Epilepsy Research 46 (2001) 139–144

T-STAR gene: fine mapping in the candidate region forchildhood absence epilepsy on 8q24 and mutational analysis

in patients

Yoshihisa Sugimoto a, Ryoji Morita a, Kenji Amano a, Pravina U. Shah b,Ignacio Pascual-Castroviejo c, Sonia Khan d, Antonio V. Delgado-Escueta e,

Kazuhiro Yamakawa a,*a Laboratory for Neurogenetics, Brain Science Institute, The Institute of Physical and Chemical Research (RIKEN), 2-1 Hirosawa,

Wako-shi, Saitama, 351-0198, Japanb K.E.M. Hospital and Seth G. S. Medical College, Bombay, Indiac Pediatric Neurology, Uni�ersity Hospital La Paz, Madrid, Spain

d Neuroscience Department, Riyadh Armed Forces Hospital, Riyadh, Saudi Arabiae Comprehensi�e Epilepsy Program, UCLA School of Medicine and West Los Angeles DVA Medical Center, 11301 Wilshire Bl�d,

Los Angeles, CA 90073, USA

Received 27 April 2001; received in revised form 4 May 2001; accepted 7 May 2001

Abstract

Childhood absence epilepsy (CAE) is one of the most common epilepsies in children. At least four phenotypicsubcategories of CAE have been proposed. Among them, a subtype persisting with tonic–clonic seizures has beenmapped to 8q24 (ECA1 MIM 600131). By constructing a physical map for the 8q24 region, we recently narrowed theECA1 locus to a 1.5-Mb region. In the present communication, we show that T-STAR gene is located within theECA1 region. T-STAR is a novel member of STAR (for signal transduction and activation of RNA) family, and ispredicted to encode a spermatogenesis related RNA-binding protein. T-STAR is located within the markers D8S2049and D8S1753 and its complete coding region spans nine exons. In addition to its known expression in testis, moderatelevel of transcripts for T-STAR gene was detected in brain, heart and is highly abundant in skeletal muscle.Mutational analysis for the T-SATR gene in CAE families did not show any sequence variation in the coding region,and this suggests that the T-STAR gene is not involved in the pathogenesis of persisting CAE. However, genomicorganization of T-STAR gene characterized in the present report might help in understanding the biological functionsof T-STAR as well as its suspected involvement in other disorders mapped on this region.

GenBank accession No. NM–006558 © 2001 Elsevier Science B.V. All rights reserved.

Keywords: T-STAR; Genomic structure; Chromosomal band 8q24; Physical map

www.elsevier.com/locate/epilepsyres

* Corresponding author. Tel.: +81-48-467-9703; fax: +81-48-467-7095.E-mail address: yamakawa@brain.riken.go.jp (K. Yamakawa).

0920-1211/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved.

PII: S 0920 -1211 (01 )00274 -1

Y. Sugimoto et al. / Epilepsy Research 46 (2001) 139–144140

1. Introduction

Epilepsies can be broadly divided into symp-tomatic and idiopathic depending on whetherseizures occur in the presence of detectable brainlesions and/or metabolic abnormalities. Idiopathicepilepsies are further categorized into partial andgeneralized. Among the idiopathic generalizedepilepsies (IGEs), childhood absence epilepsy

(CAE) is one of the most frequent; accounting for5–15% of all epilepsies. CAE is characterized by itsonset in childhood and frequent absence attackswhich number several to hundreds a day (Cavaz-zuti, 1980). CAE can be further subcategorized intoseveral classes such as (1) CAE often with absenceseizure as a sole phenotype, (2) persisting CAEassociated with tonic-clonic seizures during adoles-cence, (3) CAE with eyelid myoclonia induced byphotic stimulus (natural and electronic light as wellas by stroboscopic), and (4) CAE that persists withmyoclonic and atonic seizures during adolescence.We previously identified the locus for CAE persist-ing with grand-mal seizures on chromosome 8q24(ECA1; MIM 600131) by genetic linkage analysis(Fong et al., 1998). Recently, we narrowed downthe ECA1 region to 1.5 Mb region, between mark-ers D8S554 and D8S502, by constructing a BACand YAC-based physical map covering the entireECA1 region and by refined haplotype analysisusing accurately ordered STS markers on the map(Sugimoto et al., 2000). Several genes mapped on8q24 including the KCNQ3 gene (Charlier et al.,1998), the ionotropic NMDA-associated glutamatereceptor 1 (Lewis et al., 1996) and human homo-logue of the mouse jerky gene (JRK/JH8) (Moritaet al., 1998) have been excluded as candidates forthe ECA1 gene.

