Characterization of a Highly Conserved Human Homolog to the Chicken Neural Cell Surface Protein Bravo/Nr-CAM That Maps to Chromosome Band 7q31

The neuronal cell adhesion molecule Bravo/Nr-CAM is a cell surface protein of the immunoglobulin (Ig) superfamily and is closely related to the L1/NgCAM and neurofascin molecules, all of which contain six immunoglobulin domains, five fibronectin repeats, a transmembrane region, and an intracellular domain. Chicken Bravo/Nr-CAM has been shown to interact with other cell surface molecules of the Ig superfamily and has been implicated in specific pathfinding roles of axonal growth cones in the developing nervous system. We now report the characterization of cDNA clones encoding the human Bravo/Nr-CAM protein, which, like its chicken homolog, is composed of six V-like Ig domains and five fibronectin type III repeats. The human Bravo/Nr-CAM homolog also contains a transmembrane and intracellular domain, both of which are 100% conserved at the amino acid level compared to its chicken homolog. Overall, the human Bravo/Nr-CAM homolog is 82% identical to the chicken Bravo/Nr-CAM amino acid sequence. Independent cDNAs encoding four different isoforms were also identified, all of which contain alternatively spliced variants around the fifth fibronectin type III repeat, including one isoform that had been previously identified for chicken Bravo/Nr-CAM. Northern blot analysis reveals one mRNA species of approximately 7.0 kb in adult human brain tissue. Fluorescencein situhybridization maps the gene for human Bravo/Nr-CAM to human chromosome 7q31.1–q31.2. This chromosomal locus has been previously identified as containing a tumor suppressor candidate gene commonly deleted in certain human cancer tissues.     

Structure of a new nervous system glycoprotein, Nr-CAM, and its relationship to subgroups of neural cell adhesion molecules

We have identified and characterized a new glycoprotein in the chicken nervous system using immunological and molecular biological methods and we have examined its tissue distribution. Analysis revealed that this protein is very similar in structure to the chicken neuron-glia cell adhesion molecule, Ng-CAM, and to mouse L1. cDNA clones encompassing the entire coding sequence of this Ng-CAM related molecule, called Nr-CAM, have been isolated and sequenced. A glycoprotein containing one major component of Mr 145,000 on SDS-PAGE was purified from brain by lentil lectin affinity chromatography and FPLC, and its amino-terminal sequence was identical to that predicted from the Nr-CAM cDNA. The complete cDNA sequence encodes six Ig-like domains, five fibronectin type III repeats, a predicted transmembrane domain, and a short cytoplasmic domain. On Northern blots, nucleic acid probes for Nr-CAM recognized one major RNA species of approximately 7 kb and much lesser amounts of larger RNAs. Most of the same probes hybridized to single bands on genomic Southern blots, suggesting that Nr-CAM is encoded by a single gene that may be alternatively processed to yield several mRNAs. In support of this notion, two Nr-CAM cDNA clones had a 57-bp sequence located between the second and third Ig-like domains that was not found in two other Nr-CAM cDNA clones, and two other clones were isolated that lacked the 279-bp segment encoding the fifth fibronectin-like type III repeat. Antibodies against the purified protein and synthetic peptides in Nr-CAM both recognized a predominant Mr 145,000 species and a much less prevalent species of Mr 170,000 in neural tissues. Levels of Nr-CAM expression increased in the brain until approximately embryonic day (E) 12, followed by slightly lower levels of expression at E18 and after hatching. Immunofluorescent staining with anti-Nr-CAM antibodies showed that most neurons in the retina were positive at E7 and the pattern of expression became restricted to several layers on neuronal cell bodies and fibers during development. Anti-Nr-CAM antibodies labeled specifically cell surfaces on neurons in culture. Although the structure of Nr-CAM resembles that of chicken Ng-CAM and mouse L1, the identity with each of these neural CAMs does not exceed 40%. The differences indicate that Nr-CAM is distinct from Ng-CAM and L1, but there are sufficient similarities to suggest that all of these molecules are members of a subgroup of neural CAMs in the N-CAM superfamily.(ABSTRACT TRUNCATED AT 400 WORDS)     

Ankyrin-binding proteins related to nervous system cell adhesion molecules: candidates to provide transmembrane and intercellular connections in adult brain

