Investigating the landscape of intracellular [Ca2+] in live cells by rapid photoactivated cross-linking of calmodulin-protein interactions.

The Ca2+ sensor protein calmodulin interacts in a Ca2+-dependent manner with a large number of proteins that among them encompass a diverse assortment of functions and subcellular localizations. A method for monitoring calmodulin-protein interactions as they occur throughout a living cell would thus uniquely enable investigations of the intracellular landscape of [Ca2+] and its relationship to cell function. We have developed such a method based on capture of calmodulin-protein interactions by rapid photoactivated cross-linking (t1/2 ∼7s) in cells stably expressing a tandem affinity tagged calmodulin that have been metabolically labeled with a photoreactive methionine analog. Tagged adducts are stringently enriched, and captured calmodulin interactors are then identified and quantified based on tandem mass spectrometry data for their tryptic peptides. In this paper we show that the capture behaviors of interactors in cells are consistent with the presence of basal microdomains of elevated [Ca2+]. Ca2+ sensitivities for capture were determined, and these suggest that [Ca2+] levels are above ∼1 µM in these regions. Although the microdomains appear to affect capture of most proteins, capture of some is at an apparent Ca2+-dependent maximum, suggesting they are targeted to the domains. Removal of extracellular Ca2+ has both immediate (5 min) and delayed (30 min) effects on capture, implying that the microdomains are supported by a combination of Ca2+ influx across the cell membrane and Ca2+ derived from internal stores. The known properties of the presumptive microdomain targeted proteins suggestroles in a variety of Ca2+-dependent basal metabolism and in formation and maintenance of the domains.

Evaluating Calmodulin-Protein Interactions by Rapid Photoactivated Cross-Linking in Live Cells Metabolically Labeled with Photo-Methionine.

This work addresses the question of how the Ca2+ sensor protein calmodulin shapes cellular responses to Ca2+ signals. Proteins interacting with affinity tagged calmodulin were captured by rapid (t1/2 ≈ 7 s) photoactivated cross-linking under basal conditions, after brief removal of extracellular Ca2+ and during a cytosolic [Ca2+] transient in cells metabolically labeled with a photoreactive methionine analog. Tagged adducts were stringently enriched, and captured proteins were identified and quantified by LC-MS/MS. A set of 489 proteins including 27 known calmodulin interactors was derived. A threshold for fractional capture was applied to define a high specificity group of 170 proteins, including 22 known interactors, and a low specificity group of 319 proteins. Capture of ∼60% of the high specificity group was affected by manipulations of Ca2+, compared with ∼20% of the low specificity group. This suggests that the former is likely to contain novel interactors of physiological significance. The capture of 29 proteins, nearly all high specificity, was decreased by the removal of extracellular Ca2+, although this does not affect cytosolic [Ca2+]. Capture of half of these was unaffected by the cytosolic [Ca2+] transient, consistent with high local [Ca2+]. These proteins are hypothesized to reside in or near microdomains of high [Ca2+] supported by the Ca2+ influx.

Fluorescence quenching studies of structure and dynamics in calmodulin-eNOS complexes.

Activation of endothelial nitric oxide synthase (eNOS) by calmodulin (CaM) facilitates formation of a sequence of conformational states that is not well understood. Fluorescence decays of fluorescently labeled CaM bound to eNOS reveal four distinct conformational states and single-molecule fluorescence trajectories show multiple fluorescence states with transitions between states occurring on time scales of milliseconds to seconds. A model is proposed relating fluorescence quenching states to enzyme conformations. Specifically, we propose that the most highly quenched state corresponds to CaM docked to an oxygenase domain of the enzyme. In single-molecule trajectories, this state occurs with time lags consistent with the oxygenase activity of the enzyme.

Calmodulin-induced structural changes in endothelial nitric oxide synthase.

We have derived structures of intact calmodulin (CaM)-free and CaM-bound endothelial nitric oxide synthase (eNOS) by reconstruction from cryo-electron micrographs. The CaM-free reconstruction is well fitted by the oxygenase domain dimer, but the reductase domains are not visible, suggesting they are mobile and thus delocalized. Additional protein is visible in the CaM-bound reconstruction, concentrated in volumes near two basic patches on each oxygenase domain. One of these corresponds with a presumptive docking site for the reductase domain FMN-binding module. The other is proposed to correspond with a docking site for CaM. A model is suggested in which CaM binding and docking position the reductase domains near the oxygenase domains and promote docking of the FMN-binding modules required for electron transfer.

