www.thelancet.com/neurology Vol 13 May 2014 503
Distinct neurological disorders with ATP1A3 mutations
Erin L Heinzen, Alexis Arzimanoglou, Allison Brashear, Steven J Clapcote, Fiorella Gurrieri, David B Goldstein, Sigurður H Jóhannesson,
Mohamad A Mikati, Brian Neville, Sophie Nicole, Laurie J Ozelius, Hanne Poulsen, Tsveta Schyns, Kathleen J Sweadner, Arn van den Maagdenberg,
Bente Vilsen, for the ATP1A3 Working Group
Genetic research has shown that mutations that modify the protein-coding sequence of ATP1A3, the gene encoding the
α3 subunit of Na+/K+-ATPase, cause both rapid-onset dystonia parkinsonism and alternating hemiplegia of childhood.
These discoveries link two clinically distinct neurological diseases to the same gene, however, ATP1A3 mutations are,
with one exception, disease-specifi c. Although the exact mechanism of how these mutations lead to disease is still
unknown, much knowledge has been gained about functional consequences of ATP1A3 mutations using a range of in-
vitro and animal model systems, and the role of Na+/K+-ATPases in the brain. Researchers and clinicians are attempting
to further characterise neurological manifestations associated with mutations in ATP1A3, and to build on the existing
molecular knowledge to understand how specifi c mutations can lead to diff erent diseases.
The introduction of next-generation sequencing tech-
nology, which allows nearly complete assessment of
human exomes and genomes, has signifi cantly boosted
gene discovery.1 These gene discoveries have advanced our
knowledge of disease-specifi c pathophysiology, and have
enabled genetic connections to be made between diseases.
An example is the recent identifi cation of mutations in
ATP1A3, a gene previously associated with rapid-onset
dystonia parkinsonism (RDP), in alternating hemiplegia
of childhood (AHC). This genetic connection between
diseases off ers unique research opportunities to explore
how genetics governs specifi c clinical phenotypes, and to
study the underlying pathophysiology.
Here, we review current knowledge of the roles of
ATP1A3 in the brain, as well as the phenotypes associated
with mutations in the gene. Finally, we discuss additional
research needed to further characterise the full
phenotypic spectrum associated with mutations in
ATP1A3 and the associated changes in protein function,
to establish possible strategies for the development of
new treatments for AHC, RDP, and related disorders.
Roles of Na+/K+-ATPases
Na+/K+-ATPases are membrane-bound transporters that
harness the energy of ATP hydrolysis to move three Na+
out of the cell in exchange for two K+ ions moving
inwards. The resulting ionic gradients establish
membrane potentials that generate electrical impulses
and move neurotransmitters and Ca2+ across the plasma
membrane. Na+/K+-ATPases consist of catalytic α, β, and
regulatory γ subunits (fi gure 1). The main role of the α
subunit is to bind and transport Na+ and K+. There are
four α subunits, all encoded by diff erent genes. The α3
subunit, encoded by ATP1A3, is the predominant α
subunit expressed in neurons,2–4 although many neurons
also express α1. The α3 subunit diff ers from α1 in that it
has a comparatively low affi nity for Na+ and K+,5,6 which
enables rapid normalisation of ion gradients after intense
neuronal fi ring. Changes in the activity of Na+/K+-
ATPases in neurons (predominantly expressing α3) have
physiological consequences. For example, inhibition of
Na+/K+-ATPases in the thalamus converts neuronal
bursting responses to single spike discharges.7 In the
hippocampus, reduction of the activity of Na+/K+-ATPase
causes interictal epileptiform
Furthermore, localisation of α3 in dendritic spines has
been shown to play a part in controlling the size and
speed of the small depolarisations caused by fl uctuations
of intracellular Na+ that occur during activation of ion-
gating neurotransmitter receptors. These fl uctuations,
known as transients, are summated in dendrites and are
the basis of synaptic integration.9–11
In addition to its primary function in ion transport,
subunits of Na+/K+-ATPases have been shown to interact
with proteins that assist in localisation of enzymes to the
cell membrane, modulate the PI3K, PLCγ, and MAPK
signal transduction cascades, and regulate the activity of
other transporters and receptors.12 The protein–protein
interactions specifi c to the α3 subunit are not fully
characterised, but, as an example, the α3 isoform has been
shown in rat neurons to bind specifi cally to PSD95, a
scaff olding protein that organises proteins at the synapse.10
ATP1A3 mutations and neurological disorders
In 1999, two large families were identifi ed with several
members presenting with RDP.13,14 Linkage analyses
identifi ed the 19q13 locus as the region most likely to
harbour mutations associated with disease;15 a fi nding
that was confi rmed in additional families with RDP.16,17
RDP in all these families was inherited as an autosomal
dominant trait with incomplete penetrance. Using a
positional cloning approach, investigators identifi ed six
diff erent heterozygous missense mutations in ATP1A3
that co-segregated with the disease phenotype.18 Since
the initial discovery, 11 mutations (nine missense
mutations, a 3-bp in-frame deletion, and a 3-bp in-frame
insertion; fi gure 2; table 118–34) have been reported in
20 patients with RDP, 12 of whom had a positive family
history of the disease.17–30,35 Three mutations, one found
only in sporadic RDP cases and two found in both
sporadic and familial RDP cases, are recurrent, probably
because they are located at hypermutable methylated
CpG-dinucleotides in ATP1A3 (table 1).36
Lancet Neurol 2014; 13: 503–14
Center for Human Genome
Variation (E L Heinzen PhD,
Prof D B Goldstein PhD),
Department of Medicine,
Section of Medical Genetics
(E L Heinzen), Department of
Molecular Genetics and
(Prof D B Goldstein), Division of
(Prof M A Mikati MD), and
Department of Neurobiology
(Prof M A Mikati), Duke
University, School of Medicine,
Durham, NC, USA; Epilepsy,
Sleep and Pediatric
HFME, University Hospitals of
Lyon, and Centre de Recherche
en Neurosciences de Lyon,
Centre National de la Recherche
Scientifi que, UMR 5292,
INSERM U1028, Lyon, France
(A Arzimanoglou MD);
Department of Neurology,
Wake Forest School of
Medicine, Winston Salem, NC,
USA (Prof A Brashear MD);
School of Biomedical Sciences,
University of Leeds, Leeds, UK
(S J Clapcote PhD); Istituto
di Genetica Medica, Università
Cattolica S Cuore, Rome, Italy
(Prof F Gurrieri PhD); AHC
Federation of Europe and AHC
Association of Iceland,
(S H Jóhannesson); Institute of
Child Health, University College
London, London, UK
(Prof B Neville MB BS); Institut
National de la Santé et de la
Recherche Médicale, U975,
Centre de Recherche de
l’Institut du Cerveau et de la
Moelle, Paris, France
(S Nicole PhD); Centre National
de la Recherche Scientifi que,
UMR7225, Paris, France
(S Nicole); Université Pierre et
Marie Curie Paris VI, UMRS975,
Paris, France (S Nicole);
Department of Genetics and
Genomic Sciences and
Department of Neurology,
Icahn School of Medicine at
Mount Sinai, New York, NY,
USA (L J Ozelius PhD); Danish
Research Institute for
Nordic-EMBL Partnership of
www.thelancet.com/neurology Vol 13 May 2014
Department of Molecular
Biology and Genetics, Aarhus
University, and Centre for
Membrane Pumps in Cells and
National Research Foundation,
(H Poulsen PhD); European
Network for Research on
(ENRAH), Brussels, Belgium
(T Schyns PhD); Neurosurgery,
Hospital, Boston, MA, USA
(K J Sweadner PhD); Department
of Human Genetics and
Department of Neurology,
Leiden University Medical
Centre, Leiden, Netherlands
(Prof A van den Maagdenberg
PhD); and Department of
University, Aarhus, Denmark
(Prof B Vilsen DMSc)
Dr Erin L Heinzen, Levine Science
Research Building, Duke
University, School of Medicine,
Durham, NC 27008, USA
For the Exome Variant server
In 2012, two independent studies, one by an inter-
national consortium31 and one by German researchers,32
identifi ed de-novo mutations in ATP1A3 as the cause of
AHC. In both studies next-generation sequencing was
used to screen the protein-coding portion of the genome
of patients with sporadic AHC to look for disease-causing
mutations that were absent in their unaff ected parents.
This approach identifi ed a de-novo ATP1A3 mutation in
each of the ten initial patients screened in the two studies
combined, which defi nitively establishes ATP1A3 as the
fi rst AHC gene.31,32 This fi nding was replicated in an
independent Japanese study,33 which found ATP1A3
mutations in eight of ten patients with AHC. The German
study identifi ed mutations in all 24 German patients with
AHC.32 Notably, the international study reported ATP1A3
mutations in 82 (78%) of 105 patients with AHC,31 which
suggests that some ATP1A3 mutations might have been
missed, that other AHC genes might exist, or that the
diagnosis might not always be accurate. 27 diff erent
ATP1A3 mutations have been reported in patients with
AHC (fi gure 2, table 1). Ten mutations have been identifi ed
in more than one individual, with one mutation
(Asp801Asn) explaining more than 40% of AHC cases
with an ATP1A3 mutation (table 1).
