An Official Journal of the American Academy of Neurology Neurology.org/ng • Online ISSN: 2376-7839 Volume 3, Number 6, December 2017 Genetics

Germline and Copy number variation CDKL5 variants: improving somatic in analysis increases the our understanding of a rare STXBP1 with diverse diagnostic yield in muscle neurological disorder neurodevelopmental diseases phenotypes Table of Contents Neurology.org/ng Online ISSN: 2376-7839 Volume 3, Number 6, December 2017

ARTICLES e207 NaV channel variants in patients with painful and e197 c.207C.G in sepiapterin reductase nonpainful peripheral neuropathy causes autosomal dominant dopa-responsive S. Wadhawan, S. Pant, R. Golhar, S. Kirov, dystonia J. Thompson, L. Jacobsen, I. Qureshi, S. Ajroud-Driss, A.S. Shalash, T.W. Rösler, S.H. Müller, M. Salama, R. Freeman, D.M. Simpson, A.G. Smith, A. Hoke, and G. Deuschl, U. Müller, T. Opladen, B.-S. Petersen, L.J. Bristow, On behalf of the Foundation for A. Franke, F. Hopfner, G. Kuhlenbäumer, and Peripheral Neuropathy and the Peripheral Neuropathy G.U. Höglinger Research Registry (PNRR) Consortium

e204 Copy number variation analysis increases the e198 Characterization of a KCNB1 variant associated with autism, intellectual disability, and diagnostic yield in muscle diseases J.D. Calhoun, C.G. Vanoye, F. Kok, A.L. George, Jr., S. Välipakka, M. Savarese, M. Johari, L. Sagath, and J.A. Kearney M. Arumilli, K. Kiiski, A. Sáenz, A.L. De Munain, A.-M. Cobo, K. Pelin, B. Udd, and P. Hackman e206 Clinical features and outcome of 6 new e199 Germline and somatic mutations in STXBP1 with patients carrying de novo KCNB1 diverse neurodevelopmental phenotypes mutations M. Uddin, M. Woodbury-Smith, A. Chan, L. Brunga, C. Marini, M. Romoli, E. Parrini, C. Costa, D. Mei, F. Mari, S. Lamoureux, G. Pellecchia, R.K.C. Yuen, M. Faheem, L. Parmeggiani, E. Procopio, T. Metitieri, E. Cellini, D.J. Stavropoulos, J. Drake, C.D. Hahn, C. Hawkins, S. Virdò, D. De Vita, M. Gentile, P. Prontera, A. Shlien, C.R. Marshall, L.A. Turner, B. Minassian, P. Calabresi, and R. Guerrini S.W. Scherer, and C. Boelman

e200 CDKL5 variants: Improving our understanding of e208 Clinical heterogeneity and phenotype/genotype a rare neurologic disorder findings in 5 families with GYG1 deficiency R.D. Hector, V.M. Kalscheuer, F. Hennig, R. Ben Yaou, A. Hubert, I. Nelson, J.R. Dahlqvist, H. Leonard, J. Downs, A. Clarke, T.A. Benke, D. Gaist, N. Streichenberger, M. Beuvin, M. Krahn, J.Armstrong,M.Pineda,M.E.S.Bailey,and P. Petiot, F. Parisot, F. Michel, E. Malfatti, N. Romero, S.R. Cobb R.Y. Carlier, B. Eymard, P. Labrune, M. Duno, T. Krag, M. Cerino, M. Bartoli, G. Bonne, J. Vissing, P. Laforet, and F.M. Petit e203 Factors influencing the age at onset in familial frontotemporal lobar dementia: Important weight of genetics CLINICAL/SCIENTIFIC NOTES M. Barbier, A. Camuzat, M. Houot, F. Clot, e205 Novel NDUFS4 gene mutation in an atypical P.Caroppo,C.Fournier,D.Rinaldi,F.Pasquier, late-onset mitochondrial form of multifocal D. Hannequin, J. Pariente, K. Larcher, The French dystonia Clinical and Genetic Research Network on C. Bris, T. Rouaud, V. Desquiret-Dumas, N. Gueguen, FTD/FTD-ALS, Predict-PGRN & PrevDemAls Study D. Goudenege, M. Barth, D. Bonneau, Groups, A. Brice, E. Génin, A. Sabbagh, and P. Amati-Bonneau, G. Lenaers, P. Reynier, I. Le Ber A.-S. Lebre, and V. Procaccio Table of Contents continued e201 Novel GRN mutation presenting as an aphasic e202 Opening one’s eyes to mosaicism in progressive dementia and evolving into corticobasal syndrome external ophthalmoplegia H. Botha, N.C.A. Finch, R.H. Gavrilova, M.M. Machulda, E.W. Sommerville, R.L. Jones, S.A. Hardy, J.A. Fields, V.J. Lowe, R.C. Petersen, C.R. Jack, Jr., E.L. Blakely, A. Pyle, A.M. Schaefer, C.M. Dheel, D.J. Gearhart, D.S. Knopman, P.F. Chinnery, D.M. Turnbull, G.S. Gorman, and R. Rademakers, and B.F. Boeve R.W. Taylor

Podcast Video LOE classification LOE recommendation

tinyurl.com/NeurologyNG twitter.com/GreenJournal

Cover image: Location of the KCNB1-I199F variant mapped onto the crystal structure of a KV2.1/KV1.2 chimera (PDB 29R9). Stylized by Kaitlyn Aman Ramm, Neurology Editorial Assistant. See “Characterization of a KCNB1 variant associated with autism, intellectual disability and epilepsy.” Neurology.org/ng Online ISSN: 2376-7839 Volume 3, Number 6, December 2017

fi VISION: Neurology® Genetics will be the premier peer-reviewed Executive Of ce, American Academy of Neurology fi Catherine M. Rydell, CAE, Executive Director/CEO journal in the eld of neurogenetics. 201 Chicago Ave Minneapolis, MN 55415 MISSION: Neurology: Genetics will provide neurologists with Tel: 612-928-6100 outstanding original contributions that elucidate the role of genetic and epigenetic variations in diseases and biological traits Editorial Office of the central and peripheral nervous systems. Patricia K. Baskin, MS, Executive Editor Kathleen M. Pieper, Senior Managing Editor, Neurology Editor Lee Ann Kleffman, Managing Editor, Neurology: Genetics Neurology: Genetics Sharon L. Quimby, Managing Editor, Neurology® Clinical Practice Stefan M. Pulst, MD, Dr med, FAAN Morgan S. Sorenson, Managing Editor, Neurology® Neuroimmunology & fl Professor and Chair, Department of Neurology Neuroin ammation University of Utah Cynthia S. Abair, MA, Senior Graphics Editor Salt Lake City, UT Andrea R. Rahkola, Production Editor, Neurology [email protected] Robert J. Witherow, Senior Editorial Associate Specialties: Genetics, movement disorders Karen Skaja, Senior Editorial Associate Kaitlyn Aman Ramm, Editorial Assistant Editor-in-Chief Kristen Swendsrud, Editorial Assistant Neurology® Andrea Willgohs, Editorial Assistant Robert A. Gross, MD, PhD, FAAN Professor of Neurology and of Pharmacology and Physiology Strong Epilepsy Center Publisher University of Rochester Medical Center Wolters Kluwer Rochester, NY Baltimore, MD [email protected]

Editorial Inquiries Publishing Staff Tel: 612-928-6400 Kim Jansen, Executive Publisher Toll-free: 800-957-3182 (US) Alexandra Lazerow, Production Team Leader, AAN Journals Fax: 612-454-2748 Megen Miller, Production Editor [email protected] Steve Rose, Editorial Assistant Stacy Drossner, Production Associate Academy Officers Ralph L. Sacco, MD, MS, FAAN, President James C. Stevens, MD, FAAN, President Elect Ann H. Tilton, MD, FAAN, Vice President Carlayne E. Jackson, MD, FAAN, Secretary Janis M. Miyasaki, MD, MEd, FRCPC, FAAN, Treasurer Terrence L. Cascino, MD, FAAN, Past President Neurology.org/ng Online ISSN: 2376-7839 Volume 3, Number 6, December 2017

Editor Biostatistics Neurology® Genetics Richard J. Kryscio, PhD Stefan M. Pulst, MD, Dr med, FAAN University of Kentucky Professor and Chair, Department of Neurology Lexington, KY University of Utah Christopher A. Beck, PhD Salt Lake City, UT University of Rochester [email protected] Rochester, NY Specialties: Genetics, movement disorders Sue Leurgans, PhD Editor-in-Chief Rush University Medical Center Neurology® Chicago, IL Robert A. Gross, MD, PhD, FAAN Level of Evidence Evaluations Professor of Neurology and of Pharmacology and Physiology Gary S. Gronseth, MD, FAAN Strong Epilepsy Center University of Kansas University of Rochester Medical Center Kansas City, KS Rochester, NY [email protected] Podcasts Andrew M. Southerland, MD, MSc Deputy Editors University of Virginia Neurology Charlottesville, VA Bradford B. Worrall, MD, MSc, FAAN Ted M. Burns, MD, Deputy Podcast Editor Professor of Neurology and Public Health Sciences University of Virginia University of Virginia Charlottesville, VA Charlottesville, VA Ombudsman Neurology: Genetics David S. Knopman, MD, FAAN Nicholas Elwood Johnson, MD Mayo Clinic University of Utah Rochester, MN Salt Lake City, UT Specialties: Muscle diseases, patient-centered outcomes Scientific Integrity Advisor Robert B. Daroff, MD, FAAN Associate Editors Case Western Reserve University Neurology: Genetics Cleveland, OH Alexandra Durr, MD, PhD Editorial Board Hôpital de la Salpêtrière Hilary Coon, PhD, University of Utah, Salt Lake City, UT Paris, France Giovanni Coppola, MD, UCLA, Los Angeles, CA Specialties: Spastic paraplegia, Huntington disease, ataxia Chantal Depondt, MD, PhD, Université Libre de Bruxelles, Brussels, Belgium Massimo Pandolfo, MD, FAAN Brent L. Fogel, MD, PhD, FAAN, UCLA, Los Angeles, CA Hôpital Erasme Anthony J. Griswold, PhD, University of Miami, Miami, FL Université Libre de Bruxelles Orhun H. Kantarci, MD, Mayo Clinic, Rochester, MN Brussels, Belgium Julie R. Korenberg, PhD, MD, University of Utah, Salt Lake City, UT Specialties: Epilepsy, Friedreich ataxia, stem cells Margherita Milone, MD, PhD, Mayo Clinic, Rochester, MN Davide Pareyson, MD, C. Besta Neurological Institute IRCCS Foundation, Raymond P. Roos, MD, FAAN Milan, Italy University of Chicago Medical Center Shoji Tsuji, MD, PhD, University of Tokyo Graduate School of Medicine, Chicago, IL Tokyo, Japan Specialties: Motor disease, prion disease, neurovirology/neuroimmunology Dineke S. Verbeek, PhD, University Medical Center Groningen, Groningen, The Netherlands Jeffery M. Vance, MD, PhD David Viskochil, MD, PhD, University of Utah, Salt Lake City, UT University of Miami Juliane Winkelmann, MD, Technische Universität München, Helmholtz Miami, FL Zentrum Münche, Munich, Germany Specialties: Parkinson, Alzheimer disease, inherited neuropathies Juan I. Young, PhD, University of Miami, Miami, FL Level of Evidence Review Team Melissa J. Armstrong, MD, Gainesville, FL Richard L. Barbano, MD, PhD, FAAN, Rochester, NY Richard M. Dubinsky, MD, MPH, FAAN, Kansas City, KS Jeffrey J. Fletcher, MD, MSc, Ann Arbor, MI Gary M. Franklin, MD, MPH, FAAN, Seattle, WA David S. Gloss II, MD, MPH&TM, Charleston, WV John J. Halperin, MD, FAAN, Summit, NJ Jason Lazarou, MSc, MD, Toronto, Ontario, Canada Steven R. Messé, MD, FAAN, Philadelphia, PA Pushpa Narayanaswami, MBBS, DM, FAAN, Boston, MA Alex Rae-Grant, MD, Cleveland, OH Neurology.org/ng  Online ISSN: 2376-7839 Volume 3, Number 6, December 2017

Publication Information Neurology® is a registered trademark of the American Academy of Neurology (registration valid in the United States). Neurology® Genetics (eISSN 2376-7839) is an open access journal published online for the American Academy of Neurology, 201 Chicago Avenue, Minneapolis, MN 55415, by Wolters Kluwer Health, Inc. at 14700 Citicorp Drive, Bldg. 3, Hagerstown, MD 21742. Business offices are located at Two Commerce Square, 2001 Market Street, Philadelphia, PA 19103. Production offices are located at 351 West Camden Street, Baltimore, MD 21201-2436. © 2017 American Academy of Neurology. Neurology® Genetics is an official journal of the American Academy of Neurology. Journal website: Neurology.org/ng, AAN website: AAN.com Copyright and Permission Information: Please go to the journal website (www.neurology.org/ng) and click the Permissions tab for the relevant article. Alternatively, send an email to [email protected]. General information about permissions can be found here: https://shop.lww.com/journal-permission. Disclaimer: Opinions expressed by the authors and advertisers are not necessarily those of the American Academy of Neurology, its affiliates, or of the Publisher. The American Academy of Neurology, its affiliates, and the 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 or non-use of any of the material contained in this publication. Advertising Sales Representatives: Wolters Kluwer, 333 Seventh Avenue, New York, NY 10001. Contacts: Eileen Henry, tel: 732-778-2261, fax: 973-215-2485, [email protected] and Elizabeth S. Hall, tel: 267-804-8123, [email protected]. In Europe: Avia Potashnik, Wolters Kluwer, tel: 144 207 981 0722; 144 7919 397 933 or e-mail: [email protected]. Careers & Events: Monique McLaughlin, Wolters Kluwer, Two Commerce Square, 2001 Market Street, Philadelphia, PA 19103, tel: 215-521- 8468, fax: 215-521-8801; [email protected]. Reprints: Meredith Edelman, Commercial Reprint Sales, Wolters Kluwer, Two Commerce Square, 2001 Market Street, Philadelphia, PA 19103, tel: 215-356-2721; [email protected]; [email protected]. Special projects: US & Canada: Alan Moore, Wolters Kluwer, Two Commerce Square, 2001 Market Street, Philadelphia, PA 19103, tel: 215-521-8638, [email protected]. International: Andrew Wible, Senior Manager, Rights, Licensing, and Partnerships, Wolters Kluwer; [email protected]. c.207C.G mutation in sepiapterin reductase causes autosomal dominant dopa-responsive dystonia

Ali S. Shalash, MD* ABSTRACT * Thomas W. Rösler, PhD Objective: To elucidate the genetic cause of an Egyptian family with dopa-responsive dystonia * Stefanie H. Müller, MS (DRD), a childhood-onset dystonia, responding therapeutically to levodopa, which is caused by Mohamed Salama, MD mutations in various . Günther Deuschl, MD Methods: Rare variants in all coding exons of GCH1 were excluded by Sanger sequencing. Exome Ulrich Müller, MD sequencing was applied for 1 unaffected and 2 affected family members. To investigate the Thomas Opladen, MD functional consequences of detected genetic variants, urinary sepiapterin concentrations were Britt-Sabina Petersen, determined by high-performance liquid chromatography. PhD Andre Franke, PhD Results: A heterozygous rare nonsynonymous variant in exon 1 of sepiapterin reductase (SPR, . Franziska Hopfner, MD* c.207C G, p.Asp69Glu) was found in all affected family members. Urinary concentrations of Gregor Kuhlenbäumer, sepiapterin were above the standard of normal controls in most SPR mutation carriers, suggest- MD, PhD* ing functional biochemical consequences of the mutation. Variant filtering of all genes involved in Günter U. Höglinger, the tetrahydrobiopterin pathway, required for levodopa synthesis, revealed an additional common MD* variant in dihydrofolate reductase (DHFR, rs70991108). The presence of both variants was significantly stronger associated with the biochemical abnormality and the clinical disease state as opposed to 1 variant only. Correspondence to Conclusions: The rare SPR mutation can cause autosomal dominant DRD with incomplete pene- Prof. Dr. Höglinger: [email protected] trance. The common DHFR variant might have synergistic effects on production of tetrahydro- biopterin and levodopa, thereby increasing penetrance. Neurol Genet 2017;3:e197; doi: 10.1212/ NXG.0000000000000197

GLOSSARY DHFR 5 dihydrofolate reductase; DRD 5 dopa-responsive dystonia; MAF 5 minor allele frequency; NADPH 5 nicotinamide adenine dinucleotide phosphate; SPR 5 sepiapterin reductase.

Dopa-responsive dystonia (DRD) encompasses a group of movement disorders with heteroge- neous clinical and genetic manifestations, presenting with childhood-onset limb dystonia and response to levodopa therapy.1 Additional neurologic presentations may include parkinsonian and pyramidal features.2,3 Variants in the gene of GTP cyclohydrolase I (GCH1), a key of the tetrahydrobiopterin (BH4) pathway, required for the biosynthesis of levodopa (figure, A), are the most frequent cause of autosomal dominant DRD, also known as DYT5a.1,4 Variants of sepiapterin reductase (SPR), encoding another key protein of the BH4 pathway (figure, A), have been identified in autosomal recessive DRD.5 At present, only 1 patient with DRD without family history has been identified with a heterozygous SPR variant.6 Here, we describe the first DRD family carrying a rare SPR variant (c.207C.G, p.Asp69Glu, chr2:73114768_C.G,

*These authors contributed equally to the manuscript. From the Department of Neurology (A.S.S.), Ain Shams University, Cairo, Egypt; German Center for Neurodegenerative Diseases (DZNE) (T.W.R., G.U.H.), Munich, Germany; Department of Neurology (T.W.R., G.U.H.), Technical University of Munich, Germany; Department of Neurology (S.H.M., G.D., F.H., G.K.), University Hospital Schleswig Holstein, Kiel, Germany; Medical Experimental Research Center (MERC) (M.S.), Mansoura University, Egypt; Institute for Human Genetics (U.M.), University of Giessen, Germany; Division of Neuropediatrics and Metabolic Medicine (T.O.), University Children’s Hospital, Heidelberg, Germany; and Institute of Clinical Molecular Biology (B.-S.P., A.F.), Christian-Albrechts-Universität zu Kiel, Germany. Funding information and disclosures are provided at the end of the article. Go to Neurology.org/ng for full disclosure forms. The Article Processing Charge was funded by Klinikum rechts der Isar der TUM. 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.

Neurology.org/ng Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 Figure Sepiapterin reductase and dihydrofolate reductase in the context of dopa-responsive dystonia

(A) Synthesis and regeneration of 5,6,7,8-tetrahydrobiopterin (BH4), an essential of L-DOPA synthesis (scheme adapted from references 3, 11). of relevance are highlighted in red and blue. AR 5 aldose reductase; CR 5 carbonyl reductase; DHPR 5 dihydropteridine reductase; PCD 5 pterin- 4a-carbinolamine reductase; TH 5 tyrosine hydroxylase; 29-Oxo-TP 5 19-hydroxy-29-oxopropyltetrahydrobiopterin; 19Oxo-TP 5 29-hydroxy-19-oxopropyl- tetrahydrobiopterin. (B) Pedigree of the investigated dopa-responsive dystonia family. Sepiapterin reductase (SPR) C/G 5 chr2:73,114,768 C/G in exon 1 of SPR; dihydrofolate reductase (DHFR)-/Ins 5 chr5:79,950,163-/TGGCGCGTCCCGCCCAGGT (19 base-pair insert) intronic, located upstream of exon 5 of DHFR. Green asterisks indicate family members with urine sepiapterin values slightly above the concentration range measured in normal controls. (C) Structure of the NADP- in chain A of SPR obtained from the Research Collaboratory for Structural Bioinformatics (identifier 5 4XWY) in wild-type (left) and mutant states (right). Aspartate on position 69 forms hydrogen bonds (green dashed lines) with cofactor nicotinamide adenine dinucleotide phosphate (NADPH) (colored in black). In the case of a substitution of aspartate with glutamate, a loss of hydrogen bonds is predicted (dashed red line).

undetected in 8,599 alleles annotated in the , Lung, and Blood Institute Exome Sequencing Project, Exome Variant Server) with autosomal domi- version ESP6500siv2), were discarded. Assuming a dominant dis- order, only nonsynonymous variants were retained as candidate nant inheritance and incomplete penetrance. alleles, including missense, nonsense, stop-loss, and splice-site mutations, as well as small insertions and deletions present in METHODS Standard protocol approvals, registrations, the heterozygous state and present in both affected sisters (figure and patient consents. Affected and nonaffected individuals of B, II.2 and II.4). Hypothesis-based variant filtering was applied to a family of Egyptian origin were examined by neurologists special- all genes in the SPR pathway, also considering variants with an ized in movement disorders (A.S.S. and G.U.H.) at the MAF .1.0%. Candidate variants were validated by Sanger Department of Neurology, Ain Shams University, Cairo, Egypt. sequencing and cosegregation with disease in the family (table 1). Ethical approval was obtained at Ain Shams University (RP/42) In silico analyses. The effects of substitutions on pro- and Technical University of Munich (203/15s). All participants tein function were predicted using MutationTaster (mutationtaster. provided written informed consent. Consanguinity was excluded org), PolyPhen-2 (genetics.bwh.harvard.edu/pph2), and CADD by interview and genetic analyses. (cadd.gs.washington.edu). Mutagenesis analysis was performed with Sanger and next-generation sequencing. After excluding PyMOL (pymol.org) based on the SPR protein structure 4XWY known rare variants in all 6 coding exons of GCH1 by Sanger (PDB file). Functional protein domains were analyzed using Hap- sequencing, exome sequencing was applied to 1 unaffected and 2 loReg v4.1 (broadinstitute.org/mammals/haploreg/haploreg.php). affected family members. Genomic DNA libraries were captured Urinary sepiapterin. First-void urine was collected, protected using the Nextera Rapid Capture Expanded Exome Kit (Illumina, from light, immediately placed on dry ice, and stored at 280°C. San Diego, CA), and DNA fragments were sequenced on an Illu- For sepiapterin quantification, samples were prepared as mina HiSeq2000 system. Variants were identified by a standard described7 and subjected to high-performance liquid chromatogra- analysis pipeline. phy analysis on a C18 reversed-phase column (Spherisorb ODS1, Variant filtering and validation. Variants with a minor allele 5 mm, ID 4.6 mm, L250 mm; Waters, Milford, MA) and . . frequency (MAF) 0.1%, based on 6,500 (The National isocratic elution with 24 mM KH2PO4 (pH 5.0)/CH3OH

2 Neurology: Genetics Table 1 Primer sequences for the SPR gene encoding sepiapterin reductase (SPR) and for rs70991108 located in the DHFR gene encoding dihydrofolate reductase (DHFR)

Primer GC content, Length Annealing Oligo name Sequence Orientation length Tm % product Gene temperature, °C

SPR_Ex1_F CCCGCCGACACCCGTA Forward 16 61.19 75 466 SPR 59.9

SPR_Ex1_R CCCTTAAATTCCCCAGCGGT Reverse 20 60.03 55

SPR_Ex2_F GCATGGGAAGAGGATATCTGG Forward 21 57.65 52.38 462 SPR 59.9

SPR_Ex2_R GGACTAGGCCGCTTGC Reverse 16 57.21 68.75

SPR_Ex3_F ACCCCCGACATAAAACAGGG Forward 20 59.67 55 473 SPR 55.3

SPR_Ex3_R CACAGCACAGACTCCTGACA Reverse 20 59.61 55

DHFR_rs70991108_F AGAAAAGGGGAATCCAGTCG Forward 20 57.21 50.00 473 DHFR 60.4

DHFR_rs70991108_R GAGACATGGCAGGGCAAG Reverse 18 58.08 61.11

(85:15, vol/vol) (flow rate 1.1 mL/min over 15 minutes). A fluo- Structural gene/exon variations of SPR were checked rescence detector (RF 20A XS; Shimadzu, Kyoto, Japan; excitation in the 3 samples in which whole-exome sequencing was l 5 l 5 425 nm; emission 530 nm) was used. Sepiapterin eluted performed using the BreakDancer tool (github.com/ with a retention time of 9.9 minutes. Urine samples were masked genome/breakdancer).8 We did not detect such an before analysis. Mean values of 4 technical measurements are presented. event near SPR or elsewhere on 2. In addition, we found an intronic 19 base-pair Statistical analyses. Descriptive analyses and x2 statistics were insertion (rs70991108) in an active transcription start calculated using Statistica 8.0 (StatSoft Inc., Tulsa, OK). Data are presented as mean 6 SEM. p , 0.05 was considered statistically site of the dihydrofolate reductase (DHFR) gene in all significant. affected family members (table 2). This variant alters 8 regulatory motifs (HaploReg). RESULTS Clinical description. We examined 5 affected and 4 nonaffected family members Urinary sepiapterin. SPR deficiency increases urinary 7 (figure, B). Symptom onset was between childhood sepiapterin levels. To investigate the consequences and puberty (table 2). Patient II.4 presented with of the variants observed in this study, we measured lower limb dystonia (tiptoe walking) and rigidity urinary sepiapterin concentrations in all members since childhood, with progressive course and diurnal (table 2). All clinically affected family members variation. He improved markedly with L-DOPA, had sepiapterin concentrations slightly above the 6 however with low-threshold drug-induced dyskine- range measured in normal controls (mean SD: 6 m sia. His current medication comprises fractionated 4.2 3.12 mol/mol creatinine), but below the 6 L-DOPA doses (750 mg/d), anticholinergics, and values in proven SPR-deficiency patients (727 m 7 amantadine (400 mg/d). The other affected in- 230 mol/mol creatinine). dividuals showed diurnal lower limb dystonia, rigid- Genotype-phenotype relationship. The family member ity, and mild pyramidal signs. Patients (II.2 and II.5) without the SPR variant (II.1) had normal sepiapterin experienced symptomatic benefit with L-DOPA and concentrations and was healthy. From the 8 SPR var- continued on pramipexole. Patients III.2 and III.3 iant carriers (table 2), 7 had elevated sepiapterin con- were not treated with L-DOPA because of their young centrations; 5 of those were clinically affected. age, mild symptoms, and fear of their parents of From 2 members without the DHFR variant L-DOPA–induced dyskinesia. (I.2 and III.1), 1 had elevated sepiapterin values, Genetic analysis. We found a nonsynonymous hetero- but both were healthy. From 7 DHFR variant carriers zygous variant in exon 1 of SPR (NM_003124.4 (table 2), 6 had elevated sepiapterin concentrations; c.207C.G, p.Asp69Glu, genomic [GRCh37]: 5 of those were clinically affected. chr2:73114768_C.G) in all affected family mem- Six family members carried both SPR and DHFR bers. In SPR, aspartate 69 is involved in binding variants (table 2); all of them had elevated sepiapterin the cofactor nicotinamide adenine dinucleotide phos- concentrations; 5 of them were clinically affected. phate (NADPH). In silico mutagenesis analysis pre- There was a significantly stronger association of dicted that the longer and more flexible side chain of both variants simultaneously with the biochemical glutamate 69 in mutation carriers leads to an abnormality (x2 5.1429, p , 0.05) and the clinical increased distance and loss of hydrogen bonds to disease state (x2 5.625, p , 0.05) as opposed to only NADPH (figure, C). 1 variant.

Neurology: Genetics 3 Table 2 Clinical, genetic, and biochemical data of the investigated DRD family

Variant Family Age at Disease Phenomenology Response to Sepiapterin member Sex examination, y Dystonia onset (in chronological order) L-DOPA SPR DHFR (mmol/mol Cr) 6 SEM

I.1 MND NANANA NA NANANA

I.2 F56 22 2 ND 12 3.75 6 0.63

II.1 M35 22 2 ND 21 2.99 6 0.53

II.2 F32 1 Childhood Dystonic posturing of 11115.44 6 2.07a both feet; with diurnal fluctuation, mild lower limb bradykinesia and rigidity, and extensor planter responses

II.3 M29 22 2 ND 11 8.47 6 0.72a

II.4 F27 1 Childhood Initially; lower limb dystonia, 11128.43 6 4.02a bradykinesia, and rigidity with diurnal variation progressed to generalized parkinsonism; L-DOPA–induced dyskinesia and pyramidal signs

II.5 M21 1 Puberty Lower limb dystonia (tiptoe walking) 1117.39 6 0.88a after prolonged walking; bilateral lower limb rigidity, postural tremor, and equivocal Babinski sign

III.1 M14 22 2 ND 12 17.74 6 3.92a

III.2 F10 1 Childhood Tiptoe walking with activity; ND 11 22.67 6 4.75a fast fatigability with dystonic posturing on the left foot

III.3 M5 1 Infancy Tiptoe walking; equinovarus ND 11 19.95 6 2.40a deformity

Abbreviations: Cr 5 creatinine; DHFR 5 dihydrofolate reductase; DRD 5 dopa-responsive dystonia; NA 5 not available; ND 5 not determined; SPR 5 sepiapterin reductase; 25not present; 15present; variant DHFR 5 chr5:79,950,163-/TGGCGCGTCCCGCCCAGGT (19 base-pair insert); variant SPR 5 chr2:73,114,768 C/G in exon 1. Normal range of urine sepiapterin: 4.2 6 3.12 mmol/mol Cr (mean of 10 controls without mutations in genes involved in the synthesis of L-DOPA). a Elevated values.

DISCUSSION We identified a rare heterozygous var- allele frequency in the general population,9 we do not iant of SPR in a family with DRD segregating as assume this variant to be pathogenic per se. autosomal dominant trait (figure, B and table 2). Of interest, the presence of the common DHFR SPR is essential for levodopa synthesis. In silico anal- variant in addition to the SPR variant significantly ysis suggested pathogenicity of this variant. Consis- increased the risk of the presence of the biochemical tently, we found slightly elevated urine sepiapterin abnormality (elevated sepiapterin urine concentra- concentrations in 7 variant carriers, 5 of these pre- tions) and of clinical symptoms. The 1 sibling carry- senting clinical symptoms. Since also 3 nonaffected ing both variants (SPR and DHFR) without clinical siblings carried this variant, incomplete penetrance DRD symptoms might be either in a preclinical dis- was assumed. Other SPR variants are known to cause ease stage or might be protected by unknown resil- autosomal recessive DRD. A heterozygous SPR vari- ience factors. ant in a single patient with DRD without family While high DHFR expression in the liver may history had already indicated the possibility of auto- account for a modest biochemical phenotype in the somal dominant causality of SPR variants.6 These periphery, the very low cerebral DHFR expression observations suggest that this SPR variant has autoso- might be further reduced by the variant, leading to mal dominant pathogenicity with incomplete a clinically significant disturbance of L-DOPA synthe- penetrance. sis.10 This suggests that the presence or absence of In the same family, we also identified a DHFR the common DHFR variant affects the penetrance of variant (rs70991108). In silico analysis suggested the autosomal dominant SPR variant. With regard to a potential effect of this variant on protein expression. the sequential metabolic function of the SPR and Consistently, we found elevated urine sepiapterin DHFR gene products in the L-DOPA synthesis path- concentrations in 6 and clinical symptoms in 5 of way (figure, A), this interpretation appears plausible. the 7 variant carriers. Two nonaffected siblings car- The SPR-DHFR pathway metabolizes sepiapterin ried the variant as well. Because of its relatively high to generate tetrahydrobiopterin, essential for the

4 Neurology: Genetics formation of L-DOPA. The SPR and DHFR no disclosures. G. Deuschl has served on the scientific advisory boards of are serially catalyzing this reaction. Homozygous muta- Medtronic, Sapiens, Britannica, and Boston Scientific; has received travel funding/speaker honoraria from Medtronic and Desitin; has served on tions affecting SPR on both homologous the editorial boards of Movement Disorders, Aktuelle Neurologie, and sev- impose a massive bottleneck within this pathway and eral other journals (names not indicated); and receives publishing royal- are known to cause DRD. A heterozygous mutation ties from Thieme Publishers. U. Müller has served on the editorial board of Neurogenetics and has received research support from CurePSP. affecting SPR on only 1 chromosome imposes a minor T. Opladen has received speaker honoraria from Swedish Orphan. bottleneck for this pathway. The clinical penetrance of B.-S. Petersen, A. Franke, and F. Hopfner report no disclosure. G. this condition may thus depend on additional cofactors Kuhlenbäumer has received research support from the International affecting this pathway, such as epigenetic silencing of Essential Tremor Foundation. G.U. Höglinger has served on the scien- tific advisory boards of Asceneuron, AbbVie, Bristol-Myers Squibb, the healthy SPR allele. In our family, this cofactor seems Roche, UCB, Novartis, Sanofi, and Biogen; has received travel fund- to be the co-occurrence of an additional variant in the ing/speaker honoraria from AbbVie, Roche, UCB, and Movement Dis- downstream DHFR, causing a “tandem steno- orders Society; holds a patent for Method for the treatment of sis” within this pathway. neurodegenerative diseases (PCT/EP2015/068734); and has received research support from Sellas Life Sciences Group, Bayerische Forschung- Taken together, the present study proposes the sallianz, German Ministry of Education and Research (BMBF), German rare SPR c.207C.G variant as a cause of autosomal Academic Exchange Service (DAAD), German Research Foundation dominant DRD with incomplete penetrance and the (DFG), German Center for Neurodegenerative Diseases (DZNE), Cur- ePSP, International Parkinson Foundation, and NOMIS Foundation. Go common DHFR rs70991108 variant as a potential to Neurology.org/ng for full disclosure forms. modifier, significantly affecting the penetrance of the SPR variant. Received May 24, 2017. Accepted in final form August 21, 2017.

AUTHOR CONTRIBUTIONS REFERENCES Ali S. Shalash: study concept and design, acquisition of data, analysis and 1. Klein C. Genetics in dystonia. Parkinsonism Relat Disord interpretation of data, and critical revision of the manuscript for intellec- 2014;20(suppl 1):S137–S142. tual content. Thomas W. Rösler: study concept and design, acquisition of 2. Wijemanne S, Jankovic J. Dopa-responsive dystonia— data, analysis and interpretation of data, study supervision, and critical clinical and genetic heterogeneity. Nat Rev Neurol 2015; revision of the manuscript for intellectual content. Stefanie H. Müller: 11:414–424. acquisition of data, analysis and interpretation of data, and critical revi- 3. Friedman J, Roze E, Abdenur JE, et al. Sepiapterin reduc- sion of the manuscript for intellectual content. Mohamed Salama: study concept and design, acquisition of data, analysis and interpretation of tase deficiency: a treatable mimic of cerebral palsy. Ann – data, and critical revision of the manuscript for intellectual content. Neurol 2012;71:520 530. Günther Deuschl: study concept and design, analysis and interpretation 4. Müller U. The monogenic primary dystonias. Brain 2009; of data, and critical revision of the manuscript for intellectual content. 132:2005–2025. Ulrich Müller and Thomas Opladen: study concept and design, acquisi- 5. Bonafe L, Thony B, Leimbacher W, Kierat L, Blau N. tion of data, analysis and interpretation of data, and critical revision of the Diagnosis of dopa-responsive dystonia and other tetrahy- manuscript for intellectual content. Britt-Sabina Petersen: acquisition of drobiopterin disorders by the study of biopterin metabo- data, analysis and interpretation of data, and critical revision of the lism in fibroblasts. Clin Chem 2001;47:477–485. manuscript for intellectual content. Andre Franke: study concept and 6. Steinberger D, Blau N, Goriuonov D, et al. Heterozygous design, acquisition of data, analysis and interpretation of data, study mutation in 59-untranslated region of sepiapterin reductase supervision, and critical revision of the manuscript for intellectual con- tent. Franziska Hopfner: study concept and design, acquisition of data, gene (SPR) in a patient with dopa-responsive dystonia. – analysis and interpretation of data, study supervision, and drafting the Neurogenetics 2004;5:187 190. manuscript for intellectual content. Gregor Kuhlenbäumer: study con- 7. Carducci C, Santagata S, Friedman J, et al. Urine sepiap- cept and design, acquisition of data, analysis and interpretation of data, terin excretion as a new diagnostic marker for sepiapterin study supervision, and critical revision of the manuscript for intellectual reductase deficiency. Mol Genet Metab 2015;115: content. Günter U. Höglinger: study concept and design, acquisition of 157–160. data, analysis and interpretation of data, study supervision, and drafting 8. Fan X, Abbott TE, Larson D, Chen K. BreakDancer: the manuscript for intellectual content. identification of genomic structural variation from paired-end read mapping. Curr Protoc Bioinformatics ACKNOWLEDGMENT 2014;45:15.6.1–15.6.11. The authors thank Lena Jaschkowitz and Magda Berjas for technical 9. Dutta S, Chattopadhayay S, Dasgupta C, Sarkar S. Low assistance. molecular weight heparin: a practical approach in deep venous thrombosis in palliative care. Indian J Palliat Care STUDY FUNDING 2011;17:143–145. Supported by the German Academic Exchange Service (DAAD, 10. Blau N, Bonafe L, Thony B. Tetrahydrobiopterin defi- Grants: 56502692, 57070724, 57162390, and 57247022). Günter U. ciencies without hyperphenylalaninemia: diagnosis and Höglinger was funded by the Deutsche Forschungsgemeinschaft (DFG, HO2402/6-2). genetics of dopa-responsive dystonia and sepiapterin reductase deficiency. Mol Genet Metab 2001;74: – DISCLOSURE 172 185. A.S. Shalash reports no disclosures. T.W. Rösler has served on the edi- 11. Neville BG, Parascandalo R, Farrugia R, Felice A. Sepiap- torial board of Frontiers in Neurology and will receive royalty payments for terin reductase deficiency: a congenital dopa-responsive a patented invention (Perk activator for the treatment of neurodegener- motor and cognitive disorder. Brain 2005;128: ative diseases, WO 2016024010 A1). S.H. Müller and M. Salama report 2291–2296.

Neurology: Genetics 5 Characterization of a KCNB1 variant associated with autism, intellectual disability, and epilepsy

Jeffrey D. Calhoun, PhD* ABSTRACT * Carlos G. Vanoye, PhD Objective: To perform functional characterization of a potentially pathogenic KCNB1 variant Fernando Kok, MD identified by clinical exome sequencing of a proband with a neurodevelopmental disorder that Alfred L. George, Jr., MD included epilepsy and centrotemporal spikes on EEG. Jennifer A. Kearney, PhD Methods: Whole-exome sequencing identified the KCNB1 variant c.595A.T (p.Ile199Phe). Bio- chemical and electrophysiologic experiments were performed to determine whether this variant affected protein expression, trafficking, and channel functional properties. Correspondence to Dr. Kearney: jennifer.kearney@northwestern. Results: Biochemical characterization of the variant suggested normal protein expression and edu trafficking. Functional characterization revealed biophysical channel defects in assembled homo- tetrameric and heterotetrameric channels. Conclusions: The identification of the KCNB1 variant c.595A.T (p.Ile199Phe) in a neurodevelop- mental disorder that included epilepsy with centrotemporal spikes expands the phenotypic spec- trum of associated with KCNB1 variants. The KCNB1-I199F variant exhibited partial loss of function relative to the wild-type channel. This defect is arguably less severe than pre- viously reported KCNB1 variants, suggesting the possibility that the degree of KCNB1 protein dysfunction may influence disease severity. Neurol Genet 2017;3:e198; doi: 10.1212/ NXG.0000000000000198

GLOSSARY MW 5 molecular weight; WT 5 wild type.

De novo KCNB1 variants have been identified in patients with early-infantile epileptic enceph- alopathy-26.1–6 The KCNB1 gene is intolerant of missense variation and is ranked among the – 5 7 top 1% 2% of intolerant genes in the genome (Zmis 5.15). KCNB1 encodes KV2.1, a voltage-gated expressed in a wide variety of neuron types. KV2.1 channels underlie the principal delayed rectifier current in many and contribute to regulation of neuronal excitability, especially during high-frequency firing.8 Characterization of previously reported KCNB1 patient variants revealed a variety of functional defects, including loss of ion selectivity, reduced conductance, and dominant negative effects.4–6 Here, we report a novel KCNB1 variant in an individual with a neurodevelopmental disorder that included epilepsy and centrotemporal spikes on EEG. The proband exhibits infrequent seizures that are well controlled on sulthiame, intellectual disability, disruptive behavior, and autistic traits. Functional characterization revealed partial loss-of-function biophysical defects of assembled homotetrameric and heterotetrameric channels.

METHODS Standard protocol approvals, registrations, and patient consents. Anonymized genetic and clinical summaries were reported to us following clinical whole-exome sequencing. A review by the Northwestern University International Review Board

*These authors contributed equally to this work. From the Department of Pharmacology (J.D.C., C.G.V., A.L.G., J.A.K.), Northwestern University Feinberg School of Medicine, Chicago, IL; and Research Center (F.K.), Biosciences Institute, University of Sao Paulo, Brazil; and Mendelics Análise Genomica (F. K.), Sao Paulo, Brazil. Funding information and disclosures are provided at the end of the article. Go to Neurology.org/ng for full disclosure forms. 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.