A major class of ankyrin-binding glycoproteins have been identified in adult rat brain of 186, 155, and 140 kD that are alternatively spliced products of the same pre-mRNA. Characterization of cDNAs demonstrated that ankyrin-binding glycoproteins (ABGPs) share 72% amino acid sequence identity with chicken neurofascin, a membrane-spanning neural cell adhesion molecule in the Ig super-family expressed in embryonic brain. ABGP polypeptides have the following features consistent with a role as ankyrin-binding proteins in vitro and in vivo: (a) ABGPs and ankyrin associate as pure proteins in a 1:1 molar stoichiometry; (b) the ankyrin-binding site is located in the COOH-terminal 21 kD of ABGP186 which contains the predicted cytoplasmic domain; (c) ABGP186 is expressed at approximately the same levels as ankyrin (15 pmoles/milligram of membrane protein); and (d) ABGP polypeptides are co- expressed with the adult form of ankyrinB late in postnatal development and are colocalized with ankyrinB by immunofluorescence. Similarity in amino acid sequence and conservation of sites of alternative splicing indicate that genes encoding ABGPs and neurofascin share a common ancestor. However, the major differences in developmental expression reported for neurofascin in embryos versus the late postnatal expression of ABGPs suggest that ABGPs and neurofascin represent products of gene duplication events that have subsequently evolved in parallel with distinct roles. The predicted cytoplasmic domains of rat ABGPs and chicken neurofascin are nearly identical to each other and closely related to a group of nervous system cell adhesion molecules with variable extracellular domains, which includes L1, Nr-CAM, and Ng- CAM of vertebrates, and neuroglian of Drosophila. The ankyrin-binding site of rat ABGPs is localized to the C-terminal 200 residues which encompass the cytoplasmic domain, suggesting the hypothesis that ability to associate with ankyrin may be a shared feature of neurofascin and related nervous system cell adhesion molecules.     

Bravo/Nr-CAM is closely related to the cell adhesion molecules L1 and Ng-CAM and has a similar heterodimer structure

Diverse cell-surface molecules of the nervous system play an important role in specifying cell interactions during development. Using a method designed to generate mAbs against neural surface molecules of defined molecular weight, we have previously reported on the surface protein, Bravo, found in the developing avian retinotectal system. Bravo is immunologically detected on developing optic fibers in the retina, but absent from distal regions of the same fibers in the tectum. We have isolated cDNA clones encompassing the entire coding region of Bravo, including clones containing five alternative sequences of cDNA. These putative alternatively spliced sequences encode stretches of polypeptide ranging in length from 10-93 amino acids and are predicted to be both extra- and intracellular. The deduced primary structure of Bravo reveals that, like the cell adhesion molecules (CAMs) chicken Ng-CAM and mouse L1, Bravo is composed of six Ig-like domains, five fibronectin type III repeats, a transmembrane domain, and a short cytoplasmic region. Recently, the cDNA sequence of a related molecule, Nr-CAM, was reported and its possible identity with Bravo discussed (Grumet, M., V. Mauro, M. P. Burgoon, G. E. Edelman, and B. A. Cunningham. 1991. J. Cell Biol. 113:1399-1412). Here we confirm this identity and moreover show that Bravo is found on Müller glial processes and end-feet in the developing retina. In contrast to the single polypeptide chain structure of Nr-CAM reported previously, we show that Bravo has a heterodimer structure composed of an alpha chain of M(r) 140/130 and a beta chain of 60-80 kD. As with L1 and Ng-CAM, the two chains of Bravo are generated from an intact polypeptide by cleavage at identical locations and conserved sites within all three molecules (Ser-Arg/Lys-Arg). The similar domain composition and heterodimer structure, as well as the 40% amino acid sequence identity of these molecules, defines them as an evolutionarily related subgroup of CAMs. The relationship of Bravo to molecules known to be involved in cell adhesion and process outgrowth, combined with its pattern of expression and numerous potential isoforms, suggests a complex role for this molecule in cell interactions during neural development.     

Spliced RNA from the IR1-U2 region of Epstein-Barr virus: presence of an open reading frame for a repetitive polypeptide.

