Stabilizing function for myristoyl group revealed by the crystal structure of a neuronal calcium sensor, guanylate cyclase-activating protein 1

Guanylate cyclase-activating proteins (GCAPs) are Ca(2+)-binding proteins myristoylated at the N terminus that regulate guanylate cyclases in photoreceptor cells and belong to the family of neuronal calcium sensors (NCS). Many NCS proteins display a recoverin-like "calcium-myristoyl switch" whereby the myristoyl group, buried inside the protein in the Ca(2+)-free state, becomes fully exposed upon Ca(2+) binding. Here we present a 2.0 A resolution crystal structure of myristoylated GCAP1 with Ca(2+) bound. The acyl group is buried inside Ca(2+)-bound GCAP1. This is in sharp contrast to Ca(2+)-bound recoverin, where the myristoyl group is solvent exposed. Furthermore, we provide direct evidence that the acyl group in GCAP1 remains buried in the Ca(2+)-free state and does not undergo switching. A pronounced kink in the C-terminal helix and the presence of the myristoyl group allow clustering of sequence elements crucial for GCAP1 activity.     

The myristoylation of the neuronal Ca2+ -sensors guanylate cyclase-activating protein 1 and 2

Guanylate cyclase-activating proteins (GCAPs) are Ca(2+)-binding proteins with a fatty acid (mainly myristic acid) that is covalently attached at the N terminus. Myristoylated forms of GCAP were produced in E. coli by coexpression of yeast N-myristoyl-transferase. Proteins with nearly 100% degree of myristoylation were obtained after chromatography on a reversed phase column. Although proteins were denatured by this step, they could be successfully refolded. Nonmyristoylated GCAPs activated bovine photoreceptor guanylate cyclase 1 less efficiently than the myristoylated forms. Maximal activity of guanylate cyclase at low Ca(2+)-concentration decreased about twofold, when GCAPs lacked myristoylation. In addition, the x-fold activation of cyclase was lower with nonmyristoylated GCAPs. Myristoylation of GCAP-2 had no influence on the apparent affinity for photoreceptor guanylate cyclase 1, but GCAP-1 has an about sevenfold higher affinity for cyclase, when it is myristoylated. We conclude that myristoylation of GCAP-1 and GCAP-2 is important for fine tuning of guanylate cyclase activity.     

Irregular dimerization of guanylate cyclase-activating protein 1 mutants causes loss of target activation.

Guanylate cyclase-activating proteins (GCAPs) are neuronal calcium sensors that activate membrane bound guanylate cyclases (EC 4.6.1.2.) of vertebrate photoreceptor cells when cytoplasmic Ca2+ decreases during illumination. GCAPs contain four EF-hand Ca2+-binding motifs, but the first EF-hand is nonfunctional. It was concluded that for GCAP-2, the loss of Ca2+-binding ability of EF-hand 1 resulted in a region that is crucial for targeting guanylate cyclase [Ermilov, A.N., Olshevskaya, E.V. & Dizhoor, A.M. (2001) J. Biol. Chem.276, 48143-48148]. In this study we tested the consequences of mutations in EF-hand 1 of GCAP-1 with respect to Ca2+ binding, Ca2+-induced conformational changes and target activation. When the nonfunctional first EF-hand in GCAP-1 is replaced by a functional EF-hand the chimeric mutant CaM-GCAP-1 bound four Ca2+ and showed similar Ca2+-dependent changes in tryptophan fluorescence as the wild-type. CaM-GCAP-1 neither activated nor interacted with guanylate cyclase. Size exclusion chromatography revealed that the mutant tended to form inactive dimers instead of active monomers like the wild-type. Critical amino acids in EF-hand 1 of GCAP-1 are cysteine at position 29 and proline at position 30, as changing these to glycine was sufficient to cause loss of target activation without a loss of Ca2+-induced conformational changes. The latter mutation also promoted dimerization of the protein. Our results show that EF-hand 1 in wild-type GCAP-1 is critical for providing the correct conformation for target activation.     

