Ž .Biochimica et Biophysica Acta 1397 1998 156–160

Short sequence-paper

The exon–intron architecture of human chloride channel genes isnot conserved 1

Jan Eggermont )

Laboratorium Õoor Fysiologie, Katholieke UniÕersiteit LeuÕen, Campus Gasthuisberg O&N, B-3000 LeuÕen, Belgium

Received 30 October 1997; revised 5 January 1998; accepted 13 January 1998

Abstract

The human CLCN6 gene contains a 167 bp exon that is optionally included or excluded in ClC-6 mRNAs. TheŽ .corresponding region 3.4 kbp of the human CLCN7 gene has now been cloned and sequenced. A comparison of the

human CLCN1, CLCN5, CLCN6 and CLCN7 genes indicates that there is no homologue of the optional CLCN6 exon in theCLCN1, CLCN5 or CLCN7 genes. Thus, the CLCN6 type of alternative splicing and the ensuing structural diversity is notconserved within the CLC gene family. q 1998 Elsevier Science B.V.

Keywords: Chloride channel; Splicing; Gene structure; Exon

Ž .ClC Chloride Channel proteins form a superfam-ily which in humans contains at least 9 different

Ž . w xgenes CLCN1 to CLCN7, CLCKa and CLCKb 1 .Heterologous expression of ClC-1 and ClC-2 hasclearly shown that they form voltage-gated chloride

w xchannels located in the plasma membrane 2,3 , butthe functional characterisation of other ClC proteins,

w xe.g., ClC-6 and ClC-7, remains problematic 4,5 .Multiple sequence comparison has revealed threebranches in the ClC family: one branch containsClC-1, ClC-2 and the renal ClC-Ka and -Kb iso-forms; a second branch contains ClC-3, -4 and -5; the

w xthird branch consists of ClC-6 and -7 4 . ClC pro-teins share a common membrane topology with 10

) Fax: q32-1-6-345991.1 The CLCN7 genomic sequence is deposited at the EMBL

database with accession number AJ001910.

Ž .or less likely 12 transmembrane domains and intra-w xcellular N- and C-termini 6 . We have recently de-

scribed alternative splicing of human CLCN6 genew xwhich generates 4 ClC-6 isoforms 7 . ClC-6a has a

canonical ClC structure, whereas ClC-6b, ClC-6c andClC-6d are C-terminally truncated isoforms with a

Ž . Ž .maximum of 4 ClC-6b and -6d or 7 ClC-6cpredicted transmembrane domains. Genomic se-quence analysis revealed the presence of two regionsin the human CLCN6 gene where alternative splicing

w xoccurred 7 . The first region consists of an intronŽ .flanked by two alternative donor D1 and D2 and

Ž .two alternative acceptor A1 and A2 sites: splicingof D1 to A1 results in ClC-6arc open reading framesŽthe 6a and 6c sequences are differentiated at the

.downstream optional exon: cf. infra ; the D2-A1splice inserts 10 nucleotides and generates the ClC-6bsequence; the D1-A2 option causes a 26 nt deletion

w xand creates the ClC-6d isoform 7 . The second re-

0167-4781r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved.Ž .PII S0167-4781 98 00014-1

( )J. EggermontrBiochimica et Biophysica Acta 1397 1998 156–160 157

gion of variation corresponds to a 167 bp exon andthey are ubiquitously present. In contrast, omission ofthe 167 bp exon generates mRNAs for the C-termi-nally truncated ClC-6c isoform of which the expres-

w xsion is limited to the kidney 7 .

The exonrintron architecture is often conservedbetween closely related members of a gene familyand has been taken as evidence for functional similar-

w xity 8 . This raises the question to which extent thegene architecture and in particular the optional

Fig. 1. Nucleotide sequence of the human CLCN7 genomic fragment. A 3397 bp fragment was isolated from human genomic DNA byPCR using sets of overlapping PCR primers. The PCR fragments were subcloned in pBluescript II and sequenced bidirectionally using the

Ž .automated ALF system Pharmacia . A combination of custom-made primers and overlapping restriction fragments was used to determinethe sequence in both directions. Exon sequences are represented in capitals, intron sequences in lower case. Exon–intron junctions are

Ž <.denoted by vertical bars . Only the beginning and the end of the intron sequences are shown. The nucleotide sequence is continuouslynumbered and the numbers are given above the nucleotide sequence. Introns are numbered x to xq4 and the total length of each intronis given between the exon sequences. Exons are numbered y to yq5. The length of exons yq1 to yq4 are also shown. The

Ž .corresponding ClC-7 amino acid sequence single letter code is given underneath the exon sequences.

