Seizures, sensorineural deafness, ataxia, mental
retardation, and electrolyte imbalance (SeSAME
syndrome) caused by mutations in KCNJ10
Ute I. Scholla, Murim Choia, Tiewen Liua, Vincent T. Ramaekersb, Martin G. Ha ¨uslerc, Joanne Grimmerd,1,
Sheldon W. Tobee, Anita Farhia, Carol Nelson-Williamsa, and Richard P. Liftona,2
aDepartment of Genetics, The Howard Hughes Medical Institute, Yale University School of Medicine, 333 Cedar Street, New Haven, CT 06510;bDepartment
of Paediatrics, Centre Hospitalier Universitaire, Domaine Universitaire du Sart Tilman, Ba ˆtiment B 35, B-4000 Lie `ge 1, Belgium;cDepartment of Paediatrics,
University Hospital, RWTH Aachen, Pauwelsstrasse 30, 52074 Aachen, Germany;dDepartment of Nephrology, The Hospital for Sick Children, 555 University
Avenue, Toronto, ON, Canada M5G 1X8; andeDivision of Nephrology, Sunnybrook Health Sciences Centre, 2075 Bayview Avenue, Toronto, ON, Canada
Contributed by Richard P. Lifton, February 18, 2009 (sent for review January 29, 2009)
We describe members of 4 kindreds with a previously unrecog-
nized syndrome characterized by seizures, sensorineural deafness,
ataxia, mental retardation, and electrolyte imbalance (hypokale-
mia, metabolic alkalosis, and hypomagnesemia). By analysis of
linkage we localize the putative causative gene to a 2.5-Mb
segment of chromosome 1q23.2–23.3. Direct DNA sequencing of
KCNJ10, which encodes an inwardly rectifying K?channel, identi-
fies previously unidentified missense or nonsense mutations on
both alleles in all affected subjects. These mutations alter highly
conserved amino acids and are absent among control chromo-
somes. Many of these mutations have been shown to cause loss of
function in related K?channels. These findings demonstrate that
loss-of-function mutations in KCNJ10 cause this syndrome, which
we name SeSAME. KCNJ10 is expressed in glia in the brain and
spinal cord, where it is believed to take up K?released by neuronal
repolarization, in cochlea, where it is involved in the generation of
endolymph, and on the basolateral membrane in the distal
nephron. We propose that KCNJ10 is required in the kidney for
the need for K?recycling across the basolateral membrane to
enable normal activity of the Na?-K?-ATPase; loss of this function
accounts for the observed electrolyte defects. Mice deficient for
KCNJ10 show a related phenotype with seizures, ataxia, and
hearing loss, further supporting KCNJ10’s role in this syndrome.
These findings define a unique human syndrome, and establish the
essential role of basolateral K?channels in renal electrolyte
Gitelman syndrome ? hypokalemia ? hypomagnesemia ?
inwardly rectifying K?channel ? renal salt wasting
including neuronal signal transmission and electrolyte and vol-
ume homeostasis. In many cases, homologous electrolyte flux
processes in different tissues are mediated by the encoded
products of distinct genes, while in a few cases the identical gene
products are involved. Evidence of the latter comes from Men-
delian diseases in which mutation in a single gene produces
effects on both auditory and renal function. For example,
loss-of-function mutations in ATP6B1, which encodes a subunit
of the H?-ATPase, result in systemic acidosis because of a renal
defect in H?secretion and sensorineural hearing loss caused by
defective H?secretion into the cochlear endolymph, resulting in
impaired hair cell function and deafness (1). Similarly, mutations
in barttin, which encodes a subunit of the CLCNKA and
CLCNKB chloride channels, result in renal salt wasting and
The genetic dissection of renal diseases featuring low serum
potassium (hypokalemia) and metabolic alkalosis (high serum
ransmembrane ion flux via channels, transporters, and
pumps plays a critical role in diverse physiologic functions,
pH) has identified many components required for normal renal
electrolyte homeostasis (2–7). In all cases, this syndrome has
resulted from increased activity of the epithelial Na?channel
(ENaC) on the apical membrane, which leads to increased
secretion of K?and H?because of the more negative luminal
potential. Hypokalemia with alkalosis can result either from
primary increases in ENaC activity because of mutations in
ENaC itself (3), or from activation of ENaC by aldosterone in
response to reduced intravascular volume (7). Mutations that
cause impaired salt reabsorption in the thick ascending limb of
Henle or the distal convoluted tubule cause salt wasting that
leads to secondary increases in ENaC activity and hypokalemic
alkalosis. Identified mutations in these diseases, referred to as
Bartter and Gitelman syndromes, are in genes including 2 apical
Na?-Cl?transporters that mediate Na-Cl entry into epithelia, 2
Cl?channel subunits that mediate exit of Cl?across the
basolateral membrane, and an apical K?channel (2, 4–7). These
syndromes are distinguished clinically by marked hypomag-
nesemia and low urinary calcium in Gitelman syndrome, while
hypercalciuria with normal or modest reductions in Mg2?is
observed in Bartter syndrome.
