Calmodulin

Protein Interactions (part 2/3)

Calmodulin interactions in the Drosophila retina

In the Drosophila retina, Calmodulin is concentrated in the the rhabdomere, a microvillar structure of the photoreceptor cell. Calmodulin is also found in lower amounts in the sub-rhabdomeral cytoplasm. This Calmodulin localization is dependent on the NINAC (neither inactivation nor afterpotential C) unconventional myosins. Mutant flies lacking the rhabdomere-specific p174 NINAC protein do not concentrate Calmodulin in the rhabdomere, whereas flies lacking the sub-rhabdomeral p132 isoform have no detectable cytoplasmic Calmodulin. A defect in vision results when Calmodulin is not concentrated in the rhabdomeres, suggesting a role for Calmodulin in the regulation of fly phototransduction. A general function of unconventional myosins may be to control the subcellular distribution of Calmodulin (Porter, 1993).

The ninaC locus encodes two unconventional myosins, p132 and p174, both consisting of fused protein kinase and myosin head domains expressed in Drosophila photoreceptor cells. NinaC encodes the major Calmodulin-binding proteins in the retina and the NinaC-Calmodulin interaction is required for the normal subcellular localization of Calmodulin as well as for normal photo-transduction. There are two Calmodulin-binding sites in NinaC, C1 and C2, which have different in vitro binding properties. C1 is common to both p132 and p174 while C2 is unique to p174. To address the requirements for Calmodulin binding at each site in vivo, transgenic flies were generated expressing ninaC genes deleted for either C1 or C2. The spatial localization of Calmodulin depends on binding to both C1 and C2. Mutation of either site results in a defective photoresponse. A prolonged depolarization afterpotential (PDA) is elicited at lower light intensities than necessary to produce a PDA in wild-type flies. These results suggest that Calmodulin binding to both C1 and C2 is required in vivo for termination of phototransduction (Porter, 1995).

Phototransduction in Drosophila occurs through inositol lipid signaling that results in Ca2+ mobilization. The physiological roles of calmodulin (CaM) were studied in light adaptation and in regulation of the inward current that is brought about by depletion of cellular Ca2+ stores. Three resources providing decreased Ca-CaM content in photoreceptors were analysed: (1) transgenic Drosophila P[ninaCDeltaB] flies that have CaM-deficient photoreceptors; (2) the peptide inhibitor M5 that binds to Ca-CaM and prevents its action, and (3) Ca2+-free medium that prevents Ca2+ influx and thereby diminishes the generation of Ca-CaM. Several effects have been noted due to decrease in Ca-CaM level:

1. Fluorescence of Ca2+ indicator reveals an enhanced light-induced Ca2+ release from internal stores.
2. Measurements of the light-induced current in P[ninaCDeltaB] cells show a reduced light adaptation.
3. Internal dialysis of M5 initially enhances excitation and subsequently disrupts the light-induced current.
4. An inward dark current appears after depletion of the Ca2+ stores with ryanodine and caffeine.

Importantly, application of Ca-CaM into the photoreceptor cells prevents all of the above effects. It is proposed that negative feedback of Ca-CaM on Ca2+ release from ryanodine-sensitive stores mediates light adaptation, is essential for light excitation, and keeps the store-operated inward current under a tight control (Arnon, 1997b).

Activation of the Drosophila visual cascade is extremely rapid and results in opening of the cation influx channels transient receptor potential (TRP) and transient receptor potential-like (TRPL) within ~10-20 msec of photostimulation of rhodopsin. The G-protein-signaling cascade that leads to opening of the ion channels has been extensively characterized and is known to involve the inositol phospholipid-signaling system. Termination of the photoresponse, after cessation of the light stimulus, is also rapid and is a Ca2+-regulated process; however, understanding of the mechanism by which Ca2+ contributes to termination of the photoresponse is quite incomplete (Li, 1998 and references).

Several proteins have been identified that seem to mediate Ca2+-dependent termination of phototransduction. These include the Ca2+-binding regulatory protein Calmodulin, which functions in both light adaptation and termination of the light response. The ninaC (neither inactivation nor afterpotential C) locus, which encodes two isoforms, p132 and p174, each of which consists of a protein kinase domain fused to a myosin head domain, also functions in negative feedback regulation of the photoresponse. The two NINAC proteins differ because of unique C-terminal ends. p174 is localized exclusively to the microvillar portion of the photoreceptors, the rhabdomeres, and p132 is restricted to the cell bodies. Null mutations in ninaC cause defects in adaptation and response termination. These functions are caused by p174 because elimination of p174, but not p132, causes each of these phenotypes. Because negative feedback regulation seems to be mediated by Ca2+, it is plausible that p174 is regulated by Ca2+. However, p174 does not contain a known Ca2+-binding motif, such as an EF hand or C2 domain, and there is no evidence that it binds Ca2+ directly. Thus, p174 seems to respond to the light-dependent Ca2+ flux indirectly. One NINAC Ca2+ sensor is Calmodulin because NINAC binds to Calmodulin and the NINAC-Calmodulin interaction is required for both adaptation and termination. NINAC might also be regulated by Ca2+-dependent phosphorylation because p174 contains multiple protein kinase C (PKC) consensus sites including several in its unique C-terminal tail. Moreover, mutation of an eye-specific PKC (ePKC) causes perturbations in adaptation and termination. The role of PKC in negative feedback regulation may be more significant than that indicated by mutation of ePKC because a second PKC, brain PKC (brPKC), is known to be enriched in the Drosophila retina and a third PKC, PKC98F, is highly expressed in adult heads. Two retinal substrates for PKC have been identified. These are the TRP cation influx channel and the PSD95, DLG, and ZO-1 (PDZ)-containing protein inactivation, no afterpotential D (INAD), which binds to most of the proteins that function in phototransduction and organizes a supramolecular signaling complex. However, the consequences of disrupting PKC phosphorylation of any retinal substrate that functions in Drosophila vision have not been determined (Li, 1998 and references).

