Download Expression Pattern in Brain of TASK-1, TASK

MCN
Molecular and Cellular Neuroscience 18, 632– 648 (2001)
doi:10.1006/mcne.2001.1045, available online at http://www.idealibrary.com on
Expression Pattern in Brain of TASK-1, TASK-3,
and a Tandem Pore Domain K ⴙ Channel Subunit,
TASK-5, Associated with the Central Auditory
Nervous System
Christine Karschin,* ,1 Erhard Wischmeyer,* ,1
Regina Preisig-Müller, †,1 Sindhu Rajan, † Christian Derst, †
Karl-Heinz Grzeschik, ‡ Jürgen Daut, † and Andreas Karschin* ,§,2
*Department of Molecular Neurobiology of Signal Transduction, Max-Planck-Institute for
Biophysical Chemistry, Göttingen; §Institute of Physiology, University of Würzburg,
Würzburg; and †Institute of Physiology and ‡Institute of Human Genetics,
Marburg University, Marburg, Germany
TWIK-related acid-sensitive K ⴙ (TASK) channels contribute to setting the resting potential of mammalian neurons
and have recently been defined as molecular targets for
extracellular protons and volatile anesthetics. We have
isolated a novel member of this subfamily, hTASK-5, from
a human genomic library and mapped it to chromosomal
region 20q12–20q13. hTASK-5 did not functionally express
in Xenopus oocytes, whereas chimeric TASK-5/TASK-3
constructs containing the region between M1 and M3 of
TASK-3 produced K ⴙ selective currents. To better correlate TASK subunits with native K ⴙ currents in neurons the
precise cellular distribution of all TASK family members
was elucidated in rat brain. A comprehensive in situ hybridization analysis revealed that both TASK-1 and
TASK-3 transcripts are most strongly expressed in many
neurons likely to be cholinergic, serotonergic, or noradrenergic. In contrast, TASK-5 expression is found in olfactory bulb mitral cells and Purkinje cells, but predominantly
associated with the central auditory pathway. Thus,
TASK-5 K ⴙ channels, possibly in conjunction with auxiliary proteins, may play a role in the transmission of temporal information in the auditory system.
INTRODUCTION
Our understanding of the molecular diversity and
function of K ⫹ channel subunits in which two-pore (2P)
1
These authors contributed equally to this work.
To whom correspondence and reprint requests should be addressed. Fax: (931) 31-2741. E-mail: [email protected].
2
632
domains are arranged in tandem is progressing at a
rapid pace. To date, 13 different members of the mammalian 2P K ⫹ channel family have been characterized
(genes KCNK1–17; reviewed by Lesage and Lazdunski,
2000; Goldstein et al., 2001). The subunits of 2P K ⫹
channels that have been successfully expressed in heterologous systems show a highly complex regulation
by physical and chemical stimuli, including membrane
stretch, heat, or changes in intracellular or extracellular
pH. Some of the 2P K ⫹ channels can also be modulated
by free fatty acids, volatile anesthetics, protein kinase A,
or protein kinase C. In voltage-clamp experiments, 2P
K ⫹ channels usually show extremely rapid activation
and no time-dependent inactivation (with the exception
of TWIK-2; Patel et al., 2000). With symmetrical K ⫹
concentrations, the steady-state current voltage relations were found to be (i) linear, as predicted by the
Goldman equation; (ii) inwardly rectifying; or (iii) outwardly rectifying (Lesage and Lazdunski, 2000).
Due to their sensitivity to extracellular pH, five channel species have been classified as acid-sensitive
(TASK-) or alkaline pH-activated (TALK-) channels.
Based on their primary sequence similarities TASK-1
and TASK-3 subunits (Duprat et al., 1997; Leonoudakis
et al., 1998; Rajan et al., 2000; Kim et al., 2000; Lopes et al.,
2000) may be grouped separately from the TASK-2,
TASK-4 (TALK-1), and TALK-2 subunits expressed primarily in peripheral tissues (Reyes et al., 1998; Decher et
al., 2001; Girard et al.; 2001). Acid-sensitive 2P K ⫹ channels probably play an important role both in the brain
1044-7431/01 $35.00
© 2001 Elsevier Science
All rights reserved.
633
CNS Expression of TASK Channel Subunits
(TASK-1 and TASK-3) and peripheral tissues. TASK-1,
for example (alternatively called cTBAK-1 (Kim et al.,
1998) or OAT1 (Lopes et al., 2000)), has been suggested
to represent the structural correlate of an oxygen- and
acid-sensitive background current that controls calcium
entry and neurosecretion in chemoreceptor cells of the
rat carotid body (Buckler et al., 2000). TASK-1 and also
TASK-4 are strongly expressed in cardiac atria and
ventricle (Kim et al., 1998, 1999; Lopes et al., 2000;
Decher et al., 2001), where they may regulate action
potential duration and myocardial contraction. Renal
TASK-2 channels appear to play a critical role in the K ⫹
transport of cortical distal tubules and collecting duct of
the kidney (Reyes et al., 1998). In the brain, TASK channels primarily determine the membrane potential, input
resistance and excitability of neurons, but may also
contribute to the production of cerebrospinal fluid (Kindler et al., 2000). In addition, central TASK channels
may be one of the most important targets of volatile
hypnotics and anesthetics (Patel et al., 1999; Gray et al.,
2000; Sirois et al., 2000a,b), and possibly of cannabinoids
(Maingret et al., 2001). In cerebellar granule cells (Millar
et al., 2000) and cranial motoneurons (Talley et al., 2000)
receptor-mediated inhibition of TASK-1 has been reported, which suggests an important role of 2P K ⫹
channels in neuromodulation.
In this report we describe the cloning, chromosomal
mapping, functional characterization, and tissue localization of TASK-5, a 2P K ⫹ channel subunit closely
related to TASK-1 and TASK-3, and we supply a comprehensive profile of the distribution pattern of these
three subfamily members in rat brain.
RESULTS
Molecular Cloning of the Human TASK-5 cDNA
We have identified a novel tandem pore domain K ⫹
channel, hTASK-5 (GenBank Accession Nos. AF294350
and AF294351, Fig. 1). The entire coding region and the
3⬘ nontranslated region of the cDNA were 1195 bp in
length and were obtained by combining the sequence
information of three EST clones and our isolates from a
human ␭-phage genomic library. The deduced open
FIG. 1. cDNA and amino acid sequence of human TASK-5. Transmembrane regions (black) and pore forming regions (underlined)
are highlighted. Putative PKA (a), PKC (1), and CK2 (⌬) phosphorylation sites are also indicated. Four DNA polymorphisms found are
shown in grey and the nucleotide sequence is given as IUCB abbreviation. Two TASK-5 cDNAs were submitted to GenBank under
Accession Nos. AF294350 and AF294351.
634
Karschin et al.
FIG. 2. Sequence alignment of TASK channels. The human sequences of TASK-1 (GenBank Accession No. AF006823), TASK-3 (AF212829), and
TASK-5 (AF294350) were aligned using the program CLUSTAL W. Transmembrane regions (black) and pore-forming regions (grey) are labeled.
Asteriscs denote residues conserved in all three subunits.
reading frame encodes a protein of 330 amino acids that
is identical to three recent sequences termed KT3.3/
TASK-5, published after completion of our study (Kim
and Gnatenco, 2001; Vega-Saenz De Miera et al., 2001;
Ashmole et al., 2001). The channel protein displays the
typical features of a 2P K ⫹ channel subunit, including
four transmembrane regions (M1–M4) flanking two P
regions (Figs. 1 and 2). In addition, the primary sequence revealed a large extracellular linker region between M1 and P1 that lacks both a cysteine residue for
disulfide bond formation and a glycosylation site. Compared to TASK-1 and TASK-3 subunits, the COOHterminus (85 amino acids) is short, but at the very end
it also harbors a conserved stretch of five amino acids
possibly involved in protein–protein interaction (Fig. 2).
At the amino acid level hTASK-5 is 58% and 56% identical to hTASK-1 and hTASK-3 (irrespective of four
polymorphisms in the coding regions of three different
EST cDNAs; see below), but only moderately similar
(18 –22% identity) to TASK-2 and TASK-4 subunits.
This sequence analysis suggests that only TASK-1,
TASK-3, and TASK-5 subunits form a structurally (and
possibly functionally) homogenous group of channels.