Similarly, a number of Mendelian seizure disorders have been
described. Many of these result from mutations that depolarize
neurons, increasing neuronal excitability and reducing seizure
threshold. Examples include benign familial neonatal seizures
caused by mutations in the KCNQ2/3 K?channels (8, 9), benign
familial neonatal/infantile seizures caused by mutations in the
SCN2A gene encoding the alpha subunit of voltage gated Na?
channels (10), and several idiopathic epilepsy syndromes caused
by mutations in the SCN1A sodium channel (11).
Considering the many similarities in the mechanisms govern-
ing renal electrolyte homeostasis and neuronal function, it is
surprising that relatively few single-gene disorders that have
effects on both have been identified. Here, we describe a
previously unrecognized complex syndrome featuring seizures,
sensorineural deafness, ataxia, mental retardation and electro-
lyte imbalance (SeSAME), and demonstrate that it is caused
by mutation in KCNJ10, which encodes a K?channel expressed
in epithelia of the kidney and inner ear, as well as glial cells in
Author contributions: U.I.S., M.C., and R.P.L. designed research; U.I.S., M.C., T.L., V.T.R.,
analyzed data; and U.I.S., M.C., and R.P.L. wrote the paper.
The authors declare no conflict of interest.
Freely available online through the PNAS open access option.
1Present address: Department of Paediatrics, University of Western Ontario, 800 Commis-
sioners Road East, London, Ontario, Canada N6A 5W9.
2To whom correspondence should be addressed. E-mail: email@example.com.
April 7, 2009 ?
vol. 106 ?
no. 14 www.pnas.org?cgi?doi?10.1073?pnas.0901749106
Case Report: Kindred 441. The index case, patient 441–1, is a
24-year-old female of Afghan ancestry who is the fifth of 6
offspring of healthy first-cousins. Generalized seizures began at
3 months, occurring several times daily. Seizures were controlled
initially with phenobarbital, and later diphenylhydantoin. Sitting
was first demonstrated at age 1 year, crawling at age 20 months.
At age 5 years 7 months, the patient presented for evaluation
of developmental delay. At this time, she was unable to walk or
speak. Physical examination was notable for atrophy of the lower
extremities without contractures. Motor strength was slightly
reduced in the upper, but markedly reduced in the lower
extremities. Reflexes were normal. There was marked ataxia.
Nerve conduction studies revealed reduced conduction velocity
in the left peroneal and left tibial nerves (36 and 37 m/sec,
respectively; nl 40–44 m/sec). A muscle and nerve biopsy showed
normal muscle other than fiber-type disproportion; there was
hypomyelination of the large myelinated nerve fibers in the sural
showed normal myelination, and was normal with the exception
of slightly coarsened frontal sulci. EEG, abdominal ultrasound,
and karyotype analysis were normal.
Laboratory evaluation was remarkable for persistent hypoka-
lemia, metabolic alkalosis, and hypomagnesemia (Table 1).
Plasma renin activity (PRA) was elevated on repeated measures
(8.1 and 7.6 ng/ml per hour, nl ?2.8 ng/ml per hour). Twenty-
four-hour urinary aldosterone level was elevated (31.3 ?g; nl
5.9–17.6) and the Ca2?/creatinine ratio was low (0.1 to 0.2
mmol/mmol). The patient was treated with oral potassium
replacement, and required 30 meq per day to maintain a K?level
in the normal range. At age 18, progressive hearing loss was
noted. Brainstem-evoked response audiometry and pure-tone
threshold audiometry were performed, and moderate-to-severe
sensorineural hearing loss was documented.
The patient’s family history is notable for 2 of 5 siblings with
a male, presented with seizures at age 4 months and was never
sixth child, a male, presented with seizures and vomiting at age
5 months and was unable to walk until 16 months. He died at 18
Definition of a New Clinical Syndrome. In the review of 589 subjects
referred for evaluation of Gitelman and Bartter syndromes, we
recognized subjects from 3 additional kindreds with features
seizures, mental retardation, ataxia, hypotonia, and sensorineu-
ral hearing loss. Intention tremor was noted in several cases and
volume loss of the cerebellum in 2 cases (632–2 and 404–1).
Short stature was notable in 2 affected siblings (632–1 and
632–2), with a final height of 150 cm and 149 cm, respectively.
