Selective detection of the rotational dynamics of the protein-associated lipid hydrocarbon chains in sarcoplasmic reticulum membranes

Molecular modeling of lipid probes and their influence on the membrane

In this review a number of Molecular Dynamics simulation studies are discussed which focus on the understanding of the behavior of lipid probes in biomembranes. Experiments often use specialized probe moieties or molecules to report on the behavior of a membrane and try to gain information on the membrane as a whole from the probe lipids as these probes are the only things an experiment sees. Probes can be used to make NMR, EPR and fluorescence accessible to the membrane and use fluorescent or spin-active moieties for this purpose. Clearly membranes with and without probes are not identical which makes it worthwhile to elucidate the differences between them with detailed atomistic simulations. In almost all cases these differences are confined to the local neighborhood of the probe molecules which are sparsely used and generally present as single molecules. In general, the behavior of the bulk membrane lipids can be qualitatively understood from the probes but in most cases their properties cannot be directly quantitatively deduced from the probe behavior. This article is part of a Special Issue entitled: Biosimulations edited by Ilpo Vattulainen and Tomasz Róg.

The time-dependent distribution of phosphorylated intermediates in native sarcoplasmic reticulum Ca2+-ATPase from skeletal muscle is not compatible with a linear kinetic model

Quenched-flow mixing was used to characterize the kinetic behavior of the intermediate reactions of the skeletal muscle sarcoplasmic reticulum (SR) Ca-ATPase (SERCA1) at 2 and 21 degrees C. At 2 degrees C, phosphorylation of SR Ca-ATPase with 100 microM ATP labeled one-half of the catalytic sites with a biphasic time dependence [Mahaney, J. E., Froehlich, J. P., and Thomas, D. D. (1995) Biochemistry 34, 4864-4879]. Chasing the phosphoenzyme (EP) with 1.66 mM ADP 10 ms after the start of phosphorylation revealed mostly ADP-insensitive E2P (95% of EP(total)), consistent with its rapid formation from ADP-sensitive E1P. The consecutive relationship of the phosphorylated intermediates predicts a decrease in the proportion of E1P ([E1P]/[EP(total)]) with increasing phosphorylation time. Instead, after 10 ms the proportion of E1P increased and that of E2P decreased until they reached a constant 1:1 stoichiometry ([E1P]:[E2P] approximately 1). At 21 degrees C, phosphorylation displayed a transient overshoot associated with an inorganic phosphate (P(i)) burst, reflecting increased turnover of E2P at the higher temperature. The P(i) burst exceeded the decay of the EP overshoot, suggesting that rephosphorylation of the enzyme occurs before the recycling step (E2 --> E1). This behavior and the reversed order of accumulation of phosphorylated intermediates at 2 degrees C are not compatible with the conventional linear consecutive reaction mechanism: E1 + ATP --> E1.ATP --> E1P + ADP --> E2P --> E2.P(i) --> E1 + P(i). Solubilization of the Ca-ATPase into monomers using the nonionic detergent C(12)E(8) gave a pattern of phosphorylation in which E1P and E2P behave like consecutive intermediates. Kinetic modeling of the C(12)E(8)-solubilized SR Ca-ATPase showed that it behaves according to the conventional Ca-ATPase reaction mechanism, consistent with monomeric catalytic function. We conclude that the nonconforming features of native SERCA1 arise from oligomeric protein conformational interactions that constrain the subunits to a staggered or out-of-phase mode of operation.

The amide class herbicide 3,4-dichloropropionanilide (DCPA) alters the mobility of hydrocarbon chains in T-lymphocyte but not macrophage membranes

Previous studies in our laboratory have demonstrated that the lipophilic herbicide 3,4-dicholoropropionanilide (DCPA) adversely affects cytokine production by activated macrophages and T lymphocytes. The purpose of this study was to test the hypothesis that DCPA alters the mobility of plasma membrane lipid hydrocarbon chains, which interferes with normal T-lymphocyte activation and macrophage function. Electron spin reasonance (ESR) spectroscopy of stearic acid spin labels incorporated into each cell type was used to test the effects of DCPA on lipid hydrocarbon chain mobility in the absence and presence of specific agents that activate each cell type. The results indicated that DCPA treatment had no significant effect on hydrocarbon chain mobility in either cell type per se. However, for T lymphocytes, but not macrophages, DCPA treatment increased a small population of lipid molecules that exhibited reduced hydrocarbon chain mobility near the bilayer hydrocarbon core following cell stimulation. In contrast, there were no significant effects of DCPA on hydrocarbon chain mobility near the head group region of the bilayer for either cell type. The identity of this subpopulation of lipids and its motional properties could not be elucidated from these studies. Nevertheless, data show that DCPA alters the distribution of lipids in distinct motional environments in the membrane of activated T lymphocytes.