Analysis of the Human TASK-5 Gene KCNK15
Direct sequencing of the isolated ␭-DNA identified a
16.9-kb fragment containing the TASK-5 gene KCNK15
(Genbank Accession No. AF294352). Similar to the
genes for hTASK-1 (KCNK3) and hTASK-3 (KCNK9),
KCNK15 harbors a single intron that splits the coding
region in the selectivity filter of the first pore region.
The 3.9-kb-long intron is comparable in size to the
equivalent intron in hTASK-1 (⬃3 kb; Lopes et al., 2000),
but is significantly smaller than the intron of hTASK-3
(83.6 kb; Rajan et al., 2000). KCNK14 was assigned to
chromosomal region 20q12–20q13, with the nearest
flanking markers WI-6622 (centromeric) and D20S169
(telomeric; see Table 1).
Human TASK-5 Polymorphisms
Sequence analysis of three independent EST clones
and comparison to the human ␭-phage clones revealed
the existence of four putative polymorphisms in the
hTASK-5 coding region at codons 95 (GGG or GAG),
260 (CCC or ACC), 261 (CAC or CCC), and 323 (CTT or
CCT), respectively (Fig. 1). These polymorphisms give
rise to cDNAs encoding three different polypeptides,
TASK-5A (G95, P260, H261, P323), TASK-5B (E95, T260,
P261, L323), and TASK-5C (G95, P260, P261, P323). Of
particular interest, the nucleotide exchange at position
95 replaces the consensus glycine of the first selectivity
filter to a glutamate (GYG to EYG), which is expected to
disrupt channel function. When human genomic DNA
of 22 unrelated individuals was tested on this polymor-
635
CNS Expression of TASK Channel Subunits
TABLE 1
GeneBridge 4 Panel Radiation Hybrid Mapping Data
Gene/locus
Data vector a
KCNK14 (hTASK-5)
1000010000000011111
1000010100110000000
0011000000000011001
1100100101011001101
11000000011000001
a
Cytogenetic
map location
Genetic
map bin
Physical map
bin (GB4)
cR 3000
20 q12
—
20 q13
57.7 cM
—
61.0 cM
240.67 cR
—
251.26 cR
Flanking markers
(Whitehead
framework)
Flanking markers
(Whitehead
comprehensive
map)
8.56 cR
from WI-6622
—
5.2 cR
from D20S169
WI-4006
WI-6622
—
D20S169
A002H21
Homologous
mouse map
position
Chrom. 2
92.0 cM
—
94.0 cM
0, negative; 1, positive; 2, uncertain data.
phism, we found 6 individuals to be homozygous for
G95, 5 to be homozygous for E95, and 11 to be heterozygous. To clarify as to whether this variation was due to
a real polymorphism or possibly an RNA/DNA editing
mechanism, we also tested 24 individuals belonging to
6 families. From this population 4 individuals were
homozygous for G95, 7 were homozygous for E95, and
13 were heterozygous. All combinations found were in
agreement with a normal mendelian inheritance of this
polymorphism (P ⬍ 0.04).
Tissue Localization
In an RT-PCR assay of various rat tissues a 499-bp rat
TASK-5 PCR fragment (⬃90% identical to hTASK-5)
was amplified in significant amounts only from brain
(although faint bands were seen in liver and spleen).
Compared to rTASK-1 and rTASK-3 subunits, this is the
most restricted tissue expression. Whereas rTASK-1 is
expressed in all tissues tested, rTASK-3 was found in
brain, testis, and weakly in heart and stomach (Fig. 3).
FIG. 3. Tissue distribution of TASK-5. RT-PCR analysis of the expression of TASK-1, TASK-3, and TASK-5 channels in various rat
tissues as indicated. Intron-spanning primers were used for PCR and
amplified fragments were directly sequenced for confirmation.
In Situ Hybridizations of Rat Brain Sections
To specify in detail the CNS distribution of all TASK
subunits we have compiled comprehensive data on the
mRNA expression of TASK-1, TASK-3, and TASK-5 in
the adult rat brain (TASK-2, TASK-4/TALK-1, and
TALK-2 subunits are virtually absent from the brain).
Sagittal and coronal sections were hybridized with both
radiolabeled oligonucleotide probes (2–3 per subunit)
and digoxigenin-labeled cRNA probes. All different hybridization probes for a given subunit resulted in identical labeling profiles, thus ensuring specificities of the
probes and general hybridization conditions. Table 2
summarizes the cellular expression levels of all three
mRNA species in different regions of the rat brain.
TASK-1 and TASK-3. Detailed analysis revealed
that the widespread distribution patterns for TASK-1
and TASK-3 greatly resembled each other despite a
generally stronger and more punctate expression of
TASK-3 (Table 2 and Fig. 4) and mutually exclusive
expression in several selected nuclei (ruling out crosshybridization). Most remarkably, cholinergic structures
such as the cranial nerve motor nuclei (Figs. 4A, 4D, 5B,
and 5D), the dorsal motor nucleus of the vagus (Fig. 5D)
and spinal cord ventral horn, as well as the septal nuclei
were found to carry neuronal populations with excessive expression levels for both mRNA species. Only
TASK-3 signals were present in all cholinergic interneurons in the striatum (only ⬃2% of striatal cells; Figs. 4D
and 5C) and in tegmental nuclei of the cholinergic
system. In cranial nerve motor nuclei, the spinal cord
and striatum the positive neurons were clearly identified as being cholinergic, due to the large size of their
somata (⬎20 ␮m; Fig. 5C). In all other known cholinergic nuclei (Table 2), e.g., in the diagonal band nuclei,
septum, the parasympathetic dorsal motor nucleus of
the vagus nerve, and the tegmental nuclei, we also
636
Karschin et al.
TABLE 2
Differential Distribution of TASK-1, TASK-3, and TASK-5 Subunit mRNAs in the Adult Rat Brain
Brain region
Olfactory system
Olfactory bulb mitral cells
Granule cell layer
External plexiform layer
Periglomerular cell layer
Anterior olfactory nucleus
Olfactory tubercle*
Islands of Calleja*
Piriform cortex
Neocortex
Neocortical layers II-VI
Subiculum
Entorhinal cortex
Dorsal endopiriform nucleus
Hippocampus
Dentate gyrus granule cells
CA1 pyramidal cells
CA3 pyramidal cells
Basal nuclei
Caudate putamen*
Nucleus accumbens*
Globus pallidus*
Basal nucleus of Meynert*
Subthalamic nucleus
Septum
Bed nuclei of stria terminalis
Lateral septal nucleus
Medial septal nucleus*
Horizontal nucleus diagonal band*
Vertical nucleus diagonal band*
Amygdala
Cortical amygdala, posteromedial n.
Cortical amygdala, anterior nucleus
Central amygdaloid nuclei
Basomedial amygdaloid nucleus
Medial amygdaloid nuclei, dorsal n.
Hypothalamus
Magnocellular preoptic nucleus*
Medial preoptic nucleus
Suprachiasmatic nucleus
Lateral olfactory tract nucleus
Arcuate hypothalamic nucleus
Paraventricular hypothalamic n.