When we consider the location in the ATP1A3 protein
sequence of mutations that cause RDP or AHC, an
interesting diff erence emerges. Whereas RDP mutations
seem to be spread across the protein, AHC mutations are
located almost exclusively in particular regions of the
protein (fi gure 3). The signifi cance of the diff erent
mutation patterns in RDP and AHC is currently unknown,
but suggests that, unlike in RDP, only specifi c protein
disruptions result in AHC. Additionally, rarely the same
aminoacid is mutated in RDP and AHC, but even in these
cases the aminoacid substitution is disease-specifi c
(table 1). Only one RDP mutation (Asp923Asn) has also
been identifi ed in an unusual case of familial AHC. In
this multiplex AHC family, four individuals have the
Asp923Asn mutation, including one with a diagnosis of
AHC and three with some of the defi ning symptoms of
AHC (see below).29 This nearly perfect genotype-phenotype
correlation, with a nearly non-overlapping set of mutations
associated with AHC or RDP, strongly argues for a distinct
functional eff ect of the mutations causing AHC or RDP,
which is yet to be elucidated.
Consistent with mutations in ATP1A3 causing
neurodevelopmental diseases, only two polymorphic
missense mutations (both with low population frequencies
[minor allele frequency <0·1%]) have been reported for
ATP1A3 (fi gure 2, table 1) in the Exome Variant Server,
NHLBI GO Exome Sequencing Project (Seattle, WA,
USA). The database houses variants from protein-coding
genomes of approximately 6500 individuals who were not
identifi ed based on neurodevelopmental or neuro-
psychiatric disease phenotypes. Evaluating the relationship
of the total number of polymorphic functional variants as
a function of the total number of variants for each
Figure 1: Structure of the
(A) The Na+/K+-ATPase in the
showing K+ (red spheres), the
three protein subunits α (grey
lines), β (purple line) and
FXYDγ (green line), and
phosphorylation, which is
mimicked by MgF4²? (dark grey
spheres). Residues reported to
be mutated in AHC or RDP, or
both, are indicated by spheres
at the α carbon, yellow for one
case of AHC, orange for more
than one case of AHC, cyan for
RDP, and green for both AHC
and RDP. Key ion binding
residues are shown in stick
format. (B) A 90° degree
rotation in the membrane
plane of the representation
shown in (A), giving an
extracellular view of the
part of the Na⁺/K⁺-ATPase with
colour coded as in (A). Two of
the ion-binding residues
(shown in sticks) have been
found to be mutated both in
patients with RDP and in those
with AHC. Figures made from
pdb code 2ZXE.
of childhood. RDP=rapid-
onset dystonia parkinsonism.
Outside of cell
AHC and RDP
www.thelancet.com/neurology Vol 13 May 2014 505
sequenced gene in the database indicates that observing
two polymorphic functional variants in ATP1A3 given the
total number of variants reported is less that what is
expected. Thus, ATP1A3 is generally intolerant of
functional variation in comparison with genome-wide
expectations for a gene of its size and mutability, implying
that individuals with functional mutations in this gene
might be at high risk of developing serious diseases.37
Rapid-onset dystonia parkinsonism
Recognition of RDP began in 1993, more than 10 years
before ATP1A3 was identifi ed as the causal gene, when
Dobyns and colleagues14 reported a 15-year-old girl who
had abrupt onset of dystonia with prominent dysarthria
and dysphagia. The disorder was named RDP because of
the abrupt onset of dystonic spasms associated with
postural instability and bradykinesia that resemble signs
of parkinsonism.13,38–40 RDP is also sometimes referred to
as dystonia 12 (DYT12; OMIM: 128235). ATP1A3 is the
only known RDP gene; however, other RDP genes might
exist, as evidenced by one study reporting that, in a select
group of 14 patients referred for possible RDP, only three
had a mutation in ATP1A3.19
The clinical presentation of RDP includes three features:
its appearance, which often occurs after triggering events
such as running, alcohol binges, minor head injuries,
overheating, emotional stress, infections, or childbirth;21 a
rapid onset of typically permanent symptoms that develop
over hours to days (occasionally even over several weeks);
and involuntary movements that are characterised by
generalised dystonia with superimposed parkinsonian
features (primarily bradykinesia and postural instability
without tremor). Many patients present with a rostrocaudal
gradient of dystonia and parkinsonism, in the sense that
bulbar symptoms are more severe than arm symptoms,
and arm symptoms are more severe than leg symptoms.
The bulbar and arm symptoms rarely improve after the
primary disease onset. A few patients have reported later
episodes of abrupt worsening of symptoms that occurred
from 1 year to as late as 9 years after the initial onset. One
patient, who had transient symptoms after athletic activity,
recovered, resumed strenuous athletic activity, and then
had permanent onset of fi xed symptoms. Not all patients
report a recognised trigger, and a few report antecedent
periods of cramping. Patients typically lack other disease
features such as diurnal fl uctuation or episodes of
symptoms that are typical of patients with AHC.
A recently published cohort of 26 patients with RDP
indicated that 76% had onset of motor symptoms by the
age of 25 years.41 In addition to dystonia, patients can also
have non-motor manifestations. Recent work suggests
that patients with RDP have an increased prevalence of
mood disorders (50%) and psychosis (19%) compared
with relatives without an ATP1A3 mutation.41 These
fi ndings were observed across families with diff erent
ATP1A3 mutations, and are consistent with reports of
depression in individuals with ATP1A3 mutations
causing motor problems.16
The originally published diagnostic criteria for RDP
require a family history, an onset of the disease in teenage
years, and prominent bulbar fi ndings.19 However, as the
number of reported cases of RDP increases, these criteria
seem too restrictive for several reasons: more than half of
patients with RDP lack a positive family history because
the disease is caused by de-novo mutations in small,
single-patient families; disease onset has been reported
in children and adults;30 and recognition is growing that
there are patients in whom RDP can present with atypical
Figure 2: Location of AHC-causing or RDP-causing mutations in ATP1A3 gene, mRNA, and protein
Red dots show AHC-causing mutations and blue dots show RDP-causing mutations. The one mutation shared between disease phenotypes is located at Asp923Asn (blue dot with a red dot inside).
Two rare polymorphisms identifi ed in the general population but not associated with a disease at this time are indicated by green dots. AHC=alternating hemiplegia of childhood. RDP=rapid-onset
dystonia parkinsonism. aa=aminoacids. nt=nucleotides. bp=basepairs.
1 2 345 6789 1011 12 13141516 17 1819 20212223
www.thelancet.com/neurology Vol 13 May 2014
features, including second onsets and unusual, mild-to-
moderate improvement after the primary onset of disease
symptoms (table 2). Collectively, these fi ndings suggest
that the phenotype is broader than originally described
Drug therapy in RDP is limited; patients are
unresponsive to standard drugs for parkinsonism,
including levodopa.14,16,19,20,41 Current treatment is limited
to benzodiazepines, which have been reported to provide
symptomatic relief in some patients.
Alternating hemiplegia of childhood
Although the fi rst report of AHC was by Verret and
Steele in 1971,60 it was not until 1980 that Krageloh and
Aicardi61 fi rst defi ned the syndrome. In 1993, the specifi c
clinical criteria to diagnose AHC were proposed.42
These, named the Aicardi criteria, include seven disease
features: (1) paroxysmal episodes of hemiplegia; (2)
episodes of bilateral hemiplegia or quadriplegia; (3)
other paroxysmal manifestations, such as abnormal eye
movements, dystonia, nystagmus, intermittent strabi-
smus, tonic spells, or autonomic disturbance, which
can occur during hemiplegia or as isolated events; (4)
evidence of permanent neurological dysfunction, which
can manifest as intellectual defi ciencies, seizures,
ataxia, choreoathetosis, develop mental
persistent motor defi cits such as spastic diplegia or
quadriplegia or hypotonia; (5) inducing sleep during a
paroxysmal attack might relieve symptoms for a period
of time after awakening; (6) fi rst signs of dysfunction
occurring before age 18 months; and (7) not being
attributed to other disorders. The median age of disease
onset of the fi rst paroxysmal event is 3·5 months with a
range from the fi rst day of life to 4 years. Developmental
delay might not appear until after 12 months. Additional
clinical details are summarised in table 2.
Although nearly all cases of AHC are sporadic, some
families have an autosomal dominant inheritance of the
disorder.49–51,62 In two families a causal ATP1A3 mutation
was identifi ed,29,31,62 and in one family a disease-causing
ATP1A2 (encoding the α2 subunit of the Na+/K+-ATPase)
mutation was found.50,51 Consistent with these mutations
being highly penetrant, all ATP1A2 or ATP1A3 mutation
Disease Disease inheritance Number of unrelated
patients with the
2542+1G→A; splice site
RDP, AHCSporadic and familial
RDP, familial AHC
RDP sporadic, 2; RDP
familial, 2; AHC familial, 1
ATP1A3 mutation coordinates are defi ned based on UniProt ID P13637 and Consensus CDS ID CCDS12594·1. Mutation
658G→A, Asp220Asn, previously reported as causal by Heinzen and coworkers31 was later shown to be a rare, inherited
mutation, and the disease-causing mutation in this patient was a previously overlooked de-novo Asp801Asn ATP1A3
mutation (unpublished data); thus Asp220Asn is not shown in this table and one additional patient has been counted
as having an Asp801Asn mutation. AHC=alternating hemiplegia of childhood. RDP=rapid-onset dystonia parkinsonism.
