A Cardiac Troponin T Epitope Conserved Across Phyla*

Troponin T is a thin filament protein that is impor- tant in regulating striated muscle contraction. We have raised a monoclonal antibody against rabbit cardiac troponin T, monoclonal (mAb) 13-1 1, that recognizes its epitope in cardiac troponin T isoforms from fish, bird, and mammal but not from frog. The number of these isoforms expressed in cardiac muscle varies among species and during development. Cardiac troponin T isoforms were not found in adult skeletal muscle, while they were expressed transiently in immature skeletal muscle. We have mapped the epitope recog- nized by mAb 13-11 using rabbit cardiac troponin T isoforms. Analysis of stepwise cyanogen bromide digestion, which allowed association of the epitope to regions spanning methionine residues, coupled with immunoactivity of synthetic peptides, corresponding to sequences containing methionine residues, indicated that mAb 13-11 recognized its epitope in a 17-residue sequence containing the methionine at position 68, SKPKPRPFPNLVPPKI. Comparison of skeletal and cardiac troponin T sequences suggested that the epi- tope was contained within the sequence FWNL- VPPKI. Synthetic peptides PFMPNLVPPKI and FEPNLVPPKI were recognized by mAb 13-11 on slot-blots. Enzyme-linked immunosorbent assay demonstrated mAb 13-11 recognized, in order of descend-ing

Troponin T is a thin filament protein that is important in regulating striated muscle contraction. We have raised a monoclonal antibody against rabbit cardiac troponin T, monoclonal (mAb) 13-1 1, that recognizes its epitope in cardiac troponin T isoforms from fish, bird, and mammal but not from frog. The number of these isoforms expressed in cardiac muscle varies among species and during development. Cardiac troponin T isoforms were not found in adult skeletal muscle, while they were expressed transiently in immature skeletal muscle. We have mapped the epitope recognized by mAb 13-11 using rabbit cardiac troponin T isoforms. Analysis of stepwise cyanogen bromide digestion, which allowed association of the epitope to regions spanning methionine residues, coupled with immunoactivity of synthetic peptides, corresponding to sequences containing methionine residues, indicated that mAb 13-11 recognized its epitope in a 17-residue sequence containing the methionine at position 68, SKPKPRPFPNLVPPKI. Comparison of skeletal and cardiac troponin T sequences suggested that the epitope was contained within the sequence F W N L -VPPKI. Synthetic peptides PFMPNLVPPKI and FEPNLVPPKI were recognized by mAb 13-11 on slot-blots. Enzyme-linked immunosorbent assay demonstrated mAb 13-11 recognized, in order of descending affinity, the 17-, 11-, and 10-residue sequence. Preabsorption of mAb 13-11 with each of these sequences blocked the recognition of the 17-residue peptide by mAb 13-1 1. The domain, PFMPNLVPPKI is encoded by the 6' region of the cardiac gene exon 10 and is present in hearts across a broad range of phyla. These findings suggest that this cardiac troponin Tspecific sequence confers onto myofilaments structural and functional properties unique to the heart. Troponin T is a thin filament regulatory protein of striated muscle that binds to troponin C and I and to tropomyosin and is required for calcium-dependent ATPase activity of myofibrillar proteins (1-3). Troponin Ts of cardiac, fast skeletal, and slow skeletal muscles are encoded by different genes containing regions of both high and low homology (4-6). The functional roles of polypeptides encoded by non-homologous regions and the consequent biophysical and biochemical properties they confer on to individual muscle types are not known.
* This work was supported in part by National Institutes of Health Grants HL 37358, HL 42250, HL-20749, and RR-02722. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
** To whom correspondence should be addressed Dept. of Pediatrics, P. 0. Box  Structural differences among troponin T isoforms of different muscle types have been proposed to be functionally significant (5, 7), and correlations have been made between troponin T isoforms, expressed in a given species and muscle type, and biochemical and biophysical properties of myofilaments and myofibrillar proteins reconstituted in vitro (8)(9)(10)(11).