In the present study, we show that the T-STARgene is located within the 1.5 Mb ECA1 region. Wedetermined the genomic organization of T-STARand screened for sequence variation in affectedmembers of CAE families.

2. Patients and methods

2.1. Patients

The subjects analyzed in this study were affectedmembers of three unrelated CAE families, eachfrom India (I201), Spain (M8) and Saudi Arabia(S302), and five unrelated Japanese healthy con-trols. Each participating subject or responsibleadult signed an informed consent form as approvedby the Human Subject Protection Committee at theUniversity of California Los Angeles, UCLASchool of Medicine or at the participating institu-

Fig. 1. Chromosomal localization and genomic organization ofT-STAR gene. (a) The T-STAR cDNA clone was hybridizedto the ECA1 region BAC blot. The T-STAR gene showedpositive signals on BAC clones RKN 355 O18 and 428 E 6(arrowheads). Arrow indicates BAC vector-derived signals.The hybridization signal on RKN 323 D 5 was nonspecific(asterisk). (b) Relative location of the T-STAR gene andECA1 region BAC clones were shown. Open arrow indicateslocation and transcriptional direction of the T-STAR gene.The T-STAR is located within ECA1 region, flanking toD8S2049 and D8S1753. (c) Relationship between exon organi-zation and functional domains of the T-STAR gene. Openboxes indicate non-coding regions. Q1, QUA1 region; Q2,QUA2 region; KH, maxi-KH domain; SH, potential bindingsite for SH3 domain.

Y. Sugimoto et al. / Epilepsy Research 46 (2001) 139–144 141

Table 1Exon–intron boundary sequence of T-STAR gene

Positiona Splice acceptor bExon number Splice donorbExon length (bp)

1–4821 482 GTGAACCAAGgtgaggcgcc2 119 483–601 gcctggacagAAATAGAAAA GTTCCCTAAGgtaagacagt3 117 602–718 tttttcctagTTCAACTTTG CAAGGCCAAGgtaatattaa

719–865 ctttttccagGAAGAAGAGT147 CCTCATCCCTgttagtaccg41405 866–1005 ttttgtgtagGATTATAATG CAATAACCAGgtaggtgtaa1966 1006–1201 ttggttacagGGGAAGGGGA TGGAGAATATgtaagtgaag

1202–1284 tatgttttagGACTATGATG83 CAGCCCAAAGgtaagagtca71285–1343 tcacttgcagTGGTGCTGAT8 GATTCCTACGgtgagtgact591344–1963 ctgattgtagGGCAAGAAGA6209

a Position is according to GenBank accession No. NM–006558.b Boldface indicates splice donor and acceptor site.

tions. Clinical features and linkage data for thesefamilies were previously described (Fong et al.,1998; Sugimoto et al., 2000).

2.2. Southern hybridization analysis of cDNAclones against BAC clones

The EST clone AA112249 (IMAGE 530468)was purchased from Research Genetics. T-STARcDNA clone was isolated from human cDNAlibrary as described below. The cDNA fragmentswere labeled with AlkPhos direct labeling kit(Amersham Pharmacia Biotech) and hybridizedover ECA1 region BAC blot (Sugimoto et al.,2000), then the signals were detected by usingCDP-Star Detection Reagent (Amersham Phar-macia Biotech). The conditions for hybridization,post-hybridization washes and signal detectionwere carried out by following the manufacturer’srecommendation. Filters were exposed to X-rayfilm (X-OMAT AR; Kodak) overnight.

2.3. Screening of cDNA library

Screening of the human fetal brain cDNA li-brary cloned in the lambda ZAP II vector (Strata-gene) was performed by using the EST cloneAA112249 as probe. Phages were plated to anaverage density of 1×105 pfu per 175 cm2 plate.Plaque lifts of 20 plates (2×106 phages) weremade using duplicate Biodyne B nylon membrane(Nihon Pall). Hybridization, washing and signaldetection were same as above. Positive cDNA

clones were sequenced by dye-terminator methodas described below.

2.4. Northern blot analysis

The insert fragment of T-STAR cDNA clonewas 32P radio-labeled by Multiprime DNA Label-ing System (Amersham Pharmacia Biotech) andpurified using G-50 Sephadex columns (QuickSpin Columns; Boehringer Mannheim). The la-beled fragment was used as probe for Northern

Fig. 2. Northern blot analysis of the T-STAR gene. A radiola-beled T-STAR cDNA fragment was hybridized to humanadult multiple tissue Northern blot. A major 3 kb transcript(solid arrowhead) and a minor 2.2 kb transcript (open arrow-head) were detected. Strong signals were observed in skeletalmuscle, and relatively weak signals were observed in heart andbrain.

