JOURNAL OF EXPERIMENTAL ZOOLOGY 286:149–156 (2000)

© 2000 WILEY-LISS, INC.

Heterogeneity of Chicken Slow Skeletal MuscleTroponin T mRNA

I. YONEMURA, T. HIRABAYASHI, AND J.-I. MIYAZAKI*Institute of Biological Sciences, University of Tsukuba, Tsukuba, Ibaraki305-8572, Japan

ABSTRACT The troponin T (TnT) transcripts in chicken slow skeletal muscle were character-ized by S1 nuclease mapping and nucleotide sequencing of cDNA produced by RT-PCR and 5′-RACE. We found two kinds of transcripts in the 5′-region, one having the codon for alanine (position135–137), C (258), and A (262) and the other lacking the codon and having T (258) and G (262)instead of C and A. In the 3′-region, we found four single base substitutions at 703 (T or C), 774 (Cor T), 797 (C or T), and 827 (G or A). Four of the six substitutions lead to amino acid changes inchicken sTnT isoforms. We determined the genomic structure of the 3′-region of the chicken sTnTgene. The region includes 7 exons corresponding to position 249–891 of the chicken sTnT cDNAand no alternative exon, showing that the 3′-heterogeneity in sTnT transcripts was due to allelicvariation. J. Exp. Zool. 286:149–156, 2000. © 2000 Wiley-Liss, Inc.

Skeletal muscle cells are characterized by pre-cise organization of the contractile proteins intothe repeating units of overlapping thick and thinfilaments, i.e., sarcomeres, that organize in turnmyofibrils. The thick filaments are composed pri-marily of myosin, and major components of thethin filaments are actin, tropomyosin and tropo-nin. These contractile proteins have multipleisoforms that are produced by multigene familiesand by respective genes through alternative splic-ing (for review, see Andreadis et al., ’87) and/oruse of alternative promoters (Robert et al., ’84;Concordet et al., ’93). Functional significance ofthese multiple isoforms is not fully clarified so farin spite of much effort devoted for it.

Troponin, the key protein of Ca2+-sensitive mo-lecular switching for contraction in vertebratestriated muscle, consists of three subunits, tropo-nin T (TnT), troponin I, and troponin C (for re-views, see Zot and Potter, ’87; Schiaffino andReggiani, ’96). TnT isoforms are encoded by threegenes characteristic of slow skeletal muscle(sTnT), fast skeletal muscle (fTnT), and cardiacmuscle (cTnT) (Breitbart et al., ’85; Cooper andOrdahl, ’85; Gahlmann et al., ’87; Smillie et al.,’88; Mesnard et al., ’93). Each gene can generatea variety of transcripts by alternative splicing(Breitbart et al., ’85; Cooper and Ordahl, ’85;Gahlmann et al., ’87; Schachat et al., ’95). Mo-lecular organization of the rat fTnT gene has re-vealed its capacity to produce 128 different fTnTmRNAs by differential alternative splicing (Med-

ford et al., ’84; Breitbart et al., ’85; Breitbart andNadal-Ginard, ’87; Morgan et al., ’93). Smiillie etal. (’88) have found four variants of chicken fTnTcDNA, and Schachat et al. (’95) 16 variants ofchicken fTnT 5′-cDNA. The chicken cTnT genegenerates one embryonic and one adult tran-scripts, derived from inclusion and exclusion ofexon 5, respectively (Cooper and Ordahl, ’85). Amore complex alternative splicing pattern involv-ing exons 4 and 12 has been found in the rat cTnTgene (Jin and Lin, ’89; Jin et al., ’92).

On the other hand, very poor information isavailable on the mode of alternative splicing ofthe sTnT gene. The human sTnT gene can gener-ate three or possibly four transcripts, which dif-fer in the presence or absence of two short inserts(33 and 48 nucleotides) in the 5′- and 3′-end re-gions (Gahlmann et al., ’87; Samson et al., ’94).Chicken slow skeletal muscle, such as anteriorlatissimus dorsi (ALD), has been shown so far tocontain two kinds of sTnT transcripts differing ininclusion or exclusion of one codon encoding analanine residue in the 5′-end region (Yonemuraet al., ’96). To elucidate the alternative splicing

Abbreviations used: ALD, anterior latissimus dorsi; bp, base pair(s);cTnT, cardiac TnT; fTnT, fast muscle TnT; nt, nucleotide(s); PCR,polymerase chain reaction; 5′-RACE, 5′-rapid amplification of cDNAends; RT-PCR, reverse transcription-PCR; sTnT, slow muscle TnT;TnT, troponin T.

