insect biochemistry and molecular biologykedonglab.ent.msu.edu/publications/zhang et al.,...

lable at ScienceDirect

Insect Biochemistry and Molecular Biology xxx (2011) 1e8

Contents lists avai

Insect Biochemistry and Molecular Biology

journal homepage: www.elsevier .com/locate/ ibmb

Molecular characterization and functional expression of the DSC1 channel

Tianxiang Zhang a, Zhiqi Liu b,1, Weizhong Song b, Yuzhe Du b, Ke Dong b,*

aDepartment of Microbiology and Molecular Genetics, Michigan State University, East Lansing, MI 48824, USAbDepartment of Entomology, Neuroscience and Genetics Programs, Michigan State University, East Lansing, MI 48824, USA

a r t i c l e i n f o

Article history:Received 9 December 2010Received in revised form22 April 2011Accepted 26 April 2011

Keywords:Drosophila Sodium Channel 1RNA editingAlternative splicingVoltage-gated ion channel

* Corresponding author. 438 Giltner Hall, MichiganMI 48824, USA. Tel.: þ1 517 432 2034; fax: þ1 517 35

E-mail address: [email protected] (K. Dong).1 Present address: ChanTest, Cleveland, OH 44128,

0965-1748/$ e see front matter � 2011 Elsevier Ltd.doi:10.1016/j.ibmb.2011.04.010

Please cite this article in press as: Zhang, T., eand Molecular Biology (2011), doi:10.1016/j.

a b s t r a c t

Drosophila Sodium Channel 1 (DSC1) was predicted to encode a sodium channel based on a high sequencesimilarity with vertebrate and invertebrate sodium channel genes. However, BSC1, a DSC1 ortholog inBlattella germanica, was recently shown to encode a cation channel with ion selectivity toward Ca2þ. Inthis study, we isolated a total of 20 full-length cDNA clones that cover the entire coding region of theDSC1 gene from adults of Drosophila melanogaster by reverse transcription-polymerase chain reaction.Sequence analysis of the 20 clones revealed nine optional exons, four of which contain in-frame stopcodons; and 13 potential A-to-I RNA editing sites. The 20 clones can be grouped into eight splice typesand represent 20 different transcripts because of unique RNA editing. Three variants generated DSC1currents when expressed in Xenopus oocytes. Like the BSC1 channel, all three functional DSC1 channelsare permeable to Ca2þ and Ba2þ, and also to Naþ in the absence of external Ca2þ. Furthermore, the DSC1channel is insensitive to tetrodotoxin, a potent and specific sodium channel blocker. Our study showsthat DSC1 encodes a voltage-gated cation channel similar to the BSC1 channel in B. germanica. Extensivealternative splicing and RNA editing of the DSC1 transcripts suggest the molecular and functionaldiversity of the DSC1 channel.

� 2011 Elsevier Ltd. All rights reserved.

1. Introduction

Voltage-gated ion channels are a diverse group of integral trans-membrane proteins that are essential for electrical signaling inneurons, muscle and endocrine cells. In response to changes inmembrane potential, these channels mediate rapid ion flux throughahighly selectivepore. Voltage-gated sodiumchannels are responsiblefor inwardmovement of sodium ions during electrical signaling in cellmembranes. Sodium channels consist of a large a-subunit and a vari-able number of smaller subunits (Yu et al., 2005). The a-subunit formsthe ion conducting pore while the associated subunits modulatechannel expressionandgating. Thepore-forminga-subunit consists offour repeatedhomologous domains (IeIV), eachhaving sixmembranespanning segments (S1eS6). The S1eS4 segments serve as the voltagesensing module and the S5 and S6 segments and their connecting Ploop serve as a pore-forming module (Catterall, 2000).

Analysis of the genome of Drosophila melanogaster revealed onlytwo sodium channel-like sequences, para, and DSC1 (DrosophilaSodium Channel 1) (Littleton and Ganetzky, 2000). The para genewas

State University, East Lansing,3 4354.

USA.

All rights reserved.

t al., Molecular characterizatioibmb.2011.04.010

isolated based on temperature-sensitive paralysis phenotype; andencodes a protein having all of the structural hallmarks of voltage-gated sodium channel a subunits and exhibiting approximately 50%overall amino acid sequence identity to vertebrate sodium channela subunits (Loughney et al.,1989). It is well-established that para andpara-orthologs in other insect species encode functional sodiumchannels (Feng et al., 1995; Warmke et al., 1997; Soderlund, 2005;Dong, 2007). Extensive alternative splicing and RNA editing of thetranscripts of para and para-orthologs generate molecular and func-tional diversity of sodium channels in insects (Loughney et al., 1989;Thackeray and Ganetzky, 1994; O’Dowd et al., 1995; Thackeray andGanetzky, 1995; Park et al., 1999; Tan et al., 2002; Liu et al., 2004;Song et al., 2004; Dong, 2007; Olson et al., 2008; Lin et al., 2009).