*Compiled from references.18–34
Table 1: Disease-causing ATP1A3 mutations
Figure 3: Density plot showing the distribution of AHC and RDP mutations in
Mutations identifi ed to date in 20 patients with RDP (blue) and 118 patients
with AHC (red). In general, RDP mutations appear more evenly distributed,
whereas AHC mutations are heavily concentrated in particular sites in the
protein. AHC=alternating hemiplegia of childhood. RDP=rapid-onset dystonia
parkinsonism. Aminoacid positions are based on the ATP1A3 protein defi ned by
UniProt ID P13637.
Aminoacid position in ATP1A3
0 200 400600800 1000
www.thelancet.com/neurology Vol 13 May 2014 507
carriers in these families have symptoms of AHC, albeit
to varying severity even within families.29,31,50,51,62 In
addition to these unusual familial cases, other patients
have atypical presentations that resemble AHC, although
currently a role for ATP1A3 mutation is untested. These
include benign nocturnal alternating hemiplegia, which
occurs only in boys,52,53 mild cases with normal cognitive
development,54 cases in which dystonia is the
predominant feature,54 patients who do not have episodes
of quadriplegia, patients who have the fi rst signs of the
disorder after the age of 18 months,54,55 patients presenting
with neonatal seizures, and patients with status
epilepticus with associated long-term atrophy on MRI
and residual motor and eye movement abnormalities.56
Over the years, patients with AHC have undergone
many treatments to alleviate the frequency and severity
of hemiplegia, although success has been very limited.43,44
Flunarizine, a calcium channel blocker, performed best
because it seemed to reduce the severity and duration of
attacks, at least in some patients.45–48 Benzodiazepines,
which increase the activity of GABA, the major inhibitory
neurotransmitter in the CNS, might also have some
effi cacy either directly or by inducing sleep, which often
Although the number of patients with AHC or RDP with
an identifi ed ATP1A3 mutation is rapidly growing,
defi nitive phenotypic patterns have not been found for
patients with and without mutations, and in patients
with recurring ATP1A3 mutations.31,32,34 However, one
small study63 evaluating the phenotypes of 35 patients
with AHC with ATP1A3 mutations reports that patients
with the Glu815Lys mutation tend to have earlier onset of
symptoms, more severe motor and cognitive disabilities,
and more often report status epilepticus and respiratory
paralysis compared with patients with AHC with other
ATP1A3 mutations. This preliminary fi nding suggests
that genotype–phenotype correlations exist and that
additional studies will be needed to further evaluate
these patterns in larger sample sizes. The ATP1A3
Working Group is currently analysing genotype–
phenotype correlations in 150 patients.
Biological eff ects of ATP1A3 mutations
ATP1A3 protein expression and localisation
Ten RDP and fi ve AHC ATP1A3 mutations have been
investigated at the level of protein expression and cellular
localisation of the protein by using heterologous expression
systems.18,31 These studies revealed that for all except two of
the tested RDP mutations the ATP1A3 protein expression
was reduced, whereas none of the AHC mutations reduced
expression of ATP1A3. RDP mutations were not shown to
aff ect the maturation and localisation of the protein in
transfected cells;18 the eff ect of AHC mutations on ATP1A3
localisation has not yet been studied.
ATP1A3 mutations in platelets and fi broblasts
Platelets and fi broblasts from nine patients with AHC
have been screened for diff erences in protein expression
compared with age-matched and sex-matched controls.64
A consistent increase in the level of activated lysosomal
protein cathepsin B was observed in specimens from
patients with AHC, which was shown to increase
apoptosis. Although the mechanism remains unclear,
Age of onset 0–18 months (median 3·5 months), although some isolated cases are reported as late as 4 yearsCan occur in children and adults (9 months to
59 years); 76% of patients report symptoms by
age 25 years
Sporadic or familial
Typically abrupt onset of permanent symptoms
of generalised dystonia and parkinsonian features
including bradykinesia and postural instability
Variable progression of symptoms ranging from
hours to weeks
Diurnal fl uctuation or episodes typical of patients
with AHC are not seen in RDP
Abrupt worsening of symptoms can rarely occur
years after initial onset
Onset of symptoms typically occurs after running,
alcohol binges, minor head injuries, overheating,
emotional stress, infections, or childbirth
Can be associated with mood disorders and psychosis
Clinical presentation Paroxysmal symptoms: episodes of hemiplegia, bilateral hemiplegia, or quadriplegia with improvement in sleep and
with other paroxysmal abnormalities such as dystonia, tonic spells, epileptic seizures (seizures occur in about 50% of
patients and can include focal or generalised tonic, tonic-clonic, or myoclonic seizure types), autonomic changes, or
abnormal eye movements (including unilateral [ipsilateral to the hemiplegia] intermittent eye deviation, disconjugate
gaze, and pendular nystagmus) often with the following characteristics: common triggers include stress, excitement,
extreme heat or cold, water exposure, physical exertion, lighting changes, and foods; variable frequency ranging from
multiple episodes per day to one every few months; associated with anarthria, dysphagia, and autonomic disturbances
including bradycardia, stridor, bronchospasm, apnoea, dyspnoea, nausea, unilateral or bilateral fl ushing, hypothermia
or hyperthermia; episodes typically remit with sleep, but can return within 10–20 min after waking; episodes of
dystonia in concert or separate from episodes of hemiplegia, bilateral hemiplegia, quadriplegia, or abnormal eye
movements that are more commonly observed early in the clinical course, typically consist of head turning toward the
aff ected side with eye deviations toward the same side; very brief intense unilateral or bilateral tonic or dystonic attacks
with vibratory tremor and pain
Interictal symptoms: evidence of developmental delay (including intellectual defi ciencies, neuropsychological
defi cits), mental retardation, and interictal neurological abnormalities such as tone abnormalities, choreoathetosis
(spontaneous or movement-induced), or ataxia (sometimes associated with cerebellar atrophy or cerebellar
hypometabolism), and persistent motor defi cits (such as spastic diplegia or quadriplegia, hypotonia)
Drug therapy Flunarizine, benzodiazepines
Typically sporadic; rare cases of familial AHC
Unresponsive to levodopa
AHC=alternating hemiplegia of childhood. RDP=rapid-onset dystonia parkinsonism.
Table 2: Proposed diagnostic criteria for AHC and RDP13,14,16,19,20,29–31,38–59
www.thelancet.com/neurology Vol 13 May 2014
this work suggests that similar protein changes might
also occur in the brain and could contribute to AHC
In the established Post-Albers model for the Na+/K+-
ATPase transport mechanism (fi gure 4),65 in which three
cytoplasmic Na+ ions are exchanged for two extracellular
K+ ions for each ATP being hydrolysed, the E1 conformation
preferentially binds Na+, whereas E2 preferentially binds
K+. On the basis of this model, there are several
experimental approaches to measure the activity of the
Na+/K+-ATPase, and specifi cally the eff ects of disease-
causing mutations in ATP1A3 on the catalytic cycle.
First, using a luminescent kinase assay that measures
ADP formed from each catalytic cycle of the Na+/K+-
ATPase (fi gure 4) from cells heterologously expressing
the wild-type, RDP-causing, or AHC-causing mutant
version of the ATP1A3 cDNA, investigators have shown
that each mutation reduced the activity of the Na+/K+-
ATPase.31 These fi ndings, coupled with the afore-
mentioned protein expression analyses, suggest that
mutations that aff ect Na+/K+-ATPase function, but not
the amount of Na+/K+-ATPase per se, seem to result in
the more severe AHC phenotype.
Second, since active Na+/K+-ATPase pumps a net charge
of +1 out of the cell for each round of the catalytic cycle
(fi gure 4), electrophysiology-based approaches can be
used to measure, in a living cell or in a patch of excised
membrane, the voltage and ion sensitivity of the pump
during steady-state activity. Two disease-causing ATP1A3
mutations have been characterised in detail: Asp801Asn
and Asp923Asn. The aspartate located at aminoacid
position 801 is highly conserved in Na+/K+-ATPases in all
investigated species, and is a site that binds either the
Na+ or K+ ions during the catalytic cycle. Expression of the
AHC-causing Asp801Asn mutant in Xenopus oocytes
showed that the mutation did not generate any
measurable pump current, consistent with it being
unable to bind K+ ions.67 Analysis of the Asp923Asn
mutation, which was identifi ed in both patients with
RDP23,25,29,30 and familial
protonation at this site was crucial for the movement of
Na+ and K+ across the cell membrane.68
Third, during the ion transport process, the Na+/K+-
ATPase becomes phosphorylated by transfer of the
γ-phosphate of ATP to a conserved aspartic acid residue in
the P-type ATPase signature sequence (fi gure 4). By
incubating ATP radiolabelled at the γ-phosphate with cell
membrane fragments containing the Na+/K+-ATPase to
quantify this covalent and acid-stable phosphoryl bond,
investigators can measure the activity of the Na+/K+-ATPase
with and without disease-causing mutations.66,69–71 Because
the binding of three Na+ ions at the cytoplasmic surface of
the ATP1A3 protein is needed to activate the enzyme for
phosphorylation from ATP, the affi nity for Na+ at these
sites can be established by measurement of the Na+
dependence of phosphorylation, the affi nity being defi ned
by the Na+ concentration giving half maximum phosph-
orylation. So far, RDP-causing mutations Glu277Lys,
Thr613Met, Phe780Leu, +Tyr (an extension of the
C-terminus with an extra Tyr residue), and the RDP-
causing and AHC-causing mutation Asp923Asn have
been characterised with these methods.24,69–71 Each of the
mutations shows striking reductions of Na+ affi nity for
activation of phosphorylation. Whereas Na+ binding from
the cytoplasmic side activates phosphorylation from ATP,
binding of K+ from the external side triggers de-
phosphorylation, thereby stimulating ATP hydrolysis.
