Defining protein-protein interactions using site-directed spin-labeling: the binding of protein kinase C substrates to calmodulin

DEER experiments reveal fundamental differences between calmodulin complexes with IQ and MARCKS peptides in solution

Calmodulin (CaM) is a calcium-binding protein that regulates the function of many proteins by indirectly conferring Ca²⁺ sensitivity, and it undergoes a large conformational change on partners' binding. We compared the solution binding mode of the target peptides MARCKS and IQ by double electron-electron resonance (DEER) distance measurements and paramagnetic NMR. We combined nitroxide and Gd(III) spin labels, including specific substitution of one of the Ca²⁺ ions in the CaM mutant N60D by a Gd(III) ion. The binding of MARCKS to holo-CaM resulted neither in a closed conformation nor in a unique relative orientation between the two CaM domains, in contrast with the crystal structure. Binding of IQ to holo-CaM did generate a closed conformation. Using elastic network modeling and 12 distance restraints obtained from multiple holo-CaM/IQ DEER data, we derived a model of the solution structure, which is in reasonable agreement with the crystal structure.

Electron paramagnetic resonance spectroscopy of nitroxide-labeled calmodulin

Calmodulin (CaM) is a highly conserved calcium-binding protein consisting of two homologous domains, each of which contains two EF-hands, that is known to bind well over 300 proteins and peptides. In most cases the (Ca(2+))(4-)form of CaM leads to the activation of a key regulatory enzyme or protein in a myriad of biological processes. Using the nitroxide spin-labeling reagent, 3-(2-iodoacetamido)-2,2,5,5-tetramethyl-1-pyrrolidinyl oxyl, bovine brain CaM was modified at 2-3 methionines with retention of activity as judged by the activation of cyclic nucleotide phosphodiesterase. X-band electron paramagnetic resonance (EPR) spectroscopy was used to measure the spectral changes upon addition of Ca(2+) to the apo-form of spin-labeled protein. A significant loss of spectral intensity, arising primarily from reductions in the heights of the low, intermediate, and high field peaks, accompanied Ca(2+) binding. The midpoint of the Ca(2+)-mediated transition determined by EPR occurred at a higher Ca(2+) concentration than that measured with circular dichroic spectroscopy and enzyme activation. Recent data have indicated that the transition from the apo-state of CaM to the fully saturated form, [(Ca(2+))(4-)CaM], contains a compact intermediate corresponding to [(Ca(2+))(2-)CaM], and the present results suggest that the spin probes are reporting on Ca(2+) binding to the last two sites in the N-terminal domain, i.e. for the [(Ca(2+))(2)-CaM] → [(Ca(2+))(4-)CaM] transition in which the compact structure becomes more extended. EPR of CaM, spin-labeled at methionines, offers a different approach for studying Ca(2+)-mediated conformational changes and may emerge as a useful technique for monitoring interactions with target proteins.

The SNARE motif of synaptobrevin exhibits an aqueous-interfacial partitioning that is modulated by membrane curvature

The structure and interfacial association of the full-length vesicle SNARE, synaptobrevin, were compared in four different lipid environments using nuclear magnetic resonance and electron paramagnetic resonance spectroscopy. In micelles, segments of the SNARE motif are helical and associated with the interface. However, the fraction of helix and interfacial association decreases as synaptobrevin is moved from micelle to bicelle to bilayer environments, indicating that the tendency toward interfacial association is sensitive to membrane curvature. In bilayers, the SNARE motif of synaptobrevin transiently associates with the lipid interface, and regions that are helical in micelles are in conformational and environmental exchange in bicelles and bilayers. This work demonstrates that the SNARE motif of synaptobrevin has a significant propensity to form a helix and exchange with the membrane interface prior to SNARE assembly. This transient interfacial association and its sensitivity to membrane curvature are likely to play a role in SNARE recognition events that regulate membrane fusion.