DSC1 was first discovered two decades ago by probingaDrosophila genomic DNA library with an eel sodium channel cDNA(Salkoff et al., 1987). It was predicted to encode a sodium channelbased on its overall similarity in deduced amino acid sequence anddomain organization with eel and mammalian sodium channels(Salkoff et al., 1987; Littleton and Ganetzky, 2000). Sodium currentsin embryonic neurons from homozygotes for a chromosome defi-ciency at 60E5/6, where the DSC1 gene is located, were similar tothose from wild-type neurons, indicating that DSC1 is not theprimary sodium channel gene in embryonic neurons (Germeraadet al., 1992). A P-element insertion in the second intron of DSC1in a smell-impaired (smi) mutant, smi60E, reduced DSC1 transcript

n and functional expression of the DSC1 channel, Insect Biochemistry

mailto:[email protected]

www.sciencedirect.com/science/journal/09651748

http://www.elsevier.com/locate/ibmb

http://dx.doi.org/10.1016/j.ibmb.2011.04.010



Fig. 1. Alignment of amino acid sequences of DSC1-1.1, BSC1 and Para proteins. Dots represent amino acid residues in BSC1 and Para that are identical to those in DSC1-1.1. Dashesindicate gaps introduced to maximize sequence alignment. The four homologous domains (IeIV) and six transmembrane segments (S1eS6) in each domain are marked above thesequences. Optional exons 17B and 19Awhich are included in DSC1-1.1 are highlighted. Locations of other optional exons (which are not included in DSC1-1.1) are marked with solidarrowheads. Empty arrowheads indicate the position of introns. Amino acid residues in DSC1-1.1 and BSC1 corresponding to the MFM motif in Para (i.e., IFM motif in mammaliansodium channels) that is critical for fast inactivation are boxed. The amino acid residues which are important for ion selectivity and TTX binding are indicated with asterisks. TheGenBank accession numbers are DQ466888.1, AF312365.1 and M32078.1 for DSC1-1.1, BSC1 and Para, respectively.

T. Zhang et al. / Insect Biochemistry and Molecular Biology xxx (2011) 1e82

level by two-fold and results in a three-fold decrease in olfactoryresponse to benzaldehyde (Kulkarni et al., 2002), suggesting thatthe DSC1 channel plays an important role in olfaction. Comparativestudies revealed different spatial and temporal expression patterns

Please cite this article in press as: Zhang, T., et al., Molecular characterizatioand Molecular Biology (2011), doi:10.1016/j.ibmb.2011.04.010

of DSC1 and para transcripts and proteins, particularly in the adultperipheral nervous system, suggesting different functions of DSC1and Para channels in the nervous system. For example, in adults,expression patterns of DSC1 and para transcripts overlap in the CNS


Fig. 1. (continued).

T. Zhang et al. / Insect Biochemistry and Molecular Biology xxx (2011) 1e8 3

based on the RNA in situ hybridization study (Hong and Ganetzky,1994). However, an immunohistochemical study indicated thatDSC1 channels were only distributed in synaptic regions and axonaltracts while Para channels were also localized on cell bodies in thePNS. Onlyweak DSC1 signal could be detected from compound eyeswhere the expression level of Para channels is high. The denselocalization of the DSC1 protein was detected in the nerve endingsof motor neurons in the thorax, leg muscles and also in theproboscis, tibia and tarsi, where the mechanoreceptors or chemo-sensory organs are located (Castella et al., 2001).


No functional analysis of the DSC1 channel has been reported.However, a DSC1 ortholog, BSC1 from Blattella germanica, wasconfirmed to encode a novel Ca2þ-selective cation channel, nota sodium channel, when expressed in Xenopus oocytes (Zhou et al.,2004). The first complete cDNA sequence of DSC1 was reportedbased on the genomic sequence of a 20 kb genomic regionwhere theDSC1 gene resides; and the RT-PCR analysis of partial cDNAs(Kulkarni et al., 2002). Subsequently, with the completion ofDrosophila genome sequencing, another cDNA sequence of DSC1(NM_166696.2) was deduced from an annotated genomic sequence


Table 1Potential RNA editing events

Amino acidchange

Nucleotidechange

Position DSC1 Clones

N186S aeg IS3 2.2, 8.1W280R tec IS5wIS6 3.1, 3.2V449A tec Linker (IeII) 1.4, 7.2Q601R aeg Linker (IeII) 1.1,1.3, 2.5Q1020R aeg Linker (IIeIII) 5.1, 6.1S1575 G aeg Linker (IIeIII) 5.1, 2.3H1605R aeg Linker (IIeIII) 4.2, 2.3, 5.1, 2.1Q1687R aeg Linker (IIeIII) 1.5, 5.1, 2.2, 2.3, 8.1Y1690C aeg Linker (IIeIII) 2.1, 1.5, 5.1, 2.3, 8.1, 4.2T1752A aeg IIIS1 2.1, 1.5, 5.1, 2.2, 2.3, 8.1, 4.2I1768 V aeg IIIS1 2.1, 1.5, 5.1, 2.2, 2.3, 4.2T1782A aeg IIIS1eIIIS2 1.1, 2.1Q1864R aeg IIIS4eIIIS5 2.1,1.5M2027 V aeg Linker (IIIeV) 1.1, 1.2, 2.1, 3.1, 5.1, 4.1, 6.1, 1.6, 4.2L2163S tec IVS3eIVS4 7.1, 7.2N2388S aeg C terminus 1.2, 2.2, 1.6P2402S cet C terminus 1.2, 1.4, 1.5, 2.3, 1.6, 3.2


(NT_033778.3). However, there were substantial sequence discrep-ancies between the two reported cDNA sequences. To elucidate themolecular and functional characterization of DSC1, we isolated 20full-length cDNA clones covering the entire coding region of theDSC1gene fromadults ofD.melanogaster. Sequence analysis revealedthat DSC1 transcripts undergo extensive alternative splicing andpotential RNA editing. Functional expression in Xenopus oocytesconfirmed that, like BSC1, DSC1 encoded a Ca2þ-selective cationchannel.