*Correspondence to: Jun-Ichi Miyazaki, Institute of Biological Sci-ences, University of Tsukuba, Tsukuba, Ibaraki 305-8572, Japan. E-mail: junichi@sakura.cc.tsukuba.ac.jp

Received 22 October 1998; Accepted 11 May 1999

150 I. YONEMURA ET AL.

pattern of the sTnT transcripts and to deduce thegene structure it is very important to investigateheterogeneity of sTnT transcripts.

In the present work, we perform S1 nucleasemapping and nucleotide sequencing of cDNA pro-duced by RT-PCR and 5′-RACE to search for het-erogeneity of chicken sTnT mRNA. We show that5′ and 3′ heterogeneities in sTnT transcripts arecaused by single base substitutions at six posi-tions and by inclusion or exclusion of the codoncoding for alanine. We also show the partial ge-nomic structure of the chicken sTnT gene.

MATERIALS AND METHODSMaterials

Adult and one-day-old (H1) white leghorn chick-ens (Gallus domesticus [L]) were obtained fromcommercial sources. Slow muscle, anterior latis-simus dorsi (ALD), mixed muscle, complexus, andfast muscle, pectoralis major, were dissected outand used for RNA preparation.RNA preparation and reverse transcription-

polymerase chain reaction (RT-PCR)Total RNA was prepared by using Isogen (Nippon

gene) following its protocol. Reverse transcriptionwas performed with 10 µg of total RNA, 0.5 µg ofoligo(dT)12–18 and Superscript II (Gibco BRL) at42°C for 60 min. For PCR of the 5′-region ofchicken sTnT, the forward primer, 5′-GGCTC-GAGGAGCCAACAGGACCG-3′ (position 1 to 15of the cDNA + Xho I site), was synthesized basedon the chicken sTnT cDNA sequence. The 3′-mixedprimer (position 522 to 541) was used as the re-verse primer (Yonemura et al., ’96). PCR was per-formed by 30 cycles of 1 min denaturation at 95°C,1 min annealing at 60°C and 1 min extension withTaq DNA polymerase (Gibco BRL) at 72°C, fol-lowed by final 10 min extension at 72°C. For PCRof the 3′-region of chicken sTnT, the reverseprimer, 5′-GGGAATTCGACGAGCAGAGCTTTAT-TGG-3′ (Primer R, position 871 to 891 of the cDNA+ EcoR I site), was also synthesized based on thechicken sTnT cDNA sequence. The 5′-mixedprimer (position 248 to 269) was used as the for-ward primer (Yonemura et al., ’96). PCR was per-formed by 35 cycles of 1 min denaturation at 94°C,1 min annealing at 65°C and 1 min extension withTaq DNA polymerase at 72°C, followed by final10 min extension at 72°C. The 5′-rapid amplifica-tion of cDNA ends (5′-RACE) was performed asin Frohman et al. (’88) and Yonemura et al. (’96).Products were cloned in the pBluescript II KS+

phagemid (Stratagene) and sequenced.

Antisense DNA probes andS1 nuclease mapping

Two clones, stnt-5′R1 and sTnT-a3, were ob-tained by cloning chicken sTnT cDNA (Yonemuraet al., ’96). The former includes 1 to 541, and thelatter 78 to 898 of sTnT cDNA. The stnt-5′R1 clonewas digested with Xho I and Pst I, and the re-sulting fragment was subcloned into the Xho I-Pst I double-digested sTnT-a3 to obtain thefull-length cDNA, which did not have three bases(135–137) encoding alanine. This full-length cDNAwas cut with Xho I and BamH I and a 690 ntfragment, probe A, was obtained. The probe A waslabeled with [32P] ATP using T4 polynucleotide ki-nase (Nippon gene). Another probe, probe B, wasgenerated by PCR with sTnT-a3 as a template us-ing the 5′-mixed primer as the forward primer(Yonemura et al., ’96) and Primer R as the re-verse primer. PCR was performed by 30 cycles of30 sec denaturation at 95°C, 10 sec annealing at60°C and 30 sec extension with KOD Dash(TOYOBO) at 74°C, followed by final 10 min ex-tension at 72°C. The probe B was labeled with[32P] dATP using KOD Dash (TOYOBO). TotalRNA (5 or 10 µg) was precipitated with 105 cpmof each probe. The pellet was dissolved in 20 µl ofthe hybridization buffer. Hybridization was per-formed overnight at 55°C before digestion withS1 nuclease (Gibco BRL). Protected fragmentswere separated on 6% acrylamide/8 M urea gels,which were exposed to X-ray film (Kodak) withintensifying screen.