( )J. EggermontrBiochimica et Biophysica Acta 1397 1998 156–160158

CLCN6 exon is conserved among CLC genes. Thisquestion was addressed by cloning the correspondingregion in the human CLCN7 gene which is theclosest relative to CLCN6 within the ClC tree. Usinga combination of PCR primers a 3397 bp fragment ofthe human CLCN7 gene was sequenced. This regioncontains the coding information for bp 775 to 1249 ofthe CLCN7 cDNA sequence and it corresponds to theCLCN6 genomic region we have previously charac-

w xterised 7 . Alignment of the genomic sequence withthe cDNA sequence indicates the presence of 5 in-

Ž .trons Fig. 1 . All introns contained donor splice siteswhich matched the consensus donor sequenceŽ < . w xAG GTRAGT 9 in at least 6 of the 8 positions.Splice acceptor sites also corresponded to the consen-

Ž < . w xsus sequence ...Y CAG ... 9 .n

The exonrintron architecture of the CLCN7 ge-nomic fragment was then compared with the corre-sponding regions in the CLCN1, CLCN5 and CLCN6

Ž .genes Fig. 2 . For CLCN1 this region corresponds toŽ . Ž .exon 6 partial , exons 7,8,9,10 and exon 11 partial

w x Ž .10 . For CLCN5 it includes exon 6 partial , exon 7Ž . w xand exon 8 partial 11 . For CLCN6 this region

Ž .encompasses exon 9 partial , exons 10,11,12 andŽ . Žexon 13 partial CLCN6 exons are numbered ac-

cording to the data deposited in the EMBL database.with accession numbers AF009253 and AF009254 .

From the alignment in Fig. 2 it is clear that theexonrintron layout is not conserved between theCLC genes. Indeed, the region is interrupted by 2introns in the CLCN5 gene, by 4 introns in theCLCN6 gene and by 5 introns in the CLCN1 andCLCN7 genes. It is also clear that the optional exon

Ž .in the CLCN6 gene exon 12 is not conserved in theCLCN1, CLCN5 or CLCN7 genes. The correspond-ing region is part of a large exon in CLCN5, but it isinterrupted once in CLCN1 and twice in CLCN7. Thedegree of diversification within the ClC family canalso be inferred from the exonrintron structure sur-

Ž .rounding the D5 region Fig. 3 . Domain 5 is a highlyconserved hydrophobic region which according to therecent structural model corresponds to the 4th trans-

w xmembrane domain 6 . In spite of its high degree ofsequence identity at the amino acid level, this domainis encoded by a single exon in the CLCN5 andCLCN6 genes, but is interrupted at a different site by

Fig. 2. Partial exon–intron layout of human CLCN1, CLCN5, CLCN6 and CLCN7 genes. The exon–intron architecture of the CLCN7genomic fragment is schematically compared with the corresponding regions in the human CLCN1, CLCN5 and CLCN6 genes. These

Žregions encompass hydrophobic domains 5 to 8 indicated by grey bars; hydrophobic domains are defined according to Steinmeyer et al.w x.3 . Exons are represented by boxes and the location of introns is indicated by stepping up or down. The CLCN1 exons are numbered

w x w xaccording to Lorenz et al. 10 ; CLCN5 exons according to Fisher et al. 11 ; CLCN6 exons according to the data deposited in the EMBLŽ .database with accession numbers AF009253 and AF009254. Exon 9 of CLCN1 and exon 8 of CLCN5 are interrupted broken lines to

Ž .preserve the alignment of the coding regions. The optional exon in CLCN6 exon 12 is hatched. Vertical dashed lines indicate commonexonrintron boundaries between CLCN6 and CLCN7 in D6 and between CLCN1 and CLCN6 upstream of D8.

( )J. EggermontrBiochimica et Biophysica Acta 1397 1998 156–160 159

ŽFig. 3. Lack of correlation between exon–intron layout and conserved protein domains. Amino acid sequences of human ClC-1 aa. Ž . Ž . Ž .253–337 , ClC-5 aa 251–330 , ClC-6 aa 227–314 and ClC-7 aa 259–344 were aligned with the pileup programme of the GCG

Žcomputer programme. This region contains the highly conserved D5 domain and the less conserved D6 domain indicated on top by open.bars . Residues that are conserved in at least three of the four ClC sequences, are indicated by asterisks on top of the alignment.

Exonrintron boundaries are indicated by triangles underneath the amino acid sequence.

an intron in the CLCN1 and CLCN7 genes. Althoughthere are no absolutely conserved exonrintronboundaries between the 4 genes, common boundariescan be identified when the CLC genes are pairwise

Ž .compared Fig. 2 . This is illustrated by the intronthat interrupts the D6 domain at exactly the same sitein the CLCN6 and CLCN7 genes. Furthermore, theCLCN1 and CLCN6 genes share an identically lo-cated intron upstream of the D8 domain.