Neurology.org/ng Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 deemed this study as exempt, nonhuman subject research. Consent for release of deidentified medical information was ob- 0.011 tained by physicians in accordance with local institutional 0.001 5 policies. , p p 29; 33 26; Plasmids and transfection. I199F was introduced into 22 5 5 5 5 a human KV2.1 cDNA clone pIRES2-DsRed-MST-KV2.1- WT.6 Wild type (WT) and I199F were transiently expressed in 0.4; N 0.2; N 0.3; N 0.3; N 6

m 6 6 CHO-K1 cells by electroporation (Maxcyte STX). Briefly, 7.5 g 6 3 6

of cDNA was mixed with 10 10 viable cells resuspended in k 100 mL of electroporation buffer, transferred to a processing assembly (OC-100R10), and electroporated using the CHO protocol. Coexpression was achieved by coelectroporation with 5 m g each of pIRES2-DsRed-MST-KV2.1-I199F and pIRES2- 0.146 7.5 eGFP-K 2.1-WT using the same protocol. Following electro- 0.001 8.3 5 V , p poration, cells were incubated for 72 hours before experiments. p 29; 33 6.3 26; 22 6.1 5 5 5 Cell-surface biotinylation. Surface proteins on CHO-K1 cells 5

expressing WT or mutant KV2.1 were labeled with membrane- 2.3; N 1.4; N 1.6; N impermeable Sulfo-NHS-Biotin (Thermo Fisher Scientific, 1.4; N 6 6 6 Waltham, MA), as described.6 Total and surface fractions were 6 39.2 44.2 21.4 42.3 analyzed by immunoblotting using mouse anti-KV2.1 (1:250; ½ 2 2 2 K89/34; Neuromab, Davis, CA), mouse anti-transferrin receptor 2 (TfR) (1:500; H68.4; Invitrogen, Carlsbad, CA), and rabbit anti- calnexin (1:250; H70; Santa Cruz, Dallas, TX) primary anti- bodies. AlexaFluor680 and 790-conjugated goat anti-rabbit and 0.009 goat anti-mouse IgG secondary antibodies were used (1:20,000; 0.164 5 5 p Jackson ImmunoResearch, West Grove, PA). Blots were analyzed p using the OdysseyCLx system and ImageStudio software (LI- 31; 63 23; 26 5 5 5 COR). Calnexin immunoreactivity was absent from the cell- 5 surface fraction, confirming the selectivity of biotin labeling for 1.4; N 1.1; N 1.7; N cell-surface proteins. Normalized total, surface, and total:surface 0.7; N 6 6 6 6 ratios were compared using one-way analysis of variance. kV Electrophysiology. Automated planar-array patch-clamp recording was performed with the SyncroPatch 768 PE system (Nanion, München, Germany). Recording solutions are shown in table 1. Pulse generation was performed with PatchControl

software (Nanion), and whole-cell currents were acquired at 5 0.001 14.0 0.006 17.4 , 5 kHz and filtered at 1 kHz. Currents were not leak subtracted. Fast p p 23; 63 16.7 and slow capacitances were compensated using internal algo- 26 12.9 31; 5 5 rithms within PatchControl software. Whole-cell currents were 5 5 measured from a holding potential of 280 mV and elicited with 1.7; N 1.9; N 2 1 1.7; N depolarizing steps (500 milliseconds) from 80 to 60 mV (10 2.5; N 6 6 6 mV steps) followed by a step to 0 mV (tail currents, 250 milli- 6 23.4 1.9 12.9 10.9 seconds). Background currents were removed by digital sub- ½ Voltage dependence of activationV Voltage dependence of inactivation 1 2 2 2 traction of whole-cell currents recorded from nontransfected cells. 0.104 Table 1 Solutions for electrophysiologic 0.548 5 5 p recording p 27 40; 74; 5 185 5 External, mM Internal, mM 5 5 NaCl 140 10 26.0; N 12.6; N 12.6; N 6.6; N KCl 450 6 6 6 6 pA/pF

KF — 60 a Ip, 92.4 58.5 50.6 143.2

MgCl2 11

CaCl2 2 —

Glucose 5 — Biophysical channel properties of KCNB1-WT and KCNB1-I199F homotetramers and heterotetramers

— I199F EGTA 20 WT 1 1 HEPES 10 10 KCNB1-I199F KCNB1-WT WT WT Homotetramer Heterotetramer Table 2

pH 7.4 7.2 Peak current. a

2 Neurology: Genetics To remove bias for expression efficiency or the functional level, dependence of activation and inactivation were evaluated only homotetramer currents were analyzed from all cells with Rseal from cells with positive outward current values following back- $0.5 GV, Rseries #20MV, and Cm .2 pF. The peak current ground subtraction. Tail currents measured 5 milliseconds after was normalized for cell capacitance and plotted against voltage to stepping to 0 mV were normalized to the largest tail current generate peak current density-voltage relationships. Voltage amplitude and plotted against the depolarizing test potential

Figure Functional characterization of KCNB1-WT and KCNB1-I199F

10 (A) Location of the I199F variant mapped onto the crystal structure of a KV2.1/KV1.2 chimera (PDB 29R9). A channel tetramer is shown. I199F (magenta) lies in the S1 transmembrane segment of the voltage-sensing domain. (B) KCNB1- I199F is expressed and trafficked to the cell surface. Cell-surface expression was measured using cell-surface biotinylation of CHO-K1 cells transfected with wild-type (WT) or mutant KV2.1. Total (5 mg per lane) and surface fractions of KV2.1 were detected with the anti-KV2.1 antibody. Endogenous TfR levels were measured as a loading control. Calnexin was present in total cell lysates, but not surface fraction, demonstrating selective biotinylation of extracellular proteins. The blot shown is representative of 3 independent experiments. (C) Current-voltage relationships measured for KCNB1-I199F (n 5 74) and WT homotetrameric channels (n 5 185). (D) Voltage dependence of activation and inactivation curves calculated for KCNB1-I199F (n 5 23) and WT (n 5 63) homotetrameric channels. (E) Current-voltage relationships measured for KCNB1-I199F (n 5 40) and WT (n 5 27) heterotetrameric channels. (F) Voltage dependence of activation and inactivation curves calculated for KCNB1-I199F (n 5 31) and WT (n 5 26) heterotetrameric channels.

Neurology: Genetics 3 (280 to 160 mV). Normalized G-V curves were fit with the (p 5 0.104). Furthermore, there was a significant de- 5 1 2 Boltzmann function: G 1/(1 exp[(V V½)/k]) to deter- polarizing shift in voltage dependence of activation mine the voltage for half-maximal activation (V½) and slope relative to the WT alone, without affecting voltage factor (k). Voltage dependence of inactivation was assessed fol- dependence of inactivation (figure, F). Details of the lowing a 5-second prepulse from 290 to 140 mV (10 mV steps) followed by a 250-millisecond step to 160 mV. Normalized peak biophysical characterization are shown in table 2. currents measured at the 160 mV step were plotted against the 5 DISCUSSION prepulse voltage and fit with Boltzmann functions: I/Imax 1/(1 The nature of channel dysfunction for 1 2 exp[(V V½)/k]) to determine the voltage for half-maximal KCNB1-I199F differs from previously reported 1 inactivation (V½) and slope factor (k). For heteromeric (WT KCNB1 variants. I199F resulted in biophysical de- I199F) channels, peak current density-voltage relationships and fects predicted to confer reduced channel availability voltage dependence of channel activation and inactivation were due to shifts in the voltage dependence of activation, evaluated from cells with positive outward current values after background subtraction. Outliers (.62 SDs) were not included. consistent with partial loss of function. By contrast, Results are expressed as mean 6 SEM. p Values for comparisons previously reported variants had complete loss of between the mutant and WT (Student t test) are listed in table 2 function, dominant negative effects, or loss of ion and in figure. selectivity.4–6 Previously characterized KCNB1 var- iants were all located in the pore domain,4–6 while RESULTS Clinical whole-exome sequencing was I199F is the first reported variant located in the S1 performed in an adolescent with intellectual disabil- transmembrane segment of the voltage-sensing ity, disruptive behavior, autistic features, epilepsy and domain. KCNB1 variants have been reported in centrotemporal spikes on EEG. Seizures are infre- other voltage-sensing domain segments, but none quent and well controlled with sulthiame. Exome have been functionally characterized to date.3 How- sequencing revealed a single de novo missense variant ever, variants in this domain of other voltage-gated in KCNB1 c.595A.T (p.Ile199Phe), which was potassium channels result in disease.9 The I199F confirmed by Sanger sequencing. This variant was proband has neurodevelopmental and behavioral absent in the ExAC and gnomAD databases con- symptoms that are consistent with prior reports. taining exomes from 123,126 individuals unaffected However, seizures were reported to be infrequent and by severe pediatric disease.7 Isoleucine 199 is located well controlled with sulthiame, whereas previously in the S1 transmembrane segment of the voltage- reported cases with pore variants had frequent, drug- sensing domain (figure, A) and is evolutionarily resistant seizures.1,2,4–6 This suggests the possibility invariant through chordates. that the variant type may influence the seizure phe- Effects of the KCNB1-I199F variant on channel notype and/or drug responsiveness. Consistent with function were evaluated following transient expres- this, a recent report suggests that the variant position sion. CHO-K1 cells expressing KCNB1-I199F had (pore vs voltage sensor) influences the penetrance and total and cell-surface expression similar to the WT, severity of electroclinical and seizure phenotypes, indicating that channels were expressed and trafficked while neurodevelopmental symptoms are present in to the cell surface (figure, B). KCNB1-I199F had all cases to date.3 Larger cohorts with drug-response a larger proportion of low-molecular-weight (MW) data will be necessary to investigate this link. species relative to the WT in the total fraction (figure, The phenotype spectrum associated with KCNB1 B). Kv2.1 undergoes extensive multisite phosphory- variants will continue to expand as it becomes more lation that affects apparent MW and function.8 How- widely screened. Correlating the clinical phenotype ever, there was no apparent MW difference in surface with channel dysfunction may provide insights into fractions, suggesting that mutant and WT channels at targeted therapeutic strategies. the cell surface have similar posttranslational modifications. AUTHOR CONTRIBUTIONS Jeffrey D. Calhoun performed functional evaluation of the variant and Automated planar patch-clamp recording in parallel wrote and edited the manuscript. Carlos G. Vanoye performed functional of I199F and WT homomeric channels resulted in evaluation of the variant, performed statistical analysis of electrophysiology characteristic voltage-dependent potassium currents data, and edited the manuscript. Fernando Kok identified the KCNB1 with outward rectification and late inactivation. The variant and edited the manuscript. Alfred L. George, Jr. conceived the study and edited the manuscript. Jennifer A. Kearney conceived the study, per- current-voltage relationship was similar between formed variant interpretation, and wrote and edited the manuscript. I199F and WT (figure, C). I199F exhibited significant depolarizing shifts in the voltage dependence of activa- ACKNOWLEDGMENT tion and inactivation relative to the WT (figure, D). The authors thank Katarina Fabre for technical support and the patient To mimic the heterozygous condition, we coexpressed and family for their cooperation.

KCNB1-I199F plus WT. Coexpression of KCNB1- STUDY FUNDING I199F with the WT trended toward decreased current This work was supported by a grant from the NIH/National Institutes of amplitudes between 210 mV and 60 mV (figure, E) Neurologic Disorders and Stroke to J.A.K. (R01 NS053792). The

4 Neurology: Genetics funders had no role in study design, data collection and analysis, decision 2. de Kovel CG, Brilstra EH, van Kempen MJ, et al. Tar- to publish, or preparation of the manuscript. geted sequencing of 351 candidate genes for epileptic encephalopathy in a large cohort of patients. Mol Genet DISCLOSURE Genomic Med 2016;4:568–580. J.D. Calhoun receives fellowship support from the Dravet Syndrome 3. de Kovel CGF, Syrbe S, Brilstra EH, et al. Neurodevelop- Foundation. C.G. Vanoye consults for Nanion technologies. F. Kok mental disorders caused by de novo variants in KCNB1 gen- has served on the editorial board of Arquivos de Neuropsiquiatria; holds otypes and phenotypes. JAMA Neurol 2017;74:1228–1236. a patent for Methylmalonic acid determination by tandem mass spec- 4. Saitsu H, Akita T, Tohyama J, et al. De novo KCNB1 trometry using stable isotope; is an employee and shareholder of Men- delics Análise Genomica; and has been a speaker at NPC symposia from mutations in infantile epilepsy inhibit repetitive neuronal Actelion Pharmaceuticals. A.L. George, Jr. has served on a scientific firing. Sci Rep 2015;5:15199. advisory board of Amgen; has received travel funding from Praxis Pre- 5. Thiffault I, Speca DJ, Austin DC, et al. A novel epileptic cision Medicines, Inc.; has served on the editorial boards of the Journal of encephalopathy mutation in KCNB1 disrupts Kv2.1 ion Clinical Investigation, Heart Rhythm Journal, and the Journal of General selectivity, expression, and localization. J Gen Physiol Physiology; is funded by the NIH, Merck Sharp & Dohme, St. Baldrick’s 2015;146:399–410. ’ Foundation, AHC Foundation, and Simon s Foundation; received 6. Torkamani A, Bersell K, Jorge BS, et al. De novo KCNB1 research support from Xenon Pharmaceuticals and Dravet syndrome mutations in epileptic encephalopathy. Ann Neurol 2014; foundation; and consulted for Gilead Sciences and Xenon Pharmaceut- 76:529–540. icals. J.A. Kearney is funded by the NIH and Ovid Therapeutics; consults 7. Lek M, Karczewski KJ, Minikel EV, et al. Analysis of for Praxis Precision Medicine; received research support from Sage Ther- apeutics, Xenon Pharmaceuticals, and Pairnomix, LLC; received license protein-coding genetic variation in 60,706 humans. – fee payments from Xenon Pharmaceuticals, Novartis, and Life Splice Nature 2016;536:285 291. Technologies; and served on a scientific advisory board of Zogenix Inc. 8. Mohapatra DP, Park KS, Trimmer JS. Dynamic regulation Go to Neurology.org/ng for full disclosure forms. of the voltage-gated Kv2.1 potassium channel by multisite . Biochem Soc Trans 2007;35:1064–1068. Received May 17, 2017. Accepted in final form October 15, 2017. 9. Jen JC, Graves TD, Hess EJ, et al. Primary episodic atax- ias: diagnosis, pathogenesis and treatment. Brain 2007; REFERENCES 130:2484–2493. 1. Allen NM, Conroy J, Shahwan A, et al. Unexplained early 10. Long SB, Tao X, Campbell EB, et al. Atomic structure of onset epileptic encephalopathy: exome screening and phe- a voltage-dependent K1 channel in a lipid membrane-like notype expansion. Epilepsia 2016;57:e12–e17. environment. Nature 2007;450:376–382.

Neurology: Genetics 5 Clinical features and outcome of 6 new patients carrying de novo KCNB1 gene mutations

Carla Marini, MD, PhD ABSTRACT Michele Romoli, MD Objective: To describe electroclinical features and outcome of 6 patients harboring KCNB1 Elena Parrini, PhD mutations. Cinzia Costa, MD, PhD Methods: Clinical, EEG, neuropsychological, and brain MRI data analysis. Targeted next- Davide Mei, MSc generation sequencing of a 95 epilepsy gene panel. Francesco Mari, MD, PhD Results: The mean age at seizure onset was 11 months. The mean follow-up of 11.3 years docu- Lucio Parmeggiani, MD mented that 4 patients following an infantile phase of frequent seizures became seizure free; Elena Procopio, MD the mean age at seizure offset was 4.25 years. Epilepsy phenotypes comprised West syndrome Tiziana Metitieri, in 2 patients, infantile-onset unspecified generalized epilepsy, myoclonic and photosensitive Psyd, PhD eyelid myoclonia epilepsy resembling Jeavons syndrome, Lennox-Gastaut syndrome, and focal Elena Cellini, PhD epilepsy with prolonged occipital or clonic seizures in each and every one. Five patients had Simona Virdò, MSc developmental delay prior to seizure onset evolving into severe intellectual disability with Dalila De Vita, BSc absent speech and autistic traits in one and stereotypic hand movements with impulse control Mattia Gentile, MD, PhD disorder in another. The patient with Jeavons syndrome evolved into moderate intellectual dis- Paolo Prontera, MD, PhD ability. Mutations were de novo, 4 missense and 2 nonsense, 5 were novel, and 1 resulted from Paolo Calabresi, MD somatic mosaicism. Renzo Guerrini, Conclusions: KCNB1-related manifestations include a spectrum of infantile-onset generalized or MD, FRCP focal seizures whose combination leads to early infantile epileptic encephalopathy including West, Lennox-Gastaut, and Jeavons syndromes. Long-term follow-up highlights that following a stormy phase, seizures subside or cease and treatment may be eased or withdrawn. Cognitive and motor Correspondence to functions are almost always delayed prior to seizure onset and evolve into severe, persistent impair- Dr. Guerrini: [email protected] ment. Thus, KCNB1 mutations are associated with diffuse brain dysfunction combining seizures, motor, and cognitive impairment. Neurol Genet 2017;3:e206; doi: 10.1212/NXG.0000000000000206

GLOSSARY ACMG 5 American College of Medical Genetics and Genomics; AED 5 antiepileptic drug; ASD 5 autism spectrum disorder; CSWS 5 continuous spikes and waves during slow-wave sleep; EIEE 5 early infantile epileptic encephalopathy; ESP 5 Exome Sequencing Project; ExAC 5 Exome Aggregation Consortium; FS 5 febrile seizures; ID 5 intellectual disability; NGS 5 next-generation sequencing; TCS 5 tonic-clonic seizures; VPA 5 valproic acid.

The KCNB1 gene encodes the pore-forming and voltage-sensing b subunit of the voltage- gated potassium (K1) channel subfamily 2 (Kv2.1) that is expressed throughout the brain1–3 and plays essential roles in regulating neuronal excitability, contributing to repolarization4 and dynamic modulation of neuronal activity.5–7 De novo heterozygous missense (number 5 13) and nonsense (number 5 1) mutations of KCNB1 have been reported in 14 patients with neurodevelopmental disorders, including epilepsy of variable severity in 13 of them.8–15 Probands carrying KCNB1 mutations resulting in severe epilepsy

From the Pediatric Neurology Unit (C.M., E.P., D.M., F.M., T.M., E.C., S.V., D.D.V., R.G.), Neurogenetics and Neurobiology Laboratories, Neuroscience Department, A. Meyer Pediatric Hospital, University of Florence; Neurology Unit (M.R., C.C., P.C.), Department of Medicine, University of Perugia, Ospedale S. Maria della Misericordia; Child Neurology Service (L.P.), Hospital of Bolzano; Metabolic Unit (E.P.), A. Meyer Pediatric Hospital, Florence; Medical Genetics Unit (M.G.), Azienda Sanitaria Locale Bari; Neonatology Unit and Prenatal Diagnosis (P.P.), Medical Genetic Unit, Ospedale S. Maria della Misericordia, Perugia; Department of Experimental Neurosciences (P.C.), “Istituto di Ricovero e Cura a Carattere Scientifico,” IRCCS Santa Lucia Foundation, Rome; and IRCCS Stella Maris Foundation (R.G.), Calambrone, Pisa, Italy. Funding information and disclosures are provided at the end of the article. Go to Neurology.org/ng for full disclosure forms. The Article Processing Charge was funded by the University of Florence. 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.

Neurology.org/ng Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 phenotypes have been classified as epileptic Hospital or referred from other national or international epilepsy encephalopathy–early infantile 26 (OMIM: centers for genetic testing, we identified and studied 6 patients * carrying de novo variants of the KCNB1 gene (GenBank Acces- EIEE26, 600397). Table 1 summarizes the sion Number: NM_004975.2). Three patients were examined at clinical and genetic features of previously the Neurology unit, and 3 were referred for genetic testing to our reported patients. Neurogenetic Laboratory. Collection and analysis of retrospective Here, we describe the electroclinical fea- clinical, EEG, neuropsychological, neuroimaging data were per- formed using a specific format filled by the treating specialist tures and long-term follow-up of 6 novel pa- aiming to obtain accurate and homogenous information. We tients carrying de novo missense and classified seizure types and epilepsy/syndromes according to the nonsense KCNB1 mutations and provide an International League Against Epilepsy guidelines.16,17 overall view of the spectrum of epilepsy phe- Molecular genetic testing. Genomic DNA was extracted from notypes associated with KCNB1 mutations, peripheral blood leukocytes using a QIAsymphony SP robot ’ rangingfromLennox-Gastautsyndrometo (Qiagen, Hilden, Germany) according to the manufacturer s pro- tocol. High-quality DNA was quantified using a QuantiFluor remitting infantile spasms and to mild gener- Fluorometer (Promega, Madison, WI). Targeted resequencing of alized epilepsy with eyelid myoclonia. Our a panel including 95 genes associated with epilepsy was per- small cohort of patients also suggests that formed in all 6 patients. Target enrichment and library prepara- tion were performed using a custom-designed Nextera Rapid the KCNB1 EIEE26 manifests infantile- Capture assay (Illumina, San Diego, CA), and sequencing was onset seizures with a tendency to attenuate performed on an Illumina MiSeq (Illumina) with a 2 3 150-bp orresolvecompletelyovertimeinsomepa- paired-end protocol as previously described.18 Variants were 19 tients. Concomitant impairment of cognitive annotated and filtered using the ANNOVAR tool. Variants localized in intronic regions outside the 10-bp exon-flanking and motor functions, constant and often boundaries and in the 59- and 39-UTR regions were excluded. severe, persists over time and is at times com- Variants reported in the Exome Aggregation Consortium (ExAC) bined with absent speech, autism spectrum database (exac.broadinstitute.org/) and/or in the 1000 Genomes disorder (ASD), and psychiatric problems. Project (in the National Heart, Lung, and Blood Institute Exome Sequencing Project [ESP], ESP6500 database, evs.gs.washington. Mutations in this gene, which are not rare, edu/EVS), in the Genome Aggregation Database (gnomAD) confirm that genes coding for potassium (gnomad.broadinstitute.org/), with a minor allele frequency .0. channels are important genetic contributors 01 (1%) were dropped out. In silico prediction of mutations’ pathogenicity was obtained using the dbNSFP database (v3.3), to epilepsies and more broadly to neurodeve- which provides functional prediction scores on more than 20 lopmental disorders. different algorithms (sites.google.com/site/jpopgen/dbNSFP). Putative causative variants were analyzed by Sanger sequencing to METHODS From a highly heterogenous cohort of 873 pa- confirm the next-generation sequencing (NGS) results in pro- tients with pediatric epilepsies, either refractory or benign, with bands and investigated in the parents to check the inheritance a supposed genetic etiology, recruited at the Meyer Children’s status. We classified variants according to the international

Table 1 Clinical and genetic summary of previously published patients

Patient Ethnicity Phenotype Genotypes Genetic analysis Inheritance Reference

1 Chinese 2 y: TC sz, spasm, drug resistant; p.Gly379Val WES De novo 15 EEG: GSW; 3 y: regression

3 Caucasian EOEE p.Phe416Leu; p.Arg312His; NGS gene panel De novo 14 p.Tyr533*

1 Caucasian ID, severe speech impairment, no epilepsy p.Val378Leu° WES De novo 13

2 Caucasian EOEE p.Gly381Arg; p.Phe416Leu WES De novo 12

1 Caucasian EE p.Val378Ala° WES De novo 11

2 Asian DD and severe infantile generalized sz p.Arg306Cys; p.Gly401Arg WES De novo 10

1 Caucasian EE p.Gly379Arg^ WES De novo 9

3 Caucasian Sz onset 4.75 y; EE, hypotonia, DD, p.S347Arg WES De novo 8 intermittent agitation

Sz onset 8 mo; drug-resistant sz, DD, p.Gly379Arg^ WES De novo absent speech, stereotyped handwringing movements

Sz onset: first year of life: TC, atypical absence, atonic, p.Thr374Ile WES De novo infantile spasms, and focal dyscognitive

Abbreviations: ^, ° 5 recurrent mutations; * 5 nonsense mutation; DD 5 developmental delay; EE 5 epileptic encephalopathy; EOEE 5 early-onset epileptic encephalopathy; GSW 5 generalized spikes and waves; ID 5 intellectual disability; mo 5 months; NGS 5 next-generation sequencing; TC 5 tonic-clonic seizures; sz 5 seizures; WES 5 whole-exome sequencing; y 5 years.

2 Neurology: Genetics guidelines of the American College of Medical Genetics and heterogeneous cohort of patients with potentially Genomics (ACMG) Laboratory Practice Committee Working genetic epilepsies, studied with our target resequencing 20 Group. gene panel.18 Table 2 summarizes the main clinical – Standard protocol approvals, registrations, and patient features of each KCNB1 mutation positive patient. consents. Written informed consent to disclose clinical informa- Epilepsy. The mean age at seizure onset was 11 tion was obtained from all parents/guardians of the participants. months (median 9, range 5–23 months). Four of Written informed consent for genetic testing was obtained ac- cording to the Paediatric Ethic Committee of the Tuscany Region the 6 patients (66.6%) following an initial phase of (approval no. 2014-0000559) in the context of the DESIRE pro- active epilepsy with frequent seizures became seizure ject (grant agreement no. 602531). free at a median age of 4.25 years; the mean duration of seizure remission at the last follow-up was 6 years RESULTS The 6 patients (4 females and 2 males) har- (range 8 months–16 years). boring KCNB1 mutations had a mean age at the time Seizure types (table 2) included infantile spasms of the study of 12 years (range 7–22 years). They (patients 1 and 2), myoclonic seizures with eyelid represented the 0.75% (6 of 873) of a highly myoclonia (table 2, patient 3), febrile seizures (FS)

Table 2 Clinical and genetic summary of the 6 patients reported harboring KCNB1 mutations

ID 1 2 3 4 5 6

Age, y/sex 7/M 8/F 22/F 12/F 8/F 17/M

Sz onset/ 9 mo/18 mo 5 mo/8 mo 6 mo/10 mo; 7 y/16 y 18 mo FS; 23 mo/7 y 12 mo/ongoing, 11 mo: FS; ongoing offset yearly sz from 8 y daily sz

Sz type Spasms Spasms Myoclonic, eyelid myoclonia and FS . focal: occipital, TCS, atypical Ab Focal, Ab, self-induced Ab-eyelid myoclonia, prolonged lateralized nonconvulsive status sleep-related TCS motor; monthly epilepticus, TC, tonic frequency

Epilepsy type West syndrome West syndrome Myoclonic, photosensitive Focal symptomatic Infantile onset Lennox-Gastaut or syndrome generalized (Jeavons syndrome) generalized syndrome

EEG

Onset Hypsarrhythmia Hypsarrhythmia GSW/myoclonic jerks Focal F-T spike Bilateral SW R . LNA increased in sleep

Current Backgr: delayed Normal Normal backgr; GSW and PSW; Multifocal/diffuse Slow backgr; Slow backgr; focal L spikes eyelid myoclonia; generalized abnormalities in sleep at FC discharges prolonged photosensitivity times resembling CSWS L . R discharges of bilateral SW, at times subcontinuous

Treatment

Previous ACTH ACTH CBZ, VPA, TPM, LEV, LTG, ESM, CLN VPA, LEV, pyridoxine, VPA VPA, RFN, BDZ, STP, BDZ, PHT ACTH, LEV, LTG, primidone, CBZ, TPM, ESM

Current ——VPA 1 ESM 1 CLN LTG VPA 1 TPM VPA 1 RFN 1 perampanel lorazepam 1 neuroleptics

Development Developmental Developmental Borderline development 7 y: mild ID, Developmental delay: Developmental Developmental delay: severe ID delay: severe ID with learning disability, motor and severe ID delay: severe ID delay: severe ID verbal dyspraxia; 11 y: moderate ID; 16 y: moderate ID, major impairment of oculomanual coordination and slowness of execution

Other Hypotonia, No speech; R leg Mild hypotonia, motor and verbal Significant sleep Hypotonia, Hypotonia, language: language: single hyperchromic dyspraxia, clumsiness, enuresis problem, absent speech, nonverbal, single words, words skin spot; ASD stereotypic hand behavioral disorder bilateral movements, clinodactyly hyperlaxity, ataxia, trichotillomania, pavor nocturnus

Mutation c.1109G.Ap. c.1747C.Tp. c.916C.T p.Arg306Cys c.586A.T p.Ile196Phe c.629C.Tp. c.1045G.Tp. Trp370* Arg583* Thr210Met Val349Phe (mosaic)

Inheritance De novo De novo De novo De novo De novo De novo

Abbreviations: Ab 5 absence seizures; ACTH 5 adrenocorticotropic hormone; ASD 5 autism spectrum disorder; backgr 5 background; BDZ 5 benzodi- azepine; CBZ 5 carbamazepine; CLN 5 clonazepam; CSWS: continuous spikes and waves during slow-wave sleep; ESM 5 ethosuximide; F 5 frontal; FC 5 frontocentral; FS 5 febrile seizures; GSW 5 generalized spikes and waves; ID 5 intellectual disability; L 5 left; LEV 5 levetiracetam; LTG 5 lamotrigine; NA 5 not available; PHT 5 phenytoin; PSW 5 polyspike-wave discharges; R 5 right; RFN 5 rufinamide; STP 5 stiripentol; SW 5 spike wave; sz 5 seizures; T 5 temporal; TCS 5 tonic-clonic seizures; TPM 5 topiramate; VPA 5 valproic acid.

Neurology: Genetics 3 followed by focal seizures (patients 4 and 5), absences myoclonic seizures and absences with eyelid myoclo- and nonconvulsive status epilepticus (patient 6), and nia with persistent generalized photosensitivity. Her afebrile tonic-clonic seizures (TCS) followed by other epilepsy started at 6 months, with bilateral myoclonic generalized seizure types (patients 5 and 6). For each jerks, and was very active for 3 months when treat- patient, we considered age of symptom appearance ment with valproic acid (VPA) prompted a prolonged, and age at the time of the study and calculated a mean yet temporary, resolution of symptoms. Indeed, at age follow-up of 11.3 years when last seen (median 8.5 7 years, during primary school age, myoclonic years, range 6–21 years). seizures and absences with eyelid myoclonia with Two patients with infantile spasms (patients 1 photosensitivity were observed (figure 1A). Episodes and 2), classified as West syndrome due to their hyp- of absences with eyelid myoclonia occurred fre- sarrhythmic EEG and co-occurring developmental quently on eye closure or were autoinduced like seen delay, exhibited an identical phenotype and outcome. in Jeavons syndrome.21 Two generalized TCS during Both patients were immediately treated with adreno- sleep were also reported around age 15 years. The corticotropic hormone (ACTH) after the onset of patient was trialed with several AEDs, but only a com- spasms, with a prompt remission of spasms and bination of VPA, ethosuximide, and clonazepam was EEG improvement. After 6–7 years of follow-up, effective at age 16 years. At the time of the study, this both patients with “remitting infantile spasms” patient was 22 years old and is still seizure free on remain seizure free and off antiepileptic drugs AEDs. Patients 4 and 5 (table 2) exhibited infantile- (AEDs). Yet, their development was delayed and cur- onset multifocal and generalized seizures preceded by rently sets on severe impairment of cognitive and global developmental delay. Then, patient 4 mani- motor functions including absent speech. fested monthly seizures from age 2 to 5 years when The remaining 4 patients manifested different and seizures ceased, she is currently 12, still on AEDs, and heterogeneous phenotypes. Patient 3 (table 2) showed seizure free. Despite 5 years of seizure freedom, her an interesting and peculiar phenotype consisting of cognitive and motor functions are yet severely

Figure 1 Representative EEG findings in 3 patients

(A) Ictal videopolygraphic EEG recording of patient 3 (table 2), showing discharges of diffuse polyspike waves concomitant with myoclonic jerks evident on the leads recording from both deltoid muscles. (B) Ictal videopolygraphic EEG recording of patient 3 (table 2), showing generalized paroxysmal activity evoked by eye closure with concomitant eyelid myoclonia (video available as supplementary material). (C) Interictal EEG recording of patient 5, showing bifrontal single complexes of spikes and waves with left predominance. (D) Interictal sleep EEG recording of patient 4, showing very frequent paroxysmal activity over the fronto-centro-temporal regions (continuous spikes and waves during the slow-wave sleep pattern).

4 Neurology: Genetics impaired with absent speech and autistic traits. childhood, into mild cognitive impairment with Patient 5 has ongoing yearly seizures and severe motor and verbal dyspraxia and poor coordination. cognitive impairment. The last patient (patient 6) Despite such a cognitive profile, she attended primary exhibited the most refractory epilepsy with infantile- school unassisted. While during secondary school, she onset FS followed by focal seizures, frequent absences, required a special needs teacher, and further neuro- nonconvulsive status, tonic, and TCS configuring psychological testing documented moderate ID a phenotype resembling Lennox-Gastaut syndrome. (Wechsler Intelligence Scale for Children—Revised: He is currently 17 and has daily atypical absences and full-scale IQ 5 45). Yet, she was able to reach sleep-related TCS. Concomitant with such refractory secondary school graduation. epilepsy, he manifested early developmental delay Five of the 6 patients acquired independent walk- evolving into moderate to severe intellectual disability ing, with a variable degree of hypotonia, hyperlaxity, (ID) with a behavioral disorder requiring antipsy- and mild ataxia in 3. One patient was using a wheel- chotic drugs. chair (patient 1). EEG recordings. All patients underwent repeated MRI. All patients underwent normal brain MRI. EEG recordings from seizure onset to current age. Ac- Genetic. Heterozygous KCNB1 mutations occurred cording to their epilepsy phenotype, at onset, 2 pa- de novo (table 2, figure 2) and were classified as path- tients presented a severely abnormal EEG with ogenic according to the ACMG guidelines. Both hypsarrhythmia that overtime evolved into a normal patients 1 and 2 carried a novel de novo nonsense EEG in 1 (patient 2) and slow background activity mutation, including the c.1109G.A [p.(Trp370*)] with rare left temporal spikes in the second patient variant that is located in the pore domain of the pro- (patient 1). Video-EEG polygraphic recordings of tein and the c.1747C.T [p.(Arg583*)] variant, the patient with Jeavons syndrome (patient 3) which falls in the C-terminal domain. Patient 3 har- showed, at onset and during follow-up, generalized bored the c.916C.T [p.(Arg306Cys)] missense var- spike- and polyspike-wave discharges with a prom- iant, which is located in the S4 segment of the inent generalized photoparoxysmal response (figure voltage-sensor domain of the protein. This variant 1B). Several episodes of myoclonia and absences with had previously been reported in a patient with eyelid myoclonia were recorded (figure 1, A and B). developmental delay and infantile-onset seizures.10 Patient 4 (table 2), with infantile-onset multifocal Patient 4 carried the novel c.586A.T [p.(Ile196- epilepsy, consistently showed slow background with Phe)] missense variant, located in the S1 segment of multifocal or diffuse paroxysmal activity in sleep and the protein, and patient 5 carried the novel semiperiodic posterior fast activity runs on awaken- c.629C.T [p.(Thr210Met)] missense variant, ing. Some EEG recordings in this patient showed, in located in the first extracellular domain, between S1 early childhood, a relevant increase in paroxysmal and S2. Patient 6 carried the novel c.1045G.T activity during sleep, resembling the continuous spike [p.(Val349Phe)] missense variant, located in the S5 and wave during the slow-wave sleep (CSWS) EEG segment of the pore; both NGS and Sanger pattern (figure 1C). The EEG recordings in patient 5 sequencing data showed evidence of somatic mosai- (table 2) exhibited slow background activity and cism (figure e-1, http://links.lww.com/NXG/A1). intermittent bilateral frontocentral spike and wave None of the KCNB1 variants were reported in the discharges with left predominance (figure 1B). EEG public available allele frequency databases (table e-1, recordings of patient 6 showed slow background and http://links.lww.com/NXG/A0), and the tools frequent, at times almost continuous multifocal and included in the dbNSFP, comprising the MetaSVM diffuse spike and wave discharges. and MetaLR ensemble scores, predicted the 4 mis- Cognition and behavior. Five of the 6 patients sense variants to be damaging (table e-1, http://links. (table 2, patients 1, 2, and 4–6) had clear-cut devel- lww.com/NXG/A0). opmental delay prior to seizure onset, involving both motor and cognitive functions. Their cognitive out- DISCUSSION We describe the electroclinical fea- come evolved into severe impairment with absent tures and outcome of 6 patients harboring de novo speech and autistic traits in 1 (table 2, patient 4). In heterozygous KCNB1 mutations, confirming that addition, one of them (table 2, patient 5) manifested mutations in this gene are not rare. Indeed, in our stereotypic hand movements and impulse control highly heterogenous cohort of pediatric epilepsies disorder consisting of trichotillomania, and patient 6 with a supposed genetic etiology, nearly 1% of pa- (table 2) developed a psychiatric disorder requiring tients carried mutations in this gene. treatment with neuroleptics. Genetic studies, from the “old” linkage analyses The girl with Jeavons syndrome instead had a de- studies to the modern era of whole-exome sequenc- layed early development (sitting: 8 months, ambula- ing/whole-genome sequencing techniques, have dem- tion: 15 months; and language: 2 years) evolving, in onstrated that heterozygous mutations in several

Neurology: Genetics 5 Figure 2 Schematic representation of the Kv2.1 protein with mutation distribution

Structure of the human Kv2.1 channel including previously published mutations and those reported here (specified under each mutation). The protein topol- ogy was performed using the Protter online tool (29) with the Uniprot accession Q14721 (KCNB1_HUMAN).

genes coding for different K1 channels, including (including our 6 probands and the 14 previously KNCQ2, KCNQ3, KCNA2, KCNA1, KCNC1, published, for a total of 20, of whom 19 with seiz- KCNH1, KCNMA1, HCN2, and KCNT1, are ures), we can observe that (1) TCS are the common- involved in epileptic disorders whose severity varies est seizure type observed in 57.8% (11/19) of from benign forms to early infantile epileptic enceph- patients; (2) focal seizures are the second most com- alopathy (EIEE).22–32 The list of K1 channel genes mon type, reported in 47% (9/19); (3) infantile implicated in epilepsy includes KCNB1 that was, in spasms are also a relatively common manifestation 2014, linked to the EIEE26.8 At present, there are 14 occurring in 31.5% (6/19); (4) approximately 15% patients on record carrying mutations in this gene, (3–4/19) of patients have either absences, atonic, or and 13 of them exhibit epilepsy.8–15 tonic seizures; (5) single patients might have peculiar All patients, including the 6 reported here, present and unusual phenotypes consisting of CSWS, non- with early motor and cognitive delay and, thence- convulsive status epilepticus (NCSE) or myoclonic forth, with different types of infantile-onset seizures. seizures or eyelid myoclonia, and photosensitive seiz- Review of previously published patients shows that ures. Of course, most patients manifest more than 1 both focal and generalized seizures including tonic- seizure type that combined the result in unspecified clonic, tonic, atonic, absences, myoclonic seizures, EE or more rarely to known epilepsy syndromes and infantile spasms can be equally observed (figure including West, Lennox-Gastaut, and Jeavons 3A). Epilepsy, often classified as early-onset epileptic syndromes. encephalopathy, has been reported as severe and Likewise, previously published patients, in our refractory to AEDs in most patients. In our small small series, seizures started in infancy, around 1 year series, 2 of the 6 patients were diagnosed as West of age, and were also drug resistant at onset. Yet, over syndrome (figure 3B). Of the remaining 4, 2 had time in 4 of our 6 patients (66.6%), seizure frequency generalized epilepsy with myoclonic seizures and eye- attenuated and disappeared, still being under remis- lid myoclonia with photosensitivity, resembling Jea- sion after a seizure-free period of 6 years. Two pa- vons syndrome in 1 (figure 3B). Focal epilepsy with tients are also at present off AEDs. The reason why prolonged occipital seizures or focal clonic, possibly there is such a wide spectrum of epilepsy severity in frontal, seizures was documented only in 1, and the this and in other EIEE remains a true mystery at pres- last, sixth patient’s, had a phenotype consistent with ent. Thus, further speculative discussion is better Lennox-Gastaut syndrome (figure 3B). postponed until we improve our understanding about Combining the analysis of the seizure types of the underlying pathophysiologic mechanism and how all known patients harboring KCNB1 mutations individual genetic backgrounds influence outcomes.

6 Neurology: Genetics Figure 3 Graphic representation of seizure type distribution in patients with KCNB1 mutations

(A) Previously published patients (data from 10 of the 13 patients of previously published KCNB1 mutations and epilepsy) and (B) in our series of 6 patients. AB 5 absences; CSWS 5 continuous spikes and waves during slow-wave sleep; FS 5 febrile seizures; MY 5 myoclonia; NCSE 5 nonconvulsive status epilepticus; TCS 5 tonic-clonic seizures.

Long-term follow-up of our 6 patients showed damaging by Saitsu et al.10 Patients carrying the that following delay of milestone acquisition and p.Arg306Cys mutation appear to have different phe- despite seizure freedom, they evolved into moderate notypes. The patient we are reporting in this study to severe impairment of cognitive functions includ- has generalized epilepsy with myoclonic seizures and ing psychiatric symptoms and autistic traits. Over- eyelid myoclonia, while the patient described by Sait- all, these findings indicate that Kv2.1 impaired su et al.10 had a more severe phenotype with earlier function causes diffuse brain dysfunction express- seizure onset and heterogeneous, drug-resistant, ing, in humans, as epilepsy and developmental or seizures including infantile spasms, myoclonic, behavioral disorders. Thus, it might be more appro- tonic-clonic, and focal seizures. The latter patient priate to define patients with KCNB1 mutations as manifested also severe ID and macrocephaly unlike having developmental encephalopathy with epi- the one we are reporting, who had a normal head size lepsy. An intermingled pathophysiologic relation- and moderate ID. ship between the triad of seizures/ASD/ID and While assessing genotype-phenotype correlations, potassium channel genes is further supported by we noted that both p.Trp370* and p.Arg583* non- the identification of mutations in KCNJ10,agene sense mutations occurred in the 2 patients with remit- coding for the inwardly rectifying potassium chan- ting infantile-onset spasms and severe cognitive nel Kir4.1, in several probands exhibiting seizures, impairment. At present, there is a single previously ASD, and ID.33 reported nonsense mutation14 in a patient with EE, All mutations reported here were de novo and yet detailed clinical features are not available for fur- included 4 missense and 2 nonsense variants. None ther correlations. of the mutations appeared in allele frequency data- The remaining 2 mutations, p.Ile196Phe and the bases (table e-1, http://links.lww.com/NXG/A0), p.Thr210Met, both predicted to be damaging, are and all were predicted to be damaging by several pre- missense variants (table 1) located in the S1 and in diction tools and ensemble scores (table e-1, http:// the S1–S2 extracellular loop. The S1–S4 segments links.lww.com/NXG/A0) and novel, with the excep- serve as the voltage-sensing module, and the voltage tion of the p.Arg306Cys falling in the S4 of the sensitivity of the channel is conferred by a series of voltage-sensor domain and demonstrated to be highly conserved basic side chains in the S4 segment.

