KT3.2 and KT3.3, Two Novel Human Two-Pore K Channels Closely ...

KT3.2 and KT3.3, Two Novel Human Two-Pore K1 ChannelsClosely Related to TASK-1

ELEAZAR VEGA-SAENZ DE MIERA,1 DAVID H. P. LAU,1 MARIA ZHADINA, 1 DAVID POUNTNEY,2

WILLIAM A. COETZEE,1,2 AND BERNARDO RUDY1,3

1Department of Physiology and Neuroscience,2Department of Pediatric Cardiology, and3Department of Biochemistry,New York University School of Medicine, New York, New York 10016

Received 27 April 2000; accepted in final form 8 March 2001

Vega-Saenz de Miera, Eleazar, David H. P. Lau, Maria Zhadina,David Pountney, William A. Coetzee, and Bernardo Rudy.KT3.2and KT3.3, two novel human two-pore K1 channels closely related toTASK-1. J Neurophysiol86: 130–142, 2001. We report the cloning ofhuman KT3.2 and KT3.3 new members of the two-pore K1 channel(KT) family. Based on amino acid sequence and phylogenetic analy-sis, KT3.2, KT3.3, and TASK-1 constitute a subfamily within the KTchannel mammalian family. WhenXenopusoocytes were injectedwith KT3.2 cRNA, the resting membrane potential was brought closeto the potassium equilibrium potential. At low extracellular K1 con-centrations, two-electrode voltage-clamp recordings revealed the ex-pression of predominantly outward currents. With high extracellularK1 (98 mM), the current-voltage relationship exhibited weak outwardrectification. Measurement of reversal potentials at different [K1]orevealed a slope of 48 mV per 10-fold change in K1 concentration asexpected for a K1-selective channel. Unlike TASK-1, which is highlysensitive to changes of pH in the physiological range, KT3.2 currentswere relatively insensitive to changes in intracellular or extracellularpH within this range due to a shift in the pH dependency of KT3.2 of1 pH unit in the acidic direction. On the other hand, the phorbol es-ter phorbol 12-myristate 13-acetate (PMA), which does not affectTASK-1, produces strong inhibition of KT3.2 currents. Human KT3.2mRNA expression was most prevalent in the cerebellum. In rat, KT3.2is exclusively expressed in the brain, but it has a wide distributionwithin this organ. High levels of expression were found in the cere-bellum, medulla, and thalamic nuclei. The hippocampus has a non-homogeneous distribution, expressing at highest levels in the lateralposterior and inferior portions. Medium expression levels were foundin neocortex. The KT3.2 gene is located at chromosome 8q24 1–3,and the KT3.3 gene maps to chromosome 20q13.1.

I N T R O D U C T I O N

Potassium channels are ubiquitously expressed and playcritical roles in both excitable and nonexcitable cells. Theyhelp to determine the resting potential as well as the excitabil-ity and firing properties of excitable cells (Hille 1992; Llinas1988; Rudy 1988). A large number of mammalian potassium-channel subunits have been described (Chandy and Gutman1993; Coetzee et al. 1999; Isomoto et al. 1997; Jan and Jan1997). The latest addition to this expanding set of mammalianK1-channel subunits is a group of subunits containing twoseparate pore regions within a single subunit. The first proto-

types of this group YOK1/TOK1 was discovered in yeast(Ketchum et al. 1995; Lesage et al. 1996a) and consist ofsubunits containing eight transmembrane domains (TMDs) andtwo pore regions. K1-channel subunits having two pore re-gions within a single polypeptide (KT) have also been de-scribed in mammals. However, all of these subunits describedto date contain only four TMDs (termed M1–M4) with the twopore regions (P1 and P2), respectively, located between TMDs1 and 2 and between TMDs 3 and 4. K1-channel subunitshaving this last structural organization have been identified ininvertebrates (Goldstein et al. 1996; Wei et al. 1996), mammals(Chavez et al. 1999; Duprat et al. 1997; Fink et al. 1996, 1998;Lesage et al. 1996b; Pountney et al. 1999; Reyes et al. 1998;Salinas et al. 1999), and plants (Czempinski et al. 1997),suggesting a common ancestral origin before the evolutionarydiversification of plants and animals. Seven genes of the KTfamily have been identified in mammals. In the order of theirdiscovery, they are: TWIK1 (Lesage et al. 1996b), TREK (Finket al. 1996), TASK-1 (Duprat et al. 1997; Leonoudakis et al.1998), TRAAK/KT4.1 (Fink et al. 1998; A. Ozaita and E.Vega-Saenz de Miera, unpublished data), TASK-2 (Reyes etal. 1998), TWIK-2/TOSS (Chavez et al. 1999; Pountney et al.1999), and KCNK6/KCNK7 (Salinas et al. 1999).

Here, we describe the cloning of two new members of thisemerging group of mammalian K1-channel subunits, humanKT3.2 and KT3.3, having a sequence similarity closest toTASK-1. KT3.2 subunits express channels that differ fromTASK-1 channels in pH sensitivity, modulation by PMA, andtissue distribution.

A portion of these results was presented at the 44th annualmeeting of the Biophysical Society.

M E T H O D S

Bioinformational analysis

Initial identification of the sequences was obtained by using theBlast algorithm (http://www.ncbi.nlm.nih.gov/blast/blast.cgi). Exonidentification was performed by using Blast of two sequences (http://www.ncbi.nlm.nih.gov/gorf/bl2.html), followed by sequence transla-tion (http://www.expasy.ch/tools/dna.html) and identification of se-quences by comparison to the amino acid sequence of human TASK-1

Address for reprint requests: E. Vega-Saenz de Miera, Dept. of Physiologyand Neuroscience, New York University School of Medicine, 550 First Ave.,New York, NY 10016 (E-mail: [email protected]).

The costs of publication of this article were defrayed in part by the paymentof page charges. The article must therefore be hereby marked ‘‘advertisement’’in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

130 0022-3077/01 $5.00 Copyright © 2001 The American Physiological Society www.jn.org

by 10.220.33.3 on Septem

ber 27, 2016http://jn.physiology.org/

Dow

nloaded from

http://jn.physiology.org/

and finding the consensus sequence for intron/exon boundaries (Ste-phens and Schneider 1992). Hydropathy analysis was performed usingthe Kyte and Doolittle algorithm (Kyte and Doolittle 1982) im-plemented in ProtScale web page (http://www.expasy.ch/cgi-bin/protscale.pl). The prediction of phosphorylation sites and posttrans-lational modifications analysis was performed using ProfileScan(http://www.isrec.isb-sib.ch/software/PFSCAN_form.html).

Cloning of human KT3.2

Forward and reverse primers to amplify the full coding sequence ofKT3.2 by polymerase chain reaction, using human brain cDNA (Clon-tech) as template, were designed by comparison of a human genomicsequence (Accession No. AC007869) and the sequence of humanTASK-1 (Duprat et al. 1997). The sequence for the forward primerwas (TTG CTG GCG GCC ATG AAG AG) and for the reverseprimer was (CTA AAC GGA CTT CCG GCG TTT C). The PCRconditions were: 35 cycles, 54°C annealing, 72°C extension, and 94°Cdenaturation, each step for 1 min. The band was subcloned intopCR2.1 using TA cloning (Invitrogen), and both strands were se-quenced.

To obtain additional 59UTR sequences, rapid amplification ofcDNA ends (RACE) experiments were performed using nested prim-ers. The first round of amplification was performed using a specificoligonucleotide (AAC CTA TGG TGG TGA TGA CCG TGA TCG)complementary to the sequence that encodes for the amino acidsequence ITVITTIG in KT3.2, and the forward primer AP1 formClontech Marathon-Ready human brain cDNA Kit. A second nestedamplification was performed using the oligonucleotide complemen-tary to the area that encodes for the amino acid sequence LIVCT-FTYL in the human KT3.2 cDNA (AGC AGG TAG GTG AAG GTGCAG ACG ATG AGG) and the internal forward primer AP2 form theClontech Marathon-Ready human brain cDNA kit.

