ARTICLE OPEN ACCESS

Myasthenic congenital myopathy from recessivemutations at a single residue in NaV14Nathaniel Elia BS Johanna Palmio MD PhD Marisol Sampedro Castantildeeda PhD Perry B Shieh MD PhD

Marbella Quinonez BS Tiina Suominen PhD Michael G Hanna MD Roope Mannikko PhD

Bjarne Udd MD PhD and Stephen C Cannon MD PhD

Neurologyreg 201992e1405-e1415 doi101212WNL0000000000007185

Correspondence

Dr Cannon

sccannonmednetuclaedu

AbstractObjectiveTo identify the genetic and physiologic basis for recessive myasthenic congenital myopathy in 2families suggestive of a channelopathy involving the sodium channel gene SCN4A

MethodsA combination of whole exome sequencing and targeted mutation analysis followed by voltage-clamp studies of mutant sodium channels expressed in fibroblasts (HEK cells) and Xenopusoocytes

ResultsMissense mutations of the same residue in the skeletal muscle sodium channel R1460 ofNaV14 were identified in a family and a single patient of Finnish origin (pR1460Q) anda proband in the United States (pR1460W) Congenital hypotonia breathing difficultiesbulbar weakness and fatigability had recessive inheritance (homozygous pR1460W or com-pound heterozygous pR1460Q and pR1059X) whereas carriers were either asymptomatic(pR1460W) or had myotonia (pR1460Q) Sodium currents conducted by mutant channelsshowed unusual mixed defects with both loss-of-function (reduced amplitude hyperpolarizedshift of inactivation) and gain-of-function (slower entry and faster recovery from inactivation)changes

ConclusionsNovel mutations in families with myasthenic congenital myopathy have been identified atpR1460 of the sodium channel Recessive inheritance with experimentally established loss-of-function is a consistent feature of sodium channel based myasthenia whereas the mixed gain offunction for pR1460 may also cause susceptibility to myotonia

RELATED ARTICLE

EditorialMuscle at the junctionWhere next generationsequencing is sending us

Page 591

These authors contributed equally to this work

From the Departments of Physiology (NE MQ SCC) and Neurology (PBS) David Geffen School of Medicine at UCLA Molecular and Cellular Integrative Physiology Program atUCLA (NE SCC) Los Angeles CA Tampere Neuromuscular Center (JP TS BU) Tampere University and University Hospital Finland MRC Centre for Neuromuscular Diseases(MSC MGH RM) Department of Neuromuscular Disease UCL Institute of Neurology London UK Folkhalsan Genetic Institute (BU) Helsinki andNeurologyDepartment (BU)Vasa Central Hospital Finland

Go to NeurologyorgN for full disclosures Funding information and disclosures deemed relevant by the authors if any are provided at the end of the article The Article ProcessingCharge was funded by Medical Research Council

This is an open access article distributed under the terms of the Creative Commons Attribution License 40 (CC BY) which permits unrestricted use distribution and reproduction in anymedium provided the original work is properly cited

Copyright copy 2019 The Author(s) Published by Wolters Kluwer Health Inc on behalf of the American Academy of Neurology e1405

Several allelic disorders of skeletal muscle are caused bymutations of SCN4A that encodes the pore-forming α subunitof the voltage-gated sodium channel (NaV14)

1 Missensemutations with gain-of-function changes (GOF too muchinward Na+ current) are found in hyperkalemic periodic pa-ralysis (HyperPP) paramyotonia congenita and several var-iants of sodium channel myotonia2 Leaky channels resultingfrom mutations of arginine residues in the voltage sensordomain cause hypokalemic periodic paralysis (HypoPP)type 234 These traits are all dominantly inherited

Loss-of-function (LOF) mutations of SCN4A are encoun-tered far less frequently and are associated with recessivelyinherited phenotypes A congenital myasthenic syndromewith ptosis bulbar weakness respiratory difficulties andprolonged episodes of weakness more typical for periodicparalysis has been associated with missense mutations ofSCN4A that cause a LOF by markedly enhancing channelinactivation5ndash7 More recently congenital myopathy withneonatal hypotonia has been reported in patients with nullmutations in SCN4A8 A homozygous null is embryonic le-thal while compound heterozygous mutations (null alleleplus an LOF allele) result in congenital myopathy with sur-vival to adulthood Remarkably family members with a singleSCN4A null allele are healthy

In this report we describe the molecular and clinical con-sequences of 2 additional LOFmutations both of which are atresidue p1460 The index cases presented with congenitalhypotonia respiratory difficulties and delayed motor mile-stones plus fatigue and were found to have biallelic mutationsas either pR1460Q plus pR1059X or homozygous pR1460WExpression studies of the pR1460 mutant channels alsorevealed GOF changes that account for the myotonia in somecarriers of pR1460Q Moreover the phenotype for some car-riers of the pR1460Q mutation in the primary Finnish familywas complicated by the independent cosegregation of a knownCLCN1mutation pR894X associated with recessive myotoniacongenita

MethodsClinical examinationThe proband (III-3) and 6 of her relatives were examined ina Finnish family (F1 figure 1A) In addition one of her aunts(II-6) and hermaternal grandfather (I-1) had similar symptoms(larynx spasms) but were not available for further studiesThe patients underwent neurologic examination EMG andDNA extraction Further a single unrelated Finnish patient(P2) with myotonia was similarly examined Muscle histology

was available for 2 patients and muscle MRI for 1 patient Theproband from family 2 and her parents were examined neu-rologically and whole blood was collected for DNA analysis

