Freeze-fracture evidence for gap junctions connecting the axon terminals of dipteran photoreceptors

The six photoreceptor terminals entering each synaptic cartridge in higher Diptera (flies) are known to be connected reciprocally by small junctions with reduced membrane spacing, so-called ‘close appositions’, but the nature of these sites is currently in dispute. Claims have been made that they are gap junctions, whilst counter-claims have denied this. From freeze-fractured replicas from Lucilia, we conclude that bow-shaped particle groupings, seen on the membrane P fracture face, have been wrongly equated by previous workers to the close appositions. They do not connect between receptors, and probably correspond instead to the frequent chemically mediated ‘tetrad’ synapses to postsynaptic cells. On the opposite (E) face, different, small assemblies of particles are identified here for the first time, which actually do connect the membranes of adjacent terminals. These form roundish, depressed maculae averaging 0 · 13 μm across, containing groups of 30 – 40 irregularly packed, irregularly sized particles of mean diameter 11 · 8 nm, with a dark dot sometimes visible at their apices. These maculae thus possess all the freeze-fracture attributes considered necessary to define an arthropod gap junction. The size range and distribution of the maculae are as expected if they are to be identified with the close appositions seen in sections, and some cross-fractures suggest this correspondence directly. The gap width at the apposition, 4 8 nm, is larger than that usually quoted for a gap junction, but actually falls within the limits of the few accurate determinations in the literature on arthropods. Both sections and replicas confirm that the maculae)appositions are much more numerous than previously thought, with up to 62 on each terminal, spaced about 1 μm apart. The total junctional area coupling a pair of terminals is comparatively large, about 0 · 4 μm2, and similar to the cross-sectional area of the incoming receptor axon as this approaches the cartridge. The junctions are therefore likely candidates to account for the strong electrical coupling recently detected between terminals. The junctions are distributed throughout the synaptic cartridge and not solely confined to its distal end, as most previous workers have contended.