Therefore, the affi nity for external K+ can be established by
study of the K+ dependence of dephosphorylation or
ATPase activity at a fi xed Na+ concentration. Notably, none
of the mutants show a reduced affi nity for K+, indicating a
selective disturbance of Na+ binding that is associated with
mutations causing RDP and one mutation causing AHC
The described studies show that a selective reduction
of the affi nity of the Na+/K+-ATPase for cytoplasmic Na+
without disturbance of K+ binding is a central feature
in RDP. Consequently, increased intracellular Na+
concentration resulting from the reduced Na+ affi nity
could be a key pathogenic factor in RDP.66 A rise in
intracellular Na+ might result in a secondary increase
in intracellular Ca²+ via the Na+/Ca²+ exchange system,
which can activate signalling cascades triggered by
changes in Ca²+ concentration. Additionally, a
disturbance of the Na+ gradient could aff ect the uptake
of neurotransmitters such as dopamine and glutamate.
Because several AHC-causing mutations target the
same residue (Ile274, Asp801, Asp923) as the one
mutated in RDP, an essential question to address is
whether Na+ affi nity is also typically disturbed in AHC.
The affi nity for K+ might also be disturbed in AHC,
which could also explain why AHC is at the severe end
of the phenotypic spectrum, whereas RDP is at the
AHC,29 suggested that
Figure 4: Post-Albers model65 for the Na+/K+-ATPase reaction cycle
Reproduced from Toustrup-Jensen and coworkers66 with permission from the
American Society for Biochemistry and Molecular Biology. E1 and E2 are major
conformational states with preference for binding of Na⁺ and K⁺, respectively.
Cytoplasmic and extracellular ions are indicated by subscripts c and e,
respectively. Brackets indicate occlusion of the ions in a cavity in the protein.
P indicates the bound phosphate.
www.thelancet.com/neurology Vol 13 May 2014 509
mild end. Structural modelling of three AHC-causing
mutations (Ile274Asn, Asp801Asn, and Asp923Tyr) and
three RDP-causing mutations (Ile274Thr, Asp801Tyr,
and Asp923Asn) that aff ect identical positions in the
Na+/K+-ATPase α3 subunit72 predicted that AHC
mutations would bring about structural changes that
severely aff ect effi cient K+ movement along the narrow
K+ access pathway. Instead, RDP-causing mutations
seem to have milder structural consequences that are
likely to result in a milder impairment of K+
Notably, a mutation in the sarcoendoplasmic reticulum
Ca²+ ATPase, a member of the same type II P-type ATPase
family as the Na+/K+-ATPase, in Drosophila melanogaster
causes temperature-sensitive ionic leakage of the
transporter.74 Investigators postulated that a temperature-
sensitive gain-of-function mechanism might also
underlie the phenotypic consequences of disease-causing
mutations in other type II P-type ATPases, including
ATP1A3 in AHC and RDP. Although the eff ect of
temperature on the functional eff ects of AHC-causing
and RDP-causing mutations in Na+/K+-ATPases is
unknown, if correct, this mechanism could explain why
environmental triggers such as stress, physical exertion,
and temperature changes can lead to symptom onset in
patients with RDP or AHC (table 2).
Several animal models have been used to study the in-
vivo consequences of ATP1A3 modulation (fi gure 5). The
aminoacid sequence identity between the human and
mouse Na+/K+-ATPase α3 subunits is about 99%.
Heterozygous Myshkin (Atp1a3Myk/+; Myk/+) mutant mice
have an aminoacid change (Ile810Asn) that aff ects the
identical position to Ile810Ser in the human Na+/K+-
ATPase α3 subunit that was identifi ed in a patient with
AHC.31,75 Molecular modelling of Ile810Asn and Ile810Ser
showed that both changes bring about similarly severe
structural eff ects on the Na+/K+-ATPase α3 subunit,
including the capacity for effi cient K+ movement along
the K+ access pathway.73 Ile810Asn was generated through
N-nitroso-N-ethylurea (ENU) mutagenesis and results in
a normally expressed, but inactive, α3 protein and a
subsequent 36–42% reduction in total Na+/K+-ATPase
activity (refl ecting the combined activity of α1, α2, and α3)
in the brain.75,81
Heterozygous Myk/+ mice have an unsteady, tremulous
gait with occasional splaying of the hindlimbs, but without
an overt hemiplegia.73 Phenotypic analysis revealed a
range of other abnormalities in Myk/+ mice, including
reduction in body size, motor defi cits in the balance beam
and rotarod tests, cognitive defi cits in the fear-conditioning
Figure 5: Na+/K+-ATPase α3 genetic animal models
(A) Atp1α3 mutant mice. The locations of three mutations in the mouse Atp1a3 genomic locus are depicted. Myshkin mice carry a T→A transversion in exon 18 that results in the substitution of
asparagine for isoleucine at position 810 (Ile810Asn).73,75 Atp1a3tm1Ling mice carry a point mutation in intron 4 adjacent to the exon-intron splice site that results in aberrant splicing of the gene, adding
126 bp to the RNA transcript.75–77 Atp1a3tm2Kwk mice carry a STOP-polyA cassette that replaces exons 2–6 in Atp1a3.78 (B) Atpα mutant Drosophila. AtpαCJ10 fruit fl ies carry a G→A transition that results in
the substitution of glycine for serine at position 744 (Gly744Ser).79 Gly744Ser in the Drosophila α subunit is equivalent to mutation Gly755Ser in the human α3 subunit found in a patient with AHC.31
(C) atp1a3a/b knockdown zebrafi sh. Knockdown of atp1a3a or atp1a3b RNA transcript by around 65% in 60 hpf embryos had similar phenotypic eff ects.80 Red stars show point mutations. NKA=Na+K+-
ATPase. ↓=lower than wild-type. ↑=greater than wild-type. No change=no change from wild-type. ND=not determined. AHC= alternating hemiplegia of childhood. hpf=hours post-fertilisation.
in brain NKA
60 hpf embryo phenotypeKnockdown
Reduction in α3
in intron 4
1 23 45 6 7 8910 111213 141516 171819202122
1 23 45 6 7 8910 11 1213 1415161718 19202122
1 23 45 67 8910 11 1213 1415161718 192021 22
B Atpα mutant Drosophila
C atp1a3a/b knockdown zebrafish
A Atp1a3 mutant mice
www.thelancet.com/neurology Vol 13 May 2014
and conditioned taste aversion tests, neuronal hyper-
excitability with spontaneous convulsions, and mania-
related behaviours such as increased risk-taking and
responsiveness to treatment with lithium and valproic
acid.73,75,81 When subjected to vestibular stress, Myk/+ mice
have transient tonic attacks and staggering movements
that, in a third of mice, develop into tonic-clonic seizures
that are accompanied by epileptiform discharges.75 ¹⁴C-2-
deoxyglucose imaging of Myk/+ mice identifi ed
compromised thalamocortical functioning, including a
defi cit in frontal cortex functioning and reduced
thalamocortical functional connectivity.73 When bred to
homozygosity, Myk/Myk pups die shortly after birth.75
Strategies aimed at increasing Na+/K+-ATPase activity
have shown some therapeutic eff ects in Myk/+ mice.
Transgenic delivery of an additional copy of the wild-type
Atp1a3 gene to the X chromosome increased Na+/K+-
ATPase α3 subunit protein expression and whole brain
Na+/K+-ATPase activity, and reduced the epileptic seizure
susceptibility as well as the risk-taking behaviour of
Myk/+ mice.75,82 Chronic treatment with rostafuroxin, a
compound that antagonises the inhibitory action of
ouabain on Na+/K+-ATPase,83 was also found to reduce
the risk-taking behaviour of Myk/+ mice.81 Eff ects of this
intervention on the motor and cognitive defi cits of Myk/+
mice have not yet been established.