Molecular dynamics simulation of site-directed spin labeling: experimental validation in muscle fibers

We have developed a computational molecular dynamics technique to simulate the motions of spin labels bound to the regulatory domain of scallop myosin. These calculations were then directly compared with site-directed spin labeling experimental results obtained by preparing seven single-cysteine mutants of the smooth muscle (chicken gizzard) myosin regulatory light chain and performing electron paramagnetic resonance experiments on these spin-labeled regulatory light chains in functional scallop muscle fibers. We determined molecular dynamics simulation conditions necessary for obtaining a convergent orientational trajectory of the spin label, and from these trajectories we then calculated correlation times, orientational distributions, and order parameters. Simulated order parameters closely match those determined experimentally, validating our molecular dynamics modeling technique, and demonstrating our ability to predict preferred sites for labeling by computer simulation. In several cases, more than one rotational mode was observed within the 14-ns trajectory, suggesting that the spin label samples several local energy minima. This study uses molecular dynamics simulations of an experimental system to explore and enhance the site-directed spin labeling technique.

Phospholamban remains associated with the Ca2+- and Mg2+-dependent ATPase following phosphorylation by cAMP-dependent protein kinase

We have used fluorescence and spin-label EPR spectroscopy to investigate how the phosphorylation of phospholamban (PLB) by cAMP-dependent protein kinase (PKA) modifies structural interactions between PLB and the Ca(2+)- and Mg(2+)-dependent ATPase (Ca-ATPase) that result in enzyme activation. Following covalent modification of N-terminal residues of PLB with dansyl chloride or the spin label 4-isothiocyanato-2,2,6,6-tetramethylpiperidine-N-oxyl ('ITC-TEMPO'), we have co-reconstituted PLB with affinity-purified Ca-ATPase isolated from skeletal sarcoplasmic reticulum (SR) with full retention of catalytic function. The Ca(2+)-dependence of the ATPase activity of this reconstituted preparation is virtually identical with that observed using native cardiac SR before and after PLB phosphorylation, indicating that co-reconstituted sarcoplasmic/endoplasmic-reticulum Ca(2+)-ATPase 1 (SERCA1) and PLB provide an equivalent experimental model for SERCA2a-PLB interactions. Phosphorylation of PLB in the absence of the Ca-ATPase results in a greater amplitude of rotational mobility, suggesting that the structural linkage between the transmembrane region and the N-terminus is destabilized. However, whereas co-reconstitution with the Ca-ATPase restricts the amplitude of rotational motion of PLB, subsequent phosphorylation of PLB does not significantly alter its rotational dynamics. Thus structural interactions between PLB and the Ca-ATPase that restrict the rotational mobility of the N-terminus of PLB are retained following the phosphorylation of PLB by PKA. On the other hand, the fluorescence intensity decay of bound dansyl is sensitive to the phosphorylation state of PLB, indicating that there are changes in the tertiary structure of PLB coincident with enzyme activation. These results suggest that PLB phosphorylation alters its structural interactions with the Ca-ATPase by inducing structural rearrangements between PLB and the Ca-ATPase within a defined complex that modulates Ca(2+)-transport function.

Microsecond motions of the lipids associated with trypsinized Na,K-ATPase membranes. Progressive saturation spin-label electron spin resonance studies

The microsecond motions of spin-labeled lipids associated with the Na(+)/K(+)-transporting ATP hydrolase (Na,K-ATPase) in native and tryptically shaved membranes from Squalus acanthias have been studied by progressive saturation electron spin resonance (ESR). This includes both the segmental mobility of the lipid chains and the exchange dynamics of the lipids interacting directly with the protein. The lipids at the protein interface display a temperature-dependent chain mobility on the submicrosecond time scale. Exchange of these lipids with those in the bulk bilayer regions of the membrane takes place on the time scale of the nitroxide spin-lattice relaxation, i.e., in the microsecond regime. The off-rates for exchange directly reflect the specificity of ionized fatty acids relative to protonated fatty acids for interaction with the Na,K-ATPase. These essential features of the lipid dynamics at the intramembranous protein surface, namely, a temperature-dependent exchange on the microsecond time scale that reflects the lipid selectivity, are preserved on removing the extramembranous parts of the Na,K-ATPase by extensive trypsinization.

Lysophosphatidylcholine modulates catalytically important motions of the Ca-ATPase phosphorylation domain

