults presenting to our neuromuscular clinic and to identify the pitfalls responsible for diagnostic delays.

Methods We conducted a retrospective review of patients diagnosed with CM in adulthood in our neuromuscular clinic between 2008 and 2018. Patients with an established diagnosis of CM before age 18 years were excluded.

Results We identiﬁed 26 patients with adult-onset CM and 18 patients with pediatric-onset CM who were only diagnosed in adulthood. Among patients with adult onset, the median age at onset was 47 years, and the causative genes were RYR1 (11 families), MYH7 (3 families) and ACTA1 (2 families), and SELENON, MYH2, DNM2, and CACNA1S (1 family each). Of 33 patients who underwent muscle biopsy, only 18 demonstrated histologic abnormalities characteristic of CM. Before their diagnosis of CM, 23 patients had received other diagnoses, most commonly non-neurologic disorders. The main causes of diagnostic delays were mildness of the symptoms delaying neurologic evaluation and attribution of the symptoms to coexisting comorbidities, particularly among pediatric-onset patients.

Conclusions CMs in adulthood represent a diagnostic challenge, as they may lack the clinical and myopathologic features classically associated with CM. Our ﬁndings underscore the need for a revision of the terminology and current classiﬁcation of these disorders.

From the Department of Neurology, Mayo Clinic, Rochester, MN.

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

Copyright © 2019 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 Glossary CFTD = congenital ﬁber-type disproportion; CK = creatine kinase; CM = congenital myopathy; MRC = Medical Research Council; MUP = motor unit potential; NGS = next-generation sequencing; VUS = variants of unknown signiﬁcance.

Congenital myopathies (CMs) are a group of inherited clinicopathologic or genetic diagnosis. Because of the over- myopathies most commonly presenting in infancy with hy- lapping phenotypes of CM and other inherited neuromus- potonia and weakness.1,2 Some patients may also have facial cular disorders, patients without a molecular diagnosis or weakness, ptosis, ophthalmoparesis, respiratory insufficiency, distinctive histologic findings were not included. Patients with cardiac involvement, or skeletal abnormalities. Generally, the sporadic late-onset nemaline myopathy were excluded, as this – weakness is either static or very slowly progressive.3 6 CMs is an acquired disorder.20 Patients diagnosed with congenital – are rare disorders and, as a group, have an estimated preva- muscular dystrophies were also excluded.21 23 All patients lence of 0.6–2 per 100,000 individuals.7,8 Traditionally, CMs with a diagnosis of CM established on clinical, histologic, or have been classified on the basis of distinctive myopathologic genetic grounds before age 18 years were excluded. features, such as nemaline rods, cores, multiminicores, central nuclei, or congenital fiber-type disproportion (CFTD).9 Each Clinical and laboratory evaluation histologic subtype is genetically heterogeneous, and over Medical records were retrospectively reviewed, and data were 25 CM genes have been identified.10 In addition, a spectrum collected regarding patients’ age at onset, clinical features, and of myopathologic changes may stem from the same molecular disease course. The severity of weakness was graded accord- defect, even within the same family.11 ing to the Medical Research Council (MRC) scale and classified as absent (MRC 5/5), mild (4/5), moderate (3/5), or Increasing use of genetic testing and application of next- severe (0–2/5). Results of genetic testing, electrodiagnostic generation sequencing (NGS) has identified patients with studies, creatine kinase (CK) levels, complete blood count, variants in genes known to cause CM but who lack the his- and cardiac and pulmonary function tests were reviewed. tologic and clinical features typically associated with CM. This phenomenon is well recognized in RYR1-related myopathies, Muscle pathology Muscle biopsies performed at our institution were processed in which approximately 40% of patients present with core 24 myopathy,12 whereas others present with malignant hyper- as previously described, and biopsies performed at outside thermia or recurrent rhabdomyolysis without histopathologic institutions were reviewed. abnormalities.12,13 Moreover, CM can manifest in adulthood, Statistical methods as reported with CM caused by variants in RYR1,12 DNM2,14 Patients were divided into subgroups according to age at ACTA1,15,16 BIN1,17 MYH7,18 and KBTBD13.19 Despite onset and molecular diagnosis. Patient characteristics are these reports, the prevalence of adult-onset CM remains un- presented as counts, medians, and ranges. The number of known, and their characteristics have not been systematically patients with missing data is indicated. studied. We therefore sought to investigate the clinical, histopathologic, and genetic features of adult patients with un- Standard protocol approvals, registrations, diagnosed CM presenting to our neuromuscular clinic to (1) and patient consents characterize the spectrum of adult-onset CM and (2) identify The Mayo Clinic institutional review board approved this study. pitfalls leading to diagnostic delays among both pediatric- and adult-onset patients. Data availability Anonymized data will be shared by request from any qualified Methods investigator. Patient selection Results We searched the medical records of the Mayo Clinic in Rochester, MN, for patients evaluated in the neurology de- Patients and genetic testing partment between July 1, 2008, and June 30, 2018, and who Forty-four patients from 36 unrelated families met the in- received a diagnosis of CM in adulthood (age ≥18 years). A clusion criteria. A molecular diagnosis was established in 37 diagnosis of CM was established either on the basis of genetic patients from 29 families (figure). The most common caus- testing or muscle histopathology. Patients found to have ative genes were RYR1 (13 families), ACTA1 (5 families), a disease-causing variant in a gene known to cause CM10 were SELENON (4 families), MYH7 (3 families), MYH2, TPM2, included. Patients without a molecular diagnosis were in- DNM2, and CACNA1S (1 family each). The mode of included only if their muscle biopsy demonstrated distinctive heritance of the disorders was autosomal recessive in the histopathologic features of CM (centronuclear myopathy, patients with MYH2 and SELENON variants and 1 patient nemaline rods, central cores, combined cores and rods, mul- with ACTA1 myopathy. Inheritance was autosomal dominant timinicores, or CFTD) in the absence of an alternative in families with ACTA1, TPM2, MYH7, DNM2,andCACNA1S

2 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG another with centronuclear myopathy only underwent Sanger Figure Genes causative of congenital myopathies in our sequencing of the MTM1 and DNM2 genes. cohort A history of more than 1 affected family member was present in 22 of 36 families. In 7 families, the discovery of a molecular diagnosis in the proband led to confirmation of the diagnosis in additional family members. These were all first-degree relatives of the probands. All individuals thus identified underwent neurologic examination, with the exception of 1 affected family member who was not available for a face-to-face assessment.

Seventeen additional patients were found to harbor variants of unknown signiﬁcance (VUS) in CM genes, as classiﬁed according to the American College of Medical Genetics criteria.28 Fourteen of these 17 patients had variants in RYR1, whereas single patients had variants in MYH7, SELENON and CCDC78. There was no clinical or histologic evidence to support the possible pathogenicity of these VUS. These patients were therefore not included.

Pediatric-onset patients with delayed diagnosis Eighteen patients had onset of symptoms before age 18 years but received a diagnosis of CM in adulthood (table 1). The causative genes were ACTA1 (5 patients), RYR1 (3), SELE- NON (3), MYH7 and TPM2 (1 each). Onset of symptoms varied from infancy to adolescence, whereas the median age at diagnosis was 40 years.

Various factors contributed to the diagnostic delay in these Percentages indicate the proportion of families with the corresponding patients. Ten patients did not undergo relevant investigations genetic defect among (A) adult-onset and (B) pediatric-onset patients di- before adulthood because of the mildness of their symptoms. agnosed in adulthood. Three patients had no muscle biopsy and were diagnosed with Becker muscular dystrophy, myasthenia gravis, and spinal muscular atrophy.11 Four patients underwent muscle biopsy, variants and at least 12 families with RYR1 variants. The domi- which showed nonspecific abnormalities, and were diagnosed nant inheritance of the RYR1 variants was established by pre- with muscular dystrophy. vious reports (10 families) or occurrence of the variant in affected parent-child pairs (2 families). One patient had Adult-onset patients a single novel RYR1 frameshift variant. Although we could not Twenty-six patients had onset of symptoms in adulthood ascertain its dominant or recessive inheritance, this was the (table 2). The most common causative gene was RYR1 (13 only candidate variant identified in this patient, and there have patients), followed by ACTA1 (3), MYH7 (3), DNM2 (2), been previous reports of dominant inheritance with truncat- and SELENON, MYH2, and CACNA1S (1 each). The median ing variants in RYR1.25,26 In 2 families with dominant in- age at onset was 47 years. Eight patients retrospectively heritance, however, additional variants may have contributed reported athletic performance below their peers since child- to the clinical phenotype, as previously reported (table e-1, hood, which had not previously been thought to be abnormal. links.lww.com/NXG/A158).25,27 Sixteen patients described a progressive course, whereas 4 indicated a static course. Three patients reported only episodic The molecular diagnosis was first established by single gene symptoms, such as rhabdomyolysis or malignant hyperthermia, sequencing in 8 families, by an NGS gene panel in 18 families, and 3 patients presented with asymptomatic hyperCKemia. and by whole-exome sequencing in 3 families. In total, 29 distinct genetic variants were identified (table e-1, links.lww. Twenty of the 25 examined patients had muscle weakness, com/NXG/A158). In 7 of the 44 patients, no molecular di- which was mild (9 patients), moderate (7), or severe (4). agnosis could be established, and the diagnosis of CM was When present, weakness most commonly affected limb-girdle therefore made on histopathologic grounds. Five of these muscles. Two patients started to use a wheelchair (with patients had no genetic testing; 1 patient with CFTD un- ACTA1 and RYR1 myopathy). Two patients (1 with ACTA1 derwent an NGS panel targeting 22 known CM genes, and myopathy and 1 with RYR1 myopathy) had mild dysphagia.

Neurology.org/NG Neurology: Genetics | Volume 5, Number 4 | August 2019 3 Table 1 Clinical and laboratory features of pediatric-onset congenital myopathies diagnosed in adulthood

MYH7 ACTA1 TPM2 SELENON RYR1 None Total

No. of patients (families) 1 (1) 5 (4) 1 (1) 3 (3) 3 (3) 5 (5) 18 (17)

Age at onset—median (range) 10 2 (0–15) 17 1 (1–5) 14 (10–15) 0 (0–12) 5 (0–17)

Age at diagnosis—median (range) 67 52 (24–69) 56 34 (24–42) 34 (32–57) 37 (19–46) 40 (19–69)

Clinical course

Static 11 2 4

Progressive 14 3 1 3 12

Fluctuating/episodic 22

Degree of weakness

None 11

Mild 1225

Moderate 33 2 8

Severe 12 1 4

Pattern of weakness

Limb girdle 1121 2 7

Distal 12 1 1 5

Generalized 21 2 5

Other clinical features

Facial weakness 05 1 3 0 4 13

Ptosis 00 0 0 0 1 1

Ophthalmoparesis 00 0 0 0 0 0

Skeletal abnormalities 13 1 2 1 4 12

Cardiomyopathy 0/1 1/4 0/1 0/3 1/2 1/4 3/15

Respiratory involvement 2/4 1/1 3/3 1/1 5/5 12/14

Creatine kinase

Elevated 0/1 1/4 0/1 0/3 2/3 0/2 3/14

Median (range) N N (N-222) N N 412 (N −572) N N (N −572)

EMG

Short-duration MUPs 1/1 5/5 1/1 3/3 3/3 3/4 16/17

Long-duration MUPs 1/1 3/5 1/1 0/3 2/3 0/4 7/17

Fibrillation potentials 1/1 4/5 0/1 2/3 2/3 1/4 10/17

Muscle pathology

Distinctive abnormalities (n) CFTD (1) CFTD (1) MMC (1) Central cores (1) MMC (2), CNM (1), CFTD (1), 9 central nuclei and cores (1)

Nonspecific abnormalities 13 1 5

Normal 11

Abbreviations: CFTD = congenital fiber-type disproportion; CNM = centronuclear myopathy; MMC = multiminicore; MUP = motor unit potential; N = normal. Denominators denote the number of patients for whom data are available.

4 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG Table 2 Clinical and laboratory features of patients with adult-onset congenital myopathy

MYH2 MYH7 ACTA1 DNM2 SELENON RYR1 CACNA1S None Total

No. of patients 1 (1) 3 (3) 3 (2) 2 (1) 1 (1) 13 (11) 1 (1) 2 (2) 26 (22) (families)

Age at onset—median 26 47 34 48 32 53 (25–72) 36 49 (40–59) 47 (range) (29–49) (31–65) (25–72)

Age at 33 48 34 61 34 55 (25–84) 36 57 (55–60) 52 diagnosis—median (31–58) (51–72) (25–84) (range)

Clinical course

Asymptomatic 33

Static 11 2 4

Progressive 32216 2 16

Fluctuating/episodic 21 3

Degree of weakness

None 41 5

Mild 111 6 9

Moderate 12111 17

Severe 2114

Pattern of weakness

Limb girdle 11 1 2 1 6 1 13

Distal 11 2 4

Scapuloperoneal 11 2

Generalized 11

Other clinical features

Facial weakness 11 2 1 1 3 0 1 10

Ptosis 10 0 0 0 0 0 0 1

Ophthalmoparesis 10 0 0 0 0 0 0 1

Skeletal 11 1 1 1 2 0 1 8 abnormalities

Cardiomyopathy 0/1 2/3 1/3 1/2 0/1 3/7 0/1 0/2 7/20

Respiratory 0/1 1/2 1/2 2/2 1/1 2/2 0/2 7/12 involvement

Creatine kinase

Elevated 1/1 2/3 0/3 1/2 0/1 12/13 1/1 1/2 18/26

Median (range) 457 974 N 137 N 703 (N–2800) 895 214 (N–397) 536 (N–6685) (N–177) (N–6685)

EMG

Short-duration MUPs 1/1 2/3 3/3 1/2 1/1 9/10 0/1 2/2 19/23

Long-duration MUPs 1/1 0/3 3/3 0/2 0/1 2/10 0/1 0/2 6/23

Fibrillation 0/1 3/3 1/3 1/2 0/1 5/10 0/1 0/2 10/23 potentials

Continued

Neurology.org/NG Neurology: Genetics | Volume 5, Number 4 | August 2019 5 Table 2 Clinical and laboratory features of patients with adult-onset congenital myopathy (continued)

MYH2 MYH7 ACTA1 DNM2 SELENON RYR1 CACNA1S None Total

Muscle pathology

Distinctive NR (2) CNM (1) Central core (2), Core-like CNM (1), 9 abnormalities (n) CNM (1) structures (1) MMC (1)

Nonspecific 1a 2159 abnormalities

Abbreviations: CFTD = congenital fiber-type disproportion; CNM = centronuclear myopathy; MMC = multiminicore; MUP = motor unit potential; N = normal; NR = nemaline rod. Denominators denote the number of patients for whom data are available. a Aggregates of several sarcomeric proteins.

Rhabdomyolysis occurred in 4 patients (3 with RYR1 variants showed short-duration motor unit potentials (MUPs) in 35 and 1 with a CACNA1S variant), whereas a history of malig- patients. Thirteen patients additionally had long-duration nant hyperthermia was elicited in 3 patients, all of whom had MUPs; these were most commonly found in ACTA1 myop- RYR1 variants. athy, being present in 6/8 patients. Long-duration MUPs alone were not observed. Fibrillation potentials were seen in Skeletal abnormalities were present in 8 patients, including 20 patients. A sensorimotor polyneuropathy was found in 4 joint contractures (3 patients), pes cavus (3), lumbar hyper- patients, 2 of whom had evidence of an acquired cause. Re- lordosis (3), scoliosis (2), high-arched palate (1), and rigid petitive nerve stimulation did not show a significant (>10%) spine (1). By contrast, skeletal abnormalities were seen in 12/ decrement in any of the 15 patients tested, including 2 18 pediatric-onset patients. patients with centronuclear myopathy. All 6 patients who lacked weakness on examination did have elevated CK levels, At least 13 adult-onset patients received one or more incorrect short-duration MUPs on EMG, or both. diagnosis before the diagnosis of CM. The most common misdiagnoses were non-neurologic disorders (7 patients), Muscle pathology followed by statin-induced myopathy (3), neuropathies (2), Thirty-three patients underwent muscle biopsy. Among 26 of mitochondrial myopathy, and limb-girdle muscular dystrophy these patients in whom a molecular diagnosis was established, (1 each). Two patients received immunosuppressive treat- only 11 had the distinctive histopathologic findings of a CM, ment as a result of misdiagnoses. including central core disease (4 patients), nemaline rods (2), CFTD (2), centronuclear myopathy (2), and multiminicore Cardiac and respiratory involvement disease (1). By contrast, 14 biopsies demonstrated only Cardiomyopathy was present in 10 of 35 patients who had nonspecific myopathic findings, such as increased variation in undergone an echocardiogram, excluding patients with cor fiber size, increased numbers of internal nuclei, necrotic and pulmonale secondary to respiratory insufficiency. Cardiomy- regenerating fibers, and increased endomysial connective opathy was seen in patients with variants in MYH7, ACTA1, tissue. One biopsy was normal. Distinctive histologic findings DNM2, and RYR1. Respiratory muscle involvement was were seen in similar proportions of pediatric- and adult-onset present in 19/26 patients who had undergone overnight patients (9/15 and 9/18, respectively). oximetry or pulmonary function testing, including maximal respiratory pressures. Seven patients required noninvasive Among the 7 patients without a molecular diagnosis, muscle ventilatory support and 1 required mechanical ventilation. biopsy demonstrated multiminicore disease (3 patients), Ventilatory support was more frequently needed in pediatric- centronuclear myopathy (2), CFTD (1), and both central onset than adult-onset patients (6 and 2, respectively). Re- nuclei and cores (1). A molecular diagnosis was established in spiratory involvement was most common and most severe in 11/13 patients (85%) with specific histopathologic findings patients with SELENON myopathy. In 2/4 SELENON my- who had undergone genetic testing. opathy patients, respiratory failure was the first presenting symptom and all 4 eventually required ventilatory support. Discussion There was no association between severity of appendicular weakness and either cardiomyopathy or respiratory involvement. The present study examined patients in whom a diagnosis of CM was established in adulthood. We aimed to characterize Laboratory and electrodiagnostic features the spectrum of phenotypic variability and unravel pitfalls The CK level was elevated in 3/14 pediatric-onset patients leading to the diagnostic delays experienced by these patients. and 18/26 adult-onset patients in whom data were available. We therefore selected patients in whom a diagnosis of CM CK elevations were most common in patients with RYR1 was established after age 18 years on clinical, pathologic, and myopathy. Forty patients underwent EMG. Needle EMG genetic grounds.

6 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG Eighteen of the patients thus identified had onset of symp- cardiac disorders cannot be excluded in the remainder. Re- toms before age 18 years but were only diagnosed at a median spiratory muscle weakness and sleep disordered breathing age of 40 years. Among these patients, the most frequent were also common among our patients with CM, affecting cause of diagnostic delays was a failure to obtain any neuro- 73% of those tested. Two patients, both with SELENON logic investigations for symptoms that were considered be- myopathy, presented with symptoms of respiratory muscle nign or attributed to non-neurologic conditions. In other weakness, which preceded the onset of limb weakness and led patients, it was not recognized that the nonspecific changes on to misdiagnoses of pulmonary disorders. We also found that muscle biopsy could nonetheless signal a CM. respiratory involvement was more common and more severe among pediatric-onset than adult-onset patients. In 26 other patients, the onset of symptoms was in adulthood. Several of these patients retrospectively reported poor athletic Among patients with distinctive histologic findings, we found performance in childhood. Such data may suggest that the that genetic testing had a yield of 85%. This is in line with true age at onset of the CM in some cases preceded the previous predominantly pediatric studies, which had found reported onset without being recognized. Poor athletic per- that a molecular diagnosis could be established in 57%–79% formance, however, is a nonspecific feature, as there are many of patients with CM.7,29,31 Conversely, among patients with other more common causes of reduced athletic performance a molecular diagnosis, we found that only 42% had any of the in childhood. distinctive histopathologic features of a CM. Core myopathies (central core and multiminicore disease) were the most Among patients with onset of symptoms in adulthood, RYR1 common histologic subtypes. The frequency of nonspecific was the most commonly implicated gene. The proportion was muscle biopsies in our cohort is significantly higher than higher than observed in previous large CM cohorts mainly previously reported and indicates that muscle biopsy has comprising pediatric patients.7,29 a lower sensitivity for CM in adults. This difference is unlikely to be explained by the genetic makeup of our cohort, as it was We found that variants in CM genes were associated with similar to previously reported cohorts. Indeed, nonspecific a spectrum of adult-onset phenotypes, including patients with histologic findings were observed with variants in 5 of the 8 only episodic symptoms or asymptomatic hyperCKemia. Al- different causative genes identified (MYH2, MYH7, ACTA1, though most patients (80%) had fixed weakness, the degree SELENON, and RYR1). The absence of distinctive histo- and pattern of weakness were variable. In addition to limb pathologic features has also previously been observed in weakness, facial weakness was common, but ptosis and oph- young patients with CM, leading to the suggestion that these thalmoparesis, 2 classic features of CM, were only present in features only emerge over time as a result of muscle matura- a single patient. Although CMs are often thought to represent tion and length of disease activity. The finding of a high static or only slowly progressive conditions,6 we found that proportion of nonspecific muscle biopsies in adult patients most adult patients with CM reported a progressive rather with CM, however, indicates that this is not always the case. than static course. Contrary to pediatric-onset patients, CK Similarly, it has been suggested that CFTD may represent an elevations were common among adult-onset CM patients, early change anticipating the development of more specific likely due to the higher proportion of RYR1 myopathy among structural abnormalities, rather than a distinct histopathologic these patients. entity. Our finding of persistent CFTD in adult patients with CM, however, supports the latter possibility. Among adult-onset patients, the most frequent misdiagnoses were non-neurologic conditions, such as rheumatologic and Our data indicate that clinical phenotypes deviating from the orthopedic disorders. We suspect that the insidious onset of classical descriptions of CM are common, as are nonspecific weakness and the lack of awareness of CM among adult biopsy findings. This suggests that NGS targeting CM genes neurologists and other physicians precluded appropriate could be the initial diagnostic tool in patients with suspected investigations and contributed to diagnostic delays and mis- CM and should also be included in the investigation of adult diagnoses. In addition, the misinterpretation of electro- patients with inherited myopathies. diagnostic findings, which in chronic myopathies, can sometimes resemble a primary neurogenic process,30 also In some inherited myopathies, correlations have been estab- played a role in the misdiagnosis. Indeed, we found that long- lished between genotype and age at onset.22,32,33 Among our duration MUPs were common among our cohort and, in most cohort, such correlations were difficult to identify because of cases, could not be accounted for by a neurogenic process. the limited number of patients with variants in each specific gene. We note, however, that in 3/7 families, there were family A cardiomyopathy was present in 29% of patients who un- members exhibiting both pediatric and adult onset. This sug- derwent an echocardiogram. This proportion is higher than gests that additional factors may influence the age at onset. previously reported29 and may reflect a progression of cardiac disease over patients’ lifespans. Two patients, however, had Although not included in our analysis, we also identified 3 evidence of acquired cardiac disorders (ischemic heart disease patients with congenital muscular dystrophies caused by and tricuspid regurgitation), and a contribution from acquired variants in collagen genes (COL6A1, COL6A3, and

Neurology.org/NG Neurology: Genetics | Volume 5, Number 4 | August 2019 7 COL12A1) that were diagnosed in adulthood during the same References period. All 3 had onset of symptoms in adolescence or 1. Romero NB, Clarke NF. Congenital myopathies. Handb Clin Neurol 2013;113: 1321–1336. adulthood and faced diagnostic delays and misdiagnoses 2. Jungbluth H, Treves S, Zorzato F, et al. Congenital myopathies: disorders of similar to those experienced by patients with CM. excitation-contraction coupling and muscle contraction. Nat Rev Neurol 2018;14: 151–167. 3. North KN, Wang CH, Clarke N, et al. Approach to the diagnosis of congenital This study has some limitations. We performed a retrospec- myopathies. Neuromuscul Disord 2014;24:97–116. ’ 4. Mah JK, Joseph JT. An overview of congenital myopathies. Continuum (Minneap tive review of data collected in the course of patients clinical Minn) 2016;22:1932–1953. care, and therefore, not all parameters were available for every 5. Malfatti E, Romero NB. Nemaline myopathies: state of the art. Rev Neurol (Paris) 2016;172:614–619. patient. The course of illness was also assessed retrospectively 6. Wang CH, Dowling JJ, North K, et al. Consensus statement on standard of care for on the basis of patients’ reports, as longitudinal follow-up to congenital myopathies. J Child Neurol 2012;27:363–382. 7. Witting N, Werlauff U, Duno M, Vissing J. Phenotypes, genotypes, and prevalence of objectively assess disease progression was unavailable for congenital myopathies older than 5 years in Denmark. Neurol Genet 2017;3:e140. most patients. Finally, additional cases of CM may have gone 8. Norwood FL, Harling C, Chinnery PF, Eagle M, Bushby K, Straub V. Prevalence of undetected among those patients with nonspecific histologic genetic muscle disease in Northern England: in-depth analysis of a muscle clinic population. Brain 2009;132:3175–3186. findings who did not undergo genetic testing. 9. Cassandrini D, Trovato R, Rubegni A, et al. Congenital myopathies: clinical phenotypes and new diagnostic tools. Ital J Pediatr 2017;43:101. 10. Gonorazky HD, Bönnemann CG, Dowling JJ. The genetics of congenital myopathies. Nevertheless, our study expands the spectrum of CM by Handb Clin Neurol 2018;148:549–564. further characterizing the subgroup of patients presenting in 11. Liewluck T, Sorenson EJ, Walkiewicz MA, Rumilla KM, Milone M. Autosomal dominant distal myopathy due to a novel ACTA1 mutation. Neuromuscul Disord adulthood or diagnosed in adulthood. Compared with pre- 2017;27:742–746. vious cohorts of mainly pediatric patients, adult-onset patients 12. Snoeck M, van Engelen BG, Küsters B, et al. RYR1-related myopathies: a wide spectrum of phenotypes throughout life. Eur J Neurol 2015;22:1094–1112. with CM typically report progressive weakness and often lack 13. Voermans NC, Snoeck M, Jungbluth H. RYR1-related rhabdomyolysis: a common the characteristic histologic abnormalities associated with but probably underdiagnosed manifestation of skeletal muscle ryanodine receptor fi fi dysfunction. Rev Neurol (Paris) 2016;172:546–558. CM. This high prevalence of nonspeci c histologic ndings 14. Echaniz-Laguna A, Biancalana V, Bohm J, Tranchant C, Mandel JL, Laporte J. Adult signals the need for a revision of the current histologically centronuclear myopathies: a hospital-based study. Rev Neurol (Paris) 2013;169:625–631. fi 15. Zukosky K, Meilleur K, Traynor BJ, et al. Association of a novel ACTA1 mutation with based classi cation of CM. Last, we found that adult patients a dominant progressive scapuloperoneal myopathy in an extended family. JAMA with CM commonly experience delays in diagnosis and mis- Neurol 2015;72:689–698. 16. Lehtokari VL, Gardberg M, Pelin K, Wallgren-Pettersson C. Clinically variable diagnoses, emphasizing the need for greater awareness of these nemaline myopathy in a three-generation family caused by mutation of the skeletal disorders among adult neurologists. Indeed, the occurrence of muscle alpha-actin gene. Neuromuscul Disord 2018;28:323–326. “congenital” myopathies in adulthood may warrant the in- 17. Böhm J, Biancalana V, Malfatti E, et al. Adult-onset autosomal dominant centronuclear myopathy due to BIN1 mutations. Brain 2014;137:3160–3170. troduction of a different name for this group of disorders. 18. Fiorillo C, Astrea G, Savarese M, et al. MYH7-related myopathies: clinical, histopathological and imaging findings in a cohort of Italian patients. Orphanet J Rare Dis 2016;11:91. 19. Garibaldi M, Fattori F, Bortolotti CA, et al. Core-rod myopathy due to a novel mu- Study funding tation in BTB/POZ domain of KBTBD13 manifesting as late onset LGMD. Acta This study was funded by a donation from Mr. John N. Lawyer. Neuropathol Commun 2018;6:94. 20. Schnitzler LJ, Schreckenbach T, Nadaj-Pakleza A, et al. Sporadic late-onset nemaline myopathy: clinico-pathological characteristics and review of 76 cases. Orphanet J Rare Disclosure Dis 2017;12:86. 21. Oliveira J, Gruber A, Cardoso M, et al. LAMA2 gene mutation update: toward a more Disclosures available: Neurology.org/NG. comprehensive picture of the laminin-alpha2 variome and its related phenotypes. Hum Mutat 2018;39:1314–1337. 22. Bönnemann CG. The collagen VI-related myopathies Ullrich congenital muscular Publication history dystrophy and Bethlem myopathy. Handb Clin Neurol 2011;101:81–96. Received by Neurology: Genetics February 8, 2019. Accepted in final form 23. Bouchet-Seraphin C, Vuillaumier-Barrot S, Seta N. Dystroglycanopathies: about April 29, 2019. numerous genes involved in glycosylation of one single glycoprotein. J Neuromuscul Dis 2015;2:27–38. 24. Niu Z, Pontifex CS, Berini S, et al. Myopathy with SQSTM1 and TIA1 variants: clinical and pathological features. Front Neurol 2018;9:147. 25. Laughlin RS, Niu Z, Wieben E, Milone M. RYR1 causing distal myopathy. Mol Genet Appendix Authors Genomic Med 2017;5:800–804. 26. Rossi D, De Smet P, Lyfenko A, et al. A truncation in the RYR1 gene associated with Name Location Role Contribution central core lesions in skeletal muscle fibres. J Med Genet 2007;44:e67. 27. Brand P, Dyck PJ, Liu J, Berini S, Selcen D, Milone M. Distal myopathy with coexisting Stefan Mayo Author Acquisition, analysis, and heterozygous TIA1 and MYH7 Variants. Neuromuscul Disord 2016;26:511–515. Nicolau, MD Clinic interpretation of data and writing 28. Richards S, Aziz N, Bale S, et al. Standards and guidelines for the interpretation of of the first draft of the manuscript sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet Teerin Mayo Author Critical revision of the manuscript Med 2015;17:405–424. Liewluck, Clinic for intellectual content 29. Colombo I, Scoto M, Manzur AY, et al. Congenital myopathies: natural history of MD a large pediatric cohort. Neurology 2015;84:28–35. 30. Uncini A, Lange DJ, Lovelace RE, Solomon M, Hays AP. Long-duration polyphasic Jennifer Mayo Author Critical revision of the manuscript motor unit potentials in myopathies: a quantitative study with pathological correla- Tracy, MD Clinic for intellectual content tion. Muscle Nerve 1990;13:263–267. 31. Maggi L, Scoto M, Cirak S, et al. Congenital myopathies: clinical features and fre- Ruple Mayo Author Critical revision of the manuscript quency of individual subtypes diagnosed over a 5-year period in the United Kingdom. Laughlin, Clinic for intellectual content Neuromuscul Disord 2013;23:195–205. MD 32. Pogoryelova O, Wilson IJ, Mansbach H, Argov Z, Nishino I, Lochmuller H. GNE genotype explains 20% of phenotypic variability in GNE myopathy. Neurol Genet Margherita Mayo Author Study concept and design, 2019;5:e308. Milone, MD, Clinic analysis and interpretation of 33. Sveen ML, Schwartz M, Vissing J. High prevalence and phenotype-genotype corre- PhD data, and study supervision lations of limb girdle muscular dystrophy type 2I in Denmark. Ann Neurol 2006;59: 808–815.

8 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG ARTICLE OPEN ACCESS Genome-wide brain DNA methylation analysis suggests epigenetic reprogramming in Parkinson disease

Juan I. Young, PhD, Sathesh K. Sivasankaran, PhD, Lily Wang, PhD, Aleena Ali, BSc, Arpit Mehta, MSc, Correspondence David A. Davis, PhD, Derek M. Dykxhoorn, PhD, Carol K. Petito, MD, Gary W. Beecham, PhD, Dr Young [email protected] Eden R. Martin, PhD, Deborah C. Mash, PhD, Margaret Pericak-Vance, PhD, William K. Scott, PhD, Thomas J. Montine, MD PhD, and Jeffery M. Vance, MD PhD

Neurol Genet 2019;5:e342. doi:10.1212/NXG.0000000000000342 Abstract Objective Given the known strong relationship of DNA methylation with environmental exposure, we investigated whether brain regions aﬀected in Parkinson disease (PD) were diﬀerentially methylated between PD cases and controls.

Methods DNA chip arrays were used to perform a genome-wide screen of DNA methylation on the dorsal motor nucleus of the vagus (DMV), substantia nigra (SN), and cingulate gyrus (CG) of pathologically conﬁrmed PD cases and controls selected using the criteria of Beecham et al. Analysis examined diﬀerentially methylated regions (DMRs) between cases and controls for each brain area. RNA sequencing and pathway analysis were also performed for each brain area.

Results Thirty-eight PD cases and 41 controls were included in the analysis. Methylation studies revealed 234 signiﬁcant DMR in the DMV, 44 in the SN, and 141 in the CG between cases and controls (Sidak p < 0.05). Pathway analysis of these genes showed signiﬁcant enrichment for the Wnt signaling pathway (FDR < 0.01).

Conclusions Our data suggest that signiﬁcant DNA methylation changes exist between cases and controls in PD, especially in the DMV, one of the areas aﬀected earliest in PD. The etiology of these methylation changes is not yet known, but the predominance of methylation changes occurring in the DMV supports the hypothesis that vagus nerve function, perhaps involving the gastrointestinal system, is important in PD pathogenesis. These data also give independent support that genes involved in Wnt signaling are a likely factor in the neurodegenerative processes of PD.

From the John P. Hussman Institute for Human Genomics (J.I.Y., S.K.S., A.A., A.M., D.M.D., G.W.B., E.R.M., M.P.-V., W.K.S., J.M.V.), Miller School of Medicine, University of Miami; Department of Public Health Sciences (L.W.), Division of Biostatistics, Miller School of Medicine, University of Miami; Department of Neurology (D.A.D., D.C.M.), Miller School of Medicine, University of Miami; Department of Pathology (C.K.P.), Miller School of Medicine, University of Miami, FL; and Department of Pathology (T.J.M.), Stanford University, CA.

Deborah C. Mash is currently at NOVA Southeastern University, Ft. Lauderdale, FL.

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

Copyright © 2019 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 Glossary AAD = age at death; CG = cingulate gyrus; DEG =differentially expressed gene; DMR =differentially methylated region; DMV = dorsal motor nucleus of the vagus; DUSP22 = dual-specificity phosphatase 22; IC = independent component; iSVA = independent surrogate variable analysis; LB = Lewy body; PD = Parkinson disease; SN = substantia nigra; SV = surrogate variable; TFBS = transcription factor binding site.