Neurology: Genetics 7 These side chains are stabilized in a transmembrane AUTHOR CONTRIBUTIONS configuration by the formation of fully solvated salt Carla Marini, Michele Romoli, Elena Parrini, and Paolo Prontera: study bridges with a set of highly conserved acidic side design, data collection and analysis, and manuscript preparation. Frances- co Mari, Cinzia Costa, Lucio Parmeggiani, Tiziana Metitieri, and Mattia – 34 chains in S1 S3. The p.Ile196Phe and p. Gentile: data collection and analysis and manuscript revision. Davide Thr210Met mutations are located in the S1 con- Mei, Elena Cellini, Simona Virdò, and Dalila De Vita: genetic study, served acidic side chains and are therefore likely to data collection and analysis, and manuscript revision. Paolo Calabresi and Renzo Guerrini: study design, data analysis, and manuscript revision. impair the structural signature of voltage-sensing – 35 motion between S1 S3 and S4 segments. STUDY FUNDING NGS and Sanger sequencing showed that the p. This work was supported by a grant from the EU Seventh Framework Val349Phe variant was present as mosaicism although Programme FP7 under the project DESIRE (grant agreement no. associated with the most refractory epilepsy type in 602531) (to R.G.) and by the Italian Ministry of Health (grant agree- ment no. RF-2013-02355240) (to R.G.). our cohort. This might correlate with the location of this mutation in the pore-forming region of the DISCLOSURE channel, thus supposedly causing a severe alteration Carla Marini has received travel funding and/or speaker honoraria at VII of its function. A missense mutation located very CONGRESSO NAZIONALE SIMMESN–Firenze, December 16–18, closely, at position 347, causing severe impairment 2015, from SOBI, Swedish Orphan Biovitrum; has served on the editorial boards of Epilepsia and Epileptic Disorders; and has received research support of the channel function and associated with refractory from the Italian Ministry of Health and Cariplo Foundation. Michele epilepsy strikingly resembling our patient’s pheno- Romoli, Elena Parrini, and Cinzia Costa report no disclosures. Davide Mei type, had previously been reported.8 has received funding for travel or speaker honoraria from BioMarin Phar- maceutical and serves on the editorial board of Epilepsia. Francesco Mari, Patch-clamp studies have demonstrated that some Lucio Parmeggiani, Elena Procopio, Tiziana Metitieri, Elena Cellini, Si- KCNB1 variants cause significant alterations of Kv2.1 mona Virdò, Dalila De Vita, Mattia Gentile, Paolo Prontera, and Paolo channel function and increased excitability of cortical Calabresi report no disclosures. Renzo Guerrini has served on scientific neurons. Functional studies suggested different advisory boards of Eisai Inc., Novartis, Zogenix, and BioMarin; has received funding for travel and/or speaker honoraria from EISAI, Novartis, UCB, mechanisms including abolished voltage-activated and BioMarin; has served on the editorial boards of Epilepsia, Progress in 11 gating, loss of ion selectivity, gain of depolarizing Epileptic Disorders, Neuropediatrics, Journal of Child Neurology, Seizure, BMC inward cation,8 defects in expression and subcellular Medical Genetics, Topics in Epilepsy, Journal of Pediatric Epilepsy, Epileptic localization,8 disruption of sensitivity and cooperativ- Disorders, European Neurological Journal, Neurology,andJournal of Embry- ology & Developmental Biology; receives/has received royalties from the 10 ity of the sensor, and abolished endogenous Kv2.1 publication of Epilepsy and Movements disorders (Cambridge University 1 currents.10 The role of K channels in the neurobiol- Press, 2002), Aicardi’s Epilepsy in Children (Lippincott Williams & Wil- ogy of both epilepsy and cognitive impairment has kins, 2004), Progress in Epileptic Spasms and West Syndrome (John Libbey Eurotext, 2007), Epilepsy and Migraine (John Libbey Eurotext, been investigated also using animal models. Mice 2009), The Causes of Epilepsy (Cambridge University Press, 2011), Dravet lacking Kv2.1 are strikingly hyperactive, defective in syndrome (John Libbey Eurotext, 2011), and Epilepsy and epileptic seizures spatial learning, hypersensitive to convulsants, and (Oxford University Press, 2012); and has received research support from exhibit accelerated seizure progression.36 Moreover, Tuscany Region Research Department, EU and EC Programs, the Italian Ministry of Health, and the PISA Foundation. Go to Neurology.org/ng for 2/2 the Kcnb1 mice model showed neuronal hyperex- full disclosure forms. citability and confirmed the critical role of Kv2.1 in hippocampal neuronal network homeostatic Received June 17, 2017. Accepted in final form September 7, 2017. regulation.36 Thus, KCNB1-related manifestations comprise REFERENCES 1. Hwang PM, Fotuhi M, Bredt DS, Cunningham AM, a wide spectrum of infantile onset generalized or Snyder SH. Contrasting immunohistochemical localiza- focal seizures whose combination leads to EIEE, tions in rat brain of two novel K1 channels of the Shab including West, Lennox-Gastaut, and Jeavons syn- subfamily. J Neurosci 1993;13:1569–1576. dromes. Long-term follow-up highlights that follow- 2. Maletic-Savatic M, Lenn NJ, Trimmer JS. Differential ing a stormy phase, seizures cease and AEDs might spatiotemporal expression of K1 channel polypeptides in be withdrawn at least in some patients. Instead, cog- rat hippocampal neurons developing in situ and in vitro. – nitive and motor functions are almost always delayed J Neurosci 1995;15:3840 3851. 3. Du J, Tao-Cheng JH, Zerfas P, et al. The K1 channel, prior to seizure onset and evolve into severe, persis- Kv2.1, is apposed to astrocytic processes and is associated tent impairment. Altogether, these findings suggest with inhibitory postsynaptic membranes in hippocampal that KCNB1 mutations are associated with diffuse and cortical principal neurons and inhibitory . brain dysfunction and that the KCNB1-related clin- Neuroscience 1998;84:37–48. ical picture might also be defined as a developmental 4. Liu PW, Bean BP. Kv2 channel regulation of action “encephalopathy” with seizures. Further under- potential repolarization and firing patterns in superior cer- vical ganglion neurons and hippocampal CA1 pyramidal standing of the underlying pathophysiology by neurons. J Neurosci 2014;34:4991–5002. experimental studies including animal models is 5. Misonou H, Mohapatra DP, Park EW, et al. Regulation of required to improve treatment and prognosis of localization and phosphorylation by neuronal these disorders. activity. Nat Neurosci 2004;7:711–718.

8 Neurology: Genetics 6. Misonou H, Thompson SM, Cai X. Dynamic regulation 21. Caraballo RH, Fontana E, Darra F, et al. A study of 63 of the Kv2.1 voltage-gated potassium channel during brain cases with eyelid myoclonia with or without absences: type ischemia through neuroglial interaction. J Neurosci 2008; of seizure or an epileptic syndrome? Seizure 2009;18:440– 28:8529–8538. 445. 7. Fox PD, Loftus RJ, Tamkun MM. Regulation of Kv2.1 K 22. Biervert C, Schroeder BC, Kubisch C, et al. A potassium (1) conductance by cell surface channel density. channel mutation in neonatal human epilepsy. Science J Neurosci 2013;3:1259–1270. 1998;279:403–406. 8. Torkamani A, Bersell K, Jorge BS, et al. De novo KCNB1 23. Charlier C, Singh NA, Ryan SG, et al. A pore mutation in mutations in epileptic encephalopathy. Ann Neurol 2014; a novel KQT-like potassium channel gene in an idiopathic 76:529–540. epilepsy family. Nat Genet 1998;18:53–55. 9. Srivastava S, Cohen JS, Vernon H, et al. Clinical whole 24. Singh NA, Charlier C, Stauffer D, et al. A novel potassium exome sequencing in child neurology practice. Ann Neurol channel gene KCNQ2 is mutated in an inherited epilepsy 2014;76:473–483. of newborns. Nat Genet 1998;18:25–29. 10. Saitsu H, Akita T, Tohyama J, et al. De novo KCNB1 25. Benarroch EE. Potassium channels: brief overview and mutations in infantile epilepsy inhibit repetitive neuronal implications in epilepsy. Neurology 2009;72:664–669. firing. Sci Rep 2015;5:15199. 26. Muona M, Berkovic SF, Dibbens LM, et al. A recurrent 11. Thiffault I, Speca DJ, Austin DC, et al. A novel epileptic de novo mutation in KCNC1 causes progressive myoclo- encephalopathy mutation in KCNB1 disrupts Kv2.1 ion nus epilepsy. Nat Genet 2015;47:39–46. selectivity, expression, and localization. J Gen Physiol 27. Simons C, Rash LD, Crawford J, et al. Mutations in the 2015;146:399–410. voltage-gated potassium channel gene KCNH1 cause 12. Allen NM, Conroy J, Shahwan A, et al. Unexplained early Temple-Baraitser syndrome and epilepsy. Nat Genet onset epileptic encephalopathy: exome screening and phe- 2015;47:73–77. notype expansion. Epilepsia 2016;57:e12–e17. 28. Du W, Bautista JF, Yang H, et al. Calcium-sensitive potas- 13. Latypova X, Matsumoto N, Vinceslas-Muller C, et al. sium in human epilepsy and paroxysmal Novel KCNB1 mutation associated with non-syndromic movement disorder. Nat Genet 2005;37:733–778. intellectual disability. J Hum Genet 2017;62:569–573. 29. Nava C, Dalle C, Rastetter A, et al. De novo mutations in 14. de Kovel CGF, Brilstra EH, van Kempen MJA, et al. HCN1 cause early infantile epileptic encephalopathy. Nat Targeted sequencing of 351 candidate genes for epileptic Genet 2014;46:640–645. encephalopathy in a large cohort of patients. Mol Genet 30. Barcia G, Fleming MR, Deligniere A, et al. De novo gain- Genomic Med 2016;4:568–580. of-function KCNT1 channel mutations cause malignant 15. Miao P, Peng J, Chen C, et al. A novel mutation in KCNB1 migrating partial seizures of infancy. Nat Genet 2012;44: gene in a child with neuropsychiatric comorbidities with both 1255–1259. intellectual disability and epilepsy and review of literature. 31. Syrbe S, Hedrich UB, Riesch E, et al. De novo loss- or Zhonghua Er Ke Za Zhi 2017;55:115–119. gain-of-function mutations in KCNA2 cause epileptic 16. Commission on Classification and Terminology of the encephalopathy. Nat Genet 2015;47:393–399. International League Against Epilepsy. Proposal for revised 32. Eunson LH, Rea R, Zuberi SM, et al. Clinical, genetic, classification of epilepsies and epileptic syndromes. Epilep- and expression studies of mutations in the potassium chan- sia 1989;30:389–399. nel gene KCNA1 reveal new phenotypic variability. Ann 17. Berg AT, Berkovic SF, Brodie MJ, et al. Revised terminol- Neurol 2000;48:647–656. ogy and concepts for organization of seizures and epilep- 33. Sicca F, Imbrici P, D’Adamo MC, et al. Autism with sies: report of the ILAE Commission on Classification and seizures and intellectual disability: possible causative role Terminology, 2005–2009. Epilepsia 2010;51:676–685. of gain-of-function of the inwardly-rectifying K1 channel 18. Parrini E, Marini C, Mei D, et al. Diagnostic targeted Kir4.1. Neurobiol Dis 2011;43:239–247. resequencing in 349 patients with drug-resistant pediatric 34. Wood ML, Freites JA, Tombola F, Tobias DJ. Atomistic epilepsies identifies causative mutations in 30 different modeling of ion conduction through the voltage-sensing genes. Hum Mutat 2017;38:216–225. domain of the K1 ion channel. J Phys Chem B 19. Yang H, Wang K. Genomic variant annotation and prior- 2017;121:3804–3812. itization with ANNOVAR and wANNOVAR. Nat Protoc 35. Freites JA, Tobias DJ. Voltage sensing in membranes: 2015;10:1556–1566. from macroscopic currents to molecular motions. 20. Richards S, Aziz N, Bale S, et al. Standards and guidelines J Membr Biol 2015;248:419–430. for the interpretation of sequence variants: a joint consen- 36. Speca DJ, Ogata G, Mandikian D, et al. Deletion of the sus recommendation of the American College of Medical Kv2.1 delayed rectifier potassium channel leads to neuro- Genetics and Genomics and the Association for Molecular nal and behavioral hyperexcitability. Genes Brain Behav Pathology. Genet Med 2015;17:405–424. 2014;13:394–408.

Neurology: Genetics 9 CDKL5 variants Improving our understanding of a rare neurologic disorder

Ralph D. Hector, PhD ABSTRACT Vera M. Kalscheuer, PhD Objective: To provide new insights into the interpretation of genetic variants in a rare neurologic Friederike Hennig, MSc disorder, CDKL5 deficiency, in the contexts of population sequencing data and an updated Helen Leonard, MBChB characterization of the CDKL5 gene. Jenny Downs, PhD Methods: We analyzed all known potentially pathogenic CDKL5 variants by combining data from Angus Clarke, DM large-scale population sequencing studies with CDKL5 variants from new and all available clinical Tim A. Benke, MD, PhD cohorts and combined this with computational methods to predict pathogenicity. Judith Armstrong, PhD Mercedes Pineda, Results: The study has identified several variants that can be reclassified as benign or likely MD, PhD benign. With the addition of novel CDKL5 variants, we confirm that pathogenic missense variants Mark E.S. Bailey, PhD cluster in the catalytic domain of CDKL5 and reclassify a purported missense variant as having Stuart R. Cobb, PhD a splicing consequence. We provide further evidence that missense variants in the final 3 exons are likely to be benign and not important to disease pathology. We also describe benign splicing and nonsense variants within these exons, suggesting that isoform hCDKL5_5 is likely to have Correspondence to little or no neurologic significance. We also use the available data to make a preliminary estimate Dr. Hector: of minimum incidence of CDKL5 deficiency. [email protected] Conclusions: These findings have implications for genetic diagnosis, providing evidence for the reclassification of specific variants previously thought to result in CDKL5 deficiency. Together, these analyses support the view that the predominant brain isoform in humans (hCDKL5_1) is crucial for normal neurodevelopment and that the catalytic domain is the primary functional domain. Neurol Genet 2017;3:e200; doi: 10.1212/NXG.0000000000000200

GLOSSARY dbSNP 5 Single Nucleotide Polymorphism database; ExAC 5 Exome Aggregation Consortium; NMD 5 nonsense-mediated decay; RettBASE 5 Database; RT 5 reverse transcription; UTR 5 untranslated region; VEP 5 Variant Effect Predictor; XCI 5 X-chromosome inactivation; 1000G 5 1000 Genomes Project.

The phenotype associated with CDKL5 deficiency (MIM: 300203) has become increasingly well defined over the last decade. Its cardinal features are early-onset seizures, often presenting as infan- tile spasms and usually occurring within the first 3 months of life, global developmental delay, and severely impaired gross motor function.1 CDKL5deficiencyiscausedbydominantlyactingloss-of- function variants in the X-linked gene CDKL5 (cyclin-dependent -like 5), which plays a crucial role in brain development.2–6 The epidemiology of CDKL5 deficiency has not been studied, and no incidence or prevalence data are available. Nevertheless, the frequency of patients

From the Institute of Neuroscience & Psychology (R.D.H., S.R.C.), University of Glasgow, UK, Drs. Hector and Cobb are currently with the Patrick Wild Centre and Centre for Discovery Brain Science, University of Edinburgh, UK; Group Development and Disease (V.M.K., F.H.), Max Planck Institute for Molecular Genetics, Berlin, Germany; Telethon Kids Institute (H.L., J.D.), The University of Western Australia, Perth, Western Australia; School of Physiotherapy and Exercise Science (J.D.), Curtin University, Perth, Australia; Institute of Medical Genetics (A.C.), School of Medicine, Cardiff University, Cardiff, Wales, UK; Departments of Pediatrics, Pharmacology, Neurology and Otolaryngology (T.A.B.), University of Colorado School of Medicine, Aurora, CO; Paedriatic Neuroscience (J.A., M.P.), Institut de Recerca Sant Joan de Déu, Esplugues de Llobregat, Spain; Hospital Sant Joan de Déu Barcelona (J.A.), Esplugues de Llobregat, Spain; CIBERER (J.A.), Barcelona, Spain; Neuropediatrics (M.P.), Fundació Sant Joan de Déu, Esplugues de Llobregat, Spain; and School of Life Sciences (M.E.S.B.), College of Medical, Veterinary and Life Sciences, University of Glasgow, UK. Funding information and disclosures are provided at the end of the article. Go to Neurology.org/ng for full disclosure forms. The Article Processing Charge was funded by CDKL5 UK. 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.

Neurology.org/ng Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 diagnosed with CDKL5 deficiency is increasing Computational analysis. The effects of missense and splicing due to growing awareness of the disorder and variants in CDKL5 were predicted using several algorithms (missense: SIFT, PolyPhen, MutationTaster, and PROVEAN; the inclusion of CDKL5 in routine genetic test- splicing variants: MaxEntScan and dbscSNV), provided within 7,8 ing of early-onset epileptic encephalopathies. the Ensembl Variant Effect Predictor (VEP).15 For rare Mendelian diseases caused by de Assignment criteria. The assignment of pathogenicity was novo variants, such as CDKL5 deficiency, data based on guidelines for the interpretation of sequence variants.16 from large-scale screening of patient and pop- The following criteria were used: pathogenic—the same amino ulation samples can be mined to enhance clar- acid change as an established pathogenic variant; likely patho- genic—the allele is absent in population data sets and the pa- ification of clinically relevant variants. The tient’s phenotype is highly specific for CDKL5 disorder, and 9 Exome Aggregation Consortium (ExAC) ana- computational evidence supports a deleterious effect on the gene; lyzed high-quality exome DNA sequence data uncertain significance—evidence for benign and pathogenic clas- — of 60,706 individuals of diverse ancestries, sification is contradictory; likely benign the variant is detected in a healthy mother and/or sister, and computational evidence providing opportunities to refine the clinical suggests no effect on the gene; and benign—the variant is de- interpretation of CDKL5 variants. Here, we tected in a healthy father and/or brother, or the allele frequency is analyze all known variants observed in detected in population data sets at levels too high to explain the CDKL5 patients to date, including novel var- prevalence of a rare disorder. iants described in this study, along with data Minigene splicing assay. DNA fragments of wild-type and from ExAC, the 1000 Genomes Project mutant CDKL5 exon 14 with flanking intron sequence were synthesized (IDT, Coralville, IA) and cloned into the exon trap 10 (1000G), and the Single Nucleotide Poly- vector pET01 (MoBiTec, Göttingen, Germany). Minigenes were morphism database (dbSNP).11 All variants transfected into HEK293T cells using jetPRIME (PEQLAB, are analyzed in the context of the updated Erlangen, Germany) according to the manufacturer’s in- characterization of the CDKL5 gene12 to pro- structions. After 24 hours, total RNA was extracted using the RNeasy Plus Mini Kit (Qiagen, Hilden, Germany). Total RNA vide new insights into the clinical interpreta- was generated, and reverse transcription PCR (RT-PCR) ex- tion of variants in CDKL5 deficiency. periments were performed as described previously.17 Further details of these experiments are described in e-Methods.

METHODS Standard protocol approvals, registrations, RESULTS CDKL5 variants in the population. The and patient consents. Written informed consent was obtained number of expected and observed variants in CDKL5 from all individuals who participated in this study, and the study from the recent large-scale analysis of genetic varia- was approved by the ethics committees of the respective institu- 9 tions: University of Western Australia Human Research Ethics tion by ExAC is described in table 1. Synonymous Committee (reference # RA/4/1/5024); the Institute of Medical variants in CDKL5 are reported to occur at approxi- Genetics, University Hospital of Wales (Cardiff, UK) as part of mately the expected frequency and are found the British Isles Rett Syndrome Survey (REC reference # 15/WA/ throughout the coding region of CDKL5 (figure 1). 029); Children’s Hospital Colorado (Aurora, CO) (COMIRB By contrast, the number of missense variants 13-2020); established protocols to access clinical data of Sant observed (n 5 157) is lower than expected (table 1). Joan de Déu Children’s Hospital (Barcelona, Spain). The population sample analyzed in the ExAC study CDKL5 variant data collection. CDKL5 variants in clinical was devoid of severe pediatric disease; therefore, none cohorts (see above) were analyzed by combining data from of these 157 missense variants is, on its own, patho- the CDKL5 variation database at Rett Syndrome Database 13 genic and causative for CDKL5 deficiency. Only 15 (RettBASE), Database of Chromosomal Imbalance and Pheno- ; type in Humans Using Ensembl Resources,14 and all published ( 10%) of these missense variants are located in the reports of CDKL5 variants. Population data on CDKL5 variants catalytic domain of CDKL5, a 286-amino acid region were sourced from ExAC,9 1000G,10 and dbSNP.11 (;30% of protein length) crucial for protein func- tion2 (figures 1 and 2). No frameshift or splicing

Table 1 Expected and observed CDKL5 variants in ExAC variants were identified in ExAC; however, 2 non- sense variants (described as stop-gain in table 1) were Variant type Expected no. of variants Observed no. of variants Constraint metric identified: p.Arg952Ter and p.Arg970Ter. Both

Synonymous 99.4 111 Z 520.72 nonsense variants are found in exon 21 and therefore

Missense 237.7 157 Z 5 2.56 affect only the hCDKL5_5 isoform, which is ex- pressed almost exclusively in the testis.12,18 Stop-gain 26.4 2 pLI 5 1.00 Some known exonic coding and untranslated re- Abbreviation: ExAC 5 Exome Aggregation Consortium. gions (UTRs) of CDKL5, recently identified and The expected number of variants was calculated using a gene-size and sequencing depth- present only in minor transcript isoforms,12 are not adjusted variant probability. The positive Z score for missense variants in CDKL5 indicates increased intolerance to variation. The calculated pLI value of 1 indicates that CDKL5 is included in the exonic regions covered by ExAC. For extremely intolerant to loss-of-function variants (here referred to as “stop-gain”). example, exon 17 (originally termed 16b19) is present

2 Neurology: Genetics Figure 1 Distribution of exonic CDKL5 variants

A cartoon of the gene structure is given at the top, with exons of hCDKL5_1, the dominant brain transcript isoform, colored blue-green (coding regions) and black (UTRs). Introns are not drawn to scale. In the diagram beneath, variant types are grouped together, and individual variants are plotted according to their location in the gene. Red indicates pathogenic or likely pathogenic variants; green indicates benign or likely benign variants; and amber indicates variants of uncertain significance.

in the hCDKL5_2 isoform and is expressed only at isoform, hCDKL5_1 (figure 1). Only 1 of the 15 mis- low levels in the developing and adult brain.12 Only 2 sense variants identified within the catalytic domain in exon 17 variants have been identified, both in dbSNP ExAC has been reported in a patient with CDKL5 (rs289269000 and rs181987256), neither of which is deficiency: c.719G.C (p.Ser240Thr; figure 2). This associated with a disease phenotype. exon 9 variant has recently been reported as the first familial case of CDKL5-related disease20 in a het- CDKL5 missense variants in patients. Missense variants erozygous female, who displayed global psychomotor reported in patients are found throughout the portions delay and autistic disturbances but no epilepsy. The of CDKL5 that encode the major brain transcript same variant was detected in her asymptomatic

Figure 2 Pathogenic CDKL5 missense variants cluster in the catalytic domain

Functional domains in the CDKL5 protein are color coded. Numbers refer to the positions of amino acids. Variants in red (upper) are pathogenic or likely pathogenic. Variants in black (lower) are benign or likely benign. aVariant with a splicing consequence. bVariant of uncertain significance. NES 5 putative nuclear export signal; NLS 5 putative nuclear localization signal; ST 5 serine-threonine kinase ; TEY 5 conserved Thr-Glu-Tyr motif.

Neurology: Genetics 3 mother, and a skewed X-chromosome inactivation inclusion criteria in those studies. It is unlikely that 3 (XCI) ratio was thought to account for this phenotypic asymptomatic carriers of this variant would all have difference. We further identified 2 individuals with similarly skewed XCI, suggesting that this variant may this variant (1 in ExAC and 1 in 1000G), neither not in fact be causative. Current evidence suggests that individual having displayed signs of a neuro- the p.Ser240Thr should be classified as being of developmental disorder phenotype, consistent with the “uncertain significance.”

Table 2 Novel CDKL5 variants in our clinical cohorts

Mutation 1000G Protein CDKL5 clinical Computational type Exon Gene variant GRCh38 ExAC AF MAF dbSNP consequence Sex diagnosis prediction

Missense 2 c.37T.C 18507133 ——— p.Phe13Ser Female Yes Pathogenic

Missense 3 c.71A.G 18510826 ——— p.Tyr24Cys Female Yes Pathogenic

Missense 4 c.104C.T 18564481 ——— p.Thr35Ile Female Yes Pathogenic

Missense 5 c.200T.C 18575408 ——— p.Leu67Pro Female Yes Pathogenic

Missense 6 c.364G.A 18579929 ——— p.Ala122Thr Female Yes Uncertain significance

Missense 6 c.377G.A 18579942 ——— p.Cys126Tyr Female Yes Uncertain significance

Missense 7 c.454T.C 18581941 ——— p.Cys152Arg Female Yes Pathogenic

Missense 9 c.605G.A 18588004 ——— p.Gly202Glu Female Yes Pathogenic

Missense 10 c.784T.C 18595387 ——— p.Tyr262His Female Yes Uncertain significance

Missense 11 c.853A.G 18598489 ——— p.Arg285Gly Female Yes Pathogenic

Missense 11 c.950A.Ga 18598586 2.28E-05 — rs756537286 p.His317Arg Male — Uncertain significance

Missense 12 c.1188T.Aa 18604112 1.14E-05 — rs772076629 p.Asp396Glu Female — Uncertain significance

Missense 15 c.2243A.Ga 18613242 6.84E-05 0.0003 rs748459878 p.Asn748Ser Female — Benign

Missense 20 c.2716G.A 18646009 2.28E-05 — rs369009993 p.Gly906Ser Female — Benign

Missense 21 c.2980G.Aa 18650592 ——— p.Gly994Arg Female — Benign

Nonsense 5 c.205C.T 18575413 ——— p.Gln69Ter Female Yes —

Nonsense 5 c.258C.G 18575466 ——— p.Tyr86Ter Female Yes —

Nonsense 11 c.858C.A 18598494 ——— p.Tyr286Ter Female Yes —

Nonsense 14 c.2112C.G 18609530 ——— p.Tyr704Ter Female Yes —

Nonsense 15 c.2276G.A 18613275 ——— p.Trp759Ter Female Yes —

Nonsense 21 c.2941C.T 18650553 ——— p.Arg981Ter Female Yesb —

Splicing IVS2 c.65-2A.T 18510818 ——— — Female Yes Pathogenic

Splicing IVS2 c.65-2A.G 18510818 ——— — Female Yes Pathogenic

Splicing IVS6 c.404-1G.C 18581890 ——— — Female Yes Pathogenic

Splicing IVS10 c.82511G.A 18595429 ——— — Female Yes Pathogenic

Splicing IVS10 c.82511G.A 18595429 ——— — Female Yes Pathogenic

Splicing IVS10 c.826-1G.A 18598461 ——— — Female Yes Pathogenic

Splicing IVS15 c.227611G.T 18613276 ——— — Female Yes Pathogenic

Splicing IVS20 c.279712T.C 18664209 ——— — Male — Pathogenic

Triplication 1–15 X:18370935- ————— Female —— 18634410 x3

Duplication 3–12 X:18510568- ————— Female —— 18605032 x2

Duplication IVS1 X:18488669- ————— Female —— 18516147 x2

Abbreviations: dbSNP 5 Single Nucleotide Polymorphism database; ExAC 5 Exome Aggregation Consortium; VEP 5 Variant Effect Predictor; 1000G 5 1000 Genomes Project. Variants are grouped into types. GRCh38 indicates the X-chromosome coordinate in the GRCh38/hg38 assembly. a Variant inherited from an asymptomatic mother. Frequencies of variants in population databases are shown as ExAC allele frequency (AF) and 1000G minor allele frequency (MAF). CDKL5 clinical diagnosis was made by the clinical center to which the patient presented (bmilder-than-average CDKL5 symptoms). Computational prediction is based on analysis using the Ensembl VEP.

4 Neurology: Genetics Combining all known CDKL5 variant data, there hemizygotes. In some cases, such as p.Ile508Thr are59missensevariantsinthecatalyticdomain,which and p.Thr734Ala, the ExAC allele frequency and we consider pathogenic or likely pathogenic for CDKL5 the presence of these variants in hemizygotes allow deficiency based on our assignment criteria (figure 2). them to be reclassified as likely benign. Additional This includes 11 novel missense variants identified in variants present in dbSNP (but not in ExAC) are the clinical cohorts in this study (table 2). In each case, not associated with any clinical phenotype. In our the patient received a clinical diagnosis of CDKL5 defi- clinical cohorts, we identified further missense var- ciency, and, in most cases, VEP analysis supported the iants in patients, where a clinical diagnosis of CDKL5 prediction of a pathogenic variant. deficiency was not given (table 2). p.His317Arg, Outside the catalytic domain, there are 179 differ- p.Asp396Glu, p.Asn748Ser, and p.Asp797Asn are ent missense variants (figure 1). Upstream of the cat- present in ExAC/1000G, and all variants were iden- alytic domain, 3 variants (p.Ile3Phe, p.Asn5Asp, and tified as being inherited from a healthy mother or p.Ile6Thr) are known to be present in ExAC father (table 2), suggesting that they are all benign and 1000G, not associated with CDKL5 patients variants. In our analysis, there are 7 missense variants and predicted to be benign by VEP analysis. Fifty- outside the catalytic domain that should, at present, one different missense variants have been found in the be classified as “of uncertain significance” pending last 3 exons (20, 21, and 22). Forty of these are more information (figure 2). p.Val351Ile is recorded present in ExAC, often in hemizygous males. in dbSNP (rs587783150) and p.Thr538Ala is re- c.2995G.A (p.Val999Met), previously thought to corded in the RettBASE,13 for which no clinical or be pathogenic, is found at notable frequency in ExAC screening information is available. p.Leu302Phe and (0.01178) and should therefore be classed as a benign p.Asp422Glu are unique variants that have been asso- polymorphism. A further 10 are present in dbSNP; ciated with seizure phenotypes, although the pheno- none of these variants appear to be responsible for type of these cases is atypical in comparison to other a CDKL5-like phenotype, and VEP analysis predicts CDKL5 patients reported in these studies.21,22 p. these variants to be benign. This is consistent with Asn399Thr and p.Val793Ala have also been associ- these exons being absent in the primary brain tran- ated with epilepsy phenotypes4,23; these variants have script isoform. In our clinical cohorts, a variant in also been observed as singletons in ExAC. exon 21, c.2980G.A (p.Gly994Arg), was identified The most compelling evidence of a pathogenic in a patient with developmental delay, but was not missense variant outside the catalytic domain con- clinically diagnosed with CDKL5 deficiency (table 2). cerns the c.2152G.A (p.Val718Met) variant. This The variant was inherited from her asymptomatic variant has been identified in 3 unrelated CDKL5 mother and, combined with VEP analysis, should patients in 3 independent studies.24–26 In each case, therefore be classified as benign. the patient presented with symptoms consistent with The remaining missense variants are distributed CDKL5 deficiency and harbored no known variants between exons 11 and 19 (figure 1). Ninety-seven in other epilepsy-related genes. This variant is not of these are found in ExAC, many occurring in present in the 1000G or the ExAC population

Figure 3 CDKL5 variant c.2152G>A causes skipping of exon 14 in HEK293T cells

(A) Schematic representation of the minigene constructs used in the in vitro splicing assay (not to scale). The pET01 vector contains 59 and 39 exons separated by an intron sequence. Minigenes contain wild-type (wt) or mutant (mut; c.2152G.A) CDKL5 exon 14 sequences flanked by portions of their natural introns (thick lines). Primers used for RT-PCR experiments are indicated by arrows. Splicing events, indicated by dashed lines, would result in a 348-bp or a 242-bp product depending on inclusion or otherwise of CDKL5 exon 14. (B) Agarose gel showing RT-PCR results from the splicing assay using pET01 vector and minigenes of wild-type or mutated CDKL5 exon 14. Upper bands (348 bp) indicate the presence of exon 14, whereas lower bands (242 bp) indicate exclusion of exon 14. Marker indicates 100 bp ladder; 2RT indicates a negative control without reverse transcriptase.

Neurology: Genetics 5 databases. VEP analysis of c.2152G.A, which affects associated pathology were reported. p.Arg952Ter was the last base of exon 14, predicted a detrimental effect identified in the patient’s mother, grandmother, on splicing (MaxEntScan diff: 2.89). We assessed this and half-sister, and in 6 individuals in a control pop- variant in vitro using a minigene splicing assay, show- ulation.28 A further 12 instances of this allele are ing that splicing is disrupted, resulting in exon 14 present in ExAC, including 3 hemizygotes (allele fre- being omitted from the transcript (figure 3). We quency 0.0001367). A recent case study also identi- therefore suggest that the c.2152G.A variant be cat- fied this variant in an asymptomatic hemizygous egorized as a splicing rather than a missense variant. male.30 p.Arg970Ter occurs in a single heterozygous The consequence (r.2047_2152del) is a frameshift female in ExAC (in the case study, the variant was not variant that would generate a premature stop codon detected in the mother and the father was deceased29). in exon 16. This finding removes evidence supporting Together, these findings suggest that variants in these the existence of pathogenic missense variants outside late exons do not cause a CDKL5 deficiency the catalytic domain of CDKL5, and we therefore phenotype. conclude that caution should be applied in interpre- A c.2941C.T, p.Arg981Ter variant was identi- tations of pathogenicity whenever an apparent mis- fied in our clinical cohorts (table 2). A girl presented sense variant is found outside the catalytic domain. with a phenotype consistent with CDKL5 deficiency —profound cognitive impairment, mild dysmorphic Nonsense, frameshift, and splicing variants. Pathogenic appearance, stereotypical hand wringing, and epilepsy nonsense, frameshift, and splicing variants are found (from age 6 months). Both parents were deceased; throughout the coding region of hCDKL5_1, the pre- therefore, it is unknown whether this variant was in- dominant brain transcript isoform (figure 1). Efforts herited or occurred de novo. This variant is not pres- have been made to establish genotype-phenotype re- ent in ExAC or in 1000G, and the clinical evidence lationships based on the position of these variants in would suggest defining this as a pathogenic variant. the gene, by predicting structural and functional However, this variant lies immediately downstream of consequences.8 However, the levels of functional the 2 nonsense variants described above, both of CDKL5 in patients with truncating variants at dif- which have considerable evidence suggesting that ferent points throughout the gene are still not known. they should be classified as benign. We conclude that, Novel nonsense and splicing variants identified in the at present, p.Arg981Ter should be classified as being clinical cohorts in this study are described in table 2, of “uncertain significance.” and novel frameshift variants in table e-1, http://links. It is important that no pathogenic frameshift or lww.com/NXG/A2. splicing variants were found affecting exons 20, 21, In exon 19, the last exon utilized by the and 22 (figure 1). A novel splicing variant was hCDKL5_1 transcript, a variant described in dbSNP identified in our clinical cohorts (table 2): (rs863225289) has been identified in a patient with c.279712T.C. The patient, a boy, was not clini- early infantile epileptic encephalopathy 2 (MIM: cally diagnosed with CDKL5 deficiency and was 300672). This variant lies downstream of the internal found to have inherited the variant from his asymp- splice donor site in exon 19 and so would not be ex- tomatic mother. VEP analysis of this variant predicts pected to affect the protein product made from the the abolition of the splice donor site of exon 20, testis-specific hCDKL5_5 transcript isoform. However, which would be likely to result in a premature stop in relation to hCDKL5_1, the major brain transcript codon. isoform, this is a nonsense variant (c.2176C.T, p.Gln906Ter, with respect to hCDKL5_1). This may Copy number variations and CDKL5 duplication. Sev- trigger nonsense-mediated decay (NMD; due to the eral studies have reported duplications of Xp22 asso- presence of a long 39-UTR27) and result in typical ciated with intellectual disability and autism CDKL5 deficiency. Additional observations of this var- phenotypes,31–34 but the duplicated regions reported iant would be required to validate this conjecture, but it (spanning 8–21 Mb) included as many as 80 genes, is consistent with the idea that truncating variants in and interpretation of gene-specific overexpression exon 19 may be pathogenic. effects in such circumstances is problematic. A more Three nonsense variants have been detected in recent study described 3 unrelated families with more exon 21, which is specific to isoform hCDKL5_5: compact duplication regions incorporating CDKL5.35 c.2854C.T, p.Arg952Ter; c.2908C.T, p. Four different duplicated regions were described, Arg970Ter; and c.2941C.T, p.Arg981Ter. Both ranging from 540 to 935 kb in size. In the case of the p.Arg952Ter and p.Arg970Ter variants have 683-kb duplication harbored by three of the patients, been described in patients with CDKL5-like symp- the duplicated region does not include exon 1 of toms.28,29 However, in both cases, delay in seizure CDKL5, so the predominant adult brain isoform, onset and an absence of other features of CDKL5- hCDKL5_1, would not be expected to be

6 Neurology: Genetics overexpressed in these individuals. The authors point DISCUSSION Large-scale sequencing studies are out that the alternative hCDKL5_5 transcript may be powerful tools for the analysis of genetic variants in expressed and overexpression effects during fetal rare diseases, such as CDKL5 deficiency. Analysis of development may result in the observed phenotype. CDKL5 variants in this study reinforces the work of Three patients also harbored additional duplicated recent studies that propose hCDKL5_1 as being the regions on other chromosomes, and parents harbor- predominant functional isoform required for normal ing these same CDKL5-containing duplications were neurodevelopment and brain function. We have either phenotypically unaffected or displayed mild shown that missense variants outside the catalytic intellectual disability.35 Therefore, the pathway lead- domain are unlikely to be pathogenic. We have also ing to the phenotypes in these individuals remains highlighted specific cases that should currently be somewhat unclear, as experimental confirmation of classified as being of “uncertain significance” and have effects on CDKL5 transcript and protein levels is not functionally reclassified a pathogenic missense variant yet available. Here, we report 2 duplications and 1 as a splicing variant. Furthermore, we saw no evi- triplication in our clinical cohorts (table 2). However, dence that missense variants outside the catalytic in each of these copy number variations, only a part of domain were overrepresented in patient populations the gene is duplicated; therefore, it is extremely compared with their allele frequencies in population unlikely that an increased level of CDKL5 would be databases. However, we cannot rule out the possibil- present in these individuals. ity that this lack of association is the result of some missense variants having reduced penetrance and thus Variants in regulatory and UTRs. The recent identifica- existing in unaffected individuals in population co- tion of novel exons and transcription start sites at the horts. An analysis of male and female variants sepa- 59 end of the gene suggests the presence of multiple rately (in a compact subset of 15 missense variants in promoter regions.12 However, most brain transcripts the catalytic domain) was suggestive of no significant are driven by a putative promoter upstream of exon 1. sex differences in allele frequency (not shown). How- Several variants in the 59-UTR have been reported, ever, further analysis of all missense variants would be but so far only those associated with the deletion of recommended in a future study. exon 1 or 2, or the disruption of splicing from exon 1 There is a lack of evidence for pathogenic variants to exon 2 (c.-162-2A.G36), have been shown to be in exons 20, 21, and 22. Early studies of the CDKL5 pathogenic. Only 2 variants in the putative promoter gene had deemed hCDKL5_5 (formerly known as region, c.-440G.T and c.-189C.T, have so far hCDKL5 ) to be the predominant CDKL5 tran- been found in patients with a CDKL5-like pheno- 115 script and . However, we now know type,3 but at present their significance remains that this isoform utilizing exons 20–22 is expressed uncertain. The large 6.6 kb 39-UTR of hCDKL5_1 almost exclusively in the testis and found only at has only recently been defined,12 so the region has not extremely low levels in the adult brain.2,12,18 A recent been analyzed in patients either by targeted analysis of missense variants in these exons concluded sequencing or by exome analysis, and we found no that genetic variation in this C-terminus was likely to evidence for the presence of clinically relevant variants have little or no significance to a CDKL5 disorder in the 1000G data. phenotype.37 We find further evidence of this in our Incidence of CDKL5 deficiency. The incidence of study. Furthermore, the identification of nonsense CDKL5 deficiency is unknown. We have analyzed variants in these exons in the population is an impor- data from the International CDKL5 Disorder Data- tant observation, consistent with a previous study in base8 to provide a lower estimate of birth prevalence which patients with a 136-kb deletion lost only these in Australia of 0.21 cases per 100,000 live births 39 CDKL5 exons and the overlapping RS1 gene.38 All (95% CI 0.12–0.33) for the years 1982–2014. patients in the study showed only a retinoschisis Although a birth prevalence in this range would indi- phenotype, consistent with RS1 deficiency. Together, cate that CDKL5 deficiency is an ultra rare disorder, the data reinforces the view that variants identified at it is likely that this figure will increase, as targeted the 39 end of the CDKL5 gene should be interpreted next-generation sequencing for investigation of with caution. early-onset epileptic encephalopathy becomes more Although there is substantial evidence that common.7,25 hCDKL5_1 is the predominant brain isoform, In our analysis, the most frequent CDKL5 hCDKL5_5 continues to be cited as a reference deficiency–causing variants are 2 nonsense variants sequence. Consequently, exon 17 and full-length in exon 12: c.1648C.T (p.Arg550Ter) and exon 19 are not always routinely sequenced in tar- c.1675C.T (p.Arg559Ter). Each of these variants geted gene panels containing CDKL5 and are often affects only 3% of all patients, highlighting the high excluded from exome analysis (as in ExAC). This has degree of allelic heterogeneity in CDKL5 deficiency. potentially important consequences for molecular

Neurology: Genetics 7 diagnosis, and it is possible that pathogenic variants in Disease Center on behalf of the Loulou Foundation #CDKL5-17-103- these exon regions are therefore underreported. It is 01. F. Hennig reports no disclosures. H. Leonard has served on the editorial board of Children; received research support from an NHMRC possible that the presence of isoform hCDKL5_2 Senior Research Fellowship #1117105; has received funding for the (which contains exon 17) is not crucial for normal maintenance of the International Rett syndrome Database (InterRett) CDKL5 function, given that the levels of this isoform from Rettsyndrome.org and the International CDKL5 Disorder Database from the International Foundation for CDKL5 Research, NHMRC Pro- are only 10% (or less) of hCDKL5_1, at all stages of gram grant 572742, and an NHMRC Senior Research Fellowship 12 development. It may be critical to analyze full- #572568; has received an honorarium and travel funding to attend the length exon 19, as nonsense and frameshift variants International Rett Syndrome Symposium; and has received travel funding even at the 39 end of the gene are likely to be path- from the Italian, Russian, and Australia Rett syndrome parent associa- 9 tions, Newron Pharmaceuticals, and the LouLou Foundation. J. Downs ogenic. The presence of long 3 -UTRs is known to be has received travel funding/speaker honoraria from the Rett Syndrome 27 a major factor in triggering the NMD process, and Association of Australia and Biogen; has been an employee of Telethon the recent identification of 6.6 kb and 9.9 kb 39- Kids Institute and Curtin University; has received funding from UTRs in CDKL5 brain isoforms12 suggests that NHMRC project grants (1103745 and 1103746); and has received research support from Telethon Kids Institute and the International NMD could play an important role in the downre- Foundation for CDKL5 Research. A. Clarke has received research sup- gulation of mRNA in the event of a premature ter- port from ESRC, CDKL5 UK, and the UK FOXG1 Support Group; has mination codon, even in the last coding exon (exon received travel funding from the Nuffield Council on Bioethics and the European Society of Human Genetics; has served on the editorial boards 19 in all CDKL5 brain isoforms). Therefore, we sug- of Human Genetics, European Journal of Human Genetics, Communication gest that exon 17 and the full-length form of exon 19 & Medicine, the Journal of Community Genetics, Wiley’s Encyclopedia of should be included in all molecular diagnostic screens Life Sciences, and Genome Medicine; is a medical adviser to the (UK) for CDKL5 variants, whether by targeted gene panels Ectodermal Dysplasia Society and Rett UK; and has been supported by Edimer Pharmaceuticals in relation to a clinical trial of treatment for or by exome sequencing. a form of ectodermal dysplasia. T.A. Benke has served on the scientific The pathogenicity of CDKL5 variants can ultimately advisory boards of AveXis and Marinus; has received research support be tested in animal models or engineered human cell from Questcor Pharmaceuticals, NIH grants RO1 NS076577, lines.39 However, this study provides evidence for the U10NS077277, and U54 HD061222, unrestricted research grant from Mallinckrodt pharmaceuticals (not related to this study); has received reclassification of specific CDKL5 variants and insights research support from the Rett Syndrome Research Trust, the Citizens for genetic diagnosis. Although pathogenic CDKL5 var- United for Research in Epilepsy (CURE), and International Foundation iants are found across the majority of the coding regions for CDKL5 Research (IFCR); and has received institutional support from the University of Colorado School of Medicine, Children’s Hospital of the gene, missense variants clearly cluster in the N- Colorado, and the Children’s Hospital of Colorado Foundation Ponzio terminal catalytic domain. Missense variants outside this Family Chair for Neurology Research. J. Armstrong reports no disclo- domain and all variants in exons 20, 21, and 22 are sures. M. Pineda has received consulting fees or honoraria and payment likely to be benign. In contrast to the well-described for lectures from Actelion Pharmaceuticals Ltd and BioMarin and has been a consultant of a child neurologist on NPC for Actelion during the duplication syndromes involving closely related genes last 5 years. M.E.S. Bailey has received research support from RETTCO MECP2 and FOXG1,40,41 webelievethatmoreevidence Inc., CDKL5 UK, the Rett Syndrome Research Trust, the Rett Syn- is required to conclude that there is a well-defined drome Association Scotland, the Rosetrees Trust, the Stoneygate Trust, CDKL5 duplication syndrome. Continued evaluation and the Chief Scientist Office (Scottish Government Health Directorate; grants ETM/334 and CGA/16/3). S.R. Cobb has received research sup- of cases investigating both genotypic and phenotypic port from CDKL5 UK, the Rett Syndrome Research Trust, the Rett expressions as well as diagnoses of copy number varia- Syndrome Association Scotland, the Chief Scientist Office, the Rosetrees tions involving CDKL5 may help to elucidate this aspect Trust, and the R.S. Macdonald Charitable Trust. Go to Neurology.org/ ng for full disclosure forms. of CDKL5 biology.