Dorsomedial hypothalamic nucleus
Ventromedial hypothalamic nucleus
Premammillary nucleus, dorsal
Medial mammillary nuclei, lateral
Medial mammillary nuclei, medial
Lateral mammillary nuclei
Zona incerta, dorsal
Habenula
Thalamus
Reticular nucleus
Anterodorsal nucleus
Anteroventral nuclei
Laterodorsal nuclei
Paraventricular thal. nucleus
TASK-1
TASK-3
TASK-5
⫹⫹
⫹⫹
0
⫹
⫹/⫺
⫹/⫺
⫹⫹⫹
⫹
⫹⫹⫹
⫹⫹⫹
⫹
⫹⫹⫹
⫹⫹⫹
Ch
⫹
⫹⫹⫹
⫹⫹⫹⫹
0
⫹⫹⫹
0
0
0
0
0
⫹, Few ⫹⫹⫹
⫹, Few ⫹⫹⫹
⫹
0
⫹⫹
⫹
⫹
⫹⫹⫹
0
0
0
0
⫹/⫺
⫹/⫺
⫹/⫺
⫹
⫹
⫹/⫺
0
0
0
0
0
0
0
⫹⫹
Ch
Ch
Ch
Ch
⫹⫹⫹
0
0
0
0
0
⫹
0
⫹⫹⫹
⫹⫹⫹
⫹⫹⫹
⫹⫹⫹
⫹⫹⫹
Few ⫹⫹⫹
Few ⫹⫹⫹
Few ⫹⫹⫹
0
0
0
0
0
⫹/⫺
⫹/⫺
⫹/⫺
⫹
⫹
⫹/⫺
⫹⫹⫹
⫹/⫺
⫹⫹⫹
⫹⫹⫹⫹
0
0
0
0
0
⫹⫹
⫹/⫺
0
⫹
⫹⫹
⫹/⫺
⫹/⫺
⫹⫹
⫹⫹
⫹
⫹⫹
⫹⫹
⫹
0
⫹⫹⫹
⫹⫹⫹
⫹⫹⫹
⫹⫹⫹⫹
⫹⫹⫹
⫹⫹⫹
⫹⫹⫹
⫹⫹
⫹⫹⫹⫹
⫹⫹⫹
⫹⫹⫹
⫹⫹⫹
⫹⫹⫹
Few ⫹⫹
0
0
0
0
0
0
0
0
0
⫹
0
0
0
0
⫹⫹⫹
0
⫹/⫺
0
⫹/⫺
⫹
⫹⫹⫹⫹
⫹⫹⫹
⫹⫹⫹
⫹/⫺
0
0
0
0
0
637
CNS Expression of TASK Channel Subunits
TABLE 2—Continued
Brain region
Thalamus
Parafascicular nucleus
Central medial, lateral thal. nuclei
Paracentral thal. nucleus
Ventral nuclei
Ventroposterior nuclei
Dorsal lateral geniculate nucleus
Ventral lateral geniculate nucleus
Medial geniculate nucleus
Midbrain and brainstem
Superior colliculus
Anterior pretectal nucleus
Precommissural nucleus
Central gray, dorsal nucleus
Red nucleus
Substantia nigra, pars compacta
Substantia nigra, pars reticulata
Ventral tegmental area
Interpeduncular nuclei
Pedunculopontine tegmental n.*
Laterodorsal tegmental nucleus*
Parabigeminal nucleus*
Mesencephalic trigeminal nucleus
Raphe nuclei
Reticulotegmental nucleus
Pontine nucleus
Locus coeruleus
Dorsal tegmental central nucleus
Parabrachial nuclei
Vestibular nuclei, medial
Vestibular nuclei, spinal and lateral
Cuneate nucleus
External cuneate nucleus
Gracile nucleus
Solitary nucleus
Prepositus hypoglossus n.*
Inferior olivary nuclei
Spinal trigeminal nuclei
Principal sensory trigeminal n.
Lateral reticular nucleus
Gigantocellular reticular nuclei
Medullary reticular nuclei
Cerebellum
Deep nuclei
Molecular layer
Granule cell layer
Purkinje cells
Auditory nuclei
Cochlear nuclei, dorsal
Cochlear nuclei, ventral posterior
Cochlear nuclei, ventral anterior
Lateral superior olive
Medial superior olive
Superior paraolivary nuclei
Medioventral periolivary nuclei
Nucleus of the trapezoid body
Lateral lemniscus, dorsal nuclei
Lateral lemniscus, intermediate n.
TASK-1
TASK-3
TASK-5
⫹⫹
⫹⫹⫹
⫹⫹⫹
⫹⫹
⫹/⫺
⫹/⫺
⫹
⫹⫹
⫹/⫺
⫹
⫹⫹
⫹⫹⫹
⫹⫹
⫹⫹
⫹⫹⫹
⫹⫹
0
0
0
0
0
0
0
Few ⫹⫹⫹
⫹
⫹
⫹
⫹
⫹/⫺
⫹
⫹
⫹/⫺
⫹⫹⫹
0
⫹/⫺
0
⫹⫹
⫹/⫺
⫹⫹
⫹⫹
⫹⫹⫹
⫹
⫹/⫺
⫹
⫹/⫺
⫹/⫺
⫹
⫹
⫹
⫹
⫹/⫺
⫹
⫹/⫺
⫹⫹
⫹⫹
⫹⫹
⫹
⫹⫹
⫹⫹⫹
⫹⫹⫹
⫹⫹
Few ⫹⫹
0
Few ⫹⫹
⫹⫹⫹
⫹⫹⫹
⫹⫹⫹
⫹⫹
0
⫹⫹⫹⫹
⫹⫹
⫹⫹
⫹⫹⫹⫹
⫹⫹⫹
⫹
⫹
Few ⫹⫹
⫹⫹
0
⫹⫹⫹
⫹⫹⫹
Few ⫹⫹⫹
0
⫹⫹
⫹⫹
⫹⫹⫹
⫹⫹⫹
⫹⫹⫹
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Few ⫹⫹⫹
0
0
0
0
⫹/⫺
0
⫹⫹
0
⫹
0
⫹⫹
0
0
0
0
⫹⫹⫹⫹
⫹/⫺
⫹
⫹
0
0
0
⫹
⫹⫹
⫹/⫺
⫹/⫺
⫹⫹
⫹
⫹⫹
⫹/⫺
⫹/⫺
⫹/⫺
⫹⫹⫹
0
⫹⫹
⫹⫹
⫹⫹⫹
⫹⫹
⫹⫹
⫹⫹⫹⫹
⫹⫹⫹⫹
⫹⫹
⫹/⫺
0
⫹⫹⫹
⫹⫹⫹⫹
638
Karschin et al.
TABLE 2—Continued
Brain region
TASK-1
TASK-3
TASK-5
Lateral lemniscus, ventral nuclei
Inferior colliculus, central nucleus
Inferior colliculus, dorsal cortex
Thalamus, medial geniculate nucleus
Temporal (auditory) cortex
Cholinergic motor nuclei of cranial nerves
3 Oculomotor nucleus*
4 Trochlear nucleus*
5 Motor trigeminal nucleus*
6 Abducens nucleus*
7 Facial nucleus*
9/10 Ambiguus nucleus*
11 Accessory nucleus*
12 Hypoglossus nucleus*
10 Parasympathetic dorsal n. vagus*
Spinal cord motoneurons*
⫹/⫺
⫹/⫺
⫹/⫺
⫹⫹
⫹
⫹⫹
⫹
⫹⫹
⫹⫹
⫹⫹
⫹⫹⫹
⫹⫹⫹⫹
0
Few ⫹⫹⫹
0
⫹⫹⫹⫹
⫹⫹⫹⫹
⫹⫹⫹⫹
⫹⫹⫹⫹
⫹⫹⫹⫹
⫹⫹⫹⫹
⫹⫹⫹⫹
⫹⫹⫹⫹
⫹⫹⫹
⫹⫹⫹⫹
⫹⫹⫹⫹
⫹⫹⫹⫹
⫹⫹⫹⫹
⫹⫹⫹⫹
⫹⫹⫹⫹
⫹⫹⫹⫹
⫹⫹⫹⫹
⫹⫹⫹⫹
⫹⫹⫹
⫹⫹⫹⫹
0
0
0
0
0
0
0
0
0
0
Note. In situ hybridization signals were rated as ⫹⫹⫹⫹, very abundant; ⫹⫹⫹, abundant; ⫹⫹, moderate; ⫹, low; ⫹/⫺, just above background;
0, no expression; few, few positive cells present; *this nucleus contains cholinergic cells; Ch, only cholinergic neurons are positive; n, nucleus.
detected a population of strongly labeled cells. These
neurons most likely represent the cholinergic subset,
but simultaneous cholinacetyltransferase (ChAT)-immunoreactivity or -mRNA hybridization signals (Oh et
al., 1992) will have to provide conclusive evidence.
Apart from the cholinergic system, high expression
levels were found in the locus coeruleus (both TASK-1
and TASK-3), the major source of noradrenergic neurons (Figs. 4C, 4F, and 5A), and in all raphé nuclei (only
TASK-3), which contain the majority of all serotonergic
neurons (Fig. 5B). Also, expression was found in diverse brain nuclei or cell populations (e.g., cerebellar
and olfactory bulb granule cells and scattered neurons
in reticular areas) not related to a specific neurotransmitter system. Mainly, TASK-3 expression was prominent in distinct nuclei in the amygdala and hypothalamus, and in thalamic nuclei we observed a differential
expression of both mRNAs (Table 2). No evidence was
found for expression in glial structures.
TASK-5. A particularly unique expression pattern
was observed for TASK-5, which is almost exclusively
expressed in the nuclei of the primary central auditory
pathway (Figs. 4G, 4J, and 6). Starting most distally in
the medulla, the dorsal and ventral cochlear nuclei,
which receive input from the spiral ganglia, were found
strongly positive. Next, high expression levels were
detected in the superior olivary complex of the pons
where binaural interactions first occur. Interestingly,
the medial and lateral superior olive subnuclei within
the superior olivary complex were strongly labeled, but
the nucleus of the trapezoid body was negative. Further
downstream the ventral, dorsal and intermediate sub-
nuclei of the lateral lemniscus were strongly labeled.