As a first step in understanding how variations in troponin T structure contribute to the functional diversity of cardiac and skeletal muscle, we identified an antigenic troponin T epitope specific to cardiac muscle and have found that the epitope contains an intron-exon boundary encoded by exon 10 of the cardiac troponin T gene unique to cardiac troponin T isoforms. We followed the phylogenetic expression of this epitope in fish, frog, bird, and mammal and found that it is expressed across species being absent only in the frog. In evaluating the ontogenetic expression of this epitope in developing muscle, we found that it is present in developing heart of chicken and mammal and is expressed transiently in developing skeletal muscle of these species. The persistence of this epitope constituting a structural difference between cardiac and skeletal troponin T isoforms across such a broad range of phyla suggests that this domain confers onto cardiac myofilaments an important functional property.

MATERIALS AND METHODS
Source and Preparation of Muscle-Ventricular myocardium and thigh muscle were obtained from fetal and adult sheep, fetal, newborn, and adult rabbit, embryonic and adult chicken, newborn and adult dog, juvenile pig, adult Wistar-Kyoto rat, bullfrog, goldfish, guinea pig, and swamp sparrow following euthanasia with intravenous or intraperitoneal pentobarbital (300 mg/kg) or decapitation. Part of the muscle was frozen immediately in liquid nitrogen and stored in liquid nitrogen until used in myofilament preparations (12). The remainder of the specimen was either detergent extracted (13) and placed in sample buffer or placed directly in sample buffer and stored at -70 "C. In control experiments which assessed the effects of proteolysis on troponin T isoform numbers, myocardium was left at room temperature in the absence of protease inhibitors, and pieces were removed and placed in sample buffer after 30 min, 1, 2, 3, 4, 8, and 24 h at room temperature. These samples were examined with sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE)' and Western blots in a manner similar to experimental tissue. Rabbit cardiac and fast skeletal muscle were also prepared for immunocytochemical studies, using fluorescent light and electron microscopy (14-16, see below). Purified rabbit cardiac troponin T isoforms were obtained from rabbit myocardium by resolving individual isoforms with SDS-PAGE and transblotting them on to Immobilon or nitrocellulose membranes.
SDS-PAGE and Western Blots-Myocardial and skeletal muscle proteins were resolved by SDS-PAGE (17) on 6.5 and 7.5% polyacrylamide gels. Cyanogen bromide (CNBR) cleavage fragments were resolved on 12 and 20% SDS gels. Transblotting of proteins on to nitrocellulose and Immobilon membranes and Western blots (18) were executed as described previously (13). The sources of the reagents were as described previously (13).
Troponin T expression was examined by probing Western blots with a cardiac-specific monoclonal antibody, mAb 13-11, we raised against a rabbit cardiac troponin T isoform (TnTIR) (see below) (13, 19), or with JLT-12 which recognizes its antigenic determinants in fast skeletal muscle (20) and cardiac troponin T isoforms in rat and rabbit (13,21). The latter antibody is commercially available through Amersham Corp.
Production and Characterization of mAb 13-11-BALB/c mice were immunized with a rabbit cardiac troponin T isoform TnT4R, purified by SDS-PAGE as previously described (13). Spleen cells from a mouse whose serum recognized rabbit cardiac troponin T were fused with P3X63Ag.8.6-5.3 mouse myeloma cells as previously described (22). Hybridomas were screened by ELISA, using purified rabbit cardiac troponin T as antigen. Hybridomas that produced anti-cardiac troponin T antibody were further screened by Western blots, selecting those that produced antibodies with high affinity to cardiac troponin T. Selected hybridomas were cloned by limited dilutions. The production of monoclonal antibodies by actively growing clones was detected by ELISA and Western blots. mAb 13-11 was selected because it demonstrated high affinity and specificity to its determinant in cardiac troponin T. mAb 13-11 was isotyped using a IgG subclass kit from Binding Site Inc. and found to be an IgG1.