In calmodulin-IQ domain complexes, the Ca(2+)-free and Ca(2+)-bound forms of the calmodulin C-lobe direct the N-lobe to different binding sites.

We have investigated the roles played by the calmodulin (CaM) N- and C-lobes in establishing the conformations of CaM-IQ domain complexes in different Ca(2+)-free and Ca(2+)-bound states. Our results indicate a dominant role for the C-lobe in these complexes. When the C-lobe is Ca(2+)-free, it directs the N-lobe to a binding site within the IQ domain consensus sequence. It appears that the N-lobe must be Ca(2+)-free to interact productively with this site. When the C-lobe is Ca(2+)-bound, it directs the N-lobe to a site upstream of the consensus sequence, and it appears that the N-lobe must be Ca(2+)-bound to interact productively with this site. A model for switching in CaM-IQ domain complexes is presented in which the N-lobe adopts bound and extended positions that depend on the status of the Ca(2+)-binding sites in each CaM lobe and the compositions of the two N-lobe binding sites. Ca(2+)-dependent changes in the conformation of the bound C-lobe that appear to be responsible for directed N-lobe binding are also identified. Changes in the equilibria between extended and bound N-lobe positions may control bridging interactions in which the extended N-lobe is bound to another CaM-binding domain. Ca(2+)-dependent control of bridging interactions with CaM has been implicated in the regulation of ion channel and unconventional myosin activities.

Variations at the semiconserved glycine in the IQ domain consensus sequence have a major impact on Ca2+-dependent switching in calmodulin-IQ domain complexes.

We have replaced the semiconserved Gly in the IQ domain consensus sequence with Ala, Arg, or Met in a reference sequence and determined how this affects its complexes with calmodulin. The K(d) for the Ca(2+)-free reference complex is 2.4 +/- 0.3 microM. The Ala and Arg replacements increase this to 5.4 +/- 0.4 and 6.2 +/- 0.5 microM, while the Met increases it to 26.4 +/- 2.5 microM. When Ca(2+) is bound to both calmodulin lobes, the K(d) for the reference complex is not significantly affected, but the K(d) for the Ala variant decreases to 0.9 +/- 0.04 microM, and the values for the Arg and Met variants decrease to 0.4 +/- 0.03 microM. Using mutant calmodulins, we defined the effect of Ca(2+) binding to each lobe, with the C-terminal preceding the N-terminal (C-->N) or vice versa (N-->C). In the C-->N order the first step increases the reference K(d) approximately 5-fold, while it decreases the values for the variants approximately 2- to approximately 10-fold. The second step decreases the K(d) values for the all of the complexes approximately 5-fold, suggesting that the N-terminal lobe does not interact with the semiconserved position after the first step. In the N-->C order the first step increases the K(d) values for the reference complex and Met and Ala variants approximately 15- to approximately 200-fold but does not affect the value for the Arg variant. The second step decreases the K(d) values for the reference and Arg variant approximately 10- and approximately 15-fold and the Ala and Met variants approximately 2000-fold. Thus, both steps in the N-->C order are sensitive to variations at the semiconserved position, while only the first is in the C-->N order. Due to energy coupling, this order is followed under equilibrium conditions.

The IQ domains in neuromodulin and PEP19 represent two major functional classes.