We have constructed a cDNA library from the cytoplasmic RNAs of Raji cells, a Burkitt's lymphoma cell line latently infected with Epstein-Barr virus. We report here the characterization of a cDNA representing a spliced RNA transcribed from the IR1-U2 region of the viral genome. The cDNA is 1007 bp long. The 5' region contains three tandem repeats of two exons, 66 and 132 bp, which are transcribed from the IR1 repeats. The 3' region is formed from four exons transcribed from U2. An open reading frame extends from the 5' end to position 784, and includes the repeats. This reading frame presumably corresponds to the carboxy-terminal 261 amino acids of a polypeptide containing several repeats of a 66 amino acid sequence. Since it would be encoded by the IR1-U2 region of the viral genome, the putative polypeptide might be involved in the process of growth-transformation of B-lymphocytes.     

The carboxyl terminus of the chicken alpha 3 chain of collagen VI is a unique mosaic structure with glycoprotein Ib-like, fibronectin type III, and Kunitz modules

The primary amino acid sequence of the carboxyl-terminal portion of the alpha 3 chain of chick type VI collagen as deduced from the nucleotide sequence is reported. This carboxyl-terminal segment is not present in the alpha 1 and alpha 2 chains of chick type VI collagen and is specific for a mosaic region with extensive similarities to several other proteins. This unique segment, beginning with a stretch (73 residues) very rich in serine and threonine, is preceded by sequences analogous to the platelet glycoprotein Ib. This region is followed by one segment that closely resembles the type III domains of fibronectin. At the end of the sequence, there is a 58-residue motif very similar to sequences characteristic of the Kunitz-type proteinase inhibitors. The present findings and our recent observation that the alpha 3(VI) chain contains 11 repeats similar to type A repeats of von Willebrand factor raise interesting questions about the peculiar mosaic structure and the multiple functions that this unique collagen might play in growth and remodeling of connective tissues.     

Type XIV collagen is a variant of undulin.

We have isolated undulin, an extracellular matrix protein associated with the surface of collagen fibrils, from chicken embryos. The protein showed a molecular mass of about 600 kDa and is composed of three 210-kDa subunits linked by reducible as well as non-reducible bonds. In contrast to human undulin which reportedly is devoid of collagenous sequences, the chicken protein contained a short triple-helical segment that was sensitive to digestion by bacterial collagenase. Screening of an expression library with affinity-purified antibodies yielded two cDNA clones specific for chicken undulin. Analysis of the amino acid sequence deduced from the nucleotide sequence of these clones showed that the human and the chicken protein shared 71% sequence identity. At the amino-terminus both polypeptides contained several similar repeats related to the type III modules found in fibronectin. Towards the carboxyl terminus, however, the two sequences diverged substantially from each other. While the human sequence terminated in a proline-rich segment, the chicken sequence continued with a domain related to von Willebrand factor, with a domain similar to the noncollagenous domain NC4 of type IX collagen and with a typical collagenous triple helix. A short segment of this sequence was found to be identical with the published sequence of a bovine peptide derived from type XIV collagen. Our protein must therefore represent chicken type XIV collagen. One way to explain these results is the possibility that undulin exists in at least two alternatively spliced variants, one lacking the collagenous domain, as described initially for human undulin, and one containing the triple-helical domain, as found in type XIV collagen.     

Primary structure of rat pulmonary surfactant protein D. cDNA and deduced amino acid sequence.

Surfactant protein D (SP-D) is a carbohydrate-binding glycoprotein containing a collagen-like domain that is synthesized by alveolar type II epithelial cells. The complete primary structure of rat SP-D has been determined by sequencing of a cloned cDNA. The protein consists of three regions: an NH2-terminal segment of 25 amino acids, a collagen-like domain consisting of 59 Gly-X-Y repeats, and a COOH-terminal carbohydrate recognition domain of 153 amino acids. There are 6 cysteine residues present in rat SP-D: 2 in the NH2-terminal noncollagenous segment and 4 in the COOH-terminal carbohydrate-binding domain. The collagenous domain contains one possible N-glycosylation site. The protein is preceded by a cleaved, NH2-terminal signal peptide. SP-D shares considerable homology with the C-type mammalian lectins. Hybridization analysis demonstrates that rat SP-D is encoded by a 1.3-kilobase mRNA which is abundant in lung and highly enriched in alveolar type II cells. Extensive homology exists between rat SP-D and bovine conglutinin.     