Synaptotagmin 1 and SNAREs form a complex that is structurally heterogeneous

Synaptotagmin 1 (syt1) functions as a Ca(2+)-sensor for neuronal exocytosis. Here, site-directed spin labeling was used to examine the complex formed between a soluble fragment of syt1, which contains its two C2 domains, and the neuronal core soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) complex. Changes in electron paramagnetic resonance lineshape and accessibility for spin-labeled syt1 mutants indicate that in solution, the assembled core SNARE complex contacts syt1 in several regions. For the C2B domain, contact occurs in the polybasic face and sites opposite the Ca(2+)-binding loops. For the C2A domain, contact is seen with the SNARE complex in a region near loop 2. Double electron-electron resonance was used to estimate distances between the two C2 domains of syt1. These distances have broad distributions in solution, which do not significantly change when syt1 is fully associated with the core SNARE complex. The broad distance distributions indicate that syt1 is structurally heterogeneous when bound to the SNAREs and does not assume a well-defined structure. Simulated annealing using electron paramagnetic resonance-derived distance restraints produces a family of syt1 structures where the Ca(2+)-binding regions of each domain face in roughly opposite directions. The results suggest that when associated with the SNAREs, syt1 is configured to bind opposing bilayers, but that the syt1/SNARE complex samples multiple conformational states.

Differential interaction of spin-labeled arrestin with inactive and active phosphorhodopsin

Arrestins regulate signaling and trafficking of G protein-coupled receptors by virtue of their preferential binding to the phosphorylated active form of the receptor. To identify sites in arrestin involved in receptor interaction, a nitroxide-containing side chain was introduced at each of 28 different positions in visual arrestin, and the dynamics of the side chain was used to monitor arrestin interaction with phosphorylated forms of its cognate receptor, rhodopsin. At physiological concentrations, visual arrestin associates with both inactive dark phosphorylated rhodopsin (P-Rh) and light-activated phosphorylated rhodopsin (P-Rh*). Residues distributed over the concave surfaces of the two arrestin domains are involved in weak interactions with both states of phosphorhodopsin, and the flexible C-terminal sequence (C-tail) of arrestin becomes dynamically disordered in both complexes. A large-scale movement of the C-tail is demonstrated by direct distance measurements using a doubly labeled arrestin with one nitroxide in the C-tail and the other in the N-domain. Despite some overlap, the molecular "footprint" of arrestin bound to P-Rh and P-Rh* is different, showing the structure of the complexes to be unique. Strong immobilizing interactions with residues in a highly flexible loop between beta-strands V and VI are only observed in complex with the activated state. This result identifies this loop as a key recognition site in the arrestin-P-Rh* complex and supports the view that flexible sequences are key elements in protein-protein interactions.

Structural dynamics of the actin-myosin interface by site-directed spectroscopy

We have used site-directed spin and fluorescence labeling to test molecular models of the actin-myosin interface. Force is generated when the actin-myosin complex undergoes a transition from a disordered weak-binding state to an ordered strong-binding state. Actomyosin interface models, in which residues are classified as contributing to either weak or strong binding, have been derived by fitting individual crystallographic structures of actin and myosin into actomyosin cryo-EM maps. Our goal is to test these models using site-directed spectroscopic probes on actin and myosin. Starting with Cys-lite constructs of both yeast actin (ActC) and the Dictyostelium myosin II motor domain (S1dC), site-directed labeling (SDL) mutants were generated by mutating residues to Cys in the proposed weak and strong-binding interfaces. This report focuses on the effects of forming the strong-binding complex on four SDL mutants, two located in the proposed weak-binding interface (ActC5 and S1dC619) and two located in the proposed strong-binding interface (ActC345 and S1dC401). Neither the mutations nor labeling prevented strong actomyosin binding or actin-activation of myosin ATPase. Formation of the strong-binding complex resulted in decreased spin and fluorescence probe mobility at all sites, but both myosin-bound probes showed remarkably high mobility even after complex formation. Complex formation decreased solvent accessibility for both actin-bound probes, but increased it for the myosin-bound probes. These results are not consistent with a simple model in which there are discrete weak and strong interfaces, with only the strong interface forming under strong-binding conditions, nor are they consistent with a model in which surface residues become rigid and inaccessible upon complex formation. We conclude that all four of these residues are involved in the strong actin-myosin interface, but this interface is remarkably dynamic, especially on the surface of myosin.

