Drosophila Calmodulin differs from the mammalian protein at only three residue positions: 99, 143 and 147 (Smith, 1987).

The EF-hand motif is present in dozens of characterized proteins belong to a large family. Many, but not all EF-hand proteins bind Ca++. Identified EF-hand proteins include Calmodulin, Troponin C, the myosin Essential light chain, the myosin Regulatory light chain, nonvertebrate Troponin, Ca-dependent protein kinase, Calpain and Calcineurin.

Calpains are calcium-dependent proteases believed to participate in calcium-regulated signal pathways in cells. Calpains have been found in vertebrates in ubiquitous as well as tissue-specific locations. A highly tissue-specific calpain gene has been characterized in Drosophila: CalpA, at 56C-D on the second chromosome. The encoded protein is found in a few neurons in the central nervous system, in scattered endocrine cells in the midgut, and in blood cells. In the blood cell line mbn-2, calpain is associated with a granular component in the cytoplasm. The expression of this protein is more restricted than that of the corresponding transcripts, which are widely distributed in the central nervous system, digestive tract, and other tissues. The sequence of CalpA is closely related to that of vertebrate calpains, but an additional segment is inserted in the calmodulin-like carboxy-terminal domain. This insert contains a hydrophobic region that may be involved in membrane attachment of the enzyme. Differential splicing also gives rise to a minor transcript that lacks the calmodulin-like domain (Theopold, 1995).

Using low-stringency hybridization and polymerase chain reaction (PCR)-based DNA amplification, four Drosophila melanogaster genes have been isolated: three of these encode troponin-C isoforms, and the fourth specifies a protein closely related to calmodulin. Two of the troponin-C genes, located (respectively) within the 47D and 73F subdivisions of chromosomes 2 and 3 encode very closely related isoforms. The isoform specified by the 47D gene accumulates almost exclusively in larval muscles, while the one encoded by the 73F gene is present in both larvae and adults. The third gene, located within the 41C subdivision of chromosome 2, encodes a more distantly related troponin-C isoform that accumulates only within adults. The fourth gene, which encodes the calmodulin-related protein, is located within the 97A subdivision of chromosome three. This protein has a different primary sequence from that of conventional calmodulin, which is specified by a gene located within the 49A subdivision of chromosome 2 (Fyrberg, 1994).

Pulses of ecdysone at the end of Drosophila larval development that signal the onset of metamorphosis dramatically reprogram gene expression. Ecdysone directly induces several early puffs in the salivary gland polytene chromosomes; these, in turn, activate many late puffs. Three early puffs (at 2B5, 74EF, and 75B) have been studied at the molecular level. Each contains a single ecdysone primary-response gene that encodes a family of widely expressed transcription factors. The 63F early puff differs in significant ways from the previously characterized early puff loci: The 63F puff contains a pair of ecdysone-inducible genes that are transcribed in the larval salivary glands: E63-1 and E63-2. E63-1 induction in late third instar larvae appears to be highly tissue-specific, restricted to the salivary gland. E63-1 encodes a novel Ca(2+)-binding protein related to calmodulin. The discovery of an ecdysone-inducible Ca(2+)-binding protein provides a foundation for integrating steroid hormone and calcium second messenger signaling pathways and generates an additional level for potential regulation of the ecdysone response (Andres, 1995).

A novel EF-hand protein of Dictyostelium discoideum, termed CBP2, is composed of 168 amino acids and contains four consensus sequences that are typical for (Ca2+)-binding EF-hand domains. The protein sequence exhibits only minor similarities to other calmodulin-type proteins from Dictyostelium. The genomic DNA harbors two short introns; their positions suggest that the gene is unrelated to the EF-hand proteins from the calmodulin group. Northern blot analysis shows that the mRNA level is significantly increased during development. Polyclonal antibodies raised against the recombinant protein recognize a protein of about 20 kDa. Like the mRNA, the protein is also more abundant in developing cells. Overlay experiments with 45Ca2+ indicate that the EF-hands in fact have (Ca2+)-binding activity. The recent description of CBP1, another calmodulin-type Dictyostelium protein that is upregulated during development, suggests that D. discoideum contains a family of EF-hand proteins that have specific functions during distinct steps of development (Andre, 1996).