Received May 2, 2017. Accepted in final form September 28, 2017. AUTHOR CONTRIBUTIONS Ralph D. Hector: study concept and design, analysis and interpretation of REFERENCES data, and drafting of the manuscript. Vera M. Kalscheuer and Friederike Hen- 1. Fehr S, Wilson M, Downs J, et al. The CDKL5 disorder is nig: analysis and interpretation of data, in vitro experiments, and critical revi- sion of the manuscript for intellectual content. Helen Leonard Jenny Downs, an independent clinical entity associated with early-onset – Angus Clarke, Tim A. Benke, Judith Armstrong, and Mercedes Pineda: acqui- encephalopathy. Eur J Hum Genet 2013;21:266 273. sition of patient data, analysis and interpretation of data, and critical revision of 2. Kalscheuer VM, Tao J, Donnelly A, et al. Disruption of the manuscript for intellectual content. Mark E.S. Bailey and Stuart R. Cobb: the serine/threonine kinase 9 gene causes severe X-linked analysis and interpretation of data and drafting of the manuscript. infantile spasms and mental retardation. Am J Hum Genet 2003;72:1401–1411. STUDY FUNDING 3. Evans JC, Archer HL, Colley JP, et al. Early onset seizures The study was funded by CDKL5 UK. and Rett-like features associated with mutations in CDKL5. Eur J Hum Genet 2005;13:1113–1120. DISCLOSURE 4. Archer HL, Evans J, Edwards S, et al. CDKL5 mutations cause infantile spasms, early onset seizures, and severe R.D. Hector has received travel funding from the LouLou foundation and has received research support from CDKL5 UK and the LouLou mental retardation in female patients. J Med Genet – Foundation. V.M. Kalscheuer has received research support from the 2006;43:729 734. Max Planck Society, the German Federal Ministry of Education and 5. Tao J, Van Esch H, Hagedorn-Greiwe M, et al. Mutations Research #01EW1408A, and the University of Pennsylvania Orphan in the X-linked cyclin-dependent kinase-like 5

8 Neurology: Genetics (CDKL5/STK9) gene are associated with severe neurode- 22. Roche Martínez A, Armstrong J, Gerotinab E, Fonsa C, velopmental retardation. Am J Hum Genet 2004;75: Campistola J, Pineda M. CDKL5 in different atypical 1149–1154. Rett syndrome variants: description of the first 6. Ricciardi S, Ungaro F, Hambrock M, et al. CDKL5 eight patients from Spain. J Pediatr Epilepsy 2012;1: ensures excitatory synapse stability by reinforcing NGL- 27–35. 1-PSD95 interaction in the postsynaptic compartment and 23. Sprovieri T, Conforti FL, Fiumara A, et al. A novel muta- is impaired in patient iPSC-derived neurons. Nat Cel Biol tion in the X-linked cyclin-dependent kinase-like 5 2012;14:911–923. (CDKL5) gene associated with a severe Rett phenotype. 7. Gokben S, Onay H, Yilmaz S, et al. Targeted next gener- Am J Med Genet A 2009;149A:722–725. ation sequencing: the diagnostic value in early-onset 24. Bahi-Buisson N, Nectoux J, Rosas-Vargas H, et al. Key epileptic encephalopathy. Acta Neurol Belg 2017;117: clinical features to identify girls with CDKL5 mutations. 131–138. Brain 2008;131:2647–2661. 8. Fehr S, Wong K, Chin R, et al. Seizure variables and 25. Trump N, McTague A, Brittain H, et al. Improving their relationship to genotype and functional abilities diagnosis and broadening the phenotypes in early- in the CDKL5 disorder. Neurology 2016;87:2206– onset seizure and severe developmental delay disorders 2213. through gene panel analysis. J Med Genet 2016;53: 9. Lek M, Karczewski KJ, Minikel EV, et al. Analysis of 310–317. protein-coding genetic variation in 60,706 humans. 26. Lilles S, Talvik I, Noormets K, et al. CDKL5 gene- Nature 2016;536:285–291. related epileptic encephalopathy in Estonia: four cases, 10. Genomes Project C, Auton A, Brooks LD, et al. A global one novel mutation causing severe phenotype in a boy, reference for human genetic variation. Nature 2015;526: and overview of the literature. Neuropediatrics 2016;47: 68–74. 631–367. 11. Sherry ST, Ward MH, Kholodov M, et al. dbSNP: the 27. Yepiskoposyan H, Aeschimann F, Nilsson D, Okoniewski NCBI database of genetic variation. Nucleic Acids Res M, Muhlemann O. Autoregulation of the nonsense- 2001;29:308–311. mediated mRNA decay pathway in human cells. RNA 12. Hector RD, Dando O, Landsberger N, et al. Character- 2011;17:2108–2118. isation of CDKL5 transcript isoforms in human and 28. Intusoma U, Hayeeduereh F, Plong-On O, et al. Muta- mouse. PLoS One 2016;11:e0157758. tion screening of the CDKL5 gene in cryptogenic infantile 13. Christodoulou J, Grimm A, Maher T, Bennetts B. Re- intractable epilepsy and review of clinical sensitivity. Eur J ttBASE: the IRSA MECP2 variation database-a new Paediatr Neurol 2011;15:432–438. mutation database in evolution. Hum Mutat 2003;21: 29. Psoni S, Willems PJ, Kanavakis E, et al. A novel 466–472. p.Arg970X mutation in the last exon of the CDKL5 gene 14. Firth HV, Richards SM, Bevan AP, et al. DECIPHER: resulting in late-onset seizure disorder. Eur J Paediatr Neu- database of chromosomal imbalance and phenotype in rol 2010;14:188–191. humans using Ensembl resources. Am J Hum Genet 30. Wang T, Guo H, Xiong B, et al. De novo genic mutations 2009;84:524–533. among a Chinese autism spectrum disorder cohort. Nat 15. McLaren W, Gil L, Hunt SE, et al. The Ensembl variant Commun 2016;7:13316. effect predictor. Genome Biol 2016;17:122. 31. Froyen G, Van Esch H, Bauters M, et al. Detection of 16. Richards S, Aziz N, Bale S, et al. Standards and guide- genomic copy number changes in patients with idiopathic lines for the interpretation of sequence variants: a joint mental retardation by high-resolution X-array-CGH: consensus recommendation of the American College important role for increased gene dosage of XLMR genes. of Medical Genetics and Genomics and the Association Hum Mutat 2007;28:1034–1042. for Molecular Pathology. Genet Med 2015;17:405– 32. Tzschach A, Chen W, Erdogan F, et al. Characterization 424. of interstitial Xp duplications in two families by tiling path 17. Musante L, Kunde SA, Sulistio TO, et al. Common path- array CGH. Am J Med Genet Part A 2008;146A:197– ological mutations in PQBP1 induce nonsense-mediated 203. mRNA decay and enhance exclusion of the mutant exon. 33. Thorson L, Bryke C, Rice G, et al. Clinical and molecular Hum Mutat 2010;31:90–98. characterization of overlapping interstitial Xp21-p22 du- 18. Williamson SL, Giudici L, Kilstrup-Nielsen C, et al. A novel plications in two unrelated individuals. Am J Med Genet transcript of cyclin-dependent kinase-like 5 (CDKL5) has an Part A 2010;152A:904–915. alternative C-terminus and is the predominant transcript in 34. Sismani C, Anastasiadou V, Kousoulidou L, et al. 9 Mb brain. Hum Genet 2012;131:187–200. familial duplication in chromosome band Xp22.2-22.13 19. Fichou Y, Nectoux J, Bahi-Buisson N, Chelly J, Bienvenu associated with mental retardation, hypotonia and devel- T. An isoform of the severe encephalopathy-related opmental delay, scoliosis, cardiovascular problems and CDKL5 gene, including a novel exon with extremely high mild dysmorphic facial features. Eur J Med Genet 2011; sequence conservation, is specifically expressed in brain. 54:e510–e515. J Hum Genet 2011;56:52–57. 35. Szafranski P, Golla S, Jin W, et al. Neurodevelopmental 20. Allou L, Julia S, Amsallem D, et al. Rett-like phenotypes: and neurobehavioral characteristics in males and females expanding the genetic heterogeneity to the KCNA2 gene with CDKL5 duplications. Eur J Hum Genet 2015;23: and first familial case of CDKL5-related disease. Clin 915–921. Genet 2016;91:431–440. 36. Nemos C, Lambert L, Giuliano F, et al. Mutational spec- 21. Liang JS, Shimojima K, Takayama R, et al. CDKL5 alter- trum of CDKL5 in early-onset encephalopathies: a study ations lead to early epileptic encephalopathy in both gen- of a large collection of French patients and review of the ders. Epilepsia 2011;52:1835–1842. literature. Clin Genet 2009;76:357–371.

Neurology: Genetics 9 37. Diebold B, Delepine C, Gataullina S, Delahaye A, Nec- leading to autistic-like phenotypes in mice. Proc Natl toux J, Bienvenu T. Mutations in the C-terminus of Acad Sci USA 2012;109:21516–21521. CDKL5: proceed with caution. Eur J Hum Genet 2014; 40. Seltzer LE, Sohnee A, Paciorkowski AR, et al. Devel- 22:270–272. opmental and epilepsy follow-up of children with du- 38. Huopaniemi L, Tyynismaa H, Rantala A, Rosenberg T, Ali- plications of FOXG1 on 14q12. Ann Neurol 2013;74: talo T. Characterization of two unusual RS1 gene deletions S161. segregating in Danish retinoschisis families. Hum Mutat 41. Van Esch H, Bauters M, Ignatius J, et al. Duplication of the 2000;16:307–314. MECP2 region is a frequent cause of severe mental retarda- 39. Wang IT, Allen M, Goffin D, et al. Loss of CDKL5 tion and progressive neurological symptoms in males. Am J disrupts kinome profile and event-related potentials Hum Genet 2005;77:442–453.

10 Neurology: Genetics Factors influencing the age at onset in familial frontotemporal lobar dementia Important weight of genetics

Mathieu Barbier, PhD ABSTRACT Agnès Camuzat, PhD Objective: To quantify the effect of genetic factors and generations influencing the age at onset Marion Houot, MS (AAO) in families with frontotemporal lobar dementia (FTD) due to C9ORF72 hexanucleotide Fabienne Clot, PhD repeat expansions and GRN mutations. Paola Caroppo, MD, PhD Methods: We studied 504 affected individuals from 133 families with C9ORF72 repeat expan- Clémence Fournier, PhD sions and 90 FTD families with mutations in GRN, 2 major genes responsible for FTD and/or Daisy Rinaldi, PhD amyotrophic lateral sclerosis. Intrafamilial correlations of AAO were analyzed, and variance com- Florence Pasquier, MD, ponent methods were used for heritability estimates. Generational effects on hazard rates for PhD AAO were assessed using mixed-effects Cox proportional hazard models. Didier Hannequin, MD, PhD Results: A generational effect influencing AAO was detected in both C9ORF72 and GRN families. Jérémie Pariente, MD, Nevertheless, the estimated proportion of AAO variance explained by genetic factors was high in 5 2 PhD FTD caused by C9ORF72 repeat expansions (44%; p 1.10e 4), 62% when the AAO of 5 2 Kathy Larcher, BS dementia was specifically taken into account (p 8.10e 5),andtoalesserdegreeinGRN 5 The French Clinical and families (26%; p 0.17). Intrafamilial correlation analyses revealed a significant level of corre- Genetic Research lations in C9ORF72 families according to the degree of kinship. A pattern of intrafamilial corre- Network on FTD/ lations also suggested potential X-linked modifiers acting on AAO. Nonsignificant correlation FTD-ALS values were observed in GRN families. Predict-PGRN & Conclusions: Our results provide original evidence that genetic modifiers strongly influence the PrevDemAls Study AAO in C9ORF72 carriers, while their effects seem to be weaker in GRN families. This consti- Groups tutes a rational to search for genetic biomarkers, which could help to improve genetic counseling, Alexis Brice, MD patient care, and monitoring of therapeutic trials. Neurol Genet 2017;3:e203; doi: 10.1212/ Emmanuelle Génin, PhD NXG.0000000000000203 Audrey Sabbagh, PharmD, PhD GLOSSARY Isabelle Le Ber, MD, PhD AAO 5 age at onset; ALS 5 amyotrophic lateral sclerosis; FTD 5 frontotemporal lobar dementia.

Frontotemporal lobar dementia (FTD) is a rare neurodegenerative disease caused by neuronal Correspondence to death in the frontal and temporal lobes and represents the second cause of early-onset dementia Dr. Le Ber: [email protected] after Alzheimer disease. FTD is associated with behavioral changes and language dysfunctions and may be associated with amyotrophic lateral sclerosis (ALS). Thanks to the use of next- generation sequencing, the identification of more than 20 FTD-associated genes allowed molec- ular diagnosis in more than 60%–70% of familial cases. Among these genes, mutations in GRN and intronic hexanucleotide expansions in C9ORF72 are the major causes of FTD and/or

From the INSERM U1127 (M.B., A.C., P.C., C.F., D.R., A.B., I.L.B.), CNRS UMR 7225, UPMC Université Paris 06 UMR S1127, Sorbonne Université Institut du Cerveau et de la Moelle épinière, ICM; Ecole Pratique des Hautes Etudes—EPHE (A.C.), PSL Research University; Institute of Memory and Alzheimer’s Disease (IM2A) (M.H., P.C., D.R., I.L.B.), Centre of Excellence of Neurodegenerative Disease (CoEN), ICM, APHP Department of Neurology, Hopital Pitié-Salpêtrière, University Paris 6; Unité Fonctionnelle de Neurogénétique Moléculaire et Cellulaire (F.C., K. L.), Département de Génétique et Cytogénétique, AP-HP, Hôpitaux Universitaires La Pitié Salpêtrière-Charles Foix, Paris, France; Carlo Besta Neurological Institute (P.C.), IRCCS Foundation, Milano, Italy; Assistance Publique—Hôpitaux de Paris (D.R., I.L.B.), Hôpital Pitié-Salpêtrière, Centre de référence Démences Rares, Paris, France; Université de Lille (F.P.), Inserm U1171, CHU Lille, Labex DistAlz, LiCEND, France; Department of Neurology (D.H.), University Hospital, Rouen, France; Département de Neurologie (J.P.), CHU Toulouse, Equipe TONIC, INSERM, Place du Dr Baylac, France; INSERM (E.G.), UMR1078, CHU Brest, Université Bretagne Occidentale, France; and Institut de Recherche pour le Développement (IRD) (A.S.), UMR216—MERIT, Paris, France. Funding information and disclosures are provided at the end of the article. Go to Neurology.org/ng for full disclosure forms. 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.

Neurology.org/ng Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 ALS.1–4 In both genetic forms, the clinical Routines) computer package.9 Briefly, this method applies max- expression of the disease can vary from an early imum likelihood estimation to a mixed-effects model that in- corporates fixed effects for known covariates and variance occurrence of symptoms in the third decade of components for genetic effects. Both AAO and AAO-FTD in life to a virtually incomplete penetrance in C9ORF72 patients and AAO in GRN patients followed a normal elderly carriers. To date, only TMEM106B distribution. The heritability (h2) of a phenotype was estimated as has been described as a potential modifier gene the ratio of additive genetic variance to the total phenotypic variance unexplained by covariates. of FTD influencing the penetrance of the dis- Familial correlation coefficients were estimated using the pro- – ease.5 8 However, this association mostly con- gram FCOR of the SAGE v6.4 software package (Statistical Anal- cerned FTD cases caused by GRN mutations ysis for Genetic Epidemiology; 2016).10 Correlations were calculated using the uniform weighting scheme. Description of and did not explain all the observed variability. affected pairs of relatives per family is provided in table e-2, Therefore, the question of the variability of the http://links.lww.com/NXG/A5. As a previous study showed that age at onset (AAO) remains largely unex- penetrance may be sex dependent in patients with FTD with 11 plained, and little is known about the global C9ORF72 mutations, both familial correlations and heritability estimates were calculated using residual trait values of AAO and heritability of AAO in FTD. Here, we provide AAO-FTD after adjusting for sex. further insights into heritability estimates and Anticipation was investigated in C9ORF72 and GRN families patterns of intrafamilial correlations to evalu- using mixed-effects Cox proportional hazard regression models to ate the genetic component of AAO variability test for a generational effect on the AAO. The statistical model allows the inclusion of unaffected carriers (censored data). Unaf- in FTD because this strongly influences fected carriers or obligate unaffected carriers of C9ORF72 repeat genetic counseling, therapeutic trials, and, to expansions or GRN mutations were included in this analysis and a broader extent, patient care. censored at the age of last evaluation. Generation and sex effects were added in the model, each patient had an individual random risk, which were correlated according to the strength of relation- METHODS Standard protocol approvals, registrations, ship using the kinship matrix. The kinship matrix was assessed on and patient consents. All individuals were enrolled in agree- all individuals, but Cox models were fitted only on carriers ment with bioethics laws (Institutional Review Board: CPP Ile (including both affected and unaffected individuals) and on fam- de France II, project #RBM 02-59). Written informed consent ilies with at least 2 consecutive generations. Clinical data were was obtained from all patients (or guardians of patients) partici- available in pedigrees up to 5 consecutive generations. To get pating in the study (consent for research). more reliable estimates of generational effects, generations with Study population. This study was initially based on a global a number of informative individuals lower than 10% of total cohort of 702 probands and affected relatives from 322 indepen- individuals included in the analysis were discarded. Finally, anal- dent FTD families carrying C9ORF72 or GRN mutations, assem- yses were performed on 3 consecutive generations in C9ORF72 bled by the French clinical and genetic research network on FTD/ families (312 individuals among 79 families) and 2 consecutive FTD-ALS. AAO was defined as the age of occurrence of the first generations in GRN families (119 individuals among 42 families). symptoms (either behavioral, language, or motor), in years, as Distribution of individuals (both censored and uncensored) ac- reported by the principal informant. In most cases, a second cording to generations is indicated in table 1. Table e-3, http:// informant was questioned independently to accurately determine links.lww.com/NXG/A6, provides additional details about cen- the AAO, and only patients/families with concordant data were sored and uncensored statuses among FTD families. The effect of included in the study. After a careful review of all cases, 99 generation was tested using log-likelihood tests. Statistical analysis families were excluded because clinical information or second was performed using R 3.3.2 (The R foundation for Statistical informant’s interview was lacking, making the determination of computing) and using coxme 2.2-5 and kinship2 1.6.4 packages. AAO impossible or not reliable enough. Finally, 504 probands and affected relatives from 223 families RESULTS The heritability of traits was assayed in with robust data were used for the analyses, including 333 probands C9ORF72 families using phenotypical and pedigree and affected relatives (166 women and 167 men) from 133 C9ORF72 families as well as 171 probands and affected relatives (87 women and 84 men) from 90 GRN families. The AAO ranged from 30 to 91 years old (mean AAO 5 58.3 6 9.9 SD) in Table 1 Distribution of individuals (both C9ORF72 patients. In 236 C9ORF72 probands and relatives, the censored and uncensored) according to AAO of dementia (AAO-FTD) was specifically considered in con- mutations and generations in FTD trast to other symptoms (ALS and others) and analyzed separately, families selected for the generational as C9ORF72 repeat expansions could lead to both dementia and effect analysis ALS. Patients with mutations of GRN only developed symptoms of Generation C9ORF72 GRN dementia. The AAO ranged from 33 to 86 years old (mean AAO 5 60.8 6 8.5 SD). Most individuals (.97%) were of Caucasian Earliest born 40 —

origin. The structure of C9ORF72 and GRN pedigrees is detailed Second born 148 53 in table e-1, http://links.lww.com/NXG/A4. Latest born 124 66 Statistical analyses. The heritability of traits was estimated Total 312 119 using a variance components method as implemented in the SOLAR v8.1.4 (Sequential Oligogenic Linkage Analysis Abbreviation: FTD 5 frontotemporal lobar dementia.

2 Neurology: Genetics Table 2 Intrafamilial correlation coefficients (r) among 348 affected relative pairs from 333 cases in 133 C9ORF72 families and 114 affected pairs from 171 cases in 90 GRN families

Parent-offspring Sibling Avuncular Cousin

rpValue rpValue rpValue rpValue

C9ORF72

AAO 0.12 (111) 0.36 0.46 (145) ,1.10e24 0.20 (59) 0.39 0.28 (33) 0.24

AAO-FTD 0.33 (58) 0.06 0.66 (68) ,1.10e24 0.51 (20) 0.12 0.44 (12) 0.19

GRN

AAO 0.13 (58) 0.21 0.24 (42) 0.13 20.58 (8) 0.11 20.48 (4) 0.31

Abbreviations: AAO 5 age at onset; FTD 5 frontotemporal lobar dementia. The number of pairs is indicated in brackets.

data for the AAO (h2 5 0.44 6 0.13; p 5 1.10e24) between mother-son, brother-brother, and brother- and AAO-FTD (h2 5 0.62 6 0.17; p 5 8.10e25). sister pairs and a moderate correlation in sister-sister Table 2 resumes familial correlations among relatives. pairs (table 3). Significant correlations were observed among sibling Analyses of GRN families revealed a nonsignificant pairs (r 5 0.46), and nonsignificant correlations for heritability estimate of AAO (h2 5 0.26) and intra- second- and third-degree relatives that are more familial correlations values (table 2). However, we genetically distant (r 5 0.20 and r 5 0.28, respec- observed a common thread with the analysis of tively). Similar patterns were observed for the C9ORF72 families. The highest correlation values AAO-FTD phenotype with stronger and significant were observed for sibling pairs (r 5 0.24), and the correlation values (table 2). analysis of subtypes revealed higher correlation values We did not observe any significant correlation for mother-son and brother-sister pairs (r 5 0.36 and among first-degree relatives when parent-offspring r 5 0.34, respectively) without reaching statistical pairs were considered. This discordance between pa- significance. rents and offspring prompted us to analyze subtypes The lower parent-offspring correlation values of parent-offspring pairs, e.g., father-offspring, observed in C9ORF72 and GRN families compared mother-offspring, and sibling subtypes (table 3). A with the sibling correlation may also underline a gen- significant correlation was observed only in mother- erational effect on AAO estimates. Cumulative pro- son pairs (r 5 0.51), conversely to mother-daughter portions of disease onset in C9ORF72 and GRN or father-offspring pairs. Besides, brother-brother and families according to generations are presented in fig- sister-brother pairs were significantly correlated ure. In both C9ORF72 and GRN families, individuals among sibling subtypes (r 5 0.58 and r 5 0.56, in earliest-born generations have a reduced hazard respectively) in contrast to sister-sister pairs (r 5 ratio to develop disease compared with those in the 0.21; nonsignificant). When AAO-FTD was consid- latest-born generation (p , 0.001 and p 5 0.021, ered, we observed highly significant correlations respectively; table 4). Values of the hazard ratio were closer between C9ORF72 and GRN families when taking into account the 2 more recent generations, Table 3 Intrafamilial correlation coefficients (r) of AAO among first-degree i.e., the latest-born vs the second-born generations relative subtypes in C9ORF72 and GRN families (0.45 vs 0.61, respectively; table 4). No generational C9 AAO C9 AAO-FTD GRN AAO effect was detected when considering the AAO of ALS (data not shown). Pair subtypes rpValue rpValue rpValue

Father-son 0.03 (31) 0.89 20.15 (19) 0.65 0.19 (13) 0.54 DISCUSSION We provide the evidence for a strong Mother-son 0.51 (33) 0.008 0.57 (17) 0.03 0.36 (15) 0.18 genetic component influencing AAO in genetic

Father-daughter 0.06 (17) 0.89 0.36 (7) 0.63 0.07 (15) 0.83 FTD in C9ORF72 families. Heritability estimates

Mother-daughter 20.12 (30) 0.58 0.46 (15) 0.09 0.08 (15) 0.79 reached values closed to highly heritable diseases such as schizophrenia in which heritability estimates are Brother-brother 0.58 (37) 0.0003 0.71 (15) ,1.10e24 20.17 (9) 0.62 between 0.4 and 0.8.12 The analysis of GRN families , 2 , 2 Brother-sister 0.56 (70) 1.10e 4 0.75 (33) 1.10e 4 0.34 (23) 0.11 revealed more contrasted results. If we cannot exclude Sister-sister 0.21 (38) 0.22 0.46 (20) 0.04 0.08 (10) 0.82 a weaker effect of genetic factors on the AAO in the

Abbreviations: AAO 5 age at onset; FTD 5 frontotemporal lobar dementia. GRN cohort, the lower number of families (especially The number of pairs is indicated in brackets. extended families) and patients included compared

Neurology: Genetics 3 Figure Cumulative proportions of disease onset in familial frontotemporal lobar dementia with at least 2 successive generations

Individuals were grouped by generations: later-born individuals belong to the youngest generation, second born and earliest born are parents and grand parents, respectively. (A) Patients with C9ORF72 repeat expansions; (B) patients with GRN mutations.

with the C9ORF72 cohort may not allow a reliable role of a shared environment, which was not observed estimation of heritability. here. The same is true for parent-offspring pairs in In the C9ORF72 population, the heritability ex- which significant correlations have been observed on- plains between 44 and 62% of the phenotypic vari- ly for mother-son pairs (table 3). If we cannot com- ance, given the phenotype considered (AAO and pletely exclude an impact of a shared environment on AAO-FTD, respectively). Results from the analysis our results, this role seems to be minor. of the AAO were very likely influenced by the C9ORF72 mutations lead to pathogenic AAO-FTD. First, the AAO-FTD was the most fre- GGGGCC hexanucleotide expansions. Healthy con- quent available phenotype in our study. Second, the trols usually carry 2 to 23 copies of the repeat, coefficient values and significance of results were con- whereas patients have up to hundreds or thousands sistently improved when the appearance of dementia of hexanucleotide copies. Some studies evaluated symptoms was taken into account. Unfortunately, we whether the number of hexanucleotide repeats may did not have enough patients with isolated or associ- influence the AAO, with contradictory conclusions.13–15 ated ALS to perform similar and reliable analyses. This question is still debated, but we cannot exclude One classic criticism of heritability estimates is that that expansion length variability may explain a part of the influence of shared environment between relatives the heritability in C9ORF72 families. We were is often underestimated. However, the differences of unable to examine the residual heritability after con- correlation values among parent-offspring and sibling trolling for the C9ORF72 repeat expansion size. subtypes did not support a major role of a shared Indeed, large C9ORF72 repeat expansions cannot environment in the phenotypic correlations observed be sized by repeat-primed PCR, the methods classi- in first-degree relatives. Correlation values between cally used for molecular diagnosis. Second, the repeat brother-brother, brother-sister, and sister-sister pairs length detected in peripheral lymphocytes is different should have been homogeneous in case of a major and does not correlate with the repeat length in brain tissue,14 thus precluding accurate correlations between the size of the repeat in affected tissues and Table 4 Generational effect analysis the clinical parameters.

Hazard ratio (95% CI) We detected lower correlation values in parent- offspring pairs in C9ORF72 in our study. We showed p Earliest-born generation Second-born generation Generational effect value that specific parent-of-origin effects in transmission of C9ORF72 0.16 (0.08–0.30) 0.45 (0.30–0.67) ,0.001 pathogenic alleles may explain such discordances in GRN — 0.61 (0.40–0.93) 0.021 our cohort. Besides, a lower correlation of AAO among parent-offspring pairs could be due to the Abbreviation: CI 5 confidence interval. The hazard ratio of risk to develop disease in C9ORF72 and GRN families was estimated effect of successive generations on disease onset with the latest-born generation as reference. (anticipation) as well.

4 Neurology: Genetics Indeed, anticipation is a well-known phenomenon assessment of AAO with a define procedure used by influencing AAO in repeat expansion disorders such clinicians to reduce and standardize the relative inac- as Huntington disease or spinocerebellar ataxias.16 curacy of AAO between patients. This term implies that the unstable repeat expansion However, and despite a generational effect on size tends to increase in successive generations, lead- AAO affecting parent-offspring correlation values, it ing to an earlier and/or more severe expression of the is noteworthy that the most significant correlation disease in later-born generations. Anticipation of the was observed only among mother-son pairs. The AAO has been suggested in C9ORF72 families in length of the pathogenic expansion may be more sta- a clinically based study.17 However, evidence of ble from mother to offspring rather than from father genetic anticipation is much more delicate to demon- to offspring, as previously shown in other diseases due strate at a molecular level in C9ORF72 families than to nucleotide repeat expansions such as Huntington in other repeat expansion diseases as mentioned disease.19 This could imply particular models of above. Moreover, expansions and even contractions inheritance such as mitochondrial or X-linked inher- of the repeats can happen in successive generations, as itance as well. The fact that best correlation values it has been shown for short C9ORF72 expansions.17 were observed in mother-son, father-daughter, and For these reasons, disease anticipation due to a repeat mother-daughter pairs could imply more likely X- expansion phenomenon has not been conclusively linked genetic modifiers. Indeed, X-linked transmis- demonstrated at the molecular level up to now. sion may explain higher correlation values when the We measured the generational effect on AAO in X-chromosome is transmitted to the offspring. In C9ORF72 families using a mixed-effects Cox propor- addition, the X-chromosome inactivation that ran- tional hazard regression model that takes into account domly affects one of the 2 X-chromosomes in women the degree of kinship. It is important that this model could influence the effect of potential X-linked modi- allows the inclusion of unaffected carriers (censored fiers and, subsequently, explains the reduced correla- data) as well, which limits the right truncation bias, tion coefficients, observed in sister-sister pairs. The especially in the study of AAO compared with meth- hypothesis about an X-linked modifier based on sta- ods including only affected members.18 We detected tistical observations needs to be confirmed in larger a strong generational effect on AAO, the AAO being populations and further investigated at the molecular earlier in the latest-born generation than in the level. Notably, we and others already evoked sex- ascending generation. In contrast to dynamic mediated modifiers of C9ORF72 disease in FTD as C9ORF72 pathogenic repeat expansions, GRN muta- well as in ALS.11,20,21 Of interest, X-linked modifiers tions are stable single nucleotide variations or inser- have also been recently suggested in other repeat tion/deletion leading to haploinsufficiency.1,2 expansion disorders such as spinocerebellar ataxia Strikingly, we observed a significant generational type 2 in which the residual heritability not explained effect on AAO in GRN families with the same trend by the expansion length seems to be partially driven and the same range of the hazard ratio compared with by X-linked factors.22 This may suggest common C9ORF72 families. So, it seems that a significant pro- genetic mechanisms modulating AAO in repeat portion of the generational effect in C9ORF72 and in expansion diseases. GRN pedigrees is likely to be due to observational These results derived from a familial approach are biases rather than based on molecular mechanisms, in favor of a significant genetic component influenc- as stable mutations in GRN cannot explain earlier ing the AAO of FTD in patients with C9ORF72 AAO in successive generations. repeat expansions and in a lesser extent in patients A classic problem in studies involving AAO is with GRN mutations. This work justifies and reinfor- a biased ascertainment of the AAO among familial ces the interest of searching for biomarkers and in cases independently of mutational effects and despite particular genetic biomarkers influencing disease all precautions. Indeed, the first symptoms may be onset in FTD, thus acting on genetic counseling, detected slightly earlier in younger generations when therapeutic trials, and, to a broader extent, patient one of their parents has been affected by the same dis- care. ease previously. A more efficient detection of cases due to technical or knowledge advances could also AUTHOR CONTRIBUTIONS lead to detect cases earlier as before. In addition, it Dr. Mathieu Barbier and Dr. Isabelle Le Ber: study concept and design. Dr. Agnès Camuzat, Dr. Clémence Fournier, and Dr. Fabienne should be stressed that assessing precisely AAO, espe- Clot: acquisition of molecular data. Dr. Fabienne Clot, Dr. Paola cially the AAO of dementia, implies more inaccuracy Caroppo, Dr. Daisy Rinaldi, Dr. Florence Pasquier, Dr. Didier Hannequin, than assessing the clinical onset of motor symptoms Dr. Jérémie Pariente, and Kathy Larcher: acquisition of patients’ data. of ALS in the absence of biomarkers of disease onset. Dr.MathieuBarbier,Dr.AudreySabbagh,Dr.EmmanuelleGénin,and Marion Houot: analysis and interpretation of data. Dr. Mathieu Barbier, To reduce this bias, we first excluded familial cases Dr. Emmanuelle Génin, Dr. Audrey Sabbagh, Dr. Alexis Brice, and with unclear data. Second, we normalized the Dr. Isabelle Le Ber: drafting or revising the manuscript.

Neurology: Genetics 5 ACKNOWLEDGMENT lobar degeneration patients with GRN mutations and Predict-PGRN & PrevDemAls studies were conducted thanks to the Cen- C9orf72 repeat expansions. Neurobiol Aging 2014;35: tre d’Investigation Clinique (CIC1422) at IHU-A-ICM, Paris, France. We 2658.e1–2658.e5. thank clinicians from the French clinical and genetic research network on 7. Gallagher MD, Suh E, Grossman M, et al. TMEM106B is FTLD/FTLD-ALS and the Predict-PGRN-PrevDEMAls study groups a genetic modifier of frontotemporal lobar degeneration (file e-4, http://links.lww.com/NXG/A7). with C9orf72 hexanucleotide repeat expansions. Acta Neuropathol (Berl) 2014;127:407–418. STUDY FUNDING 8. van Blitterswijk M, Mullen B, Nicholson AM, et al. The research leading to these results has received funding from the pro- gram “Investissements d’avenir” ANR-10-IAIHU-06. This work was par- TMEM106B protects C9ORF72 expansion carriers tially funded by Assistance Publique—Hôpitaux de Paris (Clinical against frontotemporal dementia. Acta Neuropathol (Berl) – Research and Development Department) Programme Hospitalier de Re- 2014;127:397 406. cherche Clinique (PHRC) FTLD-exome RCAOM-12123 (to I.L.B.), 9. Almasy L, Blangero J. Multipoint quantitative-trait linkage The PHRC Predict-PGRN P071229-AOM-08045 (to I.L.B.) and the analysis in general pedigrees. Am J Hum Genet 1998;62: ANR-PRTS PREV-DEMALS project (to I.L.B., funding and promotion 1198–1211. by Assistance Publique—Hôpitaux de Paris and ANR). 10. S.A.G.E. [2016] Statistical Analysis for Genetic Epidemi- ology, Release 6.4. Available at: http://darwin.cwru.edu. DISCLOSURE Accessed September 5, 2017. M. Barbier, A. Camuzat, M. Houot, F. Clot, P. Caroppo, C. Fournier, 11. Le Ber I, Camuzat A, Guillot-Noel L, et al. C9ORF72 and D. Rinaldi report no disclosures. F. Pasquier has served on the scien- tific advisory board of Sanofi and Novartis; has served as a principal inves- repeat expansions in the frontotemporal dementias spec- tigator for clinical trials funded by Neuus (Piramal), Forum trum of diseases: a flow-chart for genetic testing. J Alz- – Pharmaceuticals, Probiodrug, Lilly, Ramen, DIAN-TU, EnVivo Pharma- heimers Dis 2013;34:485 499. ceutical, Roche, and Axovant; and has received research support from 12. Cardno AG, Gottesman II. Twin studies of schizophrenia: ANR Conseil Regional du Nord-Pas-de-Calais CNSA, Fondation Médé- from bow-and-arrow concordances to star wars Mx and ric Alzheimer, and Fondation Plan Alzheimer. D. Hannequin has functional genomics. Am J Med Genet 2000;97:12–17. received research support from the French Ministry of Health. J. Pariente 13. van Blitterswijk M, DeJesus-Hernandez M, Niemantsverdriet has served on the scientific advisory boards of GE Healthcare and Lilly E, et al. Association between repeat sizes and clinical and and has served on the editorial boards of the Journal of Alzheimer’s Disease pathological characteristics in carriers of C9ORF72 repeat and Revue Neurologique. K. Larcher reports no disclosures. A. Brice has expansions (Xpansize-72): a cross-sectional cohort study. served on the scientific advisory boards of the FWO (Research Founda- – tion Flanders), the ERC (European Research Council), and the BMBF Lancet Neurol 2013;12:978 988. (Bundesministerium für Bildung und Forschung—Berlin, Germany); has 14. Nordin A, Akimoto C, Wuolikainen A, et al. Extensive served on the editorial boards of Neurology and Clinical Neuroscience, size variability of the GGGGCC expansion in C9orf72 in Parkinsonism and Related Disorders, Brain, Neurodegenerative Diseases, both neuronal and non-neuronal tissues in 18 patients The , and Neurogenetics; and has received research support with ALS or FTD. Hum Mol Genet 2015;24:3133–3142. from the French Research Agency (EU), the France Parkinson Associa- 15. Gijselinck I, Van Mossevelde S, van der Zee J, et al. The tion, the RDS (Roger de Spoelberch Foundation), the FDF (Fondation C9orf72 repeat size correlates with onset age of disease, de France), and the FRM (Fondation pour la Recherche Médicale). DNA methylation and transcriptional downregulation of E. Génin has served on the editorial boards of Human Heredity and the promoter. Mol Psychiatry 2016;21:1112–1124. Genetic Epidemiology. A. Sabbagh and I. Le Ber report no disclosures. 16. Jones L, Houlden H, Tabrizi SJ. DNA repair in the Go to Neurology.org/ng for full disclosure forms. trinucleotide repeat disorders. Lancet Neurol 2017; Received July 19, 2017. Accepted in final form September 20, 2017. 16:88–96. 17. Van Mossevelde S, van der Zee J, Gijselinck I, et al. REFERENCES Clinical evidence of disease anticipation in families seg- 1. Cruts M, Gijselinck I, van der Zee J, et al. Null mutations in regating a C9orf72 repeat expansion. JAMA Neurol progranulin cause ubiquitin-positive frontotemporal dementia 2017;74:445–452. linked to chromosome 17q21. Nature 2006;442:920–924. 18. Boonstra PS, Gruber SB, Raymond VM, et al. A review of 2. Baker M, Mackenzie IR, Pickering-Brown SM, et al. Mu- statistical methods for testing genetic anticipation: looking tations in progranulin cause tau-negative frontotemporal for an answer in Lynch syndrome. Genet Epidemiol 2010; dementia linked to chromosome 17. Nature 2006;442: 34:756–768. 916–919. 19. Duyao M, Ambrose C, Myers R, et al. Trinucleotide 3. Renton AE, Majounie E, Waite A, et al. A hexanucleotide repeat length instability and age of onset in Huntington’s repeat expansion in C9ORF72 is the cause of chromosome disease. Nat Genet 1993;4:387–392. 9p21-linked ALS-FTD. Neuron 2011;72:257–268. 20. Murphy NA, Arthur KC, Tienari PJ, Houlden H, Chiò A, 4. DeJesus-Hernandez M, Mackenzie IR, Boeve BF, et al. Traynor BJ. Age-related penetrance of the C9orf72 repeat Expanded GGGGCC hexanucleotide repeat in noncoding expansion. Sci Rep 2017;7:2116. region of C9ORF72 causes chromosome 9p-linked FTD 21. Rooney J, Fogh I, Westeneng H-J, et al. C9orf72 expan- and ALS. Neuron 2011;72:245–256. sion differentially affects males with spinal onset amyotro- 5. Van Deerlin VM, Sleiman PMA, Martinez-Lage M, et al. phic lateral sclerosis. J Neurol Neurosurg Psychiatry 2017; Common variants at 7p21 are associated with frontotem- 88:281. poral lobar degeneration with TDP-43 inclusions. Nat 22. Figueroa KP, Coon H, Santos N, Velazquez L, Mederos Genet 2010;42:234–239. LA, Pulst S-M. Genetic analysis of age at onset variation in 6. Lattante S, Le Ber I, Galimberti D, et al. Defining the spinocerebellar ataxia type 2. Neurol Genet 2017;3:e155. association of TMEM106B variants among frontotemporal doi: 10.1212/NXG.0000000000000155.

