A Drosophila gene encoding a homologue of vertebrate arrestin was isolated by subtractive hybridization and identified as a member of a set of genes that are preferentially expressed in the visual system. This gene encodes a 364-amino acid protein that displays >40% amino acid sequence identity with human and bovine arrestin. Interestingly, the Drosophila homologue lacks the C-terminal sequences that were postulated to interact with rhodopsin during the quenching of the phototransduction cascade in the vertebrate visual response. These findings are discussed in terms of invertebrate phototransduction. The Drosophila gene was mapped cytogenetically to chromosomal position 36D1-2, near the ninaD locus. However, the arrestin gene does not appear to be the ninaD locus, as sequence analysis of three ethylmethane sulfateinduced ninaD mutant alleles reveals no alteration in amino acid sequence.Phototransduction is the process that converts the energy of an absorbed photon into a change of the ionic permeabilities of the photoreceptor cell membrane. This light-induced change in ionic conductances gives rise to the receptor potential and synaptic activity of the photoreceptor cell. The mechanism of visual excitation in vertebrate photoreceptors is the best understood of all sensory transduction processes (1-4). Light activation of rhodopsin is the first step in the visual response. In the vertebrate, photoactivated rhodopsin molecules activate a guanine nucleotide-binding (G) protein, transducin, which in turn activates a cGMP phosphodiesterase. The reduction of intracellular levels of cGMP leads to the transient closure of a cGMP-gated cation-selective channel and hyperpolarization of the photoreceptor cell.Unlike vertebrates, the microvillar photoreceptors of invertebrates depolarize in response to light and thus open their cation-selective channels. The identity of the intracellular transmitter(s) that mediates excitation in invertebrates has eluded firm identification. Similarly, the enzyme cascade that triggers the visual response has not been defined. However, there is a large body of physiological, genetic, and biochemical work that has strongly implicated calcium and inositol phospholipid metabolism in excitation of dipteran and Limulus photoreceptors (5-7). It is believed that photoactivated rhodopsin interacts with a G protein, which in turn activates a phospholipase C. Phospholipase C would then catalyze the generation of the second messenger inositol 1,4,5-trisphosphate and the subsequent mobilization of calcium from intracellular storage sites. The transient increase in calcium levels (or inositol 1,4,5-trisphosphate) would then lead to the opening of a cation-selective channel and the generation of a depolarizing receptor potential. Strong support for this model was recently provided by the demonstration that the Drosophila no-receptor potential A (norpA) gene encodes a phospholipase C that is abundantly expressed in the adult retina (8,9

A growing number of proteins containing PDZ 1 domains have been shown to play important roles in the organization and/or regulation of signaling events in cells. PDZ domains (or GLGF repeats) were named after three proteins identified over a decade ago: postsynaptic density-95, Drosophila Discs large, and zonula occludens-1 (3-5). These three proteins belong to the membrane-associated guanylate kinase (MAGUK) family of proteins. Most MAGUK proteins contain three PDZ domains, an Src homology 2 domain, and a guanylate kinase-like domain, each having different cellular roles. PDZ domains range from 80 to 100 amino acids in length and typically bind to the carboxyl-terminal sequence of target proteins including receptors, channels, and various signaling molecules to regulate subcellular localization, trafficking, recycling, and/or signaling (6 -10).MUPP1, a protein containing 13 putative PDZ domains, was isolated in a yeast two-hybrid screening for proteins that bound to the carboxyl-terminal tail of the 5-HT 2C R (1). MUPP1 is expressed in many tissues, whereas the 5-HT 2C R is a brainspecific protein (1, 11). The 5-HT 2C R has classically been thought to couple to G q activation; however, additional G protein families have been implicated, leading to the activation of different downstream signaling pathways including phospholipase A 2 , C, or D, and various cation channels (12-17). Since PDZ-containing proteins can scaffold many signaling molecules together into a signal transduction complex, the interaction between MUPP1 and the 5-HT 2C R was further investigated. The 5-HT 2C R contains a PDZ binding motif, Ser 458 -Ser-Val, at its extreme carboxyl terminus, which is critical for interaction with PDZ 10 of MUPP1 (18). In an alternate approach to the yeast two-hybrid system, we independently show that PDZ 10 of MUPP1 is the primary site of interaction for the 5-HT 2C R.Serotonin stimulation has previously been shown to promote phosphorylation of the two serine residues of the 5-HT 2C R PDZ binding motif, Ser 458 and Ser 459 (2). We therefore hypothesize that phosphorylation of the carboxyl-terminal serines of the 5-HT 2C R regulates receptor interaction with MUPP1. To test this hypothesis, we investigated whether a modification of Ser 458 and/or Ser 459 of the 5-HT 2C R carboxyl-terminal tail would alter PDZ 10 interaction. Ser 458 and/or Ser 459 of the receptor tail were mutated to aspartate to mimic phosphorylation (i.e. introduction of a negative charge). Next, cells expressing 5-HT 2C Rs were treated with agonist or antagonist to assess the interaction of the 5-HT 2C R with MUPP1. The results of these experiments support our hypothesis that phosphorylation is a key regulator of 5-HT 2C R interaction with MUPP1. Furthermore, the results indicated that a significant amount of basal phosphorylation of the receptor may also play a yet undetermined role in regulating PDZ-protein interactions. MATERIALS AND METHODS AntibodiesPolyclonal anti-peptide antibodies against amino acids 419 -435 (amino acids RHTNERVARKANDPE...

Arrestins belong to a family of multifunctional adaptor proteins that regulate internalization of diverse receptors including G-protein coupled receptors (GPCRs). Defects associated with endocytosis of GPCRs have been linked to human diseases. We employed eGFP-tagged arrestin 2 (Arr2) to monitor the turnover of the major rhodopsin (Rh1) in live Drosophila. We demonstrate that during degeneration of norpAP24 photoreceptors the loss of Rh1 is parallel to the disappearance of rhabdomeres, the specialized visual organelle that houses Rh1. The cause of degeneration in norpAP24 is due to a failure to activate CaMKII and RDGC because of a loss of light-dependent Ca2+ entry. A lack of activation in CaMKII, which phosphorylates Arr2, leads to hypophosphorylated Arr2, while a lack of activation of RDGC, which dephosphorylates Rh1, results in hyperphosphorylated Rh1. We investigated how reversible phosphorylation of Rh1 and Arr2 contributes to photoreceptor degeneration. To uncover the consequence underlying a lack of CaMKII activation, we characterized ala1 flies in which CaMKII was suppressed by an inhibitory peptide, and showed that morphology of rhabdomeres was not affected. In contrast, we found that expression of phosphorylation deficient Rh1s, which either lack the C-terminus or contain Ala substitution in the phosphorylation sites, was able to prevent degeneration of norpAP24 photoreceptors. This suppression is not due to a loss of Arr2 interaction. Importantly, co-expression of these modified Rh1s offered protective effects, which greatly delayed photoreceptor degeneration. Taken together, we conclude that phosphorylation of Rh1 is the major determinant that orchestrates its internalization leading to retinal degeneration.