Cloning of rat KT3.2

Two degenerated oligonucleotides were designed against areas withsignificant sequence conservation in human mouse and rat TASK-1,human KT3.2, andCaenorhabditis elegansn2p38 (Vega-Saenz deMiera and Lin 1992). The forward primer (ATG AAG AGG CARAAY RTN MGN AC) was designed using the sequences around thestarting methionine (MKRQNV/IRT). The reverse primer (AAG GCACCG ATN ACN GTN ARN CC) was designed to be complementaryto the amino acid sequence (GLTVIGA) located in the fourth TMD ofthese channels. The latter sequence is also conserved in the humanKT3.3 sequence. These primers were used for PCR amplification fromcDNAs prepared from mRNA isolated from thalamus and cortex(Vega-Saenz de Miera and Lin 1992). The PCR conditions wereidentical to those used for the cloning of the human KT3.2 except thatthe annealing temperature was 48°C.

Cloning of human KT3.3

PCR was performed using the primers GCC ATG CGG AGG CCGAGC GT (forward) and CTA TGG TGG TGA TGA CCG TGA T(reverse), under the following PCR conditions: 35 cycles of 58°C for1 min and 97°C for 1 min, using human genomic DNA as well humanbrain cDNAs as templates. The PCR products were separated in 2%agarose gels, and the amplified bands were subcloned as in the case ofhuman KT3.2. The final sequence of KT3.3 was assembled by com-bining the sequences obtained from the expressed sequence tag (EST)clones (seeRESULTS) and from the PCR band using an internalNotIsite. The integrity of this construct was confirmed by double-strandsequencing.

mRNA isolation

Total RNA was prepared using the guanidinium-thiocyanatemethod (Chomczyznski and Sacchi 1987) from tissues obtained from

adult (150–175 g) Sprague-Dawley rats. Poly(A) RNA was isolatedwith oligo(dT) cellulose columns as previously described (Rudy et al.1988).

Northern and dot blot analysis

Human multi-tissue master dot blot (Clontech) was hybridizedusing Stratagene QuickHyb solution. The complete coding sequenceof KT3.2 was labeled by random priming and used as a probe. Theblot was hybridized for 1 h at 68°C with two washes at RT in 23SSC0.1% SDS and a final wash at 60°C in 0.1 SSC, 0.1% SDS for 30 min.The blot was placed at RT in a phosphoimager cartridge for 48 h. Thevalues are reported as the values of the average signal measured by acircle that contains all the pixels in the dot minus the background,which was estimated by measuring an equal area free of signal but inthe vicinity of the dot. The value was finally corrected for mRNAconcentration applied to the blot and is reported as arbitrary digitalcounts per microgram of mRNA. Measurements of signal intensitywere performed using Scion Imager.

The mRNAs isolated from rat tissues were electrophoresed informaldehyde-denaturing gels and transferred to Duralon membranes(Stratagene). The membranes containing the bound mRNAs as well asa Clontech multi tissue rat Northern blot were hybridized with the first719 nucleotides of the coding sequence of rat KT3.2. The probe waslabeled with [32P]dCTP using the Boehringer random primers kit. Theblots were hybridized in QuikHyb (Stratagene) containing 13 106

cpm/ml (0.5–1.0 ng of DNA/ml) for 1 h at 68°C. They were thenwashed twice in 23SSC, 0.1% SDS at room temperature for 15 mineach, a final wash in 0.13SSC, 0.3% SDS at 60°C for 30 min. Theblots were exposed to X-ray film (X-OMAT AR Kodak) at270°C for1–2 days.

In situ hybridization

The in situ hybridization experiments were performed as described(Vega-Saenz de Miera et al. 1997), with the exception that a cDNAfragment made of the first 719 nucleotides of the coding sequence ofrat KT3.2 was used as template for probe synthesis. Blast search usingthis sequence shows areas with#81% identity (105 bp) and 74% (474bp) to rat Task-1 and 87% (41 bp), 81% (198 bp), and 76% (205 bp)to KT3.3 but not significant identities to other cDNAS. Specificitywas further checked by hybridization of this probe against a blotcontaining synthetic RNA of rTASK-1 (Leonoudakis et al. 1998), ratKT3.2, and human KT3.3. A strong signal was with rat KT3.2, but nosignal with rTASK or KT3.3 was obtained after 24 h (data not shown).Another evidence of specificity was obtained by the results of theNorthern blot experiments in which the distribution as well as the sizeof the mRNA band do not mach to the ones described for rTASK(Leonoudakis et al. 1998).

Two-electrode voltage clamp of Xenopus laevis oocytes

The recombinant pCR2.1 plasmid containing KT3.2 was linearizedby digestion withScaI, and cRNA was transcribed in vitro using T7RNA polymerase (Stratagene). Stage IV-VXenopusoocytes wereharvested, treated with collagenase (1 mg/ml), and micro-injectedwith 50 nl cRNA (;20 ng) (Iverson and Rudy 1990). All electro-physiology experiments were carried out at room temperature (22°C)1–2 days following injection. Recordings were made in solutionscontaining ND96 (in mM): 1.8 CaCl2, 1 MgCl2, 96 NaCl, 2 KCl, and5 HEPES, pH 7.5, or KD98, which is similar in composition to ND96except that NaCl was replaced with KCl. Intermediate concentrationsof potassium were obtained by mixing ND96 and KD98. Squarevoltage-clamp pulses were applied from a holding potential of290mV (ND96), 240 mV (KD14),230 mV (KD26) 220 mV (KD50),or 0 mV (KD98) to test potentials ranging from2150 to150 mV in10-mV increments with a 3-s inter-pulse interval. Acquisition of data

131KT3.2 AND KT3.3, TWO NOVEL KT CHANNELS RELATED TO TASK-1



Dow

nloaded from


and subsequent analysis was performed using pClamp 6.0 (AxonInstruments) and Microsoft Excel.

R E S U L T S

Human KT3.2

Screening of the GeneBank EST (expressed sequence tags)database using the Blast algorithm (Altschul et al. 1997) withthe sequence of TASK-1 cDNA (Accession No. AF006823) asa probe yielded two EST sequences, AI690321 and AI739096,distinctly different from but related to TASK-1. The cloneswere incomplete at the 59 end and are partial sequences ofKT3.3 (vide infra). Using the partial sequences of KT3.3 foundin the EST database, a search in the high throughput genomicsequences database (HTGS) identified two human genomicsequences containing two new KT genes. One of the genesencodes a KT channel that had 58% identity and 72% similar-ity to the published human TASK-1 sequence (Accession No.AC007869) These values increase to 75% identity, 86% sim-ilarity if only the core region is used for comparison. We referto this gene as KT3.2.

Analysis of the KT3.2 genomic sequence identified twoexons that cover the full coding sequence of this channel. Theexons are separated by an intron that is larger than 80 kb (Fig.1A). Using the predicted coding sequence as a query in a searchagainst the EST database revealed a sequence that contains thecarboxy terminus of a KT3.2 transcript (AA349574) andserved to confirm the position of the stop codon. A third exon,encoding the 39UTR, was also identified. The first splicejunction occurs at the first glycine of the GYG sequence in thefirst pore region. Oligonucleotide primers designed againstsequences surrounding the predicted starting methionine andthe stop codon of KT3.2 were used for PCR amplification withhuman brain cDNA (Clontech) as template. A band of 1,137 bpwas observed with gel electrophoresis, and the cDNA waspurified and subcloned. Three independent clones were iden-tical in sequence and differed only at nucleotide 636 of thecoding sequence from the putative sequence predicted by theanalysis of the genomic sequence contained in AC007869 (athymidine-to-cytosine substitution). This substitution does notalter the amino acid sequence and may represent a singlenucleotide polymorphism (Fig. 1C). To confirm that the as-signed starting methionine was the real starting methionine,five different RACE experiments using different primers wereattempted. In each case, but in the last one, we obtain sequencerelated to KT3.2 that did not reach the starting methionine. Inthe last case, using the primers for the RACE experimentsdescribed inMETHODS, we isolated 75 clones, 50 of whichcontained additional sequence related to KT3.2. From thisexperiment, we obtained 74 extra nucleotides without a stopcodon in frame. The coding sequence found in the identifiedsequence predicts a protein of 374 aa with a molecular weightof 42,264 Da. Hydrophobicity analysis (Kyte and Doolittle1982) predicts the presence of four transmembrane domains(Fig. 1B). The sequence contains two regions between the firstand second and between the third and fourth TMDs having thecanonical GYG/GFG sequence that is part of the pore ofpotassium channels (Heginbotham et al. 1994) (Fig. 1C). ApotentialN-glycosylation site (Fig. 1C, ■) (Miletich and Broze1990) is found in the linker between M1 and P1. Interestingly,as in TASK-1, this new protein does not have a cysteine in the

extracellular M1-P1 that is found in all other mammalian KTchannels and that has been suggested to be critical for channelassembly (Lesage et al. 1996c). The carboxy terminal sequenceof the predicted protein contains three consensus sequences forphosphorylation by protein kinase C (Fig. 1C,F) (Woodgett etal. 1986) and one for PKA (Fig. 1C,Œ) (Feramisco et al. 1980;Glass et al. 1986).