6 Neurology: Genetics NaV channel variants in patients with painful and nonpainful peripheral neuropathy

Samir Wadhawan, PhD ABSTRACT Saumya Pant, PhD Objective: To examine the incidence of nonsynonymous missense variants in SCN9A (NaV1.7), Ryan Golhar, PhD SCN10A (NaV1.8), and SCN11A (NaV1.9) in patients with painful and nonpainful peripheral Stefan Kirov, PhD neuropathy. John Thompson, PhD Methods: Next-generation sequencing was performed on 457 patient DNA samples provided by Leslie Jacobsen, MD the Peripheral Neuropathy Research Registry (PNRR). The patient diagnosis was as follows: 278 Irfan Qureshi, MD idiopathic peripheral neuropathy (67% painful and 33% nonpainful) and 179 diabetic distal poly- Senda Ajroud-Driss, MD neuropathy (77% painful and 23% nonpainful). Roy Freeman, MD David M. Simpson, MD Results: We identified 36 (SCN9A), 31 (SCN10A), and 15 (SCN11A) nonsynonymous missense , A. Gordon Smith, variants, with 47.7% of patients carrying a low-frequency (minor allele frequency 5%) MD, FAAN missense variant in at least 1 gene. The incidence of previously reported gain-of-function # Ahmet Hoke, PhD, MD missense variants was low ( 3%), and these were detected in patients with and without Linda J. Bristow, PhD . There were no significant differences in missense variant allele frequencies of any On behalf of the gene, or SCN9A haplotype frequencies, between PNRR patients with painful or nonpainful Foundation for peripheral neuropathy. PNRR patient SCN9A and SCN11A missense variant allele frequen- Peripheral Neuropathy cies were not significantly different from the Exome Variant Server, European American and the Peripheral (EVS-EA) reference population. For SCN10A, there was a significant increase in the alternate Neuropathy Research allele frequency of the common variant p.V1073A and low-frequency variant pS509P in Registry (PNRR) PNRR patients compared with EVS-EA and the 1000 Genomes European reference Consortium populations. Conclusions: These results suggest that identification of a genetically defined subpopulation for

testing of NaV1.7 inhibitors in patients with peripheral neuropathy is unlikely and that additional Correspondence to Dr. Bristow: factors, beyond expression of previously reported disease “mutations,” are more important for the [email protected] development of painful neuropathy than previously discussed. Neurol Genet 2017;3:e207; doi: 10.1212/NXG.0000000000000207

GLOSSARY bp 5 ; CL 5 confidence limit; EUR 5 European; EVS-EA 5 Exome Variant Server, European American; FDR 5 false discovery rate; GOF 5 gain of function; HWE 5 Hardy-Weinberg; IEM 5 inherited erythromelalgia; MAF 5 minor allele frequency; NCV 5 nerve conduction velocity; NGS 5 next-generation sequencing; NIST 5 National Institute of Standards and Technology; OR 5 odds ratio; PNRR 5 Peripheral Neuropathy Research Registry; SFN 5 small fiber neuropathy; SNP 5 single nucleotide polymorphism.

Human genetics provides compelling evidence that the voltage-gated , NaV1.7, plays a critical role in pain. Gain-of-function (GOF) mutations in SCN9A, the gene encoding NaV1.7, cause inherited erythromelalgia (IEM) and paroxysmal extreme pain disorder, rare 1–3 Supplemental data familial diseases associated with excruciating pain. By contrast, loss-of-function SCN9A at Neurology.org/ng From the Department of Translational Biomarkers and Computational Genomics (S.W., S.P., R.G., S.K., J.T.), Bristol-Myers Squibb, Hopewell Site, Pennington, NJ; Department of Genetically Defined Diseases Discovery Biology (L.J.B.), Bristol-Myers Squibb, Wallingford, CT; Department of Innovative Clinical Development (I.Q., L.J.), Bristol-Myers Squibb, Lawrenceville, Princeton, NJ; Department of Neurology (S.A.-D.), Northwestern Medical Faculty Foundation, Chicago, IL; Department of Neurology (R.F.), Beth Israel Medical Center, Harvard School of Medicine, Boston, MA; Department of Neurology (D.M.S.), Icahn School of Medicine at Mount Sinai Medical Center, New York, NY; Department of Neurology (A.G.S.), University of Utah School of Medicine, Salt Lake City, UT; Department of Neurology (A.H.), Johns Hopkins University, Baltimore, MD. S.P. is currently affiliated with Biocon Bristol-Myers Squibb Research Center, Bangalore, India. Funding information and disclosures are provided at the end of the article. Go to Neurology.org/ng for full disclosure forms. The Article Processing Charge was funded by the Bristol-Myers Squibb. 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.

Neurology.org/ng Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 mutations cause congenital insensitivity to diagnosed with diabetic distal polyneuropathy (138 painful pain, a rare autosomal recessive disease char- and 41 nonpainful). The PNRR patient sample comprised 61% of males and 83% of Caucasian ethnicity, and patients 1,2 acterized by loss of pain sensation. These with painful peripheral neuropathy were significantly younger findings have triggered renewed efforts to (by on average 6–7 years) than those without pain (table e-1 at Neurology.org/ng). develop novel, selective NaV1.7 inhibitors for 4 the treatment of pain and expanded NaV Standard protocol approvals, registrations, and patient channel sequencing to patients with more consents. At each consortium site, institutional review board prevalent chronic pain conditions. In this re- approval was obtained based on a unified protocol developed as a consensus by the consortium members. Written informed con- gard, rare missense variants have been reported sent was obtained from all patients presenting to the neurology 5,6 6,7 8 in SCN9A, SCN10A, and SCN11A, clinics that agreed to enroll in the PNRR. genes encoding NaV1.7, NaV1.8, and Nav-targeted sequencing and variant calling. Agilent Sure- NaV1.9, respectively, in patients with painful Select hybrid capture probes were designed based on GRCh 37.3 small fiber neuropathy (SFN). It is thought to span entire 50 kb flanking, exonic, and intronic regions of SCN9A, SCN10A, and SCN11A totaling to 682,262 base pairs that enhanced NaV channel activity may directly contribute to the pain experienced (bps). National Institute of Standards and Technology (NIST) standard NA12878 and NA18507 HapMap cell line were by these patients because several of the mis- used as assay and informatics controls. PNRR patient DNA sam- sense variants identified exert GOF effects in ples were randomized for processing and across NGS runs and cell-based electrophysiology assays.9 plates using criteria of age, race, pain, and several clinical criteria including numbness, weakness, walk balance, diabetes, smoking, The identification of patients carrying path- and alcohol consumption. Libraries were prepared using a stan- ologic NaV missense variants provides an dard Agilent SureSelect protocol and were sequenced on an Illu- opportunity to select subjects whose pain mina HiSeq 2000 system with 100 bp-paired end reads to attain may be more effectively treated with novel a minimal 200X coverage. The average coverage obtained was 400X per base in the targeted regions, with 10%–20% of bases NaV agents. In this study, we collaborated missing coverage across all samples. The sequencing reads were with the Foundation for Peripheral Neuropa- mapped to the human genome build hg19 (GRCh 37),10 and the thy to obtain DNA samples from patients with variant calling was performed using Genome Analysis Toolkit (GATK) best practices (https://software.broadinstitute.org/gatk/ idiopathic or diabetic peripheral neuropathy best-practices).11–13 The variants were annotated using snpEff.14 enrolled in the Peripheral Neuropathy Performance of the variant calling pipeline was assessed using Research Registry (PNRR). The objective NIST NA12878, looking at sensitivity (TP/TP 1 FN) and was to identify missense variants in SCN9A, specificity (FP/FP 1 TN), where TP are variants called and present in NIST, FP are variants called but not present in NIST, SCN10A, and SCN11A and to examine their FN are variants not called but present in NIST, and TN are frequency in patients whose peripheral neu- variants not called and not present in NIST. A 93% sensitivity ropathy was painful or nonpainful. was observed, and positive predictive value and false discovery rate (FDR) were 51% and 49%, respectively. Variant calls were filtered to include variants with FILTER 5 “PASS”j“VQSR- TrancheSNP99.90to100.00”j“VQSRTrancheSNP99.00to99.90.” METHODS Patients. The study was performed at Bristol- Myers Squibb using whole blood DNA samples obtained Variant classification and allele frequency comparisons. from patients enrolled in the PNRR from 2011 to February PNRR variants were designated as common (minor allele fre- 2015. Clinical information and patient DNA samples were quency [MAF] . 5%) or low frequency (MAF ,5%) based provided by the Neurology Departments at John Hopkins on the alternate allele frequency reported in the NHLBI Exome University, Northwestern Medical Faculty Foundation, Beth Sequencing Project Exome Variant Server, European American Israel Medical Center-Harvard Medical School, and Icahn (EVS-EA) population.15 To compare PNRR variant allele fre- School of Medicine at Mount Sinai Medical Center. Con- quencies in patients with or without pain, or to reference pop- senting patients were evaluated using a comprehensive patient ulations (EVS-EA and 1000 Genomes Project16 global or examination form and patient health questionnaire and European (EUR) populations), the Fisher exact test was per- received peripheral nerve workup including nerve conduction formed using allele counts, and the Benjamini and Hochberg17 studies and, in some cases, analysis of skin biopsies. A copy of FDR was used to correct for multiple testing. All analyses were the patient health questionnaire is provided in the supple- performed in R, and results were reported as the Fisher P and Q mental information and included questions about sensory, values. motor, and autonomic symptoms and medication. Patients were instructed to answer pain-related questions from the SCN9A haplotype analysis. PNRR single nucleotide polymor- perspective of their neuropathy only, and those with addi- phisms (SNPs) with MAF .10% were used for haplotype analysis. tional complex medical issues or neurologic diseases were Genotypes were phased using BEAGLE V4.1.18 Haplotype block excluded from the registry. A DNA sample for next- definition and frequency comparison were performed in Haplo- generation sequencing (NGS) was obtained from 457 pa- view.19 TAGGER within the Haploview package was used to find tients; 278 patients diagnosed with idiopathic peripheral tag SNPs for each haplotype blocks to refine haplotype block def- neuropathy (186 painful and 92 nonpainful) and 179 patients initions. A x2 test was used to compare haplotype frequencies.

2 Neurology: Genetics Data and statistical analysis. Height, weight, and year of birth 13.6% for SCN9A, SCN10A, and SCN11A, respec- were analyzed by a 2-tailed t test to compare patients with dia- tively. Patients were heterozygous carriers of low- betic or idiopathic peripheral neuropathy with pain vs without frequency missense variants except for 5 subjects pain (GraphPad Prism v7). Statistical analysis methods for non- who were homozygous carriers of the SCN10A var- synonymous variant and haplotype comparisons are described 5 5 above. iants p.P1045T (n 1), p.S509P (n 2), or p.R14L (n 5 1) or the SCN11A variant p.T1609I (n 5 1).

RESULTS Incidence of genetic variants in the PNRR Comparison with reference populations. For each gene, cohort. NGS of SCN9A, SCN10A, and SCN11A the analysis included all common and low-frequency identified .1000 variants in each gene, with most nonsynonymous missense variants reported in PNRR (79%–87%) occurring in intronic regions (table e- patients, and the reference population being exam- 2). Most of the protein-coding variants we identi- ined. The reference populations and the number of fied were nonsynonymous, missense variants that missense variants included in each analysis were as changed the amino acid sequence (figure 1A). Details follows: EVS-EA: SCN9A (26), SCN10A (21), and of the gene location and alternate allele frequency for SCN11A (10); 1000 Genomes (1K) global and EUR: each nonsynonymous missense variant we identified SCN9A (23), SCN10A (21), and SCN11A (9). For are reported in the supplementary section (figure e-1 SCN10A, an increase in the alternate allele frequency and tables e-3 and e-4). Variants were designated as was observed for p.V1073A (odds ratio [OR] 5 low frequency (MAF ,5%) or common (MAF 1.2519; 95% confidence limits [CLs] 1.0851– .5%) based on the alternate allele frequency re- 1.4443; p 5 0.0021) and p.S509P (OR 5 1.9225; ported in the EVS-EA database. A low-frequency 95% CL 1.3994–2.6411; p 5 0.0001) compared missense variant in at least 1 gene was present in with EVS-EA (table 1, table e-5). Similar results were 47.7% (218/457) of patients with 9.8% carrying low- also seen in comparison to the 1K-EUR population frequency variants in multiple genes (figure 1B). For (p.V1073A: OR 5 1.5783; 95% CL 1.3066– individual genes, the number of patients carrying at 1.9064; p , 0.0001; p.S509P: OR 5 2.0098; 95% least 1 low-frequency variant was 24.5%, 21%, and CL 1.2432–3.2488; p 5 0.0044) (table 1, table e-5).

Figure 1 Protein-coding variants identified in SCN9A, SCN10A, and SCN11A

(A) Variants were classified based on changes in amino acid sequence. For each gene, results show the number (%) of variants identified in each class. (B)

Number (%) of patients in the Peripheral Neuropathy Research Registry (PNRR) cohort (n 5 457) that carry a low-frequency missense variant in NaV1.7,

NaV1.8, and/or NaV1.9. Variants identified in PNRR patients were designated as low frequency (minor allele frequency ,5%) based on the alternate allele frequency reported in the Exome Variant Server, European American reference population.

Neurology: Genetics 3 in the number of missense variants carried in each Table 1 Missense variants with significant Q values compared with reference populations gene or all 3 genes combined in patients with or without pain (figure e-2). Consistent with results for Gene Variant 1K global Q value 1K-EUR Q value EVS-EA Q value the complete PNRR cohort, the number of patients SCN9A p.D1908G 1.30E-13Y 0.0283[ 0.4525 carrying at least 1 low-frequency missense variant in

SCN9A p.N1245S 0.0194[ 1 0.6291 SCN9A, SCN10A,orSCN11A was 25%, 21%, and 13%, respectively, with most of the individual var- SCN9A p.V991L 1.33E-05Y 0.042[ 0.4525 iants or variant combinations present in only 1 SCN9A p.M932L 1.33E-05Y 0.042[ 0.4525 patient (tables e-11, e-12, and e-13). For low- SCN9A p.I739V 0.0364[ 0.9584 0.7226 frequency variants present in .2% of patients, the SCN9A p.P610T 1.33E-05[ 0.9584 0.4951 incidence was similar in painful and nonpainful SCN9A p.S490N 1.28E-05Y 0.1457 0.1548 conditions (figure 2, figures e-3 and e-4) and in pa-

SCN10A p.V1073A 1.32E-09Y 4.51E-05[ 0.0323[ tients with diabetic and idiopathic peripheral neu-

SCN10A p.P1045T 2.62E-05[ 0.9116 0.8329 ropathy (figures e-4, e-5, and e-6). Finally, we selected a subset of patients with idiopathic peripheral SCN10A p.I962V 0.0145[ 1 0.9170 neuropathy with probable, painful SFN using 2 ap- SCN10A Y [ [ p.S509P 4.26E-14 0.0464 0.0059 proaches: (1) patients with idiopathic peripheral SCN10A p.I206M 0.0004Y 0.9116 1 neuropathy with pain, autonomic symptoms but SCN10A p.R14L 0.0187[ 1 0.6513 without weakness based on the patient history ques-

SCN11A p.T1609I 3.65E-07Y 0.1167 0.0823 tionnaire (71 subjects) and (2) patients with idio-

SCN11A p.V909I 0.0001[ 0.5096 0.3559 pathic peripheral neuropathy with pain, normal nerve conduction velocity (NCV) but with abnormal skin SCN11A p.G481E 0.0239[ 0.8316 1 biopsies with length-dependent patterns (19 sub- Abbreviations: 1K 5 1000 Genomes; EUR 5 European; EVS-EA 5 Exome Variant Server, jects). The incidence of low-frequency missense var- European American. iants in these groups was similar to the complete Significant Q values are shown in bold text. The arrows indicate if the alternate allele – – frequency is increased ([) or decreased (Y) in the Peripheral Neuropathy Research Registry PNRR cohort (SCN9A 20% 26%, SCN10A 20% cohort compared with the reference population. 21%, and SCN11A 14%–21%; table e-14).

Incidence of previously reported low-frequency missense Several additional variants identified in SCN9A, variants. For SCN9A, 30/457 patients (6.6%) carried SCN10A, and SCN11A had alternate allele frequen- a low-frequency missense variant previously reported cies that were significantly different from the 1K in patients with painful SFN,5,6 painful diabetic global population, but few remained significant peripheral neuropathy,20 or (primary) IEM21,22 of compared with 1K-EUR, and none remained signif- which 14 (3.1%) carried GOF variant(s) based on icant compared with the EVS-EA population (table 1, cell-based electrophysiology results (p.I228M,23 p. tables e-5, e-6, and e-7). The alternate allele fre- G616R,21 p.I739V,24,25 and p.M932L/p.V991L5/p. quencies in reference populations for previously re- W1538R22; table e-15). For SCN10A and SCN11A, ported GOF variants in SCN9A, SCN10A, and previously reported variants6–8 were detected in 4 or 2 SCN11A are shown in tables e-3 and e-4 and were not patients, respectively, with only 2 (0.4%) or 1 (0.2%) significantly different in PNRR patients compared patient(s) carrying a known GOF variant (p. with the EVS-EA population. G1662S26 (SCN10A) or p.L1158P8 (SCN11A); table e-15). Although the incidence was low, collectively, Relationship to pain status in PNRR patients. We next these variants were detected in both patients with focused on examining the incidence of nonsynony- idiopathic and diabetic peripheral neuropathy and in mous missense variants in PNRR patients with pain- both painful and nonpainful conditions (table e-15). ful or nonpainful peripheral neuropathy. Evaluation of clinical information identified a small number of SCN9A haplotypes in PNRR patients. To determine subjects (17/457; 3.7%) with discrepancies between whether specific haplotypes are enriched in PNRR patient self-reported and physician-reported pain patients with painful vs nonpainful peripheral neu- which were excluded from further analysis. The re- ropathy, we performed haplotype analysis and com- sulting distribution of patients was as follows: 270 pared haplotype frequencies between both groups. idiopathic (183 painful and 87 nonpainful) and 170 Specifically, 196 common SCN9A SNPs (MAF diabetic (134 painful and 36 nonpainful). For all .10%) sequenced within the PNRR patient popula- missense variants identified in each gene, analysis of tion were selected and phased using BEAGLE V4.1.18 allele frequency showed no significant difference One hundred eighty-five SNPs passed the Hardy- between patients with or without pain (tables e-8, Weinberg (HWE) filter and were used to define 14 e-9, and e-10). We also saw no significant difference haplotype blocks in PNRR patients, with the

4 Neurology: Genetics Figure 2 Frequency of individual NaV1.7 missense variants in Peripheral Neuropathy Research Registry patients

Results show the % of all patients, % of patients with painful peripheral neuropathy, or % of patients with nonpainful peripheral neuropathy that carry each individual SCN9A low-frequency missense variant or variant combination identified. majority consisting of 2 most frequently observed that had increased frequencies in the PNRR patient haplotypes (see Methods). No significant differences population (p , 0.01; figure 4); however, they were in haplotype frequencies were seen between patients present in less than 5% of the individuals, suggesting with painful or nonpainful peripheral neuropathy that these haplotypes may be functionally (figure 3). Because haplotype block definitions are inconsequential. altered by the number of SNPs and may affect fre- quency estimates, we also used TAGGER to find tag DISCUSSION In a cohort of 457 patients diagnosed SNPs and redefine haplotype blocks. TAGGER with peripheral neuropathy, low-frequency missense identified 30 tag SNPs that were used to define 7 variants in SCN9A, SCN10A, and/or SCN11A were haplotype blocks. Again, we observed no significant common, with almost 50% of subjects carrying differences in haplotypes between patients with a variant in at least 1 gene. While the likelihood of painful or nonpainful peripheral neuropathy (figure carrying a low-frequency variant was high, most e-7). Finally, we examined whether specific hap- (;75%) of the individual variant/variant combina- lotypes were enriched in the PNRR cohort compared tions we identified were present in only 1 patient. with the 1000 Genomes, Utah residents with Furthermore, for each gene, missense variant allele Northern and Western European ancestry (1K-CEU) frequencies were not significantly different in patients population. This reference population was selected whose peripheral neuropathy was painful or non- because of the similarity in genetic background to the painful. Finally, SCN9A haplotype analysis showed PNRR patient sample and because it provides no significant differences in haplotype frequencies a greater number of intronic and exonic SNPs for based on pain status, indicating no enrichment of analysis than the EVS-EA population for which only specific haplotypes in patients with painful peripheral exome sequencing data are available. Phased geno- neuropathy. types for 180 of the 186 PNRR SNPs were obtained Human genetics provides compelling evidence from the 1K-CEU population data set of which 69 that NaV1.7 plays a critical role in pain. Furthermore, passed HWE filtering and were used to define 8 identification of NaV1.9 mutations in rare genetic haplotype blocks. In comparison to the 1K-CEU pain disorders27–29 and GOF missense variants in reference population, there were 2 minor haplotypes NaV1.7, NaV1.8, and NaV1.9 in patients with painful

Neurology: Genetics 5 Figure 3 NaV1.7 haplotypes in Peripheral Neuropathy Research Registry patients

Results show the frequency of individual haplotypes, within 14 different haplotype blocks, in Peripheral Neuropathy Research Registry patients with painful or nonpainful peripheral neuropathy.

5–8 SFN suggest that each NaV channel subtype has biopsies, a result similar to the incidence reported in the potential to contribute toward the pain experi- patients meeting strict clinical criteria for idiopathic enced by patients harboring mutations. To date, the pure SFN (29%).5 However, the incidence in our only study reporting sequencing results for all 3 genes subset of PNRR patients was not different from the in the same patient cohort showed that missense var- complete idiopathic group (24%), patients with dia- iants in SCN9A, SCN10A,orSCN11A were detected betes (26%), or patients with painful (25%) or non- in 8.7%, 3.8%, and 2.8% of patients with painful, painful (24%) peripheral neuropathy regardless of predominantly pure SFN, respectively.8 In compari- etiology, suggesting that SCN9A missense variants son, the incidence of low-frequency missense variants do not occur more frequently in patients with idio- in PNRR patients was ; 3–5-fold higher, and 10% pathic SFN. carried variants in more than 1 gene. It is possible that With respect to the individual missense variants differences in sequencing platforms, sequencing identified in PNRR patients, many were known depth, and approaches to variant calling/identifica- SNPs, but several novel variants were also detected, tion may explain these results. Of interest, we de- most notably in SCN11A (27% novel variants). Of tected SCN9A low-frequency variants in 26% of particular interest was the identification of patients PNRR patients with probable painful idiopathic carrying known GOF variants previously reported SFN based on symptoms, NCV, and abnormal skin in patients with SFN or patients with IEM.

6 Neurology: Genetics enrichment of GOF variants in patients with idio- Figure 4 Na 1.7 haplotype comparison with a reference population V pathic SFN, which was not apparent in PNRR peripheral neuropathy patients. It should be acknowl- edged that very few PNRR patients met criteria for probable idiopathic, pure SFN and that the biological impact of most of the missense variants we identified is unknown. Further investigation of patients with strictly defined pure idiopathic SFN is therefore important, given the potential attractiveness of this

clinical population for testing of novel NaV1.7 agents. To compare missense variant allele frequencies in PNRR patients with reference populations, we selected the EVS-EA population for our primary anal- ysis.15 The EVS database contains allele frequency information for exome sequencing variants identified in 4,300 Americans of European decent, a population closest in ancestry to the PNRR cohort.15 While no differences were noted for SCN9A or SCN11A, a sig- nificant increase in the alternate allele frequency of the SCN10A variants p.V1073A and p.S509P was seen in PNRR patients. While the impact of p. S509P is unknown, the p.V1073A common variant has a GOF effect, and expression of the A1073 pro- tein produces larger peak currents, slower fast inacti- vation, and larger persistent currents compared with V1073 in cell-based electrophysiology assays.30,31 Of interest, genetic association studies also report that rs6795970 (p.V1073A) is strongly associated with prolongation of the PR interval and QRS complex of the electrocardiogram,32 atrial fibrillation,31 and Brugada syndrome.30 It should be noted that the EVS-EA population represents a mixture of pheno- types including controls, specific cardiac and lung diseases, and specific traits (low-density lipoprotein and blood pressure). However, p.V1073A and p. S509P alternate allele frequencies were also higher in PNRR patients compared with the 1000 Genomes EUR population comprising 503 self-reported healthy individuals. These results suggest an associa- tion of p.V1073A and p.S509P with the peripheral neuropathy phenotype, although there was no asso- ciation with pain status within the PNRR patient group itself. Results show the frequency of individual haplotypes, within 8 different haplotype blocks, in Previous studies have also reported an association Peripheral Neuropathy Research Registry (PNRR) patients or the 1000 Genomes CEU pop- ulation. Results were analyzed by the chi-square test, *p , 0.01. of the SCN9A missense variants p.D1908G and p. V991L/p.M932L with neuropathic pain in patients Collectively, the incidence of carrying a known GOF with painful diabetic peripheral neuropathy.20 We missense variant in the PNRR cohort was very low, also detected these variants in PNRR patients; how- and they were present in some patients with a non- ever, we did not see any difference in allele frequen- painful phenotype. These results contrast with those cies compared with the EVS-EA reference reported in a small group of patients with strictly population. It should be noted that patients with defined idiopathic SFN where all 8 SCN9A missense diabetic painful distal polyneuropathy comprised on- variants identified conferred GOF effects in cell-based ly 30% of the PNRR population, while the previous electrophysiology assays.5 Thus, while the likelihood study focused on patients with diabetes enrolled in of carrying a low-frequency SCN9A missense variant clinical trials with strict inclusion criteria including was similar (see above), previous results suggest an pain intensity scores. In addition, the reference

Neurology: Genetics 7 population selected was important; we consider EVS- Foundation, the Daniel F. and Ada L. Rice Foundation, and Jack EA to be the most appropriate comparator; however, and Goldie Wolfe Miller.

a significant difference in the alternate allele fre- DISCLOSURE quency was observed when compared with the 1K- S. Wadhawan, S. Pant, R. Golhar, and S. Kirov have been employees of EUR population. Finally, with respect to the 1K Bristol-Myers Squibb. J. Thompson has been an employee of, has received global population, we consider the marked hetero- research support from, and owns stock/stock options in Bristol-Myers geneity in MAF across the different ethnic groups Squibb. L. Jacobsen and I. Qureshi have been employees of and own stock/stock options in Bristol-Myers Squibb. S. Ajroud-Driss has served comprising this population to be the likely explana- on the scientific advisory boards of MT Pharma and Alnylam; receives pub- tion for both the increased number of variants and lishing royalties from UpToDate; has served on the speaker’s bureau of MT discrepancies in the direction of the frequency Pharma; and has received research support from Alnylam Viromed and the Foundation for Peripheral Neuropathy. R. Freeman has served on the change observed when compared with results from scientific advisory boards of Abide, Acetylon, Astellas, Biogen, Daiichi the EVS-EA analysis. Sankyo, Dong, Grace, Grunenthal, Insys, Lundbeck, Nestle, Novartis, Results from this study showed marked allelic het- Pfizer, Vertex, and Spinifex; has served on the editorial boards of Autonomic erogeneity in SCN9A, SCN10A, and SCN11A in Neuroscience, Clinical Autonomic Research,andtheClinical Journal of Pain; receives publishing royalties from UpToDate; has received research support patients with idiopathic or diabetic distal polyneur- from Impeto, Nestle, Pfizer, the NIH, and Multiple System Atrophy Con- opathy. While NaV1.7 inhibitors have the potential sortium; and owns stock/stock options in Spinifex. D.M. Simpson has for the treatment of peripheral neuropathic pain, served on the scientific advisory boards of Allergan, Merz, Ipsen, and DSMB: Axsome; has received speaker honoraria from Allergan, Merz, identification of a specific, genetically defined sub- and Ipsen; has served on the editorial board of AIDS Patient Care;has population for drug testing in patients with peripheral received publishing royalties from Oxford University Press; has been an neuropathy appears unrealistic. Additional factors, employee of the Icahn School of Medicine at Mount Sinai; has been a consultant for, has served on the speakers’ bureaus of, and has received beyond expression of an NaV channel missense vari- research support from Allergan, Merz, Ipsen; has received research support ant, appear to be important because there was no the NINDS and the Foundation for Peripheral Neuropathy; and has been relationship between the presence of missense var- involved in legal proceedings regarding Proctor and Gamble. A.G. Smith iants and pain state in PNRR patients. This approach has served on the scientific advisory board of the Celgene Data Monitoring may have utility for the exploration of other disease- Committee; has served on the editorial board of NeuroLearn;hasbeen a consultant of Regenesis, Allergan, and Viromed; and has received research related genes and the identification of druggable support from Impeto Medical SAS and the NIH. A. Hoke has served on the molecular targets in patients with neuropathic pain. editorial board of Experimental Neurology and Annals of Clinical Transla- tional Neurology; has been an employee of Johns Hopkins University; has AUTHOR CONTRIBUTIONS been a consultant of Neurocrine Inc.; has received research support from the Dr. Wadhawan: drafting of the manuscript and analysis and interpreta- NIH, the Foundation for Peripheral Neuropathy, and the Dr. Miriam and tion of data. Dr. Pant: drafting of the manuscript, study design, and anal- Sheldon G. Adelson Medical Research Foundation; and was involved in the — ysis and interpretation of data. Dr. Golhar: analysis and interpretation of following legal proceedings: National Vaccine Injury Program gave expert — data. Dr. Kirov: drafting of the manuscript and analysis and interpreta- testimony on GBS and Arnold & Porter Kaye Scholer LLP gave expert tion of data. Dr. Thompson: drafting of the manuscript, study design, testimony on peripheral neuropathy. L.J. Bristow has been an employee of, and analysis and interpretation of data. Dr. Jacobsen: study concept has received research support from, and owns stock/stock options in Bristol- and design. Dr. Qureshi: analysis and interpretation of data. Myers Squibb. Go to Neurology.org/ng for full disclosure forms. Dr. Ajroud-Driss, Dr. Freeman, Dr. Simpson, Dr. Smith, and Dr. Hoke: drafting of the manuscript, study concept, and interpretation of data. Received July 25, 2107. Accepted in final form October 2, 2017. Dr. Bristow: drafting of the manuscript, study concept and design, and analysis and interpretation of data. REFERENCES 1. Vetter I, Deuis JR, Mueller A, et al. Na 1.7 as a pain ACKNOWLEDGMENT V target—from gene to pharmacology. Pharmacol Ther The authors thank the patients, their families, consortium members, and 2017;172:73–100. the Foundation of Peripheral Neuropathy for their support of the PNRR. They also thank the NHLBI GO Exome Sequencing Project and its ongo- 2. Dib-Hajj SD, Yang Y, Black JA, Waxman SG. The Na(V) ing studies, which produced and provided exome variant calls for compar- 1.7 sodium channel: from molecule to man. Nat Rev ison: the Lung Go Sequencing Project (HL-102923), the WHI Neurosci 2013;14:49–62. Sequencing Project (HL-102924), the Broad Go Sequencing Project 3. Tang Z, Chen Z, Tang B, Jiang H. Primary erythrome- (HL-102925), the Seattle Go Sequencing Project (HL-102926), and the lalgia: a review. Orphanet J Rare Dis 2015;10:127. Heart Go Sequencing Project (HL-103010). 4. Rivara M, Zuliani V. Novel sodium channel antagonists in the treatment of neuropathic pain. Expert Opin Investig STUDY FUNDING Drugs 2016;25:215–226. Bristol-Myers Squibb provided all funding to support patient DNA 5. Faber CG, Hoeijmakers JG, Ahn HS, et al. Gain of func- sequencing, data acquisition, data analysis, and preparation of the tion NaV1.7 mutations in idiopathic small fiber neuropa- manuscript by company employees. Patient DNA samples were thy. Ann Neurol 2012;71:26–39. acquired from subjects enrolled in the Peripheral Neuropathy Research 6. Brouwer BA, Merkies IS, Gerrits MM, Waxman SG, Registry (PNRR) established by the Foundation for Peripheral Neu- Hoeijmakers JG, Faber CG. Painful neuropathies: the ropathy (FPN). Funding for the development and maintenance of emerging role of sodium . J Peripher Nerv the PNRR was provided by the FPN. Major funding for patient enroll- – ment at the consortium sites and ongoing maintenance of the database Syst 2014;19:53 65. and repository at Indiana University came from the following sources 7. Faber CG, Lauria G, Merkies IS, et al. Gain-of-function from January 1, 2014, through December 31, 2016: Bristol-Myers NaV1.8 mutations in painful neuropathy. Proc Natl Acad Squibb, INSYS Development Company, the Jack Miller Family Sci USA 2012;109:19444–19449.

8 Neurology: Genetics 8. Huang J, Han C, Estacion M, et al. Gain-of-function 21. Choi JS, Cheng X, Foster E, et al. may mutations in sodium channel Na(v)1.9 in painful neurop- contribute to time-dependent manifestation of inherited athy. Brain 2014;137:1627–1642. erythromelalgia. Brain 2010;133:1823–1835. 9. Hoeijmakers JG, Faber CG, Merkies IS, Waxman SG. 22. Cregg R, Laguda B, Werdehausen R, et al. Novel muta- Painful peripheral neuropathy and sodium channel muta- tions mapping to the fourth sodium channel domain of – tions. Neurosci Lett 2015;596:51 59. NaV1.7 result in variable clinical manifestations of primary 10. Li H. Aligning sequence reads, clone sequences and assem- erythromelalgia. Neuromolecular Med 2013;15:265–278. bly contigs with BWA-MEM. 2013. arXiv:1303.3997v2 23. Estacion M, Han C, Choi JS, et al. Intra- and interfamily [q-bio.GN]. phenotypic diversity in pain syndromes associated with

11. McKenna A, Hanna M, Banks E, et al. The Genome a gain-of-function variant of NaV1.7. Mol Pain 2011;7:92. Analysis Toolkit: a MapReduce framework for analyzing 24. Han C, Hoeijmakers JG, Liu S, et al. Functional profiles next-generation DNA sequencing data. Genome Res of SCN9A variants in dorsal root ganglion neurons and 2010;20:1297–1303. superior cervical ganglion neurons correlate with auto- 12. DePristo MA, Banks E, Poplin R, et al. A framework for nomic symptoms in small fiber neuropathy. Brain 2012; variation discovery and genotyping using next-generation 135:2613–2628. – DNA sequencing data. Nat Genet 2011;43:491 498. 25. Han C, Hoeijmakers JG, Ahn HS, et al. NaV1.7-related 13. Van der Auwera GA, Carneiro MO, Hartl C, et al. From small fiber neuropathy: impaired slow-inactivation and FastQ data to high-confidence variant calls: the Genome DRG neuron hyperexcitability. Neurology 2012;78: Analysis Toolkit best practices pipeline. Curr Protoc Bio- 1635–1643. – informatics 2013;43:1 33. 26. Han C, Vasylyev D, Macala LJ, et al. The G1662S NaV1. 14. Cingolani P, Patel VM, Coon M, et al. Using 8 mutation in small fiber neuropathy: impaired inactiva- melanogaster as a model for genotoxic chemical mutational tion underlying DRG neuron hyperexcitability. J Neurol studies with a new program, SnpSift. Front Genet 2012;3:35. Neurosurg Psychiatry 2014;85:499–505. 15. Exome Variant Server, NHLBI GO exome sequencing 27. Woods CG, Babiker MO, Horrocks I, Tolmie J, Kurth I. project (ESP). Seattle, WA. Available at: http://evs.gs. The phenotype of congenital insensitivity to pain due to

washington.edu/EVS/. Accessed March, 2017. the NaV1.9 variant p.L811P. Eur J Hum Genet 2015;23: 16. 1000 Genomes Project Consortium, Auton A, Brooks LD, 561–563. Durbin RM, et al. A global reference for human genetic 28. Zhang XY, Wen J, Yang W, et al. Gain-of-function muta- variation. Nature 2015;526:68–74. tions in SCN11A cause familial episodic pain. Am J Hum 17. Benjamini Y, Hochberg Y. Controlling the false discovery Genet 2013;93:957–966. rate: a practical and powerful approach to multiple testing. 29. Leipold E, Liebmann L, Korenke GC, et al. A de novo J R Statist Soc B 1995;57:289–300. gain-of-function mutation in SCN11A cause loss of pain 18. Browning SR, Browning BL. Rapid and accurate haplo- perception. Nat Genet 2013;45:1399–1404. type phasing and missing-data inference for whole genome 30. Behr ER, Savio-Galimberti E, Barc J, et al. Role of com- association studies by use of localized haplotype clustering. mon and rare variants in SCN10A: results from the Bru- Am J Hum Genet 2007;81:1084–1097. gada syndrome QRS gene discovery collaborative 19. Barrett JC, Fry B, Maller J, Daly MJ. Haploview: analysis study. Cardiovasc Res 2015;106:520–529. and visualization of LD and haplotype maps. Bioinfor- 31. Jabbari J, Olesen MS, Yuan L, et al. Common and rare matics 2005;21:263–265. variants in SCN10A modulate the risk of atrial fibrillation. 20. Li QS, Cheng P, Favis R, Wickenden A, Romano G, Circ Cardiovasc Genet 2015;8:64–73. Wang H. SCN9A variants may be implicated in neuro- 32. Holm H, Gudbjartsson DF, Arnar DO, et al. Several pathic pain associated with diabetic peripheral neuropathy common variants modulate , PR interval and and pain severity. Clin J Pain 2015;31:976–982. QRS duration. Nat Genet 2010;42:117–122.

Neurology: Genetics 9 Copy number variation analysis increases the diagnostic yield in muscle diseases

Salla Välipakka, MSc ABSTRACT Marco Savarese, PhD Objective: Copy number variants (CNVs) were analyzed from next-generation sequencing data, Mridul Johari, MSc with the aim of improving diagnostic yield in skeletal muscle disorder cases. Lydia Sagath, MSc Methods: Four publicly available bioinformatic analytic tools were used to analyze CNVs from Meharji Arumilli, MSc sequencing data from patients with muscle diseases. The patients were previously analyzed with Kirsi Kiiski, PhD a targeted gene panel for single nucleotide variants and small insertions and deletions, without Amets Sáenz, PhD achieving final diagnosis. Variants detected by multiple CNV analysis tools were verified with Adolfo Lopez De Munain, either array comparative genomic hybridization or PCR. The clinical significance of the verified MD, PhD CNVs was interpreted, considering previously identified variants, segregation studies, and clini- Ana-Maria Cobo, PhD cal information of the patient cases. Katarina Pelin, PhD Bjarne Udd, MD, PhD Results: Combining analysis of all different mutation types enabled integration of results and iden- Peter Hackman, PhD tified the final cause of the disease in 9 myopathy cases. Complex effects like compound hetero- zygosity of different mutation types and compound disease arising from variants of different genes were unraveled. We identified the first large intragenic deletion of the (TTN)gene Correspondence to implicated in the pathogenesis of a severe form of myopathy. Our work also revealed a “double- S. Välipakka: trouble” effect in a patient carrying a single heterozygous insertion/deletion mutation in the TTN [email protected] gene and a Becker muscular dystrophy causing deletion in the dystrophin gene. Conclusions: Causative CNVs were identified proving that analysis of CNVs is essential for increasing the diagnostic yield in muscle diseases. Complex severe muscular dystrophy pheno- types can be the result of different mutation types but also of the compound effect of 2 different genetic diseases. Neurol Genet 2017;3:e204; doi: 10.1212/NXG.0000000000000204

GLOSSARY aCGH 5 array comparative genomic hybridization; BMD 5 Becker muscular dystrophy; CK 5 creatine kinase; CNV 5 copy number variant; indels 5 insertions and deletions; LGMD 5 limb-girdle muscular dystrophy; NGS 5 next-generation sequenc- ing; SNV 5 single nucleotide variant.

