Visual receptor maintenance in Drosophila involves turnover of membrane. Coated pits derived from rhabdomere and plasmalemma, coated vesicles and multivesicular bodies (MVBs) of about 0.5 micron diameter characterize the early autophagic steps. Smaller electron dense bodies (0.15 micron) merge with MVBs. These are likely to be primary lysosomes as suggested by histochemistry for acid phosphatase in normal flies and an acid phosphatase deficient mutant. Aggregates of extracellular membranes confirm an earlier report that exocytotic shedding may also be employed in the fly with its open rhabdomeres. Microspectrophotometry was used to determine aspects of cycling of visual pigment to begin to correlate with what is known about membrane cycling at the ultrastructural level. Visual pigment decreases to about half 3 h after dawn then builds back gradually to maximum before dawn. Our fixations of tissue have been at the post-dawn period when autophagy may be high as inferred from visual pigment levels. In attempts to optimize our visualization of the constructive phase of the turnover process, we developed the paradigm of carotenoid 'replacement therapy'. Carotenoid replaced flies show an increase in visual pigment possibly associated with a streaming of membrane into the rhabdomere. Aged flies, studied to determine how effective maintenance is, have a unique accumulation of extracellular debris, and a small fraction of the receptors eventually degenerate.

Drosophila rearing media had only beta-carotene, zeaxanthin or lutein as precursors for photopigment chromophores. Zeaxanthin and lutein are potentially optimum sources of the 3-hydroxylated retinoids of visual and accessory photopigments. Mutants made the electroretinogram in white (w) eyes selective for compound eye photoreceptors R1-6, R7 and R8: R1-6 dominates w's electroretinogram; R7/8 generates w;ora's (ora = outer rhabdomeres absent); R8 generates w sev;- ora's (sev = sevenless). Microspectrophotometry revealed R1-6's visual pigment. In w, all 3 carotenoids yielded monotonic dose-responses for sensitivity or visual pigment. An ultraviolet sensitivity peak from R1-6's sensitizing pigment was present at high but not low doses. In w;ora, all 3 carotenoids gave similar spectra dominated by R7's high ultraviolet sensitivity. For w sev;ora, all spectra were the shape expected for R8, peaking around 510 nm. The sensitivity dose-response was at its ceiling except for low doses in w;ora and zero supplementation in w sev;ora. Hence, without R1-6, most of our dose range mediated maximal visual pigment formation. In Drosophila, beta-carotene, zeaxanthin and lutein mediate the formation of all major photopigments in R1-6, R7 and R8.