Human KT3.3

A Blast search using the partial EST sequences of KT3.3described in the preceding text identified a new EST sequence,AI968607, which contains 140 bp in the 59 that do not matchthe previous ESTs. Translation of this extra sequence in one ofthe three reverse frames produces a 25 amino acid sequencewith high degree of similarity to the sequences in the aminoterminal of TASK-1 and KT3.2. We speculated that this couldrepresent an inversion of the 59 sequence of the EST producedduring library construction. PCR experiments using humangenomic DNA or human brain cDNA with primers designedagainst the starting methionine predicted by the inverted 59sequence and an internal primer based on the previous KT3.3partial sequences produced a 283-bp product that was clonedand sequenced. The sequence of this product confirmed that theantisense sequence of AI968607 corresponds to the 59 se-quence of KT3.3. Combination of the sequence obtained fromthe PCR experiment with the one obtained from ESTAI690321 produced a 1,286-bp-long sequence with a 41-bppolyadenylation track. The open reading frame of this cDNApredicts a protein of 330 amino acids with a molecular weightof 36,130 Da (Fig. 1F).

Hydrophobicity analysis (Kyte and Doolittle 1982) predictsa protein containing four TMDs. The protein also includes twoloops containing the canonical GYG/GFG signature of the K1

channel pore (Heginbotham et al. 1994), located between thefirst and second TMDs and the third and fourth TMDs (Fig.1E). Two putative sites for protein kinase C (Fig. 1F, F) andone for protein kinase A (Fig. 1F,Œ) are also observed.

A Blast search of the HTGS database of GeneBank, usingKT3.3 sequence as a probe identified a genomic sequencecontaining the KT3.3 gene (Accession No. AL118522). Anal-ysis of KT3.3 sequence and the genomic clone AL118522using Blast 2 sequences shows that the cDNA sequence ismade by the participation of two different exons, separated bya 3,936-bp intron (Fig. 1D).

Expression of KT3.2 in Xenopus oocytes

Xenopusoocytes were injected with KT3.2 cRNA andstudied after 18 – 48 h in a bath solution of ND96. Theseoocytes had a resting potential of289 6 2 mV (n 5 15; P ,0.05), which was;45 mV more negative than that ofuninjected oocytes (244 6 2 mV; n 5 7). This membranepotential was close to the K1 equilibrium potential (calcu-lated to be290 mV under our experimental conditions),suggesting that the expressed channels are selectively per-meable to K1 and can play an important role in the gener-ation of the resting potential.

Using two-electrode voltage-clamping techniques (inND96 solution), mainly outward currents were observed.The currents activated almost instantaneously (Fig. 2A).

132 VEGA-SAENZ DE MIERA ET AL.



Dow

nloaded from


FIG. 1. Genomic organization, hydropathy profile, nucleotide and amino acid sequence of KT3.2 and KT3.3.A andD: genomicorganization of KT3.2 and KT3.3 genes, respectively.h, exons;o, coding areas. Start, stop, and splice sites are indicated.Connecting introns are shown by lines and their sizes are indicated. Codon for AA 95 in KT3.2 and KT3.3 created by the junctionof exons 1 and 2.B andE: hydrophobicity plot for KT3.2 and KT3.2, respectively (Kyte and Doolittle 1982). The areas of thetransmembrane domain (TMD) as well as of the 2 pores are indicated.C andF: nucleotide and predicted amino acid sequence ofKT3.2 and KT3.3, TMD, and pore regions are over lined. Consensus sites for:N-linked glycosylation (f), PKC (F), and PKA (Œ)were identified as described inMETHODS. These sequences have been deposited in GeneBank under Accession No. AF257080 forKT3.2 and AF257081 for KT3.3.




Dow

nloaded from


These currents were absent in control water-injected oocytes(Fig. 3C). At high extracellular K1 concentrations (KD98),large inward currents were detected (Fig. 2B) although aweak outward rectification can be observed; (Fig. 2E). Inthis respect, KT3.2 differs from the closely related TASK-1,which shows a linear current-voltage relationship in highextracellular K1 concentrations (Duprat et al. 1997; Leo-noudakis et al. 1998). The reversal potentials obtained fromFig. 2E, plotted as a function of the K1 concentration,yielded a linear relationship with a slope of 486 1.3 mV per10-fold change in K1 concentration (n5 7; Fig. 2F), a valuethat is close to the theoretical value of 58 mV, as predictedby the Nernst equation for a K1-selective channel. We wereunable to detect any current (recorded in ND96 with a pHthat ranges from pH 6.5 to 8.5) or shift in the restingpotential in oocytes injected with KT3.3 cRNA.

Modulation of KT3.2 currents by pharmacological agentsand by pH

Currents produced by KT3.2 were relatively insensitive toextracellular application of TEA or 4-aminopyridine (4-AP;;20% reduction in outward current was observed when allNa1 was replaced by TEA or 4-AP; data not shown). Moderateblock was observed by extracellular quinidine (;20% block

occurred with 100mM and 80% block with 1 mM). Similarresults were obtained using quinine (data not shown). Ba21 (3mM) added to the perfusion solution blocked the current. Thisblockade was released at potentials positive to the reversalpotential in a time-dependent manner (Fig. 3B). After reachingsteady state, inward currents are blocked proportionally morethan outward currents (;80% block occurred at2150 mV,whereas only;20% block occurred at150 mV; Fig. 3C). Thiseffect of Ba21 is not due to changes in membrane surfacecharge because increasing Ca21 to a final concentration of 4.8mM or Mg21 to a final concentration of 4 mM had no effect onKT3.2 currents (1126 6%, n 5 3, for Ca21 and 1066 6%,n 5 3, for Mg21).

KT3.2 currents were inhibited by the phorbol ester PMA,which is a potent PKC activator. A 50% inhibition was ob-served within 20 min of application of 10 nM PMA. Anincrease in PMA concentration to 100 nM produced similarresults (Fig. 4). At 1 h after application, an 80% inhibition wasobserved (Fig. 4). The currents were not affected by incubationwith the inactive phorbol ester 4a 4alpha-phorbol 12,13-dide-canoate (PDD) (Fukushima et al. 1996) or in the absence ofphorbol ester application (Fig. 4), suggesting that specificactivation of PKC was responsible for the decrease in currentproduced by PMA. In contrast TASK-1 has been described tobe insensitive to 100 nM PMA (Duprat et al. 1997).

KT3.2 and TASK-1 channels also differ in pH sensitivity. Acharacteristic of TASK-1 channels, from which its name wasderived (TASK: two pore acid sensitive K1 channel), is theirhigh sensitivity to pH changes around the physiological pHrange, (current magnitude depends on pH with a pK of 7.2960.03 and is nearly completely blocked at pH 6.5) (Duprat et al.1997; Leonoudakis et al. 1998). KT3.2 currents were also pHsensitive (Fig. 5), but the pH dependence is shifted toward

FIG. 3. Effect of Ba21 on KT3.2 currents.A: currents recorded in ND96under similar conditions as in Fig. 2.B: same asA but in the presence of 3 mMBa21. C: plot of the current recorded with or without Ba21 at 10 ms (‚, noBA21; E, with Ba21) and after reaching the plateau (}, no BA21; ■, withBa21). Note than‚ are on top of}.