Next-generation sequencing (NGS) methods have become the most common method for the genetic diagnosis of genetically heterogeneous disorders.1,2 We have previously developed a tar- geted NGS gene panel, MyoCap.1 An updated version used here includes probes for the exons of nearly 300 myopathy genes and candidate genes. Similar platforms are currently in use in many laboratories.2 The reported diagnostic success rates are significantly higher than those obtained by traditional gene-by-gene sequencing.2 However, over 50% patients remain undiagnosed when only concentrating on single nucleotide variants (SNVs) and small insertions and deletions (indels).2 Copy number variants (CNVs) are defined as genomic deletions or duplications greater than 1 kb in size.3 CNVs cause microdeletion and microduplication syndromes, and they have also

From the Folkhälsan Institute of Genetics (S.V., M.S., M.J., L.S., M.A., K.K., K.P., B.U., P.H.), Medicum, Faculty of Biological and Environ- mental Sciences (K.P.), University of Helsinki, Finland; Neuromuscular Research Center (B.U.), Tampere University and University Hospital, Finland; Department of Neurology (B.U.), Vaasa Central Hospital, Finland; Biodonostia Health Research Institute (A.S., A.L.D.M), Neurosciences Area, CIBERNED, University of the Basque Country, San Sebastián, Spain; and Centre de Référence Maladies Neuromusculaires (GNMH) (A.-M.C.), Hôpital Marin APHP, Hendaye, France. Funding information and disclosures are provided at the end of the article. Go to Neurology.org/ng for full disclosure forms. 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.

Neurology.org/ng Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 been associated with several complex dis- CNV assessment from NGS data. CNVs were analyzed in eases.3,4 Generally, studies aiming for the iden- smaller batches from NGS data alignment files (.bam) obtained by analyzing DNA from 791 myopathy patients with MyoCap.1 tification of causative disease variants in We used 4 CNV analysis programs: Copy Number Inference skeletal muscle disorders have not systemati- From Exome Reads (CoNIFER) v0.2.2,6 eXome-Hidden Mar- cally used CNV screening. Multiplex ligation- kov Model (XHMM) v1.1,7 ExomeDepth v1.1.10,8 and COpy 9 dependent probe amplification has a lower number Detection by EXome sequencing (CODEX) v1.4.0, with the recommended default settings of each program. throughput when the amount of investigated A minimum of 1-bp overlap was used to determine whether calls genes increases.5 Array comparative genomic intersecting between different programs originated from the same hybridization (aCGH) has long been consid- CNV. In this article, we prioritize the specific cases with a very ered the only reliable and robust platform for high clinical interest, focusing on 7 variants detected by multiple programs and verified by using independent tools. CNV discovery.4 However, in NGS studies, the diagnostic evaluation may end with a dis- CNV validation. PCR was performed to confirm CNVs in pa- tients I, IIIa, IIIb, IV, V, VI, and VII. Primers were designed covery of a single pathogenic or likely patho- using Primer3 v4.0.0 (primer3.ut.ee) (table e-1, http://links. genic mutation before the utilization of lww.com/NXG/A12), and PCR was performed with DreamTaq complementary methods, which may lead to DNA Polymerase (Thermo Fisher Scientific, Waltham, MA) (figure e-1). A custom aCGH (manuscript in preparation), an underestimation of CNV contribution to investigating 187 of the genes included in MyoCap, was used to diseases. Recently, several CNV analysis tools confirm CNV detected in patient IIa. Segregation study in the for NGS data have been developed and are in family of patient IIa was performed by PCR (table e-1). use for routine diagnosis.3,4 Here, we describe RESULTS Table 1 shows the clinical features of pa- the detection of CNVs from NGS data with tients in this study, genetic findings, amount of pro- a combination of already available bioinfor- grams that detected the CNV, and the CNV matic tools. verification method. Patient I was identified to have a heterozygous METHODS Standard protocol approvals, registrations, FINmaj mutation, an 11-bp insertion/deletion in and patient consents. DNA samples of muscle disease patients the titin (TTN) gene.10 This variant causes dominant and healthy family members were obtained from clinicians in dif- tibial muscular dystrophy, characterized by a late age ferent countries. The study was approved by the Coordinating Ethics Committee of the Hospital District of Helsinki and Uusi- at onset, normal or slightly elevated creatine kinase maa. The samples were obtained according to the Helsinki decla- (CK) levels, and a mild distal phenotype. However, ration. Written informed consent was obtained from all patients. this patient has proximal weakness and a very high

Table 1 Patient and variant features

Previous genetic Detected by n ID Sex Onset First symptom Muscle weakness CK tests Genetic findings programs/verified with

I M Early adult Weakness Proximal LL 9x CAPN3a; FKRPa; TTN: c.107780_107790 3/PCR LMNAaT; RIM32a delinsTGAAAGAAAAA1; DMD: c.(643811_6439-1) _ (c.821711_c.8218-1)del

IIa; IIb M Infancy Weakness, exercise Proximal and distal LL 11x CAPN3a; DYSFa TTN: c.107889delA:p. 4/aCGH intolerance (Lys35963Asnfs*) 1 c. (785511_7856-1)_ (970311_9704-1)del

IIIa; IIIb F Infancy Weakness, walking Proximal LL, Gowers1.10x CAPN3a; ANO5a; WES SGCD: c.(?_-1) _(50511_506-1) 4/PCR difficulties del hom

IV M Infancy Walking difficulties Proximal . distal; 20x DMDb CAPN3: c.(309 1 1_310-1) _ 2/PCR Gowers1 (*1_?) hom

V M Early adult Walking difficulties Proximal LL 5–6x None DMD: c.(643811_6439-1) _ 3/PCR (c.821711_c.8218-1)del

VI M Early adult Walking difficulties Muscle atrophy; calf 10x None DMD: c.(643811_6439-1)_ 2/PCR hypertrophy (c.709811_c.7099-1)del

VII M Infancy Walking difficulties Proximal LL, Gowers1.10x None DMD: c.(592211_5923-1) _ 2/PCR (629011_6291-1)del

Abbreviations: aCGH 5 array comparative genomic hybridization; CK 5 creatine kinase; hom 5 homozygous; LL 5 lower limbs; NA 5 not available. Genetic findings are described using the reference transcripts: NM_001267550.1 for TTN, NM_004006.1 for DMD, NM_000231.2 for SGCD, and NM_000070.2 for CAPN3. a Sanger sequencing. b Multiplex ligation-dependent probe amplification.

2 Neurology: Genetics CK level. We excluded the presence of other delete- age of 40 years, and marked hyperCKemia. The rious variants in TTN. Surprisingly, the patient was patient had no further causative SNVs or indels in found to have a previously reported dystrophin the TTN gene. We identified a large deletion in TTN (DMD) deletion (exons 45–55) known to cause (exons 34–41; figure 1A) in trans with the Iberian Becker muscular dystrophy (BMD).11 frameshift variant in the proband IIa as well as in Patient IIa has a previously reported Iberian patient IIb, a similarly affected brother. Their healthy frameshift mutation, p.(Lys35963Asnfs*), in the last relatives are heterozygous for only one of the afore- exon of TTN usually determining 1 component of mentioned TTN mutations demonstrating the reces- the recessive distal titinopathy phenotype.10 Com- sive effect of the detected deletion. The severe distal pared with the other carriers of this variant, this and proximal titinopathies were thus caused by the patient has a more severe disease progression with compound heterozygosity of the Iberian frameshift proximal weakness, loss of ambulation before the and the deletion (figure 2).

Figure 1 Visualizations of copy number variants of patients IIa and IIIa

CoNIFER visualizations of a heterozygous deletion in the TTN in patient IIa (A) and a homozygous deletion in the SGCD in patient IIIa (B) accompanied with corresponding regions visualized with Integrative Genomics Viewer (IGV). In CoNIFER visualizations, the red line corresponds to the read depths for sample with deletion and black lines correspond to control samples. In the IGV visualization, the first row shows the sample with deletion and the second row shows a control sample without deletion.

Neurology: Genetics 3 can be unraveled. It has been suggested that the Figure 2 Pedigree of patients IIa and IIb number of patients with a combination of 2 or more genetic diseases is probably underestimated.12 The phenotypic complexity in these patients may erro- neously be interpreted as a new genetic disease with unidentified genetic defect or as a phenotypic expan- sion for a single disease.12 The proximal phenotype seen in patient I is mainly due to BMD, as the typical anterior lower-leg muscle lesions of the FINmaj TTN mutation may develop only after age 60. However, the identification of multilocus genomicvariantsiscrucialforapropergenetic counseling in the family. The combination of 4 analysis tools aided us to Segregation of the TTN Iberian tibial muscular dystrophy variant, p.(Lys35963Asnfs*), and identify already known and previously unknown deletion in the TTN (exons 34–41) in the family of patients IIa and IIb. The affected brothers IIa and IIb (black squares) are compound heterozygous for the variants. Their healthy parents CNVs. CNVs can explain some of the missing her- and siblings are heterozygous for only one of the variants. Genetic findings are described itability and undiagnosed cases in skeletal muscle using the reference transcript NM_001267550.1 for TTN. disorders. An NGS-based strategy for CNV detec- tion will be of great value for increasing the diagnos- Patients IIIa and IIIb, with a severe limb-girdle tic yield in patients affected by mendelian muscle muscular dystrophy (LGMD) phenotype, are the diseases. Current technology does not adequately daughters of first-cousin parents. Sanger sequencing capture repeat expansion diseases such as myotonic of candidate genes (CAPN3 and ANO5), MyoCap, dystrophy or diseases related to other repetitive ele- and whole-exome sequencing had been performed ments. Long-read sequencing technologies may fur- without identifying the causative variant. All the ther help the identification and mapping of CNVs as CNV detection programs identified a homozygous well as of repeat expansions. However, the clinical deletion in the SGCD gene (exons 1–5) (figure 1B). interpretation of CNVs remains challenging, in par- Patient IV is the child of consanguineous parents, ticular for CNVs identified in genes without previ- and he has been suffering from lower limb muscular ously reported disease causing deletions or weakness in the lower limbs since the age of 7. Dele- duplications. The inclusion of CNV data in public tions and duplications in the DMD gene had been databases, e.g., ExAC, could help in pathogenicity excluded. Immunochemistry showed normal staining assessment of CNVs. for dystrophin as well as for the sarcoglycans. A homo- AUTHOR CONTRIBUTIONS zygous deletion in CAPN3 (exons 2–8) was detected. Salla Välipakka: study concept and design, acquisition of data, analysis In 2 males with an LGMD-like phenotype (pa- and interpretation of data, and drafting the manuscript for intellectual tients V and VI), we found previously reported content. Marco Savarese and Mridul Johari: study concept and design, DMD deletions explaining the observed proximal mus- acquisition of data, analysis and interpretation of data, and revising the cular weakness. Both patients have in-frame deletions manuscript for intellectual content. Lydia Sagath, Meharji Arumilli, and Kirsi Kiiski: acquisition of data, analysis and interpretation of data, and – – 11 (exons 45 55 and exons 45 48) causing BMD. revising the manuscript for intellectual content. Amets Sáenz, Adolfo Patient VII with a Duchenne phenotype was also Lopez De Munain, and Ana-Maria Cobo: analysis and interpretation of included in our screening. As expected, a previously data. Katarina Pelin, Bjarne Udd, and Peter Hackman: study concept and design and revising the manuscript for intellectual content. reported out-of-frame DMD deletion (exons 42–43) 11 was identified. ACKNOWLEDGMENT The authors thank Merja Soininen and Helena Luque for their technical DISCUSSION Genomes are usually analyzed for help and Sini Penttilä, Tiina Suominen, and Sara Lehtinen for acquisition SNVs and indels, but studies of CNVs are often of samples. They also thank all the patients and family members as well as underrepresented. CNVs in muscle diseases have the clinicians who provided samples. previously been studied with complementary meth- STUDY FUNDING ods like targeted aCGH, which still remains the gold This study was supported by the Folkhälsan Research Foundation, the 4,5 standard technique for CNV detection. Here, we Jane and Aatos Erkko Foundation, the Academy of Finland (no. 138491, show that combining analysis of all different muta- B.U.), the Sigrid Jusélius Foundation, the Association Française contre les tion types enables integration of results and identi- Myopathies, and the Orion Research Foundation sr. fies the final cause of the disease in several cases. DISCLOSURE Complex effects like compound heterozygosity in S. Välipakka has received research support from the Folkhälsan Research patients IIa and IIb and compound genetic disease Foundation. M. Savarese has received research support from the Association arising from variants of different genes in patient I Française contre les Myopathies and Orion Research Foundation. M. Johari,

4 Neurology: Genetics L. Sagath, M. Arumilli, and K. Kiiski report no disclosures. A. Sáenz has copy-number mutations in 245 neuromuscular disorders. received research support from Health Research Fund (PI13-00722) of the Clin Chem 2011;57:1584–1596. Spanish Ministry of Economy and Competitiveness, the European Unión 6. Krumm N, Sudmant PH, Ko A, et al. Copy number (European Regional Development Fund). A. Lopez De Munain has received variation detection and genotyping from exome sequence travel funding from Sanofi. A. Cobo and K. Pelin report no disclosures. B. data. Genome Res 2012;22:1525–1532. Udd has served on the editorial board of Neuromuscular Disorders and has 7. Fromer M, Moran JL, Chambert K, et al. Discovery and received grants from Finska Läkaresällskapet (20,000 euros 2017), the Sigrid Juselius Foundation (general grant for the study of myopathies), the Jane and statistical genotyping of copy-number variation from Aatos Erkko Foundation, and the Vasa Central Hospital Research Founda- whole-exome sequencing depth. Am J Hum Genet tion (grant partly for myopathies). P. Hackman reports no disclosures. Go to 2012;91:597–607. Neurology.org/ng for full disclosure forms. 8. Plagnol V, Curtis J, Epstein M, et al. A robust model for read count data in exome sequencing experiments and Received August 22, 2017. Accepted in final form September 28, 2017. implications for copy number variant calling. Bioinfor- matics 2012;28:2747–2754. REFERENCES 9. Jiang Y, Oldridge DA, Diskin SJ, Zhang NR. CODEX: 1. Evila A, Arumilli M, Udd B, Hackman P. Targeted next- a normalization and copy number variation detection generation sequencing assay for detection of mutations in method for whole exome sequencing. Nucleic Acids Res primary myopathies. Neuromuscul Disord 2016;26:7–15. 2015;43:e39. 2. Nigro V, Savarese M. Next-generation sequencing ap- 10. Hackman P, Marchand S, Sarparanta J, et al. Truncating proaches for the diagnosis of skeletal muscle disorders. mutations in C-terminal titin may cause more severe tibial Curr Opin Neurol 2016;29:621–627. muscular dystrophy (TMD). Neuromuscul Disord 2008; 3. Pirooznia M, Goes FS, Zandi PP. Whole-genome CNV 18:922–928. analysis: advances in computational approaches. Front 11. Bladen CL, Salgado D, Monges S, et al. The TREAT- Genet 2015;6:138. NMD DMD global database: analysis of more than 4. Tan R, Wang Y, Kleinstein SE, et al. An evaluation of 7,000 duchenne muscular dystrophy mutations. Hum copy number variation detection tools from whole-exome Mutat 2015;36:395–402. sequencing data. Hum Mutat 2014;35:899–907. 12. Posey JE, Harel T, Liu P, et al. Resolution of disease 5. Piluso G, Dionisi M, Del Vecchio Blanco F, et al. Motor phenotypes resulting from multilocus genomic variation. chip: a comparative genomic hybridization microarray for N Engl J Med 2017;376:21–31.

Neurology: Genetics 5 Germline and somatic mutations in STXBP1 with diverse neurodevelopmental phenotypes

Mohammed Uddin, PhD ABSTRACT Marc Woodbury-Smith, Objective: To expand the clinical phenotype associated with STXBP1 gene mutations and to MD, PhD understand the effect of STXBP1 mutations in the pathogenesis of focal cortical dysplasia (FCD). Ada Chan, MSc Methods: Patients with STXBP1 mutations were identified in various ways: as part of a retro- Ledia Brunga, MD spective cohort study of epileptic encephalopathy; through clinical referrals of individuals Sylvia Lamoureux, BSc (10,619) with developmental delay (DD) for chromosomal microarray; and from a collection Giovanna Pellecchia, PhD of 5,205 individuals with autism spectrum disorder (ASD) examined by whole-genome Ryan K.C. Yuen, PhD sequencing. Muhammad Faheem, PhD Results: Seven patients with heterozygous de novo mutations affecting the coding region of Dimitri J. Stavropoulos, STXBP1 were newly identified. Three cases had radiologic evidence suggestive of FCD. One PhD male patient with early infantile epileptic encephalopathy, DD, and ASD achieved complete James Drake, seizure remission following resection of dysplastic brain tissue. Examination of excised brain MB, FRCSC tissue identified mosaicism for STXBP1, providing evidence for a somatic mechanism. Cell-type Cecil D. Hahn, MD expression analysis suggested neuron-specific expression. A comprehensive analysis of the Cynthia Hawkins, published data revealed that 3.1% of severe epilepsy cases carry a pathogenic de novo mutation MD, PhD within STXBP1. By contrast, ASD was rarely associated with mutations in this gene in our large Adam Shlien, PhD cohorts. Christian R. Marshall, Conclusions: STXBP1 mutations are an important cause of epilepsy and are also rarely associated PhD with ASD. In a case with histologically proven FCD, an STXBP1 somatic mutation was identified, Lesley A. Turner, MD suggesting a role in its etiology. Removing such tissue may be curative for STXBP1-related Berge A. Minassian, MD epilepsy. Neurol Genet 2017;3:e199; doi: 10.1212/NXG.0000000000000199 Stephen W. Scherer, PhD Cyrus Boelman, MB GLOSSARY ASD 5 autism spectrum disorder; CNV 5 copy number variant; DD 5 developmental delay; ddPCR 5 droplet digital PCR; EIEE 5 early infantile epileptic encephalopathy; FCD 5 focal cortical dysplasia; ID 5 intellectual disability; SNV 5 single nucleotide variant. Correspondence to Dr. Boelman: [email protected] or Dr. Uddin: Patients with STXBP1 gene mutations live with 3 key phenotypic abnormalities: intellectual [email protected] disability (ID), epilepsy, and movement disorders.1–4 Epilepsy is often drug resistant and most commonly presents as severe early infantile epileptic encephalopathy (EIEE; Ohtahara syn- drome) or West syndrome.4–6 However, STXBP1 mutations have been identified in patients with ID without epilepsy or with autism spectrum disorder (ASD).2,4,7 This broad STXBP1 encephalopathy phenotype is perhaps unsurprising, as STXBP1 is involved in the synaptic release of , with heterozygous mutations resulting in a reduction of both STXBP1’s protein product, -binding protein 1, and syntaxin-1.8 Moreover, homozy- gous deletions in knockout models are characterized by widespread brain .9 Although

From the Mohammed Bin Rashid University of Medicine and Health Sciences (M.U.), Dubai, UAE; The Centre for Applied Genomics (M.U., M.W.), Program in Genetics and Genome Biology (GGB) (A.C., L.B., S.L., G.P., R.K.C.Y., M.F., A.S., B.A.M., S.W.S.), Genome Diagnostics (D.J.S., C.R.M.), Paediatric Laboratory Medicine, Division of Neurosurgery (J.D., B.A.M.), and Division of Neurology (C.D.H.), The Hospital for Sick Children, Toronto, Ontario, Canada; Institute of Neuroscience (M.W.), Newcastle University, UK; Department of Molecular Genetics (A.C., S.W.S.), Department of Paediatrics (C.D.H., C.H.), and McLaughlin Centre (S.W.S.), University of Toronto, Ontario, Canada; Discipline of Genetics (L.A.T.), Faculty of Medicine, Memorial University of Newfoundland, St. John’s, Canada; and Division of Neurology (C.B.), BC Children’s Hospital, Vancouver, Canada. Funding information and disclosures are provided at the end of the article. Go to Neurology.org/ng for full disclosure forms. The Article Processing Charge was funded by the Mohammed Bin Rashid University of Medicine and Health Sciences. This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloading and sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.

Neurology.org/ng Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 there is evidence for a role in neurite growth,10 decision makers, and the study was approved by The Hospital for the clinical pathophysiology remains unclear Sick Children Research Ethics Board, Toronto, Canada. and sometimes spontaneously remitting epi- DNA extraction from resected brain tissue. A portion of lepsy suggests developmental mechanisms stock brain tissue (from patient 1) kept on dry ice was removed, added directly to cell lysis solution, and then homogenized using beyond neurotransmission. a sterile probe and an Omni tissue homogenizer. The homoge- Radiologic evidence of gross brain pathol- nized tissue was digested overnight with Proteinase K at 55°C. ogy with heterozygous STXBP1 mutations is This was repeated for the remaining visible tissue fragments for variable.4,11 When present, it suggests cerebral a further 5 hours at 55°C to ensure complete digestion. The tissue digest was subsequently spun down, and the supernatant was atrophy, abnormal myelination, or hypogene- removed for extraction. High-salt Gentra protein precipitation sis of the corpus callosum.12 STXBP1 germline solution was next added, and the sample was vortexed for 20 mutation was reported only once in focal cor- seconds and then put on ice for 5 minutes for complete pre- cipitation. The sample was spun at 2,000g for 10 minutes, and all tical dysplasia (FCD) but never a somatic the supernatant was removed down to the protein pellet, leaving 13 mutation. Molecular genetic studies of a thin layer above the pellet to prevent protein contamination. FCD to date has focused on MTOR pathway The supernatant was then transferred into isopropanol and in- mutations.14 We present a case series with verted gently to mix 50 times to precipitate the DNA, which is then washed with 70% ethanol. This was then centrifuged and STXBP1 encephalopathy including a case with the ethanol was removed. Tris-EDTA buffer was next added to a radiologic diagnosis of FCD that went on to the dried DNA pellet for rehydration. After heating the DNA to have successful epilepsy surgery. 65°C for 1 hour, it was maintained on a rotator at room tem- perature overnight. DNA was then heated at 55°C for 1 hour and Our aim was twofold: first, to investigate subsequently quantitated with a fluorometer using ds dye the histopathology and molecular biology of Hoescht. the excised brain tissue from an STXBP1-pos- Immunohistochemistry. For immunohistochemical studies on itive FCD patient; and second, to characterize the resected brain tissue slice, hematoxylin and eosin staining was the phenotypic spectrum of STXBP1 through applied following a standard protocol, Olig2 staining against our case series and meta-analysis. oligodendroglia-like cells and antineuronal nuclear antigen NeuN (Cat# MAB377, Millipore, Temeuza, CA) against neu- rons with a dilution ratio of 1:50 were used following established METHODS Patients. Patient 1 was part of a retrospective protocols.22 cohort study of patients with epileptic encephalopathy recruited using a standard protocol and with patient consent at The Hospital Droplet digital PCR. We applied droplet digital PCR forSickChildreninTorontobetweenJanuary2012andJune (ddPCR) technology to validate CNVs and to quantify the exact 2014.15 All patients underwent a series of investigations as part of copy number. Copy number estimation of STXBP1 was per- their clinical workup, including chromosomal microarray and tar- formed using the QX200 ddPCR system (Bio-Rad Laboratories, geted next-generation sequencing using epilepsy panels for between Inc., Hercules, CA) using the TaqMan copy number probe 35 and 70 known epilepsy genes as previously described.15 Addi- Hs00269332_cn (Life Technologies, Carlsbad, CA). Prior to tional patients with STXBP1 deletions were identified through the copy number experiment, 250 ng of genomic DNA was m Toronto’s Hospital for Sick Children, having been referred for digested with 4U of BtgI in a 10- L reaction buffer (New 3 clinical microarray and/or exome sequencing from a variety of England Biolabs, Ipswich, MA), 1 hour 37°C incubation, m clinical services in connection with concerns regarding epilepsy and/ and no enzyme denaturation. The 20- L copy number reaction m 3 or developmental delay (DD) (10,619 microarrays).16 In addition, mix consisted of 10 Lof2 ddPCR supermix for probes m an ASD cohort, forming part of the MSSNG project (mss.ng), was (Bio-Rad Laboratories), 1 L of the copy number target assay m searched for STXBP1 mutations. This project aims to sequence (labeled with FAM), 1 L of the copy number reference assay a minimum of 10,000 families and obtain detailed phenotypes. (RNAseP, Life Technologies, part 4403326, labeled with VIC), m m m These families have been recruited from across Canada and/or are in 6 L of water, and 2 L of 25 ng/ L digested genomic DNA. the Autism Genetic Resource Exchange.17,18 Molecular and com- The copy number assay was validated by temperature gradient to putational methods have been previously described,19,20 and lollipops ensure optimal separation of target and RNAseP-containing software21 was used to map mutations. droplets. Cycling conditions for the reaction were 95°C for For meta-analysis, we have identified 162 cases with STXBP1 10 minutes, followed by 45 cycles of 94°C for 30 seconds and mutations (single nucleotide variants [SNVs] or copy number 60°C for 1 minute, 98°C for 10 minutes and finally a 4°C hold variants [CNVs]) from numerous independent scientific reports on a Life Technologies Veriti thermal cycler. Data were analyzed (tables e-1 at http://links.lww.com/NXG/A14, and e-2 at with QuantaSoft v1.4 (Bio-Rad Laboratories). Two reference http://links.lww.com/NXG/A15). For meta-analysis, we have DNA samples (NA10843 and HuRef) plus 3 nontemplate con- used PubMed to review all of the described STXBP1 scientific trols were included with the study samples. reports in the literature that were published up to September 2016. A total of 162 patients with heterozygous mutations of RESULTS We have identified 7 new patients with STXBP1 have been described so far across 35 studies. heterozygous de novo mutations affecting the coding Standard protocol approvals, registrations, and patient region of STXBP1. Patient 1 was diagnosed with consents. All patients described in this report have provided infantile spasms and global DD at age 3.5 months a written consent from the individuals involved or their substitute (table). Further evaluation at age 5 years identified

2 Neurology: Genetics Table Clinical characteristics of patients with de novo STXBP1 mutations

Chromosome Other Epilepsy coordinates Speech Motor neurologic (seizure type Anticonvulsant Phenotype Sex (hg19) Mutation type delay delay ID ASD features [s]) EEG response MRI

Patient 1 M 130415943- Deletion (exon 3-4) 1 NA 11 2 1(Infantile Bilateral Seizure-free High FLAIR 130420935 spasms; focal anterior postepilepsy signal in dyscognitive temporal surgery; anterior left seizures with epileptiform previously superior and secondary discharges refractory to mesial generalization) levetiracetam, temporal clobazam, and gyri, vigabatrin suspicious of focal cortical dysplasia (figure 1)

Patient 2 F 130427611 Splice-site G.T 1112Bilateral 1 (Unclear Diffuse slowing; NA Normal (c.66311) intention types; focal and tremor and possible generalized upper limb epileptic epileptiform dystonia spasms) activity

Patient 3 F 130430438 Missense C.T 1112Head and 1 (Tonic- Multiple Refractory to Incomplete (p.Arg292Cys) hand tremor clonic from independent valproic acid, myelination 3 mo; focal spike foci with clonazepam, of anterior clonic) left temporal- and temporal parietal focus at phenobarbital poles; 3 mo; diffuse possible left right anterior hemisphere ictal temporal focus at 6.5 y FCD

Patient 4 M 130435529 Nonsense C.T 11122 1(Infantile Hypsarrhythmia Seizure-free on Left anterior (p.Arg367*) spasms) at 5 mo; left vigabatrin temporal posterior lobe: temporal then asymmetric bifrontal myelination epileptiform or focal discharges by cortical 2y dysplasia

Patient 5 F 130286582- Deletion 1112Tremor; 1 (Brief Bilateral Partial then Normal 130806948 ataxia; generalized independent complete attention- tonic seizures anterior response to deficit in infancy and temporal, phenobarbital in hyperactivity rare nocturnal central– infancy and no disorder; and generalized temporal or anticonvulsants obsessive tonic seizures posterior epilep- after compulsive at age 6 y) tiform dis- disorder charges in early childhood; and normal at age 8y

Patient 6 F 130444715- Frameshift 1 NA 11 2 1(Absence; NA NA NA 130444717 (p.Lys526Asnfs*23) tonic-clonic)

Patient 7 M 130428536 Missense T.C 1211Fluctuating 1 (Tonic- Multiple Refractory to Normal (p.Met252Thr) gait clonic seizures independent valproic acid, abnormality from 1 y; spike foci at age topiramate, tonic; epileptic 1 y, normalized phenobarbital, spasms) then re-emerged and prednisone; at age 2.5 y; and partial response left frontal and to ketogenic right temporal diet independent ictal foci

Abbreviations: 25absent; 15present; ASD 5 autism spectrum disorder; FCD 5 focal cortical dysplasia; FLAIR 5 fluid-attenuated inversion recovery; ID 5 intellectual disability; NA 5 information unavailable.

social and communication impairments and behav- temporal, hippocampal, and parahippocampa1 gyri ioral features consistent with a diagnosis of ASD. At (figure 1A). Genetic evaluation using clinical micro- this time, he continued to present with DD, and his array and a commercial 70-gene next-generation frequent focal dyscognitive seizures (paroxysmal onset sequencing epilepsy panel (GeneDx, Gaithersburg, of altered awareness, gazing around with an open MD; table e-3 at http://links.lww.com/NXG/A14) mouth, and ending with a cough with or without identified a heterozygous partial deletion of exons 3–4 emesis), with secondary generalization, were resistant in STXBP1. Examination of parental genotypes to pharmacologic anticonvulsant treatments. EEG confirmed the de novo nature of this deletion. No demonstrated bilateral frontotemporal spikes and other clinically relevant genetic abnormality was waves, and MRI suggested FCD in the left medial identified on sequencing or microarray. In view of the temporal lobe, specifically affecting the superior intractable nature of his seizures, patient 1

Neurology: Genetics 3 tissue. The quantification (below 1 copy; figure 2E) Figure 1 MRI of patient 1 with fluid-attenuated inversion recovery signal was confirmed on 7 independent replications, con- sistent with cells mosaic for heterozygous and homo- zygous STXBP1 mutations in the dysplastic tissue. To characterize the cellular impact of mosaicism, we analyzed STXBP1 transcriptome profiles from 466 single cells24 for 6 major brain cells—astrocytes, mi- croglia, neurons, endothelial, oligodendrocytes, and oligodendrocyte precursor cells. STXBP1 encodes syntaxin-binding protein that plays a role in the release of through the regulation of syntaxin. Our analysis confirmed the principally neuronal expression of STXBP1 (figure 2F) com- pared with other primary brain cell types. This neuron-specific expression pattern and the known pathway of syntaxin imply a functional relevance of this gene into the etiology of epileptogenicity. We identified 4 additional cases with heterozygous STXBP1 mutations (patients 2, 3, and 4) or deletion (patient 5) through clinics and 10,619 samples referred for chromosomal microarray at The Hospital for Sick Children. These patients were identified with a de novo mutation or deletion effecting one or more exons of STXBP1 (table). Two (patients 2 and 5) of our reported de novo mutations are novel and 2 (pa- (A) Axial MRI of patient 1 demonstrating diffuse cortical/subcortical high fluid-attenuated inversion recovery (FLAIR) signal confined to the anterior left superior temporal gyrus and tients 3 and 4) were found to be recurrent mutations mesial temporal structures (white boxes). (B–C) Subtle increased FLAIR signal in the anterior reported in previous literature. All presented with temporal lobes of patients 2 and 3, respectively, suggestive of possible cortical dysplasia. speech delay, global DD/ID, and epilepsy. Other more variable clinical features included motor delay subsequently underwent lesionectomy of the left and upper limb tremor (patients 2 and 3). Patients 3 anterior temporal horn at the age of 6 years, with and 4 also had MRI evidence of possible FCD (figure complete remission of seizures achieved following 1, B and C). The 7-year-old female (patient 3) pre- resection and continuing as of the last follow-up 2 sented at 3 months of age with multiple seizure types years later (both Engel and International League (including myoclonic, atonic, tonic–clonic, and focal Against Epilepsy outcome Class 1). His ID and ASD dyscognitive seizures) and DD, with MRI demon- remain symptomatically unchanged after surgery. strating asymmetric increased T2 signal in the white Resected dysplastic tissue and marginal nondysplas- matter of the anterior temporal poles, consistent with tic tissue were histologically examined. While normal either delayed myelination or possibly left anterior histology was evident in most of the imaging- temporal FCD. Although interictal EEG was abnor- suggested dysplastic tissue, the superior temporal gyrus mal bilaterally, ictal EEG demonstrated seizures showed heterotopic neurons in the white matter, blur- arising from the contralateral right hemisphere. The ring the gray–white matter junction, and in the deeper 4-year-old male (patient 4) was similarly suspected to temporal white matter, there were nodular heterotopias have delayed myelination or FCD at the left anterior comprising dysplastic tissue representing collections of temporal pole based on the increased T2 signal in the cells with small round nuclei and clear cytoplasm, white matter. His interictal EEGs did not demon- possibly immature neurons (figure 2, A–D). Immu- strate epileptiform activity. nohistochemistry labeling with both NeuN and Olig2 In view of patient 1’s diagnosis of ASD and the was negative in the latter. These findings were most indication in the literature of several patients with consistent with FCD-type Ib.23 To further evaluate the “autistic features” or ASD diagnoses, we next investi- molecular basis of patient 1 with epilepsy and ID, we gated our ASD cohort for the evidence of STXBP1 undertook ddPCR on resected left lateral temporal cor- mutations. Of the currently sequenced individuals tical tissue and marginally resected normal brain tissue, (5,205), 2 were identified with novel de novo variants hypothesizing that dysplastic tissue may be character- overlapping STXBP1 (table; patients 6 and 7). A ized by mosaicism for STXBP1.Aspredicted,copy female individual (patient 6) has a de novo frameshift number quantification was consistent with haploinsuf- mutation truncating the protein’s domain 2. Patient 7 ficiency in marginal tissue but mosaicism in dysplastic is a male individual with a de novo missense mutation

4 Neurology: Genetics Figure 2 Histopathology of surgically excised dysplastic brain tissue (patient 1)

Circled areas of the left temporal lobe juxtacortical demonstrate linear neuronal heterotopia on (A) hematoxylin and eosin (H&E) and (B) NeuN stains, highlighting neurons; superior temporal gyrus nodular heterotopia are seen in boxes with (C) NeuN and (D) H&E stains, highlighting likely immature neurons; (E) Copy number variant mosaicism quantification using (ddPCR) within focal cortical dysplasia (red) and normal (black) tissue. The standard error bar shows the inferred copy number of approximately 20,000 droplets generated from each tissue type by the ddPCR; (F) single expression profile of the STXBP1 gene in primary brain cell types. RNA-seq reads were obtained (GEO GSE67835) for STXBP1 from 6 primary brain cell types (astrocytes, endothelial, microglia, neurons, oligodendrocytes, and oligodendrocyte pro- genitor cells [OPCs]). The bar showing lower 25th and upper 75th percentiles of the data, and the dots represent the outliers. Each color represents each cell types. and primarily diagnosed with ASD with severe epi- We subsequently reviewed all of the described lepsy. Both individuals were diagnosed with ASD using STXBP1 reports in the literature (up to September the ADI-R in conjunction with the ADOS. Both are 2016). There have been a total of 162 patients with nonverbal, with severe ASD in association with ID. heterozygous mutations of STXBP1 described so far Many of these individuals have also been investigated across 35 studies (table e-1). For each mutation, we for the presence of CNVs by microarray.17 have converted the mutations into hg19 track and

Neurology: Genetics 5 inferred the corresponding protein changes. The muta- of de novo variants, we excluded all case reports and tion spectrum comprises different variants spread across retained only those studies that analyzed a cohort all 3 domains of the gene, with 21 recurrent mutations (with at least 20 samples); all 19 studies were ascer- in unrelated individuals occurring in domains 2 and 3 tained by way of a primary diagnosis if epilepsy. From (figure 3A). Mutations include 121 SNVs and 15 in- these 19 studies (table e-1), a total of 2,569 cases were dels (of size ,100 bp), and a total of 26 cases were analyzed, among whom 80 were reported to have reported to have a CNV affecting STXBP1. Sex was a variant within STXBP1. More than 92.5% (74 of available for 126 cases, and we observed a higher pro- 80) carried a de novo variant, with the remainder of portion of females reported to carry an STXBP1 muta- unknown inheritance. We can compute, therefore, tion compared with males, 57.1% (72) and 42.8% a rate of 3.1% for de novo mutations within STXBP1 (54), respectively (figure 3B). Ascertainment for these for cases with epilepsy. We realize that this de novo 162 cases included 80% with a form of epilepsy syn- rate is high due to the selection of samples that drome (described as EIEE, Ohtahara syndrome, West are clinically well defined for epilepsy-related syndrome, and Dravet syndrome) and 20% with one or conditions. more of the following: ASD, DD, movement disorder, nonsyndromic ID, and atypical Rett syndrome (figure DISCUSSION Through a review of over 18,393 cases 3C). Information pertaining to comorbidity for these of ASD and DD, with molecular and/or cytogenetic 162 cases is not available. analysis and meta-analysis of the literature, we A total of 140 variants (of 162) were reported to be report the prevalence of STXBP1 gene mutations in de novo including case reports. To determine the rate epilepsy as well as present an interesting finding of

Figure 3 Mapping of mutations within the STXBP1 gene

(A) Mutations mapped within the 3 protein domains (cyan— domain 1, green—domain 2, and red—domain 3a/b) of the STXBP1 gene; (B) percentage of male and female cases reported to carry a pathogenic mutation for the STXBP1 gene. (C) Clinical spectrum associated with de novo STXBP1 mutations reported for early infantile epileptic encephalopathy (EIEE), focal cortical dysplasia (FCD), Ohtahara syndrome (OS), West syndrome (WS), Dravet syndrome, infantile spasms (ISs), neonatal refractory seizures, autism spectrum disorder, developmental delay, nonsyndromic intellectual disabilities (IDs), and atypical Rett syndrome.

