Evidence for two separate categories of junctional particle during the concurrent formation of tight and gap junctions

Abstract

The stages in the formation of vertebrate-like tight junctions and gap junctions, which coexist in tissues of the adult, have been followed in the CNS of young hatchling spiders. These junctions develop concurrently between the cells of the outer glial layer, or perineurium, that ensheathes the main ganglionic mass. In the mature state the two junctional types can be distinguished not only on the basis of the way in which their component intramembranous particles (IMPs) are arranged, but also by the size and fracturing characteristics of their respective IMPs. Hence they can be identified accurately during the initial and subsequent stages of their formation. The fully formed gap junctions, composed of clusters of 13-nm E-face (EF) IMPs, are found in the membranes between EF grooves that are complementary to a network of P-face (PF) ridges of 8- to 10-nm IMPs that compose the tight junctions. In hatchlings, the latter appear first as individual particles which become aligned initially in rows of 2 to 3. With time, the IMPs seemingly become fused together to form ridges, and the length of these particle rows is increased until they are arrayed as discontinuous strands scattered over the presumptive junctional area. The 13-nm EF IMPs, which are inserted between these tight junctional structures, are apparently restricted in their translateral movement by the strands of the latter which pinch the adjacent membranes together. As the ridge-groove system becomes interconnected into a reticular network, the gap junctions coalesce into linear, then macular aggregates, always intercalated between the strands of the network. It is clear that these two junctions are composed of IMPs which do not share a common precursor particle as has been suggested for some vertebrate tissues, but which arise as two quite distinct particle populations. The mode of formation of these anastomosing tight junctions, hitherto not explored in invertebrate tissues, seems very similar to the general pattern of events reported previously for vertebrate zonulae occludentes. It therefore seems likely that, in chordates, these tight junctional particles also arise separately from those forming the gap junctions.