Clinical electrophysiologyStandard neurography and EMG investigation was performedin 9 patients with pR1460Q mutation Compound muscleaction potential (CMAP) exercise test was carried out in 3patients A Fournier protocol was used with short (10ndash12seconds) and long (5 minutes) exercise test910 CMAPs wereevoked by supramaximal nerve stimulation The proband offamily 1 also underwent repetitive nerve stimulation at 30 Hzand single-fiber jitter examinations The proband of family 2was studied by needle EMG and repetitive nerve stimulation at3 Hz and 50 Hz Because cooperation was limited in a 6-year-old patient she did not complete a CMAP exercise test

Molecular geneticsThe DNA of the proband in family 1 was studied by wholeexome sequencing The DNA sample was used for enrichmentof a sequencing library by NimbleGen SeqCap EZ HumanExome v20 (Roche Basel Switzerland) The captured librarywas sequenced using an Illumina (San Diego CA) HiSeq 2000Sequencer Sequence reads were aligned to the human refer-ence genome (UCSC hg19) using the Burrows-WheelerAligner11 Variant calling was made with Genome AnalysisToolkit12 The data were visualized with the IntegrativeGenomics Viewer (Broad Institute Cambridge MA) DNAcapture enrichment next-generation sequencing and dataanalysis were performed at the Institute for Molecular Medi-cine Finland From the sequencing data we selected to furtheranalyze all the exons and exonndashintron borders of the followinggenes MTM1 RYR1 DNM2 SEPN1 and SCN4A For theother members of family 1 the sequencing of relevant exons ofSCN4A and CLCN1 genes was performed using traditionalSanger sequencing (table) For some of the patients 3 com-mon Finnish mutations in the CLCN1 gene (c1238TgtGpF413C c1592CgtT pA531V and c2680CgtT pR894X)were studied by targeted mutation detection using TaqManSequence Detection System (ABI Prism 7000 Applied Bio-systems Foster City CA) TaqMan Sequence Detection Sys-tem is based on fluorescent oligonucleotide probes for bothmutation and normal sequences The studied fragments wereamplified by PCR using forward and reverse primersSequences of the primers and probes are available upon re-quest DNA of patient P2 was sequenced using MYOcap tar-geted gene panel13 Whole exome sequencing was performedfor the proband in family 2 and her parents using the Agilent(Santa Clara CA) SureSelect capture kit and an IlluminaHiSeq2500 next-generation sequencer Analysis of variants wasperformed by the UCLA Clinical Genomics Center

GlossaryCMAP = compound muscle action potential CMS = congenital myasthenic syndrome GOF = gain-of-function HyperPP =hyperkalemic periodic paralysis HypoPP = hypokalemic periodic paralysis LOF = loss-of-function WT = wild-type

e1406 Neurology | Volume 92 Number 13 | March 26 2019 NeurologyorgN

Functional assessment of mutantsodium channelsSodium channels were transiently expressed in fibroblasts(HEK cells) as previously described14 except transfection wasperformed with lipofectamine Cells expressed the humanwild-type NaV14 pore-forming α subunit (WT) or mutantconstructs encoding R1460Q or R1460W plus the β1 acces-sory subunit Sodium currents were recorded using the wholecell voltage-clamp configuration The electrode contained thefollowing in mM 100 CsF 35 NaCl 5 EGTA 10HEPES pHto 73 with CsOH the bath solution contained the followingin mM 140 NaCl 4 KCl 2 CaCl2 1 MgCl2 25 glucose 10HEPES pH to 73 with NaOH After whole-cell access wasachieved cells were allowed to equilibrate for 5 to 10 minutesbefore recording For the dataset to determine the voltagedependence of gating cells with maximal peak Na+ currentslt1 nA were excluded to minimize the contribution from en-dogenous Na+ currents (typically lt01 nA) and those withpeak current gt5 nA were excluded tominimize series resistanceerrors To reduce the selection bias for the current densitymeasurements (figure 2) cells with peak Na+ currents from05 nA to 10 nA were included For the experiments to test

for gating pore currents channels were expressed in Xenopusoocytes and currents were recorded using 2-electrode voltageclamp as previously described8 The bath solution containedthe following in mM 60 Na+-methansulfonate 60 guanidinesulfate 18 CaSO4 10 HEPES plus 1 μm tetrodotoxin to blocksodium currents conducted by the central core

The functional properties of WT and mutant sodium channelswere compared quantitatively by obtaining parameter estimatesto functions describing the voltage-dependent changes of theionic currents Channel activation was quantified by fitting thepeak amplitude (Ipeak) to a linear conductance (Gmax) witha reversal potential (Erev) that was scaled with a Boltzmannfunction Ipeak = GmaxethV minus ErevTHORN=eth1 + eminusethVminus V12THORNKTHORN Thevoltage-dependence for activation of the channel is character-ized by V1=2 the voltage at which half the channels are acti-vated and K a steepness factor The relative conductance(figure 2C) was calculated as Ipeak divided by GmaxethV minus ErevTHORNThe time constant for entry to inactivation was estimatedfrom a single exponential fit of the current decay (fast in-activation) or of the change in peak current after pro-gressively longer conditioning pulses (closed-state inactivation)

Figure 1 Sodium channel mutations

(A) Segregation of clinical phenotype and ge-notype among 7 carriers of pR1460Q in family1 from Finland (B) Location of pR1460 in thepore-forming subunit (NaV14) along withestablished sites for sodium channelopathiesof skeletal muscle CMS = congenital myas-thenic syndrome HyperPP = hyperkalemicperiodic paralysis HypoPP = hypokalemic pe-riodic paralysis PAM = paramyotonia con-genita SCM = sodium channel myotonia

NeurologyorgN Neurology | Volume 92 Number 13 | March 26 2019 e1407

Table Phenotype and genotype of the patients

Patient SCN4A mutations CLCN1 mutation Symptoms Electrophysiology Other

F1 II-2 c4379 GgtA pR1460Qminus (ex 24a)

minusminusa Myotonia larynx spasms Myotonic discharges

F1 II-3 c4379 GgtA pR1460Qminus (ex 17 21 22 24a)

c2680CgtT pR894Xminus(ex 2 11 15 23)