The affinities of Ca(2+)-saturated and Ca(2+)-free calmodulin for a fluorescent reporter construct containing the PEP19 IQ domain differ by a factor of approximately 100, with K(d) values of 11.0 +/- 1.2 and 1128.4 +/- 176.5 muM, respectively, while the affinities of a reporter containing the neuromodulin IQ domain are essentially identical, with K(d) values of 2.9 +/- 0.3 and 2.4 +/- 0.3 muM, respectively. When Ca(2+) is bound only to the C-terminal pair of Ca(2+)-binding sites in calmodulin, the K(d) value for the PEP19 reporter complex is decreased approximately 5-fold, while the value for the neuromodulin reporter complex is increased by the same factor. When Ca(2+) is bound only to the N-terminal pair of Ca(2+)-binding sites, the K(d) value for the PEP19 reporter complex is unaffected, but the value for the complex with the neuromodulin reporter is increased approximately 12-fold. These functional differences are largely ascribed to three differences in the CaM-binding sequences of the two reporters. Replacement of a central Gly in the neuromodulin IQ domain with a Lys at this position in PEP19 almost entirely accounts for the distinctive patterns of Ca(2+)-dependent stability changes exhibited by the two complexes. Replacement of a Lys immediately before the "IQ" amino acid pair in the neuromodulin sequence with the Ala in PEP19 accounts for the remaining Ca(2+)-dependent differences. Replacement of an Ala in the N-terminal half of the neuromodulin sequence with the Gln in PEP19 accounts for approximately half of the Ca(2+)-independent difference in the stabilities of the two reporter complexes, with the Ca(2+)-independent effect of the Lys replacement accounting for most of the remainder. Since the central Gly in the neuromodulin sequence is conserved in half of all known IQ domains, these results suggest that the presence or absence of this residue defines two major functional classes.

Effects of combined phosphorylation at Ser-617 and Ser-1179 in endothelial nitric-oxide synthase on EC50(Ca2+) values for calmodulin binding and enzyme activation.

We have investigated the possible biochemical basis for enhancements in NO production in endothelial cells that have been correlated with agonist- or shear stress-evoked phosphorylation at Ser-1179. We have found that a phosphomimetic substitution at Ser-1179 doubles maximal synthase activity, partially disinhibits cytochrome c reductase activity, and lowers the EC(50)(Ca(2+)) values for calmodulin binding and enzyme activation from the control values of 182 +/- 2 and 422 +/- 22 nm to 116 +/- 2 and 300 +/- 10 nm. These are similar to the effects of a phosphomimetic substitution at Ser-617 (Tran, Q. K., Leonard, J., Black, D. J., and Persechini, A. (2008) Biochemistry 47, 7557-7566). Although combining substitutions at Ser-617 and Ser-1179 has no additional effect on maximal synthase activity, cooperativity between the two substitutions completely disinhibits reductase activity and further reduces the EC(50)(Ca(2+)) values for calmodulin binding and enzyme activation to 77 +/- 2 and 130 +/- 5 nm. We have confirmed that specific Akt-catalyzed phosphorylation of Ser-617 and Ser-1179 and phosphomimetic substitutions at these positions have similar functional effects. Changes in the biochemical properties of eNOS produced by combined phosphorylation at Ser-617 and Ser-1179 are predicted to substantially increase synthase activity in cells at a typical basal free Ca(2+) concentration of 50-100 nm.

Phosphorylation within an autoinhibitory domain in endothelial nitric oxide synthase reduces the Ca(2+) concentrations required for calmodulin to bind and activate the enzyme.

We have investigated the effects of phosphorylation at Ser-617 and Ser-635 within an autoinhibitory domain (residues 595-639) in bovine endothelial nitric oxide synthase on enzyme activity and the Ca (2+) dependencies for calmodulin binding and enzyme activation. A phosphomimetic S617D substitution doubles the maximum calmodulin-dependent enzyme activity and decreases the EC 50(Ca (2+)) values for calmodulin binding and enzyme activation from the wild-type values of 180 +/- 2 and 397 +/- 23 nM to values of 109 +/- 2 and 258 +/- 11 nM, respectively. Deletion of the autoinhibitory domain also doubles the maximum calmodulin-dependent enzyme activity and decreases the EC 50(Ca (2+)) values for calmodulin binding and calmodulin-dependent enzyme activation to 65 +/- 4 and 118 +/- 4 nM, respectively. An S635D substitution has little or no effect on enzyme activity or EC 50(Ca (2+)) values, either alone or when combined with the S617D substitution. These results suggest that phosphorylation at Ser-617 partially reverses suppression by the autoinhibitory domain. Associated effects on the EC 50(Ca (2+)) values and maximum calmodulin-dependent enzyme activity are predicted to contribute equally to phosphorylation-dependent enhancement of NO production during a typical agonist-evoked Ca (2+) transient, while the reduction in EC 50(Ca (2+)) values is predicted to be the major contributor to enhancement at resting free Ca (2+) concentrations.

