RESUMEN
Connexins or innexins form gap junctions, while claudins and occludins form tight junctions. In this study, statistical data, derived using novel software, indicate that these four junctional protein families and eleven other families of channel and channel auxiliary proteins are related by common descent and comprise the Tetraspan (4 TMS) Junctional Complex (4JC) Superfamily. These proteins all share similar 4 transmembrane α-helical (TMS) topologies. Evidence is presented that they arose via an intragenic duplication event, whereby a 2 TMS-encoding genetic element duplicated tandemly to give 4 TMS proteins. In cases where high resolution structural data were available, the conclusion of homology was supported by conducting structural comparisons. Phylogenetic trees reveal the probable relationships of these 15 families to each other. Long homologues containing fusions to other recognizable domains as well as internally duplicated or fused domains are reported. Large "fusion" proteins containing 4JC domains proved to fall predominantly into family-specific patterns as follows: (1) the 4JC domain was N-terminal; (2) the 4JC domain was C-terminal; (3) the 4JC domain was duplicated or occasionally triplicated and (4) mixed fusion types were present. Our observations provide insight into the evolutionary origins and subfunctions of these proteins as well as guides concerning their structural and functional relationships.
Asunto(s)
Proteínas de la Membrana/química , Secuencia de Aminoácidos , Animales , Claudinas/química , Claudinas/clasificación , Conexinas/química , Conexinas/clasificación , Uniones Comunicantes/metabolismo , Proteínas de la Membrana/clasificación , Proteínas Proteolipídicas Asociadas a Mielina y Linfocito/química , Proteínas Proteolipídicas Asociadas a Mielina y Linfocito/clasificación , Filogenia , Estructura Terciaria de Proteína , Alineación de Secuencia , Uniones Estrechas/metabolismoRESUMEN
We have designed a freely accessible program, PhyST, which allows the automated characterization of any family of homologous proteins within the Transporter Classification Database. The program performs an NCBI-PSI-BLAST search and reports (1) the average protein sequence length with standard deviations, (2) the average predicted number of transmembrane segments, (3) the total number of homologues retrieved, (4) a quantitative list of all source phyla, and (5) potential fusion proteins of sizes considerably exceeding the average size of the proteins retrieved. We have applied this program to 58 families of holins, and the results are presented. The results show that holins are very rarely fused to other protein domains, suggesting that holins form transmembrane pores as homooligomers without the participation of other proteins or protein domains.