Parkinson disease (PD) is the second most common neuro- neurofibrillary tangle stage ≤IV. Neuropathologic evaluations degenerative disorder affecting older adults. The clinical were performed in every case confirming Lewy body (LB) PD, presentation includes bradykinesia, resting tremor, and ri- and followed the filtering criteria of Beecham et al.,15 for in- gidity.1 Monogenic mutations for PD have been identified clusion into the study. Staging was performed according to the that greatly increase the risk of PD.2 However, 90% or more of Braak hypothesis of LB staging in PD. PD samples included PD cases are idiopathic. brains with Braak stages ranging from 3 to 5 (36% stage 3, 37% stage 4, and 27% stage 5). Tissue punches of 0.3 cm in diameter Epigenetics is a potentially important factor contributing to were taken from fresh, quick-frozen 1 cm coronal sections from PD risk, particularly since environmental factors have been each region studied, e.g., the medulla oblongata containing the – associated with an increased risk of developing PD.3 5 DMV, the SN, and the anterior CG. Tissue punches corre- However, little work has been done to explore the potential sponded to approximately an average of 50% of the DMV, 30% epigenetic contribution to PD. DNA methylation, the mostly of the SN, and 5% of the CG. These brain regions represent the studied form of epigenetic modification, has been primarily location of neuropathologic changes in PD at different stages of investigated in PD within select candidate genes.6,7 Several the disease. One of the earliest regions to display LBs and Lewy studies have found differential methylation in SNCA and neurites, the hallmarks of PD, is the DMV, with the charac- – MAPT.8 11 Furthermore, significant changes in DNA meth- teristic motor symptoms occurring later, when over 50% of the ylation were identified in multiple genes in both blood10,12,13 dopaminergic neurons of the SN’s pars compacta are lost.1 The and brain.10 Relevant to the current study, dysregulation of DMV has a direct connection to the environment through Wnt signaling due to methylation was observed in the frontal the vagus nerve’s innervation of the gastrointestinal tract, of cortex and midbrain sections of PD brains.14 which dysfunction in PD has been reported in multiple studies.16,17 The CG generally develops later pathologic Here, we report an initial analysis of the genome-wide changes in PD.18 A40-μm section of the tissue punch con- methylation profile in the dorsal motor nucleus of the vagus taining the DMV was used for anatomic verification. (DMV), substantia nigra (SN), and cingulate gyrus (CG) of pathology-confirmed PD cases compared with age- and sex- Standard protocol approvals, registrations, matched, pathology-confirmed controls. Each of these brain and patient consents regions represents the location of neuropathologic changes in The authorization for retention of the brain, review of medical PD at different stages of the disease. We found that patients records, and informant interviews were approved by the re- with PD have significant DNA methylation changes in these 3 spective institutional review boards. brain regions, and find the largest number of significant DNA methylation changes are found in the DMV. Furthermore, Profiling of CpG methylation using 450k/ pathway analysis in the DMV of patients with PD supports the 850k array involvement of the Wnt pathway in the pathophysiology of PD. Genomic DNA (500 ng) was bisulfite modified (EZ-96 DNA Methylation Kit; Zymo Research, Orange, CA) as per manufacturers’ instructions. For analysis of CG dinucleotide (CpG) Methods methylation, both the Illumina Infinium HumanMethylation450 BeadChip and the Infinium MethylationEPIC (850K array) Brain samples were obtained from the autopsy program of the Beadchip were used (Illumina, San Diego, CA) because of dis- University of Miami, Morris K. Udall Parkinson Disease continuation of the 450 BeadChip by the manufacturers. Center of Excellence (n = 11), the NIH Neurobiobank (n = 12), and the Pacific Udall Center Neuropathology Core (n = Statistical analysis of methylation data 22). Our initial discovery sample set consisted of 22 PD Raw intensity files were processed using the methylation pathology-confirmed cases and 24 pathology-confirmed analysis software RnBeads.19 Because all samples were male, controls. The replication data set had 16 PD pathology- we did not filter sex-specific probes. Beta-Mixture Quantile confirmed cases and 17 pathology-confirmed controls, pro- normalization was used for intra-array normalization of beta viding a total of 38 PD cases and 41 controls for the joint values, which are the ratio of the methylated signal intensity to analysis. All brains were from non-Hispanic white men, aged the sum of both methylated and unmethylated signals after >60 years at death with no premortem diagnosis of cancer, background subtraction. The beta values were then logit postmortem interval of less than 24 hours, and a Braak transformed to attain M values for statistical analysis.20

2 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG Unsupervised hierarchical clustering of the methylome data proven to induce methylation changes in cultured human revealed 1 outlier from the CG group, which was removed peripheral blood mononuclear cells.26 iPSC-derived dopa- from further analysis. minergic neurons differentiated for 70 days were cultured with varying levels of L-Dopa (0, 5, 10, and 50 uM) in Neu- For the analysis of individual CpGs, linear regression models robasal N2/B27 media (Gibco) supplemented with 1 ng/mL were used to test differential methylation between case-control of transforming growth factor beta-3, 10 ng/mL of sonic status adjusting for age at death (AAD) and methylation chip hedgehog signaling molecule, 20 ng/mL of brain-derived effects. To account for additional unmeasured confounding neurotrophic factor, and 30 ng/mL of glial cell-derived neu- factors such as cellular composition, we included surrogate rotrophic factor. After 7 days, DNA was extracted and sub- variables (SVs) estimated from independent surrogate variable jected to methylation analysis. analysis (iSVA) as covariate variables.21 iSVA estimates major independent components (ICs) in genome-wide DNA meth- RNA sequencing and statistical analysis ylation patterns. We tested each IC with status using a T test. RNA-seq (RNA integrity number ≥ 5) was performed using The significant ICs (IC3 for DMV and IC7 for SN), which a paired-end 125 bp protocol on a HiSeq 2500. Reads were could be confounders of the association between case-control aligned to the human reference genome (hg19) using the status and differential methylation, were then included in the STAR algorithm and analyzed using the “voom” method in 27 linear model: M value ;PD + AAD + array + IC. We only the Limma package. considered CpGs that showed a difference in group means in methylation M values (|DM|) of at least 25% (|DM| ≥ 1.5) and A linear model with AAD was fitted to each gene, and em- false discovery rate (FDR) <0.05 in the 2 group comparison. pirical Bayes moderated t-statistics and p values were used to assess expression differences between PD and controls. To account for underlying unknown confounding factors, we Differentially methylated region analysis 28 The majority of genome-wide methylation studies have fo- used SVAseq with default parameters to estimate SVs. None ff fi cused on single CpG sites. However, modification of single of the estimated SVs di ered signi cantly between case- CpG often produces weak correlations with gene expression control status, so we did not include them in the linear model: ; data.22 Contextualizing the methylation level of an individual log (cpm) PD + AAD. CpG by the status of neighboring CpG sites facilitates biological Pathway analysis inferences. Clusters of neighboring CpGs with coordinated dif- We used Enrichr (amp.pharm.mssm.edu/Enrichr/), which yields ferential methylation are identified as differentially methylated an FDR-adjusted p value for each pathway.29 The binding affinity regions (DMRs). Hypermethylated DMRs in promoters are of most transcription factorstoDNAisalteredbyDNACpG usually associated with silencing, whereas in the gene body, they methylation.30 Thus, we analyzed the presence of transcription associate with upregulation of expression.23 We defined a DMR factor binding sites (TFBSs) using the R package Goldmine.31 as a region including (1) 3 or more consecutive significantly ff p di erentially methylated ( < 0.05) sites between PD and con- Data availability trol groups with the same direction of methylation change; (2) ff Raw data for the primary analyses are available upon rea- each di erentially methylated CpG separated by less than 500 sonable request from the corresponding and senior author. bp; and (3) a multiple-comparison corrected p value (Sidak p) less than 0.05 for the region. Results 24 DMR analysis was performed using comb-p. Comb-p takes Samples as input unadjusted p values for each probe, identifies regions The average AAD for PD cases was 78.1 years (range 67–90 of enrichment (i.e., series of adjacent low p values), and computes years) in the discovery data set and 79.8 years (range 66–89 statistical significance of the regions using the Sidak correction. years) in the replication data set. For controls, it was 77.7 years (range 64–91 years) in the discovery and 79.6 years (range Levodopa – ff 25 67 95 years) in the replication data set. In multiple samples, we Levodopa (L-Dopa) has been shown to a ect methylation levels ffi and thus could be a contributor to the methylation changes ob- were unable to isolate all 3 regions because of insu cient material served. To address this, we used 2 approaches: (1) we examined or excessive degeneration. This is shown in tables 1 and 2, with age at onset, AAD, and the regions isolated from each donor. the dose-response relationship with L-Dopa in a line of iPSC- derived dopaminergic neurons generated from a control non- DMR analysis Hispanic white male donor. (2) We also compared methylation changes we found to those reported in a model of 7-day L-Dopa Discovery data set administration to rats rendered hemiparkinsonian through uni- Analysis in the discovery data set identified 85 DMRs in the lateral injections of 6-hydroxydopamine (6-OHDA).25 DMV, 65 in the CG, and 27 in the SN samples with Sidak p < 0.05 (table e-1, links.lww.com/NXG/A164). These DMRs We selected nontoxic concentrations ranging from 0 to 50 μM were associated with 108, 84, and 31 genes in the DMV, CG, of L-Dopa, including treatment with 10 μM, a concentration and SN, respectively. Within the discovery data set, comparison

Neurology.org/NG Neurology: Genetics | Volume 5, Number 4 | August 2019 3 Table 1 Samples investigated in the discovery methylation and RNA-seq study

Methylation RNA-seq Methylation RNA-seq Additional neuropath Control AAD DMV CG SN DMV CG SNPD AAO AAD DMV CG SN DMV CG SN diagnosis

C-001 71 Y Y Y Y P-395 65 74 Y Y Y Acute hemorrhage

C-002 75 Y Y P-346 73 81 Y Y Y AD

C-005 76 Y Y Y P-548 70 76 Y Y Y Y

C-007 78 Y Y Y P-784 48 67 Y Y

C-008 79 Y Y Y P-002 63 68 Y Y Y Y Y

C-009 80 Y Y Y P-755 65 81 Y Y Y

C-010 65 Y Y Y P-225 62 72 Y Y

C-011 66 Y Y Y P-545 62 72 Y Y Y Y

C-012 68 Y Y Y P-547 73 86 Y Y Y Y

C-013 67 Y Y P-549 68 83 Y Y Y Y

C-336 64 Y Y Y P-554 64 80 Y Y Y Y

C-132 77 Y P-812 62 76 Y Y Y

C-738 78 Y Y Y Y P-447 70 Y Y Y Y

C-598 73 Y Y Y Y P-438 84 Y Y Y Y AD

C-632 78 Y Y Y Y P-748 88 90 Y Y Y Y Subdural hematoma

C-790 83 Y Y Y Y P-524 75 78 Y Y Y Y AD

C-331 81 Y Y Y Y P-533 73 Y Y Y Y AD

C-535 86 Y Y Y Y P-080 76 Y Y

C-642 84 Y Y Y P-376 84 88 Y Y

C-353 91 Y Y P-610 81 81 Y Y

C-511 87 Y Y P-206 77 78 Y AD

C-434 84 Y Y P-698 71 85 Y Y AD

C-752 87 Y Y

C-492 87 Y Y

Total 16 11 20 6 5 12 13 11 19 5 9 13

Abbreviations: AAD = age at death; AAO = age at onset; CG = cingulate gyrus; DMR = differentially methylated region; DMV = dorsal motor nucleus of the vagus; PD = Parkinson disease; RIN = RNA integrity number; SN = substantia nigra. All brains were from non-Hispanic white men, aged >60 years at death with a Braak neurofibrillary tangle stage ≤ IV. In several samples, we were unable to isolate all 3 regions because of insufficient material and/or excessive degeneration. We analyzed the DMV in 67% of control samples and in 59% of PD samples; the CG in 46% of control samples and in 50% of PD cases; and the SN in 83% of control samples and in 86% of PD cases. RNA was used only in those samples that reached a quality cutoff (RNA integrity number, RIN ≥ 5). Samples used are indicated by a “Y.”

of the DMRs identified in the 3 tissues revealed that a region set in the CG, 4 in the DMV, and none in the SN (figure and spanning the promoter region of dual-specificity phosphatase 22 table 3). We therefore performed a gene-based analysis that (DUSP22) was hypomethylated in PD brains in the 3 tissues. revealed genes that contained DMRs in both the DMV Furthermore, another 7 genomic regions were differentially (FRMD4A and GPT) and the CG (HOXA3 and PRDM16)of hypermethylated in both the DMV and the SN (RNF5, patients with PD in which the location of the DMR is not the AGPAT1, LANCL2, LMTK3, SCAND3, SLFN12,and same in the discovery and the replication data set (figure). ZDHHC14). This analysis identified several genes, including ARFGAP1, a reported regulator of LRRK2 toxicity in PD.32,33 Replication data set We tested whether any of the significant DMRs identified in Joint data set the discovery data set were also differentially methylated As both the discovery and replication autopsy data sets were in the replication data set (Sidak p < 0.05). We found that 7 of limited in sample numbers and showed replication between the discovery DMRs were reproduced in the replication data them, to increase the power of detecting disease-associated

4 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG Table 2 Samples investigated in the replication methylation study

Methylation Methylation

Control AAD DMV CG SNPD AAO AAD DMV CG SN Additional neuropath diagnosis

C-001 84 Y Y P-593 57 66 Y

C-002 92 Y Y Y P-208 70 86 Y Focal severe gliosis

C-003 80 Y Y Y P-001 82 89 Y

C-004 67 Y Y Y P-401 53 80 Y

C-005 70 Y Y Y P-327 44 68 Y Y

C-006 70 Y Y Y P-634 68 81 Y Y Y

C-007 70 Y Y Y P-686 80 84 Y Y Y

C-008 85 Y P-443 79 89 Y Y Y

C-009 82 Y Y Y P-531 48 70 Y Y

C-010 84 Y Y Y P-045 62 80 Y Y

C-011 80 Y Y P-457 Y

C-012 85 Y Y P-678 65 84 Y Y Y

C-013 75 Y Y Y P-679 Y Y Y

C-014 90 Y Y Y P-457 73 86 Y Y

C-015 77 Y Y Y P-904 64 78 Y Y Y

C-016 68 Y Y Y P-625 65 77 Y Y Y

C-017 95 Y Y Y

Total 15 15 16 9 15 10

Abbreviations: AAD = age at death; AAO = age at onset; CG = cingulate gyrus; DMV = dorsal motor nucleus of the vagus; PD = Parkinson disease. All brains were from non-Hispanic white men, aged >60 years at death with a Braak neurofibrillary tangle stage ≤ IV. In several samples, we were unable to isolate all 3 regions because of insufficient material and/or excessive degeneration. We analyzed the DMV in 88% of control samples and in 56% of PD samples; the CG in 88% of control samples and in 93% of PD cases; and the SN in 94% of control samples and in 62% of PD cases. Samples used are indicated by a “Y”. methylation changes, we performed a single joint analysis. CG or SN, even using FDR < 0.25 as the significance cutoff,as This provided us a total of 53 DMV (22 PD and 31 controls), previously suggested for pathway analysis.35,36 52 CG (26 PD and 26 controls), and 65 SN (29 PD and 36 controls) for the analysis. In the joint analysis, we identified Integrated analysis of differential methylation 234 significant DMR in the DMV, 44 in the SN, and 141 in the and RNA-Seq CG (table e-2, links.lww.com/NXG/A165). These corre- We observed that ;80% of the DMRs identified contain spond to 266, 53, and 168 genes, respectively. The top 20 TFBS (table e-4, links.lww.com/NXG/A167), suggesting that DMRs in the joint analysis from each region are shown in the identified differences in DNA methylation are likely to table 4. In the joint analysis, a DMR in the promoter area of have transcriptional consequences. We then performed LOC100420587 is hypermethylated in the 3 brain regions. It transcriptome analysis by RNA-seq on a subset of samples is interesting to note that an SNP in this noncoding gene of (table 1). Analysis of the RNA-seq data identified 515 dif- unknown function was identified as associated with the vol- ferentially expressed genes (DEG) in the CG, 390 DEG in ume of the CG by neuroimaging and GWASs.34 the SN, and 3 DEG in the DMV associated with PD (FDR <0.05, adjusted for AAD). An overlap analysis of both Pathway analysis identified Wnt signaling as methylation data and RNA-seq revealed 6 genes with DMRs epigenetically affected in the DMV of PD brains that were also differentially expressed (NDRG4, PTPRN2, We identified physiologic pathways overrepresented among SYT7, IQSEC1, DLG4, and KCNIP1) in the CG and 1 the genes associated with DMRs identified in the joint anal- (NDRG4) in the SN. ysis. Significant enrichment was observed in the KEGG “Hippo signaling pathway” (FDR = 0.007) and “Wnt signal- Levodopa ing pathway” (FDR = 0.01) in the DMV (table e-3, links.lww. We found 14 DMRs that changed their methylation levels sig- com/NXG/A166). No pathway enrichment was observed for nificantly (Sidak p < 0.05; table e-5, links.lww.com/NXG/A168)

Neurology.org/NG Neurology: Genetics | Volume 5, Number 4 | August 2019 5 Figure PD-associated DMRs

(A) DMRs previously identified in the discovery data set were also significantly differentially methylated in the same direction in the replication data set. The SN is not included because we did not detect overlap at the DMR or gene level with any of the other 2 tissues. (B) Common PD-associated DMRs identified in the 3 analyzed tissues in the joint analysis. A DMR in LOC100420587, an SHC binding and spindle associated 1 pseudogene, is present in all 3 regions. If DMRs could be assigned to more than 1 gene, both genes are shown separated by a slash. Asterisks denote overlap at the gene level but not at the DMR level (distinct DMRs assigned to the same gene). In the Venn diagram, the numbers are the significant DMRs in the 2 data sets/brain regions and those shared between them. Brackets besides gene names indicate the genomic location of the DMR. CG = cingulate gyrus; DMR = differentially methylated region; DMV = dorsal motor nucleus of the vagus; GB = gene body; PD = Parkinson disease; Pr = promoter; SN = substantia nigra; Upst = upstream intergenic region.

on L-Dopa treatment of iPSC-derived dopaminergic neu- Discussion rons. Comparison of the DMRs identified in the joint analysis in the different tissues with the methylation changes This initial study of DNA methylation changes in the DMV, induced by L-Dopa treatment showed no overlap. Further- SN, and CG supports our hypothesis of an epigenetic contri- more, comparison of our data (all genes containing DMR bution to PD risk. Whether the identified changes are inherited, irrespective of the brain region) with the genes identified acquired de novo during development, in part due to cellular as differentially methylated in the dorsal striatum of composition changes or induced by environmental variables is 25 6-OHDA–treated rats receiving L-Dopa with FDR < 0.05 currently unknown. However, it is certainly interesting to spec- and absolute change of 10% (2703 genes) revealed an ulate that these methylation changes might be due to environ- overlap of 82 genes. Thus, the rat model–human comparison mental influences through the vagus nerve. If this were the case, data suggest that approximately 24.1% of the genes identified it would suggest that methylation is an early factor in the de- as differentially methylated in PD could be related to L-Dopa velopment of PD, as the DMV is thought to be one of the earliest administration. To attempt to identify methylation changes regions to develop characteristic PD pathologic changes. that could be induced by the presence of cell death or hypofunctioning neurons, we determined which of the Epigenetic patterns are different between cell types, specifi- DMR-containing genes (from the joint analysis) were also cally neurons and glia.37 Unlike the DMV, which does not differentially methylated in hemiparkinsonian rats not have extensive degeneration, cell loss in the SN is prominent, given L-Dopa. This revealed 60 genes with differential and thus, it is possible that a change in cellular composition methylation shared between our human data and those between controls and patients with PD contributes to some of responsive to 6-OHDA lesion in the rat striatum as pre- the changes we have seen despite the use of iSVA to correct viously identified.25 for cellular heterogeneity. Furthermore, heterogeneity in the

6 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG Table 3 Replicated DMR in the DMV and CG of PD brains

DMR Chr Start End CpGsa p Value FDR Gene Direction

Replicated in DMV chr8 144343915 144344794 4 8.69E-05 0.0075 ZFP41 −

chr5 87973439 87974548 10 0.0009715 0.0331 LOC645323 +

chr13 112724221 112724584 3 0.001156 0.0331 SOX1 +

chr20 61915437 61916280 5 0.001954 0.042 ARFGAP1 +

Replicated in CG chr9 140171765 140175394 13 0.000116 0.0074 C9orf167 −

chr15 81426347 81426821 9 0.000322 0.0074 C15orf26 −

chr19 21646006 21646782 5 0.000323 0.0074 −

chr17 81038827 81039991 6 0.000459 0.0079 METRNL −

chr16 123246 123677 5 0.001079 0.0149 RHBDF1 −

chr10 105420501 105421250 5 0.003333 0.0383 SH3PXD2A −

chr3 194030679 194030978 3 0.004104 0.0405 LOC100131551 −

Abbreviations: CG = cingulate gyrus; DMR = differentially methylated region; DMV = dorsal motor nucleus of the vagus; PD = Parkinson disease. a Probes refers to the number of CpGs included in the DMR. Direction refers to hypermethylation (+) or hypomethylation (−) in PD. amount of cell death in the different cell areas studied here is interact and appear to regulate each other’s expression.32,33 anticipated because of variable Lewy pathology Braak staging Thus, the methylation changes would suggest a wider role for (ranging from stage 3 to stage 5), potentially influencing the ARFGAP1 in PD pathophysiology. Neurexin 3, thought to be results. involved in synaptic plasticity, has been associated with multiple psychiatric disorders including Alzheimer disease.39 We analyzed the overlap between our results and changes Of interest, the promoter of DUSP22, associated with the observed in L-Dopa–treated rats. The 6-OHDA–induced par- most significant DMR in the SN in this study, was shown to be kinsonism model has limitations in terms of progression and hypermethylated in the hippocampus,40 whereas a region recapitulation of the age-related effects of PD. However, this upstream of DUSP22 was found to be hypomethylated in 41 parkinsonism model has been useful in the study of L-Dopa- the superior temporal gyrus of patients with Alzheimer 38 related dyskinesia. The small overlap in L-Dopa–associated disease. It has been recently suggested that DUSP22 is in- changes in DNA methylation and the changes we observed, with volved in both the phosphorylation of tau and CREB sig- the fact that the DMV, SN, and CG displayed minimal overlap in naling, both shown to be involved in Alzheimer disease.40 DMRs, yet are all exposed to the drug, would suggest that Furthermore, hypermethylation of the DUSP22 promoter 42 L-Dopa is not the major contributor to the changes observed here. was associated with schizophrenia. It is interesting that we did not find any DMR in SNCA that has been shown to The finding that most DMRs contain TFBS suggests that the have methylation changes previously in the cortex, SN, and PD-associated DNA methylation changes have regulatory blood in PD. The reasons for this are not clear, but may potential. We, nevertheless, did not observe a correlated reflect tissue degeneration in the earlier affected regions, change in gene expression in the DMV and only one gene in well as a larger data set reported here, as it is likely to vary the SN. Notably, in the CG, an area affected late in the course between individuals. of PD, we did detect some transcriptional dysregulation ac- companied by DNA methylation changes. This suggests the Identifying enriched functional pathways revealed epigenetic possibility that the identified DMRs constitute stable mod- perturbation in the interrelated pathways of “Wnt signaling” ifications of the epigenome, but their acute effects (tran- and “Hippo signaling” in the DMV. Wnt signaling has been scriptional changes) are not stably maintained and previously implicated in PD via expression and methylation,14 – compensated by other mechanisms. Furthermore, a subset of genetic,43 50 and modeling46,47 approaches. Wnt proteins are the differentially methylated/expressed genes in the CG has critical mediators of cell-to-cell communication and in- also DMRs in the SN, supporting the idea of stable DNA tracellular signaling associated with CNS development.43 modifications reflecting previous transcriptional changes. Particularly important for PD is their role in determining midbrain dopaminergic cell fate and functioning.48 Recently, A DMR in ADP-ribosylation factor 1 GTPase activating Wnt signaling has been implicated in propagating the innate protein 1 (ARF1GAP) is the most significant DMR in the immune function in multiple tissues. Such regulation of in- joint analysis of the DMV (table 3). ARFGAP1 and LRRK2 flammation, a process implicated in PD49 and in

Neurology.org/NG Neurology: Genetics | Volume 5, Number 4 | August 2019 7 Table 4 Top 20 DMRs identified in the joint analysis

DMR Chr Start End CpGsa SLK p value Sidak p value Associated genes Direction Ref.

DMV chr20 61915437 61916280 5 5.01E-12 2.70E-09 ARFGAP1 + 39 (PD)

chr14 79744991 79746781 12 8.80E-11 2.23E-08 NRXN3 + 43 (AD)

chr5 87973439 87974548 10 7.76E-11 3.17E-08 LOC645323 +

chr8 144343915 144344794 4 6.88E-11 3.54E-08 ZFP41 −

chr6 99278991 99280514 9 4.43E-10 1.32E-07 POU3F2 +

chr10 118892211 118894181 13 7.21E-10 1.66E-07 VAX1 +

chr2 27529325 27531536 18 5.73E-09 1.18E-06 TRIM54, UCN +

chr16 3017495 3018471 7 2.95E-09 1.37E-06 KREMEN2, PAQR4 +

chr17 79315848 79317340 7 6.05E-09 1.84E-06 ENSG00000171282, TMEM105 +

chr1 221053409 221055965 15 1.35E-08 2.40E-06 HLX +

chr11 73356316 73357397 11 1.03E-08 4.33E-06 PLEKHB1 +

chr6 33279563 33284498 99 9.04E-08 8.30E-06 TAPBP, ZBTB22 +

chr2 233924713 233925276 11 1.37E-08 1.10E-05 INPP5D + 59 (AD)

chr11 68919873 68920772 9 3.46E-08 1.74E-05 CCND1, TPCN2 + 60 (PD)

chr2 54785178 54786149 9 4.41E-08 2.06E-05 SPTBN1 + 61 (PD)

chr17 79372242 79374742 16 1.71E-07 3.09E-05 BAHCC1 +

chr11 2889602 2891496 41 1.44E-07 3.45E-05 KCNQ1DN −

chr7 1120465 1121930 9 1.36E-07 4.21E-05 C7orf50 +

CG chr15 96868857 96869221 8 1.73E-09 2.14E-06 NR2F2 −

chr16 123246 123677 5 1.96E-09 2.05E-06 RHBDF1 −

chr14 24640947 24641707 11 1.12E-08 6.63E-06 REC8 −

chr17 33825172 33825375 3 1.27E-08 2.82E-05 SLFN12L −

chr1 156610966 156612437 8 3.55E-08 1.09E-05 BCAN −

chr19 29217858 29218775 7 5.77E-08 2.84E-05 LOC100420587 +

chr3 46506104 46506865 11 6.44E-08 3.81E-05 LTF −

chr6 33560953 33561450 8 7.62E-08 6.91E-05 C6orf227 −

chr1 167682648 167683014 5 7.81E-08 9.62E-05 MPZL1, RCSD1 −

chr17 81038827 81039991 6 1.42E-07 5.50E-05 METRNL −

chr1 221060360 221061255 6 1.54E-07 7.75E-05 HLX, DUSP10 −

chr19 2041905 2042593 6 1.64E-07 0.000107 MKNK2 −

chr20 821854 822789 4 1.76E-07 8.48E-05 FAM110A −

chr19 58907184 58907510 3 2.09E-07 0.000289 ENSG00000269855 −

chr13 36048892 36051074 15 2.37E-07 4.90E-05 MAB21L1, MIR548F5, NBEA −

chr3 138655775 138656629 6 3.02E-07 0.000159 PIK3CB, FOXL2 −

chr17 38465281 38465511 7 3.45E-07 0.000676 RARA −

chr9 140171097 140175394 16 3.55E-07 3.72E-05 C9orf167 −

Continued

8 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG Table 4 Top 20 DMRs identified in the joint analysis (continued)

DMR Chr Start End CpGsa SLK p value Sidak p value Associated genes Direction Ref.

SN chr6 291687 293332 10 5.18E-13 1.42E-10 DUSP22 − 45 (AD)

chr17 33759512 33760528 12 3.25E-10 1.44E-07 SLFN12 +

chr19 29217858 29218775 7 3.75E-10 1.84E-07 LOC100420587 +

chr7 64348740 64350151 9 1.08E-08 3.46E-06 ZNF273, ZNF138 −

chr11 50257256 50258751 10 1.24E-07 3.73E-05 LOC441601 +

chr19 55972504 55973779 11 4.47E-07 0.000158 ISOC2 −

chr4 40858965 40859345 7 2.79E-07 0.00033 APBB2 − 62 (AD)

chr22 47081634 47082261 5 1.89E-06 0.001352 CERK +

chr1 2138442 2139658 7 4.09E-06 0.001513 C1orf86 +

chr2 223164459 223167618 20 1.65E-05 0.00235 PAX3, CCDC140, CCDC140 − 63 (AD)

chr7 55430948 55431277 3 1.91E-06 0.002611 LANCL2 +

chr17 7757148 7759141 20 1.35E-05 0.003041 KDM6B, TMEM88, TMEM88 +

chr6 158013621 158014656 6 8.81E-06 0.003821 ZDHHC14 +

chr5 174158195 174159904 8 1.46E-05 0.003838 MSX2, DRD1 − 64 (PD)

chr3 46792023 46792463 5 5.95E-06 0.006069 PRSS45, PRSS50 +

chr9 130955135 130955437 3 4.17E-06 0.006197 CIZ1 + 65 (AD)

chr6 28601271 28601520 12 3.58E-06 0.006444 SCAND3, TRIM27 + 66 (PD)

chr16 58534681 58535557 7 1.28E-05 0.006526 NDRG4 − 67 (AD)

Abbreviations: AD = Alzheimer disease; CG = cingulate gyrus; DMR = differentially methylated region; DMV = dorsal motor nucleus of the vagus; NRXN3 = neurexin 3; PD = Parkinson disease; SN = substantia nigra. Direction refers to hypermethylation (+) or hypomethylation (−) in PD. a CpGs refers to the number of CpGs included in the DMR. neurodegeneration in general, could be a primary source of Wnt’sinﬂuence on PD.50 Appendix Authors

Name Location Role Contribution Thus, our data support an epigenetic component to the de- ﬁ Juan I. Young, University of Author Designed and velopment of PD and t well within the growing body of evidence PhD Miami, FL conceptualized the study; involving the DMV and the vagus nerve in PD. Furthermore, our major role in the acquisition of data; analyzed data; and data support the previous studies suggesting deregulated Wnt drafted the manuscript signaling contributing to the pathogeneses of PD. Sathesh K. University of Author Analyzed data and drafted Sivasankaran, Miami, FL the manuscript Acknowledgment PhD The authors are grateful to the families and staﬀ who participated Lily Wang, PhD University of Author Analyzed data and drafted in this study. Some of the samples used in this study were Miami, FL the manuscript collected while the Udall PDRCE was based at Duke University. Deborah C. University of Author Provided samples, Mash, PhD Miami, FL performed pathologic Study funding evaluations, and This study was supported by NIH grants NS071674 and conceptualized the study NS062684. Aleena Ali, BSc University of Author Major role in the acquisition Miami, FL of data