The kinetics of Ca(2+)-dependent switching in a calmodulin-IQ domain complex.

We have performed a kinetic analysis of Ca2+-dependent switching in the complex between calmodulin (CaM) and the IQ domain from neuromodulin, and have developed detailed kinetic models for this process. Our results indicate that the affinity of the C-ter Ca2+-binding sites in bound CaM is reduced due to a approximately 10-fold decrease in the Ca2+ association rate, while the affinity of the N-ter Ca2+-binding sites is increased due to a approximately 3-fold decrease in the Ca2+ dissociation rate. Although the Ca2+-free and Ca2+-saturated forms of the CaM-IQ domain complex have identical affinities, CaM dissociates approximately 100 times faster in the presence of Ca2+. Furthermore, under these conditions CaM can be transferred to the CaM-binding domain from CaM kinase II via a ternary complex. These properties are consistent with the hypothesis that CaM bound to neuromodulin comprises a localized store that can be efficiently delivered to neuronal proteins in its Ca2+-bound form in response to a Ca2+ signal.

Dynamic changes in free Ca-calmodulin levels in adult cardiac myocytes.

Despite its multifunctional role in cardiac myocyte function, little is known about dynamic changes in activation state of calmodulin (CaM). Thus, the purpose of this study was to develop a tool to measure Ca bound CaM (Ca-CaM) levels in intact cardiac myocytes. For dynamic measurements of Ca-CaM, we generated an adenoviral vector which expresses a cyan and a yellow fluorescent protein linked by a modified version of the Ca-CaM binding domain of avian smooth muscle myosin light chain kinase. Adult rabbit cardiac myocytes were infected with the Ca-CaM sensing probe or simultaneously infected with viruses containing CaM and the Ca-CaM sensing probe for 24-48 h. Myocytes were then field stimulated (1 Hz) and excited at 440 nm with emitted fluorescence measured at 485 and 535 nm. Changes in [Ca-CaM] are expressed as the ratio of 485 nm/535 nm. Small beat-to-beat changes of [Ca-CaM] were detected, but only when CaM was co-expressed with the sensor. However, upon beta-adrenergic stimulation with isoproterenol, there was an increase in the amplitude of the signals during each beat (parallel to the shortening, which is an indirect measure of [Ca]i) and also a rise in the diastolic [Ca-CaM]. Total [CaM] measured by both competitive ELISA and semi-quantitative Western blots was 5-6 microM in isolated adult ventricular myocytes. Our results indicate that there are dynamic changes in free Ca-CaM levels (a phasic component tracking [Ca]i) as well as system memory that integrates the [Ca]i signals (a tonic component).

Biphasic Ca2+-dependent switching in a calmodulin-IQ domain complex.

The relationship between the free Ca2+ concentration and the apparent dissociation constant for the complex between calmodulin (CaM) and the neuromodulin IQ domain consists of two phases. In the first phase, Ca2+ bound to the C-ter EF hand pair in CaM increases the Kd for the complex from the Ca2+-free value of 2.3 +/- 0.1 microM to a value of 14.4 +/- 1.3 microM. In the second phase, Ca2+ bound to the N-ter EF hand pair reduces the Kd for the complex to a value of 2.5 +/- 0.1 microM, reversing the effect of the first phase. Due to energy coupling effects associated with these phases, the mean dissociation constant for binding of Ca2+ to the C-ter EF hand pair is increased approximately 3-fold, from 1.8 +/- 0.1 to 5.1 +/- 0.7 microM, and the mean dissociation constant for binding of Ca2+ to the N-ter EF hand pair is decreased by the same factor, from 11.2 +/- 1.0 to 3.5 +/- 0.6 microM. These characteristics produce a bell-shaped relationship between the apparent dissociation constant for the complex and the free Ca2+ concentration, with a distance of 5-6 microM between the midpoints of the rising and falling phases. Release of CaM from the neuromodulin IQ domain therefore appears to be promoted over a relatively narrow range of free Ca2+ concentrations. Our results demonstrate that CaM-IQ domain complexes can function as biphasic Ca2+ switches through opposing effects of Ca2+ bound sequentially to the two EF hand pairs in CaM.

Dominant affectors in the calmodulin network shape the time courses of target responses in the cell.

