New aspects of chemoreception in flies.

For studies of insect taste the sensory cells of labellar taste hairs on the proboscis of calliphorid flies (e.g., Phormia) are preferred objects. Three sensory cells are electrophysiologically characterized: a sugar cell, a water cell and a cation or salt cell. Studying hairs on legs and proboscis of other families of flies we recently obtained results that extend our knowledge about the complexity of food detection in flies. The hoverfly Eristalis tenax (Syrphidae family) feeds on nectar and pollen of flowers. While nectar is recognized via the sugar cell, the so-called salt cell is activated by low concentrations of a water-soluble substance in extracts of pollen. From several tested pollen constituents only proline, an essential amino acid for pollen germination, stimulates the salt cell. The discrimination between salt and pollen in behavioral tests can be explained by the finding that the water cell remains active in the presence of pollen extract, but is inhibited by salts. Obviously the water cell activity is involved in the feeding decision. We already described a similar situation in Phormia, where the salt cell activity only elicits feeding behavior when the water cell is active too. In contrast to the sugar cells of species of flies studied so far the sugar cell of the house fly Musca domestica (Muscidae family) is highly sensitive to lactose, a disaccharide (beta-galactoside (1-->4) glucoside) naturally occurring only in the milk of mammals. Thus sugar spectra adapt to special environmental food situations. The fifth tarsomere of the forelegs of Musca bears besides two D-hairs 46 hairs of the B-type. The B-hairs are functionally not uniform in contrast to their structural identity. Only two hairs, the 'water hairs,' contain a water cell beside a spontaneously firing cell, but no sugar cell. Other examples of diversity are hairs that contain apart from the classical sugar cells, additional cells which react either to p-nitrophenyl-beta-galactoside, but not to the chemically related lactose, or to p-nitrophenyl-alpha-glucoside. Therefore the rule established for the labellar taste hairs that each hair type contains the same physiological equipment of sensory cells, does not hold for the taste hairs of the legs.