Catalytically important motions of the Ca-ATPase, modulated by the physical properties of surrounding membrane phospholipids, have been suggested to be rate-limiting under physiological conditions. To identify the nature of the structural coupling between the Ca-ATPase and membrane phospholipids, we have investigated the functional and structural effects resulting from the incorporation of the lysophospholipid 1-myristoyl-2-hydroxy-sn-glycerol-3-phosphocholine (LPC) into native sarcoplasmic reticulum (SR) membranes. Nonsolubilizing concentrations of LPC abolish changes in fluorescence signals associated with either intrinsic or extrinsic chromophores that monitor normal conformational transitions accompanying calcium activation of the Ca-ATPase. There are corresponding decreases in the rates of calcium transport coupled to ATP hydrolysis, suggesting that LPC may increase conformational barriers associated with catalytic function. Fluorescence anisotropy measurements of the lipid analogue 1-(4-trimethylammoniumphenyl)-6-phenyl-1,3,5-hexatriene (TMA-DPH) partitioned into SR membranes indicate that LPC does not significantly modify lipid acyl chain rotational dynamics, suggesting differences in headgroup conformation between LPC and diacylglycerol phosphatidylcholines. Complementary measurements using phosphorescence anisotropy of erythrosin isothiocyanate at Lys464 on the Ca-ATPase provide a measure of the dynamic structure of the phosphorylation domain, and indicate that LPC restricts the amplitude of rotational motion. These results suggest a structural linkage between the cytosolic phosphorylation domain and the conformation of membrane phospholipid headgroups. Thus, changes in membrane phospholipid composition can modulate membrane surface properties and affect catalytically important motions of the Ca-ATPase in a manner that suggests a role for LPC generated during signal transduction.

Enhanced rotational dynamics of the phosphorylation domain of the Ca-ATPase upon calcium activation

We have used labeling conditions that permit the specific and covalent attachment of erythrosin isothiocyanate (Er-ITC) to Lys464 within the phosphorylation domain of the Ca-ATPase in skeletal sarcoplasmic reticulum membranes. These labeling conditions do not interfere with high-affinity ATP binding, phosphoenzyme formation, or phosphoenzyme hydrolysis [Huang, S., Negash, S., and Squier, T. C. (1998) Biochemistry 37, 6949-6957]. Thus, we can use frequency-domain phosphorescence spectroscopy to measure the rotational dynamics of the Ca-ATPase stabilized in different enzymatic states corresponding to the absence of bound ligands (E), calcium activation (E x Ca2), the presence of bound nucleotide (E x ATP), and formation of phosphoenzyme (E-P). We resolve three rotational correlation times corresponding to (i) a large-amplitude domain motion of the phosphorylation domain (phi1 approximately 5 +/- 1 micros), (ii) overall protein rotational motion with respect to the membrane normal (phi2 approximately 50 +/- 10 micros), and (iii) the rotational motion of the SR vesicles (phi3 approximately 1.1 +/- 0.4 ms). No differences are observed in the rotational dynamics of E, E x ATP, or E-P, indicating that phosphoenzyme formation or nucleotide binding result in no global structural changes involving the phosphorylation domain. In contrast, calcium activation enhances the amplitude of motion of the phosphorylation domain. These observed calcium-dependent changes in rotational dynamics result from structural changes within a single Ca-ATPase polypeptide chain, since protein-protein interactions do not change upon calcium binding. Thus, calcium binding induces concerted domain motions within a single Ca-ATPase polypeptide chain that may play a critical role in facilitating substrate binding and utilization.

Protein-induced vertical lipid dislocation in a model membrane system: spin-label relaxation studies on avidin-biotinylphosphatidylethanolamine interactions

The change in vertical location of spin-labeled N-biotinyl phosphatidylethanolamine in fluid-phase dimyristoyl phosphatidylcholine bilayer membranes, on binding avidin to the biotinyl headgroup, has been investigated by progressive saturation electron spin resonance measurements. Spin-labeled phospholipids were present at a concentration of 1 mol%, relative to total membrane lipids. For avidin-bound N-biotinyl phosphatidylethanolamine spin-labeled on the 8 C atom of the sn-2 chain, the relaxation enhancement induced by 30 mM Ni2+ ions confined to the aqueous phase was 2.5 times that induced by saturating molecular oxygen, which is preferentially concentrated in the hydrophobic core of the membrane. For phosphatidylcholine also spin-labeled at the 8 position of the sn-2 chain, this ratio was reversed: the relaxation enhancement by Ni2+ ions was half that induced by molecular oxygen. In the absence of avidin, the enhancement by either relaxant was the same for both spin-labeled phospholipids. For a double-labeled system, in which both N-biotinyl phosphatidylethanolamine and phosphatidylcholine were spin-labeled on the 12 C atom of the sn-2 chain, the relaxation rate in the absence of avidin was greater than that predicted from linear additivity of the corresponding singly labeled systems, because of mutual spin-spin interactions between the two labeled lipid species. On binding of avidin to the N-biotinyl phosphatidylethanolamine, this relaxation enhancement by mutual spin-spin interaction was very much decreased. These results indicate that, on binding of avidin to the lipid headgroup, N-biotinyl phosphatidylethanolamine is lifted vertically within the membrane, relative to the phosphatidylcholine host lipids. The specific binding of avidin to N-biotinyl phosphatidylethanolamine parallels the liftase activity proposed for activator proteins associated with the action of certain gangliosidases.