The inferior colliculus, the next relay station in the
midbrain receiving input via the lateral leminiscus,
showed strong levels only in its central nucleus. In the
thalamic projection area, the medial geniculate nucleus,
only few positive cells were found in the medial subnucleus, but not in the dorsal and ventral subnuclei. Of
special note, only specific neuronal subsets in these
structures were strongly labeled, whereas other cells
were only slightly labeled or not at all. Finally, TASK-5
signals were absent from primary auditory cortex regions.
Outside the auditory system, TASK-5 was only expressed in a few selected groups of neurons in the
spinal trigeminal sensory and mammillary nucleus, the
olfactory bulb, and the cerebellum (Fig. 4G). Interestingly, sagittal sections from the cerebellum show that
Purkinje cells are only positive in the posterior lobe, but
not in the anterior lobe (lobules 1–5). Also a high density of positive Purkinje cells were seen in the flocculi
and in the vermis, but very few positive cells in more
lateral regions (ansiform lobules). Within the vermis,
TASK-5-positive Purkinje cells interspersed with unlabeled cells are arranged in columns, as has been found
for a series of enzymatic markers, e.g., zebrin I and II
(Fig. 7). In the olfactory bulb, TASK-5 is present in both
output-cell populations, the mitral cells and the tufted
cells in the external plexiform layer close to the glomeruli (Fig. 4G, inset). This may be relevant for a very
specific sensory function since both cell types release
the neuropeptide transmitter corticotropin releasing
CNS Expression of TASK Channel Subunits
639
FIG. 4. Localization of TASK-1, TASK-3, and TASK-5 mRNA in the rat brain. In situ hybridization patterns were obtained with 33P-labeled
oligonucleotide probes specific for TASK-1 (A–C), TASK-3 (D–F), and TASK-5 (G–J) mRNAs. X-ray film images of sagittal and adjacent coronal
rat brain sections demonstrate that TASK-1, TASK-3, and TASK-5 transcripts are differentially distributed. Insets in A, D, G show adjacent
coronal sections through the olfactory bulb. Abbreviations: 7, facial nucleus; B, basal nucleus of Meynert; CPu, caudate putamen; Gr, cerebellar
granule cell layer; LC, locus coeruleus; LL, nuclei of the lateral lemniscus; MCPO, magnocellular preoptic nucleus; MGN, medial geniculate
nucleus; Mo5, motor trigeminal nucleus; P, cerebellar Purkinje cell layer; PMD, premammillary nucleus, dorsal; Rt, reticular thalamic nucleus;
SOC, superior olivary complex; Sp5, spinal trigeminal nucleus; STh, subthalamic nucleus; Tz, trapezoid body; VCA, ventral cochlear nucleus,
anterior; VLG, ventral lateral geniculate nucleus. Scale bars, 4 mm. Exposure time is 18 –21 days.
factor (CRF) from their terminals in the molecular layer
of the piriform cortex.
Functional Expression in Xenopus Oocytes
When hTASK-5 cRNA was injected in Xenopus oocytes, it did not induce any measurable K ⫹ current (n ⫽
20). As a first approach to decipher the cause of this
lack of functional expression, hTASK-5 and other TASK
subunits were tagged with enhanced green fluorescent
protein (EGFP) at the NH 2-terminus to trace their subcellular localization by confocal microscopy. When inspected 48 h after cRNA injection, strong membrane
fluorescence was observed for TASK-3 EGFP, but not for
640
FIG. 5. TASK-3 mRNA expression in noradrenergic (A), serotonergic (B), and cholinergic rat brain nuclei (B, C, D) as visualized by
digoxigenin-labeled cRNA probes. Shown are heavily labeled noradrenergic neurons in the locus coeruleus (A), serotonergic neurons in
the dorsal raphé nucleus (B), and cholinergic neurons in the trochlear
nucleus (B), in the caudate putamen (C), in the hypoglossus nucleus
and in the dorsal nucleus of the vagus (D). Abbreviations: 4, trochlear
motor nucleus; 10, dorsal motor nucleus of the vagus; 12, hypoglossus
nucleus; DR, dorsal raphé. Scale bar, 200 ␮m (A) and 400 ␮m (B–D).
hTASK-5 EGFP subunits, indicating inappropriate targeting to the plasma membrane (Fig. 8A). We next addressed the question whether membrane targeting of
hTASK-5 subunits could be rescued by the presence of
another TASK subunit. However, when hTASK-5 EGFP
Karschin et al.
was coinjected with wild-type TASK-3, no surface
membrane fluorescence signal could be detected, suggesting that the two subunits did not functionally interact in the membrane transport of hTASK-5 (Fig. 8A).
This view was supported by two-electrode voltage
clamp measurements of oocytes expressing TASK-3
(20.1 ⫾ 7.6 ␮A; n ⫽ 4; at ⫹30 mV and 2 mM [K ⫹] e), or
TASK-1 (1.89 ⫾ 1.1 ␮A, n ⫽ 4). For both subunits the
presence of coinjected hTASK-5 had no significant effect on current amplitude (TASK-3/5: 17.6 ⫾ 2.6 ␮A;
TASK-1/5; 1.90 ⫾ 1.1 ␮A) or their pH sensitivity, as
illustrated in Fig. 8B.
The apparent inability of TASK-5 to form heteromers
with the closely related channels TASK-1 and TASK-3
was not unexpected since there was no overlap in their
distribution in the brain. In a subsequent series of experiments we tested whether TASK-1 and TASK-3,
which are coexpressed in many regions of the brain, can
form heteromers. This was much more difficult to ascertain because both subunits give similar currents
when expressed individually. To investigate this matter
we used dominant-negative constructs of TASK subunits. A TASK-3 G95E mutant was engineered, which is
equivalent to the hTASK-5 polymorphism in which a
consensus glycine (G95) in the first selectivity filter is
replaced by a glutamate (E95), and was found to be
nonfunctional (data not shown). When wild-type
TASK-3 subunits (22.2 ⫾ 7.3 ␮A at ⫹30 mV) were then
coinjected with mutant TASK-3 G95E subunits, macro-
FIG. 6. TASK-5 expression in nuclei of the primary central auditory pathway. The arrangement shows coronal rat brain sections at the level
of the dorsal cochlear nuclei (DC), superior olivary nuclei (SOC) and ventral cochlear nuclei (VCA), lateral lemniscus (LL), and inferior colliculus
(IC). Not shown is the thalamic medial geniculate nucleus. Signals were detected with 33P-labeled oligonucleotide probes specific for TASK-5 and
graphically transposed into digitized images of cresyl-violet stained sections. Scale bar, 2 mm.
641
CNS Expression of TASK Channel Subunits
FIG. 7. TASK-5 mRNA expression in rat cerebellar Purkinje cells as
visualized by digoxigenin-labeled cRNA probes. Note that most Purkinje cell bodies are strongly labeled in cerebellar lobule 9 (A),
whereas prominent vertically aligned stripes devoid of positive cells
occur in more rostral regions, e.g., lobule 8 (B), and only very few cells
remain positive, e.g., in lobule 4 (C). Scale bars, 400 ␮m.
scopic current amplitudes were reduced to 6.4 ⫾ 0.53
␮A; n ⫽ 7 (Fig. 8C), i.e., to about 30% of control.
However, when TASK-1 subunits (1.28 ⫾ 0.41 ␮A) were
coexpressed with TASK-3 G95E, macroscopic currents
were statistically unaltered (1.53 ⫾ 0.67 ␮A), suggesting
that no heteromeric channels between the two subunits
had been formed (Fig. 8C).