The current work shows that NINAC p174, which consists of a protein kinase domain joined to the head region of myosin heavy chain, is a phosphoprotein and is phosphorylated in vitro by PKC. Mutation of either of two PKC sites in the p174 tail results in an unusual defect in deactivation that has not been detected previously for other ninaC alleles or other loci. After cessation of the light stimulus, there appeared to be a transient reactivation of the visual cascade. This phenotype suggests that a mechanism exists to prevent reactivation of the visual cascade and that p174 participates in this process. The termination mechanisms controlling Drosophila phototransduction seem to be more complicated than previously envisioned. In addition to a requirement for NINAC in facilitating rapid deactivation after cessation of the light stimulus, there is an additional requirement for this unconventional myosin in preventing transient reactivation of the plasma membrane conductances. Because p174 also functions in adaptation, it seems that NINAC has a central role in many aspects of negative feedback regulation of the visual cascade. Recently, a homolog of NINAC has been identified in the mammalian retina (D. Hillman, A. Dose, and B. Burnside, personal communication to Li, 1998). Thus, it is intriguing to speculate that vertebrate NINAC also functions in negative feedback regulation and that an active mechanism may also exist in mammalian photoreceptor cells to ensure stable termination of phototransduction (Li, 1998).

Activation of PI-PLC initiates two independent branches of protein phosphorylation cascades catalyzed by either PKC or Ca2+/calmodulin-dependent protein kinase (CaMK). Phosrestin I (PRI), a Drosophila homolog of vertebrate photoreceptor arrestin, undergoes light-induced phosphorylation on a subsecond time scale that is faster than that of any other protein in vivo. A CaMK activity is responsible for in vitro PRI phosphorylation at Ser366 in the C-terminal tryptic segment, MetLysSer(P)IleGluGlnHisArg, in which Ser(P) represents phosphoserine366. Ser366 is identified as the phosphorylation site of PRI in vivo by identifying the molecular species resulting from in-gel tryptic digestion of purified phospho-PRI. It has been concluded that the CaMK pathway, not the PKC pathway, is responsible for the earliest protein phosphorylation event following activation of PI-PLC in living Drosophila photoreceptors (Matsumoto, 1994).

The characterization of Drosophila Calmodulin mutants and the role of CAM in photoreceptor cell function have been described. In Drosophila photoreceptor neurons, light activation of rhodopsin activates a heterotrimeric G protein, which in turn activates phospholipase C (PLC). PLC catalyzes the hydrolysis of the minor membrane phospholipid phosphatidylinositol-4,5-bisphosphate (PIP2) into the second messengers inositol trisphosphate (IP3) and diacylglycerol (DAG). Activation of PLC then leads to the opening of cation-selective membrane channels encoded by the transient receptor potential (trp) and trp-like (trpl) genes. It has been hypothesized that calcium release from internal stores is required for activation of the phototransduction cascade and that the TRP channel functions as a store-operated channel gated by the light-induced emptying of the internal stores (Scott, 1997 and references).

Contrary to current models of excitation and TRP channel function, the transient phenotype of trp mutants can be explained by CAM regulation of the TRPL channel rather than by the loss of a store-operated conductance leading to depletion of the internal stores. In fact, introduction of calcium intracellularly in trp mutants does not restore responsiveness. The finding that trp mutants can maintain responsiveness in the absence of calcium suggests that there is calcium-dependent inactivation of light-induced currents in the trp mutant. Light responses were analyzed in a variety of mutant and transgenic backgrounds. The transient respone of trp mutants reflects TRPL channel function. Deletion of either of the two CAM binding sites of TRPL results in a prolonged current suggesting that CAM binding functions to inactivate TRPL. Thus, Calmodulin is essential for calcium-dependent negative regulation of phototransduction. Mutants for cam display dramatic defects in deactivation kinetics, displaying greatly prolonged deactivation times. In the absence of extracellular calcium, mutant and wild-type responses are not significantly different from each other, demonstrating that calium entry is required to reveal the cam mutant phenotype and highlighting the absolute requirement for calcium for the rapid deactivation of the phototransduction cascade. CAM ilso regulates the catalytic lifetime of activated rhodopsin by regulating the binding of arrestin to rhodopsin. Thus CAM coordinates termination of the light response by modulating receptor and ion channel activity (Scott, 1997).

To identify Calmodulin-binding proteins that may function in phototransduction and/or synaptic transmission, a screen was conducted for retinal Calmodulin-binding proteins. Twelve Calmodulin-binding proteins were found that are expressed in the Drosophila retina. The functions of Calmodulin appear to be mediated, at least in part, by four previously identified calmodulin-binding proteins: the Trp and Trp-like ion channels, NinaC and InaD. Eight calmodulin-binding proteins have been identified that have not been previously reported to be expressed in the Drosophila retina. The full-length sequences corresponding to three of the calmodulin-binding proteins have been described. These corresponded to two Calmodulin-dependent protein kinases, MLCK and CaM kinase II, as well as to one of two previously described Calcineurin proteins. A third calmodulin-dependent protein kinase is expressed in the Drosophila retina, CaM kinase I. No CaM kinase I has previously been reported from any invertebrate, raising the possibility that this protein kinase is specific to vertebrates. Nevertheless, the Drosophila CaM kinase I is highly related to vertebrate CaM kinase I: ~60% identical over 349 amino acids. The remaining four calmodulin-binding proteins have not been known to bind calmodulin prior to the current work. Six targets have been found that are related to proteins implicated in synaptic transmission. Among these six are a homolog of the diacylglycerol-binding protein (UNC13) and a protein (CRAG) related to Rab3 GTPase exchange proteins. Two other calmodulin-binding proteins found are Pollux, a protein with similarity to a portion of a yeast Rab GTPase activating protein, and Calossin, an enormous protein of unknown function conserved throughout animal phylogeny. Thus, it appears that Calmodulin functions as a Ca2+ sensor for a broad diversity of retinal proteins, some of which are implicated in synaptic transmission (Xu, 1998).