Cysteine reactivity and oligomeric structures of phospholamban and its mutants

To test models for the pentameric structure of phospholamban (PLB) and study its structure and molecular dynamics in SDS solution, we characterized recombinant PLB and several of its mutants by (a) reactivity of cysteine residues toward DTNB [5, 5'-dithiobis(2-nitrobenzoic acid)] and a thiol-reactive spin label, (b) oligomeric state on SDS-PAGE, and (c) EPR of the spin-labeled proteins. WT-PLB has three cysteine residues (36, 41, and 46), all located in the hydrophobic C-terminal transmembrane region. In SDS at pH 7.5, exhaustive reaction with either sulfhydryl reagent resulted in essentially 2 mol of cysteine reacted/mol of WT-PLB, with only slight destabilization of the native pentameric structure. When WT-PLB was denatured in guanidine at pH 8.1, all three cysteines reacted, disrupting the pentamer, which was restored upon cleavage of the disulfide bonds with DTT. In the tetrameric mutant C41L-PLB, the two remaining cysteine residues reacted, reversibly destabilizing the tetramer. In the monomeric mutant L37A-PLB, all three cysteines reacted. The pentameric double cysteine replacement mutant C36,46A-PLB showed negligible reactivity. We conclude that Cys-41 is the unreactive cysteine in PLB and is located at a crucial site for the maintenance of the pentameric structure. EPR spectra in SDS of spin-labeled WT-PLB and mutants correlate with the oligomeric state on SDS-PAGE; oligomeric proteins show decreased spin-label mobility compared with monomers. Molecular dynamics calculations were used to construct an atomic model for the transmembrane region of the PLB pentamer, constrained by previous mutagenesis results and the results of the present study. We conclude that (a) the mobilities of spin-labels attached to PLB and its mutants are sensitive to oligomeric state and (b) the pattern of cysteine reactivity, spin-label mobility, and oligomeric state supports a structural model for the PLB pentamer in which interactions between each pair of subunits are stabilized by a leucine-isoleucine zipper.

Relaxation time determinations by progressive saturation EPR: effects of molecular motion and Zeeman modulation for spin labels

The EPR spectra of nitroxide spin labels have been simulated as a function of microwave field, H1, taking into account both magnetic field modulation and molecular rotation. It is found that the saturation of the second integral, S, of the first harmonic in-phase absorption spectrum is approximated by that predicted for slow-passage conditions, that is, S approximately H1/1 + PH21, in all cases. This result is independent of the degree of inhomogeneous broadening. In general, the fitting parameter, P, depends not only on the T1 and T2 relaxation times, but also on the rate of molecular reorientation and on the modulation frequency. Calibrations for determining the relaxation times are established from the simulations. For a given modulation frequency and molecular reorientation rate, the parameter obtained by fitting the saturation curves is given by 1/P = a + 1/gamma2eT1 . Teff2, where Teff2 is the effective T2. For molecular reorientation frequencies in the range 2 x 10(7)-2 x 10(8) s-1, Teff2 is dominated by the molecular dynamics and is only weakly dependent on the intrinsic T02, allowing a direct estimation of T1. For reorientation frequencies outside this range, the (T1T2) product may be determined from the calibrations. The method is applied to determining relaxation times for spin labels undergoing different rates of rotational reorientation in a variety of environments, including those of biological relevance, and is verified experimentally by the relaxation rate enhancements induced by paramagnetic ions.

Phosphorylation of phospholamban by cAMP-dependent protein kinase enhances interactions between Ca-ATPase polypeptide chains in cardiac sarcoplasmic reticulum membranes

We have used spin-label EPR spectroscopy to examine possible alterations in protein-protein interactions that accompany the activation of the cardiac sarcoplasmic reticulum (SR) Ca-ATPase following the phosphorylation of phospholamban (PLB). Using a radioactive derivative of a maleimide spin label (MSL), we have developed conditions for the selective spin-labeling of the Ca-ATPase in both native cardiac and skeletal sarcoplasmic reticulum membranes. The rotational dynamics of the cardiac and skeletal Ca-ATPase isoforms in native SR membranes were measured using saturation transfer EPR. We report that the phosphorylation of PLB in cardiac SR results in a (1.8 +/- 0.2)-fold reduction in the overall rotational mobility of the Ca-ATPase. The alteration in the rotational dynamics of the Ca-ATPase is the direct result of the phosphorylation of PLB, and is not related to the phosphorylation of the Ca-ATPase or any other SR proteins since no alteration in the ST-EPR spectrum is observed as a result of conditions that phosphorylate the cardiac Ca-ATPase with ATP. Neither do the use of conditions that activate the Ca-ATPase in cardiac SR result in the alteration of the rotational dynamics or catalytic properties of the Ca-ATPase in skeletal SR where PLB is not expressed. Measurements of the rotational dynamics of stearic acid spin labels (SASL) incorporated into cardiac SR membranes with a nitroxide at the 5- and 12-positions using conventional EPR indicate that there is virtually no difference in the lipid acyl chain dynamics in cardiac SR membranes upon the phosphorylation of PLB. These results indicate that the decrease in the rotational dynamics of the Ca-ATPase in cardiac SR membranes associated with the phosphorylation of PLB is related to enhanced interactions between individual Ca-ATPase polypeptide chains due to (i) an alteration in the spatial arrangement of cardiac Ca-ATPase polypeptide chains within a defined oligomeric state or (ii) increased protein-protein associations. We suggest that altered interactions between Ca-ATPase polypeptide chains and PLB serves to modulate the activation barrier associated with calcium activation of the Ca-ATPase in cardiac SR membranes.