In endothelial cells nitric oxide synthase is a dominant affector in the calmodulin network by virtue of its ability to bind a significant fraction of limiting intracellular calmodulin. We have investigated how this affector function influences the kinetics of calmodulin-dependent signaling in cells co-expressing the synthase and a fluorescent calmodulin target analog similar in its interactions with calmodulin to myosin light chain kinase. The synthase binds (Ca(2+))(4)-calmodulin with a K(d) value of approximately 0.2 nM and an association rate constant of approximately 1.5 x 10(5) M(-1) s(-1). These values are, respectively, 10- and 100-fold smaller than the corresponding values for the analog. Thus, when Ca(2+) is added to a mixture of calmodulin, target analog and synthase in vitro a large fluorescence transient with a relaxation time of approximately 600 s is observed as (Ca(2+))(4)-calmodulin is rapidly bound to the analog and then slowly captured by the higher affinity synthase. A rapid increase in the free Ca(2+) concentration elicits similar transient analog responses in cells expressing the cytoplasmic target analog and either a wild-type membrane bound or mutant cytoplasmic synthase. Transient responses are not observed in cells co-expressing the fluorescent analog and a mutant T497D synthase unable to bind calmodulin. These results demonstrate that dominant affectors in the calmodulin network shape both the magnitudes and time courses of target responses in the cell.

Real-time evaluation of myosin light chain kinase activation in smooth muscle tissues from a transgenic calmodulin-biosensor mouse.

Ca(2+)/calmodulin (CaM)-dependent phosphorylation of myosin regulatory light chain (RLC) by myosin light chain kinase (MLCK) initiates smooth muscle contraction and regulates actomyosin-based cytoskeletal functions in nonmuscle cells. The net extent of RLC phosphorylation is controlled by MLCK activity relative to myosin light chain phosphatase activity. We have constructed a CaM-sensor MLCK where Ca(2+)-dependent CaM binding increases the catalytic activity of the kinase domain, whereas coincident binding to the biosensor domain decreases fluorescence resonance energy transfer between two fluorescent proteins. We have created transgenic mice expressing this construct specifically in smooth muscle cells to perform real-time evaluations of the relationship between smooth muscle contractility and MLCK activation in intact tissues and organs. Measurements in intact bladder smooth muscle demonstrate that MLCK activation increases rapidly during KCl-induced contractions but is not maximal, consistent with a limiting amount of cellular CaM. Carbachol treatment produces the same amount of force development and RLC phosphorylation, with much smaller increases in [Ca(2+)](i) and MLCK activation. A Rho kinase inhibitor suppresses RLC phosphorylation and force but not MLCK activation in carbachol-treated tissues. These observations are consistent with a model in which the magnitude of an agonist-mediated smooth muscle contraction depends on a rapid but limited Ca(2+)/CaM-dependent activation of MLCK and Rho kinase-mediated inhibition of myosin light chain phosphatase activity. These studies demonstrate the feasibility of producing transgenic biosensor mice for investigations of signaling processes in intact systems.

Monitoring the total available calmodulin concentration in intact cells over the physiological range in free Ca2+.

We describe the design, characterization and application of a new genetically encoded fluorescent biosensor for intracellular detection of both free Ca(2+)-calmodulin and apocalmodulin, which together comprise the available calmodulin concentration. The biosensor binds both forms of calmodulin with an apparent Kd value of 3 microM, and has kinetic properties making it suitable for monitoring dynamic changes on a subsecond time scale. It can be used in conjunction with the fluorescent Ca(2+)-indicator, indo-1, allowing the available calmodulin and free Ca2+ concentrations to be monitored concurrently. We have determined an intracellular available calmodulin concentration of 8.8 +/- 2.2 microM under resting conditions in a human kidney cell line stably expressing the biosensor. Elevation of the intracellular free Ca2+ concentration by agonist, store-operated Ca(2+)-entry or ionophore results in Ca(2+)-dependent consumption of the available calmodulin. A plot of normalized values for the available calmodulin concentration versus the free Ca2+ concentration fits a consumption curve with a cooperativity coefficient of 1.8 and a [Ca2+]50 of 850 nM. There is no detectible binding of calmodulin to the biosensor above a free Ca2+ concentration of approximately 4 microM, consistent with an available calmodulin concentration < or = 200 nM under these conditions, and an overall excess of calmodulin-binding sites.