Electrolyte abnormalities featured marked and persistent hy-
pokalemic metabolic alkalosis in the absence of hypertension
and striking hypomagnesemia that required electrolyte replace-
ment and, in many cases, use of pharmacologic inhibitors of the
epithelial sodium channel or aldosterone antagonists to prevent
renal K?loss. Where available, 24-h urinary electrolyte mea-
surements revealed renal K?and Mg2?wasting and high urinary
Na?levels. PRA and aldosterone levels, when measured, were
always elevated, and salt craving, enuresis, and polyuria/
laboratory findings is provided in Table 1.
Mapping the Disease Locus. The recurrent clinical features sug-
gested a previously unrecognized clinical syndrome. Moreover,
recurrence of a similar syndrome among siblings in 2 of these
kindreds, and its occurrence in the setting of parental consan-
guinity in 2 kindreds, suggested autosomal recessive transmis-
sion. To attempt to map the underlying disease locus, we
performed genome-wide analysis of linkage in the 3 informative
kindreds (2 offspring of first-cousin marriage and 2 affected
siblings from unrelated parents). The results demonstrated
complete linkage of the putative trait locus to a single chromo-
some segment, 1q23.2-q23.3, with a lod score of 3.0. The
maximum likelihood location is confined to a 2.5-Mb interval
chromosome segment showed linkage in all 3 kindreds. Signif-
icantly, the index case (441–1) did not show homozygosity at any
of the known loci for Bartter or Gitelman syndrome, and
sequencing of these genes revealed no evidence of pathogenic
Mutations in KCNJ10. The linked interval on chromosome 1
contains 70 well-defined and at least 6 hypothetical genes, none
of which has previously been implicated in human disease
phenotypes that would explain the features found in these
patients. If the disease is caused by mutation in a single gene, we
anticipate that it would likely be expressed in the CNS, inner ear,
and kidney. We considered ion channels, transporters, and
Table 1. Clinical features of affected patients
Seizures (age of onset in months)
K?, serum potassium, nl 3.5–5 mmol/l; Mg2?, serum magnesium, nl 0.8–1.2 mmol/l; HCO3?, serum bicarbonate, nl 23–26 mmol/l;
UCa/UCr, urinary calcium/creatinine ratio, nl ? 0.4 mmol/mmol); UK/UCr, urinary potasssium/creatinine ratio [nl 6–8 mmol/mmol, or 1–1.5
Gitelman syndrome (30).
Scholl et al.PNAS ?
April 7, 2009 ?
vol. 106 ?
no. 14 ?
regulators of their function to be leading candidates. We iden-
tified KCNJ10, which encodes the inwardly rectifying K?channel
Kir4.1 (also known as BIR10, BIRK1, KAB-2, Kir1.2), consisting
of 2 transmembrane segments and 1 pore, as a strong candidate
(Fig. 1B). KCNJ10 has been shown to be expressed in the CNS,
cochlea, and distal nephron, and a mouse knockout has a
neurological phenotype with many features similar to those seen
in our patients (see Discussion).
We screened available members of the 4 kindreds with this
syndrome for KCNJ10 mutations by direct sequencing (Fig. 2).
We found homozygous missense mutations in the 2 consanguin-
eous kindreds, compound heterozygous missense mutations in 1
outbred kindred, and a compound missense/premature termi-
nation mutation in 1 kindred. In the compound heterozygous
patients, cloning of the coding region on single amplicons
confirmed that the 2 mutations identified are in trans. To assess
the significance of missense mutations, we compared the
KCNJ10 amino acid sequence to orthologs in diverse vertebrate
species including mammals, Xenopus tropicalis, and zebrafish,
and closely related paralogs in chicken and Drosophila melano-
?500 million years ago, and across these species, only 27% of the
amino acids were completely conserved. All of the identified
mutations occurred at positions that were completely conserved
completely conserved through Drosophila (see Fig. 2).
Patient 327–1 was compound heterozygous for a nonsense and
a missense mutation (see Fig. 2A). The nonsense mutation
introduces a premature termination codon at position 199 in the
is known to be required for expression of Kir4.1 at the cell
surface (12). This patient also harbors an R65P substitution
immediately preceding the first transmembrane domain (Fig. 3).
This position is conserved in a related inward rectifier, Kir2.1,
and expression of Kir2.1 containing mutation at this position
abolished nearly all detectable whole-cell K?current when
expressed in Xenopus oocytes (13).
mutation (see Fig. 2B). C140 is located in the P region near the
start of the second transmembrane domain (see Fig. 3). This
position is conserved in the related channel ROMK (position
C153), and mutation of this residue to either alanine or serine
has been shown to abolish ROMK function (14).
A missense mutation was found in kindred 441, resulting in a
T164I substitution (see Fig. 2C). The index case is homozygous
for the mutant allele, while both parents are heterozygotes and
shown, with homozygous regions in patient 441–1 and 404–1 indicated by
light gray boxes, and segments that are identical by descent (IBD) in the
siblings 632–1 and 632–2 marked by dark gray boxes. The overlap of these
intervals is marked, and represents the maximum likelihood location of the
disease locus on chromosome 1q23.2-q23.3, a 4-cM interval covering 2.5 Mb
on chromosome 1q23.2. Neighbouring genes are represented by arrows in
their corresponding transcriptional orientations. KCNJ10 comprises 2 exons
indicated by boxes, with the coding sequence indicated in black.