Cloning of the chicken sTnT geneFor PCR of the chicken sTnT gene, the forward

primer, 5′-GCAGCCATGTCCGAAGCTGAG-3′(position 39–62 of the cDNA sequence), was syn-thesized based on the chicken sTnT cDNA se-quence. Primer R was used as the reverse primer.PCR was performed by 30 cycles of 30 sec dena-turation at 96°C and 10 min annealing and ex-tension at 74°C with 130 ng of chicken livergenomic DNA using KOD Dash (TOYOBO). Am-plified DNA was cloned in the pBluescript II KS+

phagemid (Stratagene) and sequenced.

RESULTSHeterogeneity in the 5¢-region of

chicken sTnT mRNAWe described previously two truncated cDNA

clones, stnt-5′R1 and sTnT-a3 (Yonemura et al.,’96). Using a 690 nt antisense probe (probe A) de-rived from the full-length cDNA, we performed

CHICKEN SLOW TROPONIN T mRNA 151

S1 nuclease mapping of total RNA from adult andone-day-old chick (H1) ALD to characterize the5′-region of chicken sTnT mRNA (Fig. 1). The 690nt fragment free from S1 nuclease digestionshowed that the transcript completely matchingwith the probe was expressed in both adult andH1 ALD. In addition, two other fragments of 558nt and about 440 nt were also generated, sug-gesting that adult and H1 chicken ALD had pos-sibly three kinds of transcripts which differed inat least two positions in the 5′-region. The con-sistent patterns of S1 nuclease mapping were ob-tained with several individuals (data not shown).As expected, no protected band was found in adultfast muscle, pectoralis major. The result is con-sistent with our previous data by Northern blot-ting (Yonemura et al., ’96).

To collect precise information on the sequenceswhich caused the variation in the 5′-region, weperformed nucleotide sequencing of RT-PCR and5′-RACE products from total RNA of adult chickenALD and neck muscle, complexus, which is knownto express sTnT isoforms (Yao et al., ’92). We ob-tained two RT-PCR products from ALD, A1-1R andAall-1R, and one 5′-RACE product from com-plexus, 20-2R (Fig. 2). The 20-2R had GCA cod-ing for alanine at 135 to 137 of chicken sTnTcDNA just as stnt-5′R20b, but A1-1R, Aall-1R andstnt-5′R1 did not (Fig. 2). Furthermore, two singlebase substitutions were found at 258 and 262. Theclones, stnt-5′R20b and 20-2R, had C and A atthose positions, respectively, while the remainingclones, stnt-5′R1, A1-1R and Aall-1R, had T andG instead of C and A. Those differences lead toamino acid substitutions of arginine (258) andlysine (262) in the former to tryptophan and argi-nine in the latter, respectively. Such nucleotidesubstitutions are also seen in the 5′-region of hu-man sTnT cDNA. Two human sTnT cDNA clones,M1 and MSL-2-27, have G and one clone, H22h,has C at the position 117, resulting in the aminoacid changes from glutamic acid in the M1 andMSL-2-27 to aspartic acid in H22h (Gahlmann etal., ’87; Samson et al., ’94). However, we could notfind the sequence corresponding to the 33 nt in-sert reported in the 5′-region of human sTnTcDNA (Gahlmann et al., ’87).

Therefore, the determination of sequencesshowed that the two kinds of transcripts, one hav-ing the codon for alanine and C (258) and A (262)and the other lacking the codon and having T (258)and G (262), were expressed in chicken slow skel-etal muscle. The results are consistent with theabove data from S1 nuclease mapping, showing

Fig. 1. S1 nuclease mapping of sTnT mRNA with anantisense probe from the 5′-region of sTnT cDNA. The up-per panel shows schematically the structure of sTnT mRNAand the position of the 690 nt antisense probe (probe A) de-rived from the cDNA clones, stnt-5′R1 and sTnT-a3, withthe 32P-labeled 3′-end (closed circle). Total RNA from adult(Ad) chicken ALD, 1-day-old chick (H1) ALD or adult pecto-ralis major was hybridized to the probe A and digested withS1 nuclease. Protected fragments were separated on the 6%acrylamide/8 M urea gel. No band was detected in fast skel-etal muscle, pectoralis major (PM), but three bands of 690nt, 558 nt and about 440 nt were found in slow skeletalmuscle, ALD (A).