2. Materials and methods

2.1. Amplification and cloning of the coding region of the DSC1 geneby reverse transcription-polymerase chain reaction (RT-PCR)

Five to seven day-oldw1118 adults were used to isolate total RNA.First-strand cDNA was synthesized using Oligo dT primers (Invi-trogen, Carlsbad, CA), and the SuperScript� II reverse transcriptionkit (Invitrogen, Carlsbad, CA). Conditions for the first-strand cDNAsynthesis reaction were: 42 �C for 2 min followed by a 60 minincubation at 48 �C. RNAwas removed by a 20 min incubation withRNaseH at 37 �C. The PCR primers for amplification of the cDNAwere, 50-CCGCTCGAGGCCACCATGGGTGATGATCAAGCGACG-30, and50-TACCCCTAGGAAATATCAGATAGAAAGTTC-30. To improve theexpression efficiency, a Kozak sequence (GCCACCATGGGT) wasadded to the forward primer and the first nucleotide of the secondcodon was substituted from A to G to meet the sequence require-ment. The Xho I and Avr II restriction enzyme sites (underlined)were added to the forward and reverse primer, respectively, tofacilitate the cloning of the PCR product into a modified pGH19vector in which an Xho I restriction site was removed by site-directed mutagenesis. The original pGH19 (a Xenopus expressionvector containing the 30 and 50 untranslated regions of the Xenopusb-globin gene at the 30 end of T7 promoter and 50 end of SP6promoter, respectively) was kindly provided by Barry Ganetzky,University of Wisconsin, Madison, WI. The PCR reaction conditionswere the same as previously described (Zhou et al., 2004). Thereaction mixture (50 ml) contained 0.5 ml cDNA, 50 pmol eachprimer, 200 mM each dNTP, 1 U eLONGase (Invitrogen), 1.5 mMMgCl2, and 1�PCR reaction buffer. PCR was carried out as follows:one cycle of 94 �C for 1 min; 33 cycles of 94 �C for 30 s, 58 �C for30 s, 68 �C for 8 min; and one cycle of 68 �C for 15 min. The PCRproducts were purified using the QIAEX II Gel Extraction Kit�

(QIAGEN Sciences, MD) and cloned into the modified pGH19. MAXEfficiency Stbl2-competent cells (Invitrogen) were used as hostcells.

The inserts of DSC1 cDNA clones were sequenced by moderatethroughput sequencing (ABI 3730xl, Research Technology SupportFacility, Michigan State University). Sequence data were analyzedusing software Lasergene 6� (DNASTAR, Inc., Madison, WI).

2.2. Expression of DSC1 channels in Xenopus oocytes

Procedures for oocyte preparation and cRNA injection weresimilar to those described by Tan et al. (Tan et al., 2002). Oocyteswere obtained surgically from healthy female Xenopus laevis(Xenopus I, Ann Arbor, MI) and incubated with 1 mg/ml type IAcollagenase (Sigma Inc. St. Louis, MO) in a Ca2þ-free ND-96medium, which contains 96 mM NaCl, 2 mM KCl, 1 mM MgCl2,and 5 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid(HEPES), pH 7.5. Follicle cells remaining on the surface of oocyteswere removed manually with forceps. Isolated oocytes were incu-bated in ND-96 culture medium containing 1.8 mM CaCl2 supple-mented with 50 mg/ml gentamicin, 5 mM pyruvate, and 0.5 mMtheophylline (Goldin, 1992). Healthy stage VeVI oocytes were used


for cRNA injection. To prepare DSC1 cRNA, plasmid DNA containingfull-length DSC1 cDNA was linearized with NotI (New EnglandBiolab, Inc., Ipswich, MA), which does not cut the insert, followedby in vitro transcription with T7 polymerase using the mMESSAGEmMACHINE Kit (Ambion, Inc., Austin, TX). Each oocyte was injectedwith 15 ng DSC1 cRNA. After injection, oocytes were incubated inND-96 culture solution containing 96 mM NaCl, 2 mM KCl, 1.8 mMCaCl2, 1 mM MgCl2, 5 mM HEPES, 2.5 mM Na-pyruvate, 0.5 mMTheophyline for 2e3 days before recording.