FIG. 4. Time course of the effect of phorbol esters on KT3.2 currents inXenopusoocytes: average normalized currents recorded at 200 ms during apulse to150 mV from a holding potential of290 mV. Results are the averageof the recorded current normalized to the control current before any treatment.Error bars represent SD,n 5 3.

FIG. 2. Functional expression of KT3.2 inXenopusoocytes.A: whole cellcurrents of KT3.2-injected oocytes recorded in ND96, from a holding potentialof 290 mV to pulses from2150 to 50 mV in 10-mV increments.3, 0 currentlevel.B: as in Abut in KD98 and a holding potential of 0 mV.C andD: waterinjected oocytes in ND96 and KD98, respectively, under identical recordingconditions as inA and B. E: plots of the currents recorded from oocytesexpressing KT3.2 at the indicated K1 concentrations.F: reversal potential ofKT3.2 currents as a function of [K1]o.




Dow

nloaded from


more acid pH’s by over 1 pH unit, such that KT3.2 currentswere relatively insensitive to changes in the pH 7.0 to 8.0 range(Fig. 5). Furthermore, KT3.2 currents were not affected byapplication of dinitrophenol (1 mM for#20 min, data notshown), a treatment that presumably acidifies the cytosol andproduces 50% reduction of TASK-1 currents after 6-min incu-bation (Leonoudakis et al. 1998).

Tissue distribution of human KT3.2

The pattern of KT3.2 tissue expression was studied by usingmultiple human tissues mRNA dot blots (Clontech). We foundexpression to occur mainly in neuronal tissue, particularly incerebellum, although hybridization signals were also detectedin samples of adrenal gland, kidney and lung mRNA (Fig. 6C).This expression pattern of KT3.2 is different from that ofhuman TASK-1, which expresses more predominantly in pla-centa and pancreas followed by brain, lung, prostate, and heart(Duprat et al. 1997).

Distribution of KT3.2 in rat tissues

To study the expression of the KT3.2 gene in rat, we generatedspecies-specific probes. We used RT-PCR with degenerate oligo-nucleotides (seeMETHODS) with cortex and thalamus rat cDNAs as

template. Two different sequences were isolated from these ex-periments. One was identical to the rat TASK-1 sequence (Leou-noudakis et al. 1998), whereas the other encoded a sequence of237 aa with 95% identity (99.6% similarity) to human KT3.2 (Fig.7). This rat ortholog of KT3.2 was used as a probe for in-situhybridization and Northern blot experiments.

Northern blot analysis revealed that the KT3.2 gene is ex-pressed in the CNS but not in heart, lung, liver, kidney,intestine, or skeletal muscle. This pattern of expression isdifferent from that of TASK-1, which in rat and mouse ex-presses mainly in heart and lung (Duprat et al. 1997; Leo-noudakis et al. 1998) (Fig. 8). In situ hybridization experimentsrevealed a wide expression pattern in rat brain (Fig. 10, Table1). The highest expression levels were found in the olfactorynuclei, piriform cortex, cerebellum, anterodorsal thalamic nu-cleus, pontine nucleus, dorsal raphe, and several nuclei in themedulla (Fig. 9). An intriguing expression pattern was found inhippocampus, with low levels of expression in the medial,anterior, and superior areas and higher levels in the lateral andventral areas (Fig. 9,A, B, F, Gvs. H andB). Diffuse expres-sion was also observed throughout the cortex (Fig. 9,A–H andTable 1). In spinal cord, the channel is expressed in the anteriorand posterior horns and more intensely in layer 9 of Rexed(Fig. 9M and Table 1).

FIG. 5. Effect of pH on KT3.2 and TASK-1.A: current-voltage relationships determined at different pH’s from KT3.2-injectedXenopusoocytes recorded during test potentials from2150 to 50 mV in 10-mV intervals from a holding potential of290 mV.Please note than the pH 8.0 values are under the pH 7.4 curve.B: current-voltage relationships of TASK-1 at different pH (modifiedfrom Leonoudakis et al. 1998).C: percentage of the current remaining at different pH’s. The currents were normalized to the currentrecorded at pH 8.5 (maximum current). The values of TASK-1 were obtained from Duprat et al. (1997).

FIG. 6. Pattern of tissue expression of human KT3.2.A: a multi-tissue master dot blot (Clontech) containing mRNAs from 50different tissues was hybridized with the full-coding cDNA sequence of human KT3.2 under the conditions described inMETHODS.The photograph shown was obtained after 5 days exposure at270 with intensifier screens.B: dot blot index of the different humanmRNAs.C: histogram of human KT3.2 concentration obtained after normalizing the signal recorded with the phosphoimager to1 mg of mRNA using the values supplied by the membrane manufacturer.




Dow

nloaded from


Chromosomal localization of KT3.2 and KT3.3

THE KT3.2 GENE IS LOCALIZED TO 8q24.1–3. A database searchusing the genomic sequence containing KT3.2 identified threesequence-tagged sites (STS), having Accession Nos. Z53064,Z53078, and G23634. These STSs correspond to markersD8S1741 and D8S1743, which are, respectively, localizedin the Geneton frame at 161.51 Mbp (536.87 cR3000 inGeneMap’99 (http://www.ncbi.nlm.nih.gov/genemap/loc.cgi?ID519131) and WI-14922, located at 698 cR from the top ofchromosome 8 in the MIT-Whitehead frame (http://carbon.wi.mit.edu:8000/cgi-bin/contig/sts_info?sts5EST275360&database5release), and 536.87 cR3000 in GeneMap’99(http://www.ncbi.nlm.nih.gov/genemap/loc.cgi?ID519130).The fact that the three markers are located in the same generalarea strongly supports the localization of this gene to this area.Although these frames are not anchored to the cytogeneticmap, it is very probable that, based on the data obtained fromGeneMap’99, this gene is located at 8q24.1–3.

THE KT3.3 GENE IS LOCATED TO CHROMOSOME 20q13. Searchingthe STS database using the genomic sequence found inAL118522, we obtained four human STS Accession Nos.,G18091, G18129, G07367, and AL031792. All of these se-quences map to human chromosome 20. The sequence G07367corresponds to the human STS marker WI-9624. The MITmap (http://carbon.wi.mit.edu:8000/cgi-bin/contig/sts_info/406)localized this marker to the chromosome 20 reference map at286.01 cR from top of the Chr20 linkage group. In GeneMap 98(http://www.ncbi.nlm.nih.gov/genemap98/map.cgi?MAP5GB4&BIN5581&MARK5WI-9624) it is localized at 56.6 cM(235.78 cR3000) from the top of chromosome 20. Although thereis no strict correlation between these frames and the cytogeneticmaps, it is very probable, based on the data contained inGeneMap98, that this gene is located at 20q12–13.1.

D I S C U S S I O N

We report here the identification of two new members of theemerging four TMDs or two-pore (KT) family of K1 channelpore forming subunits: KT3.2 and KT3.3. The KT familyconstitutes the largest family of K1-channel proteins in theC.elegansgenome, with close to 40 KT genes identified (Wang etal. 1999). Sequence analysis suggest that about half of thesegenes can be grouped into subfamilies containing at least twomembers, while the remainder appear to be unique genes.

Counting KT3.2 and KT3.3, a total of nine different KT geneshave been identified in mammals. The identities in the coreregion (TMD1–TMD4) of mammalian KT proteins range from25 to 75% (Table 2). As the number of identified genes grows,it becomes possible to identify groups within the same specieshaving particularly large similarities that can be consideredseparate subfamilies (Fig. 10).