Heterozygous Atp1a3tm1Ling/+ mice, which have a point
mutation in intron 4 of the Atp1a3 gene, show a reduction
of hippocampal α3 protein expression of around 60% and
a reduction of total brain Na+/K+-ATPase activity (of α1, α2,
and α3 combined) of around 16%.75,76,82 Non-stressed
(naive) Atp1a3tm1Ling/+ mice do not have visible neurological
defects or restricted growth, but instead show increased
locomotor activity in an open fi eld test and defi cient
spatial learning in the Morris water maze test.76 After
exposure to restraint stress for 5 days, female Atp1a3tm1Ling/+
mice show mild motor defi cits in the balance beam and
rotarod tests.77 Atp1a3tm1Ling/+ mice exposed to chronic
variable stress, consisting of one or two unpredictable
mild stressors per day for 6 weeks, have defi cits in total
brain Na+/K+-ATPase activity, sociability, and object
recognition memory, as well as increased anxiety and
depression-like behaviours, compared with non-stressed
Atp1a3tm1Ling/+ mice.82 In Atp1a3 wild-type mice, chronic
variable stress also led to depression-like behaviour and
reduced sociability, but had no eff ect on Na+/K+-ATPase
activity, anxiety, or object recognition memory compared
with non-stressed wild-type controls.82 Homozygous
Atp1a3tm1Ling pups die shortly after birth.76
Heterozygous Atp1a3tm2Kwk/+ mice have a targeted deletion
of Atp1a3 exons 2–6.78 Atp1a3tm2Kwk/+ mice do not show gross
morphological defects or apparent histological brain
anomalies. Adult male Atp1a3tm2Kwk/+ mice show increased
locomotor activity, both in the home cage and in the open
fi eld test. By contrast with Atp1a3Myk/+ mice, Atp1a3tm2Kwk/+
mice had improved performance in the balance beam and
rotarod tests compared with wild-type mice. Atp1a3tm2Kwk/+
mice do not develop dystonia spontaneously, nor after
various stressors, such as tail suspension, forced
swimming, or restraint. Dystonia can be induced
pharmacologically84 in these mice by injection of the
neuroexcitatory aminoacid kainate directly into the
cerebellum. The response to dystonia induction by kainate
injection was increased in Atp1a3tm2Kwk/+ mice, with a longer
duration of sustained dystonia compared with wild-type
mice. Electrophysiological studies showed that inhibitory
neurotransmission at molecular-layer
Purkinje cell synapses was increased in the cerebellar
cortex of Atp1a3tm2Kwk/+ mice. Homozygous Atp1a3tm2Kwk mice
show a complete lack of breathing movements and die
shortly after birth.78
Pharmacological blockade of Na+/K+-ATPase α3
Perfusion of the Na+/K+-ATPase inhibitor ouabain into the
cerebellum and basal ganglia was found to induce mild
dyskinesia in wild-type C57BL/6 mice.85 When mice were
subsequently exposed for 2 h to stress provided in the
form of random electric foot shocks in a warm
environment (38°C), 70% of the mice developed persistent
dystonia and rigidity.85 These mice show hallmark
symptoms of RDP, including dystonia and parkinsonism
induced by stress. However, this approach is limited by
the similar affi nities of the α2 and α3 isoforms for
ouabain,86 thus precluding α3 specifi city in this animal
Zebrafi sh (Danio rerio) have two ATP1A3 orthologues,
atp1a3a and atp1a3b.87 The paralogous α3a and α3b subunits
have aminoacid identities of 95% with each other and
91–92% with the human Na+/K+-ATPase α3 subunit
protein sequence. Consistent with mammalian Na+/K+-
ATPase α3 subunit protein expression, the transcripts of
atp1a3a and atp1a3b are primarily expressed in the brain,
albeit with distinct expression profi les. In 60 h post-
fertilisation zebrafi sh embryos, the atp1a3a transcript is
widely distributed throughout the brain, whereas
distribution of atp1a3b mRNA is localised to particular
brain structures. Despite having distinct expression
profi les, targeted knockdown of atp1a3a or atp1a3b by
morpholino antisense oligonucleotides results in severe
brain ventricle dilation in 60 h post-fertilisation embryos,
suggesting that both α3 paralogues are needed for brain
ventricle maintenance. The extent of brain ventricle
dilation was reduced by co-injection of the mRNA of the
knocked-down gene, but atp1a3b mRNA did not cross-
rescue the phenotype of atp1a3a-knockdown embryos.
Similarly, atp1a3a mRNA did not cross-rescue the
phenotype of atp1a3b-knockdown embryos. Both
morphants display abnormal spontaneous motility and
www.thelancet.com/neurology Vol 13 May 2014 511
an abnormal response to tactile stimulation with a
needle,80 suggesting that both α3 paralogues are needed
for embryonic motility.
The gene Atpα (FlyBase ID: FBgn0002921) in Drosophila
melanogaster fruit fl ies encodes the α subunit of the
Na+/K+-ATPase, which is orthologous to all vertebrate α
subunits, and has aminoacid sequence identities of
76–77% with the α1, α2, and α3 subunits of human beings.88
Although the Drosophila Atpα gene is not a specifi c
orthologue of ATP1A3, eight missense mutations,
generated through ethylmethanesulfonate mutagenesis at
highly conserved aminoacid residues, lead to AHC-
relevant phenotypic abnormalities in adult heterozygous
fl ies.79,89 Flies from any of six lines with mutations
(Ser201Leu, Pro262Leu, Ser348Thr, Gly528Ser, Ala588Thr,
Gly744Ser) in Atpα that were repeatedly knocked to the
bottom of a vial using a standard laboratory vortexer
showed transient mechanical stress-induced paralysis.79
Two other mutant lines (Asp981Asn, Glu928Lys) did not
show this phenotype when maintained at an ambient
temperature of 20–22°C, but showed temperature-
sensitive mechanical stress-induced paralysis when
maintained at 28°C.89 When exposed to a temperature of
37–38°C, three mutant lines (Gly744Ser, Asp981Asn,
Glu928Lys) showed temperature-sensitive paralysis that
was reversed when the ambient temperature was lowered
to 20–22°C.79,89 When the bodyweight of male fl ies of six of
the same mutant lines was measured, only Ser201Leu
mutant fl ies showed a reduction compared with wild-type
fl ies.79 Western blotting of homogenised fl y heads showed
that the expression of Na+/K+-ATPase α subunit was
reduced in two of the Atpα mutants (Ser348Thr,
Ala588Thr), but unchanged in the other four mutants.79
Mutations Gly744Ser and Asp981Asn in the Drosophila α
subunit, which led to temperature-sensitive paralysis,
aff ect equivalent aminoacid residues in the human
Na+/K+-ATPase α3 subunit, namely mutations Gly755Cys,
Gly755Ser, and Asp992Tyr, which were identifi ed in
patients with AHC (table 3).31,32 All eight missense
mutations are homozygous lethal.79
Since the original descriptions of RDP and AHC,
substantial work has been done to characterise their
clinical presentation and pathophysiology. Through the
identifi cation of disease-causing mutations in ATP1A3,
these two seemingly unrelated diseases are now linked,
allowing new opportunities to obtain insight into their
We now understand that protein-modifying genetic
variations in ATP1A3 rarely occur in the general population
and, when they do, the risk of severe neurological disease
is very high. This understanding has led to a new research
area, investigating which other diseases might be
associated with mutations in ATP1A3. One could postulate
that ATP1A3 mutations might also be found in patients
with seizures, psychiatric conditions, or other less severe
types of dystonia or ataxia. As next-generation sequencing
becomes widely used in day-to-day clinical practice, the
role of ATP1A3 mutations in a wider range of phenotypes
might become apparent. However, as this genotype–
phenotype spectrum is being defi ned, we can already
begin cataloguing ATP1A3 disease-specifi c mutations and
polymorphic protein-disrupting variants in the general
population, to establish molecular and physiological
changes associated with these DNA variations in in-vitro
and in-vivo test paradigms like those described in this
Review. Data from these studies can be used to develop
better research models with tiered complexity. We envisage
that these studies will include establishing disease models
at the transporter level in individual cells, multicellular
network models using induced pluripotent cells
diff erentiated into neurons, ex-vivo studies in brain slices
to study tissue-level eff ects, and, fi nally, evaluations at the
organism level to assess in-vivo consequences in animal
models. With these model systems we can begin to relate
molecular changes to the phenotypic presentations
associated with these disorders, including the episodic
nature of AHC, how particular stimuli lead to the onset of
symptoms, the age-dependent onset of RDP, and the
variable eff ects on organ systems and brain structures that
probably underlie the diverse phenotypic presentations.
Adapted from Ashmore and colleagues.79 AHC=equivalent aminoacid residue that is substituted in alternating
hemiplegia of childhood. Mild=when maintained at 28°C. No change=no change from wild-type. ↓=lower than
wild-type. ND=not determined.
Table 3: Drosophila Na+/K+-ATPase α subunit mutant phenotypes
Search strategy and selection criteria
References cited in this Review were identifi ed through
PubMed searches using the search terms “ATP1A3”, “alternating
hemiplegia of childhood”, “rapid-onset dystonia parkinsonism”,
and “Na,K-ATPase “, from December, 1993, until January, 2014.
Articles were identifi ed through searches of the reference lists of
the articles found with the above cited search terms and of the
authors’ own fi les. All references used in this Review were
published in English, and were selected according to originality
and relevance to the content of this Review.
www.thelancet.com/neurology Vol 13 May 2014
Importantly, once key biomarkers of disease patho-
physiology are identifi ed, we will be able to screen for
compounds to rectify the pathophysiological changes
associated with ATP1A3 mutations.
In summary, genetics has illuminated key aspects of
disease pathophysiology for both AHC and RDP.
Although extensive work is needed to disentangle the
complex biology underlying these disorders, we are
poised with evolving research approaches to rapidly
translate these genetic discoveries to detailed disease
pathophysiology, to improve understanding of develop-
mentally mediated and environmentally triggered
disease presentation, and ultimately to develop treat-
ments for these debilitating diseases.