Y. Sugimoto et al. / Epilepsy Research 46 (2001) 139–144142

Table 2Oligonucleotide primers used for heteroduplex analysis and direct sequencing

Amplicon Downstream primerUpstream primer Size (bp)

5�-CTTCAAGTCTCACCCATGCG-3�5�-GCCTGGAGTCCACATCCCG-3� 36815�-GCAGTGTAGACCTCACTGTC-3�2 2975�-GATGAAACCGGAAATGTACTTG-3�5�-CATGATACACAACTTAGCCG-3�5�-GCCATTAATCAATGATGTGTAGCA-3� 26435�-TGAGGTGATGTCCATACAGC-3� 3954 5�-ACCGCTTGCCAGATGGATTG-3�5�-AAACTGCTGCCACGGTATG-3�5�-CAGAGAAAGGTTAGTGGCTG-3� 42955�-AATATCCAGTGCTAAGTCTGG-3�6 2935�-ACTGCTAATTTGACTCTCTTGG-3�5�-CCTCTCTAGCTGACTGCCATAC-3�5�-TGACCATTTTTACCAATTGCTC-3� 38575�-AAGCTGCTGTTTCAGAGTCC-3�8 3185�-AGAACTCTGTGATCTCGGTG-3�5�-GTTTTAAGTTCCATCCATTCAG-3�5�-TGGGATTCATGGCAGTGATC-3� 2739

Table 3Polymorphysms in T-STAR gene

PositionAmplicon SequencePedigree

45 Base upstream of exon4Normal control tcatttctt{G/A}taatgacac410 Base downstream of exon44 gttagtacc{G/A}tttttcttgNormal control104 Base upstream of exon8M8 ttttcagtc{C/T}tgcccttca869 Base upstream of exon8 tttacaata{C/T}gttattgaa8 M8112 Base downstream of stop codonM8 aataaatca{A/G}aatgcttaa9

hybridization using a Multiple Tissue NorthernBlot (Clontech). Hybridization was performed byfollowing the manufacturer’s instructions. Hy-bridized membrane was washed at a final stringencyof 0.1x SSC and 0.1% SDS at 50 °C. The filter wasexposed to X-ray film at −70 °C overnight.

2.5. Sequence analysis of exon– intron junctions

The nucleotide sequences of exon– intronboundaries were determined by sequencing theBAC clones RKN428E6 and RKN 355O8 (Sugi-moto et al., 2000). The BAC DNAs were extractedby using QIAGEN Plasmid Midi Kit (QIAGEN),and sequenced with BigDye Terminator CycleSequencing Kit and ABI PRISM 377 DNA se-quencing system (PE Applied Biosystems). Theprimers for sequencing were designed from theT-STAR cDNA sequence (GenBank accession No.NM–006558).

2.6. Mutational analysis

Genomic DNA was isolated from peripheral

blood leukocyte by using QIAamp DNA bloodmidi kit (QIAGEN). To identify the nucleotidechanges in the coding region of T-STAR gene,genomic fragments which covered the entire codingregion, were analyzed by heteroduplex analysis anddirect sequencing method described elsewhere(Morita et al., 1999). Briefly, genomic fragmentswere amplified by using Pfu Turbo DNA poly-merase (Stratagene). Subsequently, denatured andgradually reannealed PCR products were analyzedby denaturing high-performance liquid chromatog-raphy using WAVE DNA fragment analysis system(Transgenomic). Direct sequencing was performedto confirm the existence of nucleotide changes.

3. Results

3.1. Mapping of T-STAR gene in the ECA1region

To identify the gene responsible for CAE, wesearched for genes located within the 1.5MbECA1 region on 8q24 that we narrowed previ-

Y. Sugimoto et al. / Epilepsy Research 46 (2001) 139–144 143

ously (Sugimoto et al., 2000). We selected severalESTs located between the markers D8S1710 andD8S272 from the GeneMap98 (http://www.ncbi.nlm.nih.gov/genemap98/) and hybridized over theECA1 region BAC blots which containing BACclones shown in Fig. 1b. One of the EST clones,AA112249 hybridized to the BAC cloneRKN428E6. Using this EST clone as a probe, wescreened a human fetal brain cDNA library andisolated five independent clones. After sequencinganalysis of these clones, we found that two of thesecDNA clones were identical to the T-STAR gene(GenBank accession No. NM–006558; Venables etal., 1999). The other three clones were T-STARgene splicing variants which contain alternative 3�noncoding region (data not shown). After hy-bridization analysis of T-STAR cDNA cloneagainst ECA1 region BAC blot, we found that theT-STAR gene is located on the BACs RKN 428E6and RKN355O18 that contain markers D8S2049and D8S1753, respectively (Fig. 1a,b).