With the cloning of the human KT3.2 and KT3.3 genes, aclear new subfamily with more than one member in the samespecies can be defined (Fig. 10). Human KT3.2 is 58% iden-tical (72% similarity) to human TASK-1. In the core region(TMD1–TMD4), these values increase to 75% identity and86% similarity, while KT3.3 shows 68% identity (81% simi-larity) to TASK-1. Thus TASK-1, KT3.2, and KT3.3 can beconsidered members of a subfamily that has at least threemembers in the same species (human). At least two membersof this subfamily also exist in rat (rat KT3.2, this paper; ratTASK-1: Leonoudakis et al. 1998). Interestingly, the membersof this subfamily are the only mammalian KT genes for which

FIG. 7. Alignment of human and ratKT3.2. The partial sequence of rat KT3.2was obtained by rtPCR from cDNAs pre-pared from brain cortex and thalamus usingdegenerated oligonucleotides and was de-posited in GenBank with Accession No.AF257082. The predicted amino acid se-quence of rat KT3.2 was aligned against thehuman KT3.2 sequence. Amino acid conser-vation is indicated by black background,TMD and pores are indicated by labeledoverlines; the localization of the primersused to amplify the full sequence are indi-cated by arrows, andN-linked glicosylationsites byŒ.

FIG. 8. Northern blot analysis of rat KT3.2.Top: X-ray autoradiogram of arat multiple tissue mRNA blot hybridized with a probe made with the first 719nucleotides of the coding sequence of rat KT3.2. The blot was exposed for 2days at270°C with intensifier screens.Bottom: photograph of the X-rayautoradiogram of the same blot following stripping and hybridization to ab-actin probe. The film was exposed for 3 h at 270°C with intensifier screens.mRNA (2mg) was applied to each lane except for the cerebellum lane in which5 mg were applied. Cx, cortex; Cer, cerebellum; Hip, hippocampus; H, heart;Lg, lung; Lv, liver; K, kidney; I, intestine; M, muscle.




Dow

nloaded from


TABLE 1. Expression of KT3.2 in brain

Olfactory bulb

Taenia Tecta 11Anterior olfactory nucleus

Medial part 1111Ventral part 1111Dorsal part 111

Glomerular cell layer 1Granular cell layer 11

Cerebral cortex

Anterior cingulate cortex 1Posterior cingulate cortex 11Piriform cortex 1111Entorrinal cortex 111Nucleus of the lateral olfactory tract 111

Amygdala

Amygdalohippocampal area 111Cortex-amygdala transition zone 111Anterior cortical amygdaloid nucleus 111Basomedial amygdaloid nucleus 11Median amygdaloid nucleus 1111Basolateral amygdaloid nucleus 1Posterolateral cortical amygdaloid nucleus 111

Septum

Lateral septal nucleus, intermediate 11Lateral septal nucleus 1

Preoptic area

Median 111Magnocellular preoptic nucleus 11Anteroventral preoptic nucleus 11Preoptic suprachiasmatic nucleus 111Nucleus of the horizontal limb of the

diagonal band 111Supraoptic nucleus 1

Hypothalamus

Dorsomedial hypothalamic nucleus 111Ventromedial hypothalamic nucleus 111Arcuate hypothalamic nucleus 111Supra chiasmatic nuclei 111Lateroanterior hypothalamic nucleus 111Anterior hypothalamic area 111

Thalamus

Anterodorsal thalamic nucleus 1111Anterior pretectal nucleus 1Anteroventral thalamic nucleus 1Anteromedial thalamic nucleus 1Zona incerta 11Medial geniculate nucleus, ventral 1Dorsal lateral geniculate nucleus 11Ventroposterior thalamic nucleus, lateral part 1Peripeduncular nucleus 11Medial geniculate nucleus 11

Hippocampus

Ant. Lat. Med. Post. Inf.Dentate gyrus 2 1/2 11 111 111CA1 2 1 1 1 111CA2 2 2 2 11 111CA3 2 2 2 1 11

Midbrain

Inferior colliculus 11Oculomotor nucleus 1111Dorsal raphe nucleus 1111Substantia nigra compacta 1Central gray 1Dorsal tegmental nucleus pericentral part 11Central gray, dorsal 111Interpeduncular nucleus, rostrolateral subnucleus 111Interpeduncular nucleus, rostral subnucleus 1Interpeduncular nucleus 1111Superficial gray layer of the superior colliculus 1

Cerebellum

Granular cell layer 1111Deep cerebellar nuclei 1

Pons

Motor trigeminal nucleus 1111Pontine nucleus 1111

Medulla Oblongata

Facial nucleus 1111Hypoglossal nucleus 1111Nucleus of the solitary tract 1111Inferior olive 1Spinal trigeminal nucleus, caudal part 1Spinal trigeminal nucleus, oral part 1Lateral vestibular nucleus 1Gelatinous layer of the caudalSpinal trigeminal nucleus 11Lateral reticular nucleus 11Ambiguus nucleus 111

Spinal cord

Grey matter 11Spinal cord layer 9 of Rexed 1111




Dow

nloaded from


a clear ortholog can be found in the C.elegansgenome, then2p38 gene (Wang et al. 1999) (Table 2; Figs. 10 and 11).

HTWIK (Lesage et al. 1996b) together with TWIK-2/TOSS(Chavez et al. 1999; Pountney et al. 1999) and TRAAK/KT4.1

(Fink et al. 1998; A. Ozaita and E. Vega-Saenz de Miera,unpublished results) together with TREK (Patel et al. 1999)may also define two additional KT subfamilies (Fig. 10),although the similarities within these subfamilies are less than

FIG. 9. Expression of KT 3.2 in the rat CNS. Autoradiograph of sagittal (AandB) and coronal sections at different levels (C–M)of rat CNS hybridized with random primed labeled35S cDNA rat KT3.2 probe. aca, anterior commissure, anterior part; ac anteriorcomisure; Acg, anterior cingulate cortex; AD, anterodorsal thalamic nucleus; AHA anterior hypothalamic area, anterior part; Ahi,amygdalohippocampal area; AMB, anteromedial thalamic nucleus, ventral part; Amb, ambiguus nucleus; AO, anterior olfactorynucleus; AOD anterior olfactory nucleus, dorsal part; AOV anterior olfactory nucleus, ventral part; APT, anterior pretectal nucleus;Arc, arcuate nucleus; BL, basolateral amygdaloid nucleus; CA1 field CA1 of hippocampus; CA2 field CA2 of hippocampus; CA3field CA3 of hippocampus; ml corpus callosum; CG, central gray; CGD, central gray, dorsal; Cli, caudal linear nucleus of the raphe;CnF, cuneiform nucleus; Cpu, caudate putamen; DCn deep cerebellar nuclei; Dco, dorsal cochlear nucleus; DG, dentate gyrus;DLG, dorsal lateral geniculate nucleus; DMC, dorsomedial hypothalamic nucleus, compact part; DR, dorsal raphe nucleus; Ent,entorrinal cortex; FrCx, frontal cortex;; Ge5, gelatinous layer of the caudal spinal trigeminal nucleus; Gr, cerebellar granular celllayer; IC, inferior colliculus; Int, interposed cerebellar nucleus, IO, inferior olive; IP, interpeduncular nucleus; LatC, lateral(dentate) cerebellar nucleus; LOT, nucleus of the lateral olfactory tract; LRt, lateral reticular nucleus; LSI, lateral septal nucleus,intermediate part; Lv, lateral ventricle; Lve, lateral vestibular nucleus; ME, median eminence; MG, medial geniculate nucleus;MGV, medial geniculate nucleus, ventral; Mi, mitral cell layer of the olfactory bulb; Mo5, motor trigeminal nucleus; Obgr, granularcell layer of the olfactory bulb; PCg, post cingulate cortex; Pir, piriform cortex; Pn, Pontine nucleus; PP, peripeduncular nucleus;PSCh, preoptic suprachiasmatic nucleus; SCh, supra chiasmatic nuclei; SNC, substantia nigra compacta; Sol, nucleus of the solitarytract; Sp5O, spinal trigeminal nucleus, oral part; SuG, superficial gray layer of the superior colliculus; TT, taenia tecta; VMH,ventromedial hypothalamic nucleus,; VP, ventral pallidum; 3, oculomotor nucleus; 3v, third ventricle; 7, facial nucleus; 9 spinalcord layer 9 of Rexed; 12, hypoglossal nucleus.