152 I. YONEMURA ET AL.

that the ca. 440 nt fragment was generated by di-gestion at 258 and/or 262 and the other fragmentof 558 nt by digestion at 135–137. Therefore, in-complete digestion at 258 and 262 might have oc-curred to produce the 558 nt fragment, becauseS1 nuclease recognizes normally more than duplexsubstitutions. Those results showed the heteroge-neity in the 5′-region of chicken sTnT mRNA.

Heterogeneity in the 3¢-region ofchicken sTnT mRNA

To search for heterogeneity in the 3′-region ofchicken sTnT mRNA, we performed S1 nucleasemapping of total RNA from adult ALD with a 658nt PCR-produced antisense probe (probe B) whichincluded the 6 nt restriction enzyme site (Xho Ior EcoR I) and 2 nt optional sequence at each end.The major protected fragment was 642 nt long,showing that the majority of transcripts matched

completely with the probe B (Fig. 3). In addition,minor and smear bands of faster mobilities werefound. Similar pattern was detected in gastrocne-mius of the 12-day-old embryo (data not shown).In order to characterize those minor protectedfragments, we sequenced RT-PCR products of to-tal RNA from adult ALD and complexus (Fig. 4).We found four single base substitutions in the 3′-region of sTnT cDNA at 703 (T or C), 774 (C orT), 797 (C or T), and 827 (G or A). The substitu-tions at 703 and 774 cause the amino acid changesfrom leucine in the five clones (b20-10, bA3, bA1,b20-4, and b20-3) to proline in the other two clonesand from arginine in the two clones (b20-1- andb20-3) to cysteine in the five remaining clones,respectively. Two single base substitutions at 797and 827 do not cause amino acid changes. Theresults showed that chicken slow skeletal musclehad at least four kinds of transcripts in the 3′-

Fig. 2. Nucleotide and deduced amino acid sequences of the5′-region of chicken sTnT cDNA. The cDNA was generated byRT-PCR and 5′-RACE with total RNA from chicken adult ALDand complexus. Two RT-PCR products from ALD, A1-1R andAall-1R, and one 5′-RACE product from complexus, 20-2R, weresequenced. The remaining stnt-5′R1 and stnt-5′R20b were de-scribed previously (Yonemura et al., ’96). The bracket and dashes

(-) indicate the absence of GCA coding for alanine and gaps,respectively. The amino acid sequences are indicated in thesingle-letter code above the nucleotide sequences. Two substitu-tions at 258 and 262 are marked with asterisks (*) with thecorresponding amino acid residues. Only single sequence isshown in 45 to 104, 165 to 224, and 285 to 344 (designated ascom.), because the sequence was identical among the clones.

CHICKEN SLOW TROPONIN T mRNA 153

region. In the 3′-region of human sTnT cDNA,single base substitutions are also seen at positions648 (G or C) and 649 (C or G) (Novelli et al., ’92).These substitutions cause the amino acid changesfrom lysine to asparagine (648) and from leucineto valine (649), respectively. However, we couldnot find the sequence corresponding to the 48 ntinsert reported in the 3′-region of human sTnTcDNA (Gahlmann et al., ’87).

Structure of the 3¢-region of thechicken sTnT gene

To investigate how the heterogeneity in thechicken sTnT transcripts is generated, we deter-