Mapping the disease locus. (A) An ideogram of chromosome 1 is
(Right) are shown. The sequence of the encoded peptide is indicated in single
with orthologs and paralogs from Mus musculus (M.m.), Gallus gallus (G.g.),
Xenopus tropicalis (X.t.), Danio rerio (D.r.), and Drosophila melanogaster
(D.m.) is shown next to each mutation. The human sequence and residues
conserved in orthologs and paralogs are marked in yellow, and the mutant
residue is indicated. (A) Patient 327–1 is compound heterozygous for a mis-
sense and a nonsense mutation in KCNJ10. (B) Patient 404–1 is homozygous
for a missense mutation, changing codon TGT (C140) to CGT (R140). (C) A
of codon ACC (T164) to ATC (I164). (D) In kindred 632, both affected siblings
are compound heterozygous for missense mutations: A167V and R297C.
Mutations in KCNJ10 in affected patients. In each panel the DNA
schematic view of the protein is shown, with intracellular N- and C-termini, 2
transmembrane helices (plasma membrane shown in shaded gray), and 1
Location of KCNJ10 mutations in patients with SeSAME syndrome. A
www.pnas.org?cgi?doi?10.1073?pnas.0901749106Scholl et al.
neither of 2 unaffected siblings are homozygous, providing
further support for linkage (total lod score for linkage after
et al. (15) have suggested that T164, which is located in the
second transmembrane domain, forms an intra-subunit H-bond
with lysine 67 in the first transmembrane domain, and that this
interaction is critical for gating of the channel. Because this
lysine is predicted to form an H-bond not only with the backbone
to isoleucine would eliminate this interaction and potentially
affect the gating properties of the channel in response to pH and
phosphatidylinositol 4,5-bisphosphate (PIP2) (see Discussion).
Finally, in kindred 632, both affected siblings are compound
heterozygotes for A167V and R297C mutations (see Fig. 2D).
A167 is located at the end of the second transmembrane domain,
close to the constriction at the inner helix bundle that likely
corresponds to the gate of the channel (16). R297 lies in a highly
conserved segment in the C terminus of the protein (see Fig. 3).
Notably, a mutation at the residue corresponding to R297 has
been found in ROMK (R292W), and was implicated as a loss of
function mutation in Bartter’s syndrome type II (17). Similarly,
mutation of the conserved position in Kir2.1 (R312) to glu-
tamine greatly reduces whole-cell currents and produces weak-
ened interaction with PIP2(see Discussion) (13).
None of the identified mutations are in the dbSNP database.
Resequencing of KCNJ10 in 103 unrelated Caucasian subjects
did not identify any of these mutations and no missense variants
at conserved residues were identified in any of the 206 alleles
We have defined a previously unrecognized human syndrome
featuring prominent neurological and renal features and have
demonstrated that in all 4 kindreds studied the disease coseg-
regates with rare mutations in KCNJ10. The finding of 6 inde-
pendent rare KCNJ10 mutations in 4 families that significantly
cosegregate with the disease under a recessive model and which
show specificity for the disease provides genetic evidence that
these mutations are the cause of this syndrome. The fact that
many of the amino acids altered by mutations are conserved in
other members of the inward rectifier K?channel family and
have been shown to be essential for their normal function lends
strong support for the functional significance of these mutations.
The genetic and biochemical evidence support these mutations
being a genetic loss of function.
Several of the identified mutations are likely to affect channel
activity via altered interaction with PIP2. Numerous functional
studies in closely related inward rectifier potassium channels
have underlined the crucial role of PIP2to sustain activity of
these channels (13). PIP2 is a membrane-delimited second
messenger, and binds proteins through electrostatic interactions
at basic amino acids. These sites have been defined in many
members of this gene family, and mutations at PIP2binding sites
have been implicated in other channelopathies, including Bart-
ter’s syndrome, caused by ROMK mutations, and Andersen’s
syndrome, caused by mutations in KCNJ2 encoding Kir2.1. A
similar mechanism likely accounts for loss of function in at least
2 of the mutations identified here (R65P and R297C), which lie
at inferred PIP2binding sites. A third residue (T164) has been
implicated in an H-bond between the 2 transmembrane helices,
which again plays an important role in the channel’s gating in
response to PIP2and pH (15).
Significant prior work has been done on KCNJ10; it appears
to function as a heteromultimer at least in some tissues. The
currents observed in native tissues have properties most similar
to those produced by coexpression of Kir4.1 and Kir5.1 in
heterologous systems, and immunofluorescence studies support
colocalization of these gene products (18, 19). This observation
raises the question of whether a related syndrome might be
caused by mutation in Kir5.1.