Electron spin resonance studies of acyl chain motion in reconstituted nicotinic acetylcholine receptor membranes

The electron spin resonance spectra of spin-label positional isomers of stearic acid (n-SASL) incorporated into nicotinic acetylcholine receptors (nAcChoR) reconstituted into dioleoylphosphatidylcholine (DOPC) were deconvoluted into bilayer- and protein-associated components by subtraction under conditions of slow exchange. The selectivity of n-SASL (n = 6, 9, 12, and 14) for the lipid-protein interface of the nAcChoR was threefold greater than that of DOPC and independent of the spin label position. The temperature at which exchange became apparent as judged from lineshape broadening of the mobile lipid component spectrum was dependent upon the position of the spin-label moiety; near the bilayer center, exchange broadening occurred at lower temperatures than it did closer to the lipid headgroup. This suggests that the lipid headgroup region of boundary lipids is relatively fixed, whereas its acyl chain whips on and off the protein with increasing frequency near the bilayer center. Motions on the microsecond time scale were examined by microwave power saturation. Each n-SASL saturated more readily when incorporated into vesicles containing the nAcChoR than when in pure DOPC liposomes. Therefore, lipid mobility is perturbed by the nAcChoR on the microsecond time scale with an apparent magnitude that is relatively modest, probably due to exchange on this time scale.

Resolution of three structural states of spin-labeled myosin in contracting muscle

We have used electron paramagnetic resonance (EPR) spectroscopy to detect ATP- and calcium-induced changes in the structure of spin-labeled myosin heads in glycerinated rabbit psoas muscle fibers in key physiological states. The probe was a nitroxide iodoacetamide derivative attached selectively to myosin SH1 (Cys 707), the conventional EPR spectra of which have been shown to resolve several conformational states of the myosin ATPase cycle, on the basis of nanosecond rotational motion within the protein. Spectra were acquired in rigor and during the steady-state phases of relaxation and isometric contraction. Spectral components corresponding to specific conformational states and biochemical intermediates were detected and assigned by reference to EPR spectra of trapped kinetic intermediates. In the absence of ATP, all of the myosin heads were rigidly attached to the thin filament, and only a single conformation was detected, in which there was no sub-microsecond probe motion. In relaxation, the EPR spectrum resolved two conformations of the myosin head that are distinct from rigor. These structural states were virtually identical to those observed previously for isolated myosin and were assigned to the populations of the M*.ATP and M**.ADP.Pi states. During isometric contraction, the EPR spectrum resolves the same two conformations observed in relaxation, plus a small fraction (20-30%) of heads in the oriented actin-bound conformation that is observed in rigor. This rigor-like component is a calcium-dependent, actin-bound state that may represent force-generating cross-bridges. As the spin label is located near the nucleotide-binding pocket in a region proposed to be pivotal for large-scale force-generating structural changes in myosin, we propose that the observed spectroscopic changes indicate directly the key steps in energy transduction in the molecular motor of contracting muscle.

Adaptive changes in lipid composition of skeletal sarcoplasmic reticulum membranes associated with aging

We have undertaken a detailed examination of changes associated with aging in lipid composition and corresponding physical properties of hindlimb skeletal sarcoplasmic reticulum (SR) membranes isolated from young (5 months), middle-aged (16 months), and old (28 months) Fischer strain 344 rats. Silica gel HPLC chromatography was used to separate phospholipid headgroup species. Subsequent reversed-phase HPLC was used to resolve fatty acid chain compositions of phosphatidylcholine, phosphatidylethanolamine, and phosphatidylinositol species. For all three phospholipid pools, significant age-related variations are observed in the abundance of multiple molecular species, particularly those having polyunsaturated fatty acid chains. Using mass spectrometry (fast atom bombardment and tandem techniques) to distinguish ester- from ether-linked phosphatidylethanolamine species, we demonstrate that overall plasmenylethanolamine content is substantially increased with age, from 48 mol% to 62 mol%. A substantial increase is also observed in the single molecular species 18:0-20:4 phosphatidylinositol suggesting implications for signalling pathways. In addition, associated with senescence we find a significant increase in the rigidifying lipid, cholesterol. Despite these changes in lipid composition of different aged animals, the average bilayer fluidity examined at several bilayer depths with stearic acid spin labels, is not altered. Neither do we find differences in the rotational mobility of maleimide spin-labeled Ca(2+)-ATPase, as determined from saturation-transfer electron paramagnetic resonance, which is sensitive to both the fluidity of lipids directly associated with the Ca(2+)-ATPase and to its association with proteins.