Three engineered mammalian calmodulins (CaMs) were constructed in which the two EF hand pairs were either substituted for one another or exchanged: CaMNN, the C-terminal EF hand pair (residues 82-148) was replaced by a duplication of the N-terminal pair (residues 9-75); CaMCC, the N-terminal pair was replaced by a duplication of the C-terminal pair; CaMCN, the two EF had pairs were exchanged. Skeletal muscle myosin light chain kinase (skMLCK) activity is activated to 75% of the maximum level by CaMCC and to 45% of the maximum level by CaMCN and is not significantly activated by CaMNN; Smooth muscle myosin light chain kinase activity (gMLCK) is fully activated by CaMCN and is not significantly activated by either CaMNN or CaMCC. Cerebellar nitric oxide synthase activity (nNOS) is fully activated by CaMNN and CaMCN and is not significantly activated by CaMCC. These results indicate that the EF hand pairs contain distinct but overlapping sets of determinants for binding and activation of enzymes, with the greater degree of overlap in determinants for binding. While the structural changes associated with swapping the EF hand pairs do not affect activation of nNOS or gMLCK activities, they significantly reduce activation of skMLCK activity, indicating that this process requires specific determinants in CaM outside the EF hand pairs (Persechini, 1996a).

Deletion of residues 2-8 from the N-terminal leader sequence in calmodulin abolishes calmodulin-dependent activation of skeletal muscle myosin light chain kinase activity and reduces calmodulin-dependent activation of smooth muscle myosin light chain kinase activity to approximately 50% of the maximum level measured at a saturating calmodulin concentration. Calmodulin-dependent activation of cerebellar nitric oxide synthase activity is not affected by this deletion. Overlapping tripeptide deletions from the leader sequence indicate that a three amino acidic cluster contains the determinants necessary for activation of myosin light chain kinase activity. Based on enzyme kinetic analyses, deletions in the leader sequence have little or no effect on the apparent affinities of calmodulin for the synthase or the two kinases. Since the N-terminal leader does not appear to play a significant structural role in the complexes between calmodulin and peptides representing the calmodulin-binding domains in the two kinases, these results indicate that it participates in secondary interactions with these enzymes that are important to activation, but not to recognition or binding of calmodulin (Persechini, 1996b).

CaM (4 cTnC) is a calmodulin-cardiac troponin C chimeric protein containing the first, second, and third calcium-binding EF-hands of calmodulin (CaM) and the fourth EF-hand of cardiac troponin C (cTnC). CaM (4 cTnC) shows 2-fold-enhanced carboxy-terminal Ca2+ affinity (relative to CaM) and also exhibits impaired activation of the CaM-regulated enzymes smooth muscle myosin light chain kinase (smMLCK), neuronal nitric oxide synthase (nNOS), and phosphodiesterase (PDE). Additional chimeras were constructed, replacing most of CaM helix 7, Ca2+-binding loop 4, and helix 8 with the corresponding helices and loops of cTnC. Point mutants in the fourth EF-hand of CaM were also constructed. Replacement of CaM's fourth loop with the corresponding loop of cTnC enhances Ca2+ affinity by over 3-fold through an increase in the Ca2+ on rate and also reduces the cooperativity of Ca2+ binding. In contrast, substitution of CaM helix 7 or 8 modestly decreases Ca2+ affinity by increasing the Ca2+ off rate, without impairment of cooperativity. All three of the helix and loop chimeras fully activate PDE, with minor shifts in Kact. CaM (helix 7 cTnC) shows a significantly impaired ability to activate smMLCK and nNOS, whereas the other two chimeras retain about 80% of the maximal smMLCK and nNOS activation observed with CaM (George, 1996).