Homophilic and heterophilic binding activities of Nr-CAM, a nervous system cell adhesion molecule

Nr-CAM is a membrane glycoprotein that is expressed on neurons. It is structurally related to members of the N-CAM superfamily of neural cell adhesion molecules having six immunoglobulin-like domains and five fibronectin type III repeats in the extracellular region. We have found that the aggregation of chick brain cells was inhibited by anti-Nr-CAM Fab' fragments, indicating that Nr-CAM can act as a cell adhesion molecule. To clarify the mode of action of Nr-CAM, a mouse fibroblast cell line L-M(TK-) (or L cells) was transfected with a DNA expression construct encoding an entire chicken Nr-CAM cDNA sequence. After transfection, L cells expressed Nr-CAM on their surface and aggregated. Aggregation was specifically inhibited by anti-Nr-CAM Fab' fragments. To check the specificity of this aggregation, a fusion protein (FGTNr) consisting of glutathione S-transferase linked to the six immunoglobulin domains and the first fibronectin type III repeat of Nr-CAM was expressed in Escherichia coli. Addition of FGTNr to the transfected cells blocked their aggregation. Further analysis using a combination of cell aggregation assays, binding of cells to FGTNr-coated substrates, aggregation of FGTNr-coated Covaspheres and binding of FGTNr-coated Covaspheres to FGTNr-coated substrates revealed that Nr-CAM mediates two types of cell interactions: a homophilic, divalent cation-independent binding, and a heterophilic, divalent cation-dependent binding. Homophilic binding was demonstrated between transfected L cells, between chick embryo brain cells and FGTNr, and between Covaspheres to which FGTNr was covalently attached. Heterophilic binding was shown to occur between transfected and untransfected L cells, and between FGTNr and primary chick embryo fibroblasts; in all cases, it was dependent on the presence of either calcium or magnesium. Primary chick embryo glia or a human glial cell line did not bind to FGTNr-coated substrates. The results indicate that Nr-CAM is a cell adhesion molecule of the nervous system that can bind by two distinct mechanisms, a homophilic mechanism that can mediate interactions between neurons and a heterophilic mechanism that can mediate binding between neurons and other cells such as fibroblasts.     

Molecular composition of the node of Ranvier: identification of ankyrin- binding cell adhesion molecules neurofascin (mucin+/third FNIII domain- ) and NrCAM at nodal axon segments

Neurofascin, NrCAM, L1, and NgCAM are a family of Ig/FNIII cell adhesion molecules that share ankyrin-binding activity in their cytoplasmic domains, and are candidates to form membrane-spanning complexes with members of the ankyrin family of spectrin-binding proteins in a variety of cellular contexts in the nervous system. Specialized forms of ankyrin, 270 kD and/or 480 kD ankyrinG are components of the membrane undercoat of axons at the node of Ranvier. This paper focuses on definition of the isoforms of ankyrin-binding cell adhesion molecules localized with ankyrinG at the nodal axon segment. The exon usage of two major forms of neurofascin was determined by isolation of full-length cDNAs and used to prepare isoform-specific antibodies. An isoform of neurofascin containing a mucin-like domain and lacking the third FNIII domain was concentrated at axon initial segments and colocalized at nodes of Ranvier with ankyrinG and the voltage-dependent sodium channel. An alternative form of neurofascin lacking the mucin-like domain and containing the third FNIII domain was present in unmyelinated axons. The antibody initially raised against neurofascin was used to screen a rat brain cDNA expression library. In addition to neurofascin, this screen yielded a clone with 80% sequence identity to NrCAM from chicken. The sequences of two full-length cDNAs are presented. NrCAM is most closely related to neurofascin among the other members of the L1/neurofascin/NgCAM family, with over 70% identity between cytoplasmic domains. NrCAM, visualized with antibodies specific for the ecto-domain, also was found to be coexpressed with neurofascin at nodes of Ranvier and at axon initial segments. This is the first characterization of defined neuronal cell adhesion molecules localized to axonal membranes at the node of Ranvier of myelinated axons.     