ELH, AA, AB, SJC, FG, DBG, SHJ, MAM, BN, SN, LJO, HP, TS, KJS,
AvdM, and BV compiled relevant information from the literature and
wrote the manuscript. The entire ATP1A3 Working Group critiqued and
edited the Review and provided scientifi c and clinical guidance regarding
ATP1A3 Working Group
Alexis Arzimanoglou (University Hospitals of Lyon, Centre National de la
Recherche Scientifi que, and Institut National de la Santé et de la
Recherche Médicale, Lyon, France), Frances M Ashcroft (University of
Oxford, Oxford, UK), Allison Brashear (Wake Forest School of Medicine,
Winston Salem, NC, USA), Knut Brockmann (University Medical Center,
Göttingen, Germany), Jaume Campistol (Hospital Sant Joan de Déu—
Barcelona University, Barcelona, Spain), Alessandro Capuano (Bambino
Gesù Paediatric Hospital, IRCSS, Rome, Italy), Inês Carrilho (Hospitalar
Center of Oporto, Oporto, Portugal), Paul Casaer (University Hospital
Gasthuisberg, Leuven, Belgium), Steven J Clapcote (University of Leeds,
Leeds, UK), Elisa De Grandis (University of Genoa, Genoa, Italy),
Boukje de Vries (Leiden University Medical Centre, Leiden, Netherlands),
Michela Di Michele (KU Leuven, Leuven, Belgium), Caroline Dion
(Canadian Association for Alternating Hemiplegia, Saint-Mathieu de
Beloeil, QC, Canada), Diane Doummar (Hôpital Trousseau, Paris, France),
Anja P Einholm (Aarhus University, Aarhus, Denmark), Carmen Fons
(Hospital Sant Joan de Déu—Barcelona University. Barcelona, Spain),
Filippo Franchini (AHC Federation of Europe, Reykjavík, Iceland, and
A.I.S.EA Onlus, Verderio Superiore, Italy), Thomas Friedrich (Technical
University of Berlin, Berlin, Germany), Kathleen Freson (KU Leuven,
Leuven, Belgium), David C Gadsby (Rockefeller University, New York, NY,
USA), Melania Giannotta (IRCCS Istituto delle Scienze Neurologiche di
Bologna, Bologna, Italy), David B Goldstein (Duke University, School of
Medicine, Durham, NC, USA), Christophe Goubau (KU Leuven and
University Hospital Leuven, Leuven, Belgium), Titiana Granata (National
Neurological Institute C Besta, Milan, Italy), Fiorella Gurrieri (Università
Cattolica S Cuore, Rome, Italy), Erin L Heinzen (Duke University, School
of Medicine, Durham, NC, USA), Shinichi Hirose (Fukuoka University,
Fukuoka, Japan), Yuki Hitomi (Duke University, School of Medicine,
Durham, NC, USA), Rikke Holm (Aarhus University, Aarhus, Denmark),
Keiko Ikeda (Hyogo College of Medicine, Hyogo, Japan), Atsushi Ishii
(Fukuoka University, Fukuoka, Japan), Sigurður H Jóhannesson (AHC
Federation of Europe and AHC Association of Iceland, Reykjavik, Iceland),
Kamran Khodakhah (Albert Einstein College of Medicine, New York, NY,
USA), Mary D King (Children’s University Hospital, Dublin, Ireland),
Greer S Kirshenbaum (Mount Sinai Hospital and University of Toronto,
Toronto, ON, Canada), Ana Kockhans (AHC-Deutschland eV,
Mönchengladbach, Germany), Jan B Koenderink (Radboud University
Nijmegen Medical Centre, Nijmegen, Netherlands), Gaetan Lesca (Centre
de Recherche en Neurosciences de Lyon, University Hospitals of Lyon, and
Claude Bernard Lyon I University, Lyon, France), Karin Lykke-Hartmann
(Centre for Membrane Pumps in Cells and Disease—PUMPKIN, Danish
National Research Foundation, and Aarhus University, Aarhus, Denmark),
Ulrike Maschke (Catholic Hospital St Johann Nepomuk, Erfurt, Germany),
Mario R Merida (Stevens-Henager College, Ogden, UT, USA),
Mohamad A Mikati (Duke University School of Medicine, Durham, NC,
USA), Ralf Müller (AHC-Deutschland eV, Blankenheim, Germany),
Giovanni Neri (Università Cattolica S Cuore, Rome, Italy), Brian Neville
(University College London, London, UK), Sophie Nicole (Centre de
Recherche de l’Institut du Cerveau et de la Moelle, Centre National de la
Recherche Scientifi que, and Université Pierre et Marie Curie Paris VI,
Paris, France), Hang N Nielsen (Aarhus University, Aarhus, Denmark),
Poul Nissen (Aarhus University, and Centre for Membrane Pumps in Cells
and Disease—PUMPKIN, Danish National Research Foundation, Aarhus,
Denmark), Tom O’Brien (Alternating Hemiplegia of Childhood Ireland,
Dublin, Ireland), Laurie J Ozelius (Icahn School of Medicine at Mount
Sinai, New York, NY, USA), Eleni Panagiotakaki (University Hospitals of
Lyon, Lyon, France), Marek Parowicz (Polish AHC Association,
Bischofsheim, Poland), Dominique Poncelin (French AHC Organisation,
AFHA, St Germain lès Arpajon, France), Hanne Poulsen (Aarhus
University, and Centre for Membrane Pumps in Cells and Disease—
PUMPKIN, Danish National Research Foundation, Aarhus, Denmark),
Sandra P Reyna (University of Utah School of Medicine, Salt Lake City, UT,
USA), John C Roder (Mount Sinai Hospital and University of Toronto,
Toronto, ON, Canada), Hendrik Rosewich (Georg August University
Göttingen, Göttingen, Germany), Masayuki Sasaki (National Center of
Neurology and Psychiatry, Tokyo, Japan), Vivien R Schack (Aarhus
University, Aarhus, Denmark), Philippe Schyns (European Network for
Research on Alternating Hemiplegia [ENRAH], Brussels, Belgium),
Tsveta Schyns (ENRAH, Brussels, Belgium), Michela Stagnaro (University
of Genoa, Genoa, Italy), Kathleen J Sweadner (Massachusetts General
Hospital, Boston, MA, USA), Kathryn J Swoboda (University of Utah
School of Medicine, Salt Lake City, UT, USA), Danilo Francesco Tiziano
(Università Cattolica S Cuore, Rome, Italy), Mads S Toustrup-Jensen
(Aarhus University, Aarhus, Denmark), Arn van den Maagdenberg (Leiden
University Medical Centre, Leiden, Netherlands), Albert Vilamala (AHC
Federation of Europe, Reykjavík, Iceland, and AHC Association of Spain,
AESHA, Barcelona, Spain), Bente Vilsen (Aarhus University, Aarhus,
Denmark), Jeff T Wuchich (Cure AHC, Rolesville, NC, USA)
Declaration of interests
AB acts as a consultant for Allergan and Concert. All other authors
declare that they have no competing interests.
We acknowledge the support of the European Network for Research on
Alternating Hemiplegia, and Duke University for organising the fi rst
meeting of the ATP1A3 Working Group in Brussels, Belgium, in
December, 2012. This Review arose from the discussions and presentations
at that meeting. Other fi nancial contributors to this meeting include RTD
Services, Austria, French Association for Alternating Hemiplegia (AFHA),
Icelandic Association for Alternating Hemiplegia (AHCAI), Italian
Association for Alternating Hemiplegia (A.I.S.E.A Onlus), Spanish
Association for Alternating Hemiplegia (AESHA), Dutch Association for
Alternating Hemiplegia (AHC Vereniging Nederland), German Association
for Alternating Hemiplegia (AHC-Deutschland eV), and the US
Foundation for Alternating Hemiplegia (AHCF). We also thank individuals
who contributed to and supported the eff orts of the ATP1A3 Working
Group: Sharon Ciccodicola (AHCF), Thierry Billette de Villemeur,
Lynn Egan (AHCF), NardoNardocci, Anne Roubergue, Francesca Ragona,
Shoji Tsuji, Rosaria Vavassori, and Federico Vigevano. Additionally, some of
the work presented in this Review was supported by the Ontario Mental
Health Foundation Research Studentship, MOP 94856 from the Canadian
Institutes of Health Research, and a grant from the Amalgamated Transit
Union. Artwork shown in fi gure 5 was produced by Samantha Sliwa.
1 Boycott KM, Vanstone MR, Bulman DE, MacKenzie AE.
Rare-disease genetics in the era of next-generation sequencing:
discovery to translation. Nat Rev Genet 2013; 14: 681–91.
2 McGrail KM, Phillips JM, Sweadner KJ. Immunofl uorescent
localization of three Na,K-ATPase isozymes in the rat central
nervous system: both neurons and glia can express more than one
Na,K-ATPase. J Neurosci 1991; 11: 381–91.
3 Bøttger P, Tracz Z, Heuck A, Nissen P, Romero-Ramos M,
Lykke-Hartmann K. Distribution of Na/K-ATPase alpha 3 isoform, a
sodium-potassium P-type pump associated with rapid-onset of
dystonia parkinsonism (RDP) in the adult mouse brain.
J Comp Neurol 2011; 519: 376–404.
www.thelancet.com/neurology Vol 13 May 2014 513
4 Dobretsov M, Stimers JR. Neuronal function and alpha3 isoform of
the Na/K-ATPase. Front Biosci 2005; 10: 2373–96.
Dobretsov M, Hastings SL, Sims TJ, Stimers JR, Romanovsky D.
Stretch receptor-associated expression of alpha 3 isoform of the
Na+, K+-ATPase in rat peripheral nervous system. Neuroscience
2003; 116: 1069–80.
Crambert G, Hasler U, Beggah AT, et al. Transport and
pharmacological properties of nine diff erent human Na, K-ATPase
isozymes. J Biol Chem 2000; 275: 1976–86.