Operation profile of zebrafish guanylate cyclase-activating protein 3

The expression pattern and property profile of the neuronal Ca(2+) sensor guanylate cyclase-activating protein 3 (zGCAP3) was studied by immunochemical approaches, biophysical methods and enzymatic assays. Using affinity purified antibodies immunoreactivity towards zGCAP3 was weakly detected in the outer and strongly in the inner segments of cone cells as well as in the outer plexiform layer, to a lesser degree also in the inner plexiform and ganglion cell layer of the zebrafish retina. This cellular distribution was independent of a dark/light cycle. Some neuronal Ca(2+) sensors are acylated (mainly myristoylated) at the amino-terminus. Probing larval and adult stages of the developing zebrafish retina indicated that zGCAP3 was first expressed in a non-myristoylated form, but was finally present in the adult retina as a myristoylated protein. While zGCAP3 did not undergo a classical Ca(2+) -myristoyl switch as investigated by surface plasmon resonance spectroscopy, myristoylation had two main other consequences: it enhanced the Ca(2+) -sensitivity of the Ca(2+) -induced conformational change and it stabilized the protein conformation. Differences between myristoylated and non-myristoylated zGCAP3 were also observed in modulating the kinetic and catalytic parameters of the GCAP-target, a membrane bound guanylate cyclase. Thus, the stabilizing effect of the myristoyl group is apparently less important in the larval than in the adult fish.     

Protection of neuronal calcium sensor 1 protein in cells treated with paclitaxel

Paclitaxel (Taxol) is one of the most effective treatment options for patients suffering from a variety of cancers. A major side effect seen in a high percentage of patients treated with paclitaxel is irreversible peripheral neuropathy. We previously reported that prolonged treatment with paclitaxel activates a calcium-dependent enzyme, calpain, which degrades neuronal calcium sensor 1 (NCS-1) and subsequent loss of intracellular calcium signaling. Because it appears that activation of calpain is an early step in this destructive cascade, we proposed that inhibition of calpain will protect against the unwanted side effects of paclitaxel treatment. First, NCS-1 levels and intracellular calcium signaling were found to be protected by the presence of lactacystin, a protesome inhibitor. To reinforce the role of calpain in this process, we showed that increased concentrations of calpastatin, a naturally occurring calpain inhibitor, were protective. Next, we tested two mutated versions of NCS-1 developed with point mutations at the P2 position of the calpain cleavage site of NCS-1 to decrease the likelihood of NCS-1 degradation. One mutant was cleaved more favorably by calpain compared with NCS-1 WT, whereas the other mutant was less favorably cleaved. Expression of either mutated version of NCS-1 in neuroblastoma cells protected intracellular calcium signals from paclitaxel-induced changes. These results support our hypothesis that it is possible to protect cells from paclitaxel-induced degradation of NCS-1 by inhibiting calpain activity.     

The cone-specific calcium sensor guanylate cyclase activating protein 4 from the zebrafish retina

Guanylate cyclase activating proteins (GCAPs) serve as neuronal Ca²⁺-sensor proteins in vertebrate rod and cone photoreceptor cells. Zebrafish express in their retina a variety of six different GCAPs, of which four are specific for cone cells. One isoform, zGCAP4, is mainly expressed in double cones and long single cones. We cloned the zGCAP4 gene, purified non-myristoylated and myristoylated forms of the protein after heterologous expression in Escherichia coli and studied its properties: zGCAP4 was a strong activator of membrane-bound guanylate cyclases from bovine and zebrafish retina, showing half-maximal activation at 520–570 nM free Ca²⁺ concentration. Furthermore, the Ca²⁺-sensitive activation properties of non-myristoylated and myristoylated zGCAP4 were similar, indicating no influence of the myristoyl moiety on Ca²⁺-sensor function. Myristoylated zGCAP4 showed low affinity for membranes and did not exhibit a Ca²⁺–myristoyl switch, a feature typical of some but not all neuronal Ca²⁺-sensor proteins. However, tryptophan fluorescence studies and Ca²⁺-dependent differences in protease accessibility revealed Ca²⁺-induced conformational changes in myristoylated and non-myristoylated zGCAP4, indicating the operation as a Ca²⁺ sensor. Thus, expression and biochemical properties of zGCAP4 are in agreement with its function as an efficient Ca²⁺-sensitive regulator of guanylate cyclase activity in cone vision.     