In addition, mice with both constitutional and selective astro-
cyte knockout of KCNJ10 have been produced, with a resultant
phenotype that is strikingly similar to the patients we describe
(20–22). The animals develop motor coordination deficits with
awkward and jerky movements and loss of balance, and drag the
hind limbs. They also suffer seizures and have sensorineural
hearing loss (20, 23). Additionally, mice with the constitutional
knockout appear to have a salt-wasting phenotype; however, this
has not been well defined. These findings strongly support the
mutations we identify as being loss of function.
In the brain, KCNJ10 appears to be primarily expressed in glial
cells (24), specifically in astrocytes surrounding synapses and
blood vessels, and oligodendrocyte cell bodies (20). Neuronal
repolarization after excitatory stimuli is achieved via efflux of
K?, and it has been proposed that KCNJ10 plays a role in
astrocyte clearance of this K?via ‘‘spatial buffering.’’ If the
resting membrane potential is set by KCNJ10, a local increase in
extracellular K?concentration close to the synapse would favor
lower extracellular K?concentrations (i.e., the rise in extracel-
lular K?would cause the local glial EKto be less negative than
the aggregate cellular membrane potential). Loss of KCNJ10
would thus result in astrocyte depolarization [which is seen in
astrocytes from KCNJ10-deficient mice (20)], loss of this K?
clearance function, prolonged neuronal depolarization, and re-
duced seizure threshold. Similarly, astrocyte depolarization
would reduce clearance of the excitatory neurotransmitter glu-
tamate, which would also reduce seizure threshold [reduced
glutamate uptake is also seen in astrocytes from KCNJ10-
deficient mice (20)]. While other mechanisms (activities of the
involved in the regulation of synaptic K?(25), the observed
seizure activity in humans deficient for KCNJ10 indicates an
important role of this channel in prevention of seizure activity.
Finally, it is of interest that common variation in the KCNJ10
gene has been suggested to be associated with seizure suscep-
tibility (26), however, the functional significance of the impli-
cated variants and the replicability of this finding has not been
Kir4.1 is expressed in intermediate cells of the stria vascularis
(27), where it is believed to contribute to the generation of the
endocochlear potential, as demonstrated by hearing loss in the
KCNJ10-knockout mouse (23) and the patients described herein.
Both mice and humans with KCNJ10 mutations have marked
ataxia and there is also lower extremity weakness in the mouse
and some affected humans. Whether the ataxia is cerebellar in
origin or sensory (due to loss of proprioception) has not been
of affected subjects. Intention tremor and volume loss in the
cerebellum, as seen in some cases, suggest cerebellar involve-
ment. However, peripheral sensory neuropathy might also con-
tribute to the ataxia.
KCNJ10-deficient mice exhibit striking pathology of the spinal
cord with dysmyelination, hypomyelination, and axonal degen-
eration along with massive spongiform vacuolation. MRI dem-
onstrates marked white-matter pathology in the spinal cord and
brainstem, while cerebellum, midbrain, and cortical regions
seem unaffected at P12. It thus appears that Kir4.1 is required
for oligodendrocyte development, and at least spinal cord my-
elination (22). The observation that sural nerve biopsy in one of
our patients showed hypomyelination suggests a possible role for
KCNJ10 in the peripheral nervous system as well, and Kir4.1 has
been shown to be expressed in satellite cells (28).
The distinct electrolyte abnormalities in our patients add
considerable new insight into the role of KCNJ10 in renal
electrolyte homeostasis. The KCNJ10 gene product has been
Scholl et al. PNAS ?
April 7, 2009 ?
vol. 106 ?
no. 14 ?
immunolocalized in the kidney. In contrast to the apical K?
channels (e.g., KCNJ1 encoding ROMK and KCNMA1 encoding
Maxi-K) that mediate K?secretion in the distal nephron, Kir4.1
localizes to the basolateral membranes of epithelia of the distal
convoluted tubule, connecting tubule, and initial collecting
tubule (29). Weak immunoreactivity in the thick ascending limb
of Henle has also been described (12). Our patients with KCNJ10
deficiency display hypokalemia, metabolic alkalosis, hypomag-
nesemia and, where studied, elevated levels of renin and aldo-
sterone. The high renin and aldosterone, along with normal
blood pressure, hypokalemia, and metabolic alkalosis strongly
point to salt wasting as an incipient event in the renal features,
and the reports of salt craving, polyuria, and enuresis are
consistent with this. Moreover, these patients have elevated
urinary sodium/creatinine ratios in the range seen in patients
with Gitelman syndrome (see Table 1) (30). Loss of KCNJ10
function can result in salt wasting by impairing the activity of the
Na?-K?-ATPase. The Na?-K?-ATPase is on the basolateral
membrane, and its activity is required for Na?reabsorption,
pumping Na?out of epithelia and K?in against their electro-
chemical potentials. Because very large amounts of filtered Na?
must be reabsorbed by renal epithelia, the K?that enters the
epithelial cell must be recycled to the interstitium by basolateral
K?channels to allow continued Na?reabsorption. Without this
mechanism, the Na?-K?-ATPase could be inhibited and the
potential across the basolateral membrane diminished. This
diminished negative intracellular potential will also attenuate
the electrical gradient for the efflux of Cl?(Fig. 4B). The
combined effects will produce impaired Na-Cl reabsorption.