TABLE 2. Percent identities between core regions of KT channels

TREK/KT2.1 KT2.2 TASK1/KT3.1 KT3.2 KT3.3 KT4.1 TASK2/KT5.1 TWIK2/KT6.1 KCNK7/KT7.1 N2p20 N2p38

TWIK1/KT1.1 31.6 33.5 30.7 31.3 31.4 33.1 28.3 48.6 41.7 25.8 26.7TREK/KT2.1 70.1 30.5 30.3 31.6 51.2 31.8 31.1 27.6 26.8 30.2KT2.2 31.3 28.4 30.4 55.2 34.4 30.1 28.9 25.5 28.4TASK1/KT3.1 75.2 68.2 32.3 31.3 28.4 26.8 45.5 59.4KT3.2 68.2 31.1 31.6 28.8 25.4 45.5 59.4KT3.3 31.4 32.6 28.7 27.7 43.1 52.7KT4.1 34.0 35.4 29.6 27.9 32.6TASK2/KT5.1 30.8 25.6 27.8 29.0TWIK2/KT6.1 37.6 26.2 23.8KCNK7/KT7.1 22.5 24.0N2p20 51.6

KT, two-pore K1 channel.




Dow

nloaded from


in the TASK group (48.6 and 51.2% identity, respectively, inthe core region; Table 2). Interestingly, another gene related toTREK, human KT2.2, obtained by analysis of genomic se-

quences (Accession No. AL122021), has 70% identities in thecore region (Vega-Saenz de Miera, personal observation).

The emergence of several KT subfamilies led us to imple-ment a nomenclature system following the model developed byChandy et al. (1991) for the pore forming subunits of voltage-gated (Kv) K1 channels. A similar nomenclature has beenapplied to the inward rectifier family (Isomoto et al. 1997).Following these models, each independent KT gene was de-nominated KTx.y where the K stands for potassium and Tstands for “two pore.” The “x” refers to the subfamily numberand the “y” to the member’s number in that subfamily. Sub-families and members within a subfamily are numbered ac-cording to their order of publication. Accordingly TASK-1would be called KT3.1 because it was the first identifiedmember of the third identified subfamily of KT genes. Hence,the newly discovered genes published in this manuscript werenamed KT3.2 and KT3.3. Although these genes could also becalled TASK-2 and TASK-3, we should point out that thename TASK-2 has already been used in the literature for a genethat also expresses pH-sensitive K1 channels (TASK: two poreacid-sensitive K1 channel) but is evolutionary more distantlyrelated (Fig. 10, Table 2).

In the case of the Kv family of K1-channel-forming sub-units, the classification into subfamilies turned out to be quiteuseful since the subfamilies, in addition to reflecting closelyrelated homologs, also appear to constitute functional groups.Kv subunits can form homomultimeric channels that are be-lieved to be tetrameric (Coetzee et al. 1999). Subunits of thesame subfamily can also interact and form functional hetero-multimeric channels (reviewed in Coetzee et al. 1999). In theKT family, it is believed that the channels are formed bydimers of KT subunits (Lesage et al. 1996c). Whether KTsubunits also form heteromeric channels and whether this is or

FIG. 10. Phylogenetic tree of the K1 channel (KT) family. Alignment ofthe KT channels was produced with Clustal W (Thompson et al. 1994), and thetree was produced by the neighbor-joining method (Saitou and Nei 1987). Thenumbers represent the fractional differences of the sequence with respect tothe node (i.e., 0.15 10% difference). TOK1 (yeast) was used as an outgroupgene. Except for n2p20 and n2p38 (Caenorhabditis elegans), all the remaininggenes are human. The names in parentheses represent the corresponding KTchannel nomenclature suggested in this paper.

FIG. 11. Alignment of of the human K1 channels, KT3.2, KT3.3, and TASK-1, and of theC. elegansn2p38 and n2p20 genes.Level of amino acid similarities indicated by the darkness of background, TMD and pore regions are overlined.N-linkedglycosylation site is indicated byŒ.




Dow

nloaded from


not restricted to members of the same subfamilies remains tobe investigated.

FUNCTIONAL ROLES OF KT CHANNELS. Several of the clonedKT subunits express K1 currents in heterologous expressionsystems, presumably by forming homodimers (Lesage et al.1996). A few others, including TOSS (Pountney et al. 1999),KCNK6, KCNK7 (Salinas et al. 1999), and KT3.3 (this paper),do not express currents when introduced into heterologousexpression systems.

Although there are some subtle differences, all KT subunitsexpressed to date form “leak”- or “open rectifier”-type K1

channels, i.e., the channels show little voltage- or time-depen-dent gating, and the currents show a dependence on extracel-lular K1 concentration close to that described by the Goldman-Hodgkin-Katz equation. Since they are open at voltages belowthe threshold for action potential generation, these channels arelikely to influence the subthreshold behavior of neurons, in-cluding helping set the resting potential and the resting imped-ance of the cell. Such channels are of great interest becausethey can dictate the degree of responsiveness of a neuron toincoming synaptic inputs. KT channels might be one the maincontributors to the K1 “leak” or resting current in neurons andone of the main determinants of the resting potential.

Of the nine KT genes known to date, seven are known to beexpressed in brain, but not much is known about the cellularpatterns of expression in this tissue. Nevertheless even fromthe little available data, it seems likely that many CNS neuronsexpress more than one KT subunit. As shown here, KT3.2 iswidely expressed within the CNS, and the same seems to betrue for TRAAK/HKT4.1 (Fink et al. 1998; A. Ozaita and E.Vega-Saenz de Miera, unpublished data), and TREK (Fink etal. 1996). TASK-1 (Talley et al. 2000) and KT3.2 (this paper)appear to overlap in the cerebellar granule cell layer (theexpression of TASK-1 in cerebellar granule cells has also beenconfirmed with specific antibodies) (Millar et al. 2000), theolfactory bulb, and several brain stem nuclei like the ambigual,motor trigeminal, facial, vagal, and hypoglossal nuclei.

Although many KT channels produce similar “leak” K1

currents, and they may have overlapping patterns of expres-sion, they differ from each in their regulation by physiologicalfactors. In this context, the strong sensitivity of TASK-1 chan-nels to changes in extracellular pH in the physiological rangeand of TRAAK and TREK channels to arachidonic acid, arefindings from heterologous expression studies of great interest.TRAAK and TREK-1 also respond to stress and might bemechanosensitive K1 channels (Maingret et al. 1999; Patel etal. 1998). This allows for multiple pathways of modulating theresting K1 conductance of neurons. Therefore the discovery offactors that modulate specific KT channels is crucial to under-stand their roles in neurons.

KT3.2 channels are notably different from those expressedby TASK-1 in their sensitivity to extracellular pH. The pHdependence of KT3.2 is shifted to more acidic pH’s by 1 pHunit, with the result that KT3.2 is much less sensitive tochanges in pH around physiological values than TASK-1 (Fig.5). On the other hand, KT3.2 currents are strongly modulatedby phorbol esters that activate PKC and do not affect TASK-1channels (Duprat et al. 1997; Leonoudakis et al. 1998). Thepresence of both TASK-1 and KT3.2 channels in the same cell(as might be the case in cerebellar granule cells and in mo-

toneurons) may allow for the independent modulation of thecell’s resting K1 conductance by extracellular pH and byneurotransmitters that activate the PLC pathway.

CHROMOSOMAL MAPPING OF KT3.2 AND KT3.3. The KT3.2 genewas mapped to human chromosome 8q24.1–3. Several diseaseshave been mapped to this area, including benign neonatalfebrile convulsion/KCNQ3 (Charlier et al. 1998; Lewis et al.1993), childhood absence epilepsy (Fong et al. 1998), idio-pathic general epilepsy (Zara et al. 1995), benign adult familialmyoclonic epilepsy (Mikami et al. 1999), Lom syndrome, anautosomal recessive peripheral neuropathy with deafness (Ka-laydjieva et al. 1996), and Meleda disease, a cardiac diseasewith cardiomegaly, electrocardiographic abnormalities, andepisodes of ventricular tachycardia (Fischer et al. 1998). On theother hand, the KT3.3 gene was mapped to 20q13.1, an area inwhich BFNC, a rare kind of epilepsy, has been mapped (Singhet al. 1999). More interestingly, several tumors (Bossolasco etal. 1999; De Angelis et al. 1999; Palmedo et al. 1999; Tanneret al. 1995) show a duplication of this chromosomal area withmultiple copies being present in the same cell, which mayexplain why all the KT3.3 ESTs found in the GenBank data-base are from tumor cells (AI690321, moderately differenti-ated endometrial adenocarcinoma; AW167075, uterus serouspapillary carcinoma; AI739096, colon tumor; AI968607,pooled germ cell tumors; and AW073155 and AI097455, glio-blastoma). A possible linkage of KT3.2 or KT3.3 to any ofthese diseases is worth further investigation.