mined the partial sequence of the chicken sTnTgene. The size of the chicken sTnT gene was onlyabout 3.5 kb (data not shown), but we encoun-tered serious difficulties in sequencing its 5′-re-gion possibly due to extensive G-C tracts inintrons. The 3′-region of the gene was composedof seven exons with the last exon including thestop codon and polyadenylation signal (Fig. 5A).We tentatively designated the exons as exon A toG. There was no alternative exon. Each one ofthe variable nucleotides at the four positions inthe transcripts was found in exon E (T at 703),exon F (C at 774) and exon G (C and G at 797and 827), suggesting that the 3′-heterogeneity inthe transcripts might be derived from allelic varia-tion. We confirmed the absence of the sequencecorresponding to the 48 nt insert reported in the3′-region of human sTnT cDNA (Gahlmann et al.,’87), showing one major difference in the genomicstructure between chicken and human sTnTgenes. On comparison of the chicken sTnT genewith the quail fTnT gene (Bucher et al., ’89), ex-ons A, B, C, D, E, F, and G in the former hadsequence homology to exons 11, 12, 13, 14, 15, 17and 18 in the latter, respectively (Fig. 5B). How-ever, the length from exons A to G (1547 bp) wasconsiderably shorter than that from exons 11 to17 (about 7.5 kb) due to highly small introns (84to 329 bp) in the chicken sTnT gene.

DISCUSSIONWe found the two kinds of transcripts in the

5′-region, one having the codon for alanine (135–137) and C (258) and A (262) and the other lack-ing the codon and having T and G instead of Cand A. In the 3′-region, we found four singlebase substitutions at 703 (T or C), 774 (C or T),797 (C or T), and 827 (G or A). Four out of sixsubstitutions in the 5′- and 3′-regions lead toamino acid changes in chicken sTnT isoforms.Such nucleotide substitutions are also seen inhuman sTnT (Gahlmann et al., ’87; Samson etal., ’90; Novelli et al., ’92; Samson et al., ’94).However, no single substitution has been re-ported in fTnT and cTnT cDNAs so far (Gar-finkel et al., ’82; Medford et al., ’84; Breitbartet al., ’85; Cooper and Ordahl, ’85; Smillie etal., ’88; Bucher et al., ’89; Jin and Lin, ’89; Jinet al., ’92; Morgan et al., ’93; Wu et al., ’94;Briggs et al., ’94; Schachat et al., ’95; Wang andJin, ’97) with two exceptions in human fetal andpathological cTnT cDNA (Mesnard et al., ’93, ’95;Townsend et al., ’95).

How is the heterogeneity in chicken sTnT then

Fig. 3. S1 nuclease mapping of sTnT mRNA with anantisense probe from the 3′-region of sTnT cDNA. The upperpanel shows schematically the structure of sTnT mRNA andthe position of the 658 nt antisense probe (probe B) gener-ated by PCR. The probe included the 6 nt restriction enzymesite and 2 nt optional sequence at each end and was 32P-labeled internally. Total RNA from adult chicken ALD (A) washybridized to the probe B and digested with S1 nuclease, andprotected fragments were separated on the 6% acrylamide/8M urea gel. The major (642 nt) and minor protected frag-ments (bracket) were found.

154 I. YONEMURA ET AL.

generated? In the 5′-region, we can consider threepossible mechanisms. First, two exons covering135 to 262 may exist, one having the codon foralanine (135–137) and C (258) and A (262) andthe other lacking the codon and having T and G.Two kinds of sTnT transcripts can be generatedby mutually exclusive alternative splicing of thetwo exons. Secondly, when the variable nucleotidesin the 5′-region are included in more than twoexons, inclusion of one exon may be tightly linkedto that of the other(s) resulting in production ofonly two transcripts. Therefore, alternative exonsof each set can be synchronously spliced. Thirdly,we can not rule out completely that two kinds oftranscripts may be derived from different allelesof this gene.

From the cDNA sequences and sTnT gene struc-ture, we showed that four single base substitu-

tions in the 3′-region might be derived from al-lelic variation. Since human sTnT cDNAs havealso several single base substitutions and dele-tions in the 3′-region (Samson et al., ’90; Novelliet al., ’92), it seems likely that the 3′-region ofthe sTnT gene could accumulate mutations, pos-sibly suggesting the moderate structural con-straints on the COOH region of the sTnT protein.Our data also showed the structural differencebetween chicken and human sTnT genes, becausethe chicken sTnT gene did not have the sequencecorresponding to the 48 nt insert reported in the3′-region of human sTnT cDNA (Gahlmann et al.,’87). The structural differences between avian andmammalian genes have also been reported in fTnT(Miyazaki et al., unpublished data; Breitbart andNadal-Ginard, ’86) and cTnT (Cooper and Ordahl,’85; Jin et al., ’92).