It is therefore concluded that the exonrintron ar-chitecture is not conserved between the ClC family

Ž .and that structurally conserved domains e.g., D5 arenot necessarily encoded by a single exon. However,the colocalisation of some exonrintron boundarieswhen pairwise compared is reminiscent of the com-mon ancestral CLC gene which has gradually beenmodified either by intron deletion or intron insertionw x12,13 . Moreover, the optional exon of CLCN6 is assuch not conserved in CLCN1, CLCN5 or CLCN7indicating that the CLCN6 alternative splice patterngenerating the truncated isoforms is a specific featureof the CLCN6 gene.

The P-type ion transport ATPase family comprisesstructurally related ion pumps such as sarcorendo-

2q Ž .plasmic reticulum Ca ATPases SERCA , plasmamembrane Ca2q ATPases and NaqrKq ATPasesw x14 . In this family exon–intron boundaries are con-served for genes encoding functionally identicalpumps, e.g., the SERCA genes and the NaqrKq

ATPase genes, but the gene architecture varies be-tween functionally different ion pumps, e.g., SERCA

q q w xversus Na rK ATPase 15–17 . Extending theseobservations to the ClC family one may conclude that

the reported differences in gene architecture betweenCLCN1, CLCN5, CLCN6 and CLCN7 reflect an asyet unidentified functional diversity, e.g., in terms oftransported substrate andror subcellular location. Thelatter point is supported by recent observations thatthe ClC-6 isoform is an intracellular membrane pro-tein most likely residing in the endoplasmic reticulumw x18 .

I thank Peter Marynen for the gift of humangenomic DNA and Diane Hermans for expert techni-cal assistance. J.E. is a Research Associate of the

ŽFlemish Fund for Scientific Research FWO-.Vlaanderen .

References

w x Ž .1 T.J. Jentsch, W. Gunther, BioEssays 19 1997 117–126.w x2 A. Thiemann, S. Grunder, M. Pusch, T.J. Jentsch, Nature

Ž .356 1992 57–60.w x Ž .3 K. Steinmeyer, C. Ortland, T.J. Jentsch, Nature 354 1991

301–304.w x Ž .4 S. Brandt, T.J. Jentsch, FEBS Lett. 377 1995 15–20.w x5 G. Buyse, T. Voets, J. Tytgat, C. De Greef, G. Droogmans,

Ž .B. Nilius, J. Eggermont, J. Biol. Chem. 272 1997 3615–3621.

w x6 T. Schmidt-Rose, T.J. Jentsch, Proc. Natl. Acad. Sci. U.S.A.Ž .A94 1997 7633–7638.

w x7 J. Eggermont, G. Buyse, T. Voets, J. Tytgat, H. De Smedt,Ž .G. Droogmans, B. Nilius, Biochem. J. 325 1997 269–276.

w x Ž .8 B. Lewin, in: Genes V, Oxford Univ. Press, Oxford 1994pp. 677–702.

w x Ž .9 A.I. Lamond, BioEssays 15 1993 595–603.w x10 C. Lorenz, C. Meyer-Kleine, K. Steinmeyer, M.C. Koch,

Ž .T.J. Jentsch, Hum. Mol. Genet. 3 1994 941–946.w x11 S.E. Fisher, I. Vanbakel, S.E. Lloyd, S.H.S. Pearce, R.V.

Ž .Thakker, I.W. Craig, Genomics 29 1995 598–606.

( )J. EggermontrBiochimica et Biophysica Acta 1397 1998 156–160160

w x12 A. Stoltzfus, D.F. Spencer, M. Zuker, J.M. Logsdon Jr.,Ž .W.F. Doolittle, Science 265 1994 202–207.

w x13 J.D. Palmer, J.M. Logsdon Jr., Curr. Opin. Genet. Dev. 1Ž .1991 470–477.

w x Ž .14 N.M. Green, Ann. New York Acad. Sci. 671 1992 104–112.

w x15 M.M. Shull, D.G. Pugh, J.B. Lingrel, J. Biol. Chem. 264Ž .1989 17532–17543.

w x Ž .16 R. Escalante, L. Sastre, J. Biol. Chem. 269 1994 13005–13012.

w x17 L. Dode, F. Wuytack, P.F. Kools, F. Baba Aissa, L. Raey-maekers, F. Brike, W.J. van de Ven, R. Casteels, Biochem.´

Ž .J. 318 1996 689–699.w x18 G. Buyse, D. Trouet, T. Voets, L. Missiaen, G. Droogmans,

Ž .B. Nilius, J. Eggermont, Biochem. J. 330 1998 1015–1021.