The structures of 13 variants of invertebrate septate junction are reviewed on the basis offreeze-fracture, lanthanum tracer and thin-section studies. In addition, a simple type ofoccluding junction in the phylum Porifera, a variation of tight junction in the phylum Tunicateand the vertebrate tight junction are covered. All the junctions considered form a belt around the apical circumference of cells lining a lumen or an exterior surface. The large number of these junctions now recognized permits discussion relating to invertebrate classification and suggested phylogenetic relationships, and to the development of intercellular junctions. The relationships revealed are discussed under three headings: Coelenterates and lower invertebrates, Proterostomia (the annelid, molluscan and arthropod lineage) and the Deuterostomia(the echinoderm and chordate lineage). It is proposed that the pleated septate junction of the lower invertebrates resembles that of the hydrozoan rather than anthozoan Coelenterates. This lower invertebrate pleated septate junction occurs in several lower invertebrate phyla including the Annelida (of the proterostome lineage), but also occurs in the Sipunculoidea, a group supposedly on the deuterostome lineage.The proterostome line includes the molluscs and the arthropods, which have the molluscarthropodpleated septate junction. Several variations of the smooth septate junction are alsoseen in Arthropoda. Among the deuterostomes the Chaetognatha have both a paired septatejunction and a pleated junction and are therefore considered to be not very far removed fromthe Sipunculoidea. The echinoderms and hemichordates also have double-septum septatejunctions. In addition however, these two phyla have anastomosing septate junctions thatare very similar, varying only in their final configuration. Of the two, the echinoderm anastomosingseptate junction most closely resembles the tight junction seen in the tunicates, and the Hemichordata are therefore considered to be a lateral development from the main lineof chordate evolution. The tunicates have a tight junction similar to that seen in vertebrates;it is however more ‘leaky’ and has distinctive freeze-fracture characteristics.In the phylum Porifera a form of simple parallel membrane junction appears to serve anoccluding function. This junction has regular intercellular spacing in the absence of any septaand it is suggested that the spacing in septate junctions is probably not dictated by the septa.This interpretation is reasonable particularly when the diversity of septal types in conjunctionwith stable intercellular spacing is considered. Finally, a theory is put forward suggesting thatin evolution a change from the septate to the tight junction could simply involve a modificationof a ‘membrane spacing factor’, which allows the membranes of adjacent cells to come together at intervals, in the normal tight junction pattern.

This content is only available via PDF.