Disclosure William K. University of Author Performed biostatistical Disclosures available: Neurology.org/NG. Scott, PhD Miami, FL review of results Thomas J. Stanford Author Provided samples, Publication history Montine, MD, University, performed pathologic Neurology: Genetics fi PhD Stanford, CA evaluations, and Received by November 2, 2018. Accepted in nal conceptualized the study form May 9, 2019. Continued Neurology.org/NG Neurology: Genetics | Volume 5, Number 4 | August 2019 9 20. Du P, Zhang X, Huang CC, et al. Comparison of beta-value and M-value methods for Appendix (continued) quantifying methylation levels by microarray analysis. BMC Bioinformatics 2010;11: 587. 21. Teschendorff AE, Zhuang J, Widschwendter M. Independent surrogate variable Name Location Role Contribution analysis to deconvolve confounding factors in large-scale microarray profiling studies. Bioinformatics 2011;27:1496–1505. Jeffery M. University of Author Designed and 22. Vanderkraats ND, Hiken JF, Decker KF, Edwards JR. Discovering high-resolution Vance, MD, Miami, FL conceptualized the study; patterns of differential DNA methylation that correlate with gene expression changes. PhD analyzed data; and edited Nucleic Acids Res 2013;41:6816–6827. the manuscript 23. Aran D, Toperoff G, Rosenberg M, Hellman A. Replication timing-related and gene body-specific methylation of active human genes. Hum Mol Genet 2011; Arpit Mehta, University of Author Analyzed data 20:670–680. MSc Miami, FL 24. Pedersen BS, Schwartz DA, Yang IV, Kechris KJ. Comb-p: software for combining, analyzing, grouping and correcting spatially correlated P-values. Bioinformatics 2012; David A. Davis, University of Author Performed pathologic 28:2986–2988. PhD Miami, FL evaluations 25. Figge DA, Eskow Jaunarajs KL, Standaert DG. Dynamic DNA methylation regulates levodopa-induced dyskinesia. J Neurosci 2016;36:6514–6524. Derek M. University of Author Coordinated iPSC studies 26. Schmitt I, Kaut O, Khazneh H, et al. L-dopa increases alpha-synuclein DNA meth- Dykxhoorn, Miami, FL ylation in Parkinson’s disease patients in vivo and in vitro. Mov Disord 2015;30: PhD 1794–1801. 27. Ritchie ME, Phipson B, Wu D, et al. Limma powers differential expression analyses for Carol K. Petito, University of Author Performed pathologic RNA-sequencing and microarray studies. Nucleic Acids Res 2015;43:e47. MD Miami, FL evaluations 28. Leek JT. svaseq: removing batch effects and other unwanted noise from sequencing data. Nucleic Acids Res 2014;42:e161. Gary W. University of Author Performed biostatistical 29. Kuleshov MV, Jones MR, Rouillard AD, et al. Enrichr: a comprehensive gene set Beecham, PhD Miami, FL review of results enrichment analysis web server 2016 update. Nucleic Acids Res 2016;44:W90–W97. 30. Yin Y, Morgunova E, Jolma A, et al. Impact of cytosine methylation on DNA binding Eden R. University of Author Performed biostatistical specificities of human transcription factors. Science 2017;356;eaaj2239. Martin, PhD Miami, FL review of results 31. Bhasin JM, Ting AH. Goldmine integrates information placing genomic ranges into meaningful biological contexts. Nucleic Acids Res 2016;44:5550–5556. Margaret University of Author Provided samples and 32. Xiong Y, Yuan C, Chen R, Dawson TM, Dawson VL. ArfGAP1 is a GTPase activating Pericak- Miami, FL conceptualized the study protein for LRRK2: reciprocal regulation of ArfGAP1 by LRRK2. J Neurosci 2012;32: Vance, PhD 3877–3886. 33. Stafa K, Trancikova A, Webber PJ, Glauser L, West AB, Moore DJ. GTPase activity and neuronal toxicity of Parkinson’s disease-associated LRRK2 is regulated by Arf- GAP1. PLoS Genet 2012;8:e1002526. 34. Bi X, Yang L, Li T, Wang B, Zhu H, Zhang H. Genome-wide mediation analysis of References psychiatric and cognitive traits through imaging phenotypes. Hum Brain Mapp 2017; 1. Kalia LV, Lang AE. Parkinson’s disease. Lancet 2015;386:896–912. 38:4088–4097. 2. Mullin S, Schapira A. The genetics of Parkinson’s disease. Br Med Bull 2015;114: 35. Subramanian A, Tamayo P, Mootha VK, et al. Gene set enrichment analysis: 39–52. a knowledge-based approach for interpreting genome-wide expression profiles. Proc 3. Allen MT, Levy LS. Parkinson’s disease and pesticide exposure: a new assessment. Natl Acad Sci USA 2005;102:15545–15550. Crit Rev Toxicol 2013;43:515–534. 36. Bueno MJ, Gómez de Cedrón M, Gómez-López G, et al. Combinatorial effects of 4. Ammal Kaidery N, Tarannum S, Thomas B. Epigenetic landscape of Parkinson’s microRNAs to suppress the Myc oncogenic pathway. Blood 2011;117: disease: emerging role in disease mechanisms and therapeutic modalities. Neuro- 6255–6266. therapeutics 2013;10:698–708. 37. Kessler NJ, Van Baak TE, Baker MS, Laritsky E, Coarfa C, Waterland RA. CpG 5. Hancock DB, Martin ER, Mayhew GM, et al. Pesticide exposure and risk of Parkin- methylation differences between neurons and glia are highly conserved from mouse to son’s disease: a family-based case-control study. BMC Neurol 2008;8:6. human. Hum Mol Genet 2016;25:223–232. 6. Wüllner U, Kaut O, deBoni L, Piston D, Schmitt I. DNA methylation in Parkinson’s 38. Duty S, Jenner P. Animal models of Parkinson’s disease: a source of novel treatments disease. J Neurochem 2016;139(suppl 1):108–120. and clues to the cause of the disease. Br J Pharmacol 2011;164:1357–1391. 7. Miranda-Morales E, Meier K, Sandoval-Carrillo A, Salas-Pacheco J, Vazquez-Carde- 39. Martinez-Mir A, Gonzalez-Perez A, Gayan J, et al. Genetic study of neurexin and nas P, Arias-Carrion O. Implications of DNA methylation in Parkinson’s disease. neuroligin genes in alzheimer’s disease. J Alzheimers Dis 2013;35:403–412. Front Mol Neurosci 2017;10:225. 40. Sanchez-Mut JV, Aso E, Heyn H, et al. Promoter hypermethylation of the phos- 8. Jowaed A, Schmitt I, Kaut O, Wüllner U. Methylation regulates alpha-synuclein phatase DUSP22 mediates PKA-dependent TAU phosphorylation and CREB acti- expression and is decreased in Parkinson’s disease patients’ brains. J Neurosci 2010; vation in alzheimer’s disease. Hippocampus 2014;24:363–368. 30:6355–6359. 41. Watson CT, Roussos P, Garg P, et al. Genome-wide DNA methylation profiling in the 9. Matsumoto L, Takuma H, Tamaoka A, et al. CpG demethylation enhances alpha- superior temporal gyrus reveals epigenetic signatures associated with Alzheimer’s synuclein expression and affects the pathogenesis of Parkinson’s disease. PLoS One disease. Genome Med 2016;8:5. 2010;5:e15522. 42. Boks MP, Houtepen LC, Xu Z, et al. Genetic vulnerability to DUSP22 promoter 10. Masliah E, Dumaop W, Galasko D, Desplats P. Distinctive patterns of DNA meth- hypermethylation is involved in the relation between in utero famine exposure and ylation associated with Parkinson disease: identification of concordant epigenetic schizophrenia. NPJ Schizophr 2018;4:16. changes in brain and peripheral blood leukocytes. Epigenetics 2013;8:1030–1038. 43. Inestrosa NC, Arenas E. Emerging roles of Wnts in the adult nervous system. Nat Rev 11. Tan YY, Wu L, Zhao ZB, et al. Methylation of alpha-synuclein and leucine-rich repeat Neurosci 2010;11:77–86. kinase 2 in leukocyte DNA of Parkinson’s disease patients. Parkinsonism Relat Disord 44. Rawal N, Corti O, Sacchetti P, et al. Parkin protects dopaminergic neurons from 2014;20:308–313. excessive Wnt/beta-catenin signaling. Biochem Biophys Res Commun 2009;388: 12. Su X, Chu Y, Kordower JH, et al. PGC-1alpha promoter methylation in Parkinson’s 473–478. disease. PLoS One 2015;10:e0134087. 45. Berwick DC, Harvey K. LRRK2 functions as a Wnt signaling scaffold, bridging cy- 13. Marsh AG, Cottrell MT, Goldman MF. Epigenetic DNA methylation profiling with tosolic proteins and membrane-localized LRP6. Hum Mol Genet 2012;21: MSRE: a quantitative NGS approach using a Parkinson’s disease test case. Front 4966–4979. Genet 2016;7:191. 46. Zhou T, Zu G, Zhang X, et al. Neuroprotective effects of ginsenoside Rg1 through the 14. Zhang L, Deng J, Pan Q, et al. Targeted methylation sequencing reveals dysregulated Wnt/beta-catenin signaling pathway in both in vivo and in vitro models of Parkinson’s Wnt signaling in Parkinson disease. J Genet Genomics 2016;43:587–592. disease. Neuropharmacology 2016;101:480–489. 15. Beecham GW, Dickson DW, Scott WK, et al. PARK10 is a major locus for sporadic 47. Liu Y, Hao S, Yang B, et al. Wnt/beta-catenin signaling plays an essential role in alpha7 neuropathologically confirmed Parkinson disease. Neurology 2015;84:972–980. nicotinic receptor-mediated neuroprotection of dopaminergic neurons in a mouse 16. Sampson TR, Debelius JW, Thron T, et al. Gut microbiota regulate motor deficits and Parkinson’s disease model. Biochem Pharmacol 2017;140:115–123. neuroinflammation in a model of Parkinson’s disease. Cell 2016;167:1469–1480.e12. 48. Prakash N, Brodski C, Naserke T, et al. A Wnt1-regulated genetic network controls 17. Stokholm MG, Danielsen EH, Hamilton-Dutoit SJ, Borghammer P. Pathological the identity and fate of midbrain-dopaminergic progenitors in vivo. Development alpha-synuclein in gastrointestinal tissues from prodromal Parkinson disease patients. 2006;133:89–98. Ann Neurol 2016;79:940–949. 49. Wang Q, Liu Y, Zhou J. Neuroinflammation in Parkinson’s disease and its potential as 18. Uribe C, Segura B, Baggio HC, et al. Patterns of cortical thinning in nondemented therapeutic target. Transl Neurodegener 2015;4:19. Parkinson’s disease patients. Mov Disord 2016;31:699–708. 50. Anderegg A, Lin HP, Chen JA, et al. An Lmx1b-miR135a2 regulatory circuit modu- 19. Assenov Y, Müller F, Lutsik P, Walter J, Lengauer T, Bock C. Comprehensive analysis lates Wnt1/Wnt signaling and determines the size of the midbrain dopaminergic of DNA methylation data with RnBeads. Nat Methods 2014;11:1138–1140. progenitor pool. PLoS Genet 2013;9:e1003973.

10 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG ARTICLE OPEN ACCESS Altered CSF levels of monoamines in hereditary spastic paraparesis 10 A case series

Mattias Andréasson, MD, Kristina Lagerstedt-Robinson, PhD, Kristin Samuelsson, MD, PhD, Correspondence ´ Goran¨ Solders, MD, PhD, Kaj Blennow, MD, PhD, Martin Paucar, MD, PhD,* and Per Svenningsson, MD, PhD* Dr. Andreasson [email protected] Neurol Genet 2019;5:e344. doi:10.1212/NXG.0000000000000344 Abstract Objective To perform a comprehensive clinical characterization and biochemical CSF profile analyses in 2 Swedish families with hereditary spastic paraparesis (HSP) 10 (SPG10) caused by 2 different mutations in the neuronal kinesin heavy chain gene (KIF5A).

Methods Structured clinical assessment, genetic studies, and neuroradiologic and electrophysiological evaluations were performed in 4 patients from 2 families with SPG10. Additional CSF analysis was conducted in 3 patients with regard to levels of neurodegenerative markers and monoamine metabolism.

Results All patients exhibited a complex form of HSP with a mild to moderate concurrent axonal polyneuropathy. The heterozygous missense mutations c.767A>G and c.967C>T in KIF5A were found. Wide intrafamilial phenotype variability was evident in both families. CSF analysis demonstrated a mild elevation of neuroﬁlament light (NFL) chain in the patient with longest disease duration. Unexpectedly, all patients exhibited increased levels of the dopamine metabolite, homovanillic acid, whereas decreased levels of the noradrenergic metabolite, 3-methoxy-4-hydroxyphenylglycol, were found in 2 of 3 patients.

Conclusions We report on CSF abnormalities in SPG10, demonstrating that NFL elevation is not a man- datory finding but may appear after long-standing disease. Impaired transportation of synaptic proteins may be a possible explanation for the increased dopaminergic turnover and noradrenergic deficiency identified. The reasons for these selective abnormalities, unrelated to ob- vious clinical features, remain to be explained. Our findings need further confirmation in larger cohorts of patients harboring KIF5A mutations.

*Equal contribution.

From the Department of Neurology (M.A., K.S., G.S., M.P., P.S.), Karolinska University Hospital; Center for Neurology (M.A., P.S.), Academic Specialist Center; Department of Molecular Medicine and Surgery (K.L.-R.), Karolinska Institutet, and Department of Clinical Genetics, Karolinska University Hospital; Department of Clinical Neurophysiology (G.S.), Karolinska University Hospital, Stockholm; Department of Clinical Neuroscience (K.B.), University of Gothenburg; and Department of Clinical Neuroscience (M.A., K.S., G.S., M.P., P.S.), Karolinska Institutet, Stockholm, Sweden.

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

Copyright © 2019 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 Glossary 5-HIAA = 5-hydroxyindoleacetic acid; ALS = amyotrophic lateral sclerosis; HSP = hereditary spastic paraparesis; HVA = homovanillic acid; KIF5A = neuronal kinesin heavy chain gene; MHPG = 3-methoxy-4-hydroxyphenylglycol; NFL = neuroﬁlament light; PNP = polyneuropathy; SPRS = Spastic Paraplegia Rating Scale.

Hereditary spastic paraparesis (HSP) comprises a large and Friedreich Ataxia Rating Scale part 1: functional staging for growing group of chronic progressive neurodegenerative ataxia, Inventory of Non-Ataxia Signs, Instituto de Pesquisa diseases with varying patterns of inheritance, age at onset, Clinica Evandro Chagas Scale, Scale for the Assessment and and disease severity. These diseases share a common aﬀec- Rating of Ataxia, and Montreal Cognitive Assessment. The tion of the corticospinal tracts. Heterozygous mutations in inclusion of rating scales assessing cerebellar function was the N-terminal motor domain of the neuronal kinesin heavy chosen based on reports of ataxia as a feature in patients with chain gene (KIF5A) are associated with autosomal domi- KIF5A mutations and other familial kinesin motor nant HSP 10 (SPG10) and less commonly with Charcot- proteinopathies.2,11 Standardized examination took place Marie-Tooth type 2, with or without pyramidal signs.1,2 between January and March of 2018. Rarely, mutations in this gene are also associated with cerebellar ataxia or cognitive impairment.2 In addition, a recent Genetic analyses genome-wide association study has identiﬁed variants in the Both families were examined with targeted genetic analyses C-terminal of KIF5A associated with amyotrophic lateral for autosomal dominant HSP (e-Methods, links.lww.com/ sclerosis (ALS).3 NXG/A160).

KIF5A encodes one of 2 heavy chain subunits that together Biochemical analyses with 2 light chain subunits make up a tetrameric kinesin-1 CSF was collected from 3 patients (III:1 in family A and II: protein.1,4,5 This kinesin is crucial for anterograde molecular 1, III:1 in family B) by standard procedures. Patient II:1, in axonal transport by binding to microtubule.4,6 At least 23 family A, declined lumbar puncture. For patient III:1, in mutations in KIF5A with HSP phenotype have been family A, CSF had been collected in 2012 and since then − reported.1,2,5,7,8 stored at 80°C. Levels of the neurodegenerative markers total tau (t-tau), phosphorylated tau (p-tau), β-amyloid β In vitro assays have demonstrated that mutant forms of the 42/40 (A 42/40) ratio, and NFL chain and monoamine kinesin-1 protein impair the transport of cargo along micro- metabolites homovanillic acid (HVA), 5-hydroxyindoleacetic tubule.6 Furthermore, 2 studies on cultured neurons from acid (5-HIAA) and 3-methoxy-4-hydroxyphenylglycol Kif5A knockout mice and mice with mutant Kif5A have (MHPG) were determined (e-Methods, links.lww.com/ demonstrated disturbed axonal bidirectional transport of NXG/A160). mitochondria and neuroﬁlaments, respectively.9,10 Thus, in patients, KIF5A mutations are believed responsible for an axonopathy damaging both the central and peripheral nervous Figure Pedigrees of the 2 Swedish families with SPG10 systems.1,5,7 Here, we hypothesized that patients with SPG10 would demonstrate an elevation of neuroﬁlament light (NFL) chain in CSF.

Methods Standard protocol approvals, registrations, and patient consents All patients have given oral and written consent to this characterization approved by the regional ethical board in Stockholm, Sweden (2016/2503-31/2).

Clinical assessments Patients with a known diagnosis of SPG10, followed at Kar- olinska University Hospital, were eligible for the study. In total, 4 patients from 2 Swedish families (A and B) with Pedigrees of family A and B harboring the c.767A>G (p.Asn256Ser) and heterozygous KIF5A mutations were included (ﬁgure). c.967C>T (p.Arg323Trp) mutations in KIF5A, respectively. Patient I:1 in family A, due to lack of comprehensive medical notes, is considered possibly Patients were assessed with standardized clinical examination symptomatic based on historical description. that included the Spastic Paraplegia Rating Scale (SPRS),

2 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG ) δ Electrophysiology ) δ Motor and sensory nerve conduction studies were compiled from all 4 patients including, at a minimum, unilateral assessment of the median, peroneal, tibial, and sural nerves. Nerve conduction studies were conducted with Natus, Viking EDX (Cephalon A/S; Denmark). Quantitative sensory testing, detecting perception thresholds for cold and heat, was assessed bilaterally in the lateral foot and unilaterally in the including small fibers (C and A PNP hand with Medusa, TSA II (Cephalon A/S; Denmark). PNP including small fibers (A

Neuroimaging Spine MRI NCS and QST Historic data from brain and spinal cord MRI were compiled NAD Mild mixed sensorimotor PNP and reviewed. Brain MRI —

Data availability statement Anonymized data will be shared by request from any qualified opathy; QST = quantitative sensory testing; SARA = Scale for the investigator. nstituto de Pesquisa Clinica Evandro Chagas Scale; mixed = axonal and Results INAS count IPEC SARA The previously reported heterozygous mutations in KIF5A, c.767A>G (p.Asn256Ser) and c.967C>T (p.Arg323Trp) 3.5 64 14 4 6 12 NAD 12.5 NAD NAD Mild axonal NAD sensory PNP Moderate axonal sensorimotor FARS stage 2 6 4 3 NAD NAD Moderate axonal sensorimotor were found in family A and B, respectively.1,5 Briefly, all the 2574 affected patients presented with a variable degree of spastic paraparesis, which is in line with previous descriptions.1,2,5,7,8 Onset was at adult age in all but one case (III:1 in family B), in which the onset was insidious during childhood. All patients had variable degrees of polyneuropathy (PNP). The index case in family B reported neuropathic symptoms many years after onset of paraparesis, and electrodiagnostic testing demonstrated a moderate axonal sensorimotor PNP. The spastic gait, and Babinski sign equivocal Babinski sign equivocal Babinski sign and Babinski sign historical rate of overall clinical progression was slow in both families. We did not find evidence of cerebellar ataxia, psychiatric symptoms, or cognitive impairment. None of the patients were treated with psychotropic medications. Neu- roimaging was normal. A summary of clinical, radiologic, and electrodiagnostic characteristics for both families is shown in s daughter (e-Clinical phenotypes, links.lww.com/NXG/A161). table 1. ’

CSF-NFL was elevated only in the patient with the longest ———— ——————— disease duration. In addition, and more unexpected, we found in all tested patients elevated CSF-HVA levels, and in 2 patients, CSF-MHPG was reduced. The serotonin metabolite (5-HIAA), Aβ42/40 ratio, and t-tau and p-tau levels 4566 c.767A>G 28 c.967C>T 19 26 Hyperreflexia, ankle clonus, 26 Pronounced scissor gait and 32 c.967C>T 30 7 Spastic gait, ankle clonus, and Age at study inclusion (y) Genotype MoCA SPRS Pyramidal signs were normal. Results from CSF analyses are presented in table 2. Detailed case descriptions are included in the supplemental data (e-Clinical phenotypes, links.lww.com/ NXG/A161). and leg cramps and imbalance and paresthesia Presenting symptoms Impaired gait Died at age 90

Discussion a 60 – 3426 Impaired gait Impaired gait 50 Childhood Impaired gait 33 Impaired gait 67 c.767A>G 28 11 Hyperreflexia, spastic gait, Age at onset (y)

There is a need for biomarkers and disease-modifying treat- Electrodiagnostic, neuroradiologic, genetic, and clinical features of 2 families with SPG10 ments for HSP diseases. The reasons for intrafamilial phenotype variability in SPG10 remain to be elucidated.1,7 This Clinical data based on the historical account provided by the patient A III:1 B II:1 A I:1 B III:1 A II:1 Table 1 Patient a Abbreviations: FARS stage =demyelinating Friedreich features Ataxia present; Rating MoCA Scale = part Montreal 1, Cognitive Functional Assessment; Staging NAD for = nothing Ataxia; abnormal INAS detected; count NCS = = Inventory nerve of conduction Non-Ataxia study; Signs; PNP IPEC = = polyneur I variation is similar to what is seen in other forms of familial Assessment and Rating of Ataxia;Results SPRS from = ancillary Spastic testing Paraplegia and Rating clinical Scale. examination. All clinical rating scales have been conducted in the spring of 2018.

Neurology.org/NG Neurology: Genetics | Volume 5, Number 4 | August 2019 3 Table 2 CSF profiles of 3 patients with SPG10

t-Tau (pg/mL) p-Tau (pg/mL) NFL (pg/mL) HVA 5-HIAA MHPG [<300 (18–45 years)] [<60 (20–60 years)] Aβ42/40 [<560 (30–39 years)] (nmol/L) (nmol/L) (nmol/L) Patient [<400 (>45 years)] [<80 (>60 years)] [>0.89] [<1850 (>60 years)] [40–170] [50–170] [65–140]

A I:1 — — —— ———

A II:1 — — —— ———

A III:1 24 26 0.90 517 208a 141 38a

B II:1 320 45 1.07 2,285a 237a 107 87

B III:1 171 32 1.02 432 272a 101 60a

Abbreviations: 5-HIAA = 5-hydroxyindoleacetic acid; HVA = homovanillic acid; MHPG = 3-methoxy-4-hydroxyphenylglycol; NFL = neurofilament light. Biochemical characteristics of 3 patients with regard to markers of neurodegeneration and monoamine metabolism. A significant elevation of NFL in the patient in family B with the longest disease duration (II:1) is demonstrated, possibly reflecting axonal damage. Elevated HVA, reflecting increased dopamine turnover, is seen in all 3 patients. Furthermore, in 2 patients, biochemical signs of decreased noradrenergic turnover are present. a Indicates value outside reference range.

kinesin motor proteinopathies such as SPG30 (KIF1A) and time (III:1 in family A) might have underestimated the values SPG58 (KIF1C); however, these diseases are biallelic and of t-tau and Aβ42/40 ratio. present with a more severe phenotype than SPG10.11,12 Previous reports on the CSF profile in patients with KIF5A An impairment of axonal transport, with resulting length- mutations are rare. Thus, future studies in larger cohorts are dependent axonal degeneration, forms the main theory of needed to better discern whether noradrenergic deficiency the underlying pathophysiology in SPG10.1 CSF levels of and increased dopaminergic neurotransmission are prevalent NFL, an important cytoskeletal component of the axon, findings in SPG10, other kinesin proteinopathies, and/or were mildly elevated in the patient with longest disease patients with ALS with KIF5A mutations. It will also be im- duration. This patient also demonstrated the highest SPRS portant to delineate potential clinical correlates to these score (table 1). Because mutated KIF5A is known to impair changes in monoaminergic neurotransmission. axonal transport of neurofilaments, at least in vitro, we were expecting a more general elevation in our patients.9 How- Acknowledgment ever, NFL elevation was not evident in the 2 younger The authors are grateful to the patients who participated in patients why such elevation cannot be viewed as an obligate the study. Funding was obtained from the ALF program at the finding in SPG10. These results are in contrast with studies Stockholm City Council. P. Svenningsson is a Wallenberg in ALS, where NFL has been proposed as a biomarker.13 Clinical Scholar. M. Paucar obtained funding from the Furthermore, elevated CSF levels of phosphorylated neu- Swedish Society for Medical Research. rofilament heavy chain in patients with HSP (n = 9) compared with controls have been reported in a previous Study funding study.14 It will be interesting to study NFL levels in patients This study was funded by the collaboration agreement be- with ALS harboring KIF5A mutations. tween Karolinska Institutet and Stockholm County Council (ALF). Per Svenningsson is a Wallenberg Clinical Scholar. Assuming that intact axonal transport is important to main- Martin Paucar obtained funding from the Swedish Society for tain synaptic supply of monoamines, we analyzed these Medical Research. metabolites. Surprisingly, CSF-HVA was elevated in all tested patients, of which none had a history of mood disturbance, Disclosure psychotic behaviors, or treatment with psychotropic drugs. M. Andréasson has received a contribution from NEURO Thus, the clinical correlates of this abnormality is unclear. In Sweden (Neuroförbundet) for another study. K. Lagerstedt- addition, 2 patients had decreased levels of the noradrenergic Robinson and K. Samuelsson report no disclosures. G. Sol- metabolite MHPG in CSF. In keeping with the proposed ders has received an unconditional grant from Sanofi/ pathophysiology of an underlying axonopathy in SPG10, Genzyme for another study. K. Blennow has served as a con- deficiency of various neurotransmitters such as noradrenaline sultant or at advisory boards for Alector, Alzheon, CogRx, may either reflect impaired transportation of synaptic proteins Biogen, Lilly, Novartis, and Roche Diagnostics and is a co- or an epiphenomenon. Regardless, the specificity of these founder of Brain Biomarker Solutions in Gothenburg AB, abnormalities remains to be explained. a GU Venture-based platform company at the University of Gothenburg, all unrelated to the work presented in this article. Small sample size is the main limitation of this study. In ad- M. Paucar and P. Svenningsson report no disclosures. Go to dition, we cannot rule out that the prolonged CSF storage Neurology.org/NG for full disclosures.

4 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG Publication history Received by Neurology: Genetics February 6, 2019. Accepted in ﬁnal form Appendix (continued)

May 13, 2019. Name Location Role Contribution

Per Karolinska Author Revision of the Svenningsson, University Hospital manuscript; MD, PhD and Karolinska analysis and Appendix Authors Institute, Stockholm interpretation of data; study Name Location Role Contribution supervision and coordination; and Mattias Karolinska Author Drafting and obtaining funding Andréasson, University Hospital, revision of the MD Karolinska Institutet manuscript; study and Academic concept and design; Specialist Center, and analysis and References Stockholm interpretation of 1. Reid E, Kloos M, Ashley-Koch A, et al. A kinesin heavy chain (KIF5A) mutation in data hereditary spastic paraplegia (SPG10). Am J Hum Genet 2002;71:1189–1194. 2. Liu Y, Laurá M, Hersheson J, et al. Extended phenotypic spectrum of KIF5A muta- Kristina Karolinska Author Interpretation of tions: from spastic paraplegia to axonal neuropathy. Neurology 2014;83:612–619. Lagerstedt- University Hospital genetic tests and 3. Nicolas A, Kenna KP, Renton AE, et al. Genome-wide analyses identify KIF5A as Robinson, PhD and Karolinska revision of the a novel ALS gene. Neuron 2018;97:1268–1283. Institutet, manuscript 4. Hirokawa N, Noda Y, Tanaka Y, Niwa S. Kinesin superfamily motor proteins and Stockholm intracellular transport. Nat Rev Mol Cell Biol 2009;10:682–696. 5. Rinaldi F, Bassi MT, Todeschini A, et al. A novel mutation in motor domain of KIF5A Kristin Karolinska Author Interpretation of associated with an HSP/axonal neuropathy phenotype. J Clin Neuromuscul Dis 2015; Samuelsson, University Hospital, data and revision of 16:153–158. MD, PhD Stockholm the manuscript 6. Ebbing B, Mann K, Starosta A, et al. Effect of spastic paraplegia mutations in KIF5A kinesin on transport activity. Hum Mol Genet 2008;17:1245–1252. Goran¨ Solders, Karolinska Author Interpretation of 7. López E, Casasnovas C, Giménez J, Santamar´ıa R, Terrazas JM, Volpini V. Identifi- MD, PhD University Hospital, neurophysiologic cation of two novel KIF5A mutations in hereditary spastic paraplegia associated with Stockholm studies and clinical mild peripheral neuropathy. J Neurol Sci 2015;358:422–427. data and revision of 8. Collongues N, Depienne C, Boehm N, et al. Novel SPG10 mutation associated with the manuscript dysautonomia, spinal cord atrophy, and skin biopsy abnormality. Eur J Neurol 2013; 20:398–401. Kaj Blennow, Clinical Author CSF analyses; 9. Wang L, Brown A. A hereditary spastic paraplegia mutation in kinesin-1A/KIF5A MD, PhD Neuroscience, interpretation of disrupts neurofilament transport. Mol Neurodegener 2010;5:52. University of data; and revision of 10. Karle KN, Möckel D, Reid E, Schöls L. Axonal transport deficit in a KIF5A(-/-) mouse Gothenburg the manuscript model. Neurogenetics 2012;13:169–179. 11. Dor T, Cinnamon Y, Raymond L, et al. KIF1C mutations in two families with hereditary Martin Paucar, Karolinska Author Revision of the spastic parapares and cerebellar dysfunction. J Med Genet 2014;51:137–142. MD, PhD University Hospital manuscript; study 12. Erlich Y, Edvardson S, Hodges E, et al. Exome sequencing and disease-network and Karolinska concept and design; analysis of a single family implicate a mutation in KIF1A in hereditary spastic para- Institute, Stockholm analysis and paresis. Genome Res 2011;21:658–664. interpretation of 13. Gaiani A, Martinelli I, Bello L, et al. Diagnostic and prognostic biomarkers in data; and study amyotrophic lateral sclerosis: neurofilament light chain levels in definite subtypes of supervision and disease. JAMA Neurol 2007;74:525–532. coordination 14. Zucchi E, Bedin R, Fasano A, et al. Cerebrospinal fluid neurofilaments may discriminate upper motor neuron syndromes: a pilot study. Neurodegener Dis 2018;18:255–261.

Neurology.org/NG Neurology: Genetics | Volume 5, Number 4 | August 2019 5 ARTICLE OPEN ACCESS MAPT p.V363I mutation A rare cause of corticobasal degeneration

Sarah Ahmed, BS, Monica Diez Fairen, MS, Marya S. Sabir, BS, Pau Pastor, MD, PhD, Jinhui Ding, PhD, Correspondence Lourdes Ispierto, MD, Ankur Butala, MD, Christopher M. Morris, PhD, Claudia Schulte, PhD, Dr. Scholz [email protected] Thomas Gasser, MD, Edwin Jabbari, MD, Olga Pletnikova, MD, Huw R. Morris, MD, PhD, Juan Troncoso, MD, Ellen Gelpi, MD, PhD, Alexander Pantelyat, MD, and Sonja W. Scholz, MD, PhD

Neurol Genet 2019;5:e347. doi:10.1212/NXG.0000000000000347 Abstract Objective Patients with corticobasal syndrome (CBS) present with heterogeneous clinical features, including asymmetric parkinsonism, dyspraxia, aphasia, and cognitive impairment; to better understand the genetic etiology of this rare disease, we undertook a genetic analysis of microtubule-associated protein tau (MAPT).

Methods We performed a genetic evaluation of MAPT mutations in 826 neurologically healthy controls and 173 cases with CBS using the Illumina NeuroChip genotyping array.

Results We identiﬁed 2 patients with CBS heterozygous for a rare mutation in MAPT (p.V363I) that is located in the highly conserved microtubule-binding domain. One patient was pathologically conﬁrmed and demonstrated extensive 4-repeat-tau-positive thread pathology, achromatic neurons, and astrocytic plaques consistent with corticobasal degeneration (CBD).

Conclusions We report 2 CBS cases carrying the rare p.V363I MAPT mutation, one of which was pathologically conﬁrmed as CBD. Our ﬁndings support the notion that this rare coding change is pathogenic.

From the Neurodegenerative Disease Research Unit (S.A., M.S.S., S.W.S.), National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, MD; Movement Disorders and Memory Unit (M.D.F., P.P.), Department of Neurology, University Hospital Mutua de Terrassa, and Fundacio´ per la Recerca Biom`edicaiSocialMutua´ Terrassa, Barcelona, Spain; Laboratory of Neurogenetics (J.D.), National Institutes on Aging, National Institutes of Health, Bethesda, MD; Neurology Service (L.I.), Hospital Universitari Germans Trias, Pujol, Badalona, Spain; Department of Neurology (A.B., A.P., S.W.S.), Johns Hopkins University Medical Center, Baltimore, MD; Newcastle Institute for Ageing (C.M.M.), Campus for Ageing and Vitality, Newcastle University, Newcastle upon Tyne, United Kingdom; Department of Neurodegenerative Diseases (C.S., T.G.), Center of Neurology, Hertie- Institute for Clinical Brain Research, University of Tuebingen, and German Center for Neurodegenerative Diseases, Germany; Department of Molecular and Clinical Neuroscience (E.J., H.R.M.), Institute of Neurology, University College London, United Kingdom; Department of Pathology (Neuropathology) (O.P., J.T.), Johns Hopkins University Medical Center, Baltimore, MD; Department of Clinical Neurosciences (H.R.M.), Royal Free Campus UCL, Institute of Neurology, London, United Kingdom; Neurological Tissue Bank (E.G.), University of Barcelona-Hospital Clinic, IDIBAPS, Barcelona, Spain; and Institute of Neurology (E.G.), Medical University of Vienna, Austria.

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 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 © 2019 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 Glossary CBD = corticobasal degeneration; CBS = corticobasal syndrome; FTD = frontotemporal dementia; MAPT = microtubule- associated protein tau; PGRN = progranulin; PSP = progressive supranuclear palsy.

Corticobasal syndrome (CBS) is a rare neurologic disease that dementia (FTD) spectrum disorders (n = 772 cases with presents with heterogeneous motor symptoms and cognitive progressive supranuclear palsy [PSP], n = 173 patients with impairment.1 A high misdiagnosis rate due to clinical het- CBS/CBD, n = 41 patients with FTD; sample characteristics erogeneity limits eﬀorts to extend disease-modifying therapy are summarized in table e-2 [links.lww.com/NXG/A162]; trials to this patient population. Improving the diagnostic source of samples and number of samples per disease are de- accuracy of complex neurodegenerative syndromes is an im- scribed in table e-3). Case 2 was clinically diagnosed with hemi- portant, yet unmet need in the research community. parkinsonism, primary progressive aphasia, and probable CBS.