3.2. Characterization of genomic structure for theT-STAR gene

The exon– intron structure of the T-STAR genewas identified by sequencing BAC clones. TheT-STAR gene was shown to consist of nine exonsinterrupted by eight introns (Fig. 1c). Sequences atthe exon-intron boundaries for all eight introns arecompatible with the consensus sequence for thesplicing junctions (Mount, 1982). The QUA1 do-main spans exon 1 and 2, maxi-KH domain spansexons 2–4, QUA2 domain and SH3-binding siteare located on exon 5 and exon 6, respectively. Thesequences of exon-intron boundary and exonlengths are given in Table 1.

3.3. Northern analysis of T-STAR gene

Expression of T-STAR gene was analyzed byNorthern blot hybridization (Fig. 2). Approxi-mately 3 and 2.4 kb transcripts were detected in theRNA samples from the heart and brain and wereexpressed more highly in skeletal muscle. Signals inother tissues were weak or negligible. The differentsize of transcripts may be due to be splicingisoforms.

3.4. Mutation analysis of the T-STAR gene inCAE patients

We performed mutation analysis of the T-STARgene entire coding region in CAE patients. Thenucleotide sequences of primers and the size ofamplicons used for mutation analysis are listed inTable 2. By screening with heteroduplex analysis,we found three aberrant chromatograms in ampli-con 4, 8 and 9. By direct sequencing of those PCRproducts, we identified five nucleotide changes aslisted in Table 3. All these SNPs were located inintrons or 3� non-coding region. There was nomutations or polymorphisms in the coding region.

4. Discussion

The T-STAR gene was previously mapped on8q22–q24 by radiation hybrid method (Venables etal., 1999). Here we demonstrated that the exactlocus of this gene is 8q24 flanked by D8S2049 andD8S1753 and that it lies within the 1.5Mb ECA1region between D8S554 and D8S502 as shown bya sequence-ready physical map (Sugimoto et al.,2000). We suspected that the T-STAR gene couldbe involved in the pathogenesis of childhood ab-sence epilepsy because (1) it is located in the criticalECA1 region, (2) it is expressed in brain, and (3)a defect in one member of the STAR gene family,quaking, leads to tonic-clonic seizures (Sidman etal., 1964).

Mutational analysis of the T-STAR gene in CAEpatients revealed only three SNPs in the intronic or3�-untranslated regions. Since there were no differ-ences in the coding region of the T-STAR genebetween CAE patients and unaffected individuals,we conclude that it is unlikely that the T-STARgene is responsible for CAE.

The STAR family members have a characteristicdomain called STAR domain. The STAR domainconsists of QUA1 region, maxi-KH region andQUA2 region and is a potential RNA-binding site(Venables et al., 1999). Most members of the STARfamily also have domains related to signal trans-duction such as SH3- and WW-binding site (Vernetand Artzt, 1997). By mediating signals and RNAmetabolism, STAR proteins exert various functions

Y. Sugimoto et al. / Epilepsy Research 46 (2001) 139–144144

in embryogenesis and myelination (Ebersole et al.,1996; Justice and Bode, 1988), muscle develop-ment in Drosophila (Baehrecke, 1997) and germ-line development in C. elegans (Francis et al.,1995). In this study, we found relatively higherlevels of T-STAR gene expression in brain, heart,and skeletal muscle, with the highest being inskeletal muscle (Fig. 2). This data suggest that theT-STAR protein may play important roles notonly in spermatogenesis but also in neuromuscu-lar organs.

Although we have demonstrated that T-STARis unlikely to be involved in CAE, this gene stillremains a candidate for other neuromusculer dis-orders, such as hereditary motor and sensoryneuropathy-Lom (HMSNL), which map to 8q24close to the ECA1 region (Kalaydjieva et al.,1996). The exon-intron organization describedhere should help in understanding the biologicalfunction of T-STAR as well as its suspected in-volvement in those disorders.

Acknowledgements

We thank Ms Emi Mazaki, Ms Azusa Nittaand Ms Yukie Tsutsumi for their technical assis-tance, and Dr Subramaniam Ganesh for helpfuldiscussion.