Two independent domains of hDlg are sufficient for subcellular targeting: the PDZ1-2 conformational unit and an alternatively spliced domain

hDlg, a human homologue of the Drosophila Dig tumor suppressor, contains two binding sites for protein 4.1, one within a domain containing three PSD-95/Dlg/ZO-1 (PDZ) repeats and another within the alternatively spliced I3 domain. Here, we further define the PDZ- protein 4.1 interaction in vitro and show the functional role of both 4.1 binding sites in situ. A single protease-resistant structure formed by the entirety of both PDZ repeats 1 and 2 (PDZ1-2) contains the protein 4.1-binding site. Both this PDZ1-2 site and the I3 domain associate with a 30-kD NH2-terminal domain of protein 4.1 that is conserved in ezrin/radixin/moesin (ERM) proteins. We show that both protein 4.1 and the ezrin ERM protein interact with the murine form of hDlg in a coprecipitating immune complex. In permeabilized cells and tissues, either the PDZ1-2 domain or the I3 domain alone are sufficient for proper subcellular targeting of exogenous hDlg. In situ, PDZ1-2- mediated targeting involves interactions with both 4.1/ERM proteins and proteins containing the COOH-terminal T/SXV motif. I3-mediated targeting depends exclusively on interactions with 4.1/ERM proteins. Our data elucidates the multivalent nature of membrane-associated guanylate kinase homologue (MAGUK) targeting, thus beginning to define those protein interactions that are critical in MAGUK function.     

Biochemical characterization of 39-kDa class I histocompatibility antigen in plasma. A secretable membrane protein derived from transmembrane domain deletion

Three human class I major histocompatibility antigens (HLA) with molecular masses of 44, 39, and 36 kDa were identified in plasma by immunoprecipitation and immunoblotting. Further biochemical characterization showed that these antigens in plasma could be fractionated by Sephacryl S-300 column chromatography into two different pools. The 44-kDa intact HLA heavy chains are detected only in pool I and have an apparent molecular weight of 200,000 as determined by calibrated gel filtration column chromatography. The 39- and 36-kDa HLA heavy chains are present only in pool II and have an apparent molecular weight of 50,000. HLA in pool I can be extracted by Triton X-114 detergent, but 39- and 36-kDa plasma HLA in pool II are water soluble and not extractable by Triton X-114. Amino acid sequences of NH2 termini for 44- and 39-kDa plasma HLA are identical to that of cellular HLA. In contrast, the NH2-terminal amino acid sequence for 36-kDa plasma HLA has not been reported previously for any other proteins. Since the loss of both transmembrane domain and cytoplasmic tail at the carboxyl terminus of HLA will generate a 36-kDa protein, the findings suggest that the 39-kDa HLA might be the product of alternatively spliced mRNA with deletion of the exon coding for transmembrane domain. By using polymerase chain reaction and DNA sequencing, the presence of alternatively spliced mRNA with deletion of the transmembrane domain exon was identified in mononuclear leukocytes of peripheral blood. This alternatively spliced HLA mRNA was not detectable in mononuclear leukocytes of an individual who had no 39-kDa plasma HLA. This finding indicates that the alternatively spliced mRNA in mononuclear leukocytes is responsible for the synthesis of a secretable class I HLA.     

Isolation and characterization of genomic DNA coding for alpha 2 type I collagen.

We have isolated and characterized a segment of the chick alpha 2 collagen gene by screening a library of chick genomic fragments using as hybridization probe an alpha 2 collagen cDNA clone. Several clones were isolated and one of them, lambda gCOL 204, was used for further studies. The DNA of lambda gCOL 204 hybridizes to a unique species of mRNA the size of alpha 2 collagen mRNA. This mRNA can be translated into a unique polypeptide which comigrates in SDS-gel electrophoresis with pro-alpha 2 collagen. Electron microscopic analysis by R-loop technique indicates that lambda gCOL 204 contains 7Kb of the alpha 2 collagen gene. This 7 Kb piece constitutes the 3' end of the gene. The same clone also contains 9 Kb of DNA that is immediately adjacent to the 3' end of the alpha 2 collagen gene. The cloned segment of the alpha 2 collagen gene is interrupted by 8 intervening sequences of various lengths. The coding sequences for collagen in this clone add up to approximately 1,800 bp, which correspond to about 1/3 of alpha 2 collagen mRNA. DNA sequence analysis of a small coding segment of lambda g COL 204 reveals a characteristic collagen type sequence which encodes for an amino acid sequence identical to a sequence found in calf alpha 2 collagen. The sequence of this region of the protein has not yet been determined for the chick alpha 2 collagen.     