Senatorov VV, Mooney D, Hu B. The electrogenic eff ects of
Na(+)-K(+)-ATPase in rat auditory thalamus. J Physiol 1997;
Vaillend C, Mason SE, Cuttle MF, Alger BE. Mechanisms of
neuronal hyperexcitability caused by partial inhibition of Na+-K+-
ATPases in the rat CA1 hippocampal region. J Neurophysiol 2002;
Shiina N, Yamaguchi K, Tokunaga M. RNG105 defi ciency impairs
the dendritic localization of mRNAs for Na+/K+ ATPase subunit
isoforms and leads to the degeneration of neuronal networks.
J Neurosci 2010; 30: 12816–30.
10 Blom H, Rönnlund D, Scott L, et al. Spatial distribution of Na+-K+-
ATPase in dendritic spines dissected by nanoscale superresolution
STED microscopy. BMC Neurosci 2011; 12: 16.
11 Rose CR, Konnerth A. NMDA receptor-mediated Na+ signals in
spines and dendrites. J Neurosci 2001; 21: 4207–14.
12 Reinhard L, Tidow H, Clausen MJ, Nissen P. Na+,K+-ATPase as a
docking station: protein-protein complexes of the Na+,K+-ATPase.
Cell Mol Life Sci 2013; 70: 205–22.
13 Brashear A, DeLeon D, Bressman SB, Thyagarajan D, Farlow MR,
Dobyns WB. Rapid-onset dystonia-parkinsonism in a second family.
Neurology 1997; 48: 1066–69.
14 Dobyns WB, Ozelius LJ, Kramer PL, et al. Rapid-onset
dystonia-parkinsonism. Neurology 1993; 43: 2596–602.
15 Kramer PL, Mineta M, Klein C, et al. Rapid-onset dystonia-
parkinsonism: linkage to chromosome 19q13. Ann Neurol 1999;
16 Pittock SJ, Joyce C, O’Keane V, et al. Rapid-onset dystonia-
parkinsonism: a clinical and genetic analysis of a new kindred.
Neurology 2000; 55: 991–95.
17 Zaremba J, Mierzewska H, Lysiak Z, Kramer P, Ozelius LJ, Brashear A.
Rapid-onset dystonia-parkinsonism: a fourth family consistent with
linkage to chromosome 19q13. Mov Disord 2004; 19: 1506–10.
18 de Carvalho Aguiar P, Sweadner KJ, Penniston JT, et al. Mutations
in the Na+/K+ -ATPase alpha3 gene ATP1A3 are associated with
rapid-onset dystonia parkinsonism. Neuron 2004; 43: 169–75.
19 Brashear A, Dobyns WB, de Carvalho Aguiar P, et al. The
phenotypic spectrum of rapid-onset dystonia-parkinsonism (RDP)
and mutations in the ATP1A3 gene. Brain 2007; 130: 828–35.
20 McKeon A, Ozelius LJ, Hardiman O, Greenway MJ, Pittock SJ.
Heterogeneity of presentation and outcome in the Irish rapid-onset
dystonia-parkinsonism kindred. Mov Disord 2007; 22: 1325–27.
21 Barbano RL, Hill DF, Snively BM, et al. New triggers and non-
motor fi ndings in a family with rapid-onset dystonia-parkinsonism.
Parkinsonism Relat Disord 2012; 18: 737–41.
22 Lee JY, Gollamudi S, Ozelius LJ, Kim JY, Jeon BS. ATP1A3 mutation
in the fi rst Asian case of rapid-onset dystonia-parkinsonism.
Mov Disord 2007; 22: 1808–09.
23 Zanotti-Fregonara P, Vidailhet M, Kas A, et al. [123I]-FP-CIT and
[99mTc]-HMPAO single photon emission computed tomography in
a new sporadic case of rapid-onset dystonia-parkinsonism.
J Neurol Sci 2008; 273: 148–51.
24 Blanco-Arias P, Einholm AP, Mamsa H, et al. A C-terminal
mutation of ATP1A3 underscores the crucial role of sodium affi nity
in the pathophysiology of rapid-onset dystonia-parkinsonism.
Hum Mol Genet 2009; 18: 2370–77.
25 Anselm IA, Sweadner KJ, Gollamudi S, Ozelius LJ, Darras BT.
Rapid-onset dystonia-parkinsonism in a child with a novel ATP1A3
gene mutation. Neurology 2009; 73: 400–01.
26 Svetel M, Ozelius LJ, Buckley A, et al. Rapid-onset dystonia-
parkinsonism: case report. J Neurol 2010; 257: 472–74.
27 Kamm C, Fogel W, Wächter T, et al. Novel ATP1A3 mutation in
a sporadic RDP patient with minimal benefi t from deep brain
stimulation. Neurology 2008; 70: 1501–03.
28 Tarsy D, Sweadner KJ, Song PC. Case records of the Massachusetts
General Hospital. Case 17-2010 - a 29-year-old woman with fl exion
of the left hand and foot and diffi culty speaking. N Engl J Med 2010;
29 Roubergue A, Roze E, Vuillaumier-Barrot S, et al. The multiple
faces of the ATP1A3-related dystonic movement disorder.
Mov Disord 2013; 28: 1457–59.
30 Brashear A, Mink JW, Hill DF, et al. ATP1A3 mutations in infants:
a new rapid-onset dystonia-Parkinsonism phenotype characterized
by motor delay and ataxia. Dev Med Child Neurol 2012; 54: 1065–67.
31 Heinzen EL, Swoboda KJ, Hitomi Y, et al, and the European
Alternating Hemiplegia of Childhood (AHC) Genetics Consortium,
and the Biobanca e Registro Clinico per l’Emiplegia Alternante
(I.B.AHC) Consortium, and the European Network for Research on
Alternating Hemiplegia (ENRAH) for Small and Medium-sized
Enterpriese (SMEs) Consortium. De novo mutations in ATP1A3 cause
alternating hemiplegia of childhood. Nat Genet 2012; 44: 1030–34.
32 Rosewich H, Thiele H, Ohlenbusch A, et al. Heterozygous de-novo
mutations in ATP1A3 in patients with alternating hemiplegia of
childhood: a whole-exome sequencing gene-identifi cation study.
Lancet Neurol 2012; 11: 764–73.
33 Ishii A, Saito Y, Mitsui J, et al. Identifi cation of ATP1A3 mutations
by exome sequencing as the cause of alternating hemiplegia of
childhood in Japanese patients. PLoS One 2013; 8: e56120.
34 Hoei-Hansen CE, Dali CI, Lyngbye TJ, Duno M, Uldall P.
Alternating hemiplegia of childhood in Denmark: clinical
manifestations and ATP1A3 mutation status. Eur J Paediatr Neurol
2014; 18: 50–54.
35 Kamphuis DJ, Koelman H, Lees AJ, Tijssen MA. Sporadic rapid-
onset dystonia-parkinsonism presenting as Parkinson’s disease.
Mov Disord 2006; 21: 118–19.
36 Brashear A, Hill DF, Snively B, Sweadner KJ, Ozelius L. Denovo
and recurrent mutations in ATP1A3 are common in rapid-onset
dystonia-parkinsonism. Neurology 2010; 74 (suppl 2): A204.
37 Petrovski S, Wang Q, Heinzen EL, Allen AS, Goldstein DB. Genic
intolerance to functional variation and the interpretation of
personal genomes. PLoS Genet 2013; 9: e1003709.
38 Brashear A, Farlow MR, Butler IJ, Kasarskis EJ, Dobyns WB.
Variable phenotype of rapid-onset dystonia-parkinsonism.
Mov Disord 1996; 11: 151–56.
39 Brashear A, Butler IJ, Ozelius LJ, et al. Rapid-onset dystonia-
parkinsonism: a report of clinical, biochemical, and genetic studies
in two families. Adv Neurol 1998; 78: 335–39.
40 Brashear A, Butler IJ, Hyland K, Farlow MR, Dobyns WB.
Cerebrospinal fl uid homovanillic acid levels in rapid-onset
dystonia-parkinsonism. Ann Neurol 1998; 43: 521–26.
41 Brashear A, Cook JF, Hill DF, et al. Psychiatric disorders in rapid-
onset dystonia-parkinsonism. Neurology 2012; 79: 1168–73.
42 Bourgeois M, Aicardi J, Goutières F. Alternating hemiplegia of
childhood. J Pediatr 1993; 122: 673–79.
43 Neville BG, Ninan M. The treatment and management of
alternating hemiplegia of childhood. Dev Med Child Neurol 2007;
44 Sweney MT, Silver K, Gerard-Blanluet M, et al. Alternating
hemiplegia of childhood: early characteristics and evolution of
a neurodevelopmental syndrome. Pediatrics 2009; 123: e534–41.
45 Silver K, Andermann F. Alternating hemiplegia of childhood:
a study of 10 patients and results of fl unarizine treatment. Neurology
1993; 43: 36–41.
46 Casaer P. Flunarizine in alternating hemiplegia in childhood.
An international study in 12 children. Neuropediatrics 1987; 18: 191–95.
47 Casaer P, Azou M. Flunarizine in alternating hemiplegia in
childhood. Lancet 1984; 2: 579.
48 Sasaki M, Sakuragawa N, Osawa M. Long-term eff ect of fl unarizine
on patients with alternating hemiplegia of childhood in Japan.
Brain Dev 2001; 23: 303–05.
49 Kanavakis E, Xaidara A, Papathanasiou-Klontza D, Papadimitriou A,
Velentza S, Youroukos S. Alternating hemiplegia of childhood: a
syndrome inherited with an autosomal dominant trait.