2.3. Electrophysiological recording

DSC1 currents were recorded using standard two-electrodevoltage clamp technique. The borosilicate glass electrodes werefilled with filtered 3 M KCl in 0.5% agarose and had resistance at4e6 MU. Currents were measured using the oocyte clamp instru-ment OC725C (Warner Instrument Corp., Hamden, CT), Digidata1200A interface (Axon Instrument, Inc., Foster City, CA), andpCLAMP 8.2 software (Axon Instrument, Inc., Foster City, CA).External recording solutions containing only Naþ, Ca2þ or Ba2þ

were used to record currents in oocytes expressing the DSC1channel as described in Zhou et al. (2004). The Naþ recordingsolution consisted of 50 mM NaOH, 45 mM TEAOH, 10 mM HEDTA,and 10 mM HEPES. The Ca2þ solution consisted of 50 mM Ca(OH)2,55 mM TEAOH, and 10 mM HEPES. The Ba2þ solution consisted of50 mM Ba(OH)2, 55 mM TEAOH, and 5 mM HEPES. All recordingsolutions were adjusted to pH 7.0 with methanesulfonic acid. Forthe relative permeability, the recording protocol consisted ofa 15-ms depolarization from �100 mV to 80 mV followed byrepolarization to �100 mV (Zhou et al., 2004). All experimentswere performed at room temperature (22e25 �C). For the appli-cation of tetrodotoxin (TTX), the disposable perfusion systemdeveloped by Tatebayashi and Narahashi (1994) was used. The TTXworking solutions were diluted from a stock solution (2 mM) usingBa2þ external recording solution.

3. Results

3.1. Comparison of deduced amino acid sequences among DSC1,BSC1 and Para proteins

A total of 20 DSC1 cDNA clones that cover the entire codingregion of the DSC1 gene were isolated by RT-PCR using total RNAisolated from adults of D. melanogaster. The nucleotide sequence of


A

B

C

Fig. 2. Molecular characterization of 20 DSC1 cDNA clones. A. The genomic organization of the coding region of the DSC1 gene deduced from the sequence of DSC1-1.1 (GenBank ID:DQ466888.1). Solid boxes represent exons. Positions of the alternative exons are labeled with arrows. B. Schematic drawing of the topology of the DSC1 protein indicating locationsof alternative exons. All alternative exons are optional. The asterisks indicate stop codons. C. Exon usage of 20 DSC1 cDNA clones. Splice variants are named according to the splicetypes. Variants of each splice type are different due to scattered amino acid differences. For example, the two variants in splice type 3 are designated as DSC1-3.1 and DSC1-3.2.


one of the clones, DSC1-1.1, has been deposited in GenBank (Gen-Bank ID: DQ466888.1) and its deduced amino acid sequence ispresented in Fig. 1 in comparison with those of BSC1 and Para. Theoverall topology of the DSC1 channel is similar to that of BSC1 andPara channels: four homologous domains each having six trans-membrane segments. The DSC1-1.1 protein shares 52% identity inthe overall sequence with BSC1, and 29% with Para. The highestsequence identities are found in the transmembrane domainswith 83e90% between DSC1-1.1 and BSC1 proteins, and 48e54%between DSC1-1.1 and Para proteins. The sequence identities of thefirst intracellular linker connecting domains I and II and the secondlinker connecting domains II and III are less than 30% betweenDSC1-1.1 and BSC1 proteins. The second linker of the DSC1-1.1channel is much longer (838 amino acids) than those of BSC1 andPara, which have 330 and 257 amino acids, respectively. The linkerconnecting domains III and IV is highly conserved between DSC1-


1.1 and BSC1 (90% identity) or Para (52% identity). It is well-established that this linker, particularly the IFM (mammals) orMFM (insects) motif in the middle of the linker, is critical for fastinactivation of sodium channels (Catterall, 2000). Corresponding tothe MFMmotif in Para, the DSC1-1.1 and BSC1 channels have “VFL”and “MFL,” respectively. The “V” in the VFL motif is the result of anA-to-I editing causing an M to V change (Hoopengardner et al.,2003; and also in our study, see Table 1). Therefore, the conven-tional motif in the DSC1 channel is also “MFL” like in the BSC1channel.

The ion selectivity of voltage-gated sodium channels is determinedby the amino acids D, E, K, and A in the pore positions of domains I, II,III, and IV, respectively (i.e., the selectivity-filter motif “DEKA”)(Catterall, 2000). However, both the DSC1 and BSC1 channels containthe DEEA motif instead of the DEKA motif. The positively chargedresidues in thevoltage sensor S4of eachdomain that are critical for the


A15 ms80 mV

-100 mV

Ba2+

B

Ba Ca2+

Na+5 ms

2µA

60

80

100

**Cur

rent

(%)

Ba2+

Ca2+

Na+

*

0

20

40

**

*

DSC1-2.2DSC1-1.2

Nor

mal

ized

DSC1-1.1

C

Control TTX5 ms

2µ A

Fig. 3. Functional expression of three DSC1 variants in Xenopus oocytes. A. DSC1currents recorded from Xenopus oocytes expressing DSC1-1.1 channels. Recordingprotocol consisted of a 15-ms depolarization from �100 mV to 80 mV followed byrepolarization to �100 mV. Superimposed Naþ, Ca2þ and Ba2þ currents recorded froman oocyte two days after injection with 15 ng DSC1-1.1 cRNA. B. DSC1 variants are morepermeable to Ca2þ and Ba2þ than to Naþ. Tail currents were normalized to the maximalcurrent from recording in Ba2þ recording solution. The data represent the mean� SDfor at least eight oocytes. The asterisks indicate significant differences from the Ba2þ

current (p< 0.05). C. The DSC1-1.1 channel is insensitive to TTX. Superimposed DSC1-1.1 currents recorded using Ba2þ recording solution from an oocyte two days afterinjection with 15 ng DSC1-1.1 cRNA before and after exposure to 10 mM TTX using theprotocol illustrated in A.


gating of sodium channels (Catterall, 2000). The number of the posi-tively charged residues in IS4 and IVS4 is conserved among Para andDSC1/BSC1 (Fig. 1). However, DSC1 and BSC1 channels have an extrapositively charged residue in IIIS4 compared to the Para channel.Furthermore, DSC1-1.1 has four positively charged residues in IIS4,whereas five in BSC1 and Para channels. These two differences areconserved in all 20 DSC1 variants.