Because much is known about the consequences of inhibition
of salt reabsorption in different nephron segments (31, 32), we
can make inferences about where the effects of KCNJ10 defi-
ciency are impairing salt reabsorption. The only site at which
inhibition of salt reabsorption produces hypomagnesemia with
reduced urinary calcium is the distal convoluted tubule. In
contrast, loss of salt reabsorption in the thick ascending limb
produces marked hypercalciuria and little hypomagnesemia (5),
while loss of ENaC activity in the collecting duct produces
hyperkalemia and acidosis rather than hypokalemia and alka-
losis (33). We consequently believe it is highly likely that
impaired salt reabsorption in the DCT plays a prominent role in
this syndrome. Because epithelial cells of the DCT have the
greatest per-cell Na?reabsorption and energy demand (34), it is
possible that sodium pump activity in other nephron segments is
also affected, but that the effect in the DCT predominates.
These considerations suggest an integrated model in which
loss of Kir4.1 activity impairs salt reabsorption in the distal
convoluted tubule (see Fig. 4). Because salt reabsorption in the
DCT comprises ?7% of the filtered load, loss of salt reabsorp-
tion here induces salt wasting, which activates the renin-
angiotensin system, increasing Na?reabsorption by the ENaC in
the connecting tubule and collecting duct. This increases the
electrical driving force for both K?and H?secretion, resulting
in hypokalemia and metabolic alkalosis. As described above, loss
of salt reabsorption in the DCT is also known to produce
hypocalciuria and hypomagnesemia. While these renal electro-
lyte defects seem relatively mild, it is noteworthy that 2 siblings
with this syndrome have died in the setting of diarrheal or other
intercurrent infections, suggesting impaired ability to defend
volume homeostasis under stress.
Little is known about the genes that underlie the most
prevalent forms of epilepsy. In the last decade, gene defects have
been identified that cause rare Mendelian forms of idiopathic
epilepsy syndromes, and most of these genes encode ion chan-
nels, consistent with their role in maintaining membrane poten-
to determine the prevalence of epilepsy caused by mutations in
the KCNJ10 gene. In addition to the characteristic neurological
features (developmental delay, ataxia, and hearing impairment),
a simple blood test might help to screen for such patients, as all
patients in this report presented with significant hypokalemia
These human findings raise the possibility that Kir4.1 could be
a useful target for pharmacologic manipulation. Similar to the
recently developed anticonvulsant drug retigabine, which opens
KCNQ2/3 channels (35), a Kir4.1 activator might have anticon-
vulsant effects; nonetheless, the expression of Kir4.1 in several
tissues raises the question of whether there might be pleiotropic
effects that could limit utility.
In summary, our data define a unique autosomal, recessive
syndrome characterized by seizures, sensorineural deafness,
ataxia, mental retardation, and electrolyte imbalance (hypokale-
mic alkalosis and hypomagnesemia), and demonstrate that it is
caused by mutations in KCNJ10. We propose the acronym
SeSAME to refer to this disorder.
Materials and Methods
the Yale Human Investigation Committee. Consent for participation was
obtained in accordance with Institutional Review Board standards. Patients
were referred for studies of hypokalemic salt-losing nephropathies, and kin-
hearing impairment. Genomic DNA was prepared from venous blood of
kindred members by standard procedures.
Genotyping. The samples were genotyped on the Illumina Human CNV370-
Duo (for 441–1) and Illumina Human 610-Quad (for 404–1, 632–1, and 632–2)
beadchips at the W.M. Keck Facility at Yale University. Sample processing and
of the 4 samples was 99.50%.
Mapping Homozygous and IBD Intervals. Because the genotype data from
441–1 and 404–1 were originated from different arrays, the data were com-
pared to generate a list of 346,073 shared SNPs to be subjected to homozy-
gosity mapping. Analysis of homozygous segments across 22 autosomes was
performed using the ‘‘Runs of homozygosity’’ tool implemented in PLINK
(v1.05). A fixed threshold of 200 consecutive SNPs and 2 Mb in length was
selected, and 1 heterozygous SNP within a segment was allowed.