Site-directed spin labeling reveals a conformational switch in the phosphorylation domain of smooth muscle myosin

We have used site-directed spin labeling and EPR spectroscopy to detect structural changes within the regulatory light chain (RLC) of smooth muscle myosin upon phosphorylation. Smooth muscle contraction is activated by phosphorylation of S19 on RLC, but the structural basis of this process is unknown. There is no crystal structure containing a phosphorylated RLC, and there is no crystal structure for the N-terminal region of any RLC. Therefore, we have prepared single-Cys mutations throughout RLC, exchanged each mutant onto smooth muscle heavy meromyosin, verified normal regulatory function, and used EPR to determine dynamics and solvent accessibility at each site. A survey of spin-label sites throughout the RLC revealed that only the N-terminal region (first 24 aa) shows a significant change in dynamics upon phosphorylation, with most of the first 17 residues showing an increase in rotational amplitude. Therefore, we focused on this N-terminal region. Additional structural information was obtained from the pattern of oxygen accessibility along the sequence. In the absence of phosphorylation, little or no periodicity was observed, suggesting a lack of secondary structural order in this region. However, phosphorylation induced a strong helical pattern (3.6-residue periodicity) in the first 17 residues, while increasing accessibility throughout the first 24 residues. We have identified a domain within RLC, the N-terminal phosphorylation domain, in which phosphorylation increases helical order, internal dynamics, and accessibility. These results support a model in which this disorder-to-order transition within the phosphorylation domain results in decreased head-head interactions, activating myosin in smooth muscle.

Electron paramagnetic resonance spectroscopy and molecular modelling of the interaction of myelin basic protein (MBP) with calmodulin (CaM)—diversity and conformational adaptability of MBP CaM-targets

The classic 18.5 kDa isoform of murine myelin basic protein (mMBP) has been shown to bind calmodulin (CaM) strongly and specifically in vitro. Here, we have used site-directed spin labelling (SDSL) and electron paramagnetic resonance (EPR) spectroscopy to map more precisely the sites of interaction of recombinant mMBP (rmMBP) with CaM. On the basis of these and previous experimental data, and the predictions of CaM-binding motifs using the Calmodulin Target Database (), three main segments of MBP were suggested for the interaction. The first site is located at the C-terminus; the second one lies in the central portion of the protein and forms an amphipathic alpha-helix in reconstituted myelin-mimetic systems; the third is quite close to the N-terminus. The murine Golli-MBP isoform J37 has also been shown to bind CaM in vitro, and an interaction site was predicted in the N-terminal Golli-specific portion of the protein. From these four segments, we selected peptide fragments of 12-14 residues in length, chosen on the bases of their amphipathicity and CaM-target characteristics. We modelled each of these peptides as alpha-helices, and performed docking simulations to investigate their interactions with the CaM peptide-binding tunnel. Different yet almost equally favourable CaM-binding modes were found for each of them. The experimental SDSL/EPR and theoretical modelling results were in good agreement, and supported the conjecture that there are several plausible CaM-binding sites in MBP, that could be induced into an alpha-helical conformation by their interaction with CaM and account for strong immobilisation of spin-labeled residues in all three segments. Phosphorylation and deimination were also emulated and simulated for known sites of MBP post-translational modification. The results obtained confirmed the appropriate utilisation of simple residue substitutions to mimic the natural modifications, and demonstrated molecular mechanisms by which MBP-CaM interactions could be modulated in vivo.