Differential calcium signaling by cone specific guanylate cyclase-activating proteins from the zebrafish retina

Zebrafish express in their retina a higher number of guanylate cyclase-activating proteins (zGCAPs) than mammalians pointing to more complex guanylate cyclase signaling systems. All six zGCAP isoforms show distinct and partial overlapping expression profiles in rods and cones. We determined critical Ca(2+)-dependent parameters of their functional properties using purified zGCAPs after heterologous expression in E.coli. Isoforms 1-4 were strong, 5 and 7 were weak activators of membrane bound guanylate cyclase. They further displayed different Ca(2+)-sensitivities of guanylate cyclase activation, which is half maximal either at a free Ca(2+) around 30 nM (zGCAP1, 2 and 3) or around 400 nM (zGCAP4, 5 and 7). Zebrafish GCAP isoforms showed also differences in their Ca(2+)/Mg(2+)-dependent conformational changes and in the Ca(2+)-dependent monomer-dimer equilibrium. Direct Ca(2+)-binding revealed that all zGCAPs bound at least three Ca(2+). The corresponding apparent affinity constants reflect binding of Ca(2+) with high (≤ 100 nM), medium (0.1-5 μM) and/or low (≥ 5 μM) affinity, but were unique for each zGCAP isoform. Our data indicate a Ca(2+)-sensor system in zebrafish rod and cone cells supporting a Ca(2+)-relay model of differential zGCAP operation in these cells.     

Identification of functional regions of guanylate cyclase-activating protein 1 (GCAP1) using GCAP1/GCIP chimeras

Guanylate cyclase-activating protein 1 (GCAP1) and guanylate cyclase-inhibitory protein (GCIP) are calmodulin-related Ca2+-binding proteins expressed in vertebrate photoreceptor cells. GCAP1 activates photoreceptor guanylate cyclase 1 (GC1) at low free [Ca2+] (<50 nM, in the light), but inhibits it at physiological high [Ca2+] (1 microM, in the dark). GCIP, a Ca2+-binding protein from frog retina, inhibits GC1 at approximately 1 microM [Ca2+], but is unable to stimulate cyclase at low [Ca2+]. In this study, we probed the interaction between GCAP1 and GC1 by producing GCAP1/GCIP chimeras and tested their capability to stimulate GC1. We prepared eight pairs of constructs in which the N-terminal portions of GCIP and GCAP1 were successively replaced by corresponding domains of GCAP1, and GCIP, respectively. The expressed proteins were purified and tested for stimulation of GC1 at 50 nM [Ca2+], and their ability to competitively inhibit GC1 stimulation by a Ca2+-insensitive GCAP1 mutant, GCAP1-tm, at high [Ca2+]. While all GCAP1/GCIP chimeras competitively inhibited GC1 stimulation at high [Ca2+] by GCAP1-tm, several of the GCIP/GCAP1 chimeras had no effect. A chimera consisting of residues 1-20 of GCIP and 21-205 of GCAP1 had no effect on GC1 at low [Ca2+], suggesting that the N-terminal region MGNIMDGKSVEELSSTECHQ, which has no sequence similarity to GCIP, is among the key components necessary for GC1 stimulation. A GCAP1/GCIP chimera consisting of residues 1-43 (including nonfunctional EF1) of GCAP1 and residues 56-206 of GCIP stimulated GC1 at low [Ca2+] and inhibited GC1 at high [Ca2+], suggesting that the essential components required to transform an inhibitory to an activating protein are contained within the N-terminal region of GCAP1 (residues 1-43).     