At least two of the four novel calmodulin-binding proteins share similarities to components implicated in synaptic transmission. One of these proteins (1441 residues), is referred to as CRAG (calmodulin-binding protein related to a Rab3 GDP/GTP exchange protein; due to its similarity to a domain in the recently identified rat Rab3 GDP/GTP exchange protein (rRab3 GEP) and the C. elegans homolog, AEX-3, which has been implicated in synaptic vesicle release. The sequences of AEX-3 and the rRab3 GEP were published contemporaneously and have therefore not been directly compared. AEX-3 and the rRab3 GEP (1409 and 1602 amino acids, respectively) contain three regions of homology, the first of which (~500 residues) is conserved in CRAG, AEX-3, rRab3 GEP (CAR domain) and the human homolog, MADD (death domain MAP kinase activator). The latter two regions, are conserved in AEX-3 and rRab3 GEP (AR1 and AR2), but not CRAG, and are shorter (~100 and 300 residues, respectively) than the CAR homology. The CAR domain in CRAG is ~36 identical over 321 amino acids (residues 95-415) to either rRab3 GEP or AEX-3. In addition, there is weak homology (16%) in the flanking sequences that extend the CAR domain in CRAG to residues 73-490. The C-terminal ~800 residues of CRAG do not share significant primary amino acid sequence homology with the rRab3 GEP, AEX-3, or any other protein in the data banks (Xu, 1998 and references).

While it remains to be determined if CRAG is also a Rab3 GEP, such a finding would have interesting implications regarding the mechanism by which GTP exchange on Rab3 is regulated. Rab3 binds to synaptic vesicles; however, this association only occurs in resting nerve terminals and requires Rab3 in the GTP bound state. Mutation of the C. elegans Rab3 GEP, known as AEX-3, causes accumulation of Rab3 in neuronal cell bodies and an impairment in the release of neurotransmitter. Thus, the Rab3 GEP appears to play a critical role in association of Rab3 with synaptic vesicles and in synaptic transmission. The observation that CRAG binds calmodulin implies that the putative GEP activity of this protein could be regulated by changes in Ca²⁺ levels, which are spatially restricted to microdomains near the active zones in presynaptic terminals. A variety of evidence suggests that the absolute level of Rab3-GTP bound to synaptic vesicles regulates the rate of exocytosis by limiting the number of vesicles that can be fused with the plasma membrane. Thus, formation of Rab3-GTP appears to be a crucial step in synaptic transmission. The mechanisms controlling the GDP-GTP exchange are not known but one possibility is that CRAG is a Rab3 GEP and the associated calmodulin provides a sensor to differentiate between the lower Ca²⁺ levels in resting nerve terminals and higher levels resulting from Ca²⁺ influx. While it remains to be determined if CRAG is a Rab3 GEP and whether the exchange activity is regulated through the associated calmodulin, an exchange factor for another small GTPase, RAS, binds to and is regulated by Ca²⁺/calmodulin (Xu, 1998 and references).

Pollux (Plx) is a protein previously reported to be 732 amino acids in length and required for viability (S. D. Zhang, 1996). The protein is predicted to have a transmembrane domain and a leucine zipper (S. D. Zhang, 1996). Plx has now been found to be 1379 amino acids in length and the formerly assigned initiator methionine corresponds to residue 648. A protein related to Plx is TBC1 (Richardson, 1995), a mouse protein which had homology to the majority of Plx. The region in Plx that contains the greatest similarity to TBC1 is a 337-amino acid segment (51% identity, residues 676-1012) that includes the putative transmembrane domain. Of particular interest, the region most highly conserved between Plx and TBC1 includes a 153-amino acid domain (residues 811-963) that displays moderate homology to the yeast Rab family GTPase-activating proteins, GYP6 or GYP7 (Strom, 1994). GYP7 is ~29% identical to this domain in either Plx or TBC1; however, if two gaps of 18 and 36 amino acids have been introduced in Plx and TBC1, the 29% homology extends to over 222 amino acids (742-963). This ~200 amino acid sequence corresponds to the domain previously referred to as a TBC domain due to its similarity to segments in the TRE-2 oncogene and the yeast regulators of mitosis, BUB2 and CDC16. TBC1 is 1141 residues and is found to be a nuclear protein. Thus, TBC1 and Plx have very disparate spatial distributions. (Xu, 1998).

The portion of the Plx protein that was isolated in the screen extends from residues 180-1379. Using a series of overlapping GST fusion proteins and the gel overlay assay, the calmodulin-binding site(s) contained in the original fusion protein was further mapped to residues 657-680. The sequence of the calmodulin-binding site is not conserved in the mouse homolog, TBC1, but is in several human ESTs. A bovine homolog of Plx (Lyncein), which was isolated from a bovine retinal library, is highly conserved in the calmodulin-binding domain despite having no higher overall sequence conservation to Plx than TBC1. Moreover, a fusion protein containing the conserved sequence in Lyncein binds calmodulin. Plx also bind to calmodulin in a pull-down assay; although this interaction is Ca²⁺ independent (Xu, 1998).