Although understanding of the genetic underpinnings of CBS Standard protocol approvals, registrations, is limited, rare mutations in the microtubule-associated pro- and patient consents tein tau (MAPT) gene are implicated as a cause of CBS and The study was approved by the respective institutional review – related tauopathy spectrum disorders.2 6 One of these MAPT boards. Written informed consent for research participation mutations is the variant p.V363I (rs63750869; c.1087G>A: was obtained from all participants. NM_005910.5), located in the MAPT microtubule-binding domain. Previously described in a small number of patients Genetic analysis and validation with clinical tauopathy phenotypes (table 1), the mutation is For each participant, DNA was extracted from blood or brain present at a very low frequency in population databases and is tissue using standard methods and followed by genotyping on hypothesized to be a disease-causing mutation with decreased the NeuroChip platform (Illumina, San Diego, CA). This – penetrance rather than a rare polymorphism.6 10 The rare affordable genotyping array contains a tagging single nucle- nature of the mutation makes it difficult to demonstrate disease otide polymorphism backbone combined with high-yield segregation, and in silico prediction algorithms of this mutation custom content that allows for rapid screening of ;180,000 are inconclusive (table e-1, links.lww.com/NXG/A162). mutations and risk variants previously implicated in neurologic diseases, including the MAPT p.V363I variant. The We describe 2 CBS cases who were found to carry the rare detailed contents of this versatile genotyping platform have p.V363I MAPT mutation. In addition, we summarize the been described elsewhere.12 The MAPT p.V363I mutation clinicopathologic features of previously reported cases with was only present in 2 patients (henceforth referred to as case 1 a coding mutation at the MAPT p.V363 residue. One of our and case 2), and we validated the mutation via direct CBS cases had postmortem confirmation, which found Sanger sequencing using the following primers: forward abundant four-repeat tau accumulations consistent with cor- 59-GTGGCCAGGTGGAAGTAAAA, reverse 59-ACATC- ticobasal degeneration (CBD). As a pathologically confirmed CAGCCAGTCAACACA. To rule out other possible patho- case with this rare missense mutation, this case provides genic mutations in these 2 patients, we assessed the NeuroChip supporting evidence for the pathogenic nature of the p.V363I data for damaging progranulin (PGRN)genemutations.Wealso MAPT mutation. performed repeat-primed PCR screening of the C9orf72 repeat using methods described elsewhere.13 APOE genotypes were determined by extracting rs7412 and rs429358 as previously Methods described.12 MAPT haplotype status was determined by imputation of the polymorphism rs1052553 (R2 = 0.99494), with the Study population “A” allele determining the H1 haplotype and the “G” allele Case 1 is a 73-year-old, right-handed, white woman who segregating with the H2 haplotype.14 presented to the NIH Clinical Center in Bethesda, MD, for participation in genetic research. She was diagnosed with Bioinformatic analysis probable CBS based on the consensus criteria.11 A commer- To better understand the effects of the p.V363I variant, cial genetic panel (Invitae, San Francisco, CA) that included a systematic literature review was conducted and summarized screening of the genes CHCHD10, DCTN, FUS, GRN, in table 1. In silico predictive tools (SIFT, PolyPhen2, TARDBP, VCP, UBQLN2, TBK1, PSEN1, PSEN2, APP, and FATHMM-XF, M-CAP, MutationTaster, CADD, ClinVar, MAPT had previously identified that she was a carrier of the and ClinPred) were applied to classify the MAPT p.V363I – MAPT p.V363I variant. Case 2 was identified by querying mutation.15 21 Sequence conservation analyses were per- a research database for the presence of the MAPT p.V363I formed in T-Coffee.22 A previously described, cryo-electron variant. This database contains genotype information on microscopy structure of the tau protofibril was used for European-ancestry individuals, including 826 neurologically 3-dimensional protein modeling (figure 1).23 Allele frequency healthy controls and 961 patients with frontotemporal differences between CBS cases and neurologically healthy

Table 1 Clinicopathologic features of patients with a mutation at the highly conserved p.V363 residue of MAPT

Clinical features Genetics

No. Clinical diagnosis Sex AAO AAD FH Neuroimaging finding(s) Mutation Haplotype Pathology Country Reference

1 CBS F 70 NA + MRI: bilateral parietal atrophy p.V363I H1/H1 NA United This report States

2 CBS and PPA F Late 62 − NA p.V363I H1/H1 CBD Spain This report 50s

3 PPA (nonfluent variant) F 69 NA + SPECT: bilateral Sylvian hypoperfusion p.V363I H1/H1 NA Spain Munoz et al.9

4 FTD (behavioral variant) F 53 61 + MRI: bilateral frontotemporal atrophy p.V363I H1/H1 NA Italy Anfossi et al.8

5 PPA (semantic variant) F 46 NA − MRI: asymmetric temporopolar atrophy p.V363I NA NA Italy Bessi et al.6

6 FTD and PPA (nonfluent F 55 NA NA SPECT: bilateral Sylvian hypoperfusion p.V363I NA NA Italy Rossi variant) et al.7

7 PCA F 54 NA NA NA p.V363I NA NA Italy Rossi et al.7

8 FTD, PPA (nonfluent F 55 NA NA MRI: mild left frontal atrophy p.V363I H1/H1 NA Italy Rossi variant), and CBS SPECT: left frontotemporal predominant hypoperfusion et al.10 FDG-PET: left parietal hypometabolism

9 PCA F 51 N/A + MRI: slight, asymmetric atrophy of posterior temporoparietal and occipital p.V363I H1/H1 N/A Italy Rossi lobes; white matter abnormalities et al.10 FDG-PET: bilateral posterior temporo-occipital and right posterior frontoparietal hypometabolism

10 PSP M 53 NA + MRI: midbrain atrophy p.V363A H1/H1 NA Italy Rossi DAT scan: bilateral dopaminergic denervation et al.10

Abbreviations: AAD = age at death; AAO = age at onset; CBS = corticobasal syndrome; CBD = corticobasal degeneration; DAT scan = dopamine transporter scan; FDG = fluorodeoxyglucose PET; FH = family history; +/− = present/ absent; FTD = frontotemporal dementia; MAPT = microtubule-associated protein tau; NA = not available or not applicable; PCA = posterior cortical atrophy; PPA = primary progressive aphasia; PSP = progressive supranuclear palsy; SPECT = single-photon emission computed tomography. 3 Figure 1 This schematic representation illustrates the location of the MAPT p.V363I mutation

A Cartesian genotype plot and electro- pherograms of this mutation are shown for both CBS cases compared with a control (A). The mutation is localized within the highly conserved microtubule-binding domain (B). (C) Position of the p.V363I mutation (arrow heads) within the microtubule-binding domain (highlighted in purple) relative to the 3-dimensional reconstruction of the tau protofibril. MAPT = microtubule- associated protein tau.

controls were determined using a Fisher exact test with Results a significance threshold of 0.05. Genetic characteristics Neuropathology In a cohort of 173 CBS cases, we identified 2 patients who The brain of case 2 was pathologically evaluated at the Neuro- were heterozygous for the rare p.V363I (c.1087G>A: NM_ logical Tissue Bank of the IDIBPAS Biobank in Barcelona, Spain, 005910.5) mutation located in the highly conserved microtubule- after obtaining written informed consent from the patient’srel- binding domain of MAPT (Fisher exact test comparing CBS cases atives for use of tissue for diagnostic and research purposes. with neurologically healthy controls p = 0.0299). Both patients Hematoxylin and eosin staining was performed after standard were homozygous for the H1 MAPT haplotype and carried no formalin fixation and paraffin block sectioning of multiple cor- pathogenic mutations in PGRN or C9orf72.Thepatients’ APOE tical and subcortical brain areas. Immunohistochemistry was genotypes were e3/e3. The MAPT p.V363I mutation was absent performed using phospho-tau (Ser202 and Thr205) monoclonal in ;1,800 additional samples, including 826 neurologically healthy antibodies (AT8; 1:2000; Thermo Scientific, Rockford, IL) and controls and 984 cases with diverse frontotemporal degeneration anti-4R-tau (RD4) antibodies. In addition, selected areas were spectrum disorders. Bioinformatic predictions demonstrated that stained for ßA4-amyloid (6F/3D 1:400; Dako, Glostrup, Den- SIFT, PolyPhen2, MutationTaster, and ClinPred categorized the mark), α-synuclein (KM51 2:200; Novocastra, Newcastle upon variant as tolerated and benign, whereas ClinVar, FATHMM-XF, Tyne, UK), and TDP43 protein (2E2-E3 1:500; Abnova, Taipei, M-CAP, and CADD predictions suggested a likely pathogenic Taiwan) for identification of concomitant pathologies. mutation (table e-1, links.lww.com/NXG/A162).

Data availability Clinicopathologic features Deidentiﬁed data are available upon request from qualiﬁed Case 1 is a 73-year-old, right-handed, white woman with investigators. a medical history of hypertension, coronary artery disease, an

4 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG old segmental left parietal stroke at age 54 years that resolved throughout. Primitive reflexes, including grasp and palmo- without residual neurologic deficits, and major depressive mental reflexes, were present. She was unable to stand without disorder. She was diagnosed with CBS at age 70 years after assistance and would spontaneously fall without support. MRI developing progressive right-sided impairment of her dex- of the brain demonstrated bilateral parietal atrophy with terity, slowed gait, and imbalance resulting in backward falls. A proportional, ex vacuo dilatation of the lateral ventricles. Her levodopa trial up to a maximum dose of 450 mg daily yielded family history was notable for parkinsonism in her father (age no benefits. Over the course of 3 years, she gradually de- at onset ;65 years). No DNA was available from her father to veloped dysarthria, severe gait dysfunction rendering her test for segregation. The patient is alive after a 3-year disease wheelchair-bound, asymmetric parkinsonism, hand dystonia, duration. apraxia, impaired word retrieval, and executive dysfunction. Her neurologic examination demonstrated bradyphrenia with Case 2 was a white woman who presented in her late 50s with a tendency to perseverate. She had severe ideomotor apraxia primary progressive aphasia, left-sided parkinsonism, and that was more prominent in her dominant hand. She was CBS. The disease progressed to complete anarthria and severe neglecting her right-sided space. Her speech was moderately dysphagia. She died at age 62 years. Clinical data on this case dysarthric. Cranial nerve examination showed slowed, hypo- were limited. She had no known family history of dementia. metric saccades (vertical more than horizontal), severe axial The patient’s neuropathologic findings were notable for and right-sided rigidity with only mild rigidity on the left, widespread, 4-repeat-tau-positive inclusions in cortical and bradykinesia, and dystonia with high-frequency/low- subcortical regions, including neurons and glial cells, consis- amplitude tremor in her right hand. She had agraphesthesia tent with CBD (figure 2). Frequent achromatic neurons were and astereognosis in her right hand. Reflexes were brisk detected in frontal, parietal, and cingular cortices. These were

Figure 2 These images showcase the pertinent neuropathologic findings of case 2

Hematoxylin and eosin staining shows superficial spongiosis in the postcentral region (A), a large achromatic or ballooned cell (highlighted by asterisk in B), and prominent nigral degeneration with severe neuronal loss and abundant extracellular neuromelanin pigment (C). (D–I) Abnormal pTau (AT8) and 4-repeat- tau-positive protein deposition on immunohistochemistry. Notable abnormal histopathologic findings included astrocytic plaques (D), frequent pretangles with some focal cytoplasmic condensations (E), tangles, pretangles, and abundant threads in the substantia nigra (F), very abundant threads (arrow heads) and coiled bodies (arrow) in the white matter (G and H), and abundant threads and pretangles in the striatum (I), overall consistent with the neuropathologic findings observed in corticobasal degeneration. Magnification scale bars are indicated in the bottom right corner of each panel.

Neurology.org/NG Neurology: Genetics | Volume 5, Number 4 | August 2019 5 associated with focal superficial spongiosis, diffuse neuronal Because of the lack of familial genetic data, we were not able to loss, astrogliosis, and microglial activation in cortical areas, test for disease segregation, and this has not yet been reported including the motor cortex, and diffuse gliosis of the un- by other investigators. derlying white matter. Prominent neuronal loss was noted in the globus pallidus and in the substantia nigra. Immunohis- We present a pathologically confirmed patient with a p.V363I tochemistry revealed astrocytic plaques, abundant pretangles, MAPT mutation. The neuropathologic findings of this case ballooned neurons, and neuropil threads in the cortex, were consistent with CBD. An additional p.V363I carrier was abundant threads and pretangles in her basal ganglia, and identified with a CBS phenotype. This mutation was absent in prominent white matter pathology with widespread threads neurologically healthy controls. Considering previous reports and coiled bodies involving also the brainstem. Remarkably, on mutation carriers with information about sequence con- there was also prominent involvement of the hippocampus, servation, functional studies, and pathologic confirmation, we including the granule cells of the dentate gyrus, without grain nominate the MAPT p.V363I change as a likely disease- pathology. Co-comitant pathologies included a moderate causing mutation. Identifying additional cases with this mu- amount of diffuse ßA4-amyloid deposits and few cored pla- tation will be important to understand the natural history and ques in cortical areas, as well as few neuronal and glial cyto- penetrance of this familial disease. plasmic TDP43 protein inclusions in the globus pallidus, without frontal, temporal, or hippocampal involvement. No Acknowledgment α-synuclein aggregates were identified. The authors thank all the subjects who donated their time and biological samples to be a part of this study. They thank the NIH NeuroBrainBank and the IDIBAPS Brain Bank, Discussion Barcelona, Spain, for contributing brain tissue samples. This study used tissue samples and data provided by the Michigan We describe 2 CBS cases carrying the rare p.V363I MAPT Brain Bank, the Michigan Alzheimer’s Disease Center (5P30 mutation located in the conserved microtubule-binding do- AG053760), and the Protein Folding Disorders Program. main. One of the 2 cases was pathologically confirmed, This study used samples from the NINDS Repository at demonstrating widespread, 4-repeat-tau-positive neuronal Coriell (catalog.coriell.org) and clinical data. The authors are and glial pathology consistent with CBD. This report grateful to the Banner Sun Health Research Institute Brain describes the pathology present in a p.V363I MAPT mutation and Body Donation Program of Sun City, Arizona, for the carrier providing further support for the notion that this provision of human brain tissue. The Brain and Body variant is likely disease causing. To date, this coding mutation Donation Program is supported by the NINDS (U24 has been described in 7 neurodegenerative disease cases with NS072026 National Brain and Tissue Resource for Parkin- heterogeneous presentations, including FTD, primary pro- son’s Disease and Related Disorders), the National Institute – gressive aphasia, and posterior cortical atrophy (table 1).6,8 10 on Aging (P30 AG19610 Arizona Alzheimer’s Disease Core Another mutation at the same residue (p.V363A) has been Center), the Arizona Department of Health Services described in a single case with clinically diagnosed PSP.10 Of (contract 211002, Arizona Alzheimer’s Research Center), interest, all cases were female, had e3/e3 APOE genotypes, the Arizona Biomedical Research Commission (contracts and were homozygous for the MAPT H1 haplotype. The 4001, 0011, 05-0901, and 1001 to the Arizona Parkinson’s average age at onset was 57 years, ranging from 46 to 70 years. Disease Consortium), and the Michael J. Fox Foundation for The 2 CBS cases presented here extend the disease onset. Parkinson’s Research. The authors thank the Kathleen Price This wide age spectrum is consistent with the pattern seen in Bryan Brain Bank at Duke University Medical Center, the tauopathies associated with MAPT mutations and could in- Harvard Brain Bank, and the Georgetown University Brain dicate a decreased, age-related penetrance.24 Bank for the provision of tissue and DNA samples. Tissue for this study was provided by the Newcastle Brain Tissue Among the cases, the initial disease manifestations were quite Resource, which is funded in part by a grant from the UK varied, including gait disturbances, memory deficits, and Medical Research Council (G0400074), by NIHR Newcastle personality changes. This heterogeneity is not unusual for Biomedical Research Centre and Unit awarded to the patients with MAPT mutations.25 The p.V363I mutation is Newcastle upon Tyne NHS Foundation Trust and Newcastle present in 3 of 62,784 people in the NHLBI TopMed Bravo University, and as part of the Brains for Dementia Research database (bravo.sph.umich.edu/freeze5/hg38/; allele fre- Programme jointly funded by Alzheimer’sResearchUK quency: 0.0000239; date accessed: October 14, 2018) and in and Alzheimer’s Society. The authors thank members of the 2 of 60,702 individuals in ExAC (allele frequency: 0.0000167, North American Brain Expression Consortium for pro- data accessed: October 14, 2018).26 The very rare presence viding DNA samples on neurologically healthy controls. within population databases might be explained by in- Tissue samples for genotyping were provided by the Johns complete penetrance and late disease onset. In addition, Hopkins Morris K. Udall Center of Excellence for limited in vitro analyses in 1 case demonstrated that this Parkinson’s Disease Research (NIH P50 NS38377) and mutation leads to an increased propensity for microtubule the Johns Hopkins Alzheimer Disease Research Center polymerization and the formation of tau protein oligomers.10 (NIH P50 AG05146).

6 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG Study funding This work was supported in part by the Intramural Research Appendix (continued)

Programs of the NINDS and the National Institute on Aging Name Location Role Contributions (NIA) (project numbers Z01-AG000949 and 1ZIA NS003154). This study was partially supported by a CurePSP Huw R. University College Author Clinical/pathologic Morris, MD, London, and Royal characterization and research grant (Cure PSP grant no.: 515-14; 2013-2015) to PhD Free Campus critical review PP. EG received support from a grant from the Marat´ode Juan C. Johns Hopkins Author Clinical/pathologic TV3 (grant no. 20141610). Troncoso, University Medical characterization and MD Center critical review

Disclosure Ellen Gelpi, University of Author Neuropathologic Disclosures available: Neurology.org/NG. MD, PhD Barcelona-Hospital assessment and Clinic critical review

Publication history Alexander Johns Hopkins Author Conceptualization Neurology: Genetics ﬁ Pantelyat, University Medical and design; clinical/ Received by March 10, 2019. Accepted in nal form MD Center pathologic May 15, 2019. characterization; and critical review

Sonja W. National Institutes of Author Drafting of the Scholz, MD, Health and Johns manuscript; Appendix Authors PhD Hopkins University conceptualization Medical Center and design; clinical/ Name Location Role Contributions pathologic characterization; and Sarah National Institutes of Author Drafting of the critical review Ahmed, BS Health manuscript; genetic assessment and analysis; and critical review References 1. Kouri N, Murray ME, Hassan A, et al. Neuropathological features of corticobasal Monica Diez University Hospital Author Genetic assessment degeneration presenting as corticobasal syndrome or Richardson syndrome. Brain Fairen, MS Mutua de Terrassa, and critical review 2011;134:3264–3275. and Fundacio´ per la 2. Poorkaj P, Bird TD, Wijsman E, et al. Tau is a candidate gene for chromosome 17 Recerca Biomèdica i frontotemporal dementia. Ann Neurol 1998;43:815–825. Social Mutua´ Terrassa 3. Hutton M, Lendon CL, Rizzu P, et al. Association of missense and 5’-splice-site mutations in tau with the inherited dementia FTDP-17. Nature 1998;393: Marya S. National Institutes of Author Genetic assessment 702–705. Sabir, BS Health and critical review 4. Spillantini MG, Murrell JR, Goedert M, Farlow MR, Klug A, Ghetti B. Mutation in the tau gene in familial multiple system tauopathy with presenile dementia. Proc Natl Pau Pastor, University Hospital Author Clinical/pathologic Acad Sci U S A 1998;95:7737–7741. MD, PhD Mutua de Terrassa, characterization and 5. Pickering-Brown SM, Rollinson S, Du Plessis D, et al. Frequency and clinical char- and Fundacio´ per la critical review acteristics of progranulin mutation carriers in the Manchester frontotemporal lobar Recerca Biomèdica i degeneration cohort: comparison with patients with MAPT and no known mutations. Social Mutua´ Terrassa Brain 2008;131:721–731. 6. Bessi V, Bagnoli S, Nacmias B, Tedde A, Sorbi S, Bracco L. Semantic dementia Jinhui Ding, National Institutes of Author Genetic assessment associated with mutation V363I in the tau gene. J Neurol Sci 2010;296: PhD Health and critical review 112–114. 7. Rossi G, Conconi D, Panzeri E, et al. Mutations in MAPT gene cause chromosome Lourdes Hospital Universitari Author Clinical/pathologic instability and introduce copy number variations widely in the genome. J Alzheimers Ispierto, MD Germans Trias characterization and Dis 2013;33:969–982. critical review 8. Anfossi M, Bernardi L, Gallo M, et al. MAPT V363I variation in a sporadic case of frontotemporal dementia: variable penetrant mutation or rare polymorphism?. Alz- Ankur Johns Hopkins Author Clinical/pathologic heimer Dis Assoc Disord 2011;25:96–99. Butala, MD University Medical characterization and 9. Munoz DG, Ros R, Fatas M, Bermejo F, de Yebenes JG. Progressive nonfluent aphasia Center critical review associated with a new mutation V363I in tau gene. Am J Alzheimers Dis Other Demen 2007;22:294–299. Christopher Newcastle University Author Clinical/pathologic 10. Rossi G, Bastone A, Piccoli E, et al. Different mutations at V363 MAPT codon are M. Morris, characterization and associated with atypical clinical phenotypes and show unusual structural and func- PhD critical review tional features. Neurobiol Aging 2014;35:408–417. 11. Armstrong MJ, Litvan I, Lang AE, et al. Criteria for the diagnosis of corticobasal Claudia University of Author Clinical/pathologic degeneration. Neurology 2013;80:496–503. Schulte, PhD Tuebingen characterization and 12. Blauwendraat C, Faghri F, Pihlstrom L, et al. NeuroChip, an updated version of the critical review NeuroX genotyping platform to rapidly screen for variants associated with neurological diseases. Neurobiol Aging 2017;57:247 e9–247 e13. Thomas University of Author Clinical/pathologic 13. Renton AE, Majounie E, Waite A, et al. A hexanucleotide repeat expansion in Gasser, MD Tuebingen characterization and C9ORF72 is the cause of chromosome 9p21-linked ALS-FTD. Neuron 2011;72: critical review 257–268. 14. Poorkaj P, Grossman M, Steinbart E, et al. Frequency of tau gene mutations in familial Edwin University College Author Clinical/pathologic and sporadic cases of non-Alzheimer dementia. Arch Neurol 2001;58:383–387. Jabbari, MD London characterization and 15. Ng PC, Henikoff S. SIFT: predicting amino acid changes that affect protein function. critical review Nucleic Acids Res 2003;31:3812–3814. 16. Jagadeesh KA, Wenger AM, Berger MJ, et al. M-CAP eliminates a majority of variants Olga Johns Hopkins Author Clinical/pathologic of uncertain significance in clinical exomes at high sensitivity. Nat Genet 2016;48: Pletnikova, University Medical characterization and 1581–1586. MD Center critical review 17. Adzhubei IA, Schmidt S, Peshkin L, et al. A method and server for predicting damaging missense mutations. Nat Methods 2010;7:248–249.

Neurology.org/NG Neurology: Genetics | Volume 5, Number 4 | August 2019 7 18. Landrum MJ, Lee JM, Benson M, et al. ClinVar: public archive of inter- 22. Rausch T, Emde AK, Weese D, Doring A, Notredame C, Reinert K. Segment-based pretations of clinically relevant variants. Nucleic Acids Res 2016;44: multiple sequence alignment. Bioinformatics 2008;24:i187–192. D862–D868. 23. Fitzpatrick AWP, Falcon B, He S, et al. Cryo-EM structures of tau ﬁlaments from 19. Rogers MF, Shihab HA, Mort M, Cooper DN, Gaunt TR, Campbell C. FATHMM- Alzheimer’s disease. Nature 2017;547:185–190. XF: accurate prediction of pathogenic point mutations via extended features. Bio- 24. Irwin DJ. Tauopathies as clinicopathological entities. Parkinsonism Relat Disord informatics 2018;34:511–513. 2016;22(suppl 1):S29–S33. 20. Kircher M, Witten DM, Jain P, O’Roak BJ, Cooper GM, Shendure J. A general 25. Ghetti B, Oblak AL, Boeve BF, Johnson KA, Dickerson BC, Goedert M. Invited framework for estimating the relative pathogenicity of human genetic variants. Nat review: frontotemporal dementia caused by microtubule-associated protein tau gene Genet 2014;46:310–315. (MAPT) mutations: a chameleon for neuropathology and neuroimaging. Neuro- 21. Alirezaie N, Kernohan KD, Hartley T, Majewski J, Hocking TD. ClinPred: prediction pathol Appl Neurobiol 2015;41:24–46. tool to identify disease-relevant nonsynonymous single-nucleotide variants. Am J 26. Lek M, Karczewski KJ, Minikel EV, et al. Analysis of protein-coding genetic variation Hum Genet 2018;103:474–483. in 60,706 humans. Nature 2016;536:285–291.

8 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG ARTICLE OPEN ACCESS Novel mutation in HTRA1 in a family with diﬀuse white matter lesions and inﬂammatory features

Amin Ziaei, MD,* Xiaohong Xu, MD, PhD,* Leila Dehghani, MSc, Carine Bonnard, PhD, Andreas Zellner, PhD, Correspondence Alvin Yu Jin Ng, PhD, Sumanty Tohari, BSc, Byrappa Venkatesh, PhD, Christof Haffner, PhD, Dr. Pouladi [email protected] Bruno Reversade, PhD, Vahid Shaygannejad, MD, and Mahmoud A. Pouladi, PhD

Neurol Genet 2019;5:e345. doi:10.1212/NXG.0000000000000345 Abstract Objective To investigate the possible involvement of germline mutations in a neurologic condition involving diﬀuse white matter lesions.

Methods The patients were 3 siblings born to healthy parents. We performed homozygosity mapping, whole-exome sequencing, site-directed mutagenesis, and immunoblotting.

Results All 3 patients showed clinical manifestations of ataxia, behavioral and mood changes, premature hair loss, memory loss, and lower back pain. In addition, they presented with inflammatory-like features and recurrent rhinitis. MRI showed abnormal diffuse demyelination lesions in the brain and myelitis in the spinal cord. We identified an insertion in high-temperature requirement A (HTRA1), which showed complete segregation in the pedigree. Functional analysis showed the mutation to affect stability and secretion of truncated protein.

Conclusions The patients’ clinical manifestations are consistent with cerebral autosomal recessive arteriopathy with subcortical infarcts and leukoencephalopathy (CARASIL; OMIM #600142), which is known to be caused by HTRA1 mutations. Because some aspects of the clinical presentation deviate from those reported for CARASIL, our study expands the spectrum of clinical consequences of loss-of-function mutations in HTRA1.

*These authors contributed equally to the manuscript.

From the Translational Laboratory in Genetic Medicine (TLGM) (A. Ziaei, X.X., M.A.P.), Agency for Science, Technology and Research (A*STAR), 8A Biomedical Grove, Immunos, Level 5; Department of Medicine (A. Ziaei, M.A.P.), National University of Singapore; Department of Neurology and Stroke Center (X.X.), the First Affiliated Hospital, Jinan University; Clinical Neuroscience Institute of Jinan University (X.X.), Guangzhou, Guangdong, China; Department of Tissue Engineering and Regenerative Medicine (L.D.), School of Advanced Technologies in Medicine, Shahid Beheshti University of Medical Sciences, Tehran, Iran; Institute of Medical Biology (IMB) (C.B., B.R.), A*STAR, 8A Biomedical Grove, Immunos, Level 5, Singapore; Institute for Stroke and Dementia Research (A. Zellner, C.H.), Klinikum der Universit¨at Munchen,¨ Ludwig Maximilians University, Munich, Germany; Comparative Genomics Laboratory (A.Y.J.N., S.T., B.V.), Institute of Molecular and Cell Biology, A*STAR, Biopolis; Department of Paediatrics (B.V.), National University of Singapore; Department of Neurology (A. Ziaei, V.S.), Isfahan Neurosciences Research Centre, Faculty of Medicine, Isfahan University of Medical Sciences, Iran; and Department of Physiology (M.A.P.), National University of Singapore.

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

Copyright © 2019 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 Glossary AMD = age-related macular degeneration; CARASIL = cerebral autosomal recessive arteriopathy with subcortical infarcts and leukoencephalopathy; cHTZ = compound heterozygous; ECM = extracellular matrix; HEK = human embryonic kidney; HMZ = homozygous; HTRA = high-temperature requirement A; IBD = identity by descent; MAF = minor allele frequency; SNP = single nucleotide polymorphism; TMAP = Torrent Mapping Alignment Program; WT = wild type.

The high-temperature requirement A (HTRA) family of Standard protocol approvals, registrations, proteins are serine proteases, which were initially identified and patient consents in Escherichia coli as heat shock proteins and appear to play The entire family was recruited under an Isfahan University of an important role in protein quality control.1 Four mam- Medical Sciences ethics-approved research protocol with in- malian HTRA proteins, HTRA1-4, have been identified to formed consent. date, and all contain a highly conserved trypsin-like protease domain.2 Human HTRA1 is located on chromosome 10q26 Whole-exome sequencing and was identified originally as being repressed in SV40- Genomic DNA from proband V.2 was collected and used transformed fibroblasts.3 The majority of this ubiquitously for whole-exome sequencing. DNA was extracted using expressedenzymeissecretedintotheextracellularspace, a QIAamp DNA extraction kit (Qiagen). Using the Ion with only approximately a fifth remaining in the cytoplasm TargetSeq Exome and Custom Enrichment Kit (Thermo where it can associate with microtubules.4 Almost all iden- Fisher), 1 μg of gDNA was processed and captured tified substrates of HTRA1 such as collagen II, clusterin, according to the manufacturer’s instructions. Captured fibronectin, biglycan, fibromodulin, vitronectin, decorin, DNA molecules were sequenced on an Ion Proton in- aggrecan,5 and latent transforming growth factor beta strument (Thermo Fisher) using the ION PI Chip Kit binding protein 16 are extracellular matrix (ECM) proteins, (Thermo Fisher). Sequence reads were aligned to the suggesting a critical role for HTRA1 in ECM homeostasis.6 GRCh37 human reference genome assembly by the Tor- rent Mapping Alignment Program, which is a sequence Altered expression or activity of HTRA1 has been linked to alignment software program optimized especially for Ion a number of diseases including age-related macular de- Torrent data. A coverage of 82.14% was achieved across the generation,7 cancer,4,8,9 and Alzheimer disease.10 More exome, with 97.45% of the targeted sequences covered at ≥ recently, inactivating mutations in HTRA1 have been 20×. The variants were then called using the Torrent shown to cause cerebral autosomal recessive arteriopathy Variant Caller plugin from the Torrent Suite (v4.2.1) and with subcortical infarcts and leukoencephalopathy were then annotated using the “annotate single sample (CARASIL; OMIM #600142),11 a hereditary white matter variants” workflow. The annotation items for the called disease characterized by subcortical infarcts with non- variants included the associated gene name, variant loca- hypertensive cerebral small vessel arteriopathy, spondy- tion, quality score, coverage, predicted functional con- losis, and alopecia, and typically manifesting in the third sequences, protein position and amino acid changes, scale – decade of life.12 14 Here, we report the discovery and invariant feature transform (SIFT), PolyPhen2, Mendelian functional characterization of a novel mutation in HTRA1 Clinically Applicable Pathogenicity (M-CAP) and Gran- in 3 siblings with neurologic symptoms and diffuse white tham prediction scores, PhyloP conservation scores, and matter lesions consistent with CARASIL and some atypical 5000 genomes Minor Allele Frequencies. To exclude features such as spinal cord myelitis and rhinitis only in common single nucleotide polymorphisms (SNPs), the affected members. annotated variants were filtered using the ClinVar database (ftp://ftp.ncbi.nlm.nih.gov/pub/clinvar/vcf_GRCh37/) Methods Research study participants Table 1 Summary of homozygous loci (>2 cm) shared Board-certified clinical neurologist evaluated family members between affected probands (V.2, V.3, and V.4) (IV.1, IV.2, and V.1-V.6) clinically. Clinical evaluation in- and not present in healthy family members (IV.1, IV.2, V.1, V.5, and V.6) volved history taking, including family history, physical examination, and neurologic examination. We enrolled 6 Chr. Start End Length

siblings from 1 consanguineous family. We isolated genomic 1 108,247,048 120,434,626 12,187,578 DNA from 8 members from this family: 3 probands, 3 un- aﬀected siblings, and 2 parents. There are 2 additional aﬀected 1 145,587,476 157,497,406 11,909,930 individuals (V.7 and IV.5) in this pedigree who were not 7 36,530,276 44,243,953 7,713,677 assessed genetically because of the lack of access to their blood 10 121,950,989 128,473,047 6,522,058 samples.

2 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG Figure 1 Family pedigree, clinical features, and results of the genetic assessment

(A) Family pedigree showing the affected patients as dark gray-filled symbols. (+) refers to wild-type allele; (−) refers to mutant allele. (B) Filtering procedure of whole-exome variants. (C) MRIs of the affected proband (V.2) with multiple frontal and parietal white matter hyperintensities on T2w flair MRI, suggestive of diffuse demyelination lesion. (D) Representative spinal cord MRIs of affected probands (left: V.2, center: V.3, and right: V.4). White arrows indicating that hyperintensity signal lesions are suggestive of myelitis. (E) Confirmation of mutation in HTRA1 and its segregation by Sanger sequencing. (F) Predicted effect of the HTRA1S270Lfs*69 identified mutation on protein sequence.HMZ = homozygous; HTRA = high-temperature requirement A; HTZ = heterozygous; IBD = identity by descent; INDELs = insertions or deletions; SNVs = single nucleotide variants.

and the Exome Aggregate Consortium data set (ftp://ftp. Illumina GenomeStudio software with call rates >99%, and broadinstitute.org/pub/ExAC_release/release0.2/) Var- then, sex and parents-offspring family relationship were iants were next compared with an in-house database con- further verified. Identity-by-descent (IBD) mapping was taining 485 previously sequenced samples, which largely performed by searching for shared homozygous (HMZ) represent participants from the Middle East including Iran, regions among the 3 affected patients using customized and those that were present in more than 1% of the pre- programs written in Mathematica (Wolfram Research, viously sequenced samples were removed. Inc.). All HMZ regions that were >2 cm were examined Identity-by-descent mapping allowing 1% error rate. Those candidate regions were fi All blood samples were genotyped using Illumina further re ned by exclusion of HMZ segments present in ff fi HumanOmni 2.5-4v1_H SNP array according to the a parents and una ected members. The identi ed shared manufacturer’s instructions. Genotypes were called using HMZ loci are summarized in table 1.

Neurology.org/NG Neurology: Genetics | Volume 5, Number 4 | August 2019 3 with Dpn1 (a methylation-dependent endonuclease) for 2 Table 2 Summary of signs and symptoms in the affected hours at 37°C. The digested PCR product was transformed siblings into E. coli DH5α and was then plated on lysogeny broth agar μ Participant V.2 V.3 V.4 plate with 100 g/mL ampicillin. Eight colonies were picked from the LB agar plate and sent to Axil ScientificPteLtdfor Age at diagnosis (y) 34 30 32 sequencing. One of the positive colonies confirmed by se- Leg involvement (spasticity) +++ ++ ++ quencing was purified for transfection using the EndoFree

Gait problem (ataxia) +++ ++ ++ Plasmid Maxi Kit (Qiagen, #12362).

Behavioral and mood changes (aggression) ++ +++ +++ Cell culture, transfection, and Western Dysarthria ++ ——blot analysis Human embryonic kidney cells (HEK293, ATCC CRL-1573) Memory loss ++ + + weregrowninhigh-glucoseDulbecco’sModified Eagle Medium, Premature hair loss ++ + + supplemented with Glutamax, sodium pyruvate, and 10% fetal fi Low back pain ++ + + bovine serum (Gibco, Thermo Fisher Scienti c) at 37°C and 5% CO2. Transient transfection was performed using Lipofectamine Spinal canal stenosis +++2000 according to the manufacturer’s instructions (Thermo Recurrent rhinitis ++ + + Fisher Scientific). Serum-free conditioned media were collected 24 hours after transfection and cleared by centrifugation at 1,000g for 5 minutes. Cells were lysed in a buffer containing 25 mM Tris- Cloning of HTRA1S270Lfs*69 HCl pH 7.6, 150 mM NaCl, 1% sodium deoxycholate, 0.1% HTRA1S270Lfs*69 plasmid sodium dodecyl sulfate, and 1% Nonidet P-40. Lysates and The wild-type (WT) HTRA1 plasmid, which contains coding conditioned cell culture supernatants were analyzed by sodium DNA sequence of the HTRA1 gene in pcDNA3.1 vector, is a gift dodecyl sulfate–polyacrylamide gel electrophoresis and electro- from Professors Hiroaki Nozaki and Osamu Onodera (Brain transfer onto 0.2 μm nitrocellulose membranes using the Mini- Research Institute, Niigata University, Niigata, Japan). To in- Protean and Trans-Blot system (Biorad). Membranes were troduce a “G” into cDNA of HTRA1 at position 805 base pair for blockedwithTris-buffered saline supplemented with 0.1% construction of HTRA1S270Lfs*69 plasmid, we used site-directed Tween 20 and 4% nonfat milk for 1 hour at room temperature mutagenesis by the PCR method. In brief, the whole HTRA1 and probed overnight at 4°C using rabbit polyclonal anti-HTRA1 plasmid template was amplified using the primers of forward 59 (HPA036655, Sigma-Aldrich, 1: 5,000) and rabbit polyclonal -GCCTGTCCTGCTGCTTGGCCGGCTCCTCA- anti-GAPDH (AB9485, Abcam, 1: 2,500) antibodies. Detection GAGCTGCGGCCGGG-39 and reverse 59- was performed with horseradish peroxidase–conjugated second- CCCGGCCGCAGCTCTGAGGAGCCGGCCAAGCAG- ary antibodies (Cell Signaling, 1: 10,000), chemiluminescence CAGGACAGGC-39 with KOD Xtreme Hot Start DNA development (Immobilon ECL detection reagent, Merck Milli- Polymerase (Novagen, #71975), and cycling conditions were as pore), and the Fusion FX7 imaging system (Vilber Lourmat). follows: initial denaturation at 95°C for 4 minutes, followed by 15 cycles of 95°C for 30 seconds, 50°C for 1 minute, and 68°C Data availability for 7 minutes; The PCR product was purified and incubated All anonymized data can be shared on a collaborative basis.