Tuning of a neuronal calcium sensor

Recoverin is a Ca(2+)-regulated signal transduction modulator expressed in the vertebrate retina that has been implicated in visual adaptation. An intriguing feature of recoverin is a cluster of charged residues at its C terminus, the functional significance of which is largely unclear. To elucidate the impact of this segment on recoverin structure and function, we have investigated a mutant lacking the C-terminal 12 amino acids. Whereas in myristoylated recoverin the truncation causes an overall decrease in Ca(2+) sensitivity, results for the non-myristoylated mutant indicate that the truncation primarily affects the high affinity EF-hand 3. The three-dimensional structure of the mutant has been determined by x-ray crystallography. In addition to significant changes in average coordinates compared with wild-type recoverin, the structure provides strong indication of increased conformational flexibility, particularly in the C-terminal domain. Based on these observations, we propose a novel role of the C-terminal segment of recoverin as an internal modulator of Ca(2+) sensitivity.     

Optimized expression and purification of myristoylated human neuronal calcium sensor 1 in E. coli

We have developed a protocol to produce large quantities of high purity myristoylated and non-myristoylated neuronal calcium sensor 1 (NCS-1) protein. NCS-1 is a member of the neuronal calcium sensor (NCS) family and plays an important role in modulating G-protein signaling and exocytosis pathways in cells. Many of these functions are calcium-dependent and require NCS-1 to be modified with an N-terminal myristoyl moiety. In our system, a C-terminally 6x His-tagged variant of NCS-1 was co-expressed with yeast N-myristoyltransferase (NMT) in ZYP-5052 auto-induction media supplemented with sodium myristate (100-200 microM). With optimized growth conditions and a high capacity metal affinity purification scheme, >50mg of homogenous myristoylated NCS-1 is obtained from 1L of culture in a single step. The properties of the C-terminally tagged NCS-1 variants are indistinguishable from those reported for untagged NCS-1. Using this system, we have also isolated and characterized mutant NCS-1 proteins that have attenuated (NCS-1 E120Q) and abrogated (NCS-1 DeltaEF) ability to bind calcium. The large quantities of NCS-1 proteins isolated from small culture volumes of auto-inducible media will provide the necessary reagents for further biochemical and structural characterization. The affinity tag at the C-terminus of the protein provides a suitable reagent for easily identifying binding partners of the various NCS-1 constructs. Additionally, this method could be used to produce other recombinant proteins of the NCS family, and may be extended to express and isolate myristoylated variants of other proteins.     

Calcium- and myristoyl-dependent properties of guanylate cyclase-activating protein-1 and protein-2

In visual transduction, guanylate cyclase-activating proteins (GCAPs) activate the membrane-bound guanylate cyclase 1 (ROS-GC1) to synthesize cGMP under conditions of low cytoplasmic [Ca2+]free. GCAPs are neuronal Ca2+-binding proteins with three functional EF-hands and a consensus site for N-terminal myristoylation. GCAP-1 and GCAP-2 regulated ROS-GC1 activities differently. The myristoyl group in GCAP-1 had a strong influence on the Ca2+-dependent regulation of ROS-GC1 (shift in IC50). In contrast, myristoylation of GCAP-2 did not change the cyclase activation profile (no shift in IC50). Thus, the myristoyl group controlled the Ca2+-sensitivity of GCAP-1, but not that of GCAP-2. The myristoyl group restricted the accessibility of one cysteine in GCAP-1 and GCAP-2 observed by measuring the time-dependent thiol reactivity of cysteines. This shielding effect was not relieved when Ca2+ was buffered by EGTA. We applied surface plasmon resonance (SPR) spectroscopy to monitor the Ca2+-dependent binding of myristoylated and nonmyristoylated GCAP-1 and GCAP-2 to immobilized phospholipid membranes. None of the GCAPs exhibited a Ca2+-myristoyl switch as observed for recoverin. Thus, the myristoyl group controls the Ca2+-sensitivity of GCAP-1 (not that of GCAP-2) by an allosteric mechanism, but this control step does not involve a myristoyl switch.     

