S-Modulin is a frog homolog of recoverin. The function and the underlying mechanism of the action of these proteins are now understood in general. However, there remain some unsolved issues including; two distinct effects of S-modulin; Ca2+-dependent binding of S-modulin to membranes and a possible target protein; S-modulin-like proteins in other neurons. These issues are considered in this commentary.

This commentary discusses the balance of phosphodiesterase and guanylate cyclase activities in vertebrate photoreceptors at moderate light intensities. The rate of cGMP hydrolysis and synthesis seem to equal each other. Ca2+ as regulator of both enzyme activities is also effectively buffered in photoreceptor cells by cytoplasmic buffer components.

Light adaptation is modulated almost exclusively by changes in intracellular Ca2+ concentration, and other Ca2+-independent mechanisms are likely to play only a minor role. Changes in Ca2+i may be not only necessary for light adaptation to take place but sufficient to cause it.

Hereditary retinal degeneration due to mutations in visual genes may be amenable to therapeutic interventions that modulate, either positively or negatively, the amount of protein product. Some of the proteins involved in phototransduction are rapidly moved by a lightdependent mechanism between the inner segment and the outer segment in rod photoreceptor cells, and this phenomenon is important in phototransduction.

The subunit structure of the cGMP-gated cation channel of rod photoreceptors is rapidly being defined, and in the process the mode of regulation by Ca2+-calmodulin unraveled. Intriguingly, early results suggest that additional subunits of unknown function are associated with the channel and remain to be identified.

Considerable progress has been made toward understanding the involvement of recoverin in a cancer-associated retinopathy (CAR) that results in blindness. We describe the expression of recoverin in tumors of individuals afflicted with CAR, characterize the immunological response towards recoverin in these patients, and demonstrate how the disease can be induced in rodents using recoverin as an immunogen.

It has been suggested that type I adenylyl cyclase may play a unique role in long-term potentiation, due to both unique regulatory properties as well as a specialized distribution within the mammalian brain. This would allow an integration of the signals wrought by increased intracellular calcium with those conveyed into the cellular milieu via increased cAMP. These results are discussed in the context of changes in cellular structure, because of changing interactions between G proteins and cytoskeletal components, which might be expected to accompany chronic synaptic activation.

Our results on hippocampal long-term potentiation are considered in the context of Xia et al.'s hypothesis. Whereas the target article proposes presynaptic PKC involvement in adenylyl cyclase activation by phosphorylation of nenromodulin, we suggest an additional postsynaptic role involving RC3/nenrogranin. Finally, we examine the possibility that the adenylyl cyclase mutant mouse may display normal learning with a selective impairment of memory.

Calcium acting through calmodulin has been shown to regulate the affinity of cyclic nucleotide-gated channels expressed in cell lines. But is calmodulin the Ca-sensor that normally regulates these channels?

Amino-terminal heteroacylation has been identified in retinal proteins including recoverin and α subunit of G-protein, transducin. The tissue-specific modification seems to mediate not only a proteinmembrane interaction but also a specific protein-protein interaction. The mechanism generating the heterogeneity and its physiological role are still unclear, but an interesting idea for the latter postulates a fine regulation of the signal transduction pathway by distinct N-acyl groups.

Inactivation of the Ca2+ extrusion mode of the retinal rod Na- Ca + K exchanger is suggested to be the mechanism that prevents lowering of cytosolic free Ca2+ to < 1 nM when rod cells are saturated for a prolonged time under bright light conditions. Under these conditions, Ca2+ fluxes across disk membranes can contribute significantly to Ca2+ homeostasis in rods.

Magic angle spinning (MAS) NMR methods provide a means of obtaining high resolution structural data on rhodopsin and its photoin termediates. Current work has focused on the structure of the retinal chromophore and its interactions with surrounding protein charges. The recent development of MAS NMR methods for measuring internuclear distances with a resolution of ∼0.2 will complement diffraction methods for addressing key mechanistic questions.

A number of proteins homologous to recoverin have been identified in the brains of the several vertebrate species. The brainderived members originally contain four EF-hand domains, but NH2- terminal domain is aberrant. Many of these proteins inhibited light-induced rhodopsin phosphorylation at high [Ca2+], suggesting that the brain-derived members may act as a Ca2+-sensitive modulator of receptor phosphorylation, as recoverin does.

Genetic abnormalities of three factors related to the photoreceptor mechanism have been reported in both animal models and humans. Apoptotic mechanism has also been suggested as a final common pathway of photoreceptor cell death. Our findings of increased level of glutamate in photoreceptor cells in rds mice suggest that amino acid might mediate between these two pathological mechanisms.

Site-specific phenotypic effects of the 73 known alleles in the rhodopsin gene that cause retinal degeneration are difficult to interpret because most alleles are documented in only one case or one family, which means variation in effects could actually arise from interactions with other loci. However, sample sizes necessary to detect epistatic interaction may place an answer to this question beyond our grasp.

Our understanding of the molecular properties and cellular role of cGMP-gated channels in outer segments of vertebrate photo-receptors has come from over a decade of studies which have continuously altered and refined ideas about these channels. Further examination of this current view may lead to future surprises and further refine the understanding of cGMP-gated channels.

Cyclic nucleotides can regulate the sensitivity of retinal rods to light through phosducin. The phosphorylation state of phosducin determines the amount of G available for activation by Rho*. Phosducin phosphorylation is regulated by cyclic nucleotides through their activation of cAMP-dependent protein kinase. The regulation of phosphodiesterase activity by the noncatalytic cGMP binding sites as well as Ca2+/calmodulin dependent regulation of cGMP binding to the cation channel are also discussed.

In the current model of visual transduction, the lifetime of active cGMP phosphodiesterase depends upon the period of its interaction with GTP-bound transducin. If recoverin regulates the lifetime of light-activated cGMP phosphodiesterase through inhibition of rhodopsin phosphorylation, rhodopsin should directly interact with cGMP phosphodiesterase and/or GTP-bound transducin complexed with cGMP phosphodiesterase. Is this true?

NMR (nuclear magnetic resonance) may be useful for determining the structure of retinal and its environment in rhodopsin, but not for determining the complete protein structure. Aggregation and low yield of fragments of rhodopsin may make them difficult to study by NMR. A long-term multidisciplinary attack on rhodopsin structure is required.

The generally positive response to our target article indicates that most of the commentators accept our contention that light adaptation consists of multiple and possibly redundant mechanisms. The commentaries fall into three general categories. The first deals with putative mechanisms that we chose not to emphasize. The second is a more extended discussion of the role of calcium in adaptation. Finally, additional aspects of cGMP involvement in adaptation are considered. We discuss each of these points in turn.