3.2. Sequence comparison of 20 DSC1 cDNA clones reveals extensivealternative splicing of DSC1 transcripts

Sequence comparison of the 20 cDNA clones with the DSC1genomic sequence (NT_033778.3) revealed 20 constitutive exons,exons 1e20 (Fig. 2A); and nine optional exons, 4A, 11A, 11B, 11C,16A, 17A, 17B, 17C, and 19A (named after the constitutive exonsupstream) (Fig. 2A and B). Most of the optional exons have theconsensus GT and AG sequences at the 50 donor and 30 acceptorsites, respectively, except for 11A and 11B which do not have the GTsequence. The size of the optional exons ranges from 15 bp to861 bp. Exon 11B is 15 bp (50-ATTTTCAAGAAGAAG-300) long. Thenucleotide sequences of other longer optional exons are depositedin GenBank (GenBank ID: HM 348600.1-348607.1). Exon 4Aencodes part of the loop connecting S5 and S6 in domain I. Exons11A, 11B, and 11C are three tandem exons that encode part of thelinker connecting domains II and III. Exons 17A, 17B, and 17C areanother group of tandem exons encoding part of the linker con-necting domains III and IV as well as part of IVS1. Exon 16A encodespart of the loop connecting S5 and S6 in domain III. Exon 19Aencodes a sequence in the C-terminal region. Intriguingly, exons 4A,11A, 16A, and 17A contain in-frame stop codons (Fig. 2B). Inclusionof these exons in DSC1 transcripts creates truncated proteins.

Based on optional exon usage, the 20 cDNA clones can begrouped into eight unique splice types, DSC1-1 to DSC1-8 (Fig. 2C).The six cDNA clones in splice type 1 are designated as DSC1-1.1 toDSC1-1.6 because each clone possesses several unique amino acidchanges (see below). For the same reason, the five clones in splicetype 2 are named DSC1-2.1 to DSC1-2.5. Two optional exons, 17Band 19A, are present in both splice type 1 and 2 variants (Fig. 2C).Splice type 2 variants also have a third optional exon, 11B. A total ofseven cDNA clones in splice types 3, 4, 5, 6 and 9 are predicted toencode truncated proteins because of inclusion of stop codon-containing alternative exons (Fig. 2B and C). DSC1-3 and DSC1-4variants would lack domain IV and the C-terminus. DSC1-5,DSC1-6 and DSC1-9 variants contain only the N-terminus plusdomain I.

3.3. Identification of potential RNA editing

Comparing the deduced amino acid sequences of the 20 variantswith the cDNA sequence (NM_166696.2) deduced fromanannotatedgenomic sequence (NT_033778.3), we found 9e27 scattered aminoacid changes in individual variants due to nucleotide changes.Because we isolated these full-length clones by RT-PCR, some ofnucleotide changes could be the results of PCRmistakes. However, ifa nucleotide change occurs in two or more variants that belong todifferent splice types, such a change is likely due to RNA editing, nota PCRmistake.13 A to G changes, three T to C changes and one C to Tchangewere detected inmore than one variant, resulting in a total of17 amino acid changes. The rest of the nucleotide changes werefound only in individual variants. Therefore, the 13 amino acidchanges are likely causedbyA-to-I RNAediting and the other four arelikely due to U-to-C or C-to U editing (Table 1). Most of the editingsites are in the linkers connecting domains or the short loops con-necting segments. One A-to-I editing resulting inM2027V change inthe linker connecting domains III and IV was previously reported in


DSC1 transcripts (Hoopengardner et al., 2003).We found this editingevent in nine variants including DSC1-1.1.

Nucleotide substitutions introduced premature stop codons infour clones, two of which also contain stop codon-containing exons.The DSC1-7.1 variant contains an insertion of a 19 bp sequence



(50-CTCAACGTAAAAAAACTAA-30) at the N-terminus. Surprisingly,the 19-bp sequence was part of exon 6 which encodes part ofthe linker connecting domains I and II. Somehow the 19 bpsequence was duplicated into the N-terminus. This insertioncauses a frame shift in the downstream sequence, resulting ina premature stop codon. Therefore, a total of 10 variants containpremature stop codons. In addition, a six nucleotide sequence,CCGTCT, was found in all 20 cDNA clones, but not in the genomicDNA sequence (NT_033778.3). This difference might be caused bysequence polymorphisms between fly lines used in the genomesequence project (y; cn bw sp; þ; þ) and our study (w1118).