Stiffness larynx spasms exercise-induced muscle weakness Myotonic discharge decrease of CMAP amplitudeafter long exercise test

F1 II-4 c3175CgtT pR1059Xminus(ex 17)

minusminus (ex 2 11 15 23) Possibly mild ptosis ND

F1 II-5 c4379GgtA pR1460Qminus (ex 24a)

c2680CgtT pR894Xminus(ex 2 11 15 23)

Exercise-induced myalgia stiffness cramps cold-induced myotoniain hands

Myotonic discharges

F1 II-7 c4379GgtA pR1460Qminus (ex 21 22 24a)

minusminus (ex 2 11 15 23) Cold-induced myotonia periodic paralysis during adolescence Increased insertional activity

F1 III-1 c4379GgtA pR1460Qminus (ex 22 24a)

minusminusa (ex 2) Myotonia Myotonic discharges

F1 III-2 c4379GgtA pR1460Qminus (ex 21 22 24a)

minusminus (ex 2 11 15 23) Exercise-induced myalgia stiffness cramps larynx spasms Increased insertional activity normal findings inCMAP exercise test

Normal muscle MRI musclebiopsy mild abnormalities

F1 III-3 c4379GgtA pR1460Qc3175CgtT pR1059Xb

c2680CgtT pR894Xminusb

(ex 2)Congenital hypotonia motor milestones delayed mild fixed ptosislarynx spasms muscle fatigue ambulant stable during adulthood

Myotonic discharges 40 decrease of CMAPamplitude after long exercise test single fibermyasthenic

Muscle biopsy myopathy

P2 II-1 c4379GgtA pR1460Qminusc

minusminusc Myalgia stiffness cramps Myotonic discharges Targeted gene panel noother pathogenic mutations

F2 II-1 c4378CgtT pR1460Wc4378CgtT pR1460W

minusminus Congenital hypotonia gastrostomy and intubation at infancy motordelay respiratory distress

No myotonic discharges repetitive stimulation nodecrement (3 Hz) or increment (50 Hz)

Parents are carriers andasymptomatic

Abbreviations CMAP = compound muscle action potentials ex = exons studied nd = not donea Three common mutations in CLCN1 gene in Finland were studied by targeted mutation detection using TaqMan Sequence Detection Systemb Studied using whole exome sequencingc Myocap gene panel ref 13Exons of SCN4A and CLCN1 genes were studied by Sanger sequencing unless otherwise indicated Exon 24 of SCN4A was sequenced only partly which is indicated by marking ldquo24ardquo

e1408Neu

rology

|Vo

lume92N

umber

13|

March

262019NeurologyorgN

Steady-state fast inactivation was quantified by fitting therelative peak current at minus10 mV after a 300 ms conditioningpulse (Vcond) to a Boltzmann function IpeakethVcondTHORN=Ipeak max =1=eth1 + eethVcondminusV12THORNKTHORN For slow inactivation a conditioningdepolarization was used to induce inactivation then a re-covery pulse at minus120 mV for 20 ms was applied to remove fastinactivation and finally the available current (ie not slow-inactivated) was measured for a test pulse to minus10 mV For thesteady-state voltage dependence a plateau term (S0) was in-cluded because slow inactivation does not reduce channelavailability to 0 at strongly depolarized potentials Estimatedvalues for parameters are presented as mean plusmn SEM Errorbars in all graphs show plusmn SEM

Data availability statementPhotomicrographs of the muscle biopsy from the proband infamily 1 are available upon reasonable request The expressionplasmids for the functional studies of mutant sodium channelsare available upon request from the corresponding author

Standard protocol approvals registrationsand patient consentsInformed written consent for genetic testing was obtainedfrom patients or their guardians Oocytes were harvested fromadult female Xenopus laevis in accordance with the UK Animal(Scientific Procedures) Act 1986

ResultsClinical presentation and assessment

Family 1The proband (III-3) a 30-year-old woman had congenitalhypotonia that was severe at birth causing difficulties inbreathing and feeding but resolved to milder nonprogressivemuscle weakness and fatigue during the first years of life Themost severe symptoms resolved quickly in the first weeks oflife therefore she did not need ventilator support or tubefeeding Motor milestones were delayed independent sittingwas achieved at 12 months but she started to walk at 18months of age From the start she had severe weakness in theneck and facial muscles mild dysphonia and fixed ptosis bi-laterally Otherwise muscle strength was 45 on the MedicalResearch Council Scale in all limbs She had marked fatiga-bility when climbing stairs otherwise there was no largevariability in her muscle strength At age 30 years she wasambulant and could walk on toes but not on heels Clinicalmyotonia was not detected

Muscle biopsy showed myopathic changes with fiber sizevariation increased internal nuclei structural abnormalitiesand necklace fibers Some non-rimmed vacuolated fibers wereseen Several EMG studies were performed and one showed

Figure 2 Sodium current density was reduced for cells transfected with R1460 mutant constructs but the voltage de-pendence of activation was not altered

(A) Traces show sodium currents elicited by depolarization from minus120mV to test potentials between minus75 and +60mV Note the higher gain for the scale bars inthe mutant channel traces (B) Peak current density is shown as a function of test potential and displays a marked reduction in current amplitude for bothmutants (C) The voltage dependence of channel activation as shown by relative conductance was indistinguishable between the wild-type (WT) andmutantchannels Black squares blue triangles and red circles correspond to WT (n = 16) R1460Q (n = 12) and R1460W (n = 16) respectively