Membrane-anchoring and charge effects in the interaction of myelin basic protein with lipid bilayers studied by site-directed spin labeling

Myelin basic protein (MBP) maintains the compaction of the myelin sheath in the central nervous system by anchoring the cytoplasmic face of the two apposing bilayers and may also play a role in signal transduction. Site-directed spin labeling was done at eight matching sites in each of two recombinant murine MBPs, qC1 (charge +19) and qC8 charge (+13), which, respectively, emulate the native form of the protein (C1) and a post-translationally modified form (C8) that is increased in multiple sclerosis. When interacting with large unilamellar vesicles, most spin-labeled sites in qC8 were more mobile than those in qC1. Depth measurement via continuous wave power saturation indicated that the N-terminal and C-terminal sites in qC1 were located below the plane of the phospholipid headgroups. In qC8, the C-terminal domain dissociated from the membrane, suggesting a means by which the exposure of natural C8 to cytosolic enzymes and ligands might increase in vivo in multiple sclerosis. The importance of two Phe-Phe pairs in MBP to its interactions with lipids was investigated by separately mutating each pair to Ala-Ala. The mobility at F42A/F43A and especially F86A/F87A increased significantly. Depth measurements and helical wheel analysis indicated that the Phe-86/Phe-87 region could form a surface-seeking amphipathic alpha-helix.

Crystal structure of a MARCKS peptide containing the calmodulin-binding domain in complex with Ca2+-calmodulin

The calmodulin-binding domain of myristoylated alanine-rich C kinase substrate (MARCKS), which interacts with various targets including calmodulin, actin and membrane lipids, has been suggested to function as a crosstalk point among several signal transduction pathways. We present here the crystal structure at 2 A resolution of a peptide consisting of the MARCKS calmodulin (CaM)-binding domain in complex with Ca2+-CaM. The domain assumes a flexible conformation, and the hydrophobic pocket of the calmodulin N-lobe, which is a common CaM-binding site observed in previously resolved Ca2+-CaM-target peptide complexes, is not involved in the interaction. The present structure presents a novel target-recognition mode of calmodulin and provides insight into the structural basis of the flexible interaction module of MARCKS.

Cross-talk unfolded: MARCKS proteins

The proteins of the MARCKS (myristoylated alanine-rich C kinase substrate) family were first identified as prominent substrates of protein kinase C (PKC). Since then, these proteins have been implicated in the regulation of brain development and postnatal survival, cellular migration and adhesion, as well as endo-, exo- and phago-cytosis, and neurosecretion. The effector domain of MARCKS proteins is phosphorylated by PKC, binds to calmodulin and contributes to membrane binding. This multitude of mutually exclusive interactions allows cross-talk between the signal transduction pathways involving PKC and calmodulin. This review focuses on recent, mostly biophysical and biochemical results renewing interest in this protein family. MARCKS membrane binding is now understood at the molecular level. From a structural point of view, there is a consensus emerging that MARCKS proteins are "natively unfolded". Interestingly, domains similar to the effector domain have been discovered in other proteins. Furthermore, since the effector domain enhances the polymerization of actin in vitro, MARCKS proteins have been proposed to mediate regulation of the actin cytoskeleton. However, the recent observations that MARCKS might serve to sequester phosphatidylinositol 4,5-bisphosphate in the plasma membrane of unstimulated cells suggest an alternative model for the control of the actin cytoskeleton. While myristoylation is classically considered to be a co-translational, irreversible event, new reports on MARCKS proteins suggest a more dynamic picture of this protein modification. Finally, studies with mice lacking MARCKS proteins have investigated the functions of these proteins during embryonic development in the intact organism.

Differences in calmodulin and calmodulin-binding proteins in phasic and tonic smooth muscles

To determine whether densities of calmodulin (CaM) and CaM-binding proteins are related to phasic and tonic behavior of smooth muscles, we quantified these proteins in the opossum esophageal body (EB) and lower esophageal sphincter (LES), which represent phasic and tonic smooth muscles, respectively. Gel electrophoresis, immunoprecipitation, Western blot, and hemagglutinin epitope-tagged CaM (HA-CaM) overlay assay with quantitative scanning densitometry and phosphorylation measurements were used. Total protein content in the two smooth muscles was similar (approximately 30 mg protein/g frozen tissue). Total tissue concentration of CaM was significantly (25%) higher in EB than in LES (P < 0.05). HA-CaM-binding proteins were qualitatively similar in LES and EB extracts. Myosin, myristoylated alanine-rich C kinase substrate protein, Ca(2+)/CaM kinase II, and calponin contents were also similar in the two muscles. However, content and total activity of myosin light chain kinase (MLCK) and content of caldesmon (CaD) were three- to fourfold higher in EB than in LES. Increased CaM and MLCK content may allow for a wide range of contractile force varying from complete relaxation in the basal state to a large-amplitude, high-velocity contraction in EB phasic muscle. Increased content of CaD, which provides a braking mechanism on contraction, may further contribute to the phasic contractile behavior. In contrast, low CaM, MLCK, and CaD content may be responsible for a small range of contractile force seen in tonic muscle of LES.