Many important enzyme activities are regulated by Ca2+-dependent interactions with calmodulin (CaM). Some of the most important targets for CaM action are in the nucleus, and Ca2+-dependent CaM translocation into this organelle has been reported. Hormone-evoked cytosolic Ca2+ signals occur physiologically as oscillations, but, so far, oscillations in CaM concentration have not been described. Fluorescent-labeled CaM was loaded into pancreatic acinar cells and the fluorescence monitored in various regions by confocal microscopy. Sustained high concentrations of the hormone cholecystokinin or the neurotransmitter acetylcholine evoke a transient movement of cytosolic CaM from the basal nonnuclear area into the secretory granule region and, thereafter, a more substantial and prolonged translocation of CaM into the nucleoplasm. About 50% of the CaM that binds Ca2+ translocates into the nucleus. At a lower hormone concentration, evoking Ca2+ oscillations, regular spikes of increased CaM concentration are seen in the secretory granule region with mirror image spikes of decreased CaM concentration in the basal nonnuclear region. The nucleus is able to integrate the Ca2+ spike-evoked pulses of CaM translocation into a sustained elevation of the nucleoplasmic concentration of this protein (Craske, 1999).

Many targets of calcium signaling pathways are activated or inhibited by binding the Ca2+-liganded form of calmodulin (Ca²⁺-CaM). A test was performed of the hypothesis that local Ca²⁺-CaM-regulated signaling processes can be selectively activated by local intracellular differences in free Ca²⁺-CaM concentration.

Energy-transfer confocal microscopy of a fluorescent biosensor was used to measure the difference in the concentration of free Ca²⁺-CaM between nucleus and cytoplasm. Strikingly, short receptor-induced calcium spikes produce transient increases in free Ca²⁺-CaM concentration that are of markedly higher amplitude in the cytosol than in the nucleus. In contrast, prolonged increases in calcium leads to equalization of the nuclear and cytosolic free Ca²⁺-CaM concentrations over a period of minutes. Photobleaching recovery and translocation measurements with fluorescently labeled CaM shows that equalization is likely to be the result of a diffusion-mediated net translocation of CaM into the nucleus. The driving force for equalization is a higher Ca²⁺-CaM-buffering capacity in the nucleus compared with the cytosol, as the direction of the free Ca²⁺-CaM concentration gradient and of CaM translocation can be reversed by expressing a Ca²⁺-CaM-binding protein at high concentration in the cytosol. It is concluded that subcellular differences in the distribution of Ca²⁺-CaM-binding proteins can produce gradients of free Ca²⁺-CaM concentration that result in a net translocation of CaM. This provides a mechanism for dynamically regulating local free Ca²⁺-CaM concentrations, and thus the local activity of Ca²⁺-CaM targets. Free Ca²⁺-CaM signals in the nucleus remain low during brief or low-frequency calcium spikes, whereas high-frequency spikes or persistent increases in calcium cause translocation of CaM from the cytoplasm to the nucleus, resulting in similar concentrations of nuclear and cytosolic free Ca²⁺-CaM. From a regulatory perspective, this suggests that cells may control the amplitude of nuclear Ca²⁺-CaM signals by increasing or decreasing the concentration and composition of CaM-binding proteins in different cellular regions (Teruel, 2000).

The spindle pole body (SPB) is the microtubule organizing center of Saccharomyces cerevisiae. Its core includes the proteins Spc42, Spc110 (kendrin/pericentrin ortholog), calmodulin (Cmd1), Spc29, and Cnm67. Each was tagged with CFP and YFP and their proximity to one another was determined by fluorescence resonance energy transfer (FRET). FRET was measured by a new metric that accurately reflected the relative extent of energy transfer. The FRET values established the topology of the core proteins within the architecture of SPB. The N-termini of Spc42 and Spc29, and the C-termini of all the core proteins face the gap between the IL2 layer and the central plaque. Spc110 traverses the central plaque and Cnm67 spans the IL2 layer. Spc42 is a central component of the central plaque where its N-terminus is closely associated with the C-termini of Spc29, Cmd1, and Spc110. When the donor-acceptor pairs were ordered into five broad categories of increasing FRET, the ranking of the pairs specified a unique geometry for the positions of the core proteins, as shown by a mathematical proof. The geometry was integrated with prior cryoelectron tomography to create a model of the interwoven network of proteins within the central plaque. One prediction of the model, the dimerization of the calmodulin-binding domains of Spc110, was confirmed by in vitro analysis (Muller, 2005).