MAGUIN, a novel neuronal membrane-associated guanylate kinase-interacting protein

Postsynaptic density (PSD)-95/Synapse-associated protein (SAP) 90 and synaptic scaffolding molecule (S-SCAM) are neuronal membrane-associated guanylate kinases. Because PSD-95/SAP90 and S-SCAM function as synaptic scaffolding proteins, identification of ligands for these proteins is important to elucidate the structure of synaptic junctions. Here, we report a novel protein interacting with the PDZ domains of PSD-95/SAP90 and S-SCAM and named it MAGUIN-1 (membrane-associated guanylate kinase-interacting protein-1). MAGUIN-1 has one sterile alpha motif, one PDZ, and one plekstrin homology domain. MAGUIN-1 is localized at the plasma membrane via the plekstrin homology domain and the C-terminal region and interacts with PSD-95/SAP90 and S-SCAM via a C-terminal PDZ domain-binding motif. MAGUIN-1 has a short isoform, MAGUIN-2, which lacks a PDZ domain-binding motif. MAGUINs are expressed in neurons and localized in the cell body and neurites and are coimmunoprecipitated with PSD-95/SAP90 and S-SCAM from rat crude synaptosome. MAGUIN-1 may play an important role with PSD-95/SAP90 and S-SCAM to assemble the components of synaptic junctions.     

Molecular tuning of calcium dependent processes by neuronal calcium sensor proteins in the retina

Vertebrate photoreceptor cells are exquisite light detectors operating under very dim and bright illumination mediated by phototransduction, which is under control of the two secondary messengers cGMP and Ca²⁺. Feedback mechanisms enable photoreceptor cells to regain their responsiveness after light stimulation and involve neuronal Ca²⁺-sensor proteins, named GCAPs (guanylate cyclase-activating proteins) and recoverins. This review compares the diversity in Ca²⁺-related signaling mediated by GCAP and recoverin variants that exhibit differences in Ca²⁺-sensing, protein conformational changes, myristoyl switch mechanisms, diversity in divalent cation binding and dimer formation. In summary, both subclasses of neuronal Ca²⁺-sensor proteins contribute to a complex signaling network in rod and cone cells, which is perfectly suited to match the requirements for sensitive cell responses and maintaining this responsiveness in the presence of different background light intensities.     

1H, 13C,and 15N chemical shift assignments of neuronal calcium sensor protein,hippocalcin

Hippocalcin, a member of the neuronal calcium sensor (NCS) subclass of the calmodulin superfamily, serves as an important calcium sensor for the slow afterhyperpolarizing (sAHP) current in the hippocampus, which underlies some forms of learning and memory. Hippocalcin is also a calcium sensor for hippocampal long-term depression (LTD) and genetically linked to neurodegenerative diseases. We report NMR chemical shift assignments of Ca²⁺-free hippocalcin (BMRB no. 18627).     

Conformational changes in guanylyl cyclase-activating protein 1 (GCAP1) and its tryptophan mutants as a function of calcium concentration.

Guanylyl cyclase-activating proteins (GCAPs are 23-kDa Ca2+-binding proteins belonging to the calmodulin superfamily. Ca2+-free GCAPs are responsible for activation of photoreceptor guanylyl cyclase during light adaptation. In this study, we characterized GCAP1 mutants in which three endogenous nonessential Trp residues were replaced by Phe residues, eliminating intrinsic fluorescence. Subsequently, hydrophobic amino acids adjacent to each of the three functional Ca2+-binding loops were replaced by reporter Trp residues. Using fluorescence spectroscopy and biochemical assays, we found that binding of Ca2+ to GCAP1 causes a major conformational change especially in the region around the EF3-hand motif. This transition of GCAP1 from an activator to an inhibitor of GC requires an activation energy Ea = 9.3 kcal/mol. When Tyr99 adjacent to the EF3-hand motif was replaced by Cys, a mutation linked to autosomal dominant cone dystrophy in humans, Cys99 is unable to stabilize the inactive GCAP1-Ca2+ complex. Stopped-flow kinetic measurements indicated that GCAP1 rapidly loses its bound Ca2+ (k-1 = 72 s-1 at 37 degrees C) and was estimated to associate with Ca2+ at a rate (k1 > 2 x 10(8) M-1 s-1) close to the diffusion limit. Thus, GCAP1 displays thermodynamic and kinetic properties that are compatible with its involvement early in the phototransduction response.     