Quantitative measurements of Ca(2+)/calmodulin binding and activation of myosin light chain kinase in cells.

Myosin II regulatory light chain (RLC) phosphorylation by Ca(2+)/calmodulin (CaM)-dependent myosin light chain kinase (MLCK) is implicated in many cellular actin cytoskeletal functions. We examined MLCK activation quantitatively with a fluorescent biosensor MLCK where Ca(2+)-dependent increases in kinase activity were coincident with decreases in fluorescence resonance energy transfer (FRET) in vitro. In cells stably transfected with CaM sensor MLCK, increasing [Ca(2+)](i) increased MLCK activation and RLC phosphorylation coincidently. There was no evidence for CaM binding but not activating MLCK at low [Ca(2+)](i). At saturating [Ca(2+)](i) MLCK was not fully activated probably due to limited availability of cellular Ca(2+)/CaM.

Intracellular coupling via limiting calmodulin.

Measurements of cellular Ca2+-calmodulin concentrations have suggested that competition for limiting calmodulin may couple calmodulin-dependent activities. Here we have directly tested this hypothesis. We have found that in endothelial cells the amount of calmodulin bound to nitric-oxide synthase and the catalytic activity of the enzyme both are increased approximately 3-fold upon changes in the phosphorylation status of the enzyme. Quantitative immunoblotting indicates that the synthase can bind up to 25% of the total cellular calmodulin. Consistent with this, simultaneous determinations of the free Ca2+ and Ca2+-calmodulin concentrations in these cells performed using indo-1 and a fluorescent calmodulin biosensor (Kd = 2 nm) indicate that increased binding of calmodulin to the synthase is associated with substantial reductions in the Ca2+-calmodulin concentrations produced and an increase in the [Ca2+]50 for formation of the calmodulin-biosensor complex. The physiological significance of these effects is confirmed by a corresponding 40% reduction in calmodulin-dependent plasma membrane Ca2+ pump activity. An identical reduction in pump activity is produced by expression of a high affinity (Kd = 0.3 nm) calmodulin biosensor, and treatment to increase calmodulin binding to the synthase then has no further effect. This suggests that the observed reduction in pump activity is due specifically to reduced calmodulin availability. Increases in synthase activity thus appear to be coupled to decreases in the activities of other calmodulin targets through reductions in the size of a limiting pool of available calmodulin. This exemplifies what is likely to be a ubiquitous mechanism for coupling among diverse calmodulin-dependent activities.

Lobe-dependent regulation of ryanodine receptor type 1 by calmodulin.

Calmodulin activates the skeletal muscle Ca(2+) release channel RYR1 at nm Ca(2+) concentrations and inhibits the channel at microm Ca(2+) concentrations. Using a deletion mutant of calmodulin, we demonstrate that amino acids 2-8 are required for high affinity binding of calmodulin to RYR1 at both nm and microm Ca(2+) concentrations and are required for maximum inhibition of the channel at microm Ca(2+) concentrations. In contrast, the addition of three amino acids to the N terminus of calmodulin increased the affinity for RYR1 at both nm and microm Ca(2+) concentrations, but destroyed its functional effects on RYR1 at nm Ca(2+). Using both full-length RYR1 and synthetic peptides, we demonstrate that the calmodulin-binding site on RYR1 is likely to be noncontiguous, with the C-terminal lobe of both apocalmodulin and Ca(2+)-calmodulin binding to amino acids between positions 3614 and 3643 and the N-terminal lobe binding at sites that are not proximal in the primary sequence. Ca(2+) binding to the C-terminal lobe of calmodulin converted it from an activator to an inhibitor, but an interaction with the N-terminal lobe was required for a maximum effect on RYR1. This interaction apparently depends on the native sequence or structure of the first few amino acids at the N terminus of calmodulin.

Calmodulin is a limiting factor in the cell.

The total intracellular concentration of calmodulin (CaM) in the cell appears to be significantly below the total concentration of its targets, making it a limiting factor in their regulation. In this review we discuss the ways in which this is likely to impact signaling. A key conclusion is that competition for a limiting pool of CaM enables cross-talk between CaM-dependent signaling pathways.