Thus it has been found that Plx is 1379 residues rather than 732 amino acids as previously reported (S. D. Zhang, 1996). The additional sequence is not due to a chimeric cDNA since multiple plx cDNAs were obtained and TBC1 shares similarity to Plx both N- and C-terminal to the formerly assigned initiating methionine at residue 648. Plx has been shown to bind calmodulin and does so in a Ca²⁺-independent manner. Although the sequence of the calmodulin-binding site is not conserved in TBC1, the region is very similar in Lyncein, a homolog isolated from a bovine retinal library. Furthermore, the Lyncein sequence also binds calmodulin. Thus, it appears that a Plx homolog is expressed in the vertebrate retina. A possible clue as to the function of Plx in the retina is that it shares some similarity to two yeast Rab GAP proteins, although no homology was found to the Rab3 GAP expressed in the rat brain. Nevertheless, the observation that Plx contains a domain related to Rab GAPs combined with the finding that it appears to be localized to the plasma membrane and lumen of the trachael system raises the possibility that Plx may be involved in exocytosis. In the Drosophila visual system, exocytosis is important not only in synaptic transmission but in turn-over of the microvillar membrane of the photoreceptor cells. Shedding of membrane does not occur uniformly during the diurnal cycle, but occurs maximally soon after dawn. Thus, an increase in the exocytotic process is correlated with the light dependent rise in Ca²⁺ and therefore might be regulated in part by a Ca²⁺ sensing component in a Rab cycle. Alternative potential functions for Plx in photoreceptor cells include other processes that involve vesicular trafficking such as insertion of new membrane in the microvilli and the budding, targeting, and fusion of rhodopsin carrier vesicles with the plasma membrane. These latter events involve a variety of Rab proteins and also appear to be regulated during the daily light cycle (Xu, 1998 and references).

A third protein, not previously known to bind calmodulin, is a Drosophila homolog of UNC13 (dUNC13), a diacylglycerol-binding protein that may be required for release of neurotransmitter from the presynaptic terminal. dUNC13 is expressed as at least two alternatively spliced forms encoding proteins of >1304 (dUNC13A) and >1724 (dUNC13B) amino acids. dUNC13A and dUNC13B shared a common C-terminal region of >1216 amino acids and differ due to unique N-terminal sequences (>88 and >508 residues, respectively). dUNC13 contains extensive homology (>68%) with the C. elegans UNC13 and rat homologs (mUNC13), beginning in the unique region of dUNC13A and extending over the entire region common between both isoforms (residues 72-1304). UNC13 and mUNC13-1 share a similar level of homology over the same region and are only weakly related over the N-terminal ~500 amino acids. The 508 amino acids specific to dUNC13B are not homologous to the UNC13 proteins or any proteins in the data banks. Features common between dUNC13 and other members of the UNC13 family include strong homology to two conserved sequence motifs, C1 and C2, originally recognized in various protein kinase C isoforms. A large variety of other signaling proteins, such as RAF, diacylglycerol kinase, RAS GTPase-activating protein, synaptotagmin, and phoshopholipase C contain these domains. C1 domains typically bind diacylglycerol, while many C2 domains are Ca²⁺-binding regulatory domains. Some C2 domains also bind phospholipids and do so in a Ca²⁺-dependent manner. Other C2 domains confer Ca²⁺ dependence to functions, such as protein kinase activity, mediated by domains distinct from C2. Biochemical analyses of UNC13 demonstrate that it is a bona fide Ca²⁺-dependent phorbol ester-binding protein. The putative C1 domain in dUNC13 (residues 182-232) include six invariant cysteines as well as a seventh cysteine conserved among all UNC13 proteins. Overall, the C1 domain is 92% identical to the corresponding region in mUNC13. The two C2 domains present in each of the three other UNC13 proteins (C2-1 and C2-2) are also found in dUNC13. C2-1 (residues 299-393) and C2-2 (residues 1170-1264) are 76 and 67% identical with the same motifs in mUNC13-1 (Xu, 1998 and references).

In addition to aex-3, several other mutations have been identified in C. elegans that appear to disrupt exocytosis of synaptic vesicles and release of neurotransmitter. One such mutation is in the gene encoding the diacylglycerol-binding protein, UNC13. Although the specific function of UNC13 remains unclear, it may operate in docking and/or fusion of synaptic vesicles since the rat brain-specific mUNC13-1 protein binds directly to two proteins, syntaxin and Doc2alpha, which function in Ca²⁺-dependent exocytosis. The C2 domains present in UNC13 homologs could potentially serve as a Ca²⁺ sensor that responds to the Ca²⁺ influx required for exocytosis. Therefore, the question arises as to the function of a potential second type of Ca²⁺ sensor provided by the binding of calmodulin to dUNC13. One possibility is that each UNC13 protein really has only one Ca²⁺ sensor and that it is supplied in some isoforms by the C2 domain and in others through Ca²⁺/calmodulin. Consistent with this proposal, the calmodulin-binding domain is not conserved in UNC13 suggesting that the C2 domain provides the only Ca²⁺ detector in this protein. The reverse may be the case in mUNC13-1 since this protein does not appear to contain Ca²⁺-binding C2 domains but does show sequence similarity to the dUNC13 calmodulin-binding site. An alternative proposal, which is favored by the authors, is that some UNC13 proteins may be regulated by Ca²⁺ via both C2 domains and calmodulin. Such dual regulation may provide a mechanism for extremely rapid as well as sustained responses to highly transient increases in Ca²⁺. The rise in Ca²⁺, resulting from opening of the voltage-gated channels in synaptic terminals, occurs in microdomains and collapses within microseconds after closing of the ion channels. C2 domains comprise an unusual Ca²⁺ binding motif in that Ca²⁺ appears to regulate this domain through a shift in electrostatic potential rather than a conformational change. As such, C2 domains have the potential to respond very quickly, but transiently, to the rapid Ca²⁺ flux in the active zones of the presynaptic terminal. Although fusion and release of neurotransmitter is extremely rapid (submilliseconds to milliseconds), there is some latency between the opening and closing of the ion channels and these latter events. Ca²⁺ binding to calmodulin, which induces a conformational change, may induce a more delayed but sustained response to Ca²⁺ than that provided by the C2 domain. Thus, dual binding of Ca²⁺ to calmodulin and C2 domains may enable UNC13 proteins to sense the Ca²⁺ rise within a few microseconds and sustain the response for several hundred microseconds to several milliseconds (Xu, 1998 and references).