To determine the possible structural constraints that
prevent hTASK-5 channel expression we constructed
hybrid channels between (functional) TASK-3 and
(nonfunctional) TASK-5 subunits. The results of this
analysis are shown in the voltage-ramp experiments
and bar graphs of Fig. 9, in which the cartoons depict
TASK-3 components in black and TASK-5 components
in white. When both subunits were initially split in half,
and proximal (NH 2-M1-P1-M2) and distal parts (M3P3-M4-COOH) were exchanged, we recognized that
both the proximal and (to a lesser extent) the distal half
of hTASK-5 impaired channel function (Figs. 9A and
9B). Analysis of a further series of chimeric constructs
showed that functional channels could still be formed
when the major part of TASK-3 was replaced by
hTASK-5, although macroscopic currents decreased
with increasing portions of hTASK-5 in the chimera. As
a minimal requirement for expression of functional
hTASK-5/TASK-3 hybrids the region in between M1
and M3 of TASK-3 had to be included into hTASK-5
(Fig. 9A, right panel and Fig. 9B, hybrid 6). Conversely,
inserting the corresponding region of hTASK-5 into
TASK-3 (hybrid 7) reduced the wild-type current to
even lower levels. Interestingly, the chimera containing
only the M1–P1 loop of hTASK-5 (hybrid 4) greatly
reduced the current level, and the chimera containing
only M2 and the M2-M3 linker of TASK-3 (hybrid 11)
was not sufficient to rescue channel function. One interpretation of these findings is that within the context
of the native protein both intra- and extracellular parts
between M1 and M3 (and their possible interactions
with other channel domains) may play an important
role in the assembly and targeting of the channel.
There are several residues in the M2-M3 linker that
differ between TASK-3 and TASK-5. We speculated that
perhaps one of the residues not conserved between
TASK-3 and TASK-5 may be particularly important in
preventing correct targeting of TASK-5 in oocytes.
Therefore we generated a number of TASK-5 mutants
in which one or more of the nonconserved residues
were changed back to TASK-1, for example, R138Y,
W151R, C153D, LAAKC141-145HRAKK. However,
none of these mutations could rescue channel function.
DISCUSSION
Localization of TASK Subtypes and Implications
for Channel Regulation
In search of the ion conductances that contribute to
the resting potential of mammalian neurons, the
TASK-1 channels have recently become the focus of
642
Karschin et al.
FIG. 8. Membrane targeting and coexpression of TASK subunits in the oocyte membrane. (A) Confocal micrographs of fluorescence signals in
the oocyte membrane 48 h after cRNA injection of EGFP-labeled TASK-3 EGFP (left panel), TASK-5 EGFP (middle panel), and TASK-5 EGFP subunits
plus unlabeled TASK-3 (right panel). (B) Whole-cell currents in response to voltage ramps between ⫺150 to ⫹60 mV recorded from Xenopus
oocytes 48 h after injection with individual hTASK-5 cRNA or combinations with TASK-1 and TASK-3 as indicated. (C) Macroscopic ramp
currents of oocytes injected with TASK-1 and TASK-3, respectively, in the absence or presence of the dominant negative subunit TASK-3 G95E.
Extracellular [K ⫹] concentration in all experiments was 2 mM.
attention (Brown, 2000; North, 2000). Correlating the
properties of native neuronal K ⫹ channels with those of
recombinant 2P K ⫹ channels is difficult because (1)
some of the cloned channels are qualitatively similar in
heterologous expression systems, and (2) their properties may be altered by auxiliary proteins expressed in
specific neurons. It is therefore crucially important to
precisely determine the cellular expression of TASK-1
and its known homologues in the brain. The mRNA
patterns presented here for TASK-1, TASK-3, and the
novel TASK-5 subunits have been obtained with several
complementary techniques and different hybridization
probes which all gave consistent results. The strong
expression of TASK-1 transcripts observed in cranial
motoneurons is in good agreement with recent immu-
nocytochemical and electrophysiological studies (Duprat et al., 1997; Talley et al., 2000; Sirois et al., 2000a,b).
We found TASK-1 signals only in neurons, but not in
fiber pathways and non-neuronal ventricular cells, suggesting a negligible role in glial function and in the
production of cerebrospinal fluid (see Kindler et al.,
2000). Our results on the TASK-3 expression also agree
in most points with the recent findings of Vega-Saenz
De Miera (2001) on KT3.2. An RT-PCR analysis of various human brain regions (Chapman et al., 2000; Medhurst et al., 2001) identified substantial levels of TASK-3
transcripts only in the cerebellum, but not in any other
brain region, which is at variance with the findings
reported here for rat brain.
The observed differential distribution of the three
CNS Expression of TASK Channel Subunits
643
FIG. 9. Heterologous expression of TASK-5/TASK-3 chimera in Xenopus oocytes. (A) Macroscopic current responses of Xenopus oocytes in
response to voltage ramps between ⫺150 mV and ⫹60 mV. Oocytes were injected with cRNAs encoding hybrid TASK-3/TASK-5 constructs, as
depicted in the cartoons (TASK-3 in black, TASK-5 in grey). (B) Bar graph summarizing macroscopic current amplitudes for the chimera
indicated measured at a holding potential of ⫹30 mV and 2 mM external K ⫹.
TASK channel subunits present in the brain is particularly relevant in light of three major aspects of channel
regulation.
(i) pH regulation. TASK subunits are inhibited by a
small decline in external pH around pKs of ⬃8.5
(TASK-2; Reyes et al., 1998), 7.2–7.4 (TASK-1; Duprat et
al., 1997) and 5.9 – 6.7 (TASK-3; Rajan et al., 2000; Kim et
al., 2000), respectively. TASK channels are thus sensitive to a wide range of physiological and pathophysiological pH changes which may occur following neuronal activity and seizures (Kaila and Ransom, 1998),
ischemia, and other metabolic states, and these changes
will be reflected in the input/output characteristics of
neurons expressing one of the TASK channels. The pH
sensitivity of TASK-1 and TASK-3 subunits (but not
TASK-2 and TASK-4) is mediated by a titratable histidine residue in the outer part of the first pore region
(Rajan et al., 2000; Kim et al., 2000; Lopes et al., 2001). A
histidine residue is also present at the corresponding
position of TASK-5, which in vivo, possibly in conjunc-
tion with an as yet unknown accessory protein, may
generate pH-sensitive conductances.
(ii) Volatile anesthetics. TASK-1 (Patel et al., 1999;
Sirois et al., 2000a); TASK-2 (Gray et al., 2000), and
TASK-3 (Meadows and Randall, 2001) are also targets
for a number of volatile anesthetics. They are opened by
halothane concentrations used in general human anesthesia and may contribute to maintaining the anesthetized state by holding the membrane potential at values
near E k. It has been proposed that the sensitivity to
volatile anesthetics is mediated by a conserved binding
site at the channels’ C-termini (VLRFMT; Patel et al.,
1999). A variant of this binding site (VLRFMV) is also
present in TASK-5, which may therefore also be modulated by anesthetics.
(iii) Modulation by neurotransmitters. Recent reports indicate that 2P K ⫹ channels (probably TASK-1)
give rise to a prominent noninactivating K ⫹ conductance in cranial motoneurons (Talley et al., 2000) and
cerebellar granule cells (Millar et al., 2000) and that this
644
conductance is modulated by neurotransmitters. The
signaling pathway between receptors and TASK channels is not yet clear, but is is likely that receptors of the
G q/11 family, e.g., muscarinic M3 receptors or thyrotropin releasing hormone type 1 receptors, are involved. Closure of K ⫹ channels by neurotransmitter
stimulation is expected to depolarize neurons, which
has pronounced effects on the integration of synaptic
potentials and the resulting firing pattern. Analyzing
the Ba 2⫹-, pH-, and voltage-dependency of the neurotransmitter-modulated resting current in hypoglossal
motoneurons, Talley et al., (2000) argue that neither
Kir2.2 nor Kir2.4 inwardly rectifying K ⫹ channels
(which are strongly expressed in all cranial nerve nuclei; Töpert et al., 1998), contribute to the underlying
conductance change. Our study shows, however, that
TASK-3 channels are expressed in all cholinergic motoneurons, and that TASK-3 is expressed more strongly
than TASK-1. Furthermore, recombinant expression of
TASK-3 channels indicated that they can also be inhibited by neurotransmitters (E. Wischmeyer, unpublished
observations). Thus, the abundantly expressed TASK-3
channels, as well as TASK-5 channels, which are restricted to the central auditory system, are likely to
represent two further targets for G protein-mediated
signaling pathways.
Structural Determinants for the Lack of Functional
Expression of TASK-5
Although the overall structure of TASK-5 is similar to
that of other known 2P K ⫹ channels, it could be functionally expressed only when combined with some domains of TASK-3. The reason for the failure of functional expression is not yet clear, but our measurements
with EGFP-tagged TASK-5 suggest that the targeting to
the plasmalemma is impaired in Xenopus oocytes (see
Ashmole et al., 2001). From our mutational analysis it
may be concluded that the structural constraints that
prevent functional expression reside primarily in the
region between M1 and M3. One possible explanation is
that an intracellular ␤-subunit or auxiliary protein that
is present in neuronal cells and absent in Xenopus oocytes needs to bind to the M2–M3 linker to enable
correct targeting of TASK-5 to the plasmalemma. Exchange of single amino acid residues in the intracellular
linker of TASK-5 to the corresponding residues of
TASK-3 did not rescue functional expression. These
negative results support the idea that steric interactions
of the M2–M3 loop with other domains of the channel
or with an intracellular ligand, are necessary to ensure
correct targeting and assembly of TASK-5 channels.
