The sarcolemmal associated proteins (SLAPs) are encoded by multiple mRNAs that are presumably generated by alternative splicing mechanisms. The amino acid sequence of the SLAP1 isoform exhibited 76% identity with TOP AP , a topographically graded antigen of the chick visual system. The regions of coiled-coil structure including an 11-heptad acidic amphipathic ␣-helical segment was conserved with a major divergence in sequence noted in the hydrophobic C termini predicted to be transmembrane domains in the two polypeptides. The genomic organization of the 3 region of the SLAP gene indicated that SLAP1 and TOP AP are generated by alternative splicing mechanisms, which are conserved among mammalian and avian species. SLAP1/TOP AP were encoded by 11 exons distributed over a minimum of 35 kilobase pairs of continuous DNA; 9 of the exons were constitutively expressed, and 2 were alternatively spliced. The exons range in size from 60 to 321 base pairs, and the predicted functional domains within the polypeptides were encompassed by single exons. The introns vary from 0.2 to 10 kilobase pairs and conform to consensus dinucleotide splicing signals. Reverse transcriptase-polymerase chain reaction studies demonstrated that alternative exons (IV and X) of SLAP were expressed in a tissue-specific fashion and developmentally regulated. The alternatively spliced exon X, which encodes the putative transmembrane anchor in TOP AP , and a constitutively expressed exon XI, which encodes the putative transmembrane domain in SLAP, were found to target these polypeptides to membrane structures. The presence and conservation of termination codons in exons X and XI render expression of the two SLAP1/TOP AP transmembrane domains mutually exclusive. These data reveal that TOP AP and SLAP are alternatively spliced products of a single gene that encodes a unique class of tail-anchored membrane proteins.

Novel splice variants of the ␣ 1 subunit of the Ca v 1.2 voltage-gated Ca 2؉ channel were identified that predicted two truncated forms of the ␣ 1 subunit comprising domains I and II generated by alternative splicing in the intracellular loop region linking domains II and III. In rabbit heart splice variant 1 (RH-1), exon 19 was deleted, which resulted in a reading frameshift of exon 20 with a premature termination codon and a novel 19-amino acid carboxyl-terminal tail. In the RH-2 variant, exons 17 and 18 were deleted, leading to a reading frameshift of exons 19 and 20 with a premature stop codon and a novel 62-amino acid carboxyl-terminal tail. RNase protection assays with RH-1 and RH-2 cRNA probes confirmed the expression in cardiac and neuronal tissue but not skeletal muscle. The deduced amino acid sequence from fulllength cDNAs encoding the two variants predicted polypeptides of 99.0 and 99.2 kDa, which constituted domains I and II of the ␣ 1 subunit of the Ca v 1.2 channel. Antipeptide antibodies directed to sequences in the second intracellular loop between domains II and III identified the 240-kDa Ca v 1.2 subunit in sarcolemmal and heavy sarcoplasmic reticulum (HSR) membranes and a 99-kDa polypeptide in the HSR. An antipeptide antibody raised against unique sequences in the RH-2 variant also identified a 99-kDa polypeptide in the HSR. These data reveal the expression of additional Ca 2؉ channel structural units generated by alternative splicing of the Ca v 1.2 gene.The voltage-gated ion channels determine membrane excitability and regulate signal transduction (1-5). These ion channels consist of multimeric complexes comprising a central ␣ subunit, which contains the structural determinants for ion selectivity, conductance, and voltage sensing, and several auxiliary subunits, which confer regulation on the ␣ subunits (6, 7). The ␣ subunits of K ϩ , Na ϩ , and Ca 2ϩ channels show overall structural similarity in that the ␣ subunit of K ϩ channels consists of a single domain, which is predicted to contain six membrane-spanning (S1-S6) 1 regions, whereas the structural unit of Na ϩ and Ca 2ϩ channels consists of four such homologous domains (1). Because the K ϩ channel genes encode only one domain, it is proposed that homo-and heterotetrameric co-assembly of single-domain K ϩ channel subunits is required to constitute the four-domain channel complexes (8 -10).The different pore-forming ␣ subunits are encoded by distinct gene families (68 Changes in amino acid composition in the extramembrane regions of the ␣ 1 subunit appear to have given rise to specialized physiological roles for the different classes of ␣ 1 subunits of the Ca 2ϩ channel family (37-41). For example, the structure of the intracellular loop connecting domains II and III appears to be a critical determinant of the mode of signal transmission in the Ca v 1.1 and Ca v 1.2 ␣ 1 subunits (42-44), because this structure of the skeletal (Ca v 1.1) but not cardiac (Ca v 1.2) ␣ 1 subunits can directly interact with the Ca 2ϩ release channel of the sa...

No abstract