3.4. Three DSC1 channel variants are functionally expressedin Xenopus oocytes

To determine whether the DSC1 channel is also a cation channellike the BSC1 channel (Zhou et al., 2004), we examined the 20 DSC1variants in Xenopus oocytes using two-electrode voltage clamp. Torecord DSC1 currents, we adopted the protocol previously used forthe BSC1 channel (Zhou et al., 2004). The membrane potential wasdepolarized to 80 mV for 15 ms from a holding potential of�100 mVand then repolarized back to �100 mV (Fig. 3A). A small outwardcurrent during depolarization followed by a large tail current asso-ciated with repolarization was recorded from oocytes expressingDSC1-1.1 in Ba2þ recording solution (Fig. 3A). Smaller outward andtail currents were detected in Ca2þ recording solution; and a largeroutward current and a much smaller tail current were detected inNaþ recording solution (Fig. 3A). The larger outward current duringthe depolarization in the Naþ solution is most likely caused by theincreased driving force resulting from the relatively high internal Kþ.This phenomenonwas also observed in the studyof the BSC1 channel(see Fig. 3A in Zhou et al., 2004). To facilitate the comparisonbetweencurrents fromdifferent recording solutions, the tail currents recordedfrom each oocyte in Ca2þ recording solution or Naþ recording solu-tion were normalized to the tail current obtained in Ba2þ recordingsolution from the same oocyte (Fig. 3B). These results indicate thatlike the BSC1 channel, the DSC1-1.1 channel is more permeable toCa2þ and Ba2þ, and also to Naþ in the absence of external Ca2þ. Asexpected, none of the 10 variants that encode truncated proteinsproduced any currents using this recording protocol. However, onlytwo other variants, DSC1-1.2 and DSC1-2.2, are also functionalgenerating currents similar to those from the DSC1-1.1 channel(Fig. 3B). We did not observe any detectable currents from otherseven DSC1 clones that do not contain any premature stop codon.

The sodium channel blocker TTX inhibits insect sodium chan-nels at nanomolar concentrations (Tan et al., 2002; Du et al., 2009).To determine whether TTX also inhibits DSC1 channels, we recor-ded DSC1 currents in Ba2þ recording solution in the presence of TTX(0.1 mM, 1 mM and 10 mM) using the recording protocol describedabove. Even at 10 mM, TTX did not reduce the amplitude of eitherthe outward current or the tail current, indicating that DSC1-1.1channel is TTX-insensitive (Fig. 3C).

4. Discussion

In this study, sequencing analysis of 20 cDNA clones revealsextensive alternative splicing and potential RNA editing of DSC1transcripts. A total of nine optional exons were identified, four ofwhich contain in-frame stop codons. It is striking that seven of 20clones are predicted to encode truncated proteins due to theinclusion of stop codon-containing optional exons. Interestingly,like the stop codon-containing exon 11A located in the linkerconnecting domains II and III in DSC1, two BSC1 optional exons inthe linker connecting domains II and III also contain a prematurestop codon (Liu et al., 2001). Transcripts containing these BSC1


optional exons were detected only in leg muscle, suggestinga possibly tissue-specific role of these truncated proteins (Liu et al.,2001). The conservation of this splicing site in two phylogeneticallydistant species raises the possibility that two-domain truncatedproteins might play a role in insect neurophysiology. Prematurestop codon-containing alternative exons are also detected involtage-gated sodium and calcium channel genes, although the roleof these truncated proteins is unclear. For example, transcriptscontaining two mutually exclusive exons, 18A and 18N, encodingIIIS3-S4, are found in mouse Nav1.6 (Plummer et al., 1997). Exon18N contains a stop codon and exon 18N-containing transcripts aredetected in fetal brain and non-neuronal cells (Plummer et al.,1997). Moreover, this stop codon-containing exon is conserved insodium channel genes of human, pufferfish (Plummer et al., 1997),and the German cockroach (Tan et al., 2002). Other sodium channelgenes also produce premature stop codon-containing alternativeexons, such as exon 17A in Nav1.2 and Nav1.3, exon 17B in Nav1.3,and exon 16A in Nav1.7 in mouse and human (Kerr et al., 2008). Likeexon 18N, the inclusion of exon 17A is tissue- or developmentalstage-specific (Kerr et al., 2008).

Only three of the 20 clones produced DSC1 currents. While it isnot surprising that the 10 truncated DSC1 proteins did not produceany current, it is not clear why seven clones that contain nopremature stop codon did not produce currents. We previouslyencountered a similar problem with functional expression of BSC1variants in Xenopus oocytes. While one BSC1 variant gave robustcurrents (Zhou et al., 2004), no currents were detected from othertwo BSC1 clones that do not contain premature stop codons (Chungand Dong, unpublished data). It is likely that unidentified accessorysubunits or other chaperone proteins are required for the expres-sion of these variants in Xenopus oocytes.

The functional DSC1 variants are permeable to Naþ, Ca2þ, andBa2þ, which is similar to those recorded from BSC1 channels (Zhouet al., 2004). Like BSC1, DSC1 encodes a voltage-gated cationchannel that is more permeable to bivalent cations. It is well-established that the ion selectivity of voltage-gated sodium chan-nels is determined by the amino acids D, E, K, and A in the porepositions of domains I, II, III, and IV, respectively (i.e., theselectivity-filter motif “DEKA”) (Catterall, 2000). In contrast, fourglutamic acids (EEEE) at the corresponding positions determine theion selectivity for voltage-gated calcium channels (Catterall, 2000).It has been shown that a KeE substitution in the DEKA motif ina mammalian sodium channel altered the sodium selectivity tomore toward calcium (Heinemann et al., 1992). Both the DSC1 andBSC1 channels contain the DEEA motif instead of DEKA or EEEEmotifs. Substitution of the second E with K in the DEEA motifreduced the ion selectivity of the BSC1 channel for Ba2þ, demon-strating its critical role in modulating ion selectivity of the BSC1channel (Zhou et al., 2004). It is likely that such a substitution willhave a similar effect on the ion selectivity of the DSC1 channel.