Karschin et al.
Quite uncommon for ion channel genes, four sequence polymorphisms were identified within the coding region of hTASK-5. Most significantly, introduction
of the G95E polymorphism into the first pore region of
TASK-3 completely abolished channel expression.
However, it cannot be excluded that the G95E mutation
in TASK-5, together with accessory subunits expressed
in neuronal cells, in fact forms a functional channel in
vivo. Interestingly, another “silent” human 2P K ⫹ channel subunit, KCNK7, exhibits an amino acid exchange
from glycine to glutamate in the second pore region
(GLE). In addition, the Caenorhabditis elegans 2P K ⫹
channel F46A9.3 harbors an EYG motif in the second
pore. The physiological relevance of the glycine to glutamate mutation is still unclear. It should be noted that
TASK-5 is unlikely to represent a pseudogene, because
(1) cDNA was found in significant amounts, (2) no stop
codons are in the coding region, and (3) exon-intron
splice sites are highly conserved.
Possible Function of TASK-5 Channels
In rat brain, TASK-5 is predominantly expressed in
neurons of the central auditory nervous system. The
cause for this cell-specific expression remains speculative. Has it simply accompanied the phylogenetic refinement of the auditory system or has it evolved to
serve specific functions in the precise transmission of
brief signals across central synapses (Oertel, 1997;
Trussell, 1999)? It is well known that voltage-activated
K ⫹ channels are crucial for shaping the onset, duration
and frequency of action potentials, and their differential
expression may be important for central auditory neurons to transmit temporal information faithfully (Grigg
et al., 2000). One may imagine what particular function
time-independent or very rapidly activating (Rajan et
al., 2000, 2001) K ⫹ currents may contribute to the precise
timing of phase locked action potentials. A substantial
“baseline” TASK conductance would reduce the firing
rate elicited in the postsynaptic cell by a given synaptic
input and would shorten action potentials. The decrease in the sensitivity to synaptic stimulation may be
useful in minimizing the noise produced by jittery,
subthreshold converging inputs (Oertel, 1997).
TASK-5 is the second ion channel gene associated
with the auditory system pathway. Recently, the novel
KCNQ4 subunit that has been implicated in the autosomal dominant deafness DFNA2 (Kubisch et al., 1999;
Holt and Corey, 1999; Trussell, 1999) has been found in
the outer hair cells of the inner ear, but also in the
cochlear nuclei, the nuclei of the lateral leminiscus and
the inferior colliculus (Kharkovets et al., 2000). By anal-
645
CNS Expression of TASK Channel Subunits
ogy to the deafness caused by mutations of KCNQ4,
TASK-5 dysfunctions may also be associated in a loss of
hearing, provided that in vivo TASK-5 channels are
indeed functional and expressed at the levels predicted
by the amount of mRNA transcripts present. Incidentally, the observed polymorphism in the GYG motif of
the first pore region of TASK-5 is exactly at the same
position as the G285S mutation that disrupts channel
activity in KCNQ4 leading to DFNA2, and a similar
mutation in KvLQT1/KCNQ1 causes dominant long
QT syndrome (Wollnik et al., 1997). Hearing disorders
have not been reported by the individuals tested in the
genetic analysis performed here, but more subtle deficits may have remained unrecognized. Future studies
of the functional significance of TASK-5 in transgenic
animals and during development are required to clarify
its potential role in the processing of auditory information.
5⬘-agggacacttgggttgtctg-3⬘; reverse: 5⬘-gaggaagttcggcttctcg-3⬘). Genomic DNA of 22 unrelated and 24 related
individuals (six families; kindly provided by Dr. M.
Hess, Dept. of Pediatrics, Marburg, Germany) was used
as template for amplification. The PCR was conducted
for 35 cycles at 94°C for 30 s (denaturation), 52°C for
30 s (annealing), and 72°C for 1 min (elongation), using
AmpliTaq Gold DNA-polymerase (Applied Biosystems).
The entire hTASK-5 DNA was subcloned both into
the mammalian expression vector pcDNA3.1 (Invitrogen) and the oocyte expression vector pSGEM (kindly
provided by Dr. M. Hollmann, Bochum, Germany). For
site-directed mutagenesis the QuickChange mutagenesis system (Stratagene) was used as described by the
manufacturer. To create chimeras between hTASK-3
and hTASK-5 SacI sites were introduced by site-directed mutagenesis at amino acid position 157/158 and
BamHI sites at position 82/83 of hTASK-3 and
hTASK-5, respectively.
EXPERIMENTAL METHODS
Identification of TASK-5
The BLAST algorithm (Altschul et al., 1997) was used
to search for novel sequences homologous to TASK-1/
TASK-3 within the expressed sequence tag database
(dbEST) and the human genomic database (htgs). As a
result three nearly identical EST-clones (Accession Nos.
A1968607, A1739096 and AA283204) with significant
homology to TASK-1/3 were identified from brain, colon and ovary tumor, respectively. The corresponding
IMAGE-EST clones were purchased from the Resource
Center of the German Human Genome Project, MaxPlanck-Institute for Molecular Genetics Berlin, and
completely sequenced.
A fragment from the EST clone A1968607 was labeled
with digoxigenin and hybridized to ⬃500.000 plaque
forming units of a human ␭ DASH II genomic library
(Stratagene). Screening was performed using standard
hybridization procedures (50% formamide) at reduced
stringency (35°C hybridization, 60°C wash) and chemoluminescense detection with CSPD as substrate
(Roche Diagnostics, Mannheim, Germany). Two further
screenings were performed to purify isolated plaques.
␭-DNA was prepared using the ␭-Midi Prep kit (Qiagen, Hilden, Germany) and directly sequenced.
To test for a putative polymorphism concerning the
first selectivity filter in hTASK-5, PCR primers were
designed that amplified a 239 bp DNA fragment containing the second splice site and the coding region of
the first pore region of the hTASK-5 gene (forward:
Chromosomal Assignment of the TASK-5 Gene
(KCNK15)
Physical map positions were determined by segregation analysis of gene markers using gene-specific amplification of cloned DNAs from the GeneBridge 4 human radiation hybrid panel (Gyapay et al., 1996). The
following primers were used for amplifications: 5⬘-cagtcagcatgtcctctcca-3⬘ and 5⬘-agcaactgctcctctttcca-3⬘. PCR
conditions were as follows: denaturation for 30 s at
94°C, annealing for 30 s at 55°C, elongation for 1 min at
72°C, each for 30 cycles. Data vectors (Table 1) based on
two independent PCR analyses of the entire panel, with
data arranged in the order specified for the Whitehead/
MIT on-line Radiation Hybrid Mapper Program, were
submitted to two-point maximum-likelihood analysis.
Tissue Localization and Functional Analysis
RNA from various rat tissues was isolated using the
TRIZOL method (Gibco BRL) and 3 ␮g each were reverse transcribed with Superscript II reverse transcriptase (Gibco). For rTASK-1 and rTASK-3 specific intronspanning primers were used to PCR amplify a 253 bp
(rTASK-1), and 327 bp (rTASK-3) DNA fragment, respectively, with AmpliTaq Gold DNA polymerase. For
amplifying rTASK-5 an intron-spanning primer combination was used with a hTASK-1/3/4 specific sense
primer and a TASK-5 specific antisense primer that
amplified a 499-bp fragment. All fragments were verified by direct sequencing.
646
For expression in Xenopus laevis oocytes capped runoff poly(A ⫹) cRNA transcripts of hTASK-1, hTASK-3,
hTASK-5 and various chimera between them were synthesized and injected individually (⬃3 ng each) or in
combination in defolliculated oocytes. Oocytes were
incubated at 19°C in ND96 solution (96 mM NaCl, 2mM
KCl, 1 mM MgCl 2, 1 mM CaCl 2, 5 mM Hepes, pH
7.4 –7.5), supplemented with 100 ␮g/ml gentamicin and
2.5 mM sodium pyruvate, and assayed 48 h after injection. Two-electrode voltage-clamp measurements were
performed with a Turbo Tec-10 C amplifier (npi, Tamm,
Germany) and sampled through an EPC9 (Heka Electronics, Lambrecht, Germany) interface using Pulse/
Pulsefit software (Heka). Oocytes were placed in a
small-volume perfusion chamber and bathed with
ND96 solution.