Evidence for the regulation of alternative splicing via complementary DNA sequence repeats

MOTIVATION: While the mechanism for regulating alternative splicing is poorly understood, secondary structure has been shown to be integral to this process. Due to their propensity for forming complementary hairpin loops and their elevated mutation rates, tandem repeated sequences have the potential to influence splicing regulation. RESULTS: An analysis of human intronic sequences reveals a strong correlation between alternative splicing and the prevalence of mono- through hexanucleotide tandem repeats that may engage in complementary pairing in introns that flank alternatively spliced exons. While only 44% of the 18 173 genes in the Human Alternative Splicing Database are known to be alternatively spliced, they contain 84% of the 694 237 intronic complementary repeat pairs. Significantly, the normalized frequency and distribution of repeat sequences, independent of their potential for pairing, are indistinguishable between alternatively spliced and non-alternatively spliced genes. Thus, the increased prevalence of repeats with pairing potential in alternatively spliced genes is not merely a consequence of more repeats or repeat composition bias. These results suggest that complementary repeats may play a role in the regulation of alternative splicing. CONTACT: harold.garner@utsouthwestern.edu.     

The amino acid sequences of the myelin-associated glycoproteins: homology to the immunoglobulin gene superfamily

The myelin associated glycoproteins (MAG) are integral plasma membrane proteins which are found in oligodendrocytes and Schwann cells and are believed to mediate the axonal-glial interactions of myelination. In this paper we demonstrate the existence in central nervous system myelin of two MAG polypeptides with Mrs of 67,000 and 72,000 that we have designated small MAG (S-MAG) and large MAG (L-MAG), respectively. The complete amino acid sequence of L-MAG and a partial amino acid sequence of S-MAG have been deduced from the nucleotide sequences of corresponding cDNA clones isolated from a lambda gt11 rat brain expression library. Based on their amino acid sequences, we predict that both proteins have an identical membrane spanning segment and a large extracellular domain. The putative extracellular region contains an Arg-Gly-Asp sequence that may be involved in the interaction of these proteins with the axon. The extracellular portion of L-MAG also contains five segments of internal homology that resemble immunoglobulin domains, and are strikingly homologous to similar domains of the neural cell adhesion molecule and other members of the immunoglobulin gene superfamily. In addition, the two MAG proteins differ in the extent of their cytoplasmically disposed segments and appear to be the products of alternatively spliced mRNAs. Of considerable interest is the finding that the cytoplasmic domain of L-MAG, but not of S-MAG, contains an amino acid sequence that resembles the autophosphorylation site of the epidermal growth factor receptor.     

Structure of the bovine elastin gene and S1 nuclease analysis of alternative splicing of elastin mRNA in the bovine nuchal ligament

Genomic clones encompassing all the translated sequences, the 3' untranslated sequence, and 1 kb flanking the ATG translation initiation codon of bovine tropoelastin have been obtained and characterized by restriction enzyme analysis and extensive DNA sequencing. These analyses demonstrated that functionally distinct hydrophobic and cross-linking domains of the protein are segregated into separate exons throughout the gene. The putative promoter region lacks a TATA box, has an extremely high G+C content, and contains several SP1 binding sites. Comprehensive S1 analyses using probes covering the entire mRNA and RNA isolated from the nuchal ligament of bovine fetuses of different ages, neonate calves, and adult cows demonstrated that while only a single exon is alternatively spliced at high frequency, many exons are alternatively spliced at limited, variable frequencies. The results also suggest that such limited splicing is increased in the adult tissue relative to fetal and neonate tissues.     

Sequences of complete cDNAs encoding four variants of chicken skeletal muscle troponin T

cDNAs containing the complete coding sequences of four isoforms of troponin T derived from 1-week-old chick skeletal muscle have been isolated and sequenced. While the 5' and 3' untranslated regions and most of the coding sequence were identical for each, dramatic differences were observed in the NH2-terminal region corresponding to amino acid residues 10-37 of rabbit skeletal troponin T. These sequence differences correspond to the alternatively spliced but not mutually exclusive exons 4 to 8 of the rat skeletal muscle troponin T gene. In addition, we observe a sequence corresponding to an extra exon or exons (between 5 and 6) present in the chicken skeletal muscle gene and not previously detected in the rat skeletal or chicken cardiac genes. This sequence of 63 nucleotides consists of an almost perfect repeat of 30 and 33 nucleotides and has previously been shown to be represented as a protein variant in chicken skeletal muscle. A difference is also present in one cDNA clone corresponding to the alternatively spliced (mutually exclusive) exons 16 and 17 of the rat gene. In the protein, this corresponds to a region implicated in the interaction of troponin T with troponin C, tropomyosin, and perhaps troponin I and F-actin.