Phosphorylation-dependent conformational changes induce a switch in the actin-binding function of MARCKS

Phosphorylation of myristoylated alanine-rich protein kinase C substrate (MARCKS) by protein kinase C eliminates actin filament cross-linking activity, but residual filament binding activity docks phosphorylated MARCKS on filamentous actin. The postulated actin-binding region of MARCKS, which includes a Ca(2+)-calmodulin-binding site, has been portrayed with alpha-helical structure, analogous to other calmodulin-binding domains. Previous speculation suggested that MARCKS may dimerize to form the two functional actin-binding sites requisite for cross-linking activity. Contrary to these hypotheses, we show that MARCKS peptide with actin-cross-linking activity has an extended structure in aqueous solution but assumes a more compact structure upon phosphorylation. We hypothesize that structural changes in the MARCKS peptide induced by phosphorylation create a dynamic structure that, on average, has only one actin-binding site. Moreover, independent of the state of phosphorylation, this peptide is monomeric rather than dimeric, implying that two distinct actin-binding sites are responsible for the actin-cross-linking activity of unphosphorylated MARCKS. These studies uniquely elucidate the mechanism by which phosphorylation of MARCKS induces structural changes and suggest how these structural changes determine biological activity.

Influence of lipid on the structure and phosphorylation of protein kinase C alpha substrate peptides

The structure and phosphorylation of two protein kinase C (PKC) alpha substrate peptides were investigated in varying lipid systems using enzyme activity assays and circular dichroism (CD) spectroscopy. The alpha-peptide, which exhibits the typical PKC alpha substrate motif and is based on the pseudosubstrate region of PKCalpha, was phosphorylated to a similar extent in bovine brain phosphatidylserine vesicles or diheptanoylphosphatidylcholine (PC7) micelles (both with 5 mol % 1,2-dioleoyl-sn-glycerol), whereas neuromodulin (NM)-peptide, which does not exhibit this motif by virtue of its primary structure, was phosphorylated to a much lesser extent in the PC7 micellar system. CD spectra of the peptides indicated that NM-peptide underwent a dramatic structural change in the presence of dimyristoylphosphatidylserine (DMPS) vesicles, whereas spectra acquired in PC7 micelles were similar to those acquired in buffer alone. No significant structural change was observed in the alpha-peptide in the presence of either lipid. PKC activity assays conducted with a series of NM-peptides successively substituted with nitroxide spin labels at each residue position suggested that several residues distal to the phosphorylation site are necessary for substrate recognition. The effect of these substitutions is not consistent with the binding of the NM-peptide to PKC in an extended structure, but is consistent with the binding of this peptide in a helical conformation. Furthermore, the docking of a helical NM-peptide to the substrate binding site of PKC suggests that the interaction is energetically feasible. These results suggest that PKC may recognize some non-linear substrate motifs and that lipid binding may convert a protein into a better PKC substrate.

MARCKS, a major protein kinase C substrate, assumes non-helical conformations both in solution and in complex with Ca2+-calmodulin

MARCKS, a major cellular substrate for protein kinase C, plays important roles in various cellular functions and its functions are regulated by calmodulin. We have studied the conformational properties of recombinant human MARCKS in solution and in complex with calmodulin. Circular dichroism (CD) spectra showed a high content of random coil in physiological solution. When MARCKS or MARCKS-derived calmodulin-binding peptide was complexed with Ca2+-calmodulin, little change was observed in the CD spectra, suggesting that MARCKS binds with calmodulin in a non-helical conformation, which is unique among the calmodulin-binding proteins.