The spindle pole body is the microtubule organizing center of Saccharomyces cerevisiae (Jaspersen, 2004). Two SPBs establish the bipolar mitotic spindle, a defining event of mitosis that allows the stable transmission of equivalent genetic material to the mother and daughter cell at the time of cell division. This role of the SPB is carried out by the centrosome in higher eukaryotes (Muller, 2005).

The structure of the SPB is reviewed by Jaspersen (2004). Briefly the ultrastructure observed by electron tomography consists of a series of stacked layers embedded in the nuclear envelope. The inner plaque is the area where the microtubules dock to the SPB; this plaque harbors the gamma-tubulin complex and the N-terminus of Spc110. The central plaque and the IL2 layer are the two core layers. This core is composed of 5 proteins. Spc29 and Cmd1 reside in the central plaque. Spc42 is thought to begin within the central plaque, but terminate in the IL2 layer. The C-terminus of Spc110 is in the central plaque where it binds Cmd1. The C-terminus of Cnm67 lies in the IL2 layer where it binds Spc42 and links the SPB core to the outer plaque. The outer plaque is the cytoplasmic boundary of the SPB where the astral microtubules nucleate from a second region of gamma-tubulin. Based on primarily two-hybrid interactions the SPB core proteins are typically depicted as components lying along a linear path that proceeds from Spc110 to Spc29 to Spc42 to Cnm67 (Muller, 2005).

The ultrastructure of the SPB is clearly quite different from the centrosome. Centrioles are not present and the SPB remains inserted in the nuclear envelope during mitosis. Yet both have in common the gamma-tubulin complex, Spc110/kendrin/AKAP-450, calmodulin, centrin, and Sfi1p. (The latter two proteins are part of the SPB half-bridge, a domain involved in SPB duplication. Despite differences in gross anatomy, the SPB and centrosome likely share an underlying structure. To date the only component of either the SPB or centrosome whose structure is solved at atomic resolution is calmodulin. The paucity of structural information has limited the understanding of the molecular functions performed by individual SPB proteins. Without crystals or well behaved soluble proteins, the available research tools to probe the SPB structure or any large macromolecular complex are few (Muller, 2005).

This study used a hybrid approach that combined in vivo live-cell FRET measurements with previous cryo-EM analysis. CFP and YFP were used as FRET donor and acceptor and attached to the components of the SPB. Initially FRET values were classified as either positive or negative for energy transfer as judged by a comparison to carefully designed controls. This binary classification system allowed mapping of the ends of proteins within the architecture of the SPB. Next the positive values were subdivided into classes. The classification specified a unique geometry for the SPB components that was not only consistent with previous structural and genetic studies, but broadened the understanding of SPB organization (Muller, 2005).

The FRET results suggest that the Il2 layer and central plaque form an integrated meshwork of proteins with Spc42 closely associated with all components of the central plaque. The general features of the core proteins of the IL2 and central plaque, based on the current results and the general literature, are as follows. The N-terminus of Spc42 begins at the inner boundary of the central plaque, forms a coiled-coil domain that defines the spacing of the gap between layers, enters the IL2 layer, and finally loops back to end at the internal face of the IL2 layer. Remarkably, even though the N-terminal domain before the coiled coil is only ~60 amino acids long, the N-terminus is in close proximity to the C-termini of Spc29, Cmd1, and Spc110. Cnm67 begins at the outer plaque, penetrates the IL2 and ends in close proximity to the C-terminus of Spc42. The N- and C-termini of Spc29 both lie on the inner face of the central plaque. Cmd1 is situated near the C-terminal end of Spc110, consistent with in vitro binding experiments, genetic and two-hybrid results. Finally Spc110, which at its N-terminus binds the gamma-tubulin complex extends from the inner plaque through the central plaque and ends in close juxtaposition to the C-terminus of Spc42. All the termini of the central plaque and IL2 layer proteins lie along the internal edges of the IL2 and central plaque layers, facing the space between the two layers (Muller, 2005).