It is well recognized that the four S4 voltage sensors in sodiumchannels, rich in positively charged amino acids, detect membranedepolarization and move outward initiating channel activation(Catterall, 2000). Most of the positively charged residues are alsoconserved in DSC1 and BSC1 channels with two minor variations(Fig. 1). First, DSC1 and BSC1 channels have an extra positivelycharged residue in IIIS4 compared to the Para channel. Whether thedifference contributes to the differences in gating propertiesbetween Para and BSC1 channels (Zhou et al., 2004) remains to bedetermined. Second, DSC1-1 has four positively charged residues inIIS4, whereas five in BSC1 and Para channels, suggesting potentialdifferences in voltage sensing betweenBSC1 andDSC1 channels.Weare currently investigating gating properties of the DSC1 variants;and possible contributions of these sequence variations in voltagesensors in BSC1 and DSC1 channel gating.



In sodium channels, the selectivity-filter motif “DEKA” and theouter-ring motif “EEMD” (mammals) or “EEID” (insects) in the poreregion of domain I-IV are critical for TTX binding (Catterall, 2000). InDSC1/BSC1 channels the amino acid sequences “DEEA” and “EEIN,”respectively, are found in the corresponding positions. Because theDSC1 channel is insensitive toTTX (Fig. 3C), we speculate that the KeEsubstitution in the DEEA motif of the DSC1/BSC1 channels might beinvolved inTTX resistance since a charge-reversal substitution of the Kresidue in the DEKA motif renders rNav1.2 channels extremely resis-tant to TTX (Catterall, 2000). On the other hand, the N residue in theEEIN motif may not be involved in TTX resistance because an S to Nsubstitutionmade varroamite sodium channelsmore sensitive toTTX(Du et al., 2009). Besides these two motifs, a non-aromatic residue,cysteine or serine, immediately after the first E of the EEMDmotif hasbeenshowntobe responsible forTTXresistanceof rNav1.5, rNav1.8andrNav1.9 channels (Satin et al., 1992; Sivilotti et al., 1997). Interestingly,both DSC1 and BSC1 channels have a non-aromatic residue (N359 inDSC1, N366 in BSC1) in this position. Therefore, N359/N366 may alsocontribute to TTX-insensitivity of DSC1/BSC1 channels.

In conclusion, DSC1 gene encodes a voltage-gated cationchannel that is permeable to Ba2þ, Ca2þ, and Naþ. DSC1 transcriptsundergo extensive alternative splicing and RNA editing. Theseposttranscriptional mechanisms likely play a critical role in regu-lating DSC1 expression level or in producing channels with distinctgating properties.

Acknowledgement

The work was supported by a grant from the National Institutesof Health (GM080255). We thank Dr. Kris Silver for reviewing anearlier version of the manuscript.

References

Castella, C., Amichot, M., Bergé, J.B., Pauron, D., 2001. DSC1 channels are expressedin both the central and the peripheral nervous system of adult Drosophilamelanogaster. Invert. Neurosci. 4, 85e94.

Catterall, W.A., 2000. From ionic currents to molecular mechanisms: the structureand function of voltage-gated sodium channels. Neuron 26, 13e25.

Dong, K., 2007. Insect sodium channels and insecticide resistance. Invert. Neurosci.7, 17e30.

Du, Y., Nomura, Y., Liu, Z., Huang, Z.Y., Dong, K., 2009. Functional expression of anarachnid sodium channel reveals residues responsible for tetrodotoxin resis-tance in invertebrate sodium channels. J. Biol. Chem. 284, 33869e33875.

Feng, G., Deák, P., Chopra, M., Hall, L.M., 1995. Cloning and functional analysis ofTipE, a novel membrane protein that enhances Drosophila para sodium channelfunction. Cell 82, 1001e1011.

Germeraad, S., O’Dowd, D., Aldrich, R.W., 1992. Functional assay of a putative Drosophilasodium channel gene in homozygous deficiency neurons. J. Neurogenet. 8, 1e16.

Goldin, A.L., 1992. Maintenance of Xenopus laevis and oocyte injection. MethodsEnzymol. 207, 266e279.

Heinemann, S.H., Terlau, H., Stühmer, W., Imoto, K., Numa, S., 1992. Calcium channelcharacteristics conferred on the sodium channel by single mutations. Nature356, 441e443.

Hong, C.S., Ganetzky, B., 1994. Spatial and temporal expression patterns of twosodium channel genes in Drosophila. J. Neurosci. 14, 5160e5169.

Hoopengardner, B., Bhalla, T., Staber, C., Reenan, R.A., 2003. Nervous system targetsof RNA editing identified by comparative genomics. Science 301, 832e836.


Kerr, N.C., Holmes, F.E., Wynick, D., 2008. Novel mRNA isoforms of the sodiumchannels Na(v)1.2, Na(v)1.3 and Na(v)1.7 encode predicted two-domain, trun-cated proteins. Neuroscience 155, 797e808.