6 Neurology: Genetics somatic STXBP1 mutation associated with FCD. Our case, even if present, we would not be able to tell detailed molecular analysis of excised brain tissue whether those were in the same subclonal population from patient 1 has revealed the observation of the as the dysplasia-specific STXBP1 mosaicism. While postzygotic STXBP1 mutation in excised focally somatic mutations have received much attention in dysplastic tissue, with sustained remission of the cancer and certain nonbrain disorders, a possible etio- epilepsy following resective surgery, which highlights pathologic connection with neurodevelopmental dis- the clinical importance of this finding. Only one orders, such as epilepsy and ASD, requires further other patient has been described in the literature with elucidation. Our own study provides some impetus a similar pattern.13 This other patient, with a germline to further explore this possibility. heterozygous deletion, presented clinically with ID From a clinical genetics perspective, our study has and epilepsy. Despite having a normal MRI, she calculated that 3.1% of clinically identified epilepsy underwent surgical management of her seizures with cases harbor a rare/de novo mutation within STXBP1. near-complete remission achieved. FCD was dem- Furthermore, our wider sampling of separate very onstrated in resected tissue within the temporal lobe, large, “classic autism” ASD and DD cohorts allows although detailed molecular biology of the resected us to conclude that STXBP1 seems to be associated tissue, as we have newly done here, is not reported. only with severe ASD and DD in the presence of Patient 1’s exemplary outcome following surgery is epilepsy. We have shown that the epilepsy observed important and supports the relevance of a previous with STXBP1 encephalopathy may be controlled study with the observation of a good outcome.13 through surgery by identifying dysplastic tissue, Treatment-resistant epilepsy significantly decreases which is associated with further mosaic changes in the quality of life and increases the risk of sudden STXBP1 limited to dysplasia. These rare observations unexpected death in an epilepsy patient.25 Our obser- help define this disorder’s phenotype and introduce vation suggests that presurgical evaluations should be interesting genotypic information underlying FCD. aggressively pursued in patients with STXBP1 when there is evidence of dominant focality and not limited AUTHOR CONTRIBUTIONS out of concern for the presence of a germline mutation. M.U. and C.B. contributed to the conception and design of the study. A.C., L.B., S.L., G.P., R.K.C.Y., M.F., and A.S. helped analyze data. Although only about 50% of cases described so far with D.J.S, J.D, C.D.H., C.H., and L.A.T. contributed to sample recruitment, heterozygous deletions of STXBP1 have MRI evidence clinical data acquisition, and analysis. M.W.-S., C.R.M, B.M., S.W.S, of brain pathology, milder type II and type I classified M.U., and C.B. contributed to drafting the text and preparing the figures. cortical dysplasias may be undiagnosed. Certainly, the ACKNOWLEDGMENT limitation of current imaging sensitivity, even 3T MRI, The authors thank The Centre for Applied Genomics (TCAG), which is funded 13 is evidenced by a previous study and the abnormal by Genome Canada and the Ontario Genomics Institute, the Canada Founda- myelination vs FCD differential reported in 2 of our tion for Innovation (CFI), and the Ontario Research Fund of the Government patients who did not have surgery. of Ontario. The project was also supported by funds from the University of Although more extensive studies are required to Toronto McLaughlin Centre and Genome Canada. M.U. holds a Banting post- doctoral fellowship from the Canadian Institutes of Health Research (CIHR). fully elucidate the genetics of FCD, the case described in this study provides a new lead. It remains possible STUDY FUNDING that the FCD was an incidental finding in our re- No targeted funding reported. ported case, but it does raise the possibility of a causal relationship between neuronal expression of mosaic DISCLOSURE STXBP1 mutations and FCD. This association is sup- M. Uddin has received travel funding from a nonprofit entity; was a Bant- ported by the homozygous STXBP1 deletion being ing postdoctoral fellow from the Canadian Institutes of Health Research only identified in the dysplastic portions of the re- (CIHR); and has applied for a US patent for classifying mutation for autism and neurodevelopmental disorders. M. Woodbury-Smith is sup- sected brain, and at low levels, perhaps just in neu- ported by a Clinical Investigatorship Award from the Canadian Institutes rons. Recent studies reported somatic mutations in of Health Research’s (CIHR) Institute of Genetics and has served on the MTOR in excised dysplastic brain tissue in 12 of 77 editorial board of the Journal of Autism and Developmental Disorders.A. patients with epilepsy.26,27 The implication is that Chan has received research support from the Government of Ontario/ Ontario Graduate Scholarship and the Government of Canada/Frederick somatic mutations may arise in brain tissue during Banting and Charles Best Scholarship. L. Brunga, S. Lamoureux, development by an as-yet undetermined mecha- G. Pellecchia, R. Yuen, M. Faheem, D.J. Stavropoulos, and J. Drake nism.28 As has been presented and discussed at length report no disclosures. C. Hahn has served on the scientific advisory boards of SAGE Therapeutics and Marinus Pharmaceuticals; has served in a previous study, even low levels of mosaicism in on the editorial boards of the Journal of Clinical Neurophysiology and PLoS MTOR pathway genes may result in significant clin- One; is the second vice president of the American Clinical Neurophysi- ical pathology.14 The MTOR gene was not sequenced ology Society and the President of the Canadian Society of Child Neu- in the 70-gene panel of patient 1 (table e-2), and there rology; and has received research support from The Hospital for Sick Children Foundation and the American Epilepsy Society and Epilepsy was insufficient dysplastic tissue to do this separately. Foundation of America. C. Hawkins has served on the scientific advisory With regard to potential MTOR mutations in our board of the Pediatric Brain Tumor Consortium (nonprofit entity); has

Neurology: Genetics 7 served on the editorial boards of Acta Neuropathologica Communications in a human iPSC model for STXBP1-related epileptic and the Journal of Neuropathology and Experimental Neurology; holds a patent encephalopathy. Epilepsia 2016;57:e81–e86. for BOC modulation and uses thereof for treating, preventing and/or diag- 11. Di Meglio C, Lesca G, Villeneuve N, et al. Epileptic nosing cancer; and has received research support from Bayer, Grants4Tar- patients with de novo STXBP1 mutations: key clinical gets, the Canadian Institutes of Health Research, Medical Sciences Proof of features based on 24 cases. Epilepsia 2015;56:1931–1940. Principle (MsC POP) Program, Genome Canada, the Canadian Cancer Society Research Institute, DIPG collaborative, and Curing Kids Cancer. 12. Barcia G, Chemaly N, Gobin S, et al. Early epileptic A. Shlien has received research support from the Terry Fox Foundation. encephalopathies associated with STXBP1 mutations: C.R. Marshall has received travel funding from Affymetrix and Life Tech- could we better delineate the phenotype? Eur J Med Genet nologies and has served on the editorial boards of Genes, Genomes, Genetics 2014;57:15–20. and the Journal of Neurodevelopmental Disorders. L.A. Turner has received 13. Weckhuysen S, Holmgren P, Hendrickx R, et al. Reduc- research support from Shire and CIHR. B. Minassian holds patents for tion of seizure frequency after epilepsy surgery in a patient diagnostic testing of the following genes: EPM2A, EPM2B, MECP2, and with STXBP1 encephalopathy and clinical description of VMA21; has received research support from NIH; and has received license six novel mutation carriers. Epilepsia 2013;54:e74–e80. fee payments/royalty payments from patents for diagnostic testing of the 14. Mirzaa GM, Campbell CD, Solovieff N, et al. Association following genes: EPM2A, EPM2B, MECP2, and VMA21. S.W. Scherer holds the GlaxoSmithKline-CIHR Chair in Genome Sciences at the Uni- of MTOR mutations with developmental brain disorders, versity of Toronto and The Hospital for Sick Children; has served on the including megalencephaly, focal cortical dysplasia, and pig- scientific advisory board of Population Diagnostics; has served on the edi- mentary mosaicism. JAMA Neurol 2016;73:836–845. torial boards of Genomic Medicine, Genes, Genomes, Genetics,theJournal of 15. Mercimek-Mahmutoglu S, Patel J, Cordeiro D, et al. Personalized Medicine,theOpen Genomics Journal,theHugo Journal, Diagnostic yield of genetic testing in epileptic encephalop- Genome Medicine,theJournal of Neurodevelopmental Disorders, Autism athy in childhood. Epilepsia 2015;56:707–716. Research, PathoGenetics, Comparative and Functional Genomics, BMC Med- 16. Uddin M, Pellecchia G, Thiruvahindrapuram B, et al. In- ical Genomics,andCytogenetics and Genome Research;andhasreceived dexing effects of copy number variation on genes involved research support from Genome Canada/Ontario Genomics Institute, Cana- in developmental delay. Scientific Rep 2016;6:28663. dian Institutes of Health Research, Canadian Institute for Advanced Research, McLaughlin Centre, Canada Foundation for Innovation, govern- 17. Pinto D, Delaby E, Merico D, et al. Convergence of genes ment of Ontario, and NIH. C. Boelman reports no disclosures. Go to and cellular pathways dysregulated in autism spectrum Neurology.org/ng for full disclosure forms. disorders. Am J Hum Genet 2014;94:677–694. 18. Uddin M, Tammimies K, Pellecchia G, et al. Brain- expressed exons under purifying selection are enriched Received April 26, 2017. Accepted in final form August 30, 2017. for de novo mutations in autism spectrum disorder. Nat Genet 2014;46:742–747. REFERENCES 19. Yuen RK, Merico D, Cao H, et al. Genome-wide charac- 1. Gburek-Augustat J, Beck-Woedl S, Tzschach A, Bauer P, teristics of de novo mutations in autism. NPJ Genomic Schoening M, Riess A. Epilepsy is not a mandatory feature Med 2016;1:160271–1602710. of STXBP1 associated ataxia-tremor-retardation syn- 20. Yuen RK, Thiruvahindrapuram B, Merico D, et al. drome. Eur J Paediatr Neurol 2016;20:661–665. Whole-genome sequencing of quartet families with autism 2. Hamdan FF, Piton A, Gauthier J, et al. De novo STXBP1 spectrum disorder. Nat Med 2015;21:185–191. mutations in mental retardation and nonsyndromic epi- 21. Jay JJ, Brouwer C. Lollipops in the clinic: information lepsy. Ann Neurol 2009;65:748–753. 3. Mignot C, Moutard ML, Trouillard O, et al. STXBP1- dense mutation plots for precision medicine. PLoS One related encephalopathy presenting as infantile spasms and 2016;11:e0160519. generalized tremor in three patients. Epilepsia 2011;52: 22. Krzisch M, Temprana SG, Mongiat LA, et al. Pre-existing 1820–1827. astrocytes form functional perisynaptic processes on neu- 4. Stamberger H, Nikanorova M, Willemsen MH, et al. rons generated in the adult . Brain Struct – STXBP1 encephalopathy: a neurodevelopmental disorder Funct 2015;220:2027 2042. including epilepsy. Neurology 2016;86:954–962. 23. Blumcke I, Thom M, Aronica E, et al. The clinicopatho- 5. Otsuka M, Oguni H, Liang JS, et al. STXBP1 mutations logic spectrum of focal cortical dysplasias: a consensus clas- cause not only Ohtahara syndrome but also West sification proposed by an ad hoc Task Force of the ILAE syndrome–result of Japanese cohort study. Epilepsia Diagnostic Methods Commission. Epilepsia 2011;52: – 2010;51:2449–2452. 158 174. 6. Saitsu H, Kato M, Mizuguchi T, et al. De novo mutations in 24. Darmanis S, Sloan SA, Zhang Y, et al. A survey of human the gene encoding STXBP1 (MUNC18-1) cause early infan- brain transcriptome diversity at the single cell level. Proc – tile epileptic encephalopathy. Nat Genet 2008;40:782–788. Natl Acad Sci USA 2015;112:7285 7290. 7. Milh M, Villeneuve N, Chouchane M, et al. Epileptic and 25. Hall-Patch L, Gayatri N, Ferrie C, Morrall MC. Clinical nonepileptic features in patients with early onset epileptic reasoning and investigations for resective surgery for med- encephalopathy and STXBP1 mutations. Epilepsia 2011; ically refractory epilepsies in children. Epilepsy Behav 52:1828–1834. 2012;23:442–446. 8. Patzke C, Han Y, Covy J, et al. Analysis of conditional 26. Lim JS, Kim WI, Kang HC, et al. Brain somatic mutations heterozygous STXBP1 mutations in human neurons. in MTOR cause focal cortical dysplasia type II leading to J Clin Invest 2015;125:3560–3571. intractable epilepsy. Nat Med 2015;21:395–400. 9. Grone BP, Marchese M, Hamling KR, et al. Epilepsy, 27. Lim ET, Uddin M, De Rubeis S, et al. Rates, distribution behavioral abnormalities, and physiological comorbidities and implications of postzygotic mosaic mutations in autism in syntaxin-binding protein 1 (STXBP1) mutant zebrafish. spectrum disorder. Nat Neurosci 2017;20:1217–1224. PLoS One 2016;11:e0151148. 28. Poduri A, Evrony GD, Cai X, Walsh CA. Somatic muta- 10. Yamashita S, Chiyonobu T, Yoshida M, et al. Mislocaliza- tion, genomic variation, and neurological disease. Science tion of syntaxin-1 and impaired neurite growth observed 2013;341:1237758.

8 Neurology: Genetics Clinical heterogeneity and phenotype/ genotype findings in 5 families with GYG1 deficiency

Rabah Ben Yaou, MD* ABSTRACT * Aurélie Hubert, PhD Objective: To describe the variability of muscle symptoms in patients carrying mutations in the Isabelle Nelson, PhD GYG1 gene, encoding glycogenin-1, an enzyme involved in the biosynthesis of glycogen, and to Julia R. Dahlqvist, MD discuss genotype-phenotype relations. David Gaist, MD, PhD Methods: We describe 9 patients from 5 families in whom muscle biopsies showed vacuoles with Nathalie Streichenberger, an abnormal accumulation of glycogen in muscle fibers, partially a-amylase resistant suggesting MD polyglucosan bodies. The patients had either progressive early-onset limb-girdle weakness or Maud Beuvin, BS late-onset distal or scapuloperoneal muscle affection as shown by muscle imaging. No clear Martin Krahn, MD, PhD definite cardiac disease was found. Histologic and protein analysis investigations were performed Philippe Petiot, MD on muscle. Frédéric Parisot, BS Fabrice Michel, MD Results: Genetic analyses by direct or exome sequencing of the GYG1 gene revealed 6 different Edoardo Malfatti, MD, GYG1 mutations. Four of the mutations were novel. They were compound heterozygous in 3 PhD families and homozygous in 2. Protein analysis revealed either the absence of glycogenin-1 or Norma Romero, MD reduced glycogenin-1 expression with impaired glucosylation. Robert Yves Carlier, MD, Conclusions: Our report extends the genetic and clinical spectrum of glycogenin-1–related PhD myopathies to include scapuloperoneal and distal affection with glycogen accumulation. Neurol Bruno Eymard, MD, Genet 2017;3:e208; doi: 10.1212/NXG.0000000000000208 PhD Philippe Labrune, MD, GLOSSARY PhD FSHD 5 facioscapulohumeral dystrophy; mRNA 5 messenger RNA; PAS 5 periodic acid–Schiff. Morten Duno, PhD Thomas Krag, PhD Mutations in genes involved in the muscular glycogen metabolism cause glycogen storage myo- Mathieu Cerino, PharmD pathies and cardiomyopathies.1,2 Defects in enzymes that control muscle glycogen breakdown Marc Bartoli, PhD are more common than diseases involved in glycogen synthesis. Glycogenin is a glycosyltransfer- Gisèle Bonne, PhD ase that catalyzes the formation of an 8–12 glucose molecule chain at its tyrosine 195 residue.3 John Vissing, MD, PhD* Two types of glycogenin have been identified in humans: glyocogenin-1 and glycogenin-2 Pascal Laforet, MD, encoded by GYG1 and GYG2 genes, respectively.4,5 Glycogenin-1 (37 kDa) is expressed pref- * PhD erentially in skeletal muscle and to a lesser extent in the liver, whereas glycogenin-2 (66 kDa) is François M. Petit, abundant in cardiac muscle and the liver but sparse in the and absent in skeletal muscle. PharmD, PhD*

*These authors contributed equally to the manuscript. Correspondence to From the Assistance Publique-Hôpitaux de Paris (APHP) (R.B.Y., B.E., P.L.), Groupe Hospitalier Universitaire la Pitié-Salpêtrière, Centre de Dr. Petit: Référence de Pathologie Neuromusculaire Paris-Est, Institut de Myologie; Sorbonne Universités (R.B.Y., I.N., M.B., G.B.), UPMC Université Paris [email protected] 06, INSERM UMRS 974, Center of Research in Myology, Institut de Myologie, Paris; Department of Molecular Genetics (A.H., F.P., F.M.P.), Assistance Publique-Hôpitaux de Paris (APHP), Hôpitaux Universitaires Paris Sud, Antoine Béclère Hospital, Clamart; Department of Pediatrics (A.H., P.L.), Assistance Publique-Hôpitaux de Paris (APHP), Hôpitaux Universitaires Paris Sud, Antoine Béclère Hospital, Referal Center for Metabolic Liver Inherited Diseases, Clamart, and Université Paris-Sud, France; Department of Neurology (J.R.D., T.K., J.V.), Copenhagen Neuromuscular Center, Rigshospitalet, University of Copenhagen; Department of Neurology (D.G.), Odense University Hospital, Denmark; Hospices Civils de Lyon (N.S.), Groupement Hospitalier Est, Centre de Pathologie et Neuropathologie Est, Service de Neuropathologie, Université Claude-Bernard Lyon 1, CNRS UMR5239, LBMC, ENS; AP-HM, Département de Génétique Médicale (M.K., M.C., M.B.), Hôpital Timone Enfants; Aix Marseille Université (M.K., M.C., M.B.), Inserm, GMGF, UMR_S910; Hospices Civils de Lyon (P.P.), Hôpital la Croix-Rousse, Département de Neurologie, Université Claude Bernard Lyon 1; CHU Besançon (F.M.), Hôpital Jean Minjoz, Centre de Référence de Pathologie Neuromusculaire; Unité de morphologie neuromusculaire (E.M., N.R.), Groupe Hospitalier Universitaire La Pitié-Salpêtrière; Sorbonne Uni- versités UPMC Université Paris 06; Assistance Publique-Hôpitaux de Paris (APHP) (R.Y.C.), Service d’imagerie Médicale, CIC-IT Handicap, Hôpital Raymond Poincaré, Garches, France; and Department of Clinical Genetics (M.D.), Rigshospitalet, University of Copenhagen, Denmark. Funding information and disclosures are provided at the end of the article. Go to Neurology.org/ng for full disclosure forms. 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.

Neurology.org/ng Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 2

Table Summary of the clinical and imaging features of the 9 patients

First clinical assessment Muscle imaging Whole phenotype/ First CPK Pulmonary procedure Muscle biopsy: symptoms of onset/ Age, y/affected Age, y/affected level (UI/ ECG/ functional Disease course (no. of GYG1 muscle/age (y)/striking Patient age, y muscles in UL muscles in LL L)/EMG echocardiography tests at last assessment assessments) mutations pathologic features

A-1 (M) Scapuloperoneal/ 32/asymmetric 32/distal weakness (TA, 1,509/ Nl/Nl Nl at 42 y Del, BH, ECR, ECU, CT (3) and (c.646C.T)1 First biopsy: Del/33/ running and climbing scapular weakness PL, and EDL) . proximal myopathic old, mild and GMa weakness, WB-MRI (1) (c.14313G.C) vacuolar myopathy stair difficulties, calf atrophy (right Del and weakness (GMe) and FVC no UL elevation since containing PAS-positive and thigh cramps, and SSp) and no scapular stepping and waddling decrease 48 y old, used inclusions, fascicular erlg:Genetics Neurology: knee drop/28–29 winging gait (by 20%) a walking frame since atrophy, mild necrosis, and since 49 y 49 y old, and outside endomysial fibrosis; EM: old wheelchair use since presence of normally 50 y old branched glycogen and no fibrillary forms

Second biopsy: Del/36/ vacuolar myopathy containing PAS-positive inclusions and partially a-amylase resistant; EM: coexistence of normal granular glycogen with abnormal filamentous glycogen and abnormal dilated mitochondria

Third biopsy: ECR/48/ vacuolar myopathy containing PAS-positive inclusions, only 1 fiber was a-amylase resistant

B-1 (M) Proximal/muscle 27/forearm muscle 27/pelvifemoral 700– Nl/Nl Nl Del, SSp, SSc, and IS CT (1) and (c.14313G.C)1 Del/26/vacuolar myopathy fatigability at running hypertrophy, mild weakness (Ps, GMa, Adds 1,100/ mild weakness at 28 y WB-MRI (1) (c.14313G.C) containing a-amylase– without myalgia/16– scapular winging, and . GMe, and Ham), thigh myopathic old resistant PAS-positive in- 17 no weakness and gluteal atrophy, calf clusions; EM: polyglucosan hypertrophy, and bodies waddling gait

B-2 (M) Proximal/muscle 27/mild scapular 27/pelvifemoral 1,016/ND Nl/Nl Nl Del, SSp, SSc, and IS WB-MRI (1) (c.14313G.C)1 ND fatigability at running winging and no weakness (GMe . Ps and mild weakness at 28 y (c.14313G.C) without myalgia/15 weakness Ham), thigh and gluteal old atrophy, calf hypertrophy, and waddling gait

B-3 (M) Asymptomatic/no 23/no muscle weakness 23/no muscle weakness Nl/ND ND/ND ND but no ND CT (1) (c.14313G.C)1 ND symptoms and atrophy or and atrophy or respiratory (c.14313G.C) hypertrophy hypertrophy symptoms

C-1 (F) Distal/right then left 58/distal weakness (ED, 58/distal weakness (TA), Nl/ AVB-II at 58 y old, Nl, OSAS at PL and EDL weakness WB-MRI (1) (c.14313G.C)1 First biopsy: TA/51/ stepping gait/47–49 EDM, ECR, and ECU), lower leg atrophy, and myopathic overt coronary 70 y old at 73 y old and MRI (3) (c.996_1005del10) vacuolar myopathy hand atrophy stepping gait artery disease requiring (interosseus, thenar, since 70 y old NnIV and hypothenar requiring stenting, muscles) acetylsalicylic acid and atorvastatin

Continued Table Continued

First clinical assessment Muscle imaging Whole phenotype/ First CPK Pulmonary procedure Muscle biopsy: symptoms of onset/ Age, y/affected Age, y/affected level (UI/ ECG/ functional Disease course (no. of GYG1 muscle/age (y)/striking Patient age, y muscles in UL muscles in LL L)/EMG echocardiography tests at last assessment assessments) mutations pathologic features

Second biopsy: ECR/59/ vacuolar myopathy, subsarcolemmal PAS- positive material, type I fiber predominance, mild necrosis, and moderate endomysial fibrosis; EM: glycogen accumulation and abnormal mitochondria

C-2 (M) Distal/left stepping 72/distal weakness (ED 72/distal weakness (TA, Nl/ Nl Nl Difficulties in index CT (1) and (c.14313G.C)1 First biopsy: TA/74/ gait/72 (but older and EDM), atrophy (ED, PL, and EDL) more than myopathic finger extension (left WB-MRI (1) (c.996_1005del10) advanced fibrofatty difficulties in starting EDM, right Tr, and right pelvifemoral weakness then right) and degeneration and efforts during Pec) (GMa, GMe, Ps, and inability to stand from myofibrillar myopathy gymnastics, better hamstrings), thigh and squatting position at after 1 minute of lower leg atrophy, 77 y old effort or alleviated by Achilles contractures, taking sugar) and stepping gait

Second biopsy: Del/77/ vacuolar myopathy containing a-amylase– resistant PAS-positive inclusions

C-3 (F) Distal/ankle twists, 81/no weakness or 81/distal weakness (TA, ND/ND Myocardial Nl Shortness of breath ND (c.14313G.C)1 ND frequent falls, atrophy and left UL PL, and EDL) . infarction at 76 y on exertion (c.996_1005del10) difficulties in climbing elevation limitation pelvifemoral weakness old requiring stairs and getting up (musculotendinous cuff (GMa, GMe, Ps, and Quad) stenting from squatting, and surgery) and inability to walk on UL weakness/79 heels and tiptoes

D-1 (F) Proximal/problems in 76/asymmetric 76/asymmetric 252/ Nl/Nl Nl Cannot walk without WB-MRI (1) (c.166G.C)1 Vastus lateralis/76/ climbing stairs/50 appearance, upper right appearance and right LLs myopathic support, (c.472delA) abnormal glycogen limbs more affected more affected than the hyperlordotic, accumulation (polyglucosan than left, marked left waddling gait, and bodies) atrophy of biceps uses rollator brachii on the right (2/ 5), mild on the left (4/5), and clear facial muscle weakness erlg:Genetics Neurology: E-1 (F) Proximal/slower in 45/pronounced anterior Normal muscle strength 183/ ND/ND ND Hyperlordotic and MRI (1) (c.487delG)1 Vastus lateralis/45/ running and cycling thigh compartment myopathic walk without help (c.487delG) myopathic with evidence of compared with her atrophy, hip flexion, abnormal glycogen peers/teen abduction, and accumulation (polyglucosan adduction and knee bodies) extension weakness to 4onMRC

Abbreviations: Add 5 thigh adductor; AVB 5 atrioventricular block; BH 5 biceps humeri; CPK 5 creatine phosphokinase; Del 5 deltoid; ECR 5 extensor carpi radialis; ECU 5 extensor carpi ulnaris; ED 5 extensor digitorum; EDL 5 extensor digitorum longus; EDM 5 extensor digiti minimi; EM 5 electronic microscopy; FVC 5 forced vital capacity; GMa 5 gluteus maximus; GMe 5 gluteus medius; Ham 5 hamstring muscle; IS 5 infraspinatus; LL 5 lower limb; MRC 5 Medical Research Council Scale for muscle power; ND 5 not done; Nl 5 normal; NnIV 5 nocturnal noninvasive ventilation; OSAS 5 obstructive sleep apnea syndrome; PAS 5 periodic acid–Schiff; Pec 5 pectoralis; PL 5 peroneus longus; Ps 5 psoas; Quad 5 quadriceps; SSc 5 subscapularis; SSp 5 supraspinatus; TA 5 tibialis anterior; Tr 5 trapezius; UL 5 upper limb; WB-MRI 5 whole- body MRI. 3 In the past 6 years, 20 patients of various Western blot. Expression of glycogenin-1 in myoblast cell ethnic backgrounds have been reported with culture from patient A-1 and frozen muscle tissue from patients B-1, C-1, and C-2 were explored by Western blot with glycogen storage myopathy due to GYG1 gene monoclonal antibodies against human glycogenin-1 N-terminal 6–12 mutations (GSDXV, OMIM #613507). 1:500 (Abnova, Taipei, Taiwan) with or without a-amylase Patients with GSDXV carry mutations leading treatment as previously described.7 Forthispurpose,samples m a to inactivation of the autoglucosylation of were treated with 10 g/mL human -amylase (Sigma-Aldrich, St. Louis, MO) for 1 hour at 37°C to remove sugar residues of glycogenin-1 or impairment of glycogenin-1 the large glycogen molecules. Protein extracts were separated – interaction with glycogen synthase.6 8,12,13 In onto 10% Mini-PROTEAN TGX precast gels (Bio-Rad Lab- reported cases of glycogenin deficiency, skele- oratories, Marnes-la-Coquette, France). Immunoblots were tal muscle involvement was always that of visualized by Immobilon Western Chemiluminescent HRP Substrate (Millipore, Guyancourt, France) on a G-Box system a limb-girdle myopathy, accompanied in some using GeneSnap software (Ozyme, Montigny-le-Bretonneux, instances by an additional distal involve- France). The glucosylated form of glycogenin-1 differs in size ment.6–13 Cardiomyopathy in patients with by approximately 1 kDa compared with the a-amylase–treated control. GSDXV is rare,6,14 but mutations in the GYG1 gene can be observed with no skeletal RESULTS Clinical and muscle imaging features. Clini- muscle signs.14 cally, the 9 patients in this study can be categorized We report here a detailed clinical and mus- into 3 subgroups according to the pattern of muscle cle imaging assessment as well as protein and affection: proximal, distal, and scapuloperoneal. The genetic features of 9 patients with GSDXV. proximal subgroup included patients B-1, B-2, and Remarkably, the clinical spectrum uncovers E-1 who showed a teenage onset of pelvifemoral (B- not only the previously reported proximal 1, B-2, and E-1) and scapular weaknesses (B-1 and B-2) (table and figure 1). This muscle affection was myopathy but also new muscle patterns, thus confirmed by muscle imaging, showing partial and extending the clinical conditions due to symmetric fatty replacement of deltoid muscles in GYG1 gene mutations. upper limbs and a more advanced involvement of gluteus magnus and medius, vastus lateralis and in- METHODS Five men and 4 women from 5 families originat- termedius, adductor magnus, and biceps femoris ing from France (families A, B, and C) and Denmark (families muscles in lower limbs (figures e-1E, e-1E9, e-1F, D and E) were included in this study (table and figure 1). Clin- e-1F9, e-1G, e-1G9, e-1O, and e-1O9, links.lww. ical, histologic, and muscle imaging data were taken from the patients’ medical records. Patient B-1 was reported as patient com/NXG/A13). P1 in a previous publication.7 During their follow-up, muscle Conversely, patients C-1, C-2, and C-3 had a pre- imaging was performed in all patients, except patient C-3: muscle dominant distal myopathy pattern, manifesting late CT in 4 patients, muscle MRI in 2, and whole-body MRI in 6 in the fifth–eighth decade of life by asymmetric (figure e-1, links.lww.com/NXG/A13). stepping gait (table). Unfortunately, muscle imaging Standard protocol approvals, registrations, and patients was performed only in patient C-1 at disease onset consents. Written informed consent was obtained from all pa- and showed a highly specific extensor tibialis anterior tients before muscle biopsies and DNA collection. muscle involvement (figure e-1I, links.lww.com/ Muscle pathology analyses. At least 1 muscle biopsy was NXG/A13). No earlier muscle imaging was available performed in 6 patients (3 in A-1, 2 in C-2, and 1 in B-1, for patients C-1 and C-2, and the oldest sister C-3 C-1, D-1, and E-1) in deltoid in 4 patients, extensor carpi radi- refused medical investigations. At more advanced alis in 2, tibialis anterior in 2, and vastus lateralis in 2. Morpho- stages, the disease affected extensor digitorum and logic and histochemical analyses of fresh-frozen muscle tissue medial gastrocnemius muscles (figures e-1I, e-1J, were performed by standard techniques, including periodic 9 9 acid–Schiff (PAS) staining before and after digestion with e-1K , e-1L, and e-1M ) but also progressed toward 9 9 a-amylase15 (figure 2). proximal muscles (figures e-1J, e-1J , e-1K, e-1K , e-1L, e-1M, and e-1M9). Molecular genetic analysis. After DNA extraction from Another unexpected presentation was observed peripheral leukocytes, coding regions of GYG1 were sequenced (DNA Blood Mini kit, Qiagen, Hilden, Germany) and compared in patient A-1, who had a scapuloperoneal distribu- with the reference sequence (accession number NM_004130). tion in the first stages of the disease. For this patient, Total messenger RNAs (mRNAs) were extracted from a fresh muscle imaging was performed 4 times over the muscle biopsy of patient B-1 (RNA Mini kit, Qiagen, Hilden, course of 20 years, allowing us to precisely track Germany) and amplified by reverse transcription PCR (Promega, muscle involvement distribution and progression. Fitchburg, WI). Exome sequencing was performed in family C at In the first 7–8 years of the disease onset, the pattern the French National Centre for Genotyping and in family D at the Broad Institute’s Genomics Platform. Pathogenic mutations was clearly peroneal in lower limbs, involving tibialis were confirmed by Sanger sequencing. anterior and extensor digitorum longus muscles and

4 Neurology: Genetics Figure 1 Pedigrees of the 5 analysed families

The photographs of patients A-1, B-1, B-2, C-1, C-2, and D-1 are depicted under the corresponding symbols. * represents non tested subjects. White squares and circles represent family members with no mutation in GYG1 gene. Black squares and circles represent homozygote or compound heterozygote family members. Half black squares and circles represent heterozygote family members.

to a lesser extent the soleus muscles (figure e-1A, weakness at onset. Her assessment 26 years later links.lww.com/NXG/A13) and scapular muscles in showed facial muscle weakness with a highly asym- upper limbs, mainly deltoid muscles in an asymmet- metric scapulohumeral and pelvifemoral myopa- ric pattern (figure e-1B). After 13–14 years of pro- thies, thus resembling facioscapulohumeral gression, the disease involved almost all calf muscles dystrophy (FSHD), which was excluded by genetic and spread to the thighs and pelvic muscles, while at testing. Muscle imaging showed a peculiar pattern this time, deltoids had become totally replaced by fat with scapulohumeral and a more diffuse lower limb (figure e-1C and e-1C9). In more advanced stages, involvement. Unfortunately, there was no muscle 23–24 years after disease onset, almost all lower limb imaging performed at disease onset to better catego- muscles, anterior arm, and posterior forearm com- rize this patient pattern of myopathy. However, the partments were involved (figure e-1D and e-1D9). fact that distal parts of the upper and lower limbs Patient D-1 is peculiar, as she was diagnosed with were more conserved than proximal parts suggests the disease at 50 years of age with difficulties climb- that patient D-1 should be categorized within the ing stairs, suggesting a proximal lower limb proximal subgroup of patients. Of note, asymmetric

Neurology: Genetics 5 Figure 2 Histological features in patients A-1, B-1, C-2, and D-1

For each patient, we included hematoxyline-eosine (HE) and periodic acid Schiff (PAS) staining before and after a-amylase digestion (magnification x200 or x400).

limb involvement was found in 4 patients (A-1, C-1, positive inclusions, partially resistant to a-amylase, C-2,andD-1)involvingdeltoidmusclesin2(A-1 representing polyglucosan bodies (figure 2). Abnor- and D-1) and distal lower limb muscles in 2 (C-1 mally structured glycogen, suggesting polyglucosan and C-2). In addition, the creatine phosphokinase bodies, was shown on electron microscopy in patients (CPK) level was either normal in the oldest patients A-1, B-1, C-2, D-1, and E-1 (data not shown). (C-1, C-2, C-3, D-1, and E-1), while as high as 7 Abnormal histologic findings, usually not associated times above the upper referencelevelintheyoungest with GSDXV, were found in a few patients, such as (A-1, B-1, and B-2). mild necrosis (patient C-1), moderate endomysial All 9 patients reported here were also regularly as- fibrosis (patients C-1 and C-2), abnormal mito- sessed for heart and respiratory involvement. Only 2 pa- chondria accumulation (patients C-1), and extensive tients (patients C-1 and C-2) had cardiac abnormalities fatty replacement and myofibrillar disarrangements in the context of coronary artery disease at age 58 and (patient C-3). 76 years, respectively. Except for patient A-1, who Molecular genetics. Results of molecular investigations had mild restrictive respiratory syndrome, and patient for tested family members are summarized in the C-1 requiring night time nasal ventilation for obstruc- table. In patient A-1, we identified a new nonsense tive sleep apnea syndrome since age 70 years, respiratory mutation in exon 6 (c.646C.T, p.Arg216*) (in- assessments were unremarkable in all other patients. herited from his mother) in association with the pre- Morphologic features. Almost all the 6 patients who viously reported7 splicing mutation in intron 2 had a muscle biopsy performed showed PAS- (c.14313G.C, p.Asp3Glufs*4). In family B, the

6 Neurology: Genetics Figure 3 Effect of the mutation c.14313G>C

The top panel (A) is a proposed model explaining variable consequences of c.14313G.C mutation on RNA splicing with production of two mRNA with or without exon 2. Reverse transcriptase PCR (RT-PCR) products using primers F (in exon 2) and R (in exon 4) separated by electrophoresis shows the expected normal band at 290 bp in a patient carrying the c.14313G.C mutation at homozygote state. Electropherogram represents the reverse RT-PCR product sequence using R primer exhibiting exon 2-exon 3 junction. Western blot studies of glycogenin-1 myoblast cell culture (Panel B, patient A-1) and fresh muscle biopsy (Panel C and D, Patients B-1, C-1 and C-2) without (-) or with (1) alpha-amylase treatment. For each panel, the corresponding sex and age matched control tissue was used. same splicing mutation in intron 2 was found in c.996_1005del10, p.Tyr332*) leading to a frame- a homozygous state in twins (B-1 and B-2) and their shift that creates a premature stop codon and may younger brother (B-3). Parents and sister were het- produce a truncated protein and the previously re- erozygous for the mutation. Total mRNAs were ex- ported splicing mutation in intron 2 tracted from a muscle biopsy from patient B-1, (c.14313G.C).7 The intron 2 mutation was allowing for the identification of 2 transcripts: the found in a heterozygous state in a nonaffected previously reported in 4 families7—1 with exon 2 brother and sister, while the 10-bp deletion was skipped and the normal one (figure 3A). Patient found in a heterozygous state in another nonaf- E-1 carried a previously reported small out-of-frame fected brother. In patient D-1, 2 compound het- deletion (c.487delG, p.Asp163Thrfs*5).6 erozygous mutations (c.166G.C, p.Asp56His/ Exome sequencing of families C and D led to c.472delA, p.Asp159Thrfs*5) were identified. the identification of 3 novel mutations in GYG1. InfamilyC,weidentified2compoundheterozy- Protein analysis. Expression of glycogenin-1 is shown gous mutations: a small 10 bp out-of-frame dele- in figure 3. Under normal conditions, glucosylated tion in the last coding exon (NM_004130.3, glycogenin-1 within glycogen particles cannot be

Neurology: Genetics 7 detected unless the epitope is exposed after digestion and pelvifemoral girdle muscles similar to our pa- with a-amylase. In patient A-1 (figure 3B), tients B-1, B-2, and E-1 (figure e-1, links.lww.com/ glycogenin-1 was absent before and after a-amylase NXG/A13). The open ring-like aspect of the right treatment, indicating that glycogenin-1 is not pro- soleus muscle observed in 1 patient8 was never duced. In patient B-1 (figure 3B), glycogenin-1 was observed in our patients. not detected before a-amylase treatment, while it was As for previously published patients,7,8,10,11 an reduced after a-amylase treatment compared with intriguing asymmetry was observed in at least 5 of control, indicating that the expressed glycogenin-1 our patients manifesting as asymmetric weaker upper protein is still able to autoglucosylate. In patients limb elevation or scapular winging, associated in 1 C-1 and C-2 (figure 3, C and D), glycogenin-1 was patient with facial weakness leading initially to a clin- detected before a-amylase treatment (2 bands with ical diagnosis of FSHD. Thus, the presence of an a slight molecular weight difference) and was still asymmetric muscle involvement in a patient with present (but lower molecular weight) at similar glycogen storage myopathy should prompt GYG1 amounts after a-amylase treatment, indicating pre- gene analysis. served protein expression but impaired glucosylation. Cardiac involvement has to be evaluated with care. A patient previously reported6 developed ven- DISCUSSION We report 9 patients who showed tricular fibrillation since age 19 years, which unexpected clinical phenotypes and variable protein required implantable cardiac defibrillator. Among consequences because of mutations in the GYG1 gene the 18 other published cases, 11 had normal cardiac of which 4 are novel. assessments, while 7 had variable and inconclusive Patient A-1 had an initial scapular and peroneal cardiac abnormalities usually at advanced ages and pattern of muscle affection, which has not been sometimes with clear coronary arteries or high described previously in patients with GSDXV. blood pressure diseases.7–12 These findings are sim- However, 1 patient described in the original descrip- ilar to those of our patients. Recently, a series of 3 6 tion of the disease likely could have had the same patients with GYG1 mutation with severe cardiac phenotype since he was described to have weakness affections but with no skeletal muscular involve- of shoulder abduction, elbow flexion and extension, ment was published,14 suggesting that distinct phe- and foot dorsiflexion. Unfortunately, no muscle notypes with cardiac or skeletal muscle imaging was performed in that patient to more accu- involvement exist, which could relate to the muta- rately classify the pattern of affection. In family C, tion type. the disease started distally and evolved to the prox- All of our biopsied patients exhibited inclusions imal limb compartments. By contrast, the 9 pub- containing PAS-positive material partially resistant 6–10 lished patients showing distal involvement had to a-amylase and abnormally structured glycogen 7,8 6,7,9,10 either an initial diffuse or a proximal pattern. suggesting polyglucosan bodies. This is similar to 7 Only 1 published case may be reminiscent of family 17 of the 20 previously reported patients, in whom C with primary distal involvement, although the a major finding was the presence of inclusions in the disease started late in lower limbs. On the other center and subsarcolemmal regions of up to 50% of hand, our 4 patients with proximal pattern are sim- the fibers, containing a partially a-amylase–resistant 7–12 ilar to almost all the 20 cases reported so far. PAS-positive material. Inclusions surrounded by nor- Indeed, these previously reported patients had a wide mal glycogen content were visualized in some fibers. range of age and symptoms at onset that generally In other fibers, a depletion of normal glycogen was 7–12 pointed toward limb-girdle weakness. In some observed around the inclusions. Similar to previously instances, symptoms of onset suggested exercise reported cases,7,12 some of our patients showed occa- 8,9,11,12 intolerance, which was modest and was not sionally atypical features such as nemaline rods and the predominant feature of the phenotype. CPK increased interstitial connective and fat tissues. It is levels were usually normal but were found mildly worthy to note that among those patients who had to moderately elevated in 3 cases, reaching almost abnormal PAS-positive inclusions, 2 had a first mus- 1,100 IU/L in the youngest patient.7,9,11 The disease cle biopsy which was either normal10 or showed non- course was progressive in all patients, and muscle specific myopathic features.9 Moreover, there are 2 weakness was most disabling in the 7 oldest patients patients who showed nonspecific myopathic changes – (aged 57 82 years), as they were unable to raise arms without inclusions or polyglucosan bodies.9,11 These above their head and required aid for walking short atypical features emphasize the benefit of repeating distances or permanent wheelchair use. Muscle muscle biopsy when the diagnosis has not formerly 8,10–12 imaging was described only in 5 patients. been established. Except for 1 patient with almost all muscles The fact that mutations in 1 gene may lead to such 10 involved, the 4 others had mainly affected scapular various clinical patterns is one of the intriguing results

8 Neurology: Genetics that arise from our study. From this point of view, the truncated mRNA. The more this shorter protein is mutation c.14313G.C seems to be particularly abundant, the more distal the phenotype seems to be. demonstrative. Clinically, all the previously reported Clearly, these mutations need to be more deeply 11 patients carrying the homozygous c.14313G.C investigated at mRNA, protein, and functional levels mutation had a proximal pattern similar to the one to help us better understand their phenotypic observed in the 2 affected patients from our family B. consequences. However, there is a great heterogeneity at least in Our findings demonstrate that GYG1 mutations terms of age at disease onset among individuals car- may lead to a wide spectrum of clinical phenotypes rying this homozygous mutation. In fact, the 2 pa- including limb-girdle and scapuloperoneal muscle tients from family B had a teenage onset, while their weaknesses but also late-onset distal myopathy. This younger brother is still asymptomatic, suggesting study also highlighted the value of glycogenin-1 pro- incomplete penetrance of this mutation. By contrast, tein and mRNA studies to better understand the phe- the age at onset of the disease in the 11 published notype/genotype correlations in this group of patients carrying this homozygous mutation ranged diseases. from childhood to 55 years old. A muscle comple- mentary DNA study from our patient B-1 revealed AUTHOR CONTRIBUTIONS not only the mRNA with exon 2 skipped as previ- R. Ben Yaou, A. Hubert, and D. Gaist: study concept and design and drafting/revising the manuscript. R. Ben Yaou, I. Nelson, P. Petiot, 7 ously described but also the normal one. This new J.R. Dahlqvist, M. Krahn, F. Michel, R.Y. Carlier, T. Krag, and J. Vissing: finding can explain faint traces of protein observed in acquisition of data. A. Hubert, I. Nelson, N. Streichenberger, M. Beuvin, patient B-1 myoblasts on Western blot and in the F. Parisot, E. Malfatti, N. Romero, M. Duno, M. Cerino, M. Bartoli, G. Bonne, and F.M. Petit: analysis and interpretation. B. Eymard and different muscle biopsies of patients of family C, sug- P. Labrune: critical revision of the manuscript for important intellectual gesting variable levels of mRNA splicing in these tis- content. G. Bonne, J. Vissing, P. Laforet, and F.M. Petit: study sues. We speculate that the level of normal transcript supervision. expression may modulate the deleterious effect of glycogenin-1 deficiency and therefore the age at onset ACKNOWLEDGMENT and symptom intensity. Moreover, this balance The authors thank Nicolas Levy for his contribution to molecular investigations. between the 2 forms seems variable among individuals. DISCLOSURE In our series, the c.14313G.C mutation was R. Ben Yaou, A. Hubert, I. Nelson, and J. Dahlqvist report no disclo- also found in a compound heterozygous state com- sures. D. Gaist has received honoraria from AstraZeneca and has received bined with the new nonsense c.646C.T mutation research support from the Danish Cancer Society. N. Streichenberger and * M. Beuvin report no disclosures. M. Krahn has served on the editorial (p.Arg216 ) or the new out-of-frame 10 bp deletion board of Cahiers de Myologie; holds patents for “Exon-skipping for dys- c.996_1005del10 (p.Tyr332*). The c.646C.T ferlinopathies therapy” and “In vitro genetic diagnostic of inherited neu- mutation may result in a complete absence of protein romuscular disorders”; and has received research support from the production through an mRNA nonsense-mediated European Community Seventh Framework Program, INSERM, and AFM-TELETHON. P. Petiot, F. Parisot, and F. Michel report no dis- decay phenomenon. The total absence of closures. E. Malfatti has served on the editorial board of Frontiers in glycogenin-1 on Western blot found in myoblasts Neurology. N. Romero reports no disclosures. R. Carlier has received from patient A-1 reinforces the notion of such a mech- speaker honoraria from Genzyme. B. Eymard has been a consultant for conferences of LFB, Biogen and BioMarin companies and has received anism. On the other hand, a Western blot study of research support from Association Française contre les myopathies muscle biopsies from patients C-1 and C-2 revealed (AFM). P. Labrune and M. Duno report no disclosures. T. Krag has the presence of 2 unglucosylated glycogenin-1 bands received research support from AP Møller Foundations, the Augustinus without a-amylase treatment, probably the normal Foundation, and the Danielsen Foundation. M. Cerino and M. Bartoli * report no disclosures. G. Bonne has served on the editorial boards of the form and the shorter form related to p.Tyr332 . This Journal of Neuromuscular Diseases and Neuromuscular Disorders and has result is comparable with a Western blot finding from received research support from the ANR French-German grant (ANR- previously reported patient 77 carrying the BMBF), Association Institut de Myologie (AIM), GIS Maladies Rares “ ” c.14313G.C mutation and the nonsense Plateforme Mutations, and the CURE-CMD Translational Award. J. Vissing has served on the scientific advisory boards of Sanofi/Genzyme, . * c.970C T (p.Arg324 ), which also showed 2 unglu- aTyr Pharma, Ultragenyx Pharmaceuticals, Santhera Pharmaceuticals, cosylated glycogenin-1 bands without a-amylase Sarepta Therapeutics, Novo Nordisk, Alexion Pharmaceuticals, and treatment. Based on this observation, we speculate Stealth BioTherapeutics; has received travel funding and speaker hono- raria from Sanofi/Genzyme, Alexion Pharmaceuticals, Ultragenyx Phar- that this shorter protein had different deleterious ef- maceuticals, Santhera Pharmaceuticals, and aTyr Pharma; has served on fects on the normal one resulting from the the editorial boards of Neuromuscular Disorders and the Journal of Neu- c.14313G.C mutation. We hypothesize that trun- romuscular Diseases; and has received research support from the Lundbeck cated GYG1 protein impairs normal autoglucosyla- Foundation, the Novo foundation, the Danish Medical Research Coun- cil, the University of Copenhagen, and the Research Committee of the tion of the normal form. Different phenotypes could National Hospital. P. Laforet is a member of the Sanofi Genzyme Pompe be explained by imbalance in the production of nor- advisory board; has received travel funding and speaker honoraria from mal mRNA, mRNA with exon 2 skipped, and Sanofi Genzyme; and has received research support from Sanofi

Neurology: Genetics 9 Genzyme, Valerion Therapeutics, and the French Glycogenosis Associa- 8. Luo S, Zhu W, Yue D, et al. Muscle pathology and whole- tion. F. Petit reports no disclosures. Go to Neurology.org/ng for full body MRI in a polyglucosan myopathy associated with disclosure forms. a novel glycogenin-1 mutation. Neuromuscul Disord 2015;25:780–785. Received May 17, 2017. Accepted in final form September 29, 2017. 9. Fanin M, Torella A, Savarese M, Nigro V, Angelini C. GYG1 gene mutations in a family with polyglucosan body REFERENCES myopathy. Neurol Genet 2015;1:e21. doi: 10.1212/NXG. 1. Roach PJ. Glycogen and its metabolism. Curr Mol Med 0000000000000021. 2002;2:101–120. 10. Colombo I, Pagliarani S, Testolin S, et al. Longitudinal 2. Oldfors A, DiMauro S. New insights in the field of muscle follow-up and muscle MRI pattern of two siblings with glycogenoses. Curr Opin Neurol 2013;26:544–553. polyglucosan body myopathy due to glycogenin-1 muta- 3. Lomako J, Lomako WM, Whelan WJ. Glycogenin: the tion. J Neurol Neurosurg Psychiatry 2016;87:797–800. primer for mammalian and yeast glycogen synthesis. Bio- 11. Akman HO, Aykit Y, Amuk OC, et al. Late-onset polyglu- chim Biophys Acta 2004;1673:45–55. cosan body myopathy in five patients with a homozygous 4. Barbetti F, Rocchi M, Bossolasco M, et al. The human mutation in GYG1. Neuromuscul Disord 2016;26:16–20. skeletal muscle glycogenin gene: cDNA, tissue expression 12. Tasca G, Fattori F, Monforte M, et al. Start codon mutation and chromosomal localization. Biochem Biophys Res of GYG1 causing late-onset polyglucosan body myopathy Commun 1996;220:72–77. with nemaline rods. J Neurol 2016;263:2133–2135. 5. Mu J, Skurat AV, Roach PJ. Glycogenin-2, a novel self- 13. Nilsson J, Halim A, Moslemi AR, et al. Molecular glucosylating protein involved in liver glycogen biosynthe- pathogenesis of a new glycogenosis caused by a glycoge- sis. J Biol Chem 1997;272:27589–27597. nin-1 mutation. Biochim Biophys Acta 2012;1822: 6. Moslemi AR, Lindberg C, Nilsson J, Tajsharghi H, 493–499. Andersson B, Oldfors A. Glycogenin-1 deficiency and in- 14. Hedberg-Oldfors C, Glamuzina E, Ruygrok P, et al. Car- activated priming of glycogen synthesis. N Engl J Med diomyopathy as presenting sign of glycogenin-1 defi- 2010;362:1203–1210. ciency-report of three cases and review of the literature. 7. Malfatti E, Nilsson J, Hedberg-Oldfors C, et al. A new J Inherit Metab Dis 2017;40:139–149. muscle glycogen storage disease associated with 15. Dubowitz V. Muscle Biopsy: A Practical Approach. 2nd glycogenin-1 deficiency. Ann Neurol 2014;76:891–898. ed. London: Bailliere Tindall; 1985.