Immunocytochemistry-Localization of the epitope recognized by mAb 13-11 in cardiac myofilaments was undertaken in rabbit left ventricular myocardium by fluorescent light and electron microscopy. Recognition by mAb 13-11 of its fixed antigenic determinants was confirmed by fluorescent microscopy. Small pieces (-1 mm3) of left ventricular muscle were excised and fixed in 1% paraformaldehyde in 0.1 M phosphate-buffered saline and 10% sucrose. Sections, 3-5 pm in thickness, were obtained with a cryostat (Harris), preblocked in normal goat serum and incubated in mAb 13-11, as a primary antibody, for 1 h (room temperature). After extensive washing, a fluorescein-labeled goat anti-mouse antibody was used as a secondary antibody. For ultrastructural immunocytochemistry,2post-embedding techniques were performed (14, 15). mAb 13-11 was used as the primary antibody and 5-nm gold-labeled goat anti-mouse antibody as an electron dense secondary antibody probe (Janssen, Life Science Products, Belgium). Briefly, pieces of tissue (1-2 mm3) were excised from hearts immediately following removal from the animal or following perfusion of the coronaries with a "relaxing" solution, (140 mM NaC1, 4 mM KCl, 1 mM MgC12, 5.6 mM dextrose, 10 mM MOPS, 10 mM EGTA, pH 7.4), at a pressure of 30 cm of HzO to obttin sarcomeres with longer I bands. The myocardial blocks were immersed immediately in 1% paraformaldehyde in 0.1 M phosphate-buffered saline, pH 7.4, 10% sucrose, and cryoprotected by gradual infiltration with increasing concentrations of sucrose (up to 2.3 M sucrose in phosphate-buffered saline). Small blocks were plunge-frozen in liquid freon in liquid nitrogen. Frozen sections, 70-nm thick, were cut on a Reichert-Jung cryoultramicrotome. The sections were collected on formvar-carbon-coated nickel grids. Free aldehydes were blocked with 0.1% glycine in Gey's buffered saline solution (GIBCO), pH 7.3, before incubation for 10 min in 2% gelatin, 5% bovine serum albumin, and 5% normal goat serum in Gey's buffer. The grids were then exposed to mAb 13-11 as a primary antibody overnight at 4 "C used at a concentration of 50 pg/ml in 1:4 diluted Gey's buffer. After extensive washing in Gey's buffer, the grids were incubated for 2 h (room temperature) with 5-nm gold particles conjugated to goat antimouse. Following extensive washing, postfixation in 2% glutaraldehyde was done in 0.1 M sodium cacodylate buffer containing 0.1 M sucrose and 0.5% osmium tetroxide. Grids were stained with 0.5% aqueous uranyl acetate and counterstained with Reynold's Lead Citrate (16).
Cyanogen Bromide Digestion of Troponin T-Cardiac troponin T isoforms, TnT2R, TnT3R, and TnT4R (13, see Fig. 1, p a n e l B) from adult myofibrils were resolved by SDS-PAGE and transblotted onto Immobilon membrane. The membrane was cut out to separate TnTzR and TnT3R from T n T 4~. Cleavage of isoforms into polypeptides was done at methionine residues with CNBR. Complete or partial cleavage was achieved by varying the concentration of CNBR and the incubation time. Briefly, immobilized troponin T isoforms were incubated separately at room temperature in the dark in CNBR (50 mg/ml) in acetonitrile (CH3CN) for 19.5 h for partial cleavage of troponin T.
Specimens were sent to Micromed, a national resource center for intermediate voltage electron microscopy at Bowman Gray School of Medicine.
Complete cleavage was achieved by increasing the incubation time to 30 h. Fragments from the supernatant and those eluted in 40% CH&N in HzO were pooled as a dried precipitate and washed extensively with distilled water. They were resolved by SDS-PAGE and transblotted onto nitrocellulose or Immobilon membranes. The polypeptide fragments containing the antigenic epitope were identified on Western blots using mAb 13-11. The smallest cleaved polypeptide fragment of troponin T4R that contained the intact antigenic determinant was sent to Yale University School of Medicine Protein and Nucleic Acid Chemistry Facility for direct sequencing off Immobilon.