Fig. 4. Nucleotide and deduced amino acid sequences ofthe 3′-region of chicken sTnT cDNA. The cDNA was gener-ated by RT-PCR with total RNA from adult chicken ALD andcomplexus. Two RT-PCR products from ALD, bA1 and bA3,and five RT-PCR products from complexus, b20-1, b20-3, b20-4, b20-9, and b20-10, were sequenced. Four single base sub-stitutions (marked with asterisks, *) are seen at 703 (T or

C), 774 (C or T), 797 (C or T), and 827 (G or A). The substitu-tions at 703 and 774 cause amino acid changes from leucineto proline and from arginine to cysteine, respectively, but thoseat 797 and 827 do not. The amino acid sequences are indi-cated above the nucleotide sequences and amino acid resi-dues are also shown below the sequences at the positionswhere amino acid changes are found.

CHICKEN SLOW TROPONIN T mRNA 155

LITERATURE CITEDAndreadis A, Gallego ME, Nadal-Ginard B. 1987. Genera-

tion of protein isoform diversity by alternative splicing:mechanistic and biological implications. Ann Rev Cell Biol3:207–242.

Breitbart RE, Nadal-Ginard B. 1986. Complete nucleotide se-quence of the fast skeletal troponin T gene. J Mol Biol188:313–324.

Breitbart RE, Nadal-Ginard B. 1987. Developmentally in-duced, muscle-specific trans factors control the differentialsplicing of alternative and constitutive troponin T exons.Cell 49:793–803.

Breitbart RE, Nguyen HT, Medford RM, Destree AT, MahdaviV, Nadal-Ginard B. 1985. Intricate combinatorial patternsof exon splicing generate multiple regulated troponin Tisoforms from a single gene. Cell 41:67–82.

Briggs MM, Maready M, Schmidt JM, Schachat F. 1994. Iden-tification of a fetal exon in the human fast troponin T gene.FEBS Lett 350:37–40.

Bucher EA, Charles de la Brousse F, Emerson CP Jr. 1989.Developmental and muscle-specific regulation of avian fastskeletal troponin T isoform expression by mRNA splicing.J Biol Chem 264:12482–12491.

Concordet J-P, Salminen M, Demignon J, Moch C, Maire P,Kahn A, Deagelen D. 1993. An opportunistic promoter shar-ing regulatory sequences with either a muscle-specific or aubiquitous promoter in the human aldolase A gene. Mol CellBiol 13:9–17.

Cooper TA, Ordahl CP. 1985. A single cardiac troponin T genegenerates embryonic and adult isoforms via developmentallyregulated alternate splicing. J Biol Chem 260:11140–11148.

Frohman MA, Dush MK, Martin GR. 1988. Rapid productionof full-length cDNAs from rare transcripts: amplificationusing a single gene-specific oligonucleotide primer. Proc NatlAcad Sci USA 85:8998–9002.

Gahlmann R, Troutt AB, Wade RP, Gunning P, Kedes L. 1987.Alternative splicing generates variants in important func-tional domains of human slow skeletal troponin T. J BiolChem 262:16122–16126.

Fig. 5. Structure of the 3′-region of the chicken sTnTgene. (A) The sequence of the 3′-region of the chickensTnT gene (1547 bp) was determined. It included sevenexons, tentatively designated as exons A to G. Capitaland lower-case letters indicate exon and intron se-quences, respectively. The amino acid sequences are in-dicated below the nucleotide sequences. Outlined lettersin exons E to G indicate the positions of four single basesubstitutions. The asterisk (*) indicates the stop codon.The bold underline indicates the putative polyadenyl-ation signal. Italic letters within thin underlines indi-cate the 5′ and 3′ splice sites. (B) Comparison of the3′-regions between chicken sTnT and quail fTnT genes(Bucher et al., ‘89). The size of the 3′-region of the formerwas considerably smaller than that of the latter. Theasterisks (*) indicate the positions of four single basesubstitutions.

156 I. YONEMURA ET AL.

Garfinkel LI, Periasamy M, Nadal-Ginard B. 1982. Cloningand characterization of cDNA sequences corresponding tomyosin light chains 1, 2, and 3, troponin-C, troponin-T, α-tropomyosin, and α-actin. J Biol Chem 257:11078–11086.

Jin J-P, Huang Q-Q, Yeh H-I, Lin JJ-C. 1992. Complete nucle-otide sequence and structural organization of rat cardiactroponin T gene. J Mol Biol 227:1269–1276.