NeurologyorgN Neurology | Volume 92 Number 13 | March 26 2019 e1409

definite myotonic discharges Single fiber test was myasthenicwith increased jitter and the long-exercise test showed 40decrease of CMAP amplitude The proband had 2 mutationsin the SCN4A gene 1 missense pR1460Q inherited from themother and a nonsense pR1059X inherited from the fatherIn addition she had inherited a heterozygous CLCN1 mutationpR894X from her mother (figure 1) Whole exome sequencingdid not reveal any other variants thatmight complicate the resultsGenes associated with channelopathies (SCN4A CACNA1SKCNJ2 CLCN1) or congenital myasthenic syndrome and re-lated disorders (CHRNA1 CHRNB1 CHRND CHRNERAPSN CHAT MUSK COLQ DOK7 AGRN GFPT1DPAGT1 LAMB2 CHRNG PLEC) were specifically checked

The probandrsquos mother 3 siblings of the mother and 2 cousinswere identified with the pR1460Q mutation In addition oneunrelated patient (P2) carried the same mutation They allhad myotonic muscle symptoms and signs starting duringyoung adulthood that either showed myotonic discharges orincreased insertional activity on EMG and clinical myotoniaor related to muscle stiffness exercise-induced myalgia andcramps (table) Most of the pR1460Q carrying patients infamily 1 had larynx spasms that were evident in cold weatheror at night TheCLCN1 pR894Xmutation was also identifiedin the mother and one of her siblings They had myotonicdischarges on EMG but that was also seen in other siblingswithout CLCN1 mutation The father who carried the non-sense SCN4A mutation found in the proband possibly hadmild fixed ptosis but was otherwise healthy

Family 2The proband was born with hypotonia She exhibited poorfeeding and respiratory insufficiency requiring gastrostomy aswell as intubation and mechanical ventilation She was hospi-talized for the first month of life Motor milestones wereattained including standing and walking albeit delayed and theparents reported day-to-day variability in strength and re-current breathing difficulty Around 6 years of age she de-veloped worseningmotor function and intermittent respiratorydistress and was referred to the UCLA Neuromuscular Clinicfor further evaluation Her examination at that time demon-strated significant weakness of axial and appendicular musclesShe was not able to stand or support her head while in a seatedpositon Reflexes were present although mildly diminishedRepetitive stimulation of the median and accessory nerves at 3Hz did not demonstrate a decrement in CMAP amplitude and50 Hz repetitive nerve stimulation of the median nerve did notelicit significant increment No myotonic discharges weredetected by needle EMG Motor performance varied sponta-neously and on subsequent visits over the next 18 months shewas able to stand and walk independently She did not requirebilevel positive airway pressure at night or other chronic ven-tilatory support The patientrsquos mother reported improvementof strength and breathing with inhaled albuterol but the benefitwas not sustained on oral therapy A trial of pyridostigmine didnot improve strength but the family reported improved mo-bility when acetazolamide and albuterol were tried separately

The parents have no motor symptoms An older sister re-portedly had bulbar weakness was never able to sit withoutsupport and died at age 1 year from respiratory complicationsThe proband is homozygous for c4378CgtT pR1460W inSCN4A and both parents are asymptomatic carriers althoughthere is no history of consanguinity in the family

Molecular resultsThe NaV14 variants in these families cause missense sub-stitutions at the same amino acid arginine 1460 which is inthe highly conserved S4 transmembrane segment in domainIV of the sodium channel (figure 1B) The allele frequency forpR1460W in the gnomAD database is 6 per 276796 and thepR1460Q allele has not been reported The S4 segments inthe voltage sensors contain positively charged basic aminoacids (arginine or lysine) at every third residue15 and R1460is the fifth of 8 such residues in S4 of domain IV Althoughmutation at R1460 in NaV14 has not previously beenreported in neuromuscular disorders all 3 SCN4A mutationsassociated with a congenital myasthenic syndrome and forwhich myopathy with fluctuating weakness is a prominentfeature are missense mutations of the voltage sensor in do-main IV and 2 are at other arginines in the S4 segment(pR1454W and pR1457H)5ndash7 All 3 previously reportedmutations are also unusual because they cause LOF changesfor the sodium channel and recessive inheritance of the syn-drome whereas GOF changes are found in SCN4Amutationsassociated with myotonia paramyotonia or HyperPP1 Wetherefore tested the functional consequences of pR1460Qand pR1460W by recording sodium currents for channelsheterologously expressed in fibroblasts or Xenopus oocytes

Functional characterization of R1460 mutantsodium channels

R1460 mutant sodium channels were expressed in theplasma membrane but current density was reducedSodium currents were detected in HEK cells transfected withWT as well as R1460W and R1460Q mutant channels (figure2) Peak current amplitudes were smaller however for cellsexpressing mutant channels The peak current density (pApF to normalize for variations in cell size figure 2B) for cellsexpressing mutant channels was about 30ndash50 that of WTAt a test potential of ndash 5 mV where the sodium current islargest the amplitude was statistically smaller for each mutantcompared to WT (p lt 001 analysis of variance) but thedifference between the 2 mutants was not distinguishable(R1460W minus738 plusmn 11 pApF [n = 9] R1460Q minus140 plusmn 26 pApF [n = 7] WT minus278 plusmn 46 pApF [n = 12])

Channel activation was not altered by R1460Qor R1460WThe voltage dependence of sodium channel activation isshown in figure 1C where the relative conductance is plottedas a function of the test potential (see Methods) The dataare completely overlapping between WT and R1460 mutantchannels Quantitatively there was no detectable difference inthe voltage midpoint of activation with V12 of WT R1460Q

e1410 Neurology | Volume 92 Number 13 | March 26 2019 NeurologyorgN

and R1460W being minus187 plusmn 11 mV (n = 16) minus176 plusmn 17 mV(n = 12) and minus190 plusmn 14 mV (n = 16) respectively Inaddition the steepness of the voltage dependence was similarfor all 3 WT R1460Q and R1460W being 77 plusmn 022 mV 91plusmn 034 mV and 92 plusmn 023 mV respectively