Synthetic Peptides-Since we found that complete CNBR cleavage of rabbit troponin T destroyed the epitope recognized by mAb 13-11, we anticipated that the epitope encompassed a methionine residue. Three peptides (A, SKPKPRPFMPNLVPPKk B, DDIHRKRMEK-DLNELQ; and C, KKALSNM-HFGGYIQKQA) were synthFsized as potential antigens by the Protein Chemistry Laboratory of the University of North Carolina at Chapel Hill-National Institute of Environmental Health Sciences. The sequences of the polypeptides correspond to regions that span methionine residues of the published amino acid sequence of a rabbit cardiac troponin T isoform (23). Using these peptides on slot-blots as antigenic candidates, mAb 13-11 identified a single peptide (A) containing its epitope. Specificity of this sequence was confirmed by preabsorption experiments of mAb 13-11 with each of the three synthetic peptides. Two shorter peptides, (D) PFMPNLVPPKI and (E) FMPNLVPPKI, which correspond to shorter sequences of peptide A, were synthesized and used on slotblots and ELISA to test for nonessential residues. Alkaline phosphatase-nitrophenyl phosphate was used as colorimetric signal to quantitate the ELISA. Signals from 96-well microtiter plates were read at a wavelength of 405.

Western Blots of Cardiac and Skeletal Myofibrillar Preparations
Cardiac Preparations-mAb 13-11 recognized its epitope on Western blots in all mammalian and avian species and the goldfish (Fig. l), but not in the frog. Troponin T isoforms in the cardiac myofilaments from the different species had electrophoretic mobilities similar to those of the rabbit cardiac troponin T isoforms purified from rabbit myocardium (13, 19). JLT-12 did not recognize any additional proteins.
Chicken heart demonstrated predominantly two troponin T isoforms in embryonic myocardium, but only the isoform with the faster electrophoretic mobility was present in adult myocardium ( Fig. 1, panel A ) . This maturational profile is consistent with findings from chicken cardiac troponin T cDNA and gene analysis reported by Cooper and Ordahl (4). Adult swamp sparrow contained one dominant isoform with an electrophoretic mobility similar to that of the dominant isoform in adult chicken heart ( Fig. 1, panels A and C). A single isoform was present in goldfish heart (Fig. 1, panel C).
Two troponin T isoforms were found in adult rat myocardium ( Fig. 1, panel C) consistent with a recent report by Schiaffino et al. (24). Fetal and adult sheep heart demonstrated a single dominant isoform and a very small amount of a second isoform with a faster electrophoretic mobility (Fig.  1, panel A ) .
In contrast to the relatively simple expression of troponin T described above, rabbit, dog, pig, and guinea pig hearts had four or more isoforms (Fig. 1, panels A X ) . The relative proportions of these isoforms expressed in the muscle changed with development (e.g. Fig. 1, panels A and B). In the rabbit, the abundance of the two isoforms with the slowest mobilities markedly decreased during postnatal life, with the isoform with the apparent largest M, becoming essentially absent from adult myocardium. JLT-12 did not identify in fetal or newborn rabbit heart its epitope in proteins with the relatively faster electrophoretic mobilities of fast skeletal muscle troponin T isoforms (20). Six isoforms were identified by mAb 13-11 in adult dog myocardium (Fig. 1, panel A ) . The  The two size markers are from Sigma: egg albumin, 45,000 Da and glyceraldehyde-3-phosphate dehydrogenase, 36,000 Da. Panel B shows that mAb 13-11 recognizes its epitope in rabbit fetal skeletal muscle but not in adult rabbit skeletal muscle. The two other lanes were loaded with fetal and adult rabbit cardiac preparations to illustrate the developmental changes in rabbit cardiac troponin T expression. The rabbit cardiac troponin T isoforms, recognized by mAb 13-11, were named TnTIR, TnT2R, TnT3R, and T n T 4~ based on their electrophoretic mobilities, as was previously done (13). T n T 4~ has the fastest and TnTIR the slowest mobilities. mAb 13-11 was raised against T n T 4~. These proteins have the same electrophoretic mobilities as those of isoforms purified from rabbit myocardium (13,19). Panel C shows that mAb 13-11 recognizes its epitope in juvenile pig, adult rat, adult guinea pig, goldfish, and adult swamp sparrow (Melospizageorgiana). Panel D shows Western blots, probed with mAb 13-11, of two-dimensional gels, 6.5% polyacrylamide concentration, loaded with cardiac preparations from adult rat, adult guinea pig, 2-day postnatal rabbit, and adult dog. The more basic part of the gel is to the right. In a study of human cardiac troponin T, we found that the more acidic protein in pairs of troponin T spots with the same M, was phosphorylated (27).