Results Table 3 Features of the novel HTRA1 mutation identified In this study, 3 affected sibling patients and their unaffected Feature parents from 1 consanguineous pedigree in Isfahan province, ff fi Chromosome number 10 Iran, were analyzed. In all a ected probands ( gure 1A: V.2, V.3, and V.4), the manifestations appeared at almost the same age at Position of mutation 124266234 onset at the beginning of their third decade of life. These 3 Minor allele frequency (ExAC, UK10, in-house 0 patients all developed clinical features such as gait problems, database of 485 exomes) aggressive behavioral and mood changes, memory loss, pre- Gene name HTRA1 mature hair loss, low back pain, and spinal canal stenosis. Fur- thermore, only the affected members presented with recurrent Nature of mutation Exonic rhinitis that is chronic and nonseasonal, appearing monthly with Exon no. 4 allergic-like symptoms such as sneezing. Moreover, the disease

Change introduced in protein product p.Ser270Leufs*69 features in the elder proband V.2 were more prominent than those in younger patients V.3 and V.4. These signs and symptoms PhyloP score 2.46 have persisted and are summarized in table 2. MRI was only ﬀ HTRA = high-temperature requirement A. performed on the 3 a ected siblings based on their neurologic presentation and showed abnormal diﬀuse demyelination lesions

4 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG To identify possible disease-related mutations, whole-exome S270Lfs*69 Figure 2 HTRA1 is poorly expressed and in- sequencing was performed on the affected proband V.2. A efficiently secreted total of 5,558 potentially pathogenic variants were identified across protein-coding exons, untranslated regions, splice sites, and flanking introns. After following an autosomal recessive model of inheritance, a final set of 91 HMZ variants and 75 compound heterozygous (cHTZ) variants (34 gene sets) was identified (figure 1D). Of those, 8 variants (3 cHTZ variants in 1 gene and 5 HMZ variants in 5 genes) were located in the loci determined by IBD mapping and shared among the 3 siblings. The cHTZ variants in the CROCC gene did not pass our filtering criteria because 2 of the 3 are intronic and not conserved. Similarly, 2 intronic and poorly conserved variants in MIR7851 and LINC01138 and 1 exonic nonconserved variant in TNRC18, predicted to be nondeleterious by SIFT and PolyPhen, were excluded. The 2 remaining shared variants were a frameshift mutation in HTRA1 and a missense mutation in CGN. The missense mutation in CGN has a mi- nor allele frequency (MAF) higher than 1% in healthy subject databases, and the affected amino acid, Gly 415, is not conserved in humans. Furthermore, CGN-deficient mice have not been reported to exhibit any immunologic or neurologic abnormalities.15 Therefore, the clinical manifestations in our patientsarenotlikelytoberelatedtothemissensemutationin CGN. The only remaining mutation consistent with the patients’ phenotype that passed our filtering criteria including MAF below 1% in public and proprietary variant databases, exonic or

Immunoblots of cell lysates and conditioned media from HEK293 cells transiently highly conserved intronic variant (based on PhyloP score), and transfected with plasmids encoding wild-type HTRA1 or the truncated variant 16 S270Lfs*69 S270Lfs*69 pathogenic score based on SIFT, PolyPhen, and M-CAP was HTRA1 . The migration behavior of HTRA1 is in agreement S270Lfs*69 with its predicted molecular weight (;37 kDa). For the detection of HTRA1, the HTRA1 mutation (table 3). It was confirmed by a polyclonal antibody against amino acids 120–179 of HTRA1 was used Sanger sequencing and showed complete segregation in the (Sigma-Aldrich). Equal volumes of conditioned media were analyzed; for the cell lysates, GAPDH (glyceraldehyde 3-phosphate dehydrogenase) was pedigree (figure 1E) as a “G” insertion mutation at position 805 used as a loading control. HTRA = high-temperature requirement A. (c.805_806insG) in HTRA1, which is predicted to result in a truncated HTRA1 protein p. Ser270Leufs*69 (figure 1F). in the brain (figure 1B), spondylotic changes of the lumbar spine, and spinal channel stenosis (data not shown). In addition, only To examine the expression and secretion of the HTRA1S270Lfs*69 the affected siblings showed lower limb weakness, which was variant, the WT and mutant HTRA1 plasmids were transfected responsive to corticosteroid (methylprednisolone) treatment and into HEK293 cells and conditioned medium and cell pellets were hyperintensity signals on T2-weighted MRIs suggestive of mye- analyzed by Western blot 24 hours post-transfection (figure 2). litis in the spinal cord (figure 1C). The truncated protein was present at the predicted molecular

Figure 3 A schematic presentation of HTRA1 mutations

A representation of the HTRA1 gene, its protein domains, and the pathogenic mutations identified to date. The novel mutation reported here is highlighted in red. Scheme adapted from reference 28. HTRA = high-temperature requirement A.

Neurology.org/NG Neurology: Genetics | Volume 5, Number 4 | August 2019 5 weight in cell lysates, albeit at strongly reduced levels compared HTRA1 has recently been shown to be enriched in mature with full-length WT HTRA1. In conditioned supernatants, the astrocytes in the brain.31 Given their well-established role in ce- mutant protein could not be detected at all, whereas WT HTRA1 rebrovascular and blood flow regulation,32 it is tempting to was clearly identified (figure 2). These findings suggest that the speculate whether astrocytes play a role in mediating the in- S270Lfs*69 mutant might be rapidly degraded intracellularly flammatory features and the cerebral arteriopathy in CARASIL. preventing its secretion. Of note, recent studies investigating Alexander disease, a leuko- dystrophy caused by mutations in the astrocyte-enriched glial fibrillary acidic protein, provide evidence of increased immune Discussion 33–35 activation, cytokine production, and inflammatory response. Multiple HMZ mutations in HTRA1 have been found in These studies suggest an important role for astrocytic dysfunction – patients with CARASIL.17 22 These mutations are concentrated in cerebrovascular and demyelinating disorders. in the protease domain, and all have been shown or are predicted to reduce the catalytic activity of HTRA1 (figure 3). In this We identified a novel frameshift mutation in the CARASIL- study, we demonstrate that the novel HTRA1 variant we iden- associated gene HTRA1 and characterized its functional tified, which is located in the middle of the protease domain, consequences. Moreover, some aspects of the clinical pre- results in a truncated protein lacking the active site serine residue sentation in affected patients described here deviate from (S328) and exhibiting significantly reduced expression levels. those previously reported in CARASIL. Therefore, our study expands the spectrum of clinical consequences of mutations in Heterozygous HTRA1 mutations have been shown to be as- HTRA1. sociated with a dominant, late-onset form of small vessel disease, without any extraneurologic symptoms.23,24 Although Acknowledgment these mutations are mostly missense mutations, the mutation The authors thank Professors Hiroaki Nozaki and Osamu identified in the present study clearly results in a loss of Onodera (Brain Research Institute, Niigata University, HTRA1 function and thus represents a typical CARASIL Niigata, Japan) for the HTRA1 vectors. mutation. Accordingly, the heterozygous parents of the 3 affected siblings were healthy despite an age well within the Study funding reported range of age at onset for the dominant small vessel This study was supported by the Agency for Science Tech- disease form (IV.1: 58 years, IV.2: 68 years). nology and Research (SPF2012/005) to M.A.P. and the German Research Foundation (DFG, HA2448/6-1) to C.H. S In this study, all 3 patients with the HTRA1 270Lfs*69 mutation demonstrated typical clinical CARASIL features and some atyp- Disclosure ical features such as recurrent rhinitis and possible spinal cord Disclosures available: Neurology.org/NG. myelitis suggestive of inflammation with lower limb weakness responsive to corticosteroid (methylprednisolone) treatment. Of Publication history note, a case report recently described myelitis in cerebral auto- Received by Neurology: Genetics October 29, 2018. Accepted in final somal dominant arteriopathy with subcortical infarcts and leu- form May 28, 2019. koencephalopathy.25 The presence of 2 additional affected participants (V.7 and IV.5) in this pedigree further supports the role of genetics in their clinical presentation. It is highly likely that Appendix Authors these individuals carry the same HTRA1 mutation as the one harbored by the 3 analyzed participants in this study. Un- Name Affiliation Role Contribution fortunately, because DNA samples from these additional indi- Amin Ziaei, MD Translational Author Role in identifying viduals are not available, it will not be possible to confirm this by Laboratory in Genetic family and clinical Medicine (TLGM), data acquisition; genetic analysis. Agency for Science, designed and Technology and conceptualized the Research (A*STAR), study; analyzed the The mechanism by which loss of HTRA1 function results in 8A Biomedical Grove, data; role in genetic CARASIL has been proposed to involve altered TGF-β signaling, Immunos, Level 5, data analysis; and 138648, Singapore drafted and revised although the exact nature of this dysregulation whether it is Department of the manuscript. – caused by excessive11,26 28 or reduced6 TGF-β pathway activity Medicine, National 29 β University of remains a subject of debate. Moreover, TGF- has well- Singapore, 117597, established immunoregulatory roles.30 We speculate that the in- Singapore fl Department of ammatory features of the probands in the present study such as Neurology, Isfahan spinal cord myelitis and recurrent rhinitis may be related to the Neurosciences β 11,28 Research Centre, dysregulation of TGF- signaling. On the other hand, these Faculty of Medicine, symptoms, which are atypical for CARASIL, may be caused by Isfahan University of ff Medical Sciences, other genetic variants shared by the 3 a ected siblings, a possi- Isfahan, Iran bility we cannot formally rule out.

6 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG Appendix (continued) Appendix (continued)

Name Affiliation Role Contribution Name Affiliation Role Contribution

Xiaohong Xu, Translational Author Designed and Byrappa Comparative Author Genetic data MD, PhD Laboratory in conceptualized the Venkatesh, Genomics analysis and revised Genetic Medicine study; analyzed the PhD Laboratory, the manuscript. (TLGM), Agency for data; and drafted Institute of Science, Technology and revised the Molecular and Cell and Research manuscript. Biology, A*STAR, (A*STAR), 8A Biopolis, Singapore Biomedical Grove, 138673, Singapore Immunos, Level 5, 138648, Singapore Christof Institute for Stroke Author Biochemical Department of Haffner, PhD and Dementia analysis; Neurology and Research, Klinikum interpreted data; Stroke Center, the der Universit¨at and revised the First Affiliated Munchen,¨ Ludwig manuscript. Hospital, Jinan Maximilians University, 613 University, Munich, Huangpu Avenue Germany West, Guangzhou, Guangdong 510632, Bruno Institute of Medical Author Genetic data China Reversade, Biology (IMB), analysis and revised Clinical Neuroscience PhD A*STAR, 8A the manuscript. Institute of Jinan Biomedical Grove, University, 613 Immunos, Level 5, Huangpu Avenue 138648, Singapore West, Guangzhou, Guangdong 510632, Vahid Department of Author Role in identifying China Shaygannejad, Neurology, Isfahan family and clinical MD Neurosciences data acquisition and Leila Department of Author Role in identifying Research Centre, revised the Dehghani, MSc Tissue Engineering family and clinical Faculty of Medicine, manuscript. and Regenerative data acquisition and Isfahan University Medicine, School revised the of Medical Sciences, of Advanced manuscript. Isfahan, Iran Technologies in Medicine, Shahid Mahmoud A. Translational Author Designed and Beheshti Pouladi, PhD Laboratory in conceptualized the University of Genetic Medicine study; analyzed the Medical Sciences, (TLGM), Agency for data; role in genetic Tehran, Iran Science, Technology data analysis; and and Research drafted and revised Carine Institute of Author Genetic data (A*STAR), 8A the manuscript. Bonnard, PhD Medical Biology analysis and revised Biomedical Grove, (IMB), A*STAR, 8A the manuscript. Immunos, Level 5, Biomedical 138648, Singapore Grove, Immunos, Department of Level 5, 138648, Medicine, National Singapore University of Singapore, 117597, Andreas Institute for Stroke Author Biochemical Singapore Zellner, PhD and Dementia analysis; Department of Research, Klinikum interpreted data; Physiology, National der Universit¨at and revised the University of Munchen,¨ Ludwig manuscript. Singapore, 117597, Maximilians Singapore University, Munich, Germany

Alvin Yu Jin Ng, Comparative Author Genetic data PhD Genomics analysis; Revised Laboratory, the manuscript. References Institute of 1. Clausen T, Southan C, Ehrmann M. The HtrA family of proteases. Mol Cell 2002;10: Molecular and 443–455. Cell Biology, 2. Clausen T, Kaiser M, Huber R, Ehrmann M. HTRA proteases: regulated proteolysis A*STAR, Biopolis, in protein quality control. Nat Rev Mol Cel Biol 2011;12:152–162. Singapore 138673, 3. Zumbrunn J, Trueb B. Primary structure of a putative serine protease speciﬁc for IGF- Singapore binding proteins. FEBS Lett 1996;398:187–192. 4. Chien J, Staub J, Hu SI, et al. A candidate tumor suppressor HtrA1 is downregulated in Sumanty Comparative Author Genetic data ovarian cancer. Oncogene 2004;23:1636–1644. Tohari, BSc Genomics analysis and revised 5. An E, Sen S, Park SK, Gordish-Dressman H, Hathout Y. Identiﬁcation of novel Laboratory, the manuscript. substrates for the serine protease HTRA1 in the human RPE secretome. Invest Institute of Ophthalmol Vis Sci 2010;51:3379. Molecular and Cell 6. Beaufort N, Scharrer E, Kremmer E, et al. Cerebral small vessel disease-related pro- Biology, A*STAR, tease HtrA1 processes latent TGF-β binding protein 1 and facilitates TGF-β signaling. Biopolis, Singapore Proc Natl Acad Sci U S A 2014;111:16496–16501. 138673, Singapore 7. DeWan A, Liu M, Hartman S, et al. HTRA1 promoter polymorphism in wet age- related macular degeneration. Science 2006;314:989–992.

Neurology.org/NG Neurology: Genetics | Volume 5, Number 4 | August 2019 7 8. Baldi A, Luca AD, Morini M, et al. The HtrA1 serine protease is down-regulated 22. Wang XL, Li CF, Guo HW, Cao BZ. A novel mutation in the HTRA1 gene identified during human melanoma progression and represses growth of metastatic melanoma in Chinese CARASIL pedigree. CNS Neurosci Ther 2012;18:867–869. cells. Oncogene 2002;21:6684–6688. 23. Verdura E, Hervé D, Scharrer E, et al. Heterozygous HTRA1 mutations are associated 9. Esposito V, Campioni M, De Luca A, et al. Analysis of HtrA1 serine protease ex- with autosomal dominant cerebral small vessel disease. Brain 2015;138:2347–2358. pression in human lung cancer. Anticancer Res 2006;26:3455–3459. 24. Nozaki H, Kato T, Nihonmatsu M, et al. Distinct molecular mechanisms of HTRA1 10. Grau S, Baldi A, Bussani R, et al. Implications of the serine protease HtrA1 in amyloid mutants in manifesting heterozygotes with CARASIL. Neurology 2016;86: precursor protein processing. Proc Natl Acad Sci U S A 2005;102:6021–6026. 1964–1974. 11. Hara K, Shiga A, Fukutake T, et al. Association of HTRA1 mutations and familial 25. Collongues N, Derache N, Blanc F, Labauge P, de Seze J, Defer G. Inflammatory-like ischemic cerebral small-vessel disease. N Engl J Med 2009;360:1729–1739. presentation of CADASIL: a diagnostic challenge. BMC Neurol 2012;12:78. 12. Maeda S, Nakayama H, Isaka K, Aihara Y, Nemoto S. Familial unusual encephalopathy of 26. Shiga A, Nozaki H, Yokoseki A, et al. Cerebral small-vessel disease protein HTRA1 Binswanger’s type without hypertension. Psychiatry Clin Neurosci 1976;30:165–177. controls the amount of TGF-β1 via cleavage of proTGF-β1. Hum Mol Genet 2011; 13. Fukutake T, Hirayama K. Familial young-adult-onset arteriosclerotic leukoence- 20:1800–1810. phalopathy with alopecia and lumbago without arterial hypertension. Eur Neurol 27. Graham JR, Chamberland A, Lin Q, et al. Serine protease HTRA1 antagonizes 1995;35:69–79. transforming growth factor-β signaling by cleaving its receptors and loss of HTRA1 in 14. Yanagawa S, Ito N, Arima K, Ikeda S. Cerebral autosomal recessive arteriopathy with vivo enhances bone formation. PLoS One 2013;8:e74094. subcortical infarcts and leukoencephalopathy. Neurology 2002;58:817–820. 28. Tikka S, Baumann M, Siitonen M, et al. CADASIL and CARASIL: CADASIL and 15. Guillemot L, Schneider Y, Brun P, et al. Cingulin is dispensable for epithelial barrier CARASIL. Brain Pathol 2014;24:525–544. function and tight junction structure, and plays a role in the control of claudin-2 29. Zellner A, Scharrer E, Arzberger T, et al. CADASIL brain vessels show a HTRA1 loss- expression and response to duodenal mucosa injury. J Cell Sci 2012;125:5005–5014. of-function profile. Acta Neuropathol 2018;136:111–125. 16. Jagadeesh KA, Wenger AM, Berger MJ, et al. M-CAP eliminates a majority of variants of 30. Li MO, Wan YY, Sanjabi S, Robertson AKL, Flavell RA. Transforming growth factor-β uncertain significance in clinical exomes at high sensitivity. Nat Genet 2016;48:1581–1586. regulation of immune responses. Annu Rev Immunol 2006;24:99–146. 17. Bayrakli F, Balaban H, Gurelik M, Hizmetli S, Topaktas S. Mutation in the HTRA1 31. Chen J, Van Gulden S, McGuire TL, et al. BMP-responsive protease HtrA1 is dif- gene in a patient with degenerated spine as a component of CARASIL syndrome. ferentially expressed in astrocytes and regulates astrocytic development and injury Turk Neurosurg 2014;24:67–69. response. J Neurosci 2018;38:3840–3857. 18. Bianchi S, Di Palma C, Gallus GN, et al. Two novel HTRA1 mutations in a European 32. MacVicar BA, Newman EA. Astrocyte regulation of blood flow in the brain. Cold CARASIL patient. Neurology 2014;82:898–900. Spring Harb Perspect Biol 2015;7:a020388. 19. Chen Y, He Z, Meng S, Li L, Yang H, Zhang X. A novel mutation of the high- 33. Olabarria M, Putilina M, Riemer EC, Goldman JE. Astrocyte pathology in Alexander temperature requirement A serine peptidase 1 (HTRA1) gene in a Chinese family disease causes a marked inflammatory environment. Acta Neuropathol 2015;130: with cerebral autosomal recessive arteriopathy with subcortical infarcts and leu- 469–486. koencephalopathy (CARASIL). J Int Med Res 2013;41:1445–1455. 34. Kondo T, Funayama M, Miyake M, et al. Modeling Alexander disease with patient 20. Mendioroz M, Fernandez-Cadenas I, del Rio-Espinola A, et al. A missense HTRA1 iPSCs reveals cellular and molecular pathology of astrocytes. Acta Neuropathol mutation expands CARASIL syndrome to the Caucasian population. Neurology Commun 2016;4:69. 2010;75:2033–2035. 35. Li L, Tian E, Chen X, et al. GFAP mutations in astrocytes impair oligodendrocyte 21. Nishimoto Y, Shibata M, Nihonmatsu M, et al. A novel mutation in the HTRA1 gene progenitor proliferation and myelination in an hiPSC model of alexander disease. Cell causes CARASIL without alopecia. Neurology 2011;76:1353–1355. Stem Cell 2018;23:239–251.e6.

8 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG ARTICLE OPEN ACCESS Genetic risk of Parkinson disease and progression: An analysis of 13 longitudinal cohorts

Hirotaka Iwaki, MD, Cornelis Blauwendraat, PhD, Hampton L. Leonard, MS, Ganqiang Liu, PhD, Correspondence Jodi Maple-Grødem, PhD, Jean-Christophe Corvol, MD, PhD, Lasse Pihlstrøm, MD, PhD, Dr. Nalls [email protected] Marlies van Nimwegen, PhD, Samantha J. Hutten, PhD, Khanh-Dung H. Nguyen, PhD, Jacqueline Rick, PhD, Shirley Eberly, MS, Faraz Faghri, MS, Peggy Auinger, MS, Kirsten M. Scott, MRCP, MPhil, Ruwani Wijeyekoon, MRCP, Vivianna M. Van Deerlin, MD, PhD, Dena G. Hernandez, PhD, Aaron G. Day-Williams, PhD, Alexis Brice, MD, Guido Alves, MD, PhD, Alastair J. Noyce, MRCP, PhD, Ole-Bjørn Tysnes, MD, PhD, Jonathan R. Evans, MRCP, PhD, David P. Breen, MRCP, PhD, Karol Estrada, PhD, Claire E. Wegel, MPH, Fabrice Danjou, MD, PhD, David K. Simon, MD, PhD, Bernard Ravina, MD, Mathias Toft, MD, PhD, Peter Heutink, PhD, Bastiaan R. Bloem, MD, PhD, Daniel Weintraub, MD, Roger A. Barker, MRCP, PhD, Caroline H. Williams-Gray, MRCP, PhD, Bart P. van de Warrenburg, MD, PhD, Jacobus J. Van Hilten, MD, PhD, Clemens R. Scherzer, MD, Andrew B. Singleton, PhD, and Mike A. Nalls, PhD

Neurol Genet 2019;5:e348. doi:10.1212/NXG.0000000000000348 Abstract Objective To determine if any association between previously identiﬁed alleles that confer risk for Parkinson disease and variables measuring disease progression.

Methods We evaluated the association between 31 risk variants and variables measuring disease progression. A total of 23,423 visits by 4,307 patients of European ancestry from 13 longitudinal cohorts in Europe, North America, and Australia were analyzed.

Results We confirmed the importance of GBA on phenotypes. GBA variants were associated with the development of daytime sleepiness (p.N370S: hazard ratio [HR] 3.28 [1.69–6.34]) and possible REM sleep behavior (p.T408M: odds ratio 6.48 [2.04–20.60]). We also replicated previously reported associations of GBA variants with motor/cognitive declines. The other genotype-phenotype associations include an intergenic variant near LRRK2 and the faster development of motor symptom (Hoehn and Yahr scale 3.0 HR 1.33 [1.16–1.52] for the C allele of rs76904798) and an intronic variant in PMVK and the development of wearing-off effects (HR 1.66 [1.19–2.31] for the C allele of rs114138760). Age at onset was associated with TMEM175 variant p.M393T (−0.72 [−1.21 to −0.23] in years), the C allele of rs199347 (intronic region of GPNMB, 0.70 [0.27–1.14]), and G allele of rs1106180 (intronic region of CCDC62, 0.62 [0.21–1.03]).

Conclusions This study provides evidence that alleles associated with Parkinson disease risk, in particular GBA variants, also contribute to the heterogeneity of multiple motor and nonmotor aspects. Accounting for genetic variability will be a useful factor in understanding disease course and in minimizing heterogeneity in clinical trials.

From the Laboratory of Neurogenetics (H.I., C.B., H.L.L., F.F., D.G.H., A.B.S., M.A.N.), National Institute on Aging, National Institutes of Health, Bethesda; Data Tecnica International (H.I., M.A.N.), Glen Echo, MD; Precision Neurology Program (G.L., C.R.S.), Harvard Medical School, Brigham and Women’s Hospital; Neurogenomics Laboratory (G.L., C.R.S.), Harvard Medical School, Brigham and Women’s Hospital; Ann Romney Center for Neurologic Diseases (G.L., C.R.S.), Brigham and Women’s Hospital, Boston, MA; The Norwegian Centre for Movement Disorders (J.M.-G., G.A.), Stavanger University Hospital; Department of Chemistry (J.M.-G., G.A.), Bioscience and Environmental Engineering, University of Stavanger, Norway; Assistance-Publique Hopitauxˆ de Paris (J.-C.C.), ICM, INSERM UMRS 1127, CNRS 7225, ICM, Department of Neurology and CIC Neurosciences, Pitié-Salpêtrière Hospital, Paris, France; Department of Neurology (L.P., M.T.), Oslo University Hospital, Norway; Department of Neurology (M.N., B.R.B., B.P.W.), Donders Institute for Brain, Cognition, and Behaviour, Radboud University Medical Centre, Nijmegen, The Netherlands; Michael J Fox Foundation (S.J.H.), New York; Translational Genome Sciences (K.-D.H.N, K.E.), Biogen, Cambridge, MA; Department of Neurology University of Pennsylvania (J.R.), Philadelphia, PA; Department of Biostatistics and Computational Biology (S.E.), University of Rochester, NY; Department of Computer Science (F.F.), University of Illinois Urbana-Champaign; Department of Neurology (P.A.), Center for Health + Technology, University of Rochester, NY; Department of Clinical Neurosciences (K.M.S., R.W.), University of Cambridge, John van Geest Centre for Brain Repair, UK; Department of Pathology and Laboratory Medicine (V.M.V.D.), Center for Neurodegenerative Disease Research, Parelman School of Medicine at the University of Pennsylvania, Philadelphia; Genetics and Pharmacogenomics (A.G.D.-W.), Merck Research Laboratory, Boston, MA; Statistical Genetics (A.G.D.-W.), Biogen, Cambridge, MA; Institut du cerveau et de la moelle épinière ICM (A.B., F.D.); Sorbonne Université SU (A.B.); INSERM UMR1127 (A.B.), Paris, France; Department of Neurology (G.A.), Stavanger University Hospital, Norway; Preventive Neurology Unit (A.J.N.), Wolfson Institute of Preventive Medicine, Queen Mary University of London; Department of Molecular Neuroscience (A.J.N.), UCL Institute of Neurology, London, UK; Department of Neurology (O.-B.T.), Haukeland University Hospital; University of Bergen (O.-B.T.), Bergen, Norway; Department of Neurology (J.R.E.), Nottingham University NHS Trust, UK; Centre for Clinical Brain Sciences (D.P.B.), University of Edinburgh; Anne Rowling Regenerative Neurology Clinic (D.P.B.), University of Edinburgh; Usher Institute of Population Health Sciences and Informatics (D.P.B.), University of Edinburgh, Scotland; Department of Medical and Molecular Genetics (C.E.W.), Indiana University, Indianapolis; Department of Neurology (D.K.S.), Beth Israel Deaconess Medical Center; Harvard Medical School (D.K.S.), Boston; Voyager Therapeutics (B.R.), Cambridge, MA; Department of Neurology (B.R.), University of Rochester School of Medicine, NY; Institute of Clinical Medicine (M.T.), University of Oslo, Norway; German Center for Neurodegenerative Diseases-Tubingen (P.H.); HIH Tuebingen (P.H.), Germany; Department of Psychiatry (D.W.), University of Pennsylvania School of Medicine; Department of Veterans Affairs (D.W.), Philadelphia, PA; and Department of Clinical Neurosciences (R.A.B., C.H.W.-G.), University of Cambridge, UK; Department of Neurology (J.J.V.H.), Leiden University Medical Center, The Netherlands.

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

Copyright © 2019 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 Glossary ESS = Epworth Sleepiness Scale; FDR = false discovery rate; GRS = genetic risk score; GWAS = genome-wide association study; HR = hazard ratio; HY = Hoehn and Yahr scale; MAF = minor allele frequency; MDS = Movement Disorder Society; MMSE = Mini-Mental State Examination; MoCA = Montreal Cognitive Assessment; MSQ = Mayo Sleep Questionnaire; NMS = nonmotor symptom; OR = odds ratio; PC = principal component; PDSS = Parkinson’s Disease Sleep Scale; PPMI = Parkinson’s Progression Markers Initiative; RBD = rapid eye movement sleep behavior disorder; RBDSQ = RBD Screening Questionnaire; RLS = restless legs syndrome; SEADL = Schwab and England Activities of Daily Living Scale; UPDRS = Uniﬁed Parkinson’s Disease Rating Scale.

Parkinson disease is one of the most common neurodegen- (PPMI), Profiling Parkinson’s disease study (ProPark), and erative diseases, with an estimated lifetime risk as high as the Morris K. Udall Centers for Parkinson’s Research (Udall). 1%–2%.1 Parkinson disease is traditionally characterized by The 4 cohorts from randomized clinical trials were Deprenyl motor features such as bradykinesia, rigidity, and tremor. and Tocopherol Antioxidative Therapy of Parkinsonism However, in addition to these motor symptoms, patients with (DATATOP), NIH Exploratory Trials in Parkinson’s Disease Parkinson disease also develop nonmotor symptoms Large Simple Study 1, ParkFit study (ParkFit), and Parkinson (NMSs), which include depression, cognitive decline, sleep Research Examination of CEP-1347 Trial with a subsequent abnormalities, reduced olfaction, and autonomic dysfunc- prospective study (PreCEPT/PostCEPT). Information on tion.2 Collectively, the combined spectrum of motor and these cohorts can be found in appendix e-1 (links.lww.com/ NMSs more accurately reflects the multisystem nature of the NXG/A169). Subsets of participants from the cohorts who disease. Patients with Parkinson disease may present with provided DNA and were nonrelated participants with PD, various combinations of symptoms and show differences in diagnosed at age 18 years or later, and of European ancestry the rates of progression.3 The application of modern molec- were included in the study. Participants’ information and ular genetic approaches over the last decade has revealed genetic samples were obtained under appropriate written a significant number of genetic risk loci for idiopathic Par- consent and with local institutional and ethical approvals. – kinson disease.4 7 However, in comparison with case-control genome-wide association study (GWAS), analyzing how ge- Genotyping SNPs and calculation of GRS netic factors influence clinical presentation and progression Oslo samples were genotyped on the Illumina Infinium requires longitudinal cohorts with much more detailed OmniExpress array, DIGPD samples were genotyped by observations. Such data are sparse, and individual cohorts are Illumina Multi-Ethnic Genotyping Array, and all other sam- 8 often small in size and quite varied, posing a challenge both in ples were genotyped on the NeuroX array. The quality sample size and heterogeneity. control process of variant calling included GenTrain score <0.7, minor allele frequency (MAF) >0.05 (for sample quality In an attempt to address these issues, we collected data from control but not in our analysis of rare risk factors), and Hardy- −6 13 distinct longitudinal Parkinson disease cohorts with de- Weinberg equilibrium test statistic >10 . Sample-specific tailed clinical data, including assessment of disease pro- quality control included a sample call rate of >0.95, confir- gression. We sought to determine whether Parkinson disease mation of sex through genotyping, homozygosity quantified genetic risk factors, either in the form of known GWAS var- by F within ± 3 SD from the population mean, European iants or an aggregate genetic risk score (GRS), are associated ancestry confirmed by principal-components analysis with with changes in clinical progression and the disease features. 1000 Genomes data as the reference, and genetic relatedness of any 2 individuals <0.125. Detailed information regarding NeuroX and the quality control process has been described Methods previously.9 In the present study, we investigated 31 single nucleotide polymorphisms (SNPs) previously shown to be – Study design and participants significantly associated with Parkinson disease.10 12 In addi- A total of 13 Parkinson disease cohorts from North America, tion, we also calculated a GRS for each participant based on Europe, and Australia participated in the study. Nine were these variants. The scores were transformed into Z-scores prospective observational cohorts and the rest were from within each cohort and treated as an exposure, with effect randomized clinical trials. The observational cohorts were estimates based on 1 SD change from the population mean. Drug Interaction with Genes in Parkinson’s Disease The list of 31 SNPs and the GRS calculation method are (DIGPD), Harvard Biomarkers Study (HBS), Oslo Parkin- provided in table e-1 (links.lww.com/NXG/A170). son’s Disease study (partly including retrospective data), The Norwegian ParkWest study (ParkWest), Parkinson’s Disease Furthermore, principal components (PCs) were created for Biomarker Program (PDBP), Parkinsonism: Incidence and each data set from genotypes using PLINK. For the PC calcu- Cognitive and Non-motor heterogeneity In CambridgeShire lation, variants were filtered for MAF (>0.05), genotype miss- − (PICNICS), Parkinson’s Progression Markers Initiative ingness (<0.05), and Hardy-Weinberg equilibrium (p ≥ 10 5).