Fast inactivation had mixed LOF and GOF defects forR1460 mutant channelsSodium channels undergo 2 forms of inactivation fast in-activation on a time scale of milliseconds which contributes totermination of the action potential and limits repetitive firingand slow inactivation over a time course of tens of secondsto minutes To test the steady-state voltage dependenceof fast invitation (also referred to as availability) we mea-sured the peak current elicited at minus10 mV after a 300 ms

conditioning pulse to potentials over a range from minus130 tominus40 mV (figure 3A) Depolarization promotes inactivation(reduces availability) and the voltage dependence wasmarkedly left-shifted for R1460 mutant channels (V12

value for R1460Q minus781 plusmn 067 mV n = 10 R1460W minus861plusmn 15 mV n = 12 and WT minus705 plusmn 091 mV n = 15) Inaddition the steepness of the voltage dependence ofsteady-state invitation was reduced for R1460 mutantchannels (k value for WT 55 plusmn 088 mV R1460Q 86 plusmn 02mV R1460W 77 plusmn 02 mV) Combined these changes inthe voltage dependence of steady-state fast inactivationproduce a substantial LOF as can be seen by the reducedavailability (figure 3A) of mutant channels (R1460Q = 07R1460W = 05 compared to WT = 09) at the resting po-tential of minus85 mV for skeletal muscle

Figure 3 Fast inactivation of R1460 mutant channels had both gain-of-function and loss-of-function defects

(A) The steady-state voltage dependence of fast inactivationwas shifted leftward (hyperpolarized) for bothR1460mutant channels relative towild-type (WT) Insetshows the voltage protocol used tomeasure inactivation produced by a 300ms conditioning pulse at variouspotentials (B) The rateof inactivationwas slower forR1460 mutant channels at depolarized potentials as shown by the superposition of amplitude normalized currents elicited at 10 mV (C) Recovery from fastinactivationwas faster for bothR1460QandR1460W compared toWTchannels (tau recovery of 47plusmn 06msplt 00001 [n= 5] 64 plusmn 06msplt 00001 [n= 7] 168plusmn 04 ms [n = 4] respectively) Data show the time course for the recovery of peak current amplitude at a holding potential of minus80 mV after channels wereinactivatedwitha conditioningpulseof 30msatminus10mV (inset) (D) Plot summarizing the voltage-dependent kinetics for entry toor recovery from fast inactivationThree separate protocols were used to measure inactivation kinetics over the entire voltage range (see text Results) Overall the changes in steady-state fastinactivation (A) are loss of function whereas the slower entry and more rapid recovery from fast inactivation are gain-of-function changes

NeurologyorgN Neurology | Volume 92 Number 13 | March 26 2019 e1411

The voltage-dependent kinetics of fast inactivation wereassessed for both entry to and recovery from inactivation Theentry rate was slower for R1460 mutant channels which can beobserved directly as a prolonged decay in a superposition ofsodium currents elicited by a step depolarization to 10 mV(figure 3B) This slower entry rate was observed over a range oftest potentials (minus20 to +60 mV) as shown by the larger am-plitude time constant of the current decay (figure 3D) Re-covery from fast inactivation was recorded using a 2-pulseprotocol Channels were fast inactivated using a 30 ms condi-tioning pulse to minus10mV then after a variable duration recoveryat a hyperpolarized potential (minus80 to minus130 mV) a test pulse tominus10 mV was delivered to determine the relative amplitude ofthe peak sodium current (figure 3C inset) The time course ofrecovery at minus80 mV is shown in figure 3C and demonstratesthe faster recovery (left shift) for R1460 mutant channels Afaster recovery rate (smaller time constant) of R1460 mutantchannels was observed over a potential range from minus50 to minus130mV (figure 3D) Finally the kinetics of closed-state fast in-activation (figure 3D minus40 to minus70mV) was measured usinga 2-pulse protocol with a 30 ms conditioning pulse to induceinactivation followed by a test pulse at minus10 mV to measurethe relative current The altered entry and recovery kinetics forfast inactivation of R1460 mutants both contribute to a GOFchange in other words more inward sodium current In thephysiologic context of an action potential mutant channelsare predicted to inactivate more slowly at the peak of de-polarization and then recover more quickly when the mem-brane potential returns to its resting value (minus85 mV)

The impact of the altered kinetics of fast inactivation can beobserved directly by measuring the peak amplitudes of so-dium currents elicited by a train of brief depolarizing pulsesNormally sodium channels do not have time to fully recover

in the interval between action potentials with high-frequencydischarges We simulated this scenario by measuring sodiumchannel availability as the relative peak current elicited bya 2 ms depolarization to 10 mV with an intervening recov-ery interval at minus80 mV over a range of pulse frequencies(figure 4) Use-dependent inactivation was pronounced forWT channels (eg 30 decrease at 60 Hz and 50 decreaseat 100 Hz) but was substantially less for R1460 mutantchannels These data demonstrate a GOF for R1460 mutantchannels that would manifest for example as myotonicbursts

Slow inactivation was unaffected for bothR1460Q and R1460W mutant channelsProlonged depolarizations lasting seconds or sustained burstsof discharges both cause sodium channels to become slowinactivated and derangements of slow inactivation are knownto cause susceptibility to periodic paralysis16 or congenitalmyasthenia6 We characterized slow inactivation by usingprolonged step depolarizations (up to 60 seconds) to induceslow inactivation followed by a 20ms recovery period at minus120mV to remove fast inactivation and then applied a test pulseto minus10 mV to detect the presence of slow inactivation asa reduced peak current The time course of entry to slowinactivation at minus10 mV was indistinguishable between WTand R1460 mutant channels (figure 5A) Similarly the rate ofrecovery was comparable for WT and R1460 mutant channelsat minus100 mV (figure 5B) The small rightward shift representsabout a 15-fold slower rate of recovery for R1460 mutantsalthough the fitted time constants were not statistically dis-tinguishable from those of WT channels The voltage de-pendence of steady-state slow inactivation as determined by30-second conditioning pulses was identical for WT andR1460 mutant channels (figure 5C)