The fourth novel calmodulin-binding protein is referred to as Calossin (CALO) due to its interaction with calmodulin and colossal molecular mass (predicted >450 kDa). Several overlapping cDNAs have been obtained resulting in the identification of a single open reading frame encoding >4118 amino acids. Several hydrophobic regions are predicted according to a computer algorithm; however, it is unclear if any is sufficiently long to span a lipid bilayer. CALO is related to a predicted C. elegans protein (cCalossin) of similar size (3864 residues) that was identified as part of the C. elegans Genome Sequencing Consortium. The homology between CALO and cCALO was not uniform but concentrated in several domains. The longest continuous region of identity begins at amino acid 2460 and extends ~1650 residues to near the C terminus. In addition, there are two shorter stretches of similarity between residues 604 and 1150. The highest levels of identity (each ~70%) are in three ~50-100 amino acid regions: (1) residues 604-649; ( 2) residues 2587-2638, and (3) residues 3276-3380. The first two of these conserved regions are cysteine-rich domains, CRD1 and CRD2, respectively, that resemble different classes of zinc finger domains. CRD1 is most similar to the zinc finger family defined by Requiem, a protein required for apoptosis, while CRD2 shares features equally well with several families of zinc family proteins and can not be included within a single group (Xu, 1998 and references).

Regulation of the Rhodopsin protein phosphatase Rdgc by Calmodulin

Hundreds of G protein-coupled receptors (GPCRs) and at least six GPCR kinases have been identified, but the only GPCR phosphatase that has been definitively demonstrated is the rhodopsin phosphatase encoded by the rdgC locus of Drosophila. Mutations in rdgC result in defects in termination of the light response and cause severe retinal degeneration. RDGC is shown to bind to Calmodulin, and a mutation in an IQ motif that eliminates the Calmodulin/RDGC interaction prevents dephosphorylation of rhodopsin in vivo and disrupts termination of the photoresponse. These data indicate that RDGC is a novel calmodulin-dependent protein phosphatase and raise the possibility that regulation of other GPCRs through dephosphorylation may be controlled by calmodulin-dependent protein phosphatases related to RDGC (Lee, 2001).

Stimulation of G protein-coupled receptors (GPCRs) by hormones, growth factors, neurotransmitters, sensory stimuli, and other agonists frequently results in an increase in intracellular Ca²⁺. Such changes in Ca²⁺ concentration regulate a variety of effects ranging from apoptosis to differentiation, cell movement, the modulation of synaptic plasticity, and visual transduction. A primary mechanism through which alterations in Ca²⁺ levels lead to discrete physiological consequences involves the control of protein phosphorylation by the Ca²⁺ sensor calmodulin. Several calmodulin-dependent serine/threonine kinases have been described, such as myosin light chain kinase and Ca²⁺/calmodulin-dependent protein kinases I, II, and IV. However, the only known calmodulin-dependent protein phosphatase is calcineurin, despite the ~20 years that have elapsed since its discovery. Calcineurin is conserved from yeast to humans and is a heterodimer consisting of a catalytic subunit, CnA, and a regulatory subunit, CnB, comprised of four Ca²⁺ binding motifs referred to as EF hands (Lee, 2001).

Given the central role of Ca²⁺ in regulating a vast array of essential processes, it seems likely that there exist additional calmodulin-dependent protein phosphatases. Candidates include phosphatases known to function in Ca²⁺-regulated signaling cascades. One such protein is the rhodopsin phosphatase, RDGC, which participates in Drosophila phototransduction (Steele, 1990; Steele, 1992; Byk, 1993; Vinós, 1997). Drosophila phototransduction culminates with Ca²⁺ and Na⁺ influx via the TRP, TRPL, and TRPgamma channels (Lee, 2001).

A key mediator of the Ca²⁺-mediated feedback regulation is calmodulin; however, there are only a few signaling proteins known to function in Drosophila phototransduction that bind to calmodulin. These include the NINAC myosin III, TRP, TRPL, and INAD. Some calmodulin binding proteins, such as TRPL, interact with calmodulin via the positive face of an amphiphilic alpha helix, while others, such as NINAC, associate through IQ motifs. IQ motifs, which contain the core consensus IQxxxRGxxxR (x denotes any amino acid), may associate with calmodulin in either a Ca²⁺-dependent or independent manner. IQ motifs are present in a wide diversity of proteins ranging from myosins to neuromodulin (GAP-43, voltage-gated Ca²⁺ channels, and the Ras guanine nucleotide exchange factor, RAS-GRF). Moreover, the IQ/calmodulin interactions regulate the activities of each of these classes of proteins (Lee, 2001).