with the slowest electrophoretic mobilities were essentially the only ones in newborn dog heart (Fig. 1, panel A ) . Pig myocardium contained four isoforms (Fig. 1, panel C). Similarly, guinea pig myocardium demonstrated four isoforms with electrophoretic mobilities similar to those of the pig and rabbit isoforms and a fifth protein with a faster mobility (Fig. 1, Western blots of two-dimensional polyacrylamide gels of rabbit, dog, guinea pig, and rat cardiac preparations demonstrated that the smaller the M, of the spots of troponin T containing mAb 13-11 epitope, the more basic they were (Fig.   1, panel D). The NH2-terminal region of troponin T contains many acidic residues and is a region in both cardiac and skeletal muscle where alternative splicing occurs. Our findings are consistent with the smaller more basic troponin T isoforms being generated by loss of acidic exons through alternative splicing, as proposed by Jin and Lin (25).
Skeletal Muscle-The epitope recognized by mAb 13-11 was absent from adult mammalian, avian, goldfish, and bullfrog skeletal muscle. In fetal rabbit skeletal muscle, mAb 13-11 recognized three isoforms with the same electrophoretic mobilities as the three cardiac troponin T isoforms predominantly expressed in fetal rabbit heart (TnTIR, TnT2R, and panel C).
TnTdR), one of these being the dominant adult rabbit cardiac isoform, TnT4R (Fig. 1, panel B ) . Interestingly, the relative proportions of the isoforms were similar in both fetal skeletal and fetal cardiac muscle. These proteins, recognized by mAb 13-11 and presumably of cardiac gene origin, made up a very minor fraction of the total amount of troponin T in skeletal muscle as estimated from immunoblots of rabbit fetal skeletal muscle probed with JLT-12. Embryonic chicken skeletal muscle also contained two proteins that comigrated with the two isoforms of embryonic heart (results not shown).

Proteolysis Control Experiments
Rabbit, dog, guinea pig, and chicken cardiac preparations were left at room temperature for up to 24 h in order to allow proteolysis to occur. Western blots performed on muscle from these preparations demonstrated that when compared to myocardium immediately placed in sample buffer, the proportion of proteins recognized by mAb 13-11 did not change with time at room temperature. No decrease in the relative amount of the largest isoforms to that of the smaller ones occurred with increased time at room temperature (results not shown), suggesting that the different bands recognized by mAb 13-11 on Western blots (see Fig. 1) are troponin T isoforms and not products of proteolysis.

Cardiac Troponin T-specific Epitope Immunocytochemistry
Recognition by mAb 13-11 of its fixed antigenic determinant was confirmed by fluorescent microscopy (Fig. 2a). Transverse fluorescent striations, seen by confocal laser scanning microscopy, corresponded to the periodicity of I bands in myofibrils.
Rabbit myocardium probed with mAb 13-11 and a goldlabeled secondary antibody and examined under the electron microscope demonstrated gold particles concentrated mostly over the I band (Fig. 2, b and c). Numerous but fewer particles were present over the A band with none over the Z and M bands. This localization of gold particles over the region of thin filaments, which extend from the Z disc into the A band, is consistent with the binding of troponin T onto tropomyosin.