This research was supported by National Institute of Neurological Disordersand Stroke Grants NS-30989 and NS-35215 to B. Rudy and by American HeartAssociation Grant-in-Aid 9951070T to E. Vega-Saenz de Miera.NOTE ADDED IN PROOF

While this paper was under review, three papers reporting thesequence and functional properties of KT3.2 (TASK-3) in rat, guineapig, and human were published (J Biol Chem275: 8340–9347;J BiolChem275: 16650–16657;Mol Brain Res82: 74–83).

REFERENCES

ALTSCHUL SF, MADDEN TL, SCHAFFER AA, ZHANG J, ZHANG Z, MILLER W,AND LIPMAN DJ. Gapped BLAST and PSI-BLAST: a new generation ofprotein database search programs.Nucleic Acids Res25: 3389–3402, 1997.

BOSSOLASCOM, LEBEL M, LEMIEUX N, AND MES-MASSON AM. The humanTDE gene homologue: localization to 20q13.1–13.3 and variable expressionin human tumor cell lines and tissue.Mol Carcinog26: 189–200, 1999.

CHANDY KG, DOUGLAS J, GUTMAN GA, JAN L, JOHO R, KACZMAREK L,MCKINNON D, NORTH RA, NUMA S. PHILIPSON L, RIBERA AB, RUDY B,SALKOFF L, SWANSON R, STEINER D, TANOUYE M, AND TEMPEL BL. Sim-plified gene nomenclature.Nature352: 26, 1991.

CHANDY KG AND GUTMAN GA. Nomenclature for mammalian potassiumchannel genes.Trends Pharmacol Sci14: 434, 1993.

CHARLIER C, SINGH NA, RYAN SG, LEWIS TB, REUS BE, LEACH RJ, AND

LEPPERTM. A pore mutation in a novel KQT-like potassium channel genein an idiopathic epilepsy family.Nat Genet18: 53–55, 1998.

CHAVEZ RA, GRAY AT, ZHAO BB, KINDLER CH, MAZUREK MJ, MEHTA Y,FORSAYETH JR, AND YOST CS. TWIK-2, a new weak inward rectifyingmember of the tandem pore domain potassium channel family.J Biol Chem274: 7887–7892, 1999.

CHOMCZYNSKI P AND SACCHI N. Single-step method of RNA isolation by acidguanidinium thiocyanate-phenol-chloroform extraction.Anal Biochem162:156–159, 1987.

COETZEEW, AMARILLO Y, CHIU J, CHOW A, LAU D, MCCORMACK T, MORENO

H, NADAL M, OZAITA A, POUNTNEY D, SAGANICH M, VEGA-SAENZ DE MIERA

E, AND RUDY B. Molecular diversity of K1 channels.Ann NY Acad Sci868:233–285, 1999.




Dow

nloaded from


CZEMPINSKI K, ZIMMERMANN S, EHRHARDT T, AND MULLER-ROBER B. Newstructure and function in plant K1 channels: KCO1, an outward rectifierwith a steep Ca21 dependency.EMBO J16: 2565–2575, 1997.

DE ANGELIS PM, CLAUSEN OP, SCHJOLBERGA, AND STOKKE T. Chromosomalgains and losses in primary colorectal carcinomas detected by CGH andtheir associations with tumour DNA ploidy, genotypes and phenotypes.Br JCancer80: 526–535, 1999.

DUPRAT F, LESAGE F, FINK M, REYES R, HEURTEAUX C, AND LAZDUNSKI M.TASK, a human background K1 channel to sense external pH variationsnear physiological pH.EMBO J16: 5464–5471, 1997.

FERAMISCO JR, GLASS DB, AND KREBS EG. Optimal spatial requirements forthe location of basic residues in peptide substrates for the cyclic AMP-dependent protein kinase.J Biol Chem255: 4240–4245, 1980.

FINK M, DUPRAT F, LESAGE F, REYES R, ROMEY G, HEURTEAUX C, AND

LAZDUNSKI M. Cloning, functional expression and brain localization of anovel unconventional outward rectifier K1 channel.EMBO J 15: 6854–6862, 1996.

FINK M, LESAGE F, DUPRAT F, HEURTEAUX C, REYES R, FOSSET M, AND

LAZDUNSKI M. A neuronal two P domain K1 channel stimulated by arachi-donic acid and polyunsaturated fatty acids.EMBO J17: 3297–308, 1998.

FISCHER J, BOUADJAR B, HEILIG R, FIZAMES C, PRUD’HOMME JF, AND WEIS-SENBACH J. Genetic linkage of Meleda disease to chromosome 8qter.Eur JHum Genet6: 542–547, 1998.

FONG GCY, SHAH PU, GEE MN, SERRATOSA JM, CASTROVIEJO IP, KHAN S,RAVAT SH, MANI J, HUANG Y, ZHAO HZ, MEDINA MT, TREIMAN LJ, PINEDA

G, AND DELGADO-ESCUETA AV. Childhood absence epilepsy with tonic-clonic seizures and electroencephalogram 3–4-Hz spike and multispike-slow wave complexes: linkage to chromosome 8q24.Am J Hum Genet63:1117–1129, 1998.

FUKUSHIMA Y, ASANO T, KATAGIRI H, AIHARA M, SAITOH T, ANAI M, FUNAKI

M, OGIHARA T, INUKAI K, MATSUHASHI N, OKA Y, YAZAKI Y, AND SUGANO

K. Interaction between the two signal transduction systems of the histamineH2 receptor: desensitizing and sensitizing effects of histamine stimulationon histamine-dependent cAMP production in Chinese hamster ovary cells.Biochem J320: 27–32, 1996.

GLASS DB, EL-MAGHRABI MR, AND PILKIS SJ. Synthetic peptides correspond-ing to the site phosphorylated in 6-phosphofructo-2-kinase/fructose-2,6-bisphosphatase as substrates of cyclic nucleotide-dependent protein kinases.J Biol Chem261: 2987–2993, 1986.

GOLDSTEIN SA, PRICE LA, ROSENTHAL DN, AND PAUSCH MH. ORK1, apotassium-selective leak channel with two pore domains cloned fromDro-sophila melanogasterby expression inSaccharomyces cerevisiae.Proc NatlAcad Sci USA93: 13256–13261, 1996.

HEGINBOTHAM L, LU Z, ABRAMSON T, AND MACKINNON R. Mutations in theK1 channel signature sequence.Biophys J66: 1061–1067, 1994.

HILLE B. Ionic Channels of Excitable Membranes(2nd ed.). Sunderland, MA:Sinauer, 1992.

IVERSON LE AND RUDY B. The role of the divergent amino and carboxyldomains on the inactivation properties of potassium channels derived fromthe Shaker gene of Drosophila.J Neurosci10: 2903–2916, 1990.

ISOMOTO S, KONDO C, AND KURACHI Y. Inwardly rectifying potassium chan-nels: their molecular heterogeneity and function.Jpn J Physiol47: 11–39,1997.

JAN LY AND JAN YN. Voltage-gated and inwardly rectifying potassium chan-nels.J Physiol (Lond)505: 267–282, 1997.

KALAYDJIEVA L, HALLMAYER J, CHANDLER D, SAVOV A, NIKOLOVA A, AN-GELICHEVA D, KING RHH, ISHPEKOVA B, HONEYMAN K, CALAFELL F,SHMAROV A, PETROVA J, TURNEV I, HRISTOVA A, MOSKOV M, STANCHEVA S,PETKOVA I, BITTLES AH, GEORGIEVA V, MIDDLETON L, AND THOMAS PK.Gene mapping in Gypsies identifies a novel demyelinating neuropathy onchromosome 8q24.Nat Genet14: 214–217, 1996.