The rhodopsin phosphatase, RDGC, is a potential target for regulation by calmodulin, since mutations in rdgC result in severe defects in the Ca²⁺-dependent termination of the photoresponse (Steele, 1990; Vinós, 1997). Disruption of rdgC function also results in age- and light-dependent retinal degeneration (Steele, 1990). Homologs of RDGC are conserved from C. elegans to humans and are collectively referred to as the PPEF family due to the protein phosphatase domain and the presence of multiple C-terminal Ca²⁺ binding motifs, EF hands (Lee, 2001).

The present work shows that RDGC is a calmodulin-regulated protein phosphatase. RDGC binds directly to calmodulin, and this interaction disrupts an association between the N-terminal domain of RDGC and the catalytic domain. Furthermore, the calmodulin/RDGC interaction is required to potentiate dephosphorylation of rhodopsin in vivo and for rapid termination of the photoresponse (Lee, 2001).

To test whether RDGC binds to calmodulin, calmodulin-agarose pull-down assays were performed. Agarose beads conjugated to calmodulin were incubated with full-length RDGC labeled in vitro with ³⁵S. As a negative control, a segment of the PDZ protein, INAD (PDZ domains 3 and 4; amino acids 346-581), which is devoid of calmodulin binding activity, was used. RDGC binds to calmodulin-agarose, although the INAD-PDZ3-4 segment does not. Binding of RDGC to calmodulin-agarose is not strictly Ca²⁺ dependent. However, a greater proportion of RDGC binds to calmodulin in the presence of Ca²⁺ (Lee, 2001).

To determine whether RDGC and calmodulin interact in vivo, whether the two proteins coimmunoprecipitate from fly heads was tested. Anti-RDGC antibodies were generated that recognize three bands (84, 78, and 76 kDa) in wild-type but not rdgC head extracts. The 78 and 76 bands appear to be eye specific, since they are not detected in the eyeless mutant, sine oculis (so). The 84 kDa isoform, which is not eye specific, may be responsible for the previously reported RDGC expression in the mushroom bodies of the central brain (Steele, 1992). To assess whether RDGC and calmodulin associate in vivo, RDGC was immunoprecipitated from wild-type or null rdgC head extracts, and Western blots of the immune complexes were probed with anti-calmodulin antibodies. Calmodulin is detected in the immune complexes from wild-type but not from null rdgC head extracts. Furthermore, calmodulin coimmunoprecipitates with RDGC in the presence or absence of Ca²⁺, although more calmodulin immunoprecipitates in the presence of Ca²⁺ (Lee, 2001).

To map the sites of interaction between RDGC and calmodulin, calmodulin overlay assays were used. Various fragments of RDGC were expressed in E. coli as GST fusion proteins. Total bacterial extracts were resolved by SDS-PAGE, transferred to PVDF membranes, and probed with [¹²⁵I]calmodulin. All the fusion proteins that bind to calmodulin contain the N-terminal 32 residues, while those proteins that lack these residues fail to associate with calmodulin. Thus, the N-terminal 32 residues contain the calmodulin binding site (Lee, 2001).

RDGC contains a sequence similar to the IQ-calmodulin binding motif, and this sequence maps to the N terminus of RDGC (residues 12-22). To address whether RDGC binds to calmodulin through the IQ motif, most of the sequence (deltaIQ, amino acids 12-22) was deleted and the effects on calmodulin binding were assessed using calmodulin overlay assays. GST-fusion proteins containing full-length RDGC or the N-terminal 73 residues of RDGC bind calmodulin (RDGC and 73^WT). However, derivatives of these fusion proteins that lacked residues 12-22 fail to associate with calmodulin (RDGC^deltaIQ and 73^deltaIQ) (Lee, 2001).

To obtain additional evidence that RDGC binds to calmodulin through the IQ motif, the effects of a variety of conservative and nonconservative substitutions of the most invariant residues within the IQ sequence were assessed. Among the point mutations generated, the only one that almost completely abolishes interaction with calmodulin is a glutamic acid substitution of the isoleucine (residue 12) that begins the motif (73^I12E). A relatively conservative alanine substitution of the same residue (73^I12A) does not eliminate calmodulin binding, but appears to result in an increase in the level of calmodulin bound to RDGC. Thus, it appears that the IQ sequence is the only calmodulin binding site in RDGC (Lee, 2001).

To confirm the effects of the single amino acid substitutions in residue 12 and to test whether these alterations influence the effect of Ca²⁺ on the RDGC/calmodulin interaction, calmodulin-agarose pull-down assays were performed. Consistent with the results of the overlay assay, 73^deltaIQ and 73^I12E virtually abolished binding whereas 73^I12A displays an increase in calmodulin binding relative to 73^WT. In addition, the I12A mutation alters the Ca²⁺ dependence for calmodulin binding. While wild-type RDGC binds calmodulin in the presence or absence of Ca²⁺, the interaction between 73^I12A and calmodulin is strictly Ca²⁺ dependent (Lee, 2001).

Other members of the RDGC/PPEF family also contain an N-terminal IQ consensus sequence, suggesting that other RDGC-related proteins may interact with calmodulin. To test whether the IQ sequence in human PPEF2 is a calmodulin binding site, calmodulin-agarose pull-down assays were performed. The core region of the IQ motif (amino acids 21-42) of human PPEF2 was fused to GST (GST-HsIQ), purified using a glutathione-Sepharose column, and incubated with calmodulin-agarose beads. GST-HsIQ binds to calmodulin-agarose, but GST alone does not. In addition, GST-HsIQ binds to calmodulin-agarose in the presence or absence of Ca²⁺. Thus, the IQ-type motif domain in human PPEF-2 is sufficient to bind calmodulin (Lee, 2001).