Dev Med Child Neurol 2003; 45: 833–36.
50 Bassi MT, Bresolin N, Tonelli A, et al. A novel mutation in the
ATP1A2 gene causes alternating hemiplegia of childhood.
J Med Genet 2004; 41: 621–28.
514 Download full-text
www.thelancet.com/neurology Vol 13 May 2014
51 Swoboda KJ, Kanavakis E, Xaidara A, et al. Alternating hemiplegia
of childhood or familial hemiplegic migraine? A novel ATP1A2
mutation. Ann Neurol 2004; 55: 884–87.
52 Andermann E, Andermann F, Silver K, Levin S, Arnold D. Benign
familial nocturnal alternating hemiplegia of childhood. Neurology
1994; 44: 1812–14.
53 Wagener-Schimmel LJ, Nicolai J. Child neurology: benign nocturnal
alternating hemiplegia of childhood. Neurology 2012; 79: e161–63.
54 Mikati MA, Kramer U, Zupanc ML, Shanahan RJ. Alternating
hemiplegia of childhood: clinical manifestations and long-term
outcome. Pediatr Neurol 2000; 23: 134–41.
55 Saito Y, Sakuragawa N, Sasaki M, Sugai K, Hashimoto T. A case of
alternating hemiplegia of childhood with cerebellar atrophy.
Pediatr Neurol 1998; 19: 65–68.
56 Saito Y, Inui T, Sakakibara T, Sugai K, Sakuma H, Sasaki M.
Evolution of hemiplegic attacks and epileptic seizures in alternating
hemiplegia of childhood. Epilepsy Res 2010; 90: 248–58.
57 Dangond F, Garada B, Murawski BJ, Rey-Casserly C, Holman BL,
Mikati MA. Focal brain dysfunction in a 41-year old man with
familial alternating hemiplegia. Eur Arch Psychiatry Clin Neurosci
1997; 247: 35–41.
58 Mikati M. Alternating hemiplegia of childhood. Pediatr Neurol 1999;
59 Panagiotakaki E, Gobbi G, Neville B, et al, and the ENRAH
Consortium. Evidence of a non-progressive course of alternating
hemiplegia of childhood: study of a large cohort of children and
adults. Brain 2010; 133: 3598–610.
60 Verret S, Steele JC. Alternating hemiplegia in childhood: a report of
eight patients with complicated migraine beginning in infancy.
Pediatrics 1971; 47: 675–80.
61 Krägeloh I, Aicardi J. Alternating hemiplegia in infants: report of
fi ve cases. Dev Med Child Neurol 1980; 22: 784–91.
62 Mikati MA, Maguire H, Barlow CF, et al. A syndrome of autosomal
dominant alternating hemiplegia: clinical presentation mimicking
intractable epilepsy; chromosomal studies; and physiologic
investigations. Neurology 1992; 42: 2251–57.
63 Sasaki M, Ishii A, Saito Y, et al. Genotype-phenotype correlations in
alternating hemiplegia of childhood. Neurology 2014; 82: 482–90.
64 Di Michele M, Goubau C, Waelkens E, et al. Functional studies and
proteomics in platelets and fi broblasts reveal a lysosomal defect
with increased cathepsin-dependent apoptosis in ATP1A3 defective
alternating hemiplegia of childhood. J Proteomics 2013; 86: 53–69.
65 Post RL, Hegyvary C, Kume S. Activation by adenosine
triphosphate in the phosphorylation kinetics of sodium and
potassium ion transport adenosine triphosphatase. J Biol Chem
1972; 247: 6530–40.
66 Toustrup-Jensen MS, Einholm AP, Schack VR, et al. Relationship
between intracellular Na+ concentration and reduced Na+ affi nity in
Na+,K+-ATPase mutants causing neurological disease. J Biol Chem
2014; 289: 3186–97.
67 Koenderink JB, Geibel S, Grabsch E, De Pont JJ, Bamberg E,
Friedrich T. Electrophysiological analysis of the mutated Na,K-
ATPase cation binding pocket. J Biol Chem 2003; 278: 51213–22.
68 Poulsen H, Khandelia H, Morth JP, et al. Neurological disease
mutations compromise a C-terminal ion pathway in the
Na(+)/K(+)-ATPase. Nature 2010; 467: 99–102.
69 Toustrup-Jensen M, Vilsen B. Importance of Glu(282) in
transmembrane segment M3 of the Na(+),K(+)-ATPase for control
of cation interaction and conformational changes. J Biol Chem 2002;
70 Rodacker V, Toustrup-Jensen M, Vilsen B. Mutations Phe785Leu
and Thr618Met in Na+,K+-ATPase, associated with familial rapid-
onset dystonia parkinsonism, interfere with Na+ interaction by
distinct mechanisms. J Biol Chem 2006; 281: 18539–48.
71 Einholm AP, Toustrup-Jensen MS, Holm R, Andersen JP, Vilsen B.
The rapid-onset dystonia parkinsonism mutation D923N of the Na+,
K+-ATPase alpha3 isoform disrupts Na+ interaction at the third Na+
site. J Biol Chem 2010; 285: 26245–54.
72 Ogawa H, Shinoda T, Cornelius F, Toyoshima C. Crystal structure
of the sodium-potassium pump (Na+,K+-ATPase) with bound
potassium and ouabain. Proc Natl Acad Sci USA 2009; 106: 13742–47.
73 Kirshenbaum GS, Dawson N, Mullins JGL, et al. Alternating
hemiplegia of childhood-related neural and behavioural phenotypes
in Na+,K+-ATPase α3 missense mutant mice. PLoS One 2013;
74 Kaneko M, Desai BS, Cook B. Ionic leakage underlies a gain-of-
function eff ect of dominant disease mutations aff ecting diverse
P-type ATPases. Nat Genet 2014; 46: 144–51.
75 Clapcote SJ, Duff y S, Xie G, et al. Mutation I810N in the α3 isoform
of Na+,K+-ATPase causes impairments in the sodium pump and
hyperexcitability in the CNS. Proc Natl Acad Sci USA 2009;
76 Moseley AE, Williams MT, Schaefer TL, et al. Defi ciency in Na,K-
ATPase alpha isoform genes alters spatial learning, motor activity,
and anxiety in mice. J Neurosci 2007; 27: 616–26.
77 DeAndrade MP, Yokoi F, van Groen T, Lingrel JB, Li Y.
Characterization of Atp1a3 mutant mice as a model of rapid-onset
dystonia with parkinsonism. Behav Brain Res 2011; 216: 659–65.
78 Ikeda K, Satake S, Onaka T, et al. Enhanced inhibitory
neurotransmission in the cerebellar cortex of Atp1a3-defi cient
heterozygous mice. J Physiol 2013; 591: 3433–49.
79 Ashmore LJ, Hrizo SL, Paul SM, Van Voorhies WA, Beitel GJ,
Palladino MJ. Novel mutations aff ecting the Na, K ATPase alpha
model complex neurological diseases and implicate the sodium
pump in increased longevity. Hum Genet 2009; 126: 431–47.
80 Doğanli C, Beck HC, Ribera AB, Oxvig C, Lykke-Hartmann K.
α3Na+/K+-ATPase defi ciency causes brain ventricle dilation and
abrupt embryonic motility in zebrafi sh. J Biol Chem 2013;
81 Kirshenbaum GS, Clapcote SJ, Duff y S, et al. Mania-like behavior
induced by genetic dysfunction of the neuron-specifi c Na+,K+-
ATPase α3 sodium pump. Proc Natl Acad Sci USA 2011;
82 Kirshenbaum GS, Saltzman K, Rose B, Petersen J, Vilsen B,
Roder JC. Decreased neuronal Na+, K+-ATPase activity in Atp1a3
heterozygous mice increases susceptibility to depression-like
endophenotypes by chronic variable stress. Genes Brain Behav 2011;
83 Ferrari P, Ferrandi M, Valentini G, Bianchi G. Rostafuroxin: an
ouabain antagonist that corrects renal and vascular Na+-K+-ATPase
alterations in ouabain and adducin-dependent hypertension.
Am J Physiol Regul Integr Comp Physiol 2006; 290: R529–35.
84 Pizoli CE, Jinnah HA, Billingsley ML, Hess EJ. Abnormal cerebellar
signaling induces dystonia in mice. J Neurosci 2002; 22: 7825–33.
85 Calderon DP, Fremont R, Kraenzlin F, Khodakhah K. The neural
substrates of rapid-onset dystonia-parkinsonism. Nat Neurosci 2011;
86 Sweadner KJ. Isozymes of the Na+/K+-ATPase. Biochim Biophys Acta
1989; 988: 185–220.
87 Rajarao SJ, Canfi eld VA, Mohideen MA, et al. The repertoire of
Na,K-ATPase alpha and beta subunit genes expressed in the
zebrafi sh, Danio rerio. Genome Res 2001; 11: 1211–20.
88 Lebovitz RM, Takeyasu K, Fambrough DM. Molecular
characterization and expression of the (Na++K+)-ATPase alpha-
subunit in Drosophila melanogaster. EMBO J 1989; 8: 193–202.
89 Palladino MJ, Bower JE, Kreber R, Ganetzky B. Neural dysfunction
and neurodegeneration in Drosophila Na+/K+ ATPase alpha subunit
mutants. J Neurosci 2003; 23: 1276–86.