Kulkarni, N.H., Yamamoto, A.H., Robinson, K.O., Mackay, T.F., Anholt, R.R., 2002. TheDSC1 channel, encoded by the smi60E locus, contributes to odor-guidedbehavior in Drosophila melanogaster. Genetics 161, 1507e1516.

Lin, W.H., Wright, D.E., Muraro, N.I., Baines, R.A., 2009. Alternative splicing in thevoltage-gated sodium channel DmNav regulates activation, inactivation, andpersistent current. J. Neurophysiol. 102, 1994e2006.

Littleton, J.T., Ganetzky, B., 2000. Ion channels and synaptic organization: analysis ofthe Drosophila genome. Neuron 26, 35e43.

Liu, Z., Chung, I., Dong, K., 2001. Alternative splicing of the BSC1 gene generatestissue-specific isoforms in the German cockroach. Insect. Biochem. Mol. Biol. 31,703e713.

Liu, Z., Song, W., Dong, K., 2004. Persistent tetrodotoxin-sensitive sodium currentresulting from U-to-C RNA editing of an insect sodium channel. Proc. Natl. Acad.Sci. U.S.A. 101, 11862e11867.

Loughney, K., Kreber, R., Ganetzky, B., 1989. Molecular analysis of the para locus,a sodium channel gene in Drosophila. Cell 58, 1143e1154.

O’Dowd, D.K., Gee, J.R., Smith, M.A., 1995. Sodium current density correlates withexpression of specific alternatively spliced sodium channel mRNAs in singleneurons. J. Neurosci. 15, 4005e4012.

Olson, R., Liu, Z., Nomura, Y., Zhang, T., Song, W., Dong, K., 2008. Molecular andfunctional characterization of Para sodium channels in Drosophila melanogaster.Insect. Biochem. Mol. Biol. 38, 604e610.

Park, Y., Taylor, M.F., Feyereisen, R., 1999. Voltage-gated sodium channel genes hscpand hDSC1 of Heliothis virescens F. genomic organization. Insect. Mol. Biol. 8,161e170.

Plummer, N.W., McBurney, M.W., Meisler, M.H., 1997. Alternative splicing of thesodium channel SCN8A predicts a truncated two-domain protein in fetal brainand non-neuronal cells. J. Biol. Chem. 272, 24008e24015.

Salkoff, L., Butler, A., Wei, A., Scavarda, N., Giffen, K., Ifune, C., Goodman, R.,Mandel, G., 1987. Genomic organization and deduced amino acid sequence ofa putative sodium channel gene in Drosophila. Science 237, 744e749.

Satin, J., Kyle, J.W., Chen, M., Bell, P., Cribbs, L.L., Fozzard, H.A., Rogart, R.B., 1992.A mutant of TTX-resistant cardiac sodium channels with TTX-sensitive prop-erties. Science 256, 1202e1205.

Sivilotti, L., Okuse, K., Akopian, A.N., Moss, S., Wood, J.N., 1997. A single serineresidue confers tetrodotoxin insensitivity on the rat sensory-neuron-specificsodium channel SNS. FEBS Lett. 409, 49e52.

Soderlund, D., 2005. Sodium channels. In: Gilbert, L.I., Iatrou, K., Gill, S.S. (Eds.),Comprehensive Insect Science. Pharmacology, vol. 5. Elsevier BV, Amsterdam,pp. 1e24.

Song, W., Liu, Z., Tan, J., Nomura, Y., Dong, K., 2004. RNA editing generates tissue-specific sodium channels with distinct gating properties. J. Biol. Chem. 279,32554e32561.

Tan, J., Liu, Z., Nomura, Y., Goldin, A.L., Dong, K., 2002. Alternative splicing of aninsect sodium channel gene generates pharmacologically distinct sodiumchannels. J. Neurosci. 22, 5300e5309.

Tatebayashi, H., Narahashi, T., 1994. Differential mechanism of action of the pyre-throid tetramethrin on tetrodotoxin-sensitive and tetrodotoxin-resistantsodium channels. J. Pharmacol. Exp. Ther. 270, 595e603.

Thackeray, J.R., Ganetzky, B., 1994. Developmentally regulated alternative splicinggenerates a complex array of Drosophila para sodium channel isoforms. J. Neu-rosci. 14, 2569e2578.

Thackeray, J.R., Ganetzky, B., 1995. Conserved alternative splicing patterns andsplicing signals in the Drosophila sodium channel gene para. Genetics 141,203e214.

Warmke, J.W., Reenan, R.A., Wang, P., Qian, S., Arena, J.P., Wang, J.,Wunderler, D., Liu, K., Kaczorowski, G.J., Van der Ploeg, L.H., Ganetzky, B.,Cohen, C.J., 1997. Functional expression of Drosophila para sodium channels.Modulation by the membrane protein TipE and toxin pharmacology. J. Gen.Physiol. 110, 119e133.

Yu, F.H., Yarov-Yarovoy, V., Gutman, G.A., Catterall, W.A., 2005. Overview ofmolecular relationships in the voltage-gated ion channel superfamily. Phar-macol. Rev. 57, 387e395.

Zhou, W., Chung, I., Liu, Z., Goldin, A.L., Dong, K., 2004. A voltage-gated calcium-selective channel encoded by a sodium channel-like gene. Neuron 42,101e112.


insect biochemistry and molecular biologykedonglab.ent.msu.edu/publications/zhang et al.,...

Documents