Cyanogen Bromide Cleavage to Map the Epitope
When complete cleavage of rabbit cardiac troponin T isoforms, TnTZR, TnTnR, and T n T 4~, was achieved with CNBR the epitope recognized by mAb 13-11 was destroyed. On the other hand, when conditions were modified to yield incomplete cleavage of the individual isoforms, the antigenic determinant was preserved (Fig. 3, panel I, lanes a, c, and d). The fragments from TnT4R had molecular mass values of approximately 20 and 30 kDa. A mixture of TnTzR and TnT3R yielded "doublet" polypeptides that migrated to positions just above the 20-kDa fragment from the TnT4R cleavage (Fig. 3  Western blots of CNRR-cleaved fragments of purified rabbit cardiac troponin T isoformn (TnTZR. TnTDR, TnTdR. see Fig. 1 legend) (panel I ) . mAh 13-1 1 recognized its epitope in these rabbit cardiac troponin T isoforms when incomplete cleavage left residual fragments of approximately 20 and 30 kDa. Imnen a and c illustrate two CNRR digestions of TnT,R. In addition, mAh 13-11 recognized its determinants in incompletely cleaved fragments of TnTZR and TnTnR (lone d ) whose electrophoretic mobilities differed slightly, paralleling the differences between those of the uncleaved isoforms (see Fig. I, panel I I , lane c) (lanes a and b ) , R (lanes c and d ) , or C (lanes e andf). Preahsorption was done using two peptide concentrations, the lower concentration yielding the hlots in ~anes a. c, and e. Succesnful inhibition of the antihody-antigen reaction is illustrated in lanes a and h which confirms the specificity of mAb 18-11 to its epitope within peptide A. recognized by mAb 13-11 is contained within a region of troponin T which spanned a methionine residue: cleavage of 2 of 3 methionine residues in rahhit cardiac troponin T ( 2 3 ) leaves this epitope intact, while cleavage of the third residue destroyed the epitope. Given that mAh 13-11 recognizes its epitope in a denatured protein transferred from SDS-polyacrylamide gels onto nitrocellulose, this suggested to us that the epitope is most likely a linear sequence rather than a secondary or tertiary conformational structure of the protein (26).
Attempts at sequencing the amino acids in the 20-kDa T n T 4~ fragment failed repeatedly. We had previously found that the NH2 terminus of TnT4R was blocked to sequencing by Edman degradation (13). Pearlstone et al. (23) had also found the NH, terminus of rabhit cardiac troponin T to he Synthetic Peptide

Rabbit Cardiac
blocked. Taken all together these findings led us to suspect that the 20-kDa fragment may contain the NH2 terminus of the uncleaved troponin T molecule and that the epitope could be localized within the region spanning the first methionine residue of the molecule at position 68. Alternatively, we hypothesized that conformational changes, which might have taken place following CNBR cleavage, prevented NHP-terminal sequencing of the 20-kDa fragment.
Use of Synthetic Peptides to Characterize mAb 13-1 1 Determinant To test which methionine-containing region of cardiac TnT contained the epitope recognized by 13-11, three synthetic peptides were made (A-C, see "Materials and Methods"). They corresponded to three partial sequences of rabbit cardiac troponin T which span methionine containing regions (23).
These peptides were immobilized separately on Immobilon membrane and used as antigens on slot blots. mAb 13-11 recognized its epitope in peptide A, SKPKPRPFMPN-LVPPKI (Fig. 3, panel II, lane a). This sequence contains the methionine residue at position 68 in a rabbit cardiac troponin T molecule (see Fig. 4). When mAb 13-11 was preabsorbed using each of the three peptides, only preabsorption with peptide A inhibited its immunoreactivity (Fig. 3, panel ZIZ,   lanes a and b). Two shorter peptides, D and E (see "Materials and Methods") that lacked, respectively, the first 6 and 7 residues of peptide A were synthesized and used in immuno-* S K P K P -R P F K P N L V P P K I * E S K P K P -R P F K P N L V P P K I P 77 1 XQ * Chicken Cardiac E S K P K P -K P F K P N L V P P K I P 94 Rat Cardiac P S K P K P S R L F K P N L V P P K I P 93 Bovine Cardiac E E K P K P -K P F K P N L V P P K I P 77 Sheep Cardiac E S K P P 4 P E P F K P N L V P P K I P 80 * *

Chicken Fast Skeletal E & K P E --------L l h P K I P 71
Rat Fast Skeletal E E K P B P -K -----L T 4 P K I P 50 2 2 Human Slow Skeletal E E B P K P S R P Y Y P E L I P P K I P 56

FIG. 4.