To check the IBD shared regions between the affected siblings, the geno-
type data from 632–1 and 632–2 were directly compared. Missing calls were
caused by mutations in the Kir4.1 inwardly rectifying potassium channel. (A)
Kir4.1/5.1 heteromultimers in the basolateral membrane of the distal convo-
luted tubule (DCT) recycle potassium entering the cell via the Na?-K?-ATPase
back into the interstitial space and contribute to the negative membrane
potential that promotes basolateral chloride exit. On the luminal surface,
sodium and potassium enter the cell via the thiazide sensitive cotransporter
NCCT, and Mg2?enters via TrpM6, using the favorable electrical gradient. (B)
Disruption of Kir4.1 function inhibits the function of the Na?-K?-ATPase via
loss of potassium recycling, reduces basolateral chloride reabsorption by
rendering the membrane potential (Em) less negative, and thereby inhibits
both apical Na?and Cl?reabsorption by NCCT and Mg2?reabsorption be-
cause of a less negative membrane potential. The resulting renal salt loss
activates the renin-angiotensin-aldosterone system. Increased amounts of
Na?and Cl?are delivered to the cortical collecting duct, where aldosterone
dependent Na?reabsoption via ENaC is coupled to K?and H?secretion (see
Discussion), thus accounting for the hypokalemic alkalosis observed.
A model of impaired ion transport in the distal convoluted tubule
www.pnas.org?cgi?doi?10.1073?pnas.0901749106 Scholl et al.
discarded. Fixed thresholds of 200 consecutive SNPs and 1 Mb in length were
used to call IBD segments.
DNA Sequencing. A primer pair (KCNJ10?F: 5?-CATGGGGTGAGGGTTAGGAG-3?
and KCNJ10?R: 5?-GGGAGTGGAGGATGGGTG-3?) was used to amplify the
coding exon of KCNJ10 using as a template genomic DNA of disease family
members or controls. PCR generated a product with a size of 1,325 bp.
Products were analyzed via gel electrophoresis, and purified amplicons were
sequenced using the KCNJ10?F, KCNJ10?R, KCNJ10MF (5?-CGGGCTGAGAC-
CATTCGTTTC-3?) and KCNJ10MR (5?-AGGCTTTTGCGCATATTGGAAC-3?) prim-
ers. Disease-causing mutations were confirmed by at least 2 independent
sequences from different primers. In addition, in the 2 kindreds in which
was amplified and cloned using the TOPO TA Cloning Kit (Invitrogen), and
the identified mutations were in cis or trans.
Orthologs and Paralogs. Full-length orthologous and paralogous protein
sequences from vertebrate and invertebrate species (including rodents, bird,
fish, and fly) were extracted from GenBank. Orthologs were confirmed based
on database identity of annotation or in a BLAST of the protein sequence
against the human protein sequence, with the requirement that human
KCNJ10 be the top hit. If an ortholog could not be identified, a paralog was
studied. Protein sequences were aligned using the ClustalW algorithm. Gen-
(mouse KCNJ10), XP?425554.2 (chicken paralog), NP?001072312.1 (frog
KCNJ10), XP?001342993.1 (zebrafish ortholog), and NP?001097884.1 (fly
ACKNOWLEDGMENTS. We thank the members of the families studied and
their physicians for their invaluable contribution to this project. We thank Dr.
Sally-Anne Hulton (The Birmingham Children’s Hospital) and the Department
of Paediatrics, Klinikum Worms, Germany for referral of patients 327–1 and
404–1, respectively; Dr. Willem Proesmans (University Hospital Gasthuisberg,
Belgium) for advice in the evaluation of patient 441–1; Dr. Gerhard Giebisch
(Yale University School of Medicine), Dr. Stephen Waxman (Department of
Neurology, Yale University School of Medicine), and members of the Lifton,
State and Gu ¨nel laboratories (Department of Genetics, Yale University School
of Medicine) for helpful discussions. This study was supported by the Leducq
Transatlantic Network in Hypertension and the Yale O’Brien Center.
1. Karet FE, et al. (1999) Mutations in the gene encoding B1 subunit of H?-ATPase cause
renal tubular acidosis with sensorineural deafness. Nat Genet 21:84–90.
2. Birkenhager R, et al. (2001) Mutation of BSND causes Bartter syndrome with sensori-
neural deafness and kidney failure. Nat Genet 29:310–314.
gamma subunit: genetic heterogeneity of Liddle syndrome. Nat Genet 11:76–82.
syndrome type III. Nat Genet 17:171–178.
is caused by mutations in the Na-K-2Cl cotransporter NKCC2. Nat Genet 13:183–188.
6. Simon DB, et al. (1996) Genetic heterogeneity of Bartter’s syndrome revealed by
mutations in the K? channel, ROMK. Nat Genet 14:152–156.