KETCHUM KA, JOINER WJ, SELLERSAJ, KACZMAREK LK, AND GOLDSTEIN SAN.A new family of outwardly rectifying potassium channel proteins with twopore domains in tandem.Nature376: 690–695, 1995.

KYTE J AND DOOLITTLE RF. A simple method for displaying the hydropathiccharacter of a protein.J Mol Biol 157: 105–132, 1982.

LEONOUDAKIS D, GRAY AT, WINEGAR BD, KINDLER CH, HARADA M, TAYLOR

DM, CHAVEZ RA, FORSAYETH JR, AND YOST CS. An open rectifier potas-sium channel with two pore domains in tandem cloned from rat cerebellum.J Neurosci18: 868–877, 1998.

LESAGE F, GUILLEMARE E, FINK M, DUPRAT F, LAZDUNSKI M, ROMEY G, AND

BARHANIN J. A pH-sensitive yeast outward rectifier K1 channel with twopore domains and novel gating properties.J Biol Chem271: 4183–4187,1996a.

LESAGE F, GUILLEMARE E, FINK M, DUPRAT F, LAZDUNSKI M, ROMEY G, AND

BARHANIN J. TWIK-1, a ubiquitous human weakly inward rectifying K1

channel with a novel structure.EMBO J15: 1004–1011, 1996b.LESAGE F, REYES R, FINK M, DUPRAT F, GUILLEMARE E, AND LAZDUNSKI M.

Dimerization of TWIK-1 K1 channel subunits via a disulfide bridge.EMBOJ 15: 6400–6407, 1996c.

LEWIS TB, LEACH RJ, WARD K, O’CONNELL P, AND RYAN SG. Geneticheterogeneity in benign familial neonatal convulsions: identification of anew locus on chromosome 8q.Am J Hum Genet53: 670–675, 1993.

LLINAS RR. The intrinsic electrophysiological properties of mammalian neu-rons: insights into central nervous system function.Science242: 1654–1664, 1988.

MAINGRET F, FOSSETM, LESAGEF, LAZDUNSKI M, AND HONOREE. TRAAK isa mammalian neuronal mechano-gated K1 channel.J Biol Chem 274:1381–1387, 1999.

MIKAMI M, YASUDA T, TERAO A, NAKAMURA M, UENO S, TANABE H, TANAKA

T, ONUMA T, GOTO Y, KANEKO S, AND SANO A. Localization of a gene forbenign adult familial myoclonic epilepsy to chromosome 8q23.3-q24.1.Am J Hum Genet65: 745–751, 1999.

MILETICH JPAND BROZE GJ. Beta protein C is not glycosylated at asparagine329. The rate of translation may influence the frequency of usage atasparagine-X-cysteine sites.J Biol Chem265: 11397–11404, 1990.

MILLAR JA, BARRATT L, SOUTHAN AP, PAGE KM, FYFFE RE, ROBERTSONB,AND MATHIE A. A functional role for the two-pore domain potassiumchannel TASK-1 in cerebellar granule neurons.Proc Natl Acad Sci USA97:3614–3618, 2000.

PALMEDO G, FISCHER J, AND KOVACS G. Duplications of DNA sequencesbetween loci D20S478 and D20S206 at 20q11.2 and between loci D20S902and D20S480 at 20q13.2 mark new tumor genes in papillary renal cellcarcinoma.Lab Invest79: 311–316, 1999.

PATEL AJ, HONORE E, MAINGRET F, LESAGE F, FINK M, DUPRAT F, AND

LAZDUNSKI M. A mammalian two pore domain mechano-gated S-like K1

channel.EMBO J17: 4283–4290, 1998.PATEL AJ, HONORE E, LESAGE F, FINK M, ROMEY G, AND LAZDUNSKI M.

Inhalational anesthetics activate two-pore-domain background K1 channels.Nat Neurosci2: 422–426, 1999.

POUNTNEY DJ, GULKAROV I, VEGA-SAENZ DE MIERA E, HOLMES D, SAGANICH

M, RUDY B, ARTMAN M, AND COETZEE WA. Identification and cloning ofTWIK-originated similarity sequence (TOSS): a novel human 2-pore K1

channel principal subunit.FEBS Lett450: 191–196, 1999.REYES R, DUPRAT F, LESAGE F, FINK M, SALINAS M, FARMAN N, AND

LAZDUNSKI M. Cloning and expression of a novel pH-sensitive two poredomain K1 channel from human kidney.J Biol Chem273: 30863–30869,1998.

RUDY B. Diversity and ubiquity of K channels.Neuroscience25: 729–749,1988.

RUDY B, HOGER JH, LESTER HA, AND DAVIDSON N. At least two mRNAspecies contribute to the properties of rat brain A-type potassium channelsexpressed inXenopusoocytes.Neuron1: 649–658, 1988.

SAITOU N AND NEI M. The neighbor-joining method: a new method forreconstructing phylogenetic trees.Mol Biol Evol 4: 406–425, 1987.

SALINAS M, REYES R, LESAGE F, FOSSET M, HEURTEAUX C, ROMEY G, AND

LAZDUNSKI M. Cloning of a new mouse two-P domain channel subunit anda human homologue with a unique pore structure.J Biol Chem 274:11751–11760, 1999.

SINGH N, CHARLIER C, PEIFFER A, STAUFFER D, MELIS R, ENOCH MA, GOLD-MAN D, ROGERSS, ANDERSONVE, AND LEPPERTM. Genes for rare idiopathicgeneralized epilepsies: BFNC.Adv Neurol79: 341–350, 1999.

STEPHENS RM AND SCHNEIDER TD. Features of spliceosome evolution andfunction inferred from an analysis of the information at human splice sites.J Mol Biol 228: 1124–1136, 1992.

TALLEY EM, LEI Q, SIROIS JE,AND BAYLISS DA. TASK-1, a two-pore domainK1 channel, is modulated by multiple neurotransmitters in motoneurons.Neuron25: 399–410, 2000.

TANNER MM, TIRKKONEN M, KALLIONIEMI A, HOLLI K, COLLINS C, KOWBEL D,GRAY JW, KALLIONIEMI OP, AND ISOLA J. Amplification of chromosomalregion 20q13 in invasive breast cancer: prognostic implications.Clin CancerRes1: 1455–1461, 1995.

THOMPSON JD, HIGGINS DG, AND GIBSON TJ. CLUSTAL W: improving thesensitivity of progressive multiple sequence alignment through sequenceweighting, positions-specific gap penalties and weight matrix choice.Nu-cleic Acids Res22: 4673–4680, 1994.




Dow

nloaded from


VEGA-SAENZ DE MIERA EC AND LIN JW. Cloning of ion channel gene familiesusing the polymerase chain reaction.Methods Enzymol207: 613–619, 1992.

VEGA-SAENZ DE MIERA E, RUDY B, SUGIMORI M, AND LLINAS R. Molecularcharacterization of the sodium channel subunits expressed in mammaliancerebellar Purkinje cells.Proc Natl Acad Sci USA94: 7059–7064, 1997.

WANG ZW, KUNKEL MT, WEI A, BUTLER A, AND SALKOFF L. Genomicorganization of nematode 4TM K1 channels.Ann NY Acad Sci868: 286–303, 1999.

WEI A, JEGLA T, AND SALKOFF L. Eight potassium channel families re-

vealed by theC. elegansgenome project.Neuropharmacology35: 805–829, 1996.

WOODGETT JR, GOULD KL, AND HUNTER T. Substrate specificity of proteinkinase C. Use of synthetic peptides corresponding to physiological sites asprobes for substrate recognition requirements.Eur J Biochem161: 177–184,1986.

ZARA F, BIANCHI A, AVANZINI G, DI DONATO S, CASTELLOTTI B, PATEL PI, AND

PANDOLFO M. Mapping of genes predisposing to idiopathic generalizedepilepsy.Hum Mol Genet4: 1201–1207, 1995.




Dow

nloaded from


KT3.2 and KT3.3, Two Novel Human Two-Pore K Channels Closely ...

Documents

Transcript of KT3.2 and KT3.3, Two Novel Human Two-Pore K Channels Closely ...