RESUMO
PURPOSE: Rhabdomere microvilli dramatically reorganize in conditions of light and dark. This reorganization involves remodeling of the microvillus actin cytoskeleton. We are using the rhabdomeric retina of Octopus bimaculoides to identify actin-binding proteins that may be involved in this remodeling. METHODS: Octopus were light-/dark-adapted, retinas separated into dorsal and ventral halves, and homogenized. Actin-binding proteins were recognized using F-actin overlay blot assays and selected proteins from the overlays were identified using N-terminal sequencing methods or mass spectroscopy. Protein concentrations were quantified and compared by statistical analysis. RESULTS: Total protein gels of light-/dark-adapted, ventral/dorsal halves were almost identical except for a protein band at 26 kD. The relative amount of this protein in the dark was almost double that found in the light. The levels of other proteins did not vary significantly between the light and dark. F-actin overlays also showed matching patterns of actin-binding proteins except for the 26 kD protein. Although the 26 kD protein from light-adapted retinas transferred to the blotting membranes, it did not bind F-actin while the 26 kD protein on overlays from dark-adapted retinas always demonstrated F-actin binding. Besides the 26 kD protein, other proteins at 200 kD, 80 kD, 40 kD appeared on the overlays. These proteins and the 26 kD protein were sequenced and identified as hemocynanin, transitional ER ATPase, arginine kinase and S-crystallin, respectively. CONCLUSIONS: The amount of S-crystallin present in the octopus retina is significantly greater in dark-adapted retinas and it binds to F-actin. In the light, the level of S-crystallin is greatly reduced and there is no apparent F-actin binding. No other studies, to our knowledge, show that S-crystallin binds to the actin cytoskeleton or that its expression is regulated by light. Arginine kinase may provide energy for cytoskeletal remodeling as it may in other neural tissues.