Molecular dynamics in mouse atrial tumor sarcoplasmic reticulum

We have determined directly the effects of the inhibitory peptide phospholamban (PLB) on the rotational dynamics of the calcium pump (Ca-ATPase) of cardiac sarcoplasmic reticulum (SR). This was accomplished by comparing mouse ventricular SR, which has PLB levels similar to those found in other mammals, with mouse atrial SR, which is effectively devoid of PLB and thus has much higher (unregulated) calcium pump activity. To obtain sufficient quantities of atrial SR, we isolated the membranes from atrial tumor cells. We used time-resolved phosphorescence anisotropy of an erythrosin isothiocyanate label attached selectively and rigidly to the Ca-ATPase, to detect the microsecond rotational motion of the Ca-ATPase in the two preparations. The time-resolved phosphorescence anisotropy decays of both preparations at 25 degrees C were multi-exponential, because of the presence of different oligomeric species. The rotational correlation times for the different oligomers were similar for the two preparations, but the total decay amplitude was substantially greater for atrial tumor SR, indicating that a smaller fraction of the Ca-ATPase molecules exists as large aggregates. Phosphorylation of PLB in ventricular SR decreased the population of large-scale Ca-ATPase aggregates to a level similar to that of atrial tumor SR. Lipid chain mobility (fluidity), detected by electron paramagnetic resonance of stearic acid spin labels, was very similar in the two preparations, indicating that the higher protein mobility in atrial tumor SR is not due to higher lipid fluidity. We conclude that PLB inhibits by inducing Ca-ATPase lateral aggregation, which can be relieved either by phosphorylating or removing PLB.

Anesthetics alter the physical and functional properties of the Ca-ATPase in cardiac sarcoplasmic reticulum

We have studied the effects of the local anesthetic lidocaine, and the general anesthetic halothane, on the function and oligomeric state of the CA-ATPase in cardiac sarcoplasmic reticulum (SR). Oligomeric changes were detected by time-resolved phosphorescence anisotropy (TPA). Lidocaine inhibited and aggregated the Ca-ATPase in cardiac SR. Micromolar calcium or 0.5 M lithium chloride protected against lidocaine-induced inhibition, indicating that electrostatic interactions are essential to lidocaine inhibition of the Ca-ATPase. The phospholamban (PLB) antibody 2D12, which mimics PLB phosphorylation, had no effect on lidocaine inhibition of the Ca-ATPase in cardiac SR. Inhibition and aggregation of the Ca-ATPase in cardiac SR occurred at lower concentrations of lidocaine than necessary to inhibit and aggregate the Ca-ATPase in skeletal SR, suggesting that the cardiac isoform of the enzyme has a higher affinity for lidocaine. Halothane inhibited and aggregated the Ca-ATPase in cardiac SR. Both inhibition and aggregation of the Ca-ATPase by halothane were much greater in the presence of PLB antibody or when PLB was phosphorylated, indicating a protective effect of PLB on halothane-induced inhibition and aggregation. The effects of halothane on cardiac SR are opposite from the effects of halothane observed in skeletal SR, where halothane activates and dissociates the Ca-ATPase. These results underscore the crucial role of protein-protein interactions on Ca-ATPase regulation and anesthetic perturbation of cardiac SR.

Self-association accompanies inhibition of Ca-ATPase by thapsigargin

Recent studies have demonstrated a relationship between the activity of the Ca-ATPase of sarcoplasmic reticulum and its state of self-association. In the present study, the effects of thapsigargin (TG), a toxin that specifically inhibits the Ca-ATPase of rabbit skeletal muscle sarcoplasmic reticulum membrane, were studied by detecting the time-resolved phosphorescence anisotropy (TPA) decay of the Ca-ATPase that had been labeled with the phosphorescent probe erythrosin-isothiocyanate (ErITC). Anisotropy decays were fit to a function that consisted of three exponential decays plus a constant background, as well as to a function describing explicitly the uniaxial rotation of proteins in a membrane. In the absence of TG, the anisotropy was best-fit by a model representing the rotation of three populations, corresponding to different-sized oligomeric species in the membrane. The addition of stoichiometric amounts of TG to the Ca-ATPase promptly decreased the overall apparent rate of decay, indicating decreased rotational mobility. A detailed analysis showed that the principal change was not in the rates of rotation but rather in the population distribution of the Ca-ATPase molecules among the different-sized oligomers. TG decreased the proportion of small oligomers and increased the proportion of large ones. Preincubation of the ErITC-SR in 1 mM Ca2+, which stabilizes the E1 conformation relative to E2, was found to protect partially against the changes in the TPA associated with the presence of the inhibitor. These results are consistent with the hypothesis that TG inhibits the Ca-ATPase by stabilizing it in an E2-like conformation, which promotes the formation of larger aggregates of the enzyme. When combined with the effects of other inhibitors on the Ca-ATPase, these results support a general model for the coupling of enzyme conformation and self-association in this system.