2 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG The remaining variants were pruned (using a 50-kb window, specifications were slightly different between cohorts (de- 2 with a 5 SNP shift per window and r threshold of 0.5), and PCs tailed in table e-3, links.lww.com/NXG/A172). Briefly, the were calculated using the pruned variants. associations between an SNP/GRS and age at onset were analyzed by linear regression modeling adjusting for pop- Measurements ulation stratification (PC1 and PC2). The association be- The following clinical measurements and binomial outcomes were tween family history of Parkinson disease and SNP/GRS was recorded longitudinally (table e-2 links.lww.com/NXG/A171): analyzed with a logistic regression model adjusting for PC1/2. total and subscores of the Unified Parkinson’s Disease Rating For continuous variables, linear regression modeling adjusting Scale (UPDRS) or the Movement Disorder Society revised for sex, education, PC1/2, age at onset, years from diagnosis, UPDRS version (MDS-UPDRS); modified Hoehn and Yahr family history, and treatment status was applied. For those scales (HY); modified Schwab and England Activities of Daily who had multiple observations, random intercept was added Living Scale; and scores for the Mini-Mental State Examination to adjust for repeated measurements of the same individual. (MMSE), The SCales for Outcomes in PArkinson's disease For binomial outcomes, the logistic regression at baseline (SCOPA)-Cognition, and Montreal Cognitive Assessment observation was applied using the same covariates as the (MoCA). Each was treated as a continuous outcome. For the continuous models. Those that were negative at baseline were UPDRS and MDS-UPDRS scores specifically, we took Z-scores further analyzed by a Cox regression with the same covariates of the total and subscores (except for part 4 at baseline) to but with treatment status as a time-varying covariate. Obser- compare the original and revised UPDRS versions. The conver- vations with missing variables were excluded from the analyses. sion was applied to the scores for all subsequent visits. For UPDRS part 4, most participants had very low scores or 0 at baseline, so we Meta-analysis normalized across all observations within each cohort. We also We applied inverse weighting (precision method) for each analyzed binomial outcomes. If we had access to the raw data, we combination of outcome-predictor association and combined used common cutoff values, which had been tested and reported the estimates from the 13 different cohorts in a fixed effect specificity of 85% or more in patients’ population. The binomial model. Multiple test correction for SNPs was controlled with outcomes include existence of family history (1st-degree relative. an overall false discovery rate (FDR) of 0.05 per outcome 1st- and 2nd-degree relatives in HBS, PreCEPT, ProPark, and being considered significant. Similarly, multiple testing of Udall), hyposmia (University of Pennsylvania Smell Identification outcomes for GRS was corrected with an FDR of 0.05, but Test < 21,13 or answering “yes” to question 2 in the NMS ques- across all traits. In addition, as a test of homogeneity, I2 indices tionnaire), cognitive impairment (SCOPA-Cognition < 23, and forest plots were used for quantitative assessment. As MMSE < 27, or MoCA < 24,14,15 or diagnosed with The a sensitivity analysis, we conducted up to 13 iterations of the Diagnostic and Statistical Manual of Mental Disorders -IV criteria for meta-analyses for the 12 cohorts excluding each cohort per dementia), wearing off (UPDRS/MDS-UPDRS part 4 off time >0 iteration. This analysis provides information regarding het- or physician’s diagnosis), dyskinesia (UPDRS/MDS-UPDRS part erogeneity of the cohorts and how one specific cohort ex- 4 dyskinesia time >0 or physician’s diagnosis), depression (Beck clusion affects the results. The range of estimates and Depression Inventory > 14 [PICNICS used 9 instead of 14], maximum p values for the iterations were included. Finally, we Hamilton Depression Rating Scale > 9, Geriatric Depression Scale conducted the 13-cohort meta-analysis in a random effects [GRS] > 5,16 or physician’s diagnosis), constipation (MDS- model with restricted maximum likelihood estimation using UPDRS part 1 item 11 > 0, or answering “yes” to question 5 in the same multiple testing correction. the NMS questionnaire), excessive daytime sleepiness (Epworth sleepiness scale > 9,17 insomnia [MDS-UPDRS part 1, item 7 > All the above analyses were conducted with PLINK version 1.9, 0], rapid eye movement sleep behavior disorder [RBD]) (an- and R version 3.4.4 (64-bit). Statistical tests were all 2 sided. swered “yes” to question 1 on the Mayo Sleep Questionnaire [MSQ],18 or RBD screening questionnaire [RBDSQ >5],19 Data availability fi restless legs syndrome [RLS]) (answered “yes” to MSQ ques- Quali ed investigators can request raw data through the ’ tion 3,20 or RLS diagnosis positive by RBDSQ), and the pro- organizations homepages (PDBP: pdbp.ninds.nih.gov/, gression to HY ≥ 3 (HY3, representing moderate to severe PPMI: ppmi-info.org/) or collaboration. disease). The individual definitions of these binomial outcomes are summarized in table e-2 (links.lww.com/NXG/A171). Age, sex, years of education, age at motor symptom onset, and Results whether the patient was treated with levodopa or dopamine A total of 23,423 visits by 4,307 patients with a median follow- agonists at each visit were alsorecordedforadjustments. up period of 2.97 years (quartile range of [1.63–4.94] years) Statistical analysis were eligible for the analysis. The baseline characteristics of the cohorts are shown in table 1. The mean ages at onset Cohort-level analysis varied from 54 to 69 years; the average disease durations at We analyzed the association between exposures and out- cohort entry ranged from less than 1 to 10 years, and the mean comes using appropriate additive models. Covariates of in- observation periods were between 1.2 and 6.8 years. All terest were not available for all cohorts; therefore, the model DATATOP, ParkWest, PPMI, and PreCEPT participants

Table 1 Summary characteristics of 13 cohorts

NET-PD PreCEPT/ DATATOP DIGPD HBS LS1 Oslo ParkFit ParkWest PDBP PICNICS PPMI PostCEPT ProPark Udall

Cohort size, n 440 311 580 406 317 335 150 422 120 357 321 296 252

Follow-up 1.22 (0.41) 2.19 (1.51) 1.53 (0.87) 4.48 (1.45) 4.64 (3.10) 1.97 (0.00) 3.04 (0.09) 2.06 (1.70) 3.04 (1.63) 4.87 (1.35) 6.79 (0.95) 4.62 (1.14) 3.77 (1.81) duration, y

Female, n (%) 146 (33.2) 121 (38.9) 201 (34.7) 148 (36.5) 107 (33.8) 110 (32.8) 57 (38.0) 174 (41.2) 43 (35.8) 121 (33.9) 106 (33.0) 105 (35.5) 73 (29.0)

Family history, 86 (20.9) 69 (22.3) 148 (25.5) 59 (14.5) 43 (14.0) — 17 (11.3) 54 (12.8) 19 (15.8) 48 (13.5) 93 (29.2) 76 (25.9) 71 (28.4) n(%)

Age at onset, y 58.65 (9.17) 59.41 (9.80) 62.16 (10.46) 60.64 (9.45) 54.33 (10.06) 60.79 (8.65) 67.27 (9.26) 58.51 (10.28) 68.94 (9.34) 61.45 (9.55) 59.47 (9.22) 53.14 (10.60) 64.26 (8.64)

Baseline from 1.14 (1.17) 2.60 (1.57) 4.09 (4.63) 1.50 (1.00) 10.13 (6.04) 5.18 (4.44) 0.13 (0.12) 5.68 (5.64) 0.23 (0.48) 0.54 (0.54) 0.80 (0.83) 6.56 (4.67) 6.21 (5.38) diagnosis, y

Levodopa use, 0 (0.0) 198 (63.9) 415 (71.6) 207 (51.2) ——0 (0.0) 255 (60.4) 36 (30.0) 0 (0.0) 0 (0.0) 202 (68.2) 215 (85.3) n(%)

Dopamine 0 (0.0) 228 (73.3) 224 (38.6) 280 (69.3) ——0 (0.0) 61 (14.5) 22 (18.3) 0 (0.0) 1 (0.3) 222 (75.0) 118 (46.8) agonist use, n (%)

Modified HY 1.61 (0.53) 1.75 (0.55) 2.14 (0.64) — 2.19 (0.64) 2.08 (0.33) 1.86 (0.58) 2.04 (0.69) 1.64 (0.67) 1.55 (0.50) 1.75 (0.48) 2.51 (0.79) 2.29 (0.68)

UPDRS1 — 7.69 (4.50) 1.70 (1.59) 1.31 (1.45) ——1.95 (1.76) 9.90 (6.11) — 5.40 (3.97) 0.84 (1.19) — 1.92 (1.99)

UPDRS2 — 7.72 (4.66) 9.21 (5.23) 7.29 (3.86) ——8.19 (4.22) 11.14 (8.01) — 5.80 (4.11) 6.11 (3.20) — 10.74 (7.13)

UPDRS3 — 20.37 (10.23) 19.30 (9.58) 17.77 (8.32) 15.42 (10.30) — 22.09 (9.77) 23.64 (13.08) — 20.88 (9.00) 18.69 (7.65) — 22.92 (11.09)

UPDRS4 — 0.66 (2.56) 2.25 (2.05) 1.34 (1.49) ——0.57 (1.14) 2.20 (3.17) ————2.02 (2.75)

MDS_UPDRS total — 36.43 (16.02) —————46.88 (24.04) 47.27 (17.97) ————

UPDRS total 24.68 (11.56) — 32.33 (14.28) 27.67 (11.62) — 32.11 (10.10) 32.79 (13.91) ———25.39 (10.10) — 32.64 (18.28)

MMSE 28.99 (1.35) 28.38 (1.73) 28.35 (2.17) ——28.09 (1.61) 27.88 (2.27) — 28.71 (1.43) — 29.29 (1.07) 27.05 (2.50) 26.83 (3.50)

MoCA ———————25.44 (3.40) — 27.17 (2.23) ——24.37 (3.63)

SEADL 91.55 (6.49) 80.55 (29.02) — 91.59 (6.06) ——89.40 (7.35) 85.11 (13.10) — 93.18 (5.91) 92.77 (5.26) — 80.53 (17.56)

Hyposmia, n (%) — 89 (28.9) ————54 (36.0) 276 (65.4) — 164 (45.9) — 173 (63.8) 69 (67.0)

Cognitive 26 (5.9) 3 (1.0) 74 (13.0) 29 (7.1) — 55 (16.4) 27 (18.0) 96 (22.7) 11 (9.2) 28 (7.8) 3 (0.9) 77 (27.0) 29 (11.5) impairment, n (%)

Motor — 40 (12.9) 228 (39.9) 103 (25.4) ——4 (2.7) 129 (48.1) 1 (0.8) —— 94 (32.4) 75 (35.4) fluctuation, n (%)

Dyskinesia, n (%) 4 (0.9) 13 (4.2) 207 (36.2) 5 (1.2) ——2 (1.3) 196 (46.4) 0 (0.0) —— 81 (27.6) 44 (22.8)

Continued were dopaminergic therapy naive at baseline; patients in the other cohorts were not. In the primary analysis of 13 cohorts, 17 associations were identiﬁed as signiﬁcant after FDR cor- — — — rection (table 2, and more information in table e-4, links.lww. com/NXG/A173). Overwhelmingly, 10 were associated with GBA variants. In particular, GBA p.E365K (rs2230288) was associated with 2.37- (1.53–3.66) (95% CI) fold higher odds 83 (28.0) 126 (42.6) 138 (46.6) − of having cognitive impairment at baseline (p = 1.09 × 10 4) and 2.78- (1.88–4.11) fold higher hazard ratio (HR) of developing cognitive impairment during follow-up among those who were negative for cognitive impairment at baseline (p = — ——— — ——— — PreCEPT/ PostCEPT ProPark Udall − 2.97 × 10 7). This SNP was also associated with a higher mean − on the HY at 0.10 (0.04–0.16) (p = 1.53 × 10 3), but the test p I2

s Disease Large Simple Study 1; Oslo = Oslo PD study; of homogeneity was rejected ( = 0.017, = 48.9%). In addition, ’ it was associated with the development of an RBD among those 93 (26.1) 23 (6.4) who did not have the disorder at baseline. Other GBA mutations, rkinson p.N370S (rs767763715) and p.T408M (rs75548401), were both associated with a higher HR of reaching HY3 (4.59 − [2.60–8.10] for p.N370S [p = 1.58 × 10 7] and 1.93 [1.34–2.78] — — s Disease; HBS = Harvard Biomarkers Study; HY = Hoehn and Yahr scale; ’ − for p.T408M [p = 4.40 × 10 4]). GBA p.N370N was also associated with a higher risk of developing wearing-oﬀ,dyskinesia, and daytime sleepiness. p.T408M was associated with a 6.48 (2.04–20.60) times higher odds ratio (OR) of having an RBD 197 (50.5) 91 (23.3) − symptom at baseline (p = 1.53 × 10 3).

Two LRRK2 variants in our 31 SNPs of interest were signiﬁ- cantly associated with outcomes. LRRK2 p.G2019S 45 (30.0) 295 (69.9) 62 (51.7) 78 (21.8) 25 (16.7) 165 (39.1) 25 (20.8) 55 (15.4) 17 (11.3) 239 (56.6) 29 (24.2) 113 (31.7) 20 (13.3) 49 (11.6) 27 (22.5) 113 (31.7) 73 (22.7) 48 (16.3) 63 (25.0) (rs34637584) was associated with higher odds of having a family history of Parkinson disease (OR 3.54 [1.72–7.29], p = 6.06 × − 10 4), and the T allele of rs76904798 (intergenic at the 59 end of LRRK2) was associated with a higher HR of reaching HY3 (HR − 1.33 [1.16–1.52] for the T allele, p = 5.27 × 10 5).

Age at onset was inversely associated with the Z value of the GRS − s Disease Biomarker Program; PICNICS = Parkinsonism: Incidence and Cognitive and Non-motor heterogeneity In CambridgeShire; ’ (−0.60 [−0.89 to −0.31] years per +1 SD, p = 5.33 × 10 5). —— TMEM175 s Disease study; RBD = REM sleep behavior disorder; RLS = restless legs syndrome; SEADL = Schwab and England Activities of Daily Living

’ Moreover, it was associated with rs34311866 ( p.M393T), the C allele of rs199347 (intronic region of GPNMB), and the G allele of rs1106180 (intronic region of CCDC62). ——— ———— NET-PD LS1 Oslo ParkFit ParkWest PDBP PICNICS PPMI The majority (14/17) of associations showed good accord across cohorts (I2 < 50%), and the forest plots (figures 1–3) also (continued) illustrate this qualitatively. Furthermore, up to 13 iterations of the leave-one-out analysis assessed 15 associations of which out- ———— ———— comes were measured in more than 2 cohorts and showed a small range of betas. The maximum p value of 13 iterations was less than 0.05 for all associations except for rs114138769 (intron of PMVK) and rs76763715 (GBA p.N370S) for wearing-off.A 44 (14.5) 37 (10.9) ff s Disease Rating Scale.

’ meta-analysis with a random e ect model also detected 9 associations after the same FDR correction, although the model is more conservative than a ﬁxed model. 0 (0.0) 4 (1.3) 71 (12.4) 12 (3.0) 22 (14.5) 17 (5.1) 11 (7.3) 71 (16.8) 13 (10.8) 1 (0.3) 0 (0.0) 117 (40.8) 57 (23.0) 11 (2.5) 107 (35.1) 202 (35.1) 5 (1.1) 138 (44.8) ——————— 9 (2.0) 62 (20.3) 12 (2.7) 97 (31.6) 35 (10.9) 40 (9.9) — DATATOP DIGPD HBS s Progression Markers Initiative; ProPark = Profiling Parkinson ’ Discussion Summary characteristics of 13 cohorts We conducted a meta-analysis with 13 longitudinal patient

3.0, n (%) ﬁ

≥ cohorts and identi ed multiple associations between geno- HY Insomnia, n (%) Daytime sleepiness, n (%) RBD, n (%) Constipation, n (%) Continuous variables were summarized as mean (SD). Scale; UPDRS = Unified Parkinson PPMI = Parkinson ParkFit = ParkFit study; ParkWest = the Norwegian ParkWest study; PDBP = Parkinson Depression, n (%) RLS, n (%) MDS = Movement Disorder Society; MMSE = Mini-Mental State Examination; MoCA = Montreal Cognitive Assessment; NET-PD LS = NIH Exploratory Trials in Pa Table 1 Abbreviations: DATATOP = Deprenyl and Tocopherol Antioxidative Therapy of Parkinsonism; DIGPD = Drug Interaction with Genes in Parkinson types and clinical phenotypic characteristics, including

Table 2 Meta-analysis for 13 cohorts and the results of sensitivity analysis

Fixed effect model Leave-one-out analysis Random effect model Known gene or nearest No. of Scale of the Test of Estimate (Min to Outcome rsNo gene cohorts effect Estimate (95% CI) p homogeneity I2 (%) Max) Max p Estimate (95% CI) p

Wearing-off rs114138760 intron_PMVK 9 Multiplicative (HR) 1.66 (1.19 to 2.31) 2.62E-03 0.322 12.58 1.66 (1.44 to 1.81) 6.22E-02 1.65 (1.14 to 2.38) 7.39E-03

Dyskinesia rs76763715 GBA:N370S 8 Multiplicative (HR) 3.01 (1.81 to 5.01) 2.17E-05 0.011 60.53 3.00 (1.98 to 4.05) 2.26E-02 2.49 (1.06 to 5.86) 3.73E-02

HY ≥ 3.0 rs76763715 GBA:N370S 6 Multiplicative (HR) 4.59 (2.60 to 8.10) 1.58E-07 0.654 0.00 4.59 (4.02 to 5.41) 2.00E-05 4.59 (2.60 to 8.10) 1.58E-07a

Wearing-off rs76763715 GBA:N370S 6 Multiplicative (HR) 2.03 (1.28 to 3.21) 2.56E-03 0.021 62.70 2.02 (1.61 to 2.65) 8.67E-02 1.92 (0.85 to 4.33) 1.14E-01

Daytime rs76763715 GBA:N370S 6 Multiplicative (HR) 3.28 (1.69 to 6.34) 4.24E-04 0.467 0.00 3.30 (2.85 to 4.38) 3.75E-03 3.28 (1.69 to 6.34) 4.24E-04a sleepiness

HY ≥ 3.0 rs75548401 GBA:T408M 8 Multiplicative (HR) 1.93 (1.34 to 2.78) 4.40E-04 0.208 32.43 1.93 (1.70 to 2.41) 1.08E-02 1.96 (1.22 to 3.14) 5.22E-03

pRBD (baseline) rs75548401 GBA:T408M 2 Multiplicative (OR) 6.48 (2.04 to 20.60) 1.53E-03 0.118 59.06 ——6.25 (1.02 to 38.20) 4.72E-02

HY rs2230288 GBA:E365K 12 Continuous 0.10 (0.04 to 0.16) 1.53E-03 0.017 48.90 0.10 (0.08 to 0.11) 1.02E-02 0.11 (0.02 to 0.21) 1.88E-02

Cognitive rs2230288 GBA:E365K 8 Multiplicative (OR) 2.37 (1.53 to 3.66) 1.09E-04 0.794 0.00 2.37 (2.20 to 2.59) 8.57E-04 2.37 (1.53 to 3.66) 1.09E-04a impairment (baseline)

Cognitive rs2230288 GBA:E365K 9 Multiplicative (HR) 2.78 (1.88 to 4.11) 2.97E-07 0.555 0.00 2.78 (2.41 to 2.98) 5.08E-05 2.78 (1.88 to 4.11) 2.97E-07a impairment

pRBD rs2230288 GBA:E365K 2 Multiplicative (HR) 2.57 (1.43 to 4.63) 1.69E-03 0.665 0.00 ——2.57 (1.43 to 4.63) 1.69E-03a

Age at onset rs34311866 TMEM175: 13 Continuous −0.72 (−1.21 to −0.23) 3.87E-03 0.515 0.00 −0.72 (−0.83 to −0.58) 2.83E-02 −0.72 (−1.21 to −0.23) M393T

Age at onset rs199347 intron_GPNMB 12 Continuous 0.70 (0.27 to 1.14) 1.42E-03 0.824 0.00 0.70 (0.60 to 0.77) 1.12E-02 0.70 (0.27 to 1.14) 1.42E-03a

HY ≥ 3.0 rs76904798 5_LRRK2 13 Multiplicative (HR) 1.33 (1.16 to 1.52) 5.27E-05 0.049 43.15 1.33 (1.26 to 1.43) 1.64E-03 1.34 (1.11 to 1.63) 2.80E-03a

Family history rs34637584 LRRK2:G2019S 8 Multiplicative (OR) 3.54 (1.72 to 7.29) 6.06E-04 0.856 0.00 3.53 (2.78 to 3.98) 1.66E-02 3.54 (1.72 to 7.29) 6.06E-04a

Age at onset rs11060180 intron_CCDC62 13 Continuous 0.62 (0.21 to 1.03) 3.32E-03 0.054 42.60 0.62 (0.49 to 0.75) 2.74E-02 0.55 (−0.00 to 1.11) 5.14E-02

Age at onset GRS 13 Continuous −0.60 (−0.89, −0.31) 5.33E-05 0.749 0.00 −0.60 (−0.65, −0.52) 9.02E-04 −0.60 (−0.89, −0.31) 5.33E-05a

Abbreviations: FDR = false discovery rate; GRS = genetic risk score; HR = hazard ratio; HY = Hoehn and Yahr scale; OR = odds ratio; pRBD = possible REM sleep behavior disorder. pRBD was only available in 2 cohorts and a leave-one-out analysis was not conducted for this outcome. a Significant after FDR adjustment in a random effect model. Figure 1 Forest plots for GBA (p.N370S and p.T408M) variants and symptoms of Parkinson disease

DATATOP = Deprenyl and Tocopherol Antioxidative Therapy of Parkinsonism; DIGPD = Drug Interaction with Genes in Parkinson’s Disease; HBS = Harvard Biomarkers Study; NET-PD_LS1 = NIH Exploratory Trials in Parkinson’s Disease Large Simple Study 1; Oslo = Oslo PD study; ParkFit = ParkFit study; ParkWest = the Norwegian ParkWest study; PDBP = Parkinson’s Disease Biomarker Program; PICNICS = Parkinsonism: Incidence and Cognitive and Non-motor heterogeneity In CambridgeShire; PPMI = Parkinson’s Progression Markers Initiative; PreCEPT/PostCEPT = Parkinson Research Examination of CEP-1347 Trial with a subsequent prospective study; ProPark = Profiling Parkinson’s Disease study; Udall = Morris K. Udall Centers for Parkinson’s Research. * Indicates Beta in a Cox model; ** indicates Beta in a logistic model at baseline. progression rates. Among these, GBA coding variants showed In addition, we found associations between GBA variants and clear associations with the rate of cognitive decline (binomial RBD and daytime sleepiness. A previous cross-sectional study outcome or UPDRS part 1 score) and motor symptom pro- with 120 Ashkenazi-Jewish patients reported a higher frequency – gression (HY, HY3), consistent with previous studies.12,21 25 of RBDSQ-detected RBD symptoms in GBA variant carriers.26

Neurology.org/NG Neurology: Genetics | Volume 5, Number 4 | August 2019 7 Figure 2 Forest plots for GBA (p.E365K) variants and symptoms of Parkinson disease

Our finding suggests that GBA is associated not only with note that with 63 carriers for p.N370S, 166 for p.T408M, and baseline clinical presentation but also with disease progression. 217 for p.E365K, we have a reasonable power, but the number is yet not enough. And this may affect the results in seemingly An association between GBA and daytime sleepiness has been different magnitudes of associations and the association for rarely documented. One study reported an association between different traits per variants (e.g., motor complications with sleep problems (as assessed by the Parkinson’s Disease Sleep p.N370S and cognitive impairment with p.E365K). Another Scale) and GBA.27 However, this scale is a combined measure of possible explanation is that although the effects are associated daytime sleepiness and other aspects of sleep problems. with the same gene, the biological activity or molecular mechanism could be different. Such an example has already Finally, a GBA variant (p.N370S) was also associated with been reported for LRRK2 p.G2019S and p.G2385R.30 treatment-related complications of wearing-off and dyskinesia. Two studies have reported the association of GBA variants Aside from GBA variants, the associations between close with these complications, with 1 positive and 1 negative intergenic (59_end) variant of LRRK2, rs76904798, and the result.28,29 The negative result may be due to insufficient faster development of motor symptom, and the intronic re- power with only 19 patients with GBA mutations. gion variant of PMVK, rs114138760, and the development of wearing-off, were significant. This variant is 4.3 kb upstream Overall, our study provides a distinct clinical profile of from the 59 end of LRRK2 and reported to be associated with patients with GBA variants compared with those without. We LRRK2 gene expression changes in recent blood cis-

8 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG Figure 3 Forest plots for non-GBA risk variants/genetic risk score and symptoms or features of Parkinson disease

expression quantitative trait loci (eQTL) study from the is possible that its association was through a similar mechanism eQTLGen Consortium.31 In contrast, we did not find an as- as GBA. Including the results of cross-sectional analysis, the sociation between rs34637584, LRRK2 coding mutation associations of age at onset with rs34311866 (TMEM175, (p.G2019S) and motor progression. The p.G2019 variant is p.M393T), rs199347 (intron of GPNMB), and rs11060180 a rare variant (MAF 0.5% in our study), and our sample size (intron of CCDC62) were found. TMEM175 has been repor- was not adequate barring an extremely large effect size. The ted to impair lysosomal and mitochondrial function and in- intronic region variant of PMVK, rs114138760, and the de- crease α-synuclein aggregation,32 although no functional data velopment of wearing-off was another finding. The biological for this missense variant were studied. Of interest, the variant effect of PMVK on PD has not been reported, but the variant has recently been reported in another study as being associated is also located at close proximity of the GBA-SYT11 locus, so it with the age at onset.33 rs199347 is an eQTL increasing the

Neurology.org/NG Neurology: Genetics | Volume 5, Number 4 | August 2019 9 brain expression of GPNMB,34 suggesting a causal link. Re- Trials in Parkinson’s Disease Large Simple Study 1; Oslo PD garding rs1160180, no functional data are available in this locus. study; ParkFit study; The Norwegian ParkWest study (ParkWest); Parkinson’s Disease Biomarker Program (PDBP); We also evaluated the association between genetic risk var- Parkinsonism: Incidence and Cognitive and Non-motor iants and clinical outcomes by 2-step meta-analysis. This heterogeneity In CambridgeShire (PICNICS); Parkinson’s analysis is exploratory, and we acknowledge that this is biased progression markers initiative (PPMI); Parkinson Study toward the null due to power issues when partitioning studies Group: Parkinson Research Examination of CEP-1347 Trial randomly. However, we believe that it is helpful to assess the (PreCEPT) and its following study (PostCEPT); Profiling rigorousness of the associations we found in the primary Parkinson’s disease study (ProPark); and Morris K. Udall analysis and to explore potential missed associations. Centers for Parkinson’s Research (Udall). They also thank the following grants and financial supporters of above studies; A strength of the current study was its design, incorporating DATATOP was supported by a Public Health Service grant multiple distinct independent Parkinson disease cohorts with (NS24778) from the NINDS; by grants from the General longitudinal follow-ups. Although the cohorts contained Clinical Research Centers Program of the NIH at Columbia patients at different disease stages, and some of the definition University (RR00645), the University of Virginia (RR00847), of outcomes were not identical, we analyzed each cohort the University of Pennsylvania (RR00040), the University of separately and combined the results. Thus, the significant Iowa (RR00059), Ohio State University (RR00034), Massa- findings are consistent and applicable to the wider Parkinson chusetts General Hospital (RR01066), the University of disease populations. The forest plots showed that most of the Rochester (RR00044), Brown University (RR02038), Oregon estimates agree with each other despite the relative differences Health Sciences University (RR00334), Baylor College of in the cohort characteristics. Another strength is the size of the Medicine (RR00350), the University of California (RR00827), study. The total number of genotyped and phenotyped Johns Hopkins University (RR00035), the University of patients with Parkinson disease (N = 4,307) is one of the Michigan (RR00042), and Washington University largest to date for an investigation of disease progression. (RR00036), the Parkinson’s Disease Foundation at Columbia-Presbyterian Medical Center, the National Parkin- The limitations of our study were as follows. First, we only son Foundation, the Parkinson Foundation of Canada, the included patients of European ancestry. It is uncertain whether United Parkinson Foundation, Chicago, the American Parkin- the associations in the current study are also applicable to people son’s Disease Association, New York, and the University of from different ethnic backgrounds and further research is Rochester; DIGPD is supported by Assistance Publique needed. Second, the current analysis could not distinguish cau- Hopitauxˆ de Paris, funded by a grant from the French Ministry sality, only basic associations. Different approaches, such as of Health (PHRC 2008, AOM08010) and a grant from the molecular-level assessment and Mendelian randomization, are Agence Nationale pour la SécuritédesMédicaments (ANSM crucial. Third, interaction effects between genes and other factors 2013); HBS is supported by the Harvard NeuroDiscovery are another important research target not addressed in this re- Center, Michael J Fox Foundation, NINDS U01NS082157, port because of power constraints. For example, gene-by- U01NS100603, and the Massachusetts Alzheimer’sDisease smoking interactions for Parkinson disease were indicated re- Research Center NIA P50AG005134; NET-PD_LS1 was cently35 and highlight the importance of correctly modeling supported by NINDS grants U01NS043128; OSLO is gene-environment interactions. Finally, compared with the typ- supported by the Research Council of Norway and South- ical GWAS analysis (which includes tens of thousands of cases), Eastern Norway Regional Health Authority; ParkFit is the number of participants was small, and the outcomes of in- supported by ZonMw (the Netherlands Organization for terest were not as simple or easily defined as with case-control Health Research and Development [75020012]) and the distinctions in GWAS. Acknowledging the limitations, the list of Michael J Fox Foundation for Parkinson’s research, VGZ associations provided here is valuable as a foundation for further (health insurance company), GlaxoSmithKline, and the studies and as an example that illustrates the potential of efforts National Parkinson Foundation; ParkWest is supported by to define the genetic basis of variability in presentation and the Research Council of Norway, the Western Norway course. Accounting for this variability, even in part, has the po- Regional Health Authority, Stavanger University Hospital tential to positively affect etiology-based clinical trials by re- Research Funds, and the Norwegian Parkinson’sDisease ducing variability between placebo and treatment groups and by Association; PDBP is a consortium with NINDS initiative; providing better predictions of expected individual progression. PICNICS has received funding from the Cure Parkinson’s Trust, the Van Geest Foundation and is supported by the NIH Acknowledgment Research Cambridge Biomedical Research Centre; PPMI is The authors thank all study participants and their family, supported by the Michael J Fox Foundation for Parkinson’s investigators, and members of the following studies: Parkinson research; PreCEPT and PostCEPT were funded by NINDS Study Group: Deprenyl and Tocopherol Antioxidative 5U01NS050095-05, Department of Defense Neurotoxin Therapy of Parkinsonism (DATATOP); Drug Interaction Exposure Treatment Parkinson’sResearchProgram(Grant with Genes in Parkinson’s Disease (DIGPD); Harvard Number: W23RRYX7022N606), the Michael J Fox Founda- Biomarkers Study (HBS); NET-PD_LS1, NIH Exploratory tion for Parkinson’s Research, Parkinson’s Disease Foundation,

10 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG Lundbeck Pharmaceuticals. Cephalon Inc, Lundbeck Inc, John Vereniging. D. Weintraub—consultancies: Acadia, Alkahest, Blume Foundation, Smart Family Foundation, RJG Founda- Anavex Life Sciences, BlackThorn Therapeutics, Bracket, tion, Kinetics Foundation, National Parkinson Foundation, Clintrex LLC, Sunovion, Theravance Biopharma, and the Amarin Neuroscience LTD, CHDI Foundation Inc, NIH CHDI Foundation. R.A. Barker—consultancies: CDI and (NHGRI and NINDS), and Columbia Parkinson’sDisease Oxford Biomedica; royalties: Springer and Wiley; grants: EU, Research Center; ProPARK is funded by the Alkemade-Keuls NIHR, PUK, CPT, Rosetrees Trust, MRC, Wellcome Trust, Foundation, Stichting Parkinson Fonds, Parkinson Vereniging, and Evelyn Trust. C.H. Williams-Gray—grants: MRC Clini- and The Netherlands Organization for Health Research and cian Scientist fellowship, the NIHR Cambridge Biomedical Development; Udall is supported by the NINDS. Research Centre, the Michael J Fox Foundation, the Rosetr- ees Trust, the Evelyn Trust, and Addenbrookes Charitable Study funding Trust. B.P. van de Warrenburg—advisory boards: member of This study is supported by the Intramural Research Program, medical advisory boards and patient organizations; royalties: the National Institute on Aging (NIA, Z01-AG000949-02), Reed Elsevier (for chapter in Dutch Neurology textbook); Biogen Idec, and the Michael J Fox Foundation for Parkin- grants: Radboud University Medical Centre, ZonMW, Her- son’s Research. The funders of the study had no role in study senstichting, and Bioblast Pharma. J.J. Van Hilten—grants: design, data collection, data analysis, data interpretation, or Alkemade-Keuls Foundation, Stichting Parkinson Fonds, writing of the report. The authors had full access to the data in Parkinson Vereniging, and The Netherlands Organisation for the study and had final responsibility for the decision to Health Research and Development. C.R. Scherzer—grants: submit for publication. NIH grants U01NS082157, U01NS095736, and U01NS100603. M.A. Nalls—consultancies: Lysosomal Therapies Inc., Vivid Disclosure Genomics Inc., Kleiner Perkins Caufield & Byers, and Michael J. H. Iwaki—grants: Michael J Fox Foundation. J. Maple- Fox Foundation. Go to Neurology.org/NG for full disclosures. Grødem—grants: Norwegian Parkinson’s Disease Associa- tion. J.-C. Corvol—advisory boards: Biogen, Air Liquide, Publication history BrainEver, Theranexus, BMS, Zambon, Pfizer, Ipsen, and Received by Neurology: Genetics November 13, 2018. Accepted in final AbbVie; grants: MJFF, Actelion, and Ipsen. L. Pihlstrøm— form April 30, 2019. grants: Norwegian Health Association, South-East Norway Regional Health Authority, Norwegian Parkinson Research Fund, and Michael J. Fox Foundation. K.-D.H. Nguyen— stock ownership in medically related fields: Biotech/ Pharmaceutical Industry. K.M. Scott—grants: Wellcome Appendix Authors Trust PhD Fellowship. V.M. Van Deerlin—grants: NIH NS- Name Location Role Contributions 053488. A.G. Day-Williams—stock ownership in medically related fields: Biogen and Merck. A. Brice—advisory boards: Hirotaka Iwaki, Laboratory of Author Literature search; MD, PhD Neurogenetics, study design; data FWO and ERC; grants: JPND, ANR, Eranet Neuron, and National Institute on analysis; data Association France Parkinson. A.J. Noyce—honoraria: Bri- Aging, National interpretation; and ’ Institutes of Health, writings tannia Pharmaceuticals; grants: Parkinson s UK (G-1606). Bethesda, MD J.R. Evans—advisory boards: AbbVie, Global Kinetics, and Cornelis Laboratory of Author Literature search; Allergan; honoraria: UCB, Allergan, and AbbVie. K. Blauwendraat, Neurogenetics, data analysis; data Estrada—stock ownership in medically related fields: Biogen. PhD National Institute on interpretation; and — Aging, National critical review D.K. Simon consultancies: Lysosomal Therapeutics, Inc.; Institutes of Health, advisory boards: Weston Brain Institute; honoraria: Parkin- Bethesda, MD son Study Group, Harvard Medical School, Michael J Fox Hampton L. Laboratory of Author Critical review Foundation, and Biogen; grants: NIH, Weston Brain In- Leonard, MS Neurogenetics, National Institute on stitute, Mission Therapeutics, Inc., and BioElectron Tech- Aging, National nologies. B. Ravina—stock ownership in medically related Institutes of Health, fields: Voyager Therapeutics; consultancies: Michael J Fox Bethesda, MD Foundation. M. Toft—honoraria: Roche; grants: Research Ganqiang Liu, Precision Neurology Author Data collection and Council of Norway, South-Eastern Norway Regional Health PhD Program, Harvard critical review Medical School, Authority, and Michael J. Fox Foundation. B.R. Bloem— Brigham and consultancies: AbbVie and Zambon; advisory boards: Michael Women’s Hospital, Boston, MA J Fox Foundation; honoraria and speaker fees: AbbVie, Zambon, and Bial; grants: The Netherlands Organization for Jodi Maple- The Norwegian Author Data collection and fi Grødem, PhD Centre for Movement critical review Scienti c Research, the Michael J Fox Foundation, UCB, Disorders, Stavanger AbbVie, the Stichting Parkinson Fonds, the Hersenstichting University Hospital, Nederland, the Parkinson’s Foundation, Verily Life Sciences, Stavanger, Norway the Topsector Life Sciences and Health, and the Parkinson Continued