Targeted gene disruption of the neuronal calcium sensor 1 homologue in rice blast fungus,Magnaporthe grisea

We isolated a neuronal calcium sensor 1/frequenin-like gene, Mg-NCS-1, from Magnaporthe grisea and evaluated the phenotypes of null-mutants of the gene. The putative Mg-NCS-1 protein showed high similarity to the other NCS-1 proteins. The null-mutants had normal growth and pathogenicity similar to the parental strain, but their growth was suppressed in high concentrations of Ca2+ or acidic conditions.     

Immunocytochemical localization and crystal structure of human frequenin (neuronal calcium sensor 1)

Frequenin, a member of a large family of myristoyl-switch calcium-binding proteins, functions as a calcium-ion sensor to modulate synaptic activity and secretion. We show that human frequenin colocalizes with ARF1 GTPase in COS-7 cells and occurs in similar cellular compartments as the phosphatidylinositol-4-OH kinase PI4Kbeta, the mammalian homolog of the yeast kinase PIK1. In addition, the crystal structure of unmyristoylated, calcium-bound human frequenin has been determined and refined to 1.9 A resolution. The overall fold of frequenin resembles those of neurocalcin and the photoreceptor, recoverin, of the same family, with two pairs of calcium-binding EF hands and three bound calcium ions. Despite the similarities, however, frequenin displays significant structural differences. A large conformational shift of the C-terminal region creates a wide hydrophobic crevice at the surface of frequenin. This crevice, which is unique to frequenin and distinct from the myristoyl-binding box of recoverin, may accommodate a yet unknown protein ligand.     

Intestinal inflammation modulates expression of the synaptic vesicle protein neuronal calcium sensor-1

The calcium-binding protein neuronal calcium sensor 1 (NCS-1) is involved in modulation of neurotransmitter release in the peripheral and central nervous systems. Since intestinal inflammation impairs neurotransmitter release, we evaluated the expression of NCS-1 in the normal rat colon and in dinitrobenzene sulfonic acid (DNBS)-induced colitis. Immunocytochemistry and Western blots showed high levels of NCS-1 in the myenteric plexus and in axons in the smooth muscle layers; 23 +/- 2% of myenteric neurons were NCS-1 positive, with staining restricted to the largest neurons. NCS-1-positive axons decreased to 13.3 +/- 0.4% of total axons by day 2 and dropped further to 7.0 +/- 0.1% by day 4, returning to control levels by day 16. Dual-label Western blot analysis showed that the expression of NCS-1 relative to PGP 9.5 decreased by 50% on day 4 but returned to control by day 16. The selective loss of NCS-1 during colitis may underlie the altered neural function seen in the inflamed intestine.     

Intracellular neuronal calcium sensor (NCS) protein VILIP-1 modulates cGMP signalling pathways in transfected neural cells and cerebellar granule neurones

The family of intracellular neuronal calcium-sensors (NCS) belongs to the superfamily of EF-hand proteins. Family members have been shown by in vitro assays to regulate signal cascades in retinal photoreceptor cells. To study the functions of NCS proteins not expressed in photoreceptor cells we examined Visinin-like protein-1 (VILIP-1) effects on signalling pathways in living neural cells. Visinin-like protein-1 expression increased cGMP levels in transfected C6 and PC12 cells. Interestingly, in transfected PC12 cells stimulation was dependent on the subcellular localization of VILIP-1. In cells transfected with membrane-associated wild-type VILIP-1 particulate guanylyl cyclase (GC) was stimulated more strongly than soluble GC. In contrast, deletion of the N-terminal myristoylation site resulted in cytosolic localization of VILIP-1 and enhanced stimulation of soluble GC. To study the molecular mechanisms underlying GC stimulation VILIP-1 was examined to see if it can physically interact with GCs. A direct physical interaction of VILIP-1 with the recombinant catalytic domain of particulate GCs-A, B and with native GCs enriched from rat brain was observed in GST pull-down as well as in surface plasmon resonance interaction studies. Furthermore, following trituration of recombinant VILIP-1 protein into cerebellar granule cells the protein influenced only signalling by GC-B. Together with the observed colocalization of GC-B, but not GC-A, with VILIP-1 in cerebellar granule cells, these results suggest that VILIP-1 may be a physiological regulator of GC-B.