10 Neurology: Genetics Clinical/Scientific Notes

Celine Bris, PharmD NOVEL NDUFS4 GENE MUTATION IN AN enzyme activities revealed, however, a moderate Tiphaine Rouaud, MD ATYPICAL LATE-ONSET MITOCHONDRIAL FORM reduction of complex I enzyme activity in both mus- Valerie Desquiret-Dumas, OF MULTIFOCAL DYSTONIA cle and fibroblasts (figure, C). The patient was given PhD coenzyme Q10 (150 mg/d) since the age of 22 years Naig Gueguen, PhD Mitochondrial complex I, the largest component of and received routine injections of botulinum toxin David Goudenege, PhD the mitochondrial respiratory chain, comprises 44 into dystonic muscles. Magalie Barth, MD subunits of which 7 are encoded by the mitochondrial The complex I enzyme deficiency identified in Dominique Bonneau, genome and the remainder by the nuclear genome.1 muscle prompted us to look for molecular abnormal- MD, PhD Isolated complex I deficiencies represent a major con- ities. Blood sample was obtained from the patient Patrizia Amati-Bonneau, tribution within the group of respiratory chain de- after written informed consent was obtained. The MD fects.2 We report an atypical case carrying entire mitochondrial genome sequence did not reveal Guy Lenaers, PhD a homozygous NDUFS4 missense mutation, with pathogenic mutations. Thereafter, targeted rese- Pascal Reynier, MD, PhD late-onset multifocal dystonia, in contrast to expected quencing of 179 nuclear-encoded mitochondrial Anne-Sophie Lebre, clinical phenotypes due to other NDUFS4 mutations, genes, including all known structural and assembly PharmD, PhD which have been constantly reported to be responsi- complex I genes, was performed (table e-1, http:// Vincent Procaccio, MD, ble for Leigh syndrome of early onset and death.3 links.lww.com/NXG/A10), revealing a novel homo- PhD zygous c.369C.A mutation in exon 4 of NDUFS4, confirmed by Sanger sequencing, changing an evolu- Neurol Genet Case report. A 29-year-old woman was followed up 2017;3:e205; doi: 10.1212/ for years in our clinics who had progressive multifocal tionary conserved asparagine in lysine at position 123 NXG.0000000000000205 dystonia and parkinsonism since the age of 12 years (p.Asn123Lys) (figure, E). No other candidate muta- when first symptoms appeared with hand tremor tion was identified in complex I genes. The . and muscular contractures of both arms. She was c.369C A mutation has robust bioinformatics dam- the only child born to first-degree consanguineous aging predictions absent from publicly available data- asymptomatic parents with unremarkable family bases and classified as a likely pathogenic variant history (figure, A). Dystonia and bradykinesia did not according to the American College of Medical Genet- progress over the following years. Examination at age ics and Genomics criteria (table e-2, http://links.lww. 29 years revealed fixed dystonia involving shoulders com/NXG/A11). and both arms and the lower face responsible for mild To date, NDUFS4 has been implicated only in dysarthria, associated with severe bradykinesia of the individuals having Leigh syndrome who were com- face, amyotrophic arms and hands, and a mild pound heterozygous for null mutations.4 The late restrictive respiratory failure due to chronic mild onset and the mild phenotype observed in the patient myopathy. The trunk and legs were not affected by could be due to the homozygozity for the c.369C.A dystonia. Gait and cognitive functions were pre- missense mutation. served. The patient gave birth to a healthy boy at the To further support the pathogenicity of this novel age of 25 years with an uneventful pregnancy. variant, the effects of the mutation on complex I Brain MRI revealed bilateral atrophy of the poste- assembly were investigated in the skin fibroblasts of rior part of the putamen with T2 hyperintensities the patient. Blue native polyacrylamide gel revealed related to the dystonic features, without lactate peak a moderate assembly defect with the presence of (figure, B). Needle EMG showed diffuse myogenic assembly intermediates of 830 kDa and 300 kDa, polyphasic waves, compatible with a chronic muscu- respectively, of the N module of complex I in the pa- lar defect without signs of peripheral neuropathy. tient’s cells compared with control cells. By contrast, Lactate levels were normal both in blood and CSF. a patient with typical severe Leigh syndrome carrying The coenzyme Q10 level was low in CSF. Electrocar- NDUFS4 compound heterozygous nonsense muta- diogram and ophthalmologic examination were unre- tions showed complete absence of the complex I holo- markable. Muscle biopsy did not reveal structural enzyme and stronger accumulation of complex I abnormalities. Measurement of mitochondrial assembly intermediates (figure, D).

Neurology.org/ng Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 Figure Patient’s pedigree, paraclinical tests, and novel NDUFS4 gene mutation

(A) Pedigree of the proband III1. (B) Patient brain MRI. T2-weighted MRI showed bilateral posterior putaminal atrophy and hyperintensities. (C) Mitochondrial respiratory enzyme activities of muscle and fibroblasts from the NDUFS4 patient and controls. Mitochondrial respiratory chain and citrate synthase (CS) activities were determined by spectrophotometric analysis. CS activity is expressed in nanomoles per minute per milligram. Mitochondrial respiratory chain data are expressed as the ratio of complexes/CS activities. (D) Blue native polyacrylamide gel (BN-PAGE) analysis from NDUFS4 patients: a proband and a patient with severe Leigh syndrome carrying nonsense mutations (c.262C.T; p.Gln88X and c.316C.T; p.Arg106X) and control using antibodies against NDUFB6 and NDUFS3 subunits detecting complex I and SDHA (70 kDa) revealing complex II. Primary fibroblasts were cultured from the patient and controls, and BN-PAGE analysis was performed as described elsewhere.7 (E) Identification of the c.369C.A mutation in the proband compared with wild-type and inter- species amino acid comparison, showing that the asparagine (N) at position 123 is highly conserved (in orange).

2 Neurology: Genetics Discussion. It has been recently stated that Study funding: Study funded by Angers University Hospital, GIS NDUFS4 patients are invariably affected with neo- Institut des Maladies Rares A12119KS, and Association contre les Maladies Mitochondriales. natal and early-onset Leigh syndrome characterized Disclosure: C. Bris and T. Rouaud report no disclosures. V. by a severe complex I deficiency leading to early death Desquiret-Dumas and N. Gueguen have received research support occurring before the age of 3 years.3 Moreover, in from Fondation Maladies Rares. D. Goudenege, M. Barth, D. Bon- previous studies, all patient cells carrying NDUFS4 neau, P. Amati-Bonneau, G. Lenaers, P. Reynier, and A.-S. Lebre report no disclosures. V. Procaccio has received research support from mutations had a much more severe complex I enzyme Angers University Hospital, GIS Institut des Maladies Rares deficiency with a constant and similar abnormal (A12119KS), and Association contre les Maladies Mitochondriales. complex I assembly profile.4 Go to Neurology.org/ng for full disclosure forms. The Article Process- In addition, brain MRI was also atypical in our ing Charge was funded by the authors. patient, without brain stem abnormalities considered This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives Li- as one of the most specific signs pointing to a nuclear- cense 4.0 (CC BY-NC-ND), which permits downloading and shar- – encoded complex I deficiency.4 6 ing the work provided it is properly cited. The work cannot be Dystonia appears to be a common feature in com- changed in any way or used commercially without permission from the journal. plex I–deficient patients due to mutations in nuclear- encoded genes, present in approximately 32% of 130 Received May 23, 2017. Accepted in final form September 26, 2017. patients from a recently reported series.2 Our results indicate that dystonia and movement disorders can Correspondence to Dr. Procaccio: [email protected] also be caused by NDUFS4 mutations, emphasizing 1. Fassone E, Rahman S. Complex I deficiency: clinical fea- the need for extensive genetic investigation, thanks to tures, biochemistry and molecular genetics. J Med Genet the new sequencing technologies. These findings 2012;49:578–590. underscore the role of NDUFS4 in the dysfunction 2. Koene S, Rodenburg RJ, van der Knaap MS, et al. Natural of respiratory complex I in patients having mitochon- disease course and genotype-phenotype correlations in drial disorders and expand the clinical phenotype due Complex I deficiency caused by nuclear gene defects: what we learned from 130 cases. J Inherit Metab Dis 2012;35: to NDUFS4 mutations. 737–747. 3. Ortigoza-Escobar JD, Oyarzabal A, Montero R, et al. From the UMR CNRS 6015-INSERM U1083 (C.B., V.D.-D., N. Ndufs4 related Leigh syndrome: a case report and review G., D.G., M.B., D.B., P.A.-B., G.L., P.R., V.P.), Mitovasc Insti- of the literature. Mitochondrion 2016;28:73–78. tute, Angers University, France; Department of Biochemistry and Genetics (C.B., V.D.-D., N.G., D.G., M.B., D.B., P.A.-B., G.L., 4. Assouline Z, Jambou M, Rio M, et al. A constant and P.R., V.P.), University Hospital of Angers, France; Department of similar assembly defect of mitochondrial respiratory chain Neurology (T.R.), University Hospital of Nantes, France; and complex I allows rapid identification of NDUFS4 muta- CHU Reims (A.-S.L.), Hôpital Maison Blanche, Pole de biologie, tions in patients with Leigh syndrome. Biochim Biophys Service de génétique, France. Acta 2012;1822:1062–1069. Author contributions: Celine Bris: acquisition and interpretation of 5. Leshinsky-Silver E, LebreAS,MinaiL,etal.NDUFS4 data. Tiphaine Rouaud and Magalie Barth: clinical investigation mutations cause Leigh syndrome with predominant and analysis and interpretation of data. David Goudenege: in silico brainstem involvement. Mol Genet Metab 2009;97: analysis and analysis and interpretation of data. Valerie Desquiret- 185–189. Dumas and Naig Gueguen: study concept and design and acquisition 6. Lebre AS, Rio M, Faivre d’Arcier L, et al. A common of data. Patrizia Amati-Bonneau: analysis and interpretation of pattern of brain MRI imaging in mitochondrial diseases data. Guy Lenaers, Dominique Bonneau, and Pascal Reynier: crit- with complex I deficiency. J Med Genet 2011;48:16–23. ical revision of the manuscript for intellectual content. Anne-Sophie Lebre: study concept and design and critical revision of the manu- 7. Leman G, Gueguen N, Desquiret-Dumas V, et al. Assem- script for intellectual content. Vincent Procaccio: study concept and bly defects induce oxidative stress in inherited mitochon- design, critical revision of the manuscript for intellectual content, and drial complex I deficiency. Int J Biochem Cell Biol 2015; study supervision. 65:91–103.

Neurology: Genetics 3 Clinical/Scientific Notes

Hugo Botha, MBChB NOVEL GRN MUTATION PRESENTING AS AN rigidity. She had minimal behavioral disturbance NiCole A. Finch, MS APHASIC DEMENTIA AND EVOLVING INTO and was still able to sing, despite being essentially Ralitza H. Gavrilova, MD CORTICOBASAL SYNDROME nonverbal in conversation. Mary M. Machulda, PhD Her family history was notable for Parkinson dis- Julie A. Fields, PhD Mutations in the granulin (GRN) gene on chromo- ease and dementia (figure 2). Genetic testing was Val J. Lowe, MD some 17 most commonly result in behavioral variant offered in light of the positive family history (Gold- 4 Ronald C. Petersen, MD, frontotemporal dementia (FTD) or primary progres- man score 2). Full sequencing of the GRN gene PhD sive aphasia (PPA), although a wide range of pheno- revealed a previously unreported mutation in exon Clifford R. Jack, Jr., MD types have been described.1,2 At the time of 12 (c.1535delC, Pro512LeufsX5), resulting in a pre- Christina M. Dheel, BS publication, 172 mutations have been described mature stop codon. Both MAPT sequencing and Debra J. Gearhart, AA (molgen.vib-ua.be/FTDMutations), 79 of which are molecular analysis of the C9orf72 gene were normal. David S. Knopman, MD thought to be pathogenic, with no clear genotype- Plasma progranulin levels were quantified and com- Rosa Rademakers, PhD phenotype correlation. Here, we describe a novel pared with subjects with known pathogenic muta- Bradley F. Boeve, MD mutation presenting as a dysexecutive, aphasic tions as well as controls (figure 1C). Controls had dementia and evolving into a corticobasal syndrome levels more than double that of known mutation car- Neurol Genet riers. The level in our case was far below than that 2017;3:e201; doi: 10.1212/ (CBS) phenotype. NXG.0000000000000201 seen in controls, albeit slightly higher than other Case report. A 61-year-old right-handed woman known mutation carriers, supporting the pathogenic- presented with difficulty expressing herself in writing ity of the mutation. more so than speech. Her problems started around the age of 60, with deterioration of her penmanship. Discussion. It has been a little more than a decade This was followed by difficulty with simple arith- since the first report linking mutations in the GRN metic, impairing her ability to work, as well as trouble gene to cases of tau-negative familial FTD was pre- with tasks reliant on sequencing, such as preparing sented.2,5 Despite important advances in our a sandwich or making coffee. Closer to the time of understanding of the role granulin plays in the evaluation, the patient and her family noticed word nervous system, including as a growth factor and finding difficulty, yes-no confusion, word sub- modulator of inflammation, the exact mechanism by stitutions from semantically related categories, and which the haploinsufficiency that results from mu- mild gait imbalance. tations causes neurodegeneration has not been Her initial cognitive evaluation revealed deficits in elucidated. calculation and digit span, but no trouble with nam- Our case better illustrates the heterogeneity in ing, recall, registration, or construction. Her repeti- GRN-related disease. Although related to a novel tion was spared, but she had difficulty following 3 mutation in exon 12, a relatively rare site for GRN step commands, and her writing was considerably mutations, her presentation shares features of previ- impaired. She had clear left hemispheric atrophy ously reported mutations. Her initial complaint of and hypometabolism (figure 1A) and was amyloid- deteriorating penmanship was likely due to apraxic PET negative. Formal neuropsychometric testing 18 agraphia, well reported in CBS, but reported only months after symptom onset demonstrated impaired once in GRN-related CBS previously.6 Her pheno- executive, letter/category fluency, and visuospatial type at the time of initial evaluation did not qualify skills (figure 1B). Over the following year, she devel- for a diagnosis of PPA based on her impairment in oped more generalized cognitive, language, and nonlanguage domains, but the prominent language motor impairment. She had right hemibody parkin- difficulty is in keeping with PPA being the second sonism on examination and met the criteria for CBS most common presentation of GRN mutations and during her second visit at age 62.3 At her last follow- a common early feature in CBS.1,7 Over time, a clear up at age 63, she had minimal meaningful language CBS picture emerged, another common manifesta- output and little use of her right upper extremity, tion of GRN mutations.1,7 Her prominent parkinson- which was held in a flexed posture with marked ism, including marked rigidity, raises the possibility

Neurology.org/ng Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 Figure 1 Results of imaging, neuropsychological and molecular analyses

(A) MRI (rows 1 and 2) and FDG-PET (rows 3–5) findings at presentation (age 61 years) and follow-up (ages 62 and 63 years). Note moderate-to-severe, asymmetric left frontal-temporal-parietal atrophy, with progression at follow-up, and relative hippocampal sparing. The same pattern is present on fludeoxyglucose PET (FDG-PET) imaging, with almost exclusively left-sided hypometabolism even at follow-up, and little-to-no anterior and medial temporal involvement. (B) Performance on key tests in the neuropsychological battery is shown graphically, with performance on each test displayed using the Mayo Older American Normative Studies (MOANS) standard score as reference. Scores at or below 6 are usually considered abnormal. Impaired performance was found on fluency measures, attention/executing control measures, and one of the visuospatial measures. (C) Plasma progranulin levels quantified by ELISA in controls (CN) and affected mutation carriers (GRN1). Values in CN (mean 46.53 ng/mL, SD 3.9 ng/mL) were significantly higher than those in GRN1 (mean 14.48 ng/mL, SD 1.38 ng/mL). The level in our case (23.9 ng/mL) is shown in red. AVLT 5 Auditory Verbal Learning Test; BNT 5 Boston Naming Test; CF 5 category fluency; DRS-2 5 Dementia Rating Scale 2; GRN15progranulin mutation cases; JLO 5 judgment of line orientation; L 5 left; LF 5 letter fluency; R 5 right; TMT A 5 Trial Making Task Part A; TMT B 5 Trial Making Task Part B; WAIS-BD 5 Wechsler Adult Intelligence Scale Block Design; WAIS-DS 5 Wechsler Adult Intelligence Scale Digit Span; WMS-R LM 5 Wechsler Memory Scale–Revised Logical Memory.

2 Neurology: Genetics Disclosure: H. Botha and N.A. Finch report no disclosures. R.H. Figure 2 Outline of family pedigree Gavrilova receives research support from the NIH. M.M. Machulda receives research support from the NIH and the Robert H. and Clarice Smith and Abigail Van Buren Alzheimer’s Disease Research Program of the Mayo Foundation. J.A. Fields has received speaker honoraria from the American Academy of Clinical Neuropsychology; has served on the editorial board of the International Journal of Neuroscience; and receives research support from the NIH, the Robert H. and Clarice Smith and Abigail Van Buren Alzheimer’s Disease Research Program of the Mayo Foundation, and the Patient- Centered Outcomes Research Institute. V.J. Lowe has served on the scientific advisory boards of Piramal Imaging and Merck Research, Inc.; is a consultant for Bayer Schering Pharma, Merck Research, and Piramal Imaging Inc; and receives research support from GE Health care, Siemens Molecular Imaging, AVID Radiopharmaceuticals, the Triangles represent individuals, and shaded triangles represent individuals affected by NIH (NIA, NCI), the Elsie, and Marvin Dekelboum Family Foun- a degenerative disease. Triangles with diagonal lines through them represent deceased indi- dation, the Liston Family Foundation, and the MN Partnership for viduals. The proband is indicated by an asterisk. An elipsis in a triangle represents multiple Biotechnology and Medical Genomics. R.C. Petersen serves on data unaffected offspring not shown to maintain confidentiality. One parent was diagnosed with monitoring committees for Pfizer Inc. and Janssen Alzheimer Immu- Parkinson disease (II.2) and the other with dementia (II.3), both late in life. A sibling of the par- notherapy; works as a consultant for Merck Inc, Roche Inc., Biogen ent with dementia was diagnosed with Alzheimer disease dementia late in life (II.1). One of Inc., Eli Lilly and Company, and Genentech Inc.; receives publishing the patient’s siblings was suspected elsewhere to have Pick disease (III.3), based on behav- royalties for Mild Cognitive Impairment (Oxford University Press, ioral disturbance, aphasia, and cognitive impairment, and this person passed away in the 2003); and receives research support from the NIH, the National early 60s. No postmortem examination was performed. Multiple other siblings were cogni- Advisory Council on Aging (National Institute on Aging), and the tively normal (all older than 45 years). Robert H. and Clarice Smith and Abigail Van Buren Alzheimer’s Disease Research Program of the Mayo Foundation. C.R. Jack has that the family member with parkinsonism may in provided consulting services for Eli Lilly; owns stock in Johnson and Johnson; and receives research funding from the NIH and the Alex- fact have carried the same mutation. Although no ander Family Alzheimer’s Disease Research Professorship at Mayo imaging features are pathognomonic, GRN mutations Clinic. C.M. Dheel and D.J. Gearhart report no disclosures. D.S. tend to cause more asymmetric atrophy and hypome- Knopman receives research support from the NIH and the Robert H. ’ tabolism than is seen in sporadic FTD or in MAPT or and Clarice Smith and Abigail Van Buren Alzheimer s Disease Research Program of the Mayo Foundation; has received speaker C9orf72 mutations, as well as more parietal involve- honoraria for lectures at the Behavioral Neurology Conference, Hy- ment and higher rates of atrophy.7–9 The reasons for derabad, India, April 2016; served on the editorial board of Neu- the asymmetry, particularly marked in our case, rology; serves on Data Safety Monitoring Boards for Lundbeck Pharmaceuticals, the DIAN study, and the Consultant Bluefield remain a mystery, especially in light of the fact that project; and is an investigator in clinical trials sponsored by Biogen, the haploinsufficiency would be thought to affect TauRX Pharmaceuticals, Lilly Pharmaceuticals, and the Alzheimer’s both hemispheres to a similar degree. Disease Treatment and Research Institute, University of Southern Our case illustrates the importance of considering California. R. Rademakers holds patents for Detecting and Treating Dementia, 12/302,691 (2008) and Methods and Materials for De- GRN mutations in cases with markedly asymmetric tecting and Treating Dementia 12/413,869 (2009) and has received involvement and a positive family history for demen- research support from NIH, the Mayo Clinic Udall Center of Excel- tia or parkinsonism. lence, the ALS Therapy Alliance, the Consortium for Frontotemporal Dementia, and the Florida State Alzheimers Disease Research grant. From the Department of Neurology (H.B., R.C.P., D.S.K., B.F.B.), B.F. Boeve has served as an investigator for clinical trials sponsored by Department of Clinical Genomic and Neurology (R.H.G.), Depart- GE Health care and Axovant; receives royalties from the publication ment of Psychiatry and Psychology (M.A.M., J.A.F.), Department of of a book entitled Behavioral Neurology Of Dementia (Cambridge Nuclear Medicine (V.J.W.), Department of Radiology (C.R.J.), and Medicine, 2009); serves on the scientific advisory board of the Tau Alzheimer’s Disease Research Center (C.M.D., D.J.G.), Mayo Consortium; and receives research support from the NIH, the Robert Clinic, Rochester, MN; and Department of Neuroscience (N.A.F., H. and Clarice Smith and Abigail Van Buren Alzheimer’s Disease R.R.), Mayo Clinic, Jacksonville, FL. Research Program of the Mayo Foundation, the Little Family Foun- Author contributions: Hugo Botha: acquisition of data, analysis and dation, and the Mangurian Foundation. The Article Processing interpretation of data, and manuscript preparation. NiCole A. Finch Charge was funded by the authors. Go to Neurology.org/ng for full and Ralitza H. Gavrilova: acquisition of data and analysis and disclosure forms. interpretation of data. Mary M. Machulda and Julie A. Fields: This is an open access article distributed under the terms of the acquisition of data. Val J. Lowe: acquisition of data, analysis and Creative Commons Attribution-NonCommercial-NoDerivatives Li- interpretation of data, and study supervision. Ronald C. Petersen: cense 4.0 (CC BY-NC-ND), which permits downloading and shar- analysis and interpretation of data. Clifford R. Jack: analysis and ing the work provided it is properly cited. The work cannot be interpretation of data and study supervision. Christina M. Dheel and changed in any way or used commercially without permission from Debra J. Gearhart: acquisition of data. David S. Knopman: acqui- the journal. sition of data and analysis and interpretation of data. Rosa Rade- Received August 10, 2017. Accepted in final form September 14, 2017. makers and Bradley F. Boeve: analysis and interpretation of data, study supervision, and critical revision of the manuscript for intellec- Correspondence to Dr. Boeve: [email protected] tual content. Acknowledgment: The authors thank the patient and her family for 1. Yu CE, Bird TD, Bekris LM, et al. The spectrum of participating in aging and dementia research. mutations in progranulin: a collaborative study screening Study funding: This study was funded by R35 NS097261, U01 545 cases of neurodegeneration. Arch Neurol 2010;67: AG045390, U54 NS092089, and P50 AG016574. 161–170.

Neurology: Genetics 3 2. Baker M, Mackenzie IR, Pickering-Brown SM, et al. Mu- 6. Passov V, Gavrilova RH, Strand E, Cerhan JH, Josephs KA. tations in progranulin cause tau-negative frontotemporal Sporadic corticobasal syndrome with progranulin mutation dementia linked to chromosome 17. Nature 2006;442: presenting as progressive apraxic agraphia. Arch Neurol 916–919. 2011;68:376–380. 3. Armstrong MJ, Litvan I, Lang AE, et al. Criteria for the 7. Le Ber I, Camuzat A, Hannequin D, et al. Phenotype variabil- diagnosis of corticobasal degeneration. Neurology 2013;80: ity in progranulin mutation carriers: a clinical, neuropsycholog- 496–503. ical, imaging and genetic study. Brain 2008;131:732–746. 4. Goldman JS, Farmer JM, Wood EM, et al. Comparison of 8. Rohrer JD, Ridgway GR, Modat M, et al. Distinct profiles family histories in FTLD subtypes and related tauopathies. of brain atrophy in frontotemporal lobar degeneration Neurology 2005;65:1817–1819. caused by progranulin and tau mutations. Neuroimage 5. Gass J, Cannon A, Mackenzie IR, et al. Mutations in 2010;53:1070–1076. progranulin are a major cause of ubiquitin-positive fron- 9. Kelley BJ, Haidar W, Boeve BF, et al. Prominent pheno- totemporal lobar degeneration. Hum Mol Genet 2006;15: typic variability associated with mutations in Progranulin. 2988–3001. Neurobiol Aging 2009;30:739–751.

4 Neurology: Genetics Clinical/Scientific Notes

Ewen W. Sommerville, OPENING ONE’S EYES TO MOSAICISM IN RRM2B, SLC25A4, TWNK, POLG2, TK2, and PhD PROGRESSIVE EXTERNAL OPHTHALMOPLEGIA RNASEH1) were apparently negative, prompting Rachel L. Jones, BSc whole-exome sequencing with blood DNA. The (Hons) Autosomal dominant progressive external ophthalmo- mean depth per exome consensus coding sequence Steven A. Hardy, PhD plegia (adPEO) is a mendelian disorder of mitochon- (CCDS) was 71-fold, while the mean percentage of Emma L. Blakely, PhD, drial DNA (mtDNA) maintenance characterized by CCDS bases at 20-fold coverage was 87.01%. Anal- FRCPath restricted eye movements, ptosis, and skeletal muscle– ysis of rare exonic variants in nuclear genes encoding Angela Pyle, PhD restricted multiple mtDNA deletions.1 Dominantly in- mitochondrial-localized proteins or DNA transcrip- Andrew M. Schaefer, herited pathogenic variants of TWNK (GenBank: tion, replication, or maintenance machinery unex- MRCP NM_021830), encoding twinkle helicase, an essential pectedly revealed a known pathogenic heterozygous Patrick F. Chinnery, 2,3 protein required to unwind mtDNA during replication, c.1121G.A p.(Arg374Gln) TWNK variant. Analysis FRCP, FMedSci are among the most common cause of adult-onset of additional genes associated with mtDNA mainte- Douglass M. Turnbull, PEO,2,3 with patients manifesting relatively indolent nance disorders failed to detect pathogenic variants. PhD, FRCP PEO phenotypes, often with proximal muscle weakness, Examination of read coverage at the c.1121 base (UCSC Gráinne S. Gorman, PhD, ataxia, visual loss or impairment, and neurologic symp- hg19, g.102749088) showed 30 reads that passed qual- FRCP toms.1 However, after excluding pathogenic variants in ity score filtering, with an unfiltered allele depth of 20 Robert W. Taylor, PhD, known genes, the genetic etiology remains undeter- reads indicating wild-type and 10 reads indicating the FRCPath mined in approximately 50% of patients. We describe variant (c.1121G.A, p.Arg374Gln) (figure e-1, http://

Neurol Genet thediagnosticodysseyofapatientwhopresentedinthe links.lww.com/NXG/A8). Re-examination of the diag- 2017;3:e202; doi: 10.1212/ fourth decade of life with PEO and harbored a mosaic nostic electropherograms for TWNK from blood DNA NXG.0000000000000202 known pathogenic TWNK variant, initially undetected revealed a small peak for the c.1121G.Ap. by conventional diagnostic Sanger sequencing but iden- (Arg374Gln) variant, which was confirmed by tified using next-generating sequencing. repeated Sanger sequencing of both blood and mus- A 41-year-old woman initially presented with cle DNA, suggestive of low-level mosaicism (figure a 9-year history of progressive ptosis associated with 1C). Targeted next-generation sequencing was sub- intermittent diplopia. Clinical examination revealed sequently performed to achieve a higher depth of bilateral, asymmetric ptosis (right . left), external coverage at the variant nucleotide and allow a more ophthalmoplegia, mild facial weakness, and mild accurate estimation of the levels of mosaicism in proximal myopathy (Medical Research Council 4/ these tissues. A read depth of .1,700 was observed 5). Family history was unremarkable. Initially, she in both tissues, and the levels of mosaicism were was investigated for myasthenia gravis that proved determined to be 13% in blood and 18% in muscle negative (tensilon test, repetitive nerve stimulation, (figure e-2, http://links.lww.com/NXG/A9). and acetylcholine receptor antibodies). Electromyog- Somatic mosaicism is the phenomenon of a single raphy showed only slight myopathic changes; the variant occurring in 2 or more populations of soma muscle creatine kinase level was elevated at 994 U/ cells in 1 individual and has been increasingly recog- L (range 0–160); and lactate levels were normal (1.1 nized as an important pathogenic mechanism in mmol/L; range 0.7–2.1). Muscle biopsy showed 20% genetic disease.5 Probable germline mosaicism of cytochrome c oxidase–deficient fibers, many of which dominant pathogenic TWNK and SLC25A4 variants showed increased succinate dehydrogenase reactivity, causing adult-onset PEO with multiple mtDNA de- typical of “ragged-blue” fibers (figure 1A). Southern letions have been previously described.6,7 By contrast, blotting of muscle DNA revealed multiple mtDNA we present evidence of somatic mosaicism for a path- deletions (figure 1B), which were also previously ogenic TWNK variant associated with this disorder. confirmed by a quantitative real-time PCR assay4; our Due to the low level of mosaicism (18%) in skeletal patient corresponds to subject M1. muscle, we speculate that the variant may have arose Candidate screening of nuclear genes associated during early embryogenesis,5 manifesting in a rela- with PEO and multiple mtDNA deletions (POLG, tively indolent clinical phenotype. Despite low-level

Neurology.org/ng Copyright © 2017 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1 Figure 1 Histopathologic, molecular and genetic characterization of the mosaic c.1121G.A p.(Arg374Gln) TWNK variant

(A) Diagnostic muscle biopsy was subjected to (A.a) H&E staining in addition to histochemical reactions for (A.b) NADH-tetrazolium reductase, (A.c) SDH, and (A.d) COX. Asterisk denotes COX-deficient, SDH-reactive fibers. Images were taken at 310 magnifi- cation. Scale bar (solid black line) denotes 100 mM. (B) Southern blotting of skeletal muscle DNA using a PCR-generated D-loop (a-32P dCTP-labeled) probe, demonstrating multiple mtDNA deletions in the patient (lane 3). Also shown are wild-type DNA (lane 1), a patient with a single, large-scale mtDNA deletion (lane 2) and multiple mtDNA deletions in a patient harboring a pathogenic, heterozygous c.2864A.Gp.(Tyr955Cys)POLG variant (lane 4). (C) Sanger sequencing electropherograms also suggested mosa- icism in both (C.a) blood and (C.b) muscle DNA. COX 5 cytochrome c oxidase; SDH 5 succinate dehydrogenase.

mosaicism, the variant could be highly detrimental to diagnosis and the power of next-generation hexamer formation and helicase activity. Further- sequencing. more, diagnostic Sanger sequencing of TWNK ini- From the Wellcome Centre for Mitochondrial Research (E.W.S., R.L. tially failed to identify the mosaic variant, and J., S.A.H., E.L.B., A.M.S., D.M.T., G.S.G., R.W.T.), Institute of hence, genetic analysis was reported as negative. Neuroscience, The Medical School, Newcastle University, United Kingdom; Department of Molecular and Human Genetics (E.W. While autosomal dominant disorders have a 50% S.), Baylor College of Medicine, Houston, TX; NHS Highly Special- recurrence risk in offspring, mosaicism also presents ised Mitochondrial Diagnostic Laboratory (R.L.J., S.A.H., E.L.B., a considerable recurrence risk. For our patient, it was R.W.T.), Newcastle upon Tyne Hospitals NHS Foundation Trust, United Kingdom; Wellcome Centre for Mitochondrial Research (A. not possible to determine the extent of mosaicism in P.), Institute of Genetic Medicine, Newcastle University, United the germline, which provides further challenges for Kingdom; and Department of Clinical Neurosciences (P.F.C.), genetic counseling. Nonetheless, we provided School of Clinical Medicine, and MRC Mitochondrial Biology Unit a genetic diagnosis in a longstanding case, demon- (P.F.C.), University of Cambridge, United Kingdom. Author contributions: E.W.S.: analysis and interpretation of whole- strating the use of next-generation sequencing for exome sequencing, clinical and molecular data, drafting and revision detecting and quantifying low-level mosaicism. of the manuscript, and figure preparation. R.L.J. and S.A.H.: anal- In conclusion, we suggest that physicians should ysis and interpretation of targeted next-generation sequencing and revisit suspected adPEO patients with apparently diagnostic molecular genetic data. E.L.B.: analysis and interpretation of targeted next-generation sequencing and diagnostic molecular negative genetic testing of mtDNA maintenance genetic data. A.P.: analysis and interpretation of whole-exome disorder genes and to consider mosaicism. With sequencing data. A.M.S.: clinical care of the patient and revision expanded reproductive options and ongoing devel- of the manuscript. P.F.C.: study concept and design and analysis and interpretation of whole-exome sequencing data. D.M.T.: clinical care opment of therapeutic strategies, our case high- of the patient and revision of the manuscript. G.S.G. and R.W.T.: lights the importance of attaining a genetic study concept and design, analysis and interpretation of clinical,

2 Neurology: Genetics molecular genetic, and biochemical data, and drafting and revision of upon Tyne Foundation Hospitals NHS Trust. The Article Processing the manuscript. Charge was funded by the Newcastle University. Go to Neurology. Acknowledgment: The clinical and diagnostic mitochondrial service org/ng for full disclosure forms. in Newcastle upon Tyne is funded by the UK NHS Highly Speci- This is an open access article distributed under the terms of the alised Commissioners to provide the “Rare Mitochondrial Disorders Creative Commons Attribution License 4.0 (CC BY), which permits of Adults and Children” Service. unrestricted use, distribution, and reproduction in any medium, pro- Study funding: This study was funded by the Wellcome Centre for vided the original work is properly cited. Mitochondrial Research (203105/Z/16/Z) and a Medical Research Received June 8, 2017. Accepted in final form August 27, 2017. Council PhD studentship to E.W.S. Correspondence to Dr. Gorman: [email protected] Disclosure: E.W. Sommerville, R.L. Jones, S.A. Hardy, E.L. Blakely, A. Pyle, and A.M. Schaefer report no disclosures. P.F. Chinnery is 1. Sommerville EW, Chinnery PF, Gorman GS, Taylor RW. a Wellcome Trust Senior Fellow in Clinical Science (101876/Z/13/ Adult-onset mendelian PEO associated with mitochondrial Z) and a UK NIHR Senior Investigator, who receives support from – the Medical Research Council Mitochondrial Biology Unit disease. J Neuromuscul Dis 2014;1:119 133. (MC_UP_1501/2); in addition, he has served on the editorial board 2. Spelbrink JN, Li FY, Tiranti V, et al. Human mitochondrial of BRAIN. D.M. Turnbull is supported by the Wellcome Centre for DNA deletions associated with mutations in the gene en- Mitochondrial Research (203105/Z/16/Z), the Medical Research coding Twinkle, a phage T7 gene 4-like protein localized in Council Centre for Translational Research in Neuromuscular Disease mitochondria. Nat Genet 2001;28:223–231. Mitochondrial Disease Patient Cohort (UK) (G0800674), the Lily 3. Fratter C, Gorman GS, Stewart JD, et al. The clinical, “ Foundation, the UK NHS Highly Specialised Rare Mitochondrial histochemical, and molecular spectrum of PEO1 (Twin- ” Disorders of Adults and Children Service, and the UK NIHR Bio- kle)-linked adPEO. Neurology 2010;74:1619–1626. medical Research Centre for Ageing and Age-related disease award to 4. He L, Chinnery PF, Durham SE, et al. Detection and the Newcastle upon Tyne Foundation Hospitals NHS Trust. G.S. quantification of mitochondrial DNA deletions in individ- Gorman is supported by the Wellcome Centre for Mitochondrial Research (203105/Z/16/Z), the Medical Research Council Centre ual cells by real-time PCR. Nucleic Acids Res 2002;30:e68. for Translational Research in Neuromuscular Disease Mitochondrial 5. Ju YS, Martincorena I, Gerstung M, et al. Somatic muta- Disease Patient Cohort (UK) (G0800674), the Lily Foundation, the tions reveal asymmetric cellular dynamics in the early UK NHS Highly Specialised “Rare Mitochondrial Disorders of human embryo. Nature 2017;543:714–718. Adults and Children” Service, and the UK NIHR Biomedical 6. Deschauer M, Hudson G, Muller T, Taylor RW, Chinnery Research Centre for Ageing and Age-related disease award to the PF, Zierz S. A novel ANT1 gene mutation with probable Newcastle upon Tyne Foundation Hospitals NHS Trust. R.W. Tay- germline mosaicism in autosomal dominant progressive lor is supported by the Wellcome Centre for Mitochondrial Research external ophthalmoplegia. Neuromuscul Disord 2005;15: (203105/Z/16/Z), the Medical Research Council Centre for Trans- 311–315. lational Research in Neuromuscular Disease Mitochondrial Disease Patient Cohort (UK) (G0800674), the Lily Foundation, the UK 7. Hudson G, Deschauer M, Busse K, Zierz S, Chinnery PF. NHS Highly Specialised “Rare Mitochondrial Disorders of Adults Sensory ataxic neuropathy due to a novel C10Orf2 muta- and Children” Service, and the UK NIHR Biomedical Research tion with probable germline mosaicism. Neurology 2005; Centre for Ageing and Age-related disease award to the Newcastle 64:371–373.

Neurology: Genetics 3