In Situ Hybridization of Rat Brain Sections
Wistar rats were decapitated under ether anesthesia,
their brains removed and frozen on powdered dry ice.
Tissue was stored at ⫺20°C until cutting. Sixteen-micrometer sections were cut on a cryostat, thaw-mounted
onto silane-coated slides and airdried. After fixation for
10 –30 min in 4% paraformaldehyde dissolved in PBS,
slides were washed in PBS, dehydrated, and stored in
ethanol until hybridization. Both synthetic oligonucleotide probes and cRNA probes were used for in situ
hybridizations. Synthetic oligonucleotides were chosen
from the UTR and ORF with least homology to other
known sequences to minimize cross-hybridization. Antisense oligonucleotides designed for a minimal tendency of forming hairpins and self-dimers were as follows (base position on coding strand indicated).
TASK-1: (i) (103 to 143) 5⬘ agc tcc agc tgc cgc agc tcc agc
cgc tgc cgc tcg atc at 3⬘; (ii) (1354 to 1400) 5⬘ gat tta ggg
gct ggg gag gag tag agg tgg cgg tgg tga tgg gct gg 3⬘ and
TASK-3: (i) (⫺92 to ⫺138) 5⬘ gga gtc gaa gtt gca ggc ggt
ggt agc agc aga tgc cac agc tga tg 3⬘, and (ii) (1367 to
1409) 5⬘ gga tgg gcc tct att tgc cct ttg gac agc tct gcc tgg
ctg g 3⬘; TASK-5: (i) (553 to 597) 5⬘ ggt ggt gag ggt gat
gaa gca gta gta gta ggc gtg gaa gaa ggc 3⬘ and (ii) (684
to 730) 5⬘ gga cca cca ggt tga gga aag cac caa tga ctg tga
gcc cca gca gg 3⬘. Oligonucleotides were 3⬘ end-labeled
with 35S-dATP or 33P-dATP (New England Nuclear,
Boston, MA; 1200/1000 Ci/mmol) by terminal deoxynucleotidyl transferase (Roche Diagnostics) and used
for hybridization at concentrations of 2–10 pg/␮l
(4 䡠 10 5 cpm/100 ␮l hybridization buffer per slide).
Slides were airdried and hybridized for 20 –24 h at 43°C
in 100 ␮l buffer containing 50% formamide, 10% dextran sulfate, 50 mM DTT, 0.3 M NaCl, 30 mM Tris–HCl,
Karschin et al.
4 mM EDTA, 1⫻ Denhardt’s solution, 0.5 mg/ml denatured salmon sperm DNA, 0.5 mg/ml polyadenylic
acid, and the labeled probe. After hybridization slides
were washed 2⫻ 30 min in 1⫻ SSC plus 50 mM ␤-mercaptoethanol, 1 h in 1⫻ SSC at 60°C, and 10 min in 0.1⫻
SSC at room temperature. Specimens were then dehydrated, air-dried, and exposed to Kodak BIOMAX X-ray
film for 14 –28 days. For cellular resolution, selected
slides were dipped in photographic emulsion Kodak
NTB2, incubated for 4 –12 weeks, and then developed in
Kodak D-19 for 2.5 min.
For nonradioactive hybridizations digoxigenin-labeled sense and antisense cRNA probes were transcribed from cDNA fragments cloned into pBluescript
according to the manufacturers protocol (Roche Diagnostics). The fragments corresponded to bp 574 –1242 of
rTASK-1 (ORF), bp 502–1431 of rTASK-3 (ORF-3⬘UTR),
and to a 647-bp C-terminal fragment of the rTASK-5
ORF (corresponding to bp 311–1036 of hTASK-5). Hybridizations were carried out following the protocol
used by Bartsch et al. (1992) and the label was detected
by alkaline phosphatase-coupled antibodies to digoxigenin (Nucleic Acid Detection kit, Roche Diagnostics).
For identification and confirmation of brain structures with bright- and darkfield optics, sections were
Nissl-counterstained with cresyl violet (Paxinos and
Watson, 1986; Paxinos et al., 1994). Control sections
were (a) digested with RNase A (50 ng/ml) for 30 min
at 37°C before hybridization, (b) hybridized with sense
oligonucleotide and cRNA probes, or (c) hybridized
with a probe containing a 20 to 50-fold excess of unlabeled oligonucleotides. These control hybridizations resulted in a complete loss of specific hybridization signal.
ACKNOWLEDGMENTS
This work was supported in part by the Deutsche Forschungsgemeinschaft, grants Ka1175/1–3 and Da177/7–3. We thank Dr. M.
Hess for supplying human cDNA samples, Dr. F. Döring for ongoing
intellectual input, as well as Dirk Reuter, H. Wegener, Hartmut Engel,
and Anette Hennighausen for excellent technical help.
REFERENCES
Altschul, S. F., Madden, T. L., Shaffer, A. A., Zhang, J., Miller, W., and
Lipman, D. J. (1997). Gapped BLAST and PSI-PLAST: A new generation of protein database search programs. Nucleic Acids Res. 25:
3389 –3402.
Ashmole, I., Goodwin, P. A., and Stanfield, P. R. TASK-5, a novel
CNS Expression of TASK Channel Subunits
member of the tandem pore K⫹ channel family. Pflügers Arch., in
press.
Bartsch, S., Bartsch, U., Dörries, U., Faissner, A., Weller, A., Akblom,
P., and Schachner, M. (1992). Expression of tension in the developing and adult cerebellar cortex. J. Neurosci. 12: 736 –749.
Brown, D. A. (2000). The acid test for testing potassium channels.
Curr. Biol. 10: R456 –R459.
Buckler, K. J., Williams, B. A., and Honoré, E. (2000). An oxygen-,
acid- and anaesthetic-sensitive TASK-like background potassium
channel in rat arterial chemoreceptor cells. J. Physiol. 525: 135–142.
Chapman, C. G., Meadows, H. J., Godden, R. J., Campbell, D. A.,
Duckworth, M., Kelsell, R. E., Murdock, P. R., Randall, A. D.,
Rennie, G. I., and Gloger, I. S. (2000). Cloning, localisation and
functional expression of a novel human, cerebellum specific, two
pore domain potassium channel. Mol. Brain Res. 82: 74 – 83.
Decher, N., Maier, M., Dittrich, W., Gassenhuber, J., Brüggemann, A.,
Busch, A. E., and Steinmeyer, K. (2001). Characterization of
TASK-4, a novel member of the pH-sensitive, two-pore domain
potassoim channel family. FEBS Lett. 492: 84 – 89.
Duprat, F., Lesage, F., Fink, M., Reyes, R., Heurteaux, C., and Lazdunski, M. (1997). TASK, a human background K ⫹ channel to sense
external pH variations near physiological pH. EMBO J. 16: 5464 –
5471.
Girard, C., Duprat, F., Terrenoire, C., Tinel, N., Fosset, M., Romey, G.,
Lazdunski, M., and Lesage, F. (2001). Biochem. Biophys. Res. Commun. 282: 249 –256.
Goldstein, S. A. N., Bockenhauer, D., O’Kelly, I., and Zilberberg, N.
(2001). Potassium leak channels and the KCNK family of two-Pdomain subunits. Nature Rev. Neurosci. 2: 175–184.
Gray, A. T., Zhao, B. B., Kindler, C. H., Winegar, B. D., Mazurek, M. J.,
Xu, J., Chavez, R. A., Forsayeth, J. R., and Yost, C. S. (2000). Volatile
anesthetics activate the human tandem pore domain baseline K ⫹
channel KCNK5. Anesthesiology 92: 1722–1730.
Grigg, J. J., Brew, H. M., and Tempel, B. L. (2000). Differential expression of voltage-gated potassium channel genes in auditory nuclei of
the mouse brainstem. Hearing Res. 140: 77–90.
Gyapay, G., Schmitt, K., Fizames, C., Jones, H., Vega-Czarny, N.,
Spillett, D., Muselet, D., Prud’Homme, J. F., Dib, C., Auffray, C.,
Morissette, J., Weissenbach, J., and Goodfellow, P. N. (1996). A
radiation hybrid map of the human genome. Hum. Mol. Genet. 5:
339 – 46.
Holt, J. R., and Corey, D. P. (1999). Ion channel defects in hereditary
hearing loss. Neuron 22: 217–219.
