Connexins are chordate-specific transmembrane proteins that can form gap junctional channels between adjacent cells. With the progress in vertebrate genome sequencing, it is now possible to reconstruct the main lines in the evolution of the connexin family from fishes to mammals. Four connexin groups are only found in fishes. Otherwise, the differences between fishes and mammals can be explained by two gene losses (Cx39.9 and Cx43.4) after the divergence of the Reptilia, and three gene duplications (the generation of Cx26 and 30 from a preCx26/30 sequence, Cx30.3 and 31.1 from a preCx30.3/ 31.1 sequence, and Cx31.3 from an uncertain origin). Orthologs of most connexins can be found throughout the vertebrates from fishes to mammals. As judged from the recently defined connexins in tunicates, the original connexin might be related to the ortholog groups of Cx36, 39.2, 43.4, 45 or 47.

We suggest an extension of connexin orthology relationships across the major vertebrate lineages. We first show that the conserved domains of mammalian connexins (encoding the N-terminus, four transmembrane domains and two extracellular loops) are subjected to a considerably more strict selection pressure than the full-length sequences or the variable domains (the intracellular loop and C-terminal tail). Therefore, the conserved domains are more useful for the study of family relationships over larger evolutionary distances. The conserved domains of connexins were collected from chicken, Xenopus tropicalis, zebrafish, pufferfish, green spotted pufferfish, Ciona intestinalis and Halocynthia pyriformis (two tunicates). A total of 305 connexin sequences were included in this analysis. Phylogenetic trees were constructed, from which the orthologies and the presumed evolutionary relationships between the sequences were deduced. The tunicate connexins studied had the closest, but still distant, relationships to vertebrate connexin 36, 39.2, 43.4, 45 and 47. The main structure in the connexin family known from mammals pre-dates the divergence of bony fishes, but some additional losses and gains of connexin sequences have occurred in the evolutionary lineages of subsequent vertebrates. Thus, the connexin gene family probably originated in the early evolution of chordates, and underwent major restructuring with regard to gene and subfamily structures (including the number of genes in each subfamily) during early vertebrate evolution.

A large number of experimental studies have linked the S100A4 gene product to the metastatic phenotype of cancer cells and clinical evidence indicates a correlation between S100A4 expression and poor prognosis in several cancer types. The aim of the present study was to analyse the expression of the S100A4 protein in colorectal cancer. Paraffin-embedded samples from 277 colorectal cancer patients were immunostained with anti-S100A4 antibody. Cytoplasmic staining was observed in 178 of 277 samples (64%), whereas, unexpectedly, nuclear expression of S100A4 was found in 88 of 277 of the samples (32%). This novel finding was confirmed by western blot analysis of nuclear fractions isolated from frozen tumour tissue. Statistical analysis revealed a significant correlation between nuclear expression of S100A4 and tumour stage at diagnosis, while there was no such correlation between cytoplasmic staining and tumour stage. The nuclear localization of S100A4 in colorectal cancer and its relationship to tumour stage suggest that this protein may be involved in gene regulatory pathways of relevance to the metastatic phenotype of cancer cells.

A number of kinases and signal transduction pathways are known to affect gap junctional intercellular communication and/or phosphorylation of connexins. Most of the information is available for protein kinase A, protein kinase C, mitogen-activated protein kinase, and the tyrosine kinase Src. Much less is known for protein kinase G, Ca(2+)-calmodulin dependent protein kinase, and casein kinase. However, the present lack of knowledge is not necessarily synonymous with lack of importance in the regulation of intercellular communication and phosphorylation of connexins. Kinases and the phosphorylation of connexins may be involved in the regulation of gap junctional intercellular communication at all levels ranging from the expression of connexin genes to the degradation of the gap junction channels. The exact role of the phosphorylation depends both on the kinase and the connexin involved, as well as the cellular context.

Aim: Gap junction intercellular communication (GJIC) and hemichannel permeability may have important roles during an ischemic insult. Our aim was to evaluate the effect of ischemia on gap junction channels and hemichannels. Methods: We used neonatal rat heart myofibroblasts and simulated ischemia with a HEPES buffer with high potassium, low pH, absence of glucose, and oxygen tension was reduced by dithionite. Microinjection, western blot, immunofluorescence, cell viability and dye uptake were used to evaluate the effects induced by dithionite. Isolated perfused rat hearts were used to analyse infarct size. Results: Short period with simulated ischemia reduced the ability to transfer a dye between neighbouring cells, which indicated reduced GJIC. Prolonged exposure to simulated ischemia caused opening of hemichannels, and cell death was apparent while gap junction channels remained closed. Connexin 43 became partially dephosphorylated and the total amount decreased during simulated ischemia. We were not able to detect the alternative hemichannel-forming protein, Pannexin 1, in these cells. The potential importance of Connexin 43 or Pannexin 1 hemichannels in ischemia-induced infarct in the intact heart was studied by perfusion of the heart in the presence of peptides that block one or the other type of hemichannels. The connexin-derived peptide, Gap26, significantly reduced the infract/risk zone ratio (control 48.7±4.2% and Gap26 19.4±4.1%, p<0.001), while the pannexin-derived peptide, ¹⁰Panx1, did not change infarct/risk ratio. Conclusion: Connexin 43 is most likely responsible for both closure of gap junction channels and opening of hemichannels during simulated ischemia in neonatal rat heart myofibroblasts. Opening of connexin 43 hemichannels during ischemia-reperfusion seems to be an important mechanism for ischemia-reperfusion injury in the heart. By preventing the opening of these channels during early ischemia-reperfusion the infarct size becomes significantly reduced.