Figure 4 The use-dependent reduction in sodium current peak amplitude was less pronounced in R1460 mutants than inwild-type (WT) channels

Data show relative peak sodium current elicited by 2 ms depolarizing pulses to 10 mV applied at 60 Hz (A) or 100 Hz (B) from a holding potential of minus80 mV

e1412 Neurology | Volume 92 Number 13 | March 26 2019 NeurologyorgN

R1460 mutant channels do not conducta gating pore leakage currentMissense mutations at arginine residues in the S4 segments ofthe voltage sensor domains of sodium or calcium channelsaccount for almost all cases of HypoPP3 and have a commonfunctional defect the gating pore leakage current caused bya mutation-induced anomalous ion conduction pathway4

Since R1460 is an arginine in the S4 segment of domain IV inNaV14 (figure 1B) we tested whether R1460Q or R1460Wmutant channels have a detectable gating pore leakage cur-rent Channels were studied in Xenopus oocytes to achievehigh expression levels sufficient to observe gating pore cur-rents when the main pore was blocked by tetrodotoxinCurrents recorded from oocytes expressing R1460Q orR1460W channels were no different from WT (figure 6) ascompared to the positive control with the HypoPP mutationR1132Q for which a large gating pore current was detected(inward current a negative potentials)

DiscussionThis report identifies a new residue in NaV14 (R1460) wheremissense mutations were identified in patients with recessivemyasthenic congenital myopathy The clinical manifestationsoverlap considerably with prior reports of a recessive congenitalmyasthenia syndrome plus periodic paralysis that is associatedwith LOF mutations in SCN4A5ndash7 The consistent features arehypotonia at birth respiratory distress and feeding difficultiesoften requiring tube feeds Motor function at infancy and earlychildhood is notable for delayed milestone attainments fixedptosis myopathic weakness especially of the face and neckepisodes of hypoventilation myalgia and fatigability

Figure 5 Slow inactivation was not altered by eitherR1460W or R1460Q

(A) Onset of slow inactivation showed identical kinetics betweenwild-type (WT)and mutant channels Inset shows the voltage protocol to characterize theonset of slow inactivation by stepping to minus10 mV for a variable duration (entrytime) and then measuring the decline in relative loss current that fails to re-cover within 20 ms at minus120mV (B) The rate of recovery from slow inactivationwascomparablebetweenWTandR1460mutant channels Channelswere slowinactivated by a 30-second step to minus10 mV (inset) and the data show recoveryas the relative increase in current after repolarizing to minus80 mV for a variableduration (recovery time) (C) The voltage dependence of steady-state slow in-activation was indistinguishable between WT and R1460 mutant channelsThirty-second conditioning pulses to various conditioning potentials (inset)wereused todetermine thevoltagedependenceof inactivation Smooth curvesshow best fits to the data with V12 of minus591 plusmn 19mV for WT minus609 plusmn 28mV forR1460Q and minus634 plusmn 26 mV for R1460W The number of cells for the meanvalues shown in the data was 5ndash8 for WT and both R1460 mutant channels

Figure 6 The R1460 mutations did not cause a gating poreleakage current

The steady-state current (without subtraction of the nonspecific leak)recorded from oocytes expressing high levels of wild-type (WT) or R1460mutant channels is plotted as a function of the membrane potential Te-trodotoxin was added to block sodium currents through themain pore TheIndashV relation was identical for WT and R1460 mutant channels while thepositive control for the pR1132Q mutation in hypokalemic periodic paral-ysis showed a gating pore leakage current (large negative current formembrane potentials more negative than minus80 mV) Data are from 6 to 8oocytes per construct

NeurologyorgN Neurology | Volume 92 Number 13 | March 26 2019 e1413

Electrophysiologic signs of myasthenia are variable The firstpublished report (pS246LpV1442E) was notable for pro-nounced CMAP decrement during repetitive nerve stimula-tion (at 2 Hz after a 10 Hz load and at all frequencies of 10 Hzor higher)5 For the 2 subsequent single-case reports ofSCN4A-associated congenital myasthenic syndrome (CMS)one had increased jitter7 (as observed for patient F1 III-3herein) but neither had a CMAP decrement with repetitivestimulation Conversely in the series of 6 families with con-genital myopathy associated with SCN4A recessive mutations1 of 4 available surviving patients had a 60 CMAP decre-ment with 10 Hz stimulation8

A consistent pattern is emerging for the genotypendashphenotypeassociations with LOF mutations in SCN4A A single mutantallele whether a partial loss or complete null is asymptomaticalthough a mouse model revealed a latent susceptibility tomyasthenia17 A biallelic deficit with homozygous LOF muta-tions leads to moderate congenital myopathy with fatigue andvariable electrodiagnostic signs of myasthenia Compoundheterozygous mutations with a null plus a LOF defect causemoderate to severe congenital myopathy including an in-creased risk of infant mortality Finally biallelic null mutationswhether from nonsense mutations or missense mutations withno detectable sodium current are lethal in utero or at birth8

In hindsight this paradigm suggests the original report ofSCN4A-associated CMS was indeed caused by a recessivecompound heterozygote (pS246LpV1442E)5 Because thefast-inactivation defect for pV1442E was much greater (minus33mV shift) than for pS246L (minus73 mV) we initially consideredpS246L to possibly be a benign polymorphism In retrospectthe substantial enhancement of slow inactivation for pS246L(minus20 mV shift) will reduce NaV14 availability at Vrest by30 Recognizing now that a single LOF allele is asymp-tomatic and that pS246L does not appear in the gnomADdatabase strongly supports the notion that S246L is a func-tionally relevant LOF mutation