Neurology.org/NG Neurology: Genetics | Volume 5, Number 4 | August 2019 11 Appendix (continued) Appendix (continued)

Name Location Role Contributions Name Location Role Contributions

Jean- Assistance-Publique Author Data collection and Vivianna M. Department of Author Data collection and Christophe Hopitauxˆ de Paris, critical review Van Deerlin, Pathology and critical review Corvol, MD, ICM, INSERM UMRS MD, PhD Laboratory Medicine, PhD 1127, CNRS 7225, Center for ICM, Department of Neurodegenerative Neurology and CIC Disease Research, Neurosciences, Pitié- Parelman School of Salpêtrière Hospital, Medicine at the Paris, France University of Pennsylvania, Lasse Department of Author Data collection and Philadelphia, PA Pihlstrøm, MD, Neurology, Oslo critical review PhD University Hospital, Dena G. Laboratory of Author Data collection and Oslo, Norway Hernandez, Neurogenetics, critical review PhD National Institute on Marlies van Department of Author Data collection and Aging, National Nimwegen, Neurology, Donders critical review Institutes of Health, PhD Institute for Brain, Bethesda, MD Cognition, and Behaviour, Radboud Aaron G. Day- Genetics and Author Data collection and University Medical Williams, PhD Pharmacogenomics, critical review Centre, Nijmegen, Merck Research The Netherlands Laboratory, Boston, MA Samantha J. Michael J Fox Author Data collection and Hutten, PhD Foundation, New critical review Alexis Brice, Institut du cerveau et Author Data collection and York, NY MD de la moelle épinière critical review ICM, Paris, France Khanh-Dung H. Translational Author Data collection and Nguyen, PhD Genome Sciences, critical review Guido Alves, The Norwegian Author Data collection and Biogen, Cambridge, MD, PhD Centre for Movement critical review MA Disorders, Stavanger University Hospital, Jacqueline Department of Author Data collection and Stavanger, Norway Rick, PhD Neurology University critical review of Pennsylvania, Alastair J. Preventive Neurology Author Data collection and Philadelphia, PA Noyce, MRCP, Unit, Wolfson critical review PhD Institute of Shirley Eberly, Department of Author Data collection and Preventive Medicine, MS Biostatistics and critical review Queen Mary Computational University of London, Biology, University of London, UK Rochester, Rochester, NY Ole-Bjørn Department of Author Data collection and Tysnes, MD, Neurology, critical review Faraz Faghri, Laboratory of Author Data collection and PhD Haukeland University MS Neurogenetics, critical review Hospital, Bergen, National Institute on Norway Aging, National Institutes of Health, Jonathan R. Department of Author Data collection and Bethesda, MD Evans, MRCP, Neurology, critical review PhD Nottingham Peggy Auinger, Department of Author Data collection and University NHS Trust, MS Neurology, Center for critical review Nottingham, UK Health + Technology, University of David P. Breen, Centre for Clinical Author Data collection and Rochester, Rochester, MRCP, PhD Brain Sciences, critical review NY University of Edinburgh, Kirsten M. Department of Author Data collection and Edinburgh, Scotland Scott, MRCP, Clinical critical review MPhil Neurosciences, Karol Estrada, Translational Author Data collection and University of PhD Genome Sciences, critical review Cambridge, John van Biogen, Cambridge, Geest Centre for MA Brain Repair, Cambridge, UK Claire E. Wegel, Department of Author Data collection and MPH Medical and critical review Ruwani Department of Author Data collection and Molecular Genetics, Wijeyekoon, Clinical critical review Indiana University, MRCP Neurosciences, Indianapolis, IN University of Cambridge, John van Fabrice Institut du cerveau Author Data collection and Geest Centre for Danjou, MD, et de la moelle critical review Brain Repair, PhD épinière ICM, Paris, Cambridge, UK France

12 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG Appendix (continued) Appendix (continued)

Name Location Role Contributions Name Location Role Contributions

David K. Simon, Department of Author Data collection and Andrew B. Laboratory of Author Study design and MD, PhD Neurology, Beth critical review Singleton, PhD Neurogenetics, critical review Israel Deaconess National Institute on Medical Center, Aging, National Boston, MA Institutes of Health, Bethesda, MD Bernard Voyager Author Data collection and Ravina, MD Therapeutics, critical review Mike A. Nalls, Laboratory of Author Study design; data Cambridge, PhD Neurogenetics, analysis; data MA National Institute on interpretation; and Aging, National critical review Mathias Toft, Department of Author Data collection and Institutes of Health, MD, PhD Neurology, Oslo critical review Bethesda, MD University Hospital, Oslo, Norway

Peter Heutink, German Center for Author Data collection and References PhD Neurodegenerative critical review 1. Elbaz A, Bower JH, Maraganore DM, et al. Risk tables for parkinsonism and Par- Diseases-Tubingen, kinson’s disease. J Clin Epidemiol 2002;55:25–31. Tuebingen, Germany 2. Chaudhuri KR, Healy DG, Schapira AHV; National Institute for Clinical Excellence. Non-motor symptoms of Parkinson’s disease: diagnosis and management. Lancet Bastiaan R. Department of Author Data collection and Neurol 2006;5:235–245. Bloem, MD, Neurology, Donders critical review 3. Lewis SJG, Foltynie T, Blackwell AD, Robbins TW, Owen AM, Barker RA. Het- PhD Institute for Brain, erogeneity of Parkinson’s disease in the early clinical stages using a data driven Cognition, and approach. J Neurol Neurosurg Psychiatry 2005;76:343–348. Behaviour, Radboud 4. Satake W, Nakabayashi Y, Mizuta I, et al. Genome-wide association study identifies University Medical common variants at four loci as genetic risk factors for Parkinson’s disease. Nat Genet Centre, Nijmegen, 2009;41:1303–1307. The Netherlands 5. Simón-Sánchez J, Schulte C, Bras JM, et al. Genome-wide association study reveals genetic risk underlying Parkinson’s disease. Nat Genet 2009;41:1308–1312. Daniel Department of Author Data collection and 6. Chang D, Nalls MA, Hallgr´ımsdóttir IB, et al. A meta-analysis of genome-wide as- Weintraub, MD Psychiatry, University critical review sociation studies identifies 17 new Parkinson’s disease risk loci. Nat Genet 2017;49: of Pennsylvania 1511–1516. School of Medicine, 7. International Parkinson Disease Genomics Consortium, Nalls MA, Plagnol V, Her- Philadelphia, nandez DG, et al. Imputation of sequence variants for identification of genetic risks for PA Parkinson’s disease: a meta-analysis of genome-wide association studies. Lancet 2011; 377:641–649. Roger A. Department of Author Data collection and 8. Nalls MA, Bras J, Hernandez DG, et al. NeuroX, a fast and efficient genotyping Barker, MRCP, Clinical critical review platform for investigation of neurodegenerative diseases. Neurobiol Aging 2015;36: PhD Neurosciences, 1605.e7–1605.e12. University of 9. Nalls MA, Escott-Price V, Williams NM, et al. Genetic risk and age in Parkinson’s Cambridge, disease: continuum not stratum. Mov Disord 2015;30:850–854. Cambridge, 10. Nalls MA, McLean CY, Rick J, et al. Diagnosis of Parkinson’s disease on the basis of UK clinical and genetic classification: a population-based modelling study. Lancet Neurol 2015;14:1002–1009. Caroline H. Department of Author Data collection and 11. Pankratz N, Beecham GW, DeStefano AL, et al. Meta-analysis of Parkinson’s disease: Williams-Gray, Clinical critical review identification of a novel locus, RIT2. Ann Neurol 2012;71:370–384. MRCP, PhD Neurosciences, 12. Davis AA, Andruska KM, Benitez BA, Racette BA, Perlmutter JS, Cruchaga C. Var- University of iants in GBA, SNCA, and MAPT influence Parkinson disease risk, age at onset, and Cambridge, progression. Neurobiol Aging 2016;37:209.e1–209.e7. Cambridge, 13. Picillo M, Pellecchia MT, Erro R, et al. The use of university of Pennsylvania Smell identi- UK fication test in the diagnosis of Parkinson’s disease in Italy. Neurol Sci 2014;35:379–383. 14. Hoops S, Nazem S, Siderowf AD, et al. Validity of the MoCA and MMSE in the Bart P. van de Department of Author Data collection and detection of MCI and dementia in Parkinson disease. Neurology 2009;73: Warrenburg, Neurology, Donders critical review 1738–1745. MD, PhD Institute for Brain, 15. Verbaan D, van Rooden SM, van Hilten JJ, Rijsman RM. Prevalence and clinical Cognition, and profile of restless legs syndrome in Parkinson’s disease. Mov Disord 2010;25: Behaviour, Radboud 2142–2147. University Medical 16. Goodarzi Z, Mrklas KJ, Roberts DJ, Jette N, Pringsheim T, Holroyd-Leduc J. Centre, Nijmegen, Detecting depression in Parkinson disease: a systematic review and meta-analysis. The Netherlands Neurology 2016;87:426–437. 17. Simuni T, Caspell-Garcia C, Coffey C, et al. Correlates of excessive daytime Jacobus J. Van Department of Author Data collection and sleepiness in de novo Parkinson’s disease: a case control study. Mov Disord Hilten, MD, Neurology, Leiden critical review 2015;30:1371–1381. PhD University Medical 18. Boeve BF, Molano JR, Ferman TJ, et al. Validation of the Mayo sleep questionnaire to Center, Leiden, The screen for REM sleep behavior disorder in an aging and dementia cohort. Sleep Med Netherlands 2011;12:445–453. 19. Nomura T, Inoue Y, Kagimura T, Uemura Y, Nakashima K. Utility of the REM sleep Clemens R. Precision Neurology Author Data collection and behavior disorder screening questionnaire (RBDSQ) in Parkinson’s disease patients. Scherzer, MD Program, Harvard critical review Sleep Med 2011;12:711–713. Medical School, 20. Boeve BF, Molano JR, Ferman TJ, et al. Validation of the Mayo sleep questionnaire to Brigham and screen for REM sleep behavior disorder in a community-based sample. J Clin Sleep Women’s Hospital, Med 2013;9:475–480. Boston, 21. Davis MY, Johnson CO, Leverenz JB, et al. Association of GBA mutations and the MA E326K polymorphism with motor and cognitive progression in Parkinson disease. JAMA Neurol 2016;73:1217–1224.

Neurology.org/NG Neurology: Genetics | Volume 5, Number 4 | August 2019 13 22. Winder-Rhodes SE, Evans JR, Ban M, et al. Glucocerebrosidase mutations influence 29. Jesús S, Huertas I, Bernal-Bernal I, et al. GBA variants influence motor and non-motor the natural history of Parkinson’s disease in a community-based incident cohort. Brain features of Parkinson’s disease. PLoS One 2016;11:e0167749. 2013;136:392–399. 30. Marras C, Alcalay RN, Caspell-Garcia C, et al. Motor and nonmotor heteroge- 23. Brockmann K, Srulijes K, Pflederer S, et al. GBA-associated Parkinson’s disease: neity of LRRK2-related and idiopathic Parkinson’s disease. Mov Disord 2016; reduced survival and more rapid progression in a prospective longitudinal study. Mov 31:1192–1202. Disord 2015;30:407–411. 31. Võsa U, Claringbould P, Westra HJ, et al. Unraveling the polygenic architecture of 24. Liu G, Boot B, Locascio JJ, et al. Specifically neuropathic Gaucher’s mutations ac- complex traits using blood eQTL meta-analysis. bioRxiv Epub 2018 Oct 19. celerate cognitive decline in Parkinson’s. Ann Neurol 2016;80:674–685. 32. Jinn S, Drolet RE, Cramer PE, et al. TMEM175 deficiency impairs lysosomal and 25. Liu G, Locascio JJ, Corvol JC, et al. Prediction of cognition in Parkinson’s disease with mitochondrial function and increases α-synuclein aggregation. Proc Natl Acad Sci U S a clinical-genetic score: a longitudinal analysis of nine cohorts. Lancet Neurol 2017; A 2017;114:2389–2394. 16:620–629. 33. Blauwendraat C, Heilbron K, Vallerga CL, Bandres-Ciga S, Coelln Rvon, Pihlstrøm L. 26. Gan-Or Z, Mirelman A, Postuma RB, et al. GBA mutations are associated with rapid Parkinson disease age at onset GWAS: defining heritability, genetic loci and α-syn- eye movement sleep behavior disorder. Ann Clin Transl Neurol 2015;2:941–945. uclein mechanisms. Mov Disord Epub 2019 Apr 7. 27. Brockmann K, Srulijes K, Hauser AK, et al. GBA-associated PD presents with non- 34. UKBEC; Murthy MN, Blauwendraat C, Guelfi S, et al. Increased brain expression of motor characteristics. Neurology 2011;77:276–280. GPNMB is associated with genome wide significant risk for Parkinson’s disease on 28. Oeda T, Umemura A, Mori Y, et al. Impact of glucocerebrosidase mutations on motor chromosome 7p15.3. Neurogenetics 2017;18:121–133. and nonmotor complications in Parkinson’s disease. Neurobiol Aging 2015;36: 35. Lee PC, Ahmed I, Loriot MA, et al. Smoking and Parkinson disease: evidence for 3306–3313. gene-by-smoking interactions. Neurology 2018;90:e583–e592.

14 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG ARTICLE OPEN ACCESS New family with HSPB8-associated autosomal dominant rimmed vacuolar myopathy

Sejad Al-Tahan, DO,* Lan Weiss, MD, PhD,* Howard Yu, BS, Sha Tang, PhD, Mario Saporta, MD, PhD, Correspondence Anna Vihola, PhD, Tahseen Mozaffar, MD, Bjarne Udd, MD, PhD, and Virginia Kimonis, MD Dr. Kimonis [email protected] Neurol Genet 2019;5:e349. doi:10.1212/NXG.0000000000000349 Abstract Objective We clinically and molecularly characterize a new family with autosomal dominant rimmed vacuolar myopathy (RVM) caused by mutations in the HSPB8 gene.

Methods We performed whole-exome and whole-genome sequencing in the family. Western blot and immunocytochemistry were used to analyze 3 patient ﬁbroblasts, and ﬁndings were compared with their age- and sex-matched controls.

Results Affected patients have distal and proximal myopathy, with muscle biopsy showing rimmed vacuoles, muscle fiber atrophy, and endomysial fibrosis typical of RVM. Muscle MRI showed severe relatively symmetric multifocal fatty degenerative changes of the lower extremities. We identified a duplication of C at position 515 of the HSPB8 gene (c.515dupC) by whole-genome sequencing, which caused a frameshift with a predicted alternate stop codon p.P173SFS*43 in all affected individuals, resulting in an elongated protein product. Western blot and immunocytochemistry studies revealed reduced expression of heat shock protein beta 8 in patient fibroblasts compared with control fibroblasts, in addition to disrupted autophagy pathology.

Conclusions We report a novel family with autosomal dominant RVM caused by the c.515dupC mutation of the HSPB8 gene, causing a translational frameshift that results in an elongated protein. Un- derstanding the mechanism for the RVM pathology caused by mutated chaperone will permit novel targeted strategies to alter the natural history progression. As next-generation sequencing becomes more available, additional myopathic families will be identiﬁed with HSPB8 mutations.

*These individuals are co-first authors.

From the Division of Genetics and Genomic Medicine (S.A.-T., L.W., H.Y.), Department of Pediatrics, University of California, Irvine; Opti West (S.A.-T.), West Anaheim Medical Center, Anaheim; Ambry Genetics (S.T.), Mission Viejo, CA; Miller School of Medicine (M.S.), University of Miami, FL; Folkh¨alsan Institute of Genetics and the Department of Medical Genetics (A.V., B.U.), Medicum, University of Helsinki; Neuromuscular Research Center (A.V., B.U.), Tampere University and University Hospital, Neurology, Finland; Neuromuscular Program (T.M.), Department of Neurology, University of California-Irvine, Orange; and Neurology Department (B.U.), Vasa Central Hospital, Finland.

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 Volo Foundation and the IBM research fund. 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 © 2019 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 Glossary CASA = chaperone-assisted selective autophagy; CMT = Charcot-Marie-Tooth; FLNC = ﬁlamin C; HSPB = heat shock protein beta; MRI = Medical Research Council; NCS = nerve conduction study; RVM = rimmed vacuolar myopathy.

Heat shock protein beta 8 (HSPB8), also known as heat shock at Emory University was negative for the following 35 genes: protein 22, is involved in chaperone-assisted selective autophagy AN05, CAPN3, CAV3, COL6A1, COL6A2, COL6A3, DAG1, (CASA) in cells.1 CASA is a process that maintains muscle DES, DMD, DYSF, EMO, FHL1, FKRP, FKTN, FLNC, GAA, cytoskeleton by facilitating the degradation of damaged com- GNE, ISPD, LMNA, MYOT, PLEC, POMGNT1, POMT1, ponents. Disruption of genes in the chaperone-assisted degra- POMT2, SGCA, SGCB, SGCD, SGCG, SMCHD1, SYNE1, dation pathway is implicated in several myopathies. CASA SYNE2, TCAP, TRIM32, TTN, and VCP. interacts with filamin C (FLNC), a flexible actin cross-linker that is present within muscle Z-disks2 and along cytoskeleton- Whole-exome sequencing and segregation analysis was per- associated integrin where it functions to anchor actin during formed by Ambry Genetics.16 Briefly, the sample was pre- mechanical stress.3 DnaJ Heat Shock Protein Family (Hsp40) pared using the IDT xGen Exome Research Panel V1.0 Member B6 (DNAJB6) thought to be the first protein that (Integrated DNA Technologies), sequenced using the Illu- interacts with Heat Shock Protein Family A (Hsp70) Member 8 mina HiSeq4000 sequencer (Illumina, San Diego, CA). Data (HSPA8)4 then complexes with HSPB8. BCL2 Associated processing and interpretation were performed as previously Athanogene 3 (BAG3) is a coupling factor that mediates HSPA8 described.16 and HSPB8 interaction.1 Mutations of several CASA-associated proteins are known to be associated with myopathies: DNAJB6 Subsequent genome sequencing was performed by Otoge- is associated with limb-girdle muscular dystrophy 1D,5 BAG3,6,7 netics Corporation (Atlanta, GA) and the data analyzed by and FLNC with skeletal and cardiac myopathies.8,9 Variantyx, Inc. (Framingham, MA). For sequencing, libraries were generated using the Truseq Nano DNA Library Sample Previous studies associated HSPB8 mutations with Charcot- preparation Kit and sequenced on the Illumina HiSeq X Ten Marie-Tooth (CMT) type 2L10,11 and distal hereditary motor system (Illumina, USA) to achieve average read depth of neuropathy type IIa.12 It was only after an exome analysis that about 30X. Variants were detected, annotated, analyzed, and the association with distal myofibrillar and rimmed vacuolar reported using the Genomic Intelligence platform (version 1/ neuromyopathy was established,13 identifying a heterozygous .13.2.0) integrating the knowledge in a variety of databases change in HSPB8 in 2 families: c.421A>G; p.K141E in family and tools (Variantyx Inc). Sanger sequencing was used to 1 and the c.515 insC p.P173SfsX43 in family 2. Affected confirm the HSPB8 variant in the proband and for segregation patients had a distal neuromyopathy that showed myofibrillar studies in his family. aggregates and rimmed vacuoles combined with a clear neurogenic component on biopsy and EMG studies. Recently, HSPB8 protein study and autophagic activity of a c.508_509delCA mutation was identified, predicted to mutant HSPB8 produce a frameshift mutation leading to alteration of the last Because of the unavailability of patient myoblasts, we studied 27 amino acids and elongation of the polypeptide by 17 res- patient fibroblasts from the proband, his mother, and mater- idues in 3 families.14 Finally, a family was identified with nal uncle derived from skin biopsy samples (P1: 64-001, P2: c.421A>G p.K141E HSPB8 mutations manifesting with pes 64-002, P3: 64-005). Fibroblasts of the 3 age- and sex- cavus, hammer toes, and muscle atrophy with biopsy findings matched healthy controls from Coriell Repository (C1: including rimmed vacuoles and myopathic changes.15 Here, GM22246, C2: AG12786, C3: AM03529A) were used to test we describe a new family with an HSPB8 p.Pro173fs mutation the HSPB8 expression level for comparison. Western blotting manifesting as rimmed vacuolar myopathy (RVM). and immunocytochemistry studies using HSPB8 antibody (Millipore Sigma HPA015876) were performed as previously described.17 We also monitored the autophagic activity of Methods mutant HSPB818 using autophagosomal markers LC3B Standard protocol approvals, registrations, (Abcam, ab48394), P62/SQSTM1 (ab56416), and BAG3 and patient consents (ab47124) antibodies. This study was approved by the University of California Irvine Institutional Review Board (#2009-1005). All subjects pro- HSPB8 protein expression under heat vided approved consent for the studies. The study is also listed shock conditions in ClinicalTrials.gov (Identifier: NCT01353430). Under stress conditions such as heat shock, the concentration of small heat shock proteins such as HSPB8 would be Molecular studies expected to increase to perform their chaperone activity of Extensive gene panel of the genes involved in neuromuscular binding to misfolded proteins and blocking the formation of diseases performed by EGL Genetics Diagnostic Laboratory aggregates.19,20 The heat shock experiment was performed as

2 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG follows: fibroblasts were seeded on chambered slides (Ibidi, lordosis. The Medical Research Council (MRC) scale score of IbiTreat 80826) in subconfluent density for 24 hours. The his foot dorsiflexors was 2/5, and that of the extensor hal- slides were sealed with parafilm, placed in an aluminum foil lucis longus was 1/5 on the right and 2/5 on the left. The box, and immersed in 43°C water bath for 30 minutes. Un- MRC scale score of the infraspinatus, deltoid, finger heated control slides were also sealed for 30 minutes but kept extensors, hip flexors, hip extensors, knee flexors, hip in an incubator (37°C). Then, all slides were unsealed, and adductors, and hip abductors was 4/5. Reflexes were di- their recovery was followed for 24 hours in the incubator minished 1/4 bilaterally of the biceps, triceps, brachior- before immunocytochemistry was performed.19,20 adialis, knee, and absent at the ankle. T1-weighted MRI showed severe relatively symmetric multifocal fatty de- Data availability generative changes of the lower extremities, preferentially All the data from this study will be made available. affecting the vastus medialis and intermedius, adductor magnus and semitendinosus at the thigh level, and predominantly anterior and lateral compartments more than Results calf muscle of the lower legs (figure 2, B and C). Laboratory fi Clinical studies studies revealed normal comprehensive metabolic pro le, normal alkaline phosphatase 28 U/L, and elevated total Case 1 cholesterol 237 mg/dL (normal < 150 mg/dL). His medical The proband (IV:8) is a 44-year-old man born to non- history included asymptomatic right bundle branch block, consanguineous Northern European parents of French an- pseudogout, benign prostatic hyperplasia, and rosacea. Al- cestry with a family history of autosomal dominant though he was asymptomatic, pulmonary function studies inheritance of muscle weakness in his mother and his ma- revealed respiratory insufficiency with decreased Forced ternal uncle (figure 1). He was an all-state athlete in his 20s. Vital Capacity (sitting) of 4 L (43% of predicted). His − He had persistent elevated liver enzymes from age 28 years maximum inspiratory pressure was 72 cm H2O. BiPAP was and first noticed lower limb weakness in the mid-30s with recommended for his respiratory insufficiency. The patient progressive bilateral foot drop and occasional falls. At age 35 is taking nonprescribed nutritional supplements. years, he developed foot pain and ankle dorsiflexion weakness for which he wore ankle foot orthoses. At age 42 years, he Case 2 noticed difficulty raising his arms above his head, and at age 43 The proband’s mother (III:14) aged 64 years first presented years, difficulty rising from a chair and climbing stairs. with difficulty raising her arms and weakness in her legs and feet not until age 56 years. She then progressed to having On physical examination, there was mild scapular winging difficulty walking up stairs at age 59 years and developed (figure 2), diffuse muscle weakness noted predominantly in difficulty walking on flat ground and weakness in her hands at the lower extremities, and no sensory loss. He had a wide age 60 years. Her medical history included hyperuricemia, steppage gait related to his bilateral foot drop, was unable to hypertension, and type 2 diabetes mellitus and tendon surgery tandem walk or walk on his heels or toes, and had a lumbar of her right thumb.

Figure 1 Family pedigree showing autosomal dominant inheritance

The filled upper right quadrant indicates myopathy, and the filled upper left quadrant indicates nonspecific neurologic symptoms. The arrow indicates the proband.

Neurology.org/NG Neurology: Genetics | Volume 5, Number 4 | August 2019 3 muscle fiber size variation, numerous atrophic muscle fibers, Figure 2 Muscular involvement of the HSPB8 disease increase of internalized nuclei, and rimmed vacuoles. There were no necrotic fibers, regenerating muscle fibers, or evidence of inflammation (figure 3). No medications were prescribed for this patient. Case 3 The proband’s maternal uncle (III:15) is a 66-year-old man who first became aware of a steppage gait at 42 years and weakness of his great toe and right hand at age 44 years. This progressed to using a cane and difficulty going up stairs at age 46–47 years. At age 52 years, he started swimming instead of running for exercise and starting to use Canadian crutches. At age 59 years, he started using a manual wheelchair, and at age 64 years, he transitioned to a power wheelchair. At age 61 years, he has noticed that his breathing became more difficult.

His medical history includes neurogenic bladder, mild ob- structive sleep apnea, gout and hearing loss of his right ear, and 20 incidences of recurrences of basal cell nevi and squa- mous cell carcinoma of the skin.

On physical examination, there was no pes cavus, skeletal dysplasia, camptocormia, scapular winging, or signs of cranial nerve or cerebellar involvement. He had reduced sensation to light touch below knees bilaterally, and muscle strength examination revealed distal > proximal weakness. The MRC scales were neck flexion 4−/5, neck extension 4/5, shoulder abduction 4−/5, elbow flexion 4−/5 and extension 3+/5, finger abduction 5−/5, hip flexion 2/5 and extension 3+/5, knee flexion 3+/5 and extension 2/5, ankle dorsiflexion 1/5, and plantarflexion 3−/5. EMG and nerve conduction studies (NCSs) revealed abnormal findings suggestive of a myopathic pattern primarily in distal musculature with evidence of a motor neuropathy.

Chest x-ray revealed low lung volumes without evidence of cardiovascular disease. Pulmonary function tests revealed re- strictive lung disease with poor lung capacity with a forced vital capacity of 43%, and BiPAP was recommended. Cardiac CT showed an elevated coronary calcium score of 244 but no (A) Photograph of the proband showing scapular winging and bilateral cardiomyopathy. wasting of deltoids. (B) T1-weighted MRI of the proband showing severe relatively symmetric multifocal fatty degenerative changes of bilateral lower extremities and (C) proximal lower extremities preferentially affecting the Body CT revealed atrophy and fatty infiltration of his psoas vastus medialis and intermedius together with the adductor magnus and semitendinosus on the thigh level, as well as predominately anterior and muscle, benign prostatic hyperplasia, atelectasis with prom- lateral compartments more than soleus muscle of the distal legs. inent interstitial markings of his lung bases bilaterally, renal cyst, gallstone, and a calcified granuloma in his liver. No medications were prescribed for this patient. Physical examination revealed mild distal extremity weakness and diminished 1/4 reflexes bilaterally of her biceps, triceps, Family history of other relatives brachioradialis, and patella. Her Achilles tendon reflex was Family history was significant for the proband’s maternal absent. She had a wide based and unbalanced gait with a lor- grandfather losing his ability to walk at age 38 years and for dotic posture. There was no scapular winging, pes cavus, developing cardiac problems in his fifties. On review of the skeletal dysplasia, respiratory distress, signs of cranial nerve extended family, his grandfather’s brother lost his ability to involvement, or sensory loss. CT of the lung revealed right walk at age 35 years, and the great grandmother lost her ability basilar consolidative infiltrate vs atelectasis at the left lung base. to walk at age 56 years. Five other family members reported nonspecific neurologic symptoms; however, these individuals Muscle biopsy of her left vastus lateralis revealed adipose tested negative for the familial mutation. Figure 4 depicts the tissue infiltration, moderate to severe endomysial fibrosis, pedigree with HSPB8 mutation status indicated with a “+.”

4 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG Figure 3 Muscle biopsy shows rimmed vacuoles

(A) Muscle histology of a biopsy of the left vastus lateralis from case 2 shows the presence of rimmed vacuoles (in 3 fibers), adipose replacement, moderate to severe endomysial fibrosis, muscle fiber size variation, numerous atrophic muscle fibers, and increase in central nuclei. (B) Semithin section stained with toluidine blue showing a single fiber with accumulated autophagic vacuoles, corresponding to rimmed vacuolar changes in muscle cryosections.

Molecular studies inclusion body myopathy. Full exome sequencing and Extensive neuromuscular 35-gene panel testing was negative genotype-phenotype correlation based on exome data in the proband and uncle notably for facioscapulohumeral analysis revealed notable alterations in CCDC78, HSPB8, muscular dystrophy, GNE and VCP associated hereditary MUSK, MYH14, TRIM32, and TTN—none of which was

Figure 4 Sequencing of the HSPB8 gene

(A) Chromatogram from the proband revealing the frameshift mutation c.515dupC p.Pro173fs, which was similarly identified in all affected individuals. (B) The mRNA that results from this mutation has 5 splice variants with size ranging from 27 to 244 aa. Only the first 3 contain an α-crystallin domain and are coding. The red boxes are exons with their size being proportional to their length. The wide colored boxes are introns. Similarly colored introns are identical. The purple introns are upstream open reading frames. The solid black vertical lines indicate validated cap sites at the 59 end, and the dotted black lines indicate validated polyadenylation sites at the 39 end. The clear boxes are untranslated regions. (C) Conservation of the sequence of amino acids within HSPB8 across species reveals a very well conserved protein. (D) The predicted sequence of amino acids in the mutant HSPB8 protein (214 amino acids) is longer than its wild-type counterpart (196 amino acids).

Neurology.org/NG Neurology: Genetics | Volume 5, Number 4 | August 2019 5 considered contributory based on the known molecular and tissue in the proband showed increased internal nuclei, nu- clinical spectrum of these genes at the time of testing. Sub- merous rimmed vacuolar fibers, splitting, cytoplasmic bodies, 13 sequent genome sequencing identifiedamutationintheHSPB8 and moth-eaten fibers. gene identical to the French family 2, which had recently been reported by Ghaoui et al.13 after whole-exome analysis. HSPB8 and autophagy protein expression in mutant HSPB8 fibroblasts HSPB8 p.Pro173fs Western blot densitometry exhibited up to 50% reduction of The HSPB8 gene mutation identified consists of a duplication HSPB8 protein expression level in fibroblasts from patients with of C at position 515, located in coding exon 3 (the last coding the HSPB8 mutation c.515dupC, p.P173SFS*43 mutation com- exon) causing a translational frameshift with a predicted al- pared with control fibroblasts (figure 5, A and B). The expression ternate stop codon (c.515dupC, p.P173Sfs*43). The alter- of autophagosomal marker LC3B and autophagy receptor p62/ ation changes the remaining 25 amino acids of the protein and SQSTM1 was more pronounced in all patients’ fibroblasts in elongates the protein by a further 18 amino acids (figure 4). comparison to those from healthy subjects (figure 5). This variant is not observed in any of the control population databases including gnomAD. The calculated severity score is HSPB8 protein expression under heat 1.0 on a scale of 0.0–1.0. According to Human Gene Mutation shock conditions Database/ClinVar, the HSPB8 gene is associated with CMT As expected, heat shock led to a significant induction of disease 2L; neuropathy, distal hereditary motor, type II; distal HSPB8; however, following the recovery phase, an excessive myopathy and motor neuropathy; and distal hereditary motor amount of HSPB8 protein aggregates was still present in all 3 neuronopathy type 2. Recently, the 515dupC, p.P173Sfs*43 heated fibroblasts from patients compared with the unheated variant had been associated with a progressive distal myopa- ones. This phenomenon was not seen in healthy control cells thy and neuropathy in a proband and his brother. Muscle (figure 6).

Figure 5 Expression of HSPB8, LC3B, p62/SQSTM1, and BAG3 proteins comparison in patients versus controls

(A) Western blot revealed low HSPB8, increased autophagy (LC3, p62) protein expression in patient fibroblasts compared with age- and sex-matched controls. GAPDH was used as a positive loading control. (B) Densitometry analyses of the Western blot results (C) Immunocytochemistry images by confocal microcopy confirmed Western blotting data. Pair 1: patient (P1) and control (C1) fibroblasts; pair 2: P2 and C2; pair 3: P3 and C3.