Jin J-P, Lin JJ-C. 1989. Isolation and characterization ofcDNA clones encoding embryonic and adult isoforms of ratcardiac troponin T. J Biol Chem 264:14471–14477.

Medford RM, Nguyen HT, Destree AT, Summers E, Nadal-Ginard B. 1984. A novel mechanism of alternative RNAsplicing for the developmentally regulated generation oftroponin T isoforms from a single gene. Cell 38:409–421.

Mesnard L, Samson F, Espinasse I, Durand J, Neveux J-Y,Mercadier J-J. 1993. Molecular cloning and developmen-tal expression of human cardiac troponin T. FEBS Lett328:139–144.

Mesnard L, Logeart D, Taviaux S, Diriong S, Mercadier J-J,Samson F. 1995. Human cardiac troponin T: cloning andexpression of new isoforms in the normal and failing heart.Circ Res 76:687–692.

Morgan MJ, Earnshaw JC, Dhoot GK. 1993. Novel develop-mental regulated exon identified in the rat fast skeletalmuscle troponin T gene. J Cell Sci 106:903–908.

Novelli G, Gennarelli M, Zelano G, Sangiuolo F, Cicero SL,Samson F, Dallapiccola B. 1992. Polymerase chain reactionin the detection of mRNA transcripts from the slow skel-etal troponin T (TNNT1) gene in myotonic dystrophy andnormal muscle. Cell Biochem Funct 10:251–256.

Robert B, Daubas P, Akimenko M-A, Cohen A, Garner I,Guenet J-L, Buckingham M. 1984. A single locus in themouse encodes both myosin light chains 1 and 3, a secondlocus corresponds to a related pseudogene. Cell 39:129–140.

Samson F, Lee JE, Hung W-Y, Potter TG, Herbstreith M,Roses AD, Gilbert JR. 1990. Isolation and localization of aslow troponin T (TnT) gene on chromosome 19 by subtrac-tion hybridization of a cDNA muscle library using myotonicdystrophy muscle cDNA. J Neurosci Res 27:441–451.

Samson F, Mesnard L, Mihovilovic M, Potter TG, MercadierJ-J, Roses AD, Gilbert JR. 1994. A new human slow skel-etal troponin T (TnTs) mRNA isoform derived from alter-native splicing of a single gene. Biochem Biophys ResCommun 199:841–847.

Schachat F, Schmidt JM, Maready M, Briggs MM. 1995.Chicken perinatal troponin Ts are generated by a combina-tion of novel and phylogenetically conserved alternativesplicing pathways. Dev Biol 171:233–239.

Schiaffino S, Reggiani C. 1996. Molecular diversity of myo-fibrillar proteins: gene regulation and functional signifi-cance. Phys Rev 76:371–423.

Smillie LB, Golosinska K, Reinach FC. 1998. Sequences ofcomplete cDNAs encoding four variants of chicken skeletalmuscle troponin T. J Biol Chem 263:18816–18820.

Townsend PJ, Barton PJR, Yacoub MH, Farza H. 1995. Mo-lecular cloning of human cardiac troponin T isoforms: ex-pression in developing and failing heart. J Mol Cell Cardiol27:2223–2236.

Wang J, Jin J-P. 1997. Primary structure and develop-mental acidic to basic transition of 13 alternativelyspliced mouse fast skeletal muscle troponin T isoforms.Gene 193:105–114.

Wu Q-L, Jha PK, Raychowdhury MK, Du Y, Leavis PC, SarkarS. 1994. Isolation and characterization of human fast skel-etal β troponin T cDNA: comparative sequence analysis ofisoforms and insight into the evolution of members of amultigene family. DNA Cell Biol 13:217–233.

Yao Y, Nakamura M, Miyazaki J-I, Kirinoki M, Hirabaya-shi T. 1992. Expression pattern of skeletal muscle tro-ponin T isoforms is fixed in cell lineage. Dev Biol151:531–540.

Yonemura I, Watanabe T, Kirinoki M, Miyazaki J-I, Hira-bayashi T. 1996. Cloning of chicken slow muscle troponin Tand its sequence comparison with that of human. BiochemBiophys Res Commun 226:200–205.

Zot AS, Potter JD. 1987. Structural aspects of troponin-tro-pomyosin regulation of skeletal muscle contraction. Ann RevBiophys Chem 16:535–559.