Comparison of cardiac and skeletal muscle troponin T sequences with synthetic peptide A containing the epitope recognized by mAb 13-11 (see Fig. 3 logical assays of mAb 13-11. Immunoassays consisting of slotblots and ELISA showed that both peptide D and E were recognized by mAb 13-11, but the intensity of the signals produced by peptides D and E were less than that produced by peptide A. Coomassie staining of Immobilon sheets on which the peptides were immobilized demonstrated that peptides D and E were much less adherent to this surface than the longer parental peptide A, potentially explaining the smaller colorimetric signals produced by peptides D and E on slot-blots. On ELISAs peptide D had a signal that was 10-25% of the peptide A signal while peptide E had a smaller signal, which was still above background (background was the same as the signal produced by peptide B that is not recognized by mAb 13-11). Preabsorption of mAb 13-11 with peptide D or E blocked the recognition by mAb 13-11 of the parental peptide A. These results suggest that the sequence PFMPNLVPPKI, present in peptide D, contains critical residues of the epitope and that the proline in position 66, absent from peptide E, may be important in the formation of the epitope.

Molecular Basis of Epitope
A search in the EMBLbank, Genbank, and the Protein Identification Resource (Protein-Swiss) data bases demonstrated that the synthetic peptide containing the epitope recognized by mAb 13-11 was best matched by cardiac troponin T of the rat, chicken, cow, and rabbit (Fig. 4). Although the search did not yield a match with fast skeletal muscle troponin, some homology was found with slow skeletal muscle troponin T (see Fig. 4). Comparison of protein sequences obtained directly or derived from cDNAs of sheep, rat, rabbit, cow, and chicken cardiac troponin T demonstrated that a portion of the sequence in peptide A, namely PFMPNLV-PPKI, is identical and ubiquitous among cardiac troponin T isoforms of all species (a single substitution, leucine for proline at the amino terminus, is present in the rat). Analysis of the chicken cardiac troponin T gene (4) and that of rat fast skeletal muscle troponin T (6) indicates that 5 residues within the sequence, namely PFMPN, are encoded by the 5' region of cardiac troponin T exon 10 and that these residues are missing from exon 10 of the fast skeletal muscle troponin T gene. Examination of protein sequences obtained directly or derived from cDNAs of fast skeletal muscle troponin T of mammalian and avian species shows that these 5 residues are absent from all skeletal troponin T (Fig. 4). Taking into consideration the immunoreactivity of mAb 13-11 with cardiac muscle (Fig. l), lack of reactivity with fast and slow skeletal muscle (27, Fig. l), the optimal sequence alignment of the skeletal muscle isoforms with cardiac isoforms (Fig. 4), the missing residues (PFMPN) from fast skeletal troponin T (Fig. 4), the contrasting 5 residues that precede the highly conserved COOH-terminal sequence of the synthetic peptide in slow skeletal muscle (Fig. 4), the differences among cardiac troponin T isoform sequences comparable to the NH2-termi-na16 residues of the synthetic peptide (Fig. 4), the immunoreactivity of peptides A, D, and E, and the recognized occurrence of epitopes at intron-exon boundaries (28, 29), we conclude that the critical residues recognized by mAb 13-11 most likely lie within the sequence delineated by PFMPNLVPPKI.