6 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG Discussion including paraspinal muscles and lower limb muscles including the vastus muscles, tibialis anterior, and extensor HSPB8 is part of the chaperone-assisted selective autophagy digitorum longus similar to those reported previously13 (CASA) complex previously associated with CMT type (table). Most recently, triplets and their mother were reported 10,11 12 2L and distal hereditary motor neuropathy type IIa. with distal and proximal muscle weakness, EMG showed Only recently has the new association of hereditary rimmed mixed myopathic and neurogenic changes, and NCSs showed vacuolar neuromyopathy been described with mutations in findings of an axonal neuropathy.15 Muscle biopsy findings 13–15 the HSPB8 gene. We report 3 patients from a new family were that of a primary myopathy with rimmed vacuoles in with histologic features of myofibrillar myopathy with 20% of patients, and MRI revealed fatty degeneration mostly aggregates and rimmed vacuoles. The onset of the disease is in the quadriceps, adductor magnus, medial gastrocnemius, 13–15 very similar to the other reported families and manifests and peroneus muscles (table). as a myopathy before age 50 years affecting the lower extremities more commonly than the upper extremities. The The family of small heat shock proteins (HSPBs1–10) (22 MRI distribution of muscle involvement in our cohort kDa) represents an important group of ubiquitous proteins includes muscles of the distal lower extremities similar to involved in several key processes of cellular physiology, such previous studies13 that described T1-weighed MRI findings as cell survival, cytoskeletal remodeling, and protein degra- of the lower limb muscles and demonstrated diffuse tissue dation. Aberrant chaperone function is the likely mechanisms changes early in the disease stage progressing later for the pathologic protein aggregation seen in HSPB8 disease. to fatty replacement typical of a myopathy. In addition, The HSPB8 c.K141E mutation associated with CMT impairs there was a previous report of 5 cases from 2 families and chaperone activity,21 which suggests that the defective or a sporadic case, all with the same HSPB8 c.508_509delCA decreased activity is associated with the deteriorating motor (p.Gln170Glyfs*45) mutation predicted to induce a frame- neuron function. HSPB8 expression within motor neurons shift from amino acid 170 leading to modification of the last declines with age,20 which has been hypothesized as the 27 amino acids and to 17 additional amino acids.14 Individuals mechanism of the disease presenting in older ages and pro- had bilateral foot drop starting at age 40 years, histology of gressing with age.18 A contributing factor to the susceptibility muscle demonstrated a dystrophic and myofibrillar myopathy of these patients may be impaired formation of neurites in pattern, and MRI showed fatty changes in several muscles HSPB8 mutants.22 The HSPB8 pathology is similar to the

Figure 6 Expression of HSPB8 protein under heat shock condition

(A) No heat shock condition, (B) heat shock condition. An increase in the amount of HSPB8 protein was present in 3 heated fibroblasts from patients (B1) compared with nonheated fibroblasts (A1). This phenomenon was not observed in control cells (A2 and B2). Pair 1: patient (P1) and control (C1) fibroblasts; pair 2: P2 and C2; pair 3: P3 and C3.

Table Reported families with HSPB8-related myopathy

Age Age at RVM Individual Reference Mutation Relationship Ancestry Sex (y) onset (y) CPK (U/L) EMGa NCS phenotype

1 This study c.515dupC, Proband French M 44 35 430 +/− Normal − p.P173SFS*43

2 Mother F 64 56 370 −/− Normal +

3 Uncle M 66 42 327 +/+ Motor neuropathy with both axonal and demyelinating − neuroapathy

4 13 c.421A. G, Proband—family I English M 30 22 850–2,000 +/+ Length-dependent axonal motor neuropathy predominantly + p.K141E affecting the lower limbs

5 Mother—family I F 56 Childhood 269 +/+ Predominantly lower limb axonal motor neuropathy +

6 c.151insC Proband—family II French M 62 46 369 −/+ Normal − p.P173SfsX43

7 14 c.508_509delCA, Proband—family A II2 French M 57 40 449 +/− Normal NA pGln170Glyfs*45

8 Brother—family A II8 M 55 35 900 +/− +

9 Daughter, family A III2 F 45 40 186 +/− +

10 Proband, family B III1 M 64 40 214 +/− NA

Proband, family C II1 F 74 40 282 +/− NA

12 15 c.421A>G, Proband II1 Italian F 31 20 274 +/+ Reduced compound motor action potential amplitudes in + p.K141E lower limbs

13 Proband sibling II2 Italian M 31 Early 20s 656 +/+ Absent tibial compound motor action potential bilaterally and + reduced peroneal motor compound action potential amplitude. Normal conduction velocities.

14 Proband sibling II3 Italian F 31 Late 20s a −/− Normal at 15 years. −

15 Proband mother I1 Italian F 58 30 b +/+ Axonal motor neuropathy −

Mean 8M/ 51.2 33.3 517.4 42.8% abnormal 47% 7F positive

Abbreviations: CPK = creatinine phosphokinase; NCS = nerve conduction study; RVM = rimmed vacuolar myopathy. a Raised. b Moderately increased CPK. pathology in myopathies due to gene defects in other com- Study funding ponents of the CASA complex such as BAG3 and DNAJB6. The authors thank the Volo Foundation and the Hereditary Inclusion Body Myopathy Foundation for funding this Based on results from animal studies, there has been suggestion project. of a toxic gain-of-function mechanism for HSPB8 associated CMT.23 However, our Western blot densitometry studies Disclosure exhibited a 50% reduction of HSPB8 protein expression level in S. Al-Tahan, L. Weiss, H. Yu, S. Tang, M. Saporta, A. Vihola, patients compared with control fibroblasts (figure 5). Recently, T. Mozaffar, B. Udd, and V. Kimonis report no disclosures. Go it was noted14 that N-terminal targeting of the protein did not to Neurology.org/NG for full disclosures. detect this elongated protein suggesting mRNA decay or protein degradation.24 In addition, HSPB8 was decreased by 60% Publication history by Western blot quantification. The consequences of the effects Received by Neurology: Genetics November 19, 2018. Accepted in final of the mutant protein are difficult to predict based on current form May 16, 2019. knowledge because HSPB8 acts in complexes with other chaperones, especially in muscle tissue. The negative effects on the multiprotein complexes may be explained by a dominant Appendix Authors ff ffi negative e ect, but haploinsu ciency cannot be excluded. The Name Location Role Contribution clear increase of p62 and LC3 in the patient muscle suggests ff Sejad Al- University of Author Major role in the acquisition a harmful e ect of the mutant protein rather than rapid deg- Tahan, California, and analysis of data and radation of the mutant. To clarify this in more detail, a mutant DO Irvine, CA drafted the manuscript for protein–specific antibody would be required to check for intellectual content. eventual accumulation or degradation of the mutant. Lan University of Author Designed and conceptualized Weiss, California, the study; major role in the Most reported mutations, including the c.515dupC mutation, MD, PhD Irvine, CA acquisition and interpretation have occurred within the α-crystallin domain that manifests of data; and drafted the 25 manuscript for intellectual chaperone activity. Previous studies have shown greater binding content. of these HSPB8 mutants to the interacting partner HSPB1. Ex- Howard University of Author Acquisition of data. pression of mutant HSPB8 containing the K141C terminal Yu, BS California, mutations found in hereditary motor and sensory neuropathies in Irvine, CA fi cultured cells had signi cantly reduced chaperone activity and Sha Tang, Ambry Author Acquisition and interpretation 12 increased aggregation compared with wild-type protein. PhD Genetics, of data and revised the Mission Viejo, manuscript for intellectual CA content. In stress conditions, like heat shock, small heat shock proteins such as HSPB8 increase in concentration to perform their Mario University of Author Acquisition and interpretation Saporta, Miami, Miami, of data and revised the chaperone activity by binding to misfolded protein and MD, PhD FL manuscript for intellectual blocking the formation of aggregates. During recovery from content. heat shock under normal conditions, the transcription factor Anna University of Author Acquisition and interpretation NF-κB (Nuclear factor kappa-light-chain-enhancer of acti- Vihola, Helsinki, of data and revised the PhD Helsinki, manuscript for intellectual vated B cells) activates the removal of misfolded or aggregated Finland content. proteins by governing the expression of HSPB8, thereby in- 26 Tahseen University of Author Acquisition and interpretation creasing cell survival. Our studies indicate that patient cells Mozaffar, California, of data and revised the could not recover after the heat shock stress and displayed an MD Irvine, CA manuscript for intellectual increased HSPB8 expression, also noted by Irobi et al.27 content. Bjarne University of Author Acquisition and interpretation HSPB8 Udd, MD, Helsinki, of data and revised the We report a new family with -associated autosomal PhD Helsinki, manuscript for intellectual dominant RVM, caused by a c.515dupC frameshift mutation Finland content. resulting in an elongated protein. The clinical features in these Virginia University of Author Designed and conceptualized individuals bear similarity with recently reported cases. We Kimonis, California, the study; major role in the note reduced expression of HSPB8 in patient fibroblasts MD Irvine, CA acquisition and analysis of data; and drafted the compared with age- and sex-matched control fibroblasts manuscript for intellectual supporting our proposal of haploinsufficiency as the mecha- content nism of disease in this newly described disorder. References Acknowledgment 1. Arndt V, Dick N, Tawo R, et al. Chaperone-assisted selective autophagy is essential for The authors thank the family for their participation in the muscle maintenance. Curr Biol 2010;20:143–148. 2. Fujita M, Mitsuhashi H, Isogai S, et al. Filamin C plays an essential role in the clinical studies and Variantyx Inc. (Framingham, MA) for maintenance of the structural integrity of cardiac and skeletal muscles, revealed by the their help with the genome analysis. medaka mutant zacro. Dev Biol 2012;361:79–89.

Neurology.org/NG Neurology: Genetics | Volume 5, Number 4 | August 2019 9 3. Ehrlicher AJ, Nakamura F, Hartwig JH, Weitz DA, Stossel TP. Mechanical strain in 16. Farwell Hagman KD, Shinde DN, Mroske C, et al. Candidate-gene criteria for clinical actin networks regulates FilGAP and integrin binding to filamin A. Nature 2011;478: reporting: diagnostic exome sequencing identifies altered candidate genes among 8% 260–263. of patients with undiagnosed diseases. Genet Med 2017;19:224–235. 4. Mitra A, Menezes ME, Pannell LK, et al. DNAJB6 chaperones PP2A mediated de- 17. Llewellyn KJ, Nalbandian A, Weiss LN, et al. Myogenic differentiation of VCP disease- phosphorylation of GSK3beta to downregulate beta-catenin transcription target, induced pluripotent stem cells: a novel platform for drug discovery. PLoS One 2017; osteopontin. Oncogene 2012;31:4472–4483. 12:e0176919. 5. Sarparanta J, Jonson PH, Golzio C, et al. Mutations affecting the cytoplasmic func- 18. Rusmini P, Cristofani R, Galbiati M, et al. The role of the heat shock protein B8 tions of the co-chaperone DNAJB6 cause limb-girdle muscular dystrophy. Nat Genet (HSPB8) in motoneuron diseases. Front Mol Neurosci 2017;10:176. 2012;44:450–455. 19. Crippa V, Cicardi ME, Ramesh N, et al. The chaperone HSPB8 reduces the accu- 6. Kostera-Pruszczyk A, Suszek M, Ploski R, et al. BAG3-related myopathy, poly- mulation of truncated TDP-43 species in cells and protects against TDP-43-mediated neuropathy and cardiomyopathy with long QT syndrome. J Muscle Res Cell Motil toxicity. Hum Mol Genet 2016;25:3908–3924. 2015;36:423–432. 20. Crippa V, Sau D, Rusmini P, et al. The small heat shock protein B8 (HspB8) promotes 7. Selcen D, Muntoni F, Burton BK, et al. Mutation in BAG3 causes severe dominant autophagic removal of misfolded proteins involved in amyotrophic lateral sclerosis childhood muscular dystrophy. Ann Neurol 2009;65:83–89. (ALS). Hum Mol Genet 2010;19:3440–3456. 8. Kley RA, Serdaroglu-Oflazer P, Leber Y, et al. Pathophysiology of protein aggregation 21. Kwok AS, Phadwal K, Turner BJ, et al. HspB8 mutation causing hereditary distal and extended phenotyping in filaminopathy. Brain 2012;135:2642–2660. motor neuropathy impairs lysosomal delivery of autophagosomes. J Neurochem 9. Ferrer I, Olive M. Molecular pathology of myofibrillar myopathies. Expert Rev Mol 2011;119:1155–1161. Med 2008;10:e25. 22. Irobi J, Almeida-Souza L, Asselbergh B, et al. Mutant HSPB8 causes motor neuron- 10. Tang BS, Zhao GH, Luo W, et al. Small heat-shock protein 22 mutated in auto- specific neurite degeneration. Hum Mol Genet 2010;19:3254–3265. somal dominant Charcot-Marie-Tooth disease type 2L. Hum Genet 2005;116: 23. Bouhy D, Juneja M, Katona I, et al. A knock-in/knock-out mouse model of HSPB8- 222–224. associated distal hereditary motor neuropathy and myopathy reveals toxic gain-of- 11. Nakhro K, Park JM, Kim YJ, et al. A novel Lys141Thr mutation in small heat shock function of mutant Hspb8. Acta Neuropathol 2017;135:131–148. protein 22 (HSPB8) gene in Charcot-Marie-Tooth disease type 2L. Neuromuscul 24. Echaniz-Laguna A, Geuens T, Petiot P, et al. Axonal neuropathies due to mutations in Disord 2013;23:656–663. small heat shock proteins: clinical, genetic, and functional insights into novel muta- 12. Irobi J, Van Impe K, Seeman P, et al. Hot-spot residue in small heat-shock protein 22 tions. Hum Mutat 2017;38:556–568. causes distal motor neuropathy. Nat Genet 2004;36:597–601. 25. Carra S, Sivilotti M, Chavez Zobel AT, Lambert H, Landry J. HspB8, a small heat 13. Ghaoui R, Palmio J, Brewer J, et al. Mutations in HSPB8 causing a new phenotype of shock protein mutated in human neuromuscular disorders, has in vivo chaperone distal myopathy and motor neuropathy. Neurology 2016;86:391–398. activity in cultured cells. Hum Mol Genet 2005;14:1659–1669. 14. Echaniz-Laguna A, Lornage X, Lannes B, et al. HSPB8 haploinsufficiency causes 26. Nivon M, Fort L, Muller P, et al. NFkappaB is a central regulator of protein quality dominant adult-onset axial and distal myopathy. Acta Neuropathol 2017;134: control in response to protein aggregation stresses via autophagy modulation. Mol 163–165. Biol Cell 2016;27:1712–1727. 15. Cortese A, Laura M, Casali C, et al. Altered TDP-43-dependent splicing in HSPB8- 27. Irobi J, Holmgren A, De Winter V, et al. Mutant HSPB8 causes protein aggregates and related distal hereditary motor neuropathy and myofibrillar myopathy. Eur J Neurol a reduced mitochondrial membrane potential in dermal fibroblasts from distal he- 2018;25:154–163. reditary motor neuropathy patients. Neuromuscul Disord 2012;22:699–711.

10 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG CLINICAL/SCIENTIFIC NOTES OPEN ACCESS Missense mutations in DYT-TOR1A dystonia

Zafar Iqbal, PhD, Jeanette Koht, MD, PhD,* Lasse Pihlstrøm, MD, PhD, Sandra P. Henriksen, MSc, Correspondence Chiara Cappelletti, MSc, Michael Bjørn Russel, MD, PhD, Osmar Norberto de Souza, PhD, Dr. Iqbal [email protected] Inger Marie Skogseid, MD, PhD,* and Mathias Toft, MD, PhD*

Neurol Genet 2019;5:e343. doi:10.1212/NXG.0000000000000343

DYT-TOR1A dystonia is caused by dominant mutations in the TOR1A gene, most frequently a heterozygous in-frame deletion in exon 5 (c.904_906delGAG; p.302/303delE).1 The most frequent phenotype has childhood onset in a limb, spreading to generalized dystonia within a few years. However, also mild focal forms and onset in the cervical and even cranial region have been described. Age at onset varies from 3 to 64 years,2 and penetrance is only 30%. Other in-frame deletions and point mutations in TOR1A have been associated with dystonia in a limited number of patients.3,4 Here, we report 2 new patients with missense mutations in TOR1A.

Patient 1 Patient 1 (IV-1; figure, A) was born from nonconsanguineous Caucasian parents. Focal dystonia started at age 40 years with painful dystonic writer’s cramp affecting the right wrist and finger flexors. She was treated with botulinum toxin injections, but after 10 years, she dis- continued treatment because of very mild symptoms. At age 60 years, only increased blink rate was noted. Two of her children were identified with hyperkinetic movements from adolescence (figure, A). A daughter (V-2; figure, A) had an increased blink rate and a mild head tremor. Her son (V-3; figure, A) was treated with botulinum toxin for a dystonic head tremor and mild torticollis. Gene panel sequencing was performed as described previously,5 revealing a likely pathogenic mutation, c.934A>G; p.R312G in TOR1A (figure, B), which is conserved down to zebrafish (figure, C), has a Combined Annotation Dependent Depletion (cadd.gs.washington. edu) score of 19.5, and is found only once in 246266 alleles (0.0004061%) in the gnomAD database (gnomad.broadinstitute.org). Sanger sequencing revealed segregation of the mutation with the phenotype (figure, A). Using homology modeling, a possible deleterious effect of p.R312G was assessed.6 The basic torsinA structure should be unaffected by the variant. In the wild type, however, the highly flexible arginine allows R312 to come as close as 2.7 Å to one of the adenosine triphosphate ribose hydroxyls and make hydrogen bonds. In the R312G mutant, this interaction is lost, thus possibly causing protein malfunctioning (figure, D).

Patient 2 Patient 2 (II-1; ﬁgure, E) was born from nonconsanguineous Caucasian, healthy parents. Pre- and postnatal periods and psychomotor development were normal. Dystonia started cervically at age 15 years, with phasic torti- and retrocollis, which within 2 years generalized to involve the trunk and upper extremities. Cervical botulinum toxin injections provided some relief until age 23 years. Oral trihexyphenidyl, and levodopa and clonazepam added 2 years later, improved his severe axial dystonia only slightly. At age 26 years, brain and spine MRI and neuropsychological testing were normal. Testing for the classic 3-basepair deletion in TOR1A was negative. He had no impairment of voice, speech, or swallowing, but pathologic face grimacing and moderate

*These authors contributed equally to this study as senior author.

From the Department of Neurology (Z.I., L.P., S.P.H., C.C,. I.M.S,. M.T), Oslo University Hospital; Institute of Clinical Medicine (J.K., M.T), University of Oslo; Department of Neurology (J.K), Drammen Hospital, Vestre Viken Hospital Trust; Head and Neck Research Group, Research Center (M.B.R), Akershus University Hospital; Campus Akershus University Hospital (M.B.R), University of Oslo, Norway; and Faculty of Informatics, Laboratory for Bioinformatics, Modelling & Simulation of Biosystems (O.N.D.S), Pontifical Catholic University of Rio Grande do Sul (PUCRS), Porto Alegre, RS, Brazil.

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

Copyright © 2019 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 Figure Pedigrees of patient 1 and 2, partial Sanger chromatograms, amino acid conservation, and homology modeling

(A) Pedigree structure of the family of patient 1 (IV-1) with the c.934A>G (M) mutation in TOR1A, demonstrating segregation of the mutation. The proband is indicated by the sign of arrow. (B) Partial chromatogram indicating the c.934A>G mutation in TOR1A in patient 1. (C) Evolutionary conservation of amino acid at position 312, revealing high conservation down to zebrafish. (D) Homology model of the human torsinA p.R312G mutant. The α-helices and β-sheet of the AAA+ domain are colored in blue and purple, respectively. The C-terminal domain is colored in light blue. The ATP molecule and G312 are represented as sticks and colored by CPK. The mutation region is highlighted at the top right of this figure. Image generated with PyMOL (pymol.org/ 2/). (E) Pedigree structure of the family of patient 2 (II-1) with the c.863G>A (M) mutation in TOR1A. The proband is indicated by the sign of arrow. Three siblings are indicated by diamond (un- disclosed sex) symbols. The siblings have not been tested for the mutation. (F) Partial chromatogram indicating the c.863G>A mutation in TOR1A in patient 2. ATP = adenosine triphosphate.

distal extremity dystonia. Since age 27 years, his generalized T321M. We report a patient carrying p.R288Q, with dystonia has been treated successfully with deep brain stim- adolescent-onset, isolated generalized dystonia and marked ulation (DBS) of the internal globus pallidus bilaterally, with axial and little cranial involvement. This mutation was pre- a marked and sustained effect, allowing medication to be ta- viously reported in a patient with very early childhood-onset pered off. At age 42 years, he has residual mild retrocollis and lower limb dystonia and severe generalization, who at age 18 thoracic kyphoscoliotic posture. Gene panel sequencing years had dysphagia, dysarthria, joint contractures, pyramidal revealed a previously reported missense mutation, c.863G>A; signs, and cerebellar atrophy.7 Treatment with DBS was not p.R288Q in TOR1A,7 with paternal inheritance and reduced mentioned. Additional experiments strongly indicated path- penetrance, the latter also witnessed by Zirn et al.7 (figure, E ogenicity of the p.R288Q variant,7 which is further confirmed and F). by our present finding.

Herein, we report a novel TOR1A missense mutation, Most reported patients with TOR1A missense mutations p.R312G, which segregated with mild isolated segmental presented with adult onset, including the p.R312G index dystonia in a small family. Multiple lines of bioinformatic patient in our study. This may indicate that missense muta- predictions indicate possible deleterious effects on protein tions have a less profound effect on torsinA function than the function. However, functional analyses and identification of common deletion. However, the phenotypic spectrum of genetic recurrence are warranted to confirm its pathogenicity. TOR1A mutations is very broad, ranging from nonpenetrance to isolated focal, segmental, or generalized dystonia in carriers Of the published missense mutations in TOR1A (mdsgene. of different types of mutations, which is highlighted in our and org), genetic recurrence has so far only been reported for p. previous reports.3,4,7 The causes of this large phenotypic

2 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG variation in TOR1A mutation carriers still largely remain grant from The Norwegian Research Council (grant number elusive. 229654).

Author contributions Disclosure Z. Iqbal: bioinformatic analysis and interpretation of data, wet Disclosures available: Neurology.org/NG. laboratory work, and drafting and revision of the manuscript. J. Koht: ascertaining the patients and clinical data, study Publication history Neurology: Genetics ﬁ concept and design, and drafting and revision of the manu- Received by November 28, 2018. Accepted in nal script. S.P. Henriksen: preparation of the samples for se- form May 9, 2019. quencing and arrangement of the samples. C. Cappelletti: wet References laboratory work. M.B. Russell: ascertaining the patients. O. 1. Ozelius LJ, Hewett JW, Page CE, et al. The early-onset torsion dystonia gene (DYT1) Norberto de Souza: homology modeling and drafting the encodes an ATP-binding protein. Nat Genet 1997;17:40–48. 2. Ozelius LJ, Bressman SB. Genetic and clinical features of primary torsion dystonia. manuscript. L. Pihlstrøm: bioinformatic analysis, in- Neurobiol Dis 2011;42:127–135. terpretation of data, and revision of the manuscript. I.M. 3. Calakos N, Patel VD, Gottron M, et al. Functional evidence implicating a novel TOR1A mutation in idiopathic, late-onset focal dystonia. J Med Genet 2010;47: Skogseid: ascertaining the patients and clinical data, study 646–650. concept and design, and drafting and revision of the manu- 4. Leung JC, Klein C, Friedman J, et al. Novel mutation in the TOR1A (DYT1) gene in atypical early onset dystonia and polymorphisms in dystonia and early onset par- script. M. Toft: study concept and design, obtained funding, kinsonism. Neurogenetics 2001;3:133–143. study supervision, and revision of the manuscript. 5. Iqbal Z, Rydning SL, Wedding IM, et al. Targeted high throughput sequencing in hereditary ataxia and spastic paraplegia. PLoS One 2017;12:e0174667. 6. Hettich J, Ryan SD, de Souza ON, et al. Biochemical and cellular analysis of human Study funding variants of the DYT1 dystonia protein, TorsinA/TOR1A. Hum Mutat 2014;35: 1101–1113. The study was funded by an internal grant from Oslo Uni- 7. Zirn B, Grundmann K, Huppke P, et al. Novel TOR1A mutation p.Arg288Gln in versity Hospital. Dr. Skogseid was supported by a research early-onset dystonia (DYT1). J Neurol Neurosurg Psychiatry 2008;79:1327–1330.

Neurology.org/NG Neurology: Genetics | Volume 5, Number 4 | August 2019 3 CLINICAL/SCIENTIFIC NOTES OPEN ACCESS Adult-onset variant ataxia-telangiectasia diagnosed by exome and cDNA sequencing

Martin Krenn, MD, Ivan Milenkovic, MD, PhD, Gertrud Eckstein, PhD, Fritz Zimprich, MD, PhD, Correspondence Thomas Meitinger, MD, Thomas Foki, MD, and Matias Wagner, MD Dr. Wagner [email protected] Neurol Genet 2019;5:e346. doi:10.1212/NXG.0000000000000346

ATM Ataxia-telangiectasia (A-T) is an autosomal recessive disorder caused by mutations in , MORE ONLINE encoding a serine-threonine protein kinase that is crucially involved in DNA repair mechanisms. Clinical features include cerebellar degeneration, telangiectasia, immunodeficiency, and Video an increased risk of malignancies.1 The classic form of A-T is characterized by infantile, rapidly progressing neurodegeneration and can be differentiated from variant A-T with a comparably milder disease course.2,3 However, only a tiny fraction of patients first present with symptoms in adulthood.4 The broad phenotypic spectrum of A-T now becomes gradually disentangled owing to the increased availability of comprehensive genetic testing.5

Here, we point out possible diagnostic pitfalls with an example of an adult-onset A-T, in which exome sequencing (ES) complemented by transcriptome analysis, complementary de- oxyribonucleotide acid (cDNA) sequencing, and family genotyping eventually led to the definite genetic diagnosis A-T.

Clinical and genetic findings Our female index patient (currently aged 45 years) first developed gait instability at age 34 years. Furthermore, dizziness and affective lability were reported. At this time, a neurologic examination was unremarkable and did not reveal any signs of cerebellar dysfunction.

Symptoms were slowly progressive over the following 10 years, and she first presented at our department at age 43 years. Clinical examination revealed bilateral ataxia with wide-based gait, downbeat nystagmus, mild dysarthria, and brisk reflexes. Brain MRI showed moderate to severe bilateral cerebellar atrophy (figure, A). Electrooculography confirmed a downbeat nystagmus, saccadic pursuit, and slowed and hypometric saccades with absent vestibulo-ocular reflex suppression indicating cerebellar dysfunction (pathologic eye movements are shown in video 1). Dizziness was most pronounced during head-turning movements.

Given a positive family history with the patient’s elder sister even being more severely affected with ataxia (starting at age 28 years), a genetic etiology was suspected. Subsequently performed ES revealed 1 pathogenic nonsense variant (c.6205C>T, p.Gln2069*) and 1 intronic variant (c.1235+3A>G) of uncertain significance (VUS) in ATM as classified by the American College of Medical Genetics and Genomics criteria.6 The nonsense variant was absent from Genome Aggregation Database and our in-house database comprising more than 16,500 exome data sets, and the intronic variant, which is located in proximity to exon 9, was present twice in each database.

From the Department of Neurology (M.K., I.M., F.Z.), Medical University of Vienna, Austria; Institute of Human Genetics (M.K., T.M., M.W.), Technical University Munich; Institute of Human Genetics (G.E., T.M., M.W.), Helmholtz Zentrum Munchen,¨ Neuherberg, Germany; Department of Neurology (T.F.), Karl Landsteiner University of Health Sciences, Tulln, Austria; and Institute of Neurogenomics (M.W.), Helmholtz Zentrum Munchen,¨ Neuherberg, Germany.

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 German Research Foundation (DFG) and the Technical University of Munich (TUM) in the framework of the Open Access Publishing Program. 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 © 2019 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 Figure Brain MRT and molecular characterization of the splice-site variant c.1235+3A>G

(A) Brain MRI. Sagittal T2-weighted MRI of the reported index patient demonstrating moderate to severe cerebellar atrophy (white arrow). (B) Gel electrophoresis of PCR-amplified ATM cDNA showing 2 distinct splice variants as a result of the mutation c.1235+3A>G. (C) Subsequent Sanger sequencing of dissected bands further characterized the misspliced transcript with skipping of exon 9 (band II). (D) Photograph of the index patient’s right cheek showing telangiectasia.

To further delineate the pathogenicity of the intronic variant, a rare phenomenon.4 It has been suggested that mainly the we sought to investigate a potential splicing defect using mutation type predicts the clinical course with missense var- polyA-enriched RNA sequencing from whole blood using iants leading to a milder phenotype but an increased risk of PAXgene Blood RNA tubes (Qiagen, Hilden, Germany).7 cancer.5 With an mRNA length of 13,147 nucleotides and due to the location of exon 9 toward the 59 end of the ATM transcript, The patient reported here is affected by an adult-onset mild RNA sequencing coverage was insufficient to evaluate form of variant A-T despite carrying biallelic truncating variants. a splicing effect. Subsequently, we specifically targeted the 59 At the time of the last follow-up visit, she was still ambulatory and end with a primer binding to exon 25 of ATM (39-GAATG- there was no evidence for malignant disease by age 45 years. GATTAGAACCTCACC-59) for reverse transcription. With primers binding to exons 8 and 10 (sequences available upon Such atypically mild clinical expressions may be misleading, request), we could demonstrate skipping of exon 9 leading to and it is not far to seek that single gene testing or narrowly a frameshift and the premature termination of protein trans- targeted gene panels might significantly prolong the di- lation (p.Val356Alafs*17; figure, C and D). agnostic odyssey. Nonetheless, our report highlights that even comprehensive genomic approaches may not be sufficient to The 2 variants segregated with the disease phenotype within establish a molecular diagnosis. In our case, ES paved the way the family. Once a genetic diagnosis of A-T had been estab- for subsequent rephenotyping (confirming telangiectasia) lished, subtle but evident telangiectasia of the conjunctiva, and a more detailed molecular workup (RNA and cDNA face (figure, B), and chest was clinically confirmed by a der- sequencing) eventually corroborating disease causation. matologist. Blood levels of alpha-fetoprotein were within the normal range. There has been no evidence for malignancy or We conclude that A-T should be taken into consideration as immunodeficiency. Currently, the described index patient and differential diagnosis of adult-onset ataxia, even in the absence her affected sister are still ambulatory despite gait ataxia with of apparent systemic features. The here encountered di- a marked tendency to fall. agnostic pitfalls support the necessity of a stepwise clinical and molecular reconsideration of genetic testing. Discussion Study funding Variant A-T represents a subgroup of A-T characterized by No targeted funding reported. a milder disease course compared with the classic form. De- spite slower disease progression, the vast majority of patients Disclosure first manifest in early childhood, and adult onset remains Disclosures available: Neurology.org/NG.

2 Neurology: Genetics | Volume 5, Number 4 | August 2019 Neurology.org/NG Publication history Received by Neurology: Genetics March 14, 2019. Accepted in ﬁnal form Appendix (continued)

May 23, 2019. Name Location Role Contributions

Thomas Karl Landsteiner Author Major role in data Appendix Authors Foki, MD University of acquisition and Health Sciences, clinical Name Location Role Contributions Tulln, Austria management and critical revision of Martin Department of Author Genetic data the manuscript Krenn, MD Neurology, analysis and Medical drafting of the Matias Institute of Corresponding RNA and cDNA University of manuscript Wagner, MD Human author sequencing and Vienna, Austria Genetics, proposed and Technical supervised the Ivan Department of Author Acquisition of University manuscript Milenkovic, Neurology, clinical data Munich, MD, PhD Medical Germany University of Vienna, Austria

Gertrud Institute of Author RNA and cDNA Eckstein, Human sequencing References PhD Genetics, 1. Rothblum-Oviatt C, Wright J, Lefton-Greif MA, McGrath-Morrow SA, Craw- Helmholtz ford TO, Lederman HM. Ataxia telangiectasia: a review. Orphanet J Rare Dis Center Munich, 2016;11:159. Germany 2. Verhagen MMM, Abdo WF, Willemsen MAAP, et al. Clinical spectrum of ataxia- telangiectasia in adulthood. Neurology 2009;73:430–437. Fritz Department of Author Acquisition of 3. Verhagen MMM, Last JI, Hogervorst FBL, et al. Presence of ATM protein and Zimprich, Neurology, clinical data and residual kinase activity correlates with the phenotype in ataxia-telangiectasia: a geno- MD, PhD Medical critical revision of type-phenotype study. Hum Mutat 2012;33:561–571. University of the manuscript 4. Worth PF, Srinivasan V, Smith A, et al. Very mild presentation in adult with Vienna, Austria classical cellular phenotype of ataxia telangiectasia. Mov Disord 2013;28: 524–528. Thomas Institute of Author Supervised 5. Schon K, van Os NJH, Oscroft N, et al. Genotype, extrapyramidal features, and Meitinger, Human genetic data severity of variant ataxia-telangiectasia. Ann Neurol 2019;85:170–180. MD Genetics, analysis 6. Richards S, Aziz N, Bale S, et al. Standards and guidelines for the interpretation of Technical sequence variants: a joint consensus recommendation of the American College of University Medical Genetics and Genomics and the Association for Molecular Pathology. Genet Munich, Med 2015;17:405–423. Germany 7. Kremer LS, Bader DM, Mertes C, et al. Genetic diagnosis of Mendelian disorders via RNA sequencing. Nat Commun 2017;8:15824.

Neurology.org/NG Neurology: Genetics | Volume 5, Number 4 | August 2019 3 CORRECTION Missense mutations in DYT-TOR1A dystonia Neurol Genet 2019;5:e350. doi:10.1212/NXG.0000000000000350

In the article “Missense mutations in DYT-TOR1A dystonia” by Iqbal et al.,1 first published online June 6, 2019, the genotype under patient IV-1 in panel A of the figure should have read “M/-.” Additionally, the third label in the second row of panel A in the same figure should read “II-3.” The editorial office regrets the errors.

Figure

Reference 1. Iqbal Z, Koht J, Pihlstrøm L, et al. Missense mutations in DYT-TOR1A dystonia. Neurol Genet 2019;5:e343.

Copyright © 2019 American Academy of Neurology 1 CORRECTION Genetic risk of Parkinson disease and progression: An analysis of 13 longitudinal cohorts Neurol Genet 2019;5:e354. doi:10.1212/NXG.0000000000000354

In the article “Genetic risk of Parkinson disease and progression: An analysis of 13 longitudinal cohorts" by Iwaki et al.,1 ﬁrst published online July 9, 2019, in the abstract’s results, the phrase should be “T allele of rs114128760.” The authors regret the error.

Reference 1. Iwaki H, Blauwendraat C, Leonard HL, et al. Genetic risk of Parkinson disease and progression: An analysis of 13 longitudinal cohorts. Neurol Genet 2019;5:e348.

Copyright © 2019 American Academy of Neurology 1 CORRECTION Genome-wide Brain DNA methylation analysis suggests epigenetic reprogramming in Parkinson disease Neurol Genet 2019;5:e355. doi:10.1212/NXG.0000000000000355

In the article “Genome-wide Brain DNA methylation analysis suggests epigenetic reprogramming in Parkinson disease" by Young et al.,1 ﬁrst published online June 24, 2019, Dr. Sathesh K. Sivasankaran should have been listed as co-ﬁrst author. The authors regret the error.

Reference 1. Young JI, Sivasankaran SK, Wang L, et al. Genome-wide brain DNA methylation analysis suggests epigenetic reprogramming in Parkinson disease. Neurol Genet 2019;5:e342.