The clinical phenotypes associated with mutations at pR1460described herein were more varied which we attribute to themixed LOF and GOF changes observed for both R1460Q andR1460W mutant channels In the primary Finnish familyinheritance of myasthenic congenital myopathy was recessive(pR1460QpR1059X) but myotonia was a partially domi-nant trait Latent myotonia was detected by needle EMG inpR1460Q carriers and some had symptomatic myotoniaespecially with the coincident occurrence of an independentlysegregating mutation of the chloride channel gene CLCN1(pR894X associated with recessive myotonia congenita)Combined GOF mutations in SCN4A and LOF mutations inCLCN1 are known to synergistically aggravate the severity ofmyotonia1819 The LOF defects for R1460W channels weremore severe than those of R1460Q (smaller amplitude figure2 greater left shift of inactivation figure 3) whichmay explainwhy the parents shown to be pR1460W carriers did nothave myotonia Unusual clinical phenotypes have also beenreported for missense mutations of the second arginine in the

voltage sensor of domain IV (pR1451L and pR1451C)2021

In the heterozygous state pR1451L is associated with myo-tonia aggravated by cold and rare episodes of periodic paral-ysis typical for paramyotonia congenita which is consistentwith GOF defects manifest as slower entry and faster recoveryfrom inactivation An individual homozygous for pR1451Lhad more frequent and severe episodes of weakness one ofwhich with serum K+ 28 mM was suggestive of HypoPP21

Expression studies of R1451L did not reveal a gating poreleakage currentmdashthe canonical defect in HypoPPmdashbut in-stead showed LOF changes (reduced peak current and en-hanced inactivation) that may cause this unusual form ofperiodic paralysis in homozygous patients21 Unlike the casesherein however the LOF aspects of the mixed channeldefects for pR1451L did not result in a CMS syndrome in thehomozygous proband

Our report is the third case of a homozygous recessivemutation in SCN4A found in CMScongenital myopathyand all 3 missense mutations are at arginine residues in theS4 segment of the domain IV voltage sensor (pR1454W6pR1457H7 and pR1460W) Expression studies for all 3mutations show a similar pattern of LOF changes with ab-errant enhancement of inactivation consistent with theestablished structurendashfunction behavior of voltage-gated so-dium channels the domain IV sensor is coupled to in-activation22 whereas domains IndashIII control activation All 11mutations of NaV14 that cause HypoPP type 2 are alsomissense substitutions of arginines in S4 segments withinvoltage sensors13 These mutations all create anomalousldquogating porerdquo leakage currents that are mechanistically impli-cated in episodic attacks of depolarization and weakness inHypoPP4 The HypoPP mutations of S4 are in domains IndashIIIwhereas a homologous mutation in domain IV (pR1448C)causes paramyotonia congenita and does not create a gatingpore leak23 Scanning mutagenesis studies have shown thata missense substitution of any single arginine in S4DIV isinsufficient to create a gating pore leak24 and we have con-firmed herein that the pR1460 mutations found in recessivecongenital myopathy do not produce a leak (figure 6)

Further investigations are required to understand the mech-anism for the fatigue in CMScongenital myopathy caused bymutations in SCN4A Prolonged episodes of weakness remi-niscent of periodic paralysis are not likely to be caused bysustained fiber depolarization as with attacks of HyperPP orHypoPP because this clinical symptom occurs with CMSmutations that lack GOF changes (eg pR1454W6) Fatiguein this recessive phenotype is more likely to be a consequenceof impaired generation and conduction of action potentialscaused by a marginal sodium current density that may beexacerbated by use-dependent trapping of sodium channels inan inactive state

AcknowledgmentThe authors thank Dr Fenfen Wu for assistance with HEKcell recordings

e1414 Neurology | Volume 92 Number 13 | March 26 2019 NeurologyorgN

Study fundingThis work was supported by the NIH T32-GM065823(NE) NIAMS of the NIH AR42703 (SCC NE) UKMedical Research Council MRM0069481 UCLH Bio-medical Research Centre (MSC MGH RM) andTampere University Hospital Research Funds (JP BU)

DisclosureN Elia J Palmio M Sampedro-Castaneda P Shieh MQuinonez T Suominen M Hanna R Mannikko and B Uddreport no disclosures relevant to the manuscript S Cannon isa consultant to Strongbridge Biopharma Go to NeurologyorgN for full disclosures

Publication historyReceived by Neurology August 10 2018 Accepted in final formNovember 4 2018

References1 Cannon SC Sodium channelopathies of skeletal muscle Handb Exp Pharmacol 2018

246309ndash3302 Lehmann-Horn F Rudel R Jurkat-Rott K Nondystrophic myotonias and periodic

paralyses In Engel AG Franzini-Armstrong C eds Myology 3rd ed New YorkMcGraw-Hill 20041257ndash1300

3 Matthews E Labrum R Sweeney MG et al Voltage sensor charge loss accountsfor most cases of hypokalemic periodic paralysis Neurology 2009721544ndash1547

4 Cannon SC Voltage-sensor mutations in channelopathies of skeletal muscle J Physiol20105881887ndash1895

5 Tsujino A Maertens C Ohno K et al Myasthenic syndrome caused by mu-tation of the SCN4A sodium channel Proc Natl Acad Sci USA 20031007377ndash7382

6 Habbout K Poulin H Rivier F et al A recessive Nav14 mutation underliescongenital myasthenic syndrome with periodic paralysis Neurology 201686161ndash169

7 Arnold WD Feldman DH Ramirez S et al Defective fast inactivation recov-ery of Nav 14 in congenital myasthenic syndrome Ann Neurol 201577840ndash850

8 Zaharieva IT Thor MG Oates EC et al Loss-of-function mutations in SCN4Acause severe foetal hypokinesia or rsquoclassicalrsquo congenital myopathy Brain 2016139674ndash691