The SPB is organized around a hexagonal lattice of Spc42. The arrangement of Spc42 in the Il2 layer was suggested by analysis of cryoelectron micrographs of both SPB cores and two-dimensional crystals of Spc42 that arise in vivo upon Spc42 overexpression. Because the N-terminus of Spc42 is situated in the central plaque, the arrangement of Spc42 in the IL2 layer necessarily imposes the same organization on the location of Spc42 in the central plaque. Cryoelectron microscopy has not revealed this implied organization of the central plaque. However the visualization of the Spc42 arrangement in the IL2 relied upon the contrast between regions of high protein density and pockets of low or no density. If, as supported by the FRET results, the components of the central plaque are densely packed, a uniform and high protein density would mask the organization in electron micrographs (Muller, 2005).

The Spc42 lattice provided a template that enabled the FRET-based geometry of the core proteins to be taken and and a model to be generated for the organization of the central plaque. The model suggests that Spc42 and Spc29 form the heart of the central plaque. A strong association between Spc29 and Spc42 is well documented. Spc29 has a robust two-hybrid interaction with the N-terminus of Spc42. In an Spc110-226 mutant, Spc29 remains associated with Spc42 under the conditions in which Spc110-226, calmodulin, and the gamma-tubulin complex pull away from the SPB. Finally, Spc29 is seen with Spc42 at the satellite of the SPB. In this model Spc29 lies along the path of Spc42 and together they form a ring of protein around the center of the hexagonal unit in the central plaque (Muller, 2005).

At the center of the hexagonal unit is placed a trimer of Spc110 dimers as they unravel from their coiled coil motif. In the model Spc110 enters the central plaque through the ring of Spc29 and Spc42. Two-hybrid analysis suggested that Spc29 binds to Spc110 between the end of the coiled coil and the start of the Cmd1-binding domain, from positions 811 to 898. This region overlaps Region II of Spc110 (position 772-836), a domain that plays a role in locking Spc110 in place during mitosis. The FRET model is consistent with Spc29 and Spc42 acting as a clasp to surround and lock Spc110 in place. However the central plaque must not only lock Spc110 in place to withstand the push and pull of mitosis, but also must be organized in a way that facilitates the remodeling of the SPB during G1/S-phase when 50% of Spc110 turns over. Therefore any locking mechanism must be reversible and the interaction between Spc110 and Spc29 must be dynamic (Muller, 2005).

Calmodulin and the C-terminal domain of Spc110 are positioned to reinforce lateral stability of the central plaque. This is evident when the hexagonal unit is tessellated to form a mosaic lattice of the central plaque components. Calmodulin and the C-terminal domain of Spc110 from one hexagonal unit are juxtaposed with their counterparts in the adjoining hexagonal units. The dimerization of the C-terminal Spc110/Cmd1 domain was confirmed in vitro. Surprisingly even though calmodulin is a highly conserved component of the SPB, it is not required. An SPC110-407 mutant of S. cerevisiae that lacks the calmodulin-binding domain is still viable. One explanation is that the integrity of the SPB is maintained through structurally redundant lateral connections in IL2 layer and central plaque (Muller, 2005).

The tessellation of the repeat unit prompts the question of what determines the lateral limits of the SPB. How is the repeat symmetry broken and the boundary with the nuclear envelope established? One clue may come from a comparison of the dimensions of the SPB with the cluster of nuclear microtubules that originate at the SPB. The SPB is circular with an average diameter of ~165 nm for the central plaque from a diploid and therefore an area of ~2.1 x 10⁶Ų. A diploid would have ~35 microtubules emanating from the SPB (32 kinetochore microtubules and a three pole-to-pole microtubules. Microtubules have a cross-sectional diameter of 25 nm, so the minimal total area occupied by 35 microtubules (hexagonal packing with a packing density of 91% is 1.9 x 10⁶Ų. Even assuming some spread at the inner plaque, the SPB has almost the minimal area required to attach the nuclear microtubules. One mechanism that could minimize both the size of the SPB and the size of the bundle of microtubules would be feedback between microtubule attachment and Spc110 turnover. A removal of Spc110 molecules that are not nucleating microtubules would break the lattice symmetry, leaving Spc42 and Spc29 to interact with other proteins of the nuclear envelope. Spc110 is only added to the SPB after the insertion of Spc42 and Spc29 into the nuclear envelope, so the edge of the SPB does not require Spc110. The mechanism and role of Spc110 turnover is an area of continued research (Muller, 2005).