7. Simon DB, et al. (1996) Gitelman’s variant of Bartter’s syndrome, inherited hypokal-
aemic alkalosis, is caused by mutations in the thiazide-sensitive Na-Cl cotransporter.
Nat Genet 12:24–30.
8. Biervert C, et al. (1998) A potassium channel mutation in neonatal human epilepsy.
9. Charlier C, et al. (1998) A pore mutation in a novel KQT-like potassium channel gene
in an idiopathic epilepsy family. Nat Genet 18:53–55.
10. Heron SE, et al. (2002) Sodium-channel defects in benign familial neonatal-infantile
seizures. Lancet 360:851–852.
12. Tanemoto M, Abe T, Onogawa T, Ito S (2004) PDZ binding motif-dependent localiza-
tion of K? channel on the basolateral side in distal tubules. Am J Physiol Renal Physiol
13. Lopes CM, et al. (2002) Alterations in conserved Kir channel-PIP2 interactions underlie
channelopathies. Neuron 34:933–944.
14. Schulte U, et al. (1998) pH-dependent gating of ROMK (Kir1.1) channels involves
conformational changes in both N and C termini. J Biol Chem 273:34575–34579.
15. Rapedius M, et al. (2007) Control of pH and PIP2gating in heteromeric Kir4.1/Kir5.1
channels by H-Bonding at the helix-bundle crossing. Channels (Austin) 1:327–330.
16. Nishida M, Cadene M, Chait BT, MacKinnon R (2007) Crystal structure of a Kir3.1-
prokaryotic Kir channel chimera. EMBO J 26:4005–4015.
17. Ji W, et al. (2008) Rare independent mutations in renal salt handling genes contribute
to blood pressure variation. Nat Genet 40:592–599.
18. Lachheb S, et al. (2008) Kir4.1/Kir5.1 channel forms the major K? channel in the
19. Lourdel S, et al. (2002) An inward rectifier K(?) channel at the basolateral membrane
of the mouse distal convoluted tubule: similarities with Kir4-Kir5.1 heteromeric chan-
nels. J Physiol 538:391–404.
20. Djukic B, et al. (2007) Conditional knock-out of Kir4.1 leads to glial membrane
depolarization, inhibition of potassium and glutamate uptake, and enhanced short-
term synaptic potentiation. J Neurosci 27:11354–11365.
21. Kofuji P, et al. (2000) Genetic inactivation of an inwardly rectifying potassium channel
(Kir4.1 subunit) in mice: phenotypic impact in retina. J Neurosci 20:5733–5740.
22. Neusch C, et al. (2001) Kir4.1 potassium channel subunit is crucial for oligodendrocyte
development and in vivo myelination. J Neurosci 21:5429–5438.
23. Rozengurt N, et al. (2003) Time course of inner ear degeneration and deafness in mice
lacking the Kir4.1 potassium channel subunit. Hear Res 177:71–80.
24. Takumi T, et al. (1995) A novel ATP-dependent inward rectifier potassium channel
expressed predominantly in glial cells. J Biol Chem 270:16339–16346.
25. Kofuji P, Newman EA (2004) Potassium buffering in the central nervous system.
26. Buono RJ, et al. (2004) Association between variation in the human KCNJ10 potassium
ion channel gene and seizure susceptibility. Epilepsy Res 58:175–183.
27. Ando M, Takeuchi S (1999) Immunological identification of an inward rectifier K?
gerbils and rats. Cell Tissue Res 298:179–183.
28. Hibino H, et al. (1999) Expression of an inwardly rectifying K(?) channel, Kir4.1, in
satellite cells of rat cochlear ganglia. Am J Physiol 277:C638–C644.
29. Ito M, et al. (1996) Immunolocalization of an inwardly rectifying K? channel, K(AB)-2
(Kir4.1), in the basolateral membrane of renal distal tubular epithelia. FEBS Lett
30. Cruz DN, et al. (2001) Mutations in the Na-Cl cotransporter reduce blood pressure in
humans. Hypertension 37:1458–1464.
31. Nijenhuis T, et al. (2005) Enhanced passive Ca2? reabsorption and reduced Mg2?
channel abundance explains thiazide-induced hypocalciuria and hypomagnesemia.
J Clin Invest 115:1651–1658.
32. Schultheis PJ, et al. (1998) Phenotype resembling Gitelman’s syndrome in mice lacking
the apical Na?-Cl- cotransporter of the distal convoluted tubule. J Biol Chem
wasting with hyperkalaemic acidosis, pseudohypoaldosteronism type 1. Nat Genet
molecular anatomy. Physiol Rev 80:277–313.
35. Rundfeldt C, Netzer R (2000) The novel anticonvulsant retigabine activates M-currents
36. Halperin ML, Kamel KS (1998) Potassium. Lancet 352:135–140.
Scholl et al.PNAS ?
April 7, 2009 ?
vol. 106 ?
no. 14 ?