9 Fournier E Arzel M Sternberg D t al Electromyography guides towardsubgroups of mutations in muscle channelopathies Ann Neurol 200456650ndash661

10 Fournier E Viala K Gervais H et al Cold extends electromyography distinc-tion between ion channel mutations causing myotonia Ann Neurol 200660356ndash365

11 Li H Durbin R Fast and accurate short read alignment with Burrows-Wheelertransform Bioinformatics 2009251754ndash1760

12 McKenna A Hanna M Banks E et al The Genome Analysis Toolkit a MapReduceframework for analyzing next-generation DNA sequencing data Genome Res 2010201297ndash1303

13 Evila A Palmio J Vihola A et al Targeted next-generation sequencing reveals novelTTN mutations causing recessive distal titinopathy Mol Neurobiol 2017547212ndash7223

14 Hayward LJ Brown RH Jr Cannon SC Inactivation defects caused by myotonia-associated mutations in the sodium channel III-IV linker J Gen Physiol 1996107559ndash576

15 Gamal El-Din TM Lenaeus MJ Catterall WA Structural and functional analysis ofsodium channels viewed from an evolutionary perspective Handb Exp Pharmacol201824653ndash72

16 Hayward LJ Brown RH Jr Cannon SC Slow inactivation differs among mutant Nachannels associated with myotonia and periodic paralysis Biophys J 1997721204ndash1219

17 Wu F Mi W Fu Y Struyk A Cannon SC Mice with an NaV14 sodium channel nullallele have latent myasthenia without susceptibility to periodic paralysis Brain 20161391688ndash1699

18 Skov M Riisager A Fraser JA Nielsen OB Pedersen TH Extracellular magnesiumand calcium reduce myotonia in ClC-1 inhibited rat muscle Neuromuscul Disord201323489ndash502

19 Kato H Kokunai Y Dalle C et al A case of non-dystrophic myotonia with con-comitant mutations in the SCN4A and CLCN1 genes J Neurol Sci 2016369254ndash258

20 Poulin H Gosselin-Badaroudine P Vicart S et al Substitutions of the S4DIV R2residue (R1451) in NaV14 lead to complex forms of paramyotonia congenita andperiodic paralyses Sci Rep 201882041

21 Luo S Sampedro CastanedaMMatthews E et al Hypokalemic periodic paralysis andmyotonia in a patient with homozygous mutation pR1451L in NaV14 Sci Rep 201889714

22 Capes DL Goldschen-OhmMP Arcisio-Miranda M Bezanilla F Chanda B DomainIV voltage-sensor movement is both sufficient and rate limiting for fast inactivation insodium channels J Gen Physiol 2013142101ndash112

23 Francis DG Rybalchenko V Struyk A Cannon SC Leaky sodium channels fromvoltage sensor mutations in periodic paralysis but not paramyotonia Neurology2011761635ndash1641

24 Gosselin-Badaroudine P Delemotte L Moreau A Klein ML ChahineM Gating porecurrents and the resting state of Nav14 voltage sensor domains Proc Natl Acad SciUSA 201210919250ndash19255

Appendix Authors

Name Location Role Contribution

NathanielElia

UCLA LosAngelesCA

First author Study design recordedsodium currents wrotemanuscript

JohannaPalmio

TampereUniversityFinland

Co-first author Clinical assessmentwrote manuscript

MarisolSampedroCastaneda

UniversityCollegeLondon UK

Author Recorded sodiumcurrents

PerryShieh

UCLA LosAngeles CA

Author Clinical assessmentwrote manuscript

MarbellaQuinonez

UCLA LosAngelesCA

Author Site-directedmutagenesis editedmanuscript

TiinaSuominen

TampereUniversityFinland

Author Clinical assessment

MichaelHanna

UniversityCollegeLondon UK

Author Study design editedmanuscript

RoopeMannikko

UniversityCollegeLondon UK

Author Study design recordedsodium currents wrotemanuscript

BjarneUdd

TampereUniversityFinland

Author Study design clinicalassessment supervisedgenetic studies wrotemanuscript

StephenCannon

UCLA LosAngeles CA

Correspondingauthor

Study design dataanalysiswrotemanuscript

NeurologyorgN Neurology | Volume 92 Number 13 | March 26 2019 e1415

DOI 101212WNL0000000000007185201992e1405-e1415 Published Online before print March 1 2019Neurology

Nathaniel Elia Johanna Palmio Marisol Sampedro Castantildeeda et al 14

VMyasthenic congenital myopathy from recessive mutations at a single residue in Na

This information is current as of March 1 2019

ServicesUpdated Information amp

httpnneurologyorgcontent9213e1405fullincluding high resolution figures can be found at

References httpnneurologyorgcontent9213e1405fullref-list-1

This article cites 23 articles 8 of which you can access for free at

Citations httpnneurologyorgcontent9213e1405fullotherarticles

This article has been cited by 3 HighWire-hosted articles

Subspecialty Collections

httpnneurologyorgcgicollectionmuscle_diseaseMuscle disease

httpnneurologyorgcgicollectionion_channel_gene_defectsIon channel gene defectsfollowing collection(s) This article along with others on similar topics appears in the

Permissions amp Licensing

httpwwwneurologyorgaboutabout_the_journalpermissionsits entirety can be found online atInformation about reproducing this article in parts (figurestables) or in

Reprints

httpnneurologyorgsubscribersadvertiseInformation about ordering reprints can be found online

ISSN 0028-3878 Online ISSN 1526-632XWolters Kluwer Health Inc on behalf of the American Academy of Neurology All rights reserved Print1951 it is now a weekly with 48 issues per year Copyright Copyright copy 2019 The Author(s) Published by

reg is the official journal of the American Academy of Neurology Published continuously sinceNeurology