DISCUSSION
We have raised a monoclonal antibody against a rabbit cardiac troponin T epitope and found that this epitope is conserved across phyla. The transient expression of this epitope in developing skeletal muscle from rabbit and chicken is

Cardiac Troponin
T-specific Epitope consistent with Ordahl's results (30) which showed that cardiac proteins are ontogenically expressed in developing skeletal muscle during fetal and embryonic life. In contrast, the expression of cardiac isoforms in skeletal muscle during such fetal and embryonic development was not accompanied by expression of fast skeletal muscle troponin T in the heart. A comparison of protein sequences of rabbit and bovine cardiac troponin T and those deduced from cardiac troponin T cDNA of chicken, sheep, and rat demonstrates that a region which corresponds to the sequence PFMPNLVPPKI, encoded by the 5' portion of exon 10, is highly conserved among cardiac troponin T of these species (see Fig. 4). The recognition by mAb 13-11 of its epitope in cardiac troponin T isoforms in these various species and the apparent large amount of variability present among the 6 residues encoded by exon 9 (Fig.  4) that make up the NH2-terminal of peptide A argue against including within the epitope the first 6residues of peptide A. The immunoreactivity of peptide D with mAb 13-11 indicates that critical residues constituting the epitope are continued within the sequence PFMPNLVPPKI. The presence of this epitope in mammalian, avian, and piscine cardiac muscle suggests that this domain of cardiac TnT encoded by exon 10 was introduced early during vertebrate evolution.
The lack of immunoreactivity of adult skeletal muscle with mAb 13-11 can be explained by comparing cardiac and fast skeletal muscle troponin T genes (see Refs. 4 and 6 ) and sequences of cardiac and skeletal muscle troponin T isoforms derived from cDNAs or obtained directly (Fig. 4). The structure of exon 10 of the rat fast skeletal muscle troponin T gene differs markedly from that of the cardiac gene in that the first 15 nucleotides of the cardiac exon 10 are absent from skeletal muscle exon 10 (4, 6). The first 5 residues encoded by exon 10 of the cardiac gene are present in all mammalian and avian troponin T isoforms and are absent from fast skeletal muscle troponin T isoforms of mammal and bird (Fig. 4). Slow skeletal muscle troponin T contains 5 residues with some homology to the sequence encoded by the 5' region of exon 10 of the cardiac troponin T gene (see Fig. 4). These structural differences among skeletal and cardiac troponin T isoforms are consistent with our finding that mAb 13-11 recognizes its epitope in all mammalian and avian cardiac troponin T isoforms and that this epitope is absent from adult fast and slow skeletal muscle (27, Fig. 1).
Thus, we have mapped the epitope to a domain encoded by a single exon and shown that this epitope is a sequential one (26). These residues are encompassed in a sequence of rabbit cardiac troponin T where all prolines reside, namely a stretch of 26 amino acids encoded by exons 8-10 (e.g. amino acids This proline-rich region falls within the region of troponin T that binds strongly to tropomyosin in a calcium insensitive manner in skeletal muscle and is just downstream from the hypervariable NHz-terminal region of troponin T (25,31-33). The concentration of prolines in this region has been noted to be characteristic for all troponin T isoforms (25). A major difference between proline-rich regions of cardiac and fast skeletal muscle troponin Ts is the absence in fast skeletal muscle troponin T of 5 residues including 2 of the prolines in this region. Based on this observation and our results which show that mAb 13-11 recognizes its epitope within this region 62-77). in its native fresh or fixed and denatured states, it is tempting to speculate that this region of cardiac troponin T may be an exposed flexible domain of the molecule important for function. Alterations in the sequence of this domain could affect the binding of troponin T to tropomyosin, and the sensitivity of the myofilaments to calcium, and shed light on the specificity of the interactions of troponin T with tropomyosin in different muscle types.
Conservation of the region encoded by exon 10 of the cardiac troponin T gene over millions of years of evolution, its absence in skeletal muscle troponin T isoforms, and the functional differences between cardiac and skeletal muscle suggest that this sequence confers on to cardiac myofilaments biophysical and biochemical properties unique to the heart.