Hexanol and lidocaine affect the oligomeric state of the Ca-ATPase of sarcoplasmic reticulum

Hexanol at 7 degrees C stimulates the activity of the Ca-ATPase of sarcoplasmic reticulum (SR). Time-resolved phosphorescence spectroscopy studies of SR whose Ca-ATPase is covalently labeled with erythrosin isothiocyanate (ERITC) indicate that at 7 degrees C hexanol (1) cause a concentration-dependent increase in the rate of decay of phosphorescence anisotropy, (2) causes larger oligomers of Ca-ATPase to dissociate into smaller oligomers, and (3) increases the rotational mobility of Ca-ATPase in all its oligomeric states. Electron paramagnetic resonance (EPR) spectroscopy of spin-labeled stearic acid (SASL) in SR suggests that at 7 degrees C hexanol diminishes the fraction of SR lipids in the boundary lipid domain and disorders and fluidizes both the boundary lipid and the unrestricted lipid domain. In protein-free liposomes of extracted SR lipids hexanol increases fluidity and decreases order to a greater extent near the center of the lipid bilayer than near the polar head groups. At 25 degrees C hexanol has biphasic effects on Ca-ATPase activity: at 10 and 20 mM hexanol increases activity, but at 30 mM and especially at 40 mM there is inhibition of Ca-ATPase activity. The influence of hexanol at 25 degrees C on the oligomeric state of Ca-ATPase is also biphasic. At 10 and 20 mM, hexanol promotes the dissociation of larger oligomers into smaller ones, whereas at higher concentrations, 30 and 40 mM, hexanol causes larger oligomers to be formed from smaller ones. Lidocaine at 25 degrees C inhibits Ca-ATPase activity and causes dramatic slowing of the decay of phosphorescence anisotropy of ERITC-labeled SR by causing the formation of larger oligomers of Ca-ATPase from smaller ones. In protein-free liposomes of SR lipids at 25 degrees C, lidocaine disorders and fluidizes the acyl chains near the center of the bilayer (as did hexanol), but has opposite effects near the polar head groups. The opposite effects of hexanol and lidocaine on the oligomeric state of the SR Ca-ATPase provide a new molecular explanation for the opposite effects of hexanol and lidocaine on the activity of the Ca-ATPase. We conclude that the biphasic effects of hexanol on the activity of Ca-ATPase can be accounted for by biphasic effects of hexanol on the oligomeric state of the Ca-ATPase. This study supports the view that anesthetics can alter interactions between membrane proteins.

Orientation and rotational dynamics of spin-labeled phalloidin bound to actin in muscle fibers

We have used electron paramagnetic resonance spectroscopy (EPR) to investigate the orientational distribution of actin in thin filaments of glycerinated muscle fibers in rigor, relaxation, and contraction. A spin-labeled derivative of a mushroom toxin, phalloidin (PHSL), was bound to actin in the muscle fibers (PHSL-fibers). The EPR spectrum of unoriented PHSL-labeled myofibrils consisted of three sharp lines with a splitting between the outer extrema (2T parallel') of 42.8 +/- 0.1 G, indicating that the spin labels undergo restricted nanosecond rotational motion within an estimated half-cone angle of 76 degrees. When the PHSL-fiber bundle was oriented parallel to the magnetic field, the splitting between the zero-crossing points (2T') was 42.7 +/- 0.1 G. When the fiber bundle was perpendicular to the magnetic field, 2T' decreased to 34.5 +/- 0.2 G. This anisotropy shows that the motion of the probe is restricted in orientation by its binding site on actin, so that the EPR spectrum of PHSL-fiber bundles would be sensitive to small changes in the mean axial orientation of the PHSL-actin interface. No differences in the EPR spectra were observed in fibers during rigor, relaxation, or contraction, indicating that the mean axial orientation of the PHSL binding site changes by less than 5 degrees, and that the amplitude of nanosecond probe rotational motion, which should be quite sensitive to the local environment of the phalloidin, changes by no more than 1 degree.(ABSTRACT TRUNCATED AT 250 WORDS)

Molecular mechanism of Ca-ATPase activation by halothane in sarcoplasmic reticulum

We have studied the molecular mechanism of Ca-ATPase activation in sarcoplasmic reticulum (SR) by the volatile anesthetic halothane. Using time-resolved phosphorescence anisotropy, we determined the rotational correlation times and mole fractions of different oligomeric states of the enzyme, as a function of halothane and temperature. Lipid fluidity was measured independently, using EPR of spin-labeled lipids. At 4 and 7 degrees C, the principal effects of halothane were to increase the activity of the Ca-ATPase and to promote the formation of monomers and dimers of the enzyme from larger aggregates. At higher temperatures (up to 25 degrees C), halothane activated the enzyme, but to a lesser extent than observed at lower temperatures. While the functional effects of halothane were temperature dependent, the effects of halothane on lipid fluidity and protein aggregation state were similar at all temperatures tested. We conclude that at low temperatures Ca-ATPase activity is dominated by aggregation state, so halothane activates the enzyme primarily by promoting the formation of monomers and dimers of the enzyme from larger aggregates. At higher temperatures, the activity of the enzyme is dominated by lipid fluidity, so halothane activates the enzyme by increasing the lipid fluidity. The physical mechanism of Ca-ATPase activation, dominated by aggregation state at low temperature and lipid fluidity at higher temperature, provides an explanation for the break in the Arrhenius plot of Ca-ATPase activity (in the absence of halothane) at approximately 20 degrees C.