References

Baehrecke, E.H., 1997. Who encodes a KH RNA bindingprotein that functions in muscle development. Develop-ment 124, 1323–1332.

Cavazzuti, G.B., 1980. Epidemiology of different types ofepilepsy in school age children of Modena, Italy. Epilepsia21, 57–62.

Charlier, C., Singh, N.A., Ryan, S.G., Lewis, T.B., Reus, B.E.,Leach, R.J., Leppert, M., 1998. A pore mutation in a novelKQT-like potassium channel gene in an idiopathic epilepsyfamily. Nature Genet. 18, 53–55.

Ebersole, T.A., Chen, Q., Justice, M.J., Artzt, K., 1996. Thequaking gene product necessary in embryogenesis andmyelination combines features of RNA binding and signaltransduction proteins. Nat. Genet. 12, 260–265.

Fong, C.Y., Shah, P.U., Gee, M.N., Serratosa, J.M.,Castroviejo, I.P., Khan, S., Ravat, S.H., Mani, J., Huang,Y., Zhao, H.Z., Medina, M.T., Treiman, L.J., Pineda, G.,Delgado-Escueta, A.V., 1998. Childhood absence epilepsywith tonic-clonic seizures and electroencephalogram 3–4-Hz spike and multispike-slow wave complexes: linkage tochromosome 8q24. Am. J. Hum. Genet. 63, 1117–1129.

Francis, R., Barton, M.K., Kimble, J., Schedl, T., 1995. gld-1,a tumor suppressor gene required for oocyte developmentin Caenorhabditis elegans. Genetics 139, 579–606.

Justice, M.J., Bode, V.C., 1988. Three ENU-induced alleles ofthe murine quaking locus are recessive embryonic lethalmutations. Genet. Res. 51, 95–102.

Kalaydjieva, L., Hallmayer, J., Chandler, D., Savov, A.,Nikolova, A., Angelicheva, D., King, R.H., Ishpekova, B.,Honeyman, K., Calafell, F., Shmarov, A., Petrova, J.,Turnev, I., Hristova, A., Moskov, M., Stancheva, S.,Petkova, I., Bittles, A.H., Georgieva, V., Middleton, L.,Thomas, P.K., 1996. Gene mapping in Gypsies identifies anovel demyelinating neuropathy on chromosome 8q24.Nature Genet. 14, 214–217.

Lewis, T.B., Wood, S., Michaelis, E.K., DuPont, B.R., Leach,R.J., 1996. Localization of a gene for a glutamate bindingsubunit of a NMDA receptor (GRINA) to 8q24. Genom-ics 32, 131–133.

Morita, R., Miyazaki, E., Fong, C.Y., Chen, X.N., Koren-berg, J.R., Delgado-Escueta, A.V., Yamakawa, K., 1998.JH8, a gene highly homologous to the mouse jerky gene,maps to the region for childhood absence epilepsy on 8q24.Biochem. Biophys. Res. Commun. 248, 307–314.

Morita, R., Miyazaki, E., Shah, P.U., Castroviejo, I.P., Del-gado-Escueta, A.V., Yamakawa, K., 1999. Exclusion ofthe JRK/JH8 gene as a candidate for human childhoodabsence epilepsy mapped on 8q24. Epilepsy Res. 37, 151–158.

Mount, S.M., 1982. A catalogue of splice junction sequences.Nucleic Acids Res. 10, 459–472.

Sidman, R.L., Dickie, M.M., Appel, S.H., 1964. Mutant mice(quaking and jimpy) with deficient myelination in thecentral nervous system. Science 144, 309–311.

Sugimoto, Y., Morita, R., Amano, K., Fong, C.Y., Shah,P.U., Castroviejo, I.P., Khan, S., Delgado-Escueta, A.V.,Yamakawa, K., 2000. Childhood absence epilepsy in 8q24:refinement of candidate region and construction of physi-cal map. Genomics 68, 264–272.

Venables, J.P., Vernet, C., Chew, S.L., Elliott, D.J.,Cowmeadow, R.B., Wu, J., Cooke, H.J., Artzt, K.,Eperon, I.C., 1999. T-STAR/ETOILE: a novel relative ofSAM68 that interacts with an RNA-binding protein impli-cated in spermatogenesis. Hum. Mol. Genet. 8, 959–969.

Vernet, C., Artzt, K., 1997. STAR, a gene family involved insignal transduction and activation of RNA. Trends Genet.13, 479–484.