To determine the physiological role of the RDGC/calmodulin interaction, transgenic flies were generated that express full-length derivatives of RDGC that incorporate the I12A mutation, the I12E mutation, and the IQ deletion of amino acids 12-22. All of the mutations were introduced into the 7.1 kb genomic DNA shown to restore wild-type visual function to rdgC-mutant flies (Steele, 1992). The wild-type (P[rdgC⁺]) and mutant transgenes (P[rdgC^I12A], P[rdgC^I12E], and P[rdgC^deltaIQ]) were expressed in a null rdgC (rdgC^co6) background so that the only RDGC proteins originated from the transgenes. The wild-type transgene restored normal levels of both eye-specific isoforms of RDGC. In addition, the point mutations (I12A and I12E) do not affect the stability of RDGC, since the level of the proteins was similar to that observed in wild-type (y,w) and P[rdgC⁺] flies. However, the deletion of the IQ motif appears to render the protein unstable, since the RDGC^deltaIQ is undetectable (Lee, 2001).

To test whether calmodulin binding is critical for RDGC function in vivo, the retinal morphology of the transgenic flies was examined. Drosophila compound eyes consist of ~800 ommatidia, each of which contains eight photoreceptor cells, though only seven are present in any given plane of section. The photoreceptor cells include a microvillar structure, the rhabdomere, which contains the proteins critical for visual transduction. The morphology of the wild-type and P[rdgC⁺] rhabdomeres do not change with age or depending on the light conditions. However, rdgC flies are characterized by age- and light-dependent retinal degeneration (Steele, 1990). One-day-old rdgC flies maintained under a 12 hr light/12 hr dark cycle show little if any decrease in the size of the rhabdomeres. After rearing rdgC flies for 7 days under a light-dark cycle, the rhabdomeres of the R1-6 cells are almost completely degenerated. The rhabdomeres of rdgC^I12E flies show a pattern of degeneration similar to rdgC null flies. By contrast, no degeneration is observed in rdgC^I12A flies aged for 7 days under the light-dark cycle. These data indicated that the I12E but not the I12A mutation disrupted RDGC function (Lee, 2001).

To test whether association of calmodulin with RDGC is critical for rapid inactivation, the light response was examined using electroretinograms (ERGs). ERGs are extracellular recordings that measure the summed responses of the retinal cells to light. Exposure of rdgC⁺ flies to a light stimulus results in a corneal negative deflection in the ERG. Upon termination of the light response, there was a rapid return to the baseline. The deactivation of the light response is significantly delayed in rdgC null mutant flies. There was no delay in termination in rdgC^I12A; rather, the deactivation kinetics in these flies is slightly faster than rdgC⁺. However, the deactivation rate of rdgC^I12E decreases significantly compared with rdgC⁺, although the delay is slightly less severe than in the null mutant (Lee, 2001).

When a substantial amount of rhodopsin is photoconverted from rhodopsin to the light activated metarhodopsin state, a sustained photoresponse or prolonged depolarization afterpotential (PDA) persists after cessation of the light stimulus. In wild-type, intense light is required to produce a PDA. Blue rather than orange or white light is most effective in producing a PDA, since rhodopsin, but not metarhodopsin, is maximally activated by blue light. A second photon of light, white or orange, is required to convert metarhodopsin to rhodopsin and terminate the PDA (Lee, 2001).

A PDA can be generated in rdgC with considerably less light than in rdgC⁺. A PDA results from an excess of metarhodopsin relative to arrestin. Therefore, mutants that express less arrestin than wild-type also exhibit a PDA with less intense light. It has been suggested that rdgC flies display a PDA with less light than wild-type, since mutations in rdgC cause hyperphosphorylation of rhodopsin, which in turn impairs arrestin function (Lee, 2001).

To address whether the rdgC transgenic flies display defective PDAs, the relative intensity of light needed to produce a PDA was assayed. Like rdgC null mutants, rdgC^I12E enters a PDA with <10% the light intensity required in rdgC⁺. In contrast, rdgC^I12A shows a PDA similar to rdgC⁺. The observations that rdgC^I12E flies display retinal degeneration, a delay in deactivation kinetics, and a low light PDA indicates that calmodulin binding is necessary for normal function of RDGC in vivo (Lee, 2001).

The simplest hypothesis to account for the requirement of the RDGC/calmodulin interaction is that calmodulin regulates the phosphatase activity of RDGC in vivo. To test this proposal, the relative levels of phosphorylated rhodopsin (Rh1) were measured in rdgC⁺ and rdgC^I12E eyes. Dark-reared flies were fed ³²P and then exposed to light. As expected, rhodopsin is hyperphosphorylated in rdgC relative to rdgC⁺. Of primary importance here, rhodopsin is also hyperphosphorylated in the rdgC^I12E eyes. However, the level of rhodopsin phosphorylation in rdgC^I12A was similar to control flies (rdgC⁺). ninaE^P332 flies, which express only 0.1% of rhodopsin, were used as a negative control. Thus, association of calmodulin with RDGC appears to be required for RDGC phosphatase activity in vivo (Lee, 2001).

Whether calmodulin potentiates RDGC activity in vitro was tested using purified recombinant RDGC and p-nitrophenyl phosphate as a pseudosubstrate. Consistent with studies of a human homolog of RDGC, the activity of RDGC is increased by Ca²⁺, possibly through interaction of Ca²⁺ with the EF hands. Similar results were obtained with RDGC^I12E. Significantly, the addition of calmodulin in the presence of 100 µM Ca²⁺ further increases the phosphatase activity of wild-type RDGC but not that of RDGC^I12E. Thus, the enzymatic activity of RDGC seems to be augmented by calmodulin in vitro, although the level of calmodulin-dependent potentiation is much greater in vivo than in vitro (Lee, 2001).