Kaila, K., and Ransom, B. R. (1998). Concept of pH and its importance
in neurobiology. In PH and Brain Function (K. Kaila and B. R.
Ransom, Eds.), pp. 3–10. Wiley-Liss, New York.
Kharkovets, T., Hardelin, J. P., Safieddine, S., Schweizer, M., ElAmraoui, A., Petit, C., and Jentsch, T. J. (2000). KCNQ4, a K ⫹
channel mutated in a form of dominant deafness, is expressed in
the inner ear and the central auditory pathway. Proc. Natl. Acad. Sci.
USA 97: 4333– 4338.
Kim, D., and Gnatenco, C. (2001). TASK-5, a new member of the
tandem-pore K ⫹ channel family. Biochem. Biophys. Res. Commun.
284: 923–930.
Kim, D., Fujita, A., Horio, Y., and Kurachi, Y. (1998). Cloning and
functional expression of a novel cardiac two-pore background K ⫹
channel (cTBAK-1). Circ. Res. 82: 513–518.
Kim, Y., Bang, H., and Kim, D. (1999). TBAK-1 and TASK-1, two-pore
K ⫹ channel subunits: Kinetic properties and expression in rat heart.
Am. J. Physiol. 277: H1669 –H1678.
Kim, Y., Bang, H., and Kim, D. (2000). TASK-3, a new member of the
tandem pore K ⫹ channel family. J. Biol. Chem. 275: 9340 –9347.
647
Kindler, C. H., Pietruck, C., Yost, C. S., Sampson, E. R., and Gray, A. T.
(2000). Localization of the tandem pore domain K ⫹ channel TASK-1
in the rat central nervous system. Mol. Brain Res. 80: 99 –108.
Kubisch, C., Schroeder, B. C., Friedrich, T., Lütjohann, B., El-Amraoui,
A., Marlin, S., Petit, C., and Jentsch, T. J. (1999). KCNQ4, a novel
potassium channel expressed in sensory outer hair cells, is mutated
in dominant deafness. Cell 96: 437– 446.
Leonoudakis, D., Gray, A. T., Winegar, B. D., Kindler, C. H., Harada,
M., Taylor, D. M., Chavez, R. A., Forsayeth, J. R., and Yost, C. S.
(1998). An open rectifier potassium channel with two pore domains
in tandem cloned from rat cerebellum. J. Neurosci. 18: 868 – 877.
Lesage, F., and Lazdunski, M. (2000). Molecular and functional properties of two-pore-domain potassium channels. Am. J. Physiol. Renal
Physiol. 279: F793–F801.
Lopes, C. M. B., Gallagher, P. G., Buck, M. E., Butler, M. H., and
Goldstein, S. A. N. (2000). Proton block and voltage gating are
potassium-dependent in the cardiac leak channel Kenk3. J. Biol.
Chem. 275: 16969 –16978.
Lopes, C. M. B., Zilberberg, N., and Goldstein, S. A. N. (2001). Block
of Kenk3 by protons: evidence that 2-P-domain potassium channel
subunits function as homodimers. J. Biol. Chem. 276: 24449 –24452.
Maingret, F., Patel, A. J., Lazdunski, M., and Honoré, E. (2001). The
endocannbinoid anandamide is a direct and selective blocker of the
background K ⫹ channel TASK-1. EMBO J. 20: 47–54.
Meadows, H. J., and Randall, A. D. (2001). Functional characterisation
of human TASK-3, an acid-sensitive two-pore domain potassium
channel. Neuropharmacology 40: 551–559.
Medhurst, A. D., Rennie, G., Chapman, C. G., Meadows, H., Duckworth, M. D., Kelsell, R. E., Gloger, I. I., and Pangalos, M. N. (2001).
Distribution analysis of human two pore domain potassium channels in tissues of the central nervous system and periphery. Mol.
Brain Res. 86: 101–114.
Millar, J. A., Barratt, L., Southan, A. P., Page, K. M., Fyffe, R. E. W.,
Robertson, B., and Mathie, A. (2000). A functional role for the
two-pore domain potassium channel TASK-1 in cerebellar granule
neurons. Proc. Natl. Acad. Sci. USA 97: 3614 –3618.
North, R. A. (2000). Potassium-channel closure taken to TASK. Trends
Neurosci. 23: 234 –235.
Oertel, D. (1997). Encoding of timing in the brain stem auditory nuclei
of vertebrates. Neuron 19: 959 –962.
Oh, J. D., Woolf, N. J., Roghani, A., Edwards, R. H., and Butcher, L. L.
(1992). Cholinergic neurons in the rat central nervous system demonstrated by in situ hybridization of choline acetyltransferase
mRNA. Neuroscience 47: 807– 822.
Patel, A. J., Honoré, E., Lesage, F., Fink, M., Romey, G., and Lazdunski, M. (1999). Inhalational anesthetics activate two-pore-domain
background K ⫹ channels. Nature Neurosci. 2: 422– 426.
Patel, A. J., Maingret, F., Magnone, V., Fosset, M., Lazdunski, M., and
Honoré, E. (1999). TWIK-2, an inactivating 2P domain K ⫹ channel.
J. Biol. Chem. 275: 28722–28730.
Paxinos, G., and Watson, C. (1986). The Rat Brain in Stereotaxic Coordinates, 2nd ed. Academic Press, San Diego.
Paxinos, G., Ashwell, K. W. S., and Törk, I. (1994). Atlas of the Developing Rat Nervous System, 2nd ed. Academic Press, San Diego.
Rajan, S., Wischmeyer, E., Liu, G. X., Preisig-Müller, R., Daut, J.,
Karschin, A., and Derst, C. (2000). TASK-3, a novel tandem pore
domain acid-sensitive K ⫹ channel. J. Biol. Chem. 275: 16650 –
16657.
Rajan, S., Wischmeyer, E., Karschin, C., Preisig-Müller, R., Grzeschik,
K. H., Daut, J., Karschin, A., and Derst, C. (2001). THIK-1 and
THIK-2, a novel subfamily of tandem pore domain K ⫹ channels.
J. Biol. Chem. 276: 7302–7311.
648
Reyes, R., Duprat, F., Lesage, F., Fink, M., Salinas, M., Farman, N., and
Lazdunski, M. (1998). Cloning, and expression of a novel pHsensitive two pore domain K ⫹ channel from human kidney. J. Biol.
Chem. 273: 30863–30869.
Sirois, J. E., Lei, Q., Talley, E. M., Lynch III, C., and Bayliss, D. A.
(2000a). The TASK-1 two-pore domain K ⫹ channel is a molecular
substrate for neuronal effects of inhalation anesthetics. J. Neurosci.
20: 6347– 6354.
Sirois, J. E., Pancrazio, J. J., Lynch III, C., and Bayliss, D. A. (2000b).
Multiple ionic mechanisms mediate inhibition of rat motoneurones
by inhalation anaesthetics. J. Physiol. 512: 851– 862.
Talley, E. M., Lei, Q., Sirois, J. E., and Bayliss, D. A. (2000). TASK-1, a
two-pore domain K ⫹ channel, is modulated by multiple neurotransmitters in motoneurons. Neuron 25: 399 – 410.
Töpert, C., Döring, F., Wischmeyer, E., Karschin, C., Brockhaus, J.,
Karschin et al.
Ballanyi, K., Derst, C., and Karschin, A. (1998). Kir2.4: A novel K ⫹
inward rectifier channel associated with motoneurons of cranial
nerve nuclei. J. Neurosci. 18: 4096 – 4105.
Trussell, L. O. (1999). Synaptic mechanisms for coding timing in
auditory neurons. Annu. Rev. Physiol. 61: 477– 496.
Trussell, L. (2000). Mutant ion channel in cochlear hair cells cause
deafness. Proc. Natl. Acad. Sci. USA 97: 3786 –3788.
Vega-Saenz De Miera, E., Lau, D. H. P., Zhadina, M., Pountney, D.,
Coetzee, W. A., and Rudy, B. (2001). KT3.2 and KT3.3, two novel
human two-pore K ⫹ channels closely related to TASK-1. J. Neurophysiol. 86: 130 –142.
Wollnik, B., Schroeder, B. C., Kubisch, C., Esperer, H. D., Wieacker, P.,
and Jentsch, T. J. (1997). Pathophysiological mechanisms of dominant and recessive KVLQT K ⫹ channel mutations found in inherited cardiac arrythmias. Hum. Mol. Genet 6: 1943–1949.
Received July 10, 2001
Revised August 23, 2001
Accepted September 4, 2001