Calmodulin might augment RDGC phosphatase activity by relieving a putative intramolecular interaction that inhibits the phosphatase activity. As a first test of this model, an examination was carried out to see whether the N-terminal region of RDGC, which includes the calmodulin binding site, associates with the catalytic domain. ³⁵S-labeled N-terminal RDGC (amino acids 1-253) binds to a GST-catalytic domain fusion protein (amino acids 153-423) but not to GST alone. Moreover, neither the middle nor the C-terminal portions of RDGC associate with the catalytic domain (Lee, 2001).

To test whether calmodulin binding to RDGC affects the interaction between the N-terminal portion of RDGC and the catalytic domain, pull-down assays were performed. ³⁵S-labeled N-terminal RDGC was incubated with the GST-catalytic domain fusion in the presence or absence of Ca²⁺/calmodulin. The addition of Ca²⁺/calmodulin decreases the interaction between N-terminal RDGC and the GST-catalytic domain fusion nearly 5-fold. However, the interaction between the catalytic domain and the N-terminal fragment of RDGC containing the I12E is not decreased by Ca²⁺/calmodulin. The addition of only BSA or Ca²⁺ does not change the binding of N-terminal RDGC and the catalytic domain (Lee, 2001).

Thus, several lines of evidence support the conclusion that RDGC is a calmodulin-regulated protein phosphatase. RDGC binds to calmodulin in vitro and in vivo, and the interaction is through an established calmodulin binding sequence, the IQ motif. An I to E substitution in this motif disrupts calmodulin binding and interfers with dephosphorylation of rhodopsin in vivo. Calmodulin also potentiates dephosphorylation of a pseudosubstrate in vitro, although the effect is smaller than that observed in vivo, possibly due to a contribution of one or more cofactors that remain to be identified. Finally, addition of calmodulin interfers with an intramolecular interaction between the N-terminal region and the RDGC catalytic domain (Lee, 2001).

It had been assumed previously that RDGC is not regulated by calmodulin since an inhibitor of calmodulin, M5, does not perturb the RDGC-dependent phosphatase activity in fly head extracts (Byk, 1993). However, in addition to RDGC, the activities of several other calmodulin-regulated proteins, such as a Ca²⁺-activated K⁺ channel and an L-type Ca²⁺ channel, are unaffected by application of calmodulin inhibitors. Thus, a lack of effect of calmodulin inhibitors does not rule out the possibility that a given protein interacts with and is regulated by calmodulin (Lee, 2001).

The identification of RDGC as a calmodulin-regulated protein phosphatase addresses a key question in Drosophila phototransduction, the identity of in vivo targets for negative feedback regulation by Ca²⁺/calmodulin. Calmodulin is present at ~0.5 mM concentration in the rhabdomeres and a decrease in rhabdomeral calmodulin has a profound effect on termination of the photoresponse. However, the only rhabdomeral proteins previously shown to be regulated in vivo by calmodulin are the NINAC myosin III, the TRPL cation channel, and Arrestin 2. As is the case for the major rhodopsin, Rh1, Arrestin 2 also undergoes rapid light-dependent phosphorylation. The phosphorylation of Arrestin 2, which is mediated by Ca²⁺/calmodulin-dependent protein kinase II, appears to regulate the release of Arrestin 2 from Rh1 (Lee, 2001).

Despite the observation that Rh1 undergoes light-dependent phosphorylation, the role of this phosphorylation event is controversial. In mammals, desensitization or inactivation of rhodopsin and other GPCRs is initiated by phosphorylation by GPCR kinases, which increases the affinity for arrestin. Arrestin binding disrupts signaling by interfering with engagement of the receptor with the G protein. The GPCR must be subsequently dephosphorylated before the inactivation and recycling is complete (Lee, 2001).

Evidence that phosphorylation of Rh1 is required for the photoresponse is that mutations in rdgC result in hyperphosphorylation of the receptor and a defect in response termination (Steele, 1990; Byk, 1993; Vinós, 1997). However, a C-terminal truncation of Rh1 (Rh1^delta356), or a combination of mutations that eliminate the phosphorylation sites, has no apparent effect on the photoresponse or on Arrestin 2 binding. Nevertheless, the defect in termination in rdgC flies appears to be due to hyperphosphorylation of the receptor, since the rdgC phenotype is suppressed in rdgC/Rh1^delta356 double-mutants. To reconcile these findings, it has been proposed that the C terminus of Rh1 is an autoinhibitory domain for Arrestin 2 binding and phosphorylation relieves the intramolecular interaction. The mutations that prevent phosphorylation of Rh1 may also eliminate the putative autoinhibitory interaction, thereby abrogating the requirement for phosphorylation (Lee, 2001).

Rhodopsin cannot be the only substrate for RDGC, since it is also expressed in a region of the brain, the mushroom bodies, implicated in learning and memory. Furthermore, while one of the two human RDGC homologs, PPEF-2, is highly enriched in the retina, the other, PPEF-1, is expressed primarily in a variety of sensory neurons of neural crest origin. Thus, PPEF-1 must also engage substrates other than rhodopsin. Likely candidates are other GPCRs that initiate signaling cascades that culminate in a rise in intracellular Ca²⁺. The identification of such substrates should provide valuable insights into additional roles for this new class of calmodulin-regulated protein phosphatases (Lee, 2001).

Miscellaneous interactions

Continued: see Calmodulin Protein interactions part 3/3 | back to part 1/3

Home page: The Interactive Fly © 1997 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.