Transient detection of spin-labeled myosin subfragment 1 conformational states during ATP hydrolysis

We have used time-resolved electron paramagnetic resonance spectroscopy and caged ATP to detect nucleotide-induced changes in the conformational state of spin-labeled myosin heads (IASL-S1). Changes in the internal rotational dynamics of IASL-S1 were monitored with millisecond time resolution during the pre-steady-state phase of ATP hydrolysis. The changes in the internal protein dynamics were rigorously correlated with specific biochemical kinetic transitions, allowing us to observe directly the dynamic sequence of structural changes in IASL-S1 during the binding and hydrolysis of ATP. When caged ATP was photolyzed (producing 500 microM ATP) in the presence of 100 microM IASL-S1, the EPR signal intensity rose transiently to the steady-state ATPase level, indicating increased rotational motion about the SH1 region of the myosin head. Kinetic and spectral analyses have resolved two phases of this transient, one representing the population of the M*.ATP state and the other representing the population of the M**.ADP.Pi state. We conclude that two motionally distinct states of the myosin head are present during ATP hydrolysis and that these states represent distinct conformational states that can be correlated with specific biochemical intermediates. Since specific labeling of myosin heads with IASL has been achieved in skinned muscle fibers, this study establishes the feasibility for the first direct detection and resolution of myosin's conformational transients during muscle contraction.

Exchange rates at the lipid-protein interface of the myelin proteolipid protein determined by saturation transfer electron spin resonance and continuous wave saturation studies

The microwave saturation properties of various spin-labeled lipids in reconstituted complexes of the myelin proteolipid protein with dimyristoyl phosphatidylcholine have been studied both by conventional and saturation transfer electron spin resonance (ESR) spectroscopy. In the fluid phase, the conventional ESR spectra consist of a fluid and a motionally restricted (i.e., protein-associated) component, whose relative proportions can be determined by spectral subtractions and depend on the selectivity of the particular spin-labeled lipid for the protein. At 4 degrees C when the bulk lipid is in the gel phase, the integrated intensity of the saturation transfer ESR spectra displays a linear dependence on the fraction of motionally restricted lipid that is deduced from the conventional ESR spectra in the fluid phase, indicating the presence of distinct populations of free and protein-interacting lipid with no exchange between them on the saturation transfer ESR time scale in the gel phase. At 30 degrees C when the bulk lipid is in the fluid phase, the saturation transfer integral displays a nonlinear dependence on the fraction of motionally restricted lipid, consistent with exchange between the two lipid populations on the saturation transfer ESR time scale in the fluid phase. For lipid spin labels with different selectivities for the protein in complexes of fixed lipid/protein ratio, the data in the fluid phase are consistent with a constant (diffusion-controlled) on-rate for exchange at the lipid-protein interface. Values ranging between 1 and 9 x 10(6) s-1 are estimated for the intrinsic off-rates for exchange of spin-labeled stearic acid and phosphatidylcholine, respectively, at 30 degrees C. Conventional continuous wave saturation experiments lead to similar conclusions regarding the lipid exchange rates in the fluid and gel phases of the lipid/protein recombinants. The ESR saturation studies therefore demonstrate exchange on the time scale of the nitroxide spin-lattice relaxation at the lipid-protein interface of myelin proteolipid/dimyristoyl phosphatidylcholine complexes in the fluid phase but not in the gel phase.

Effects of melittin on lipid-protein interactions in sarcoplasmic reticulum membranes

To investigate the physical mechanism by which melittin inhibits Ca-adenosine triphosphatase (ATPase) activity in sarcoplasmic reticulum (SR) membranes, we have used electron paramagnetic resonance spectroscopy to probe the effect of melittin on lipid-protein interactions in SR. Previous studies have shown that melittin substantially restricts the rotational mobility of the Ca-ATPase but only slightly decreases the average lipid hydrocarbon chain fluidity in SR. Therefore, in the present study, we ask whether melittin has a preferential effect on Ca-ATPase boundary lipids, i.e., the annular shell of motionally restricted lipid that surrounds the protein. Paramagnetic derivatives of stearic acid and phosphatidylcholine, spin-labeled at C-14, were incorporated into SR membranes. The electronic paramagnetic resonance spectra of these probes contained two components, corresponding to motionally restricted and motionally fluid lipids, that were analyzed by spectral subtraction. The addition of increasing amounts of melittin, to the level of 10 mol melittin/mol Ca-ATPase, progressively increased the fraction of restricted lipids and increased the hyperfine splitting of both components in the composite spectra, indicating that melittin decreases the hydrocarbon chain rotational mobility for both the fluid and restricted populations of lipids. No further effects were observed above a level of 10 mol melittin/mol Ca-ATPase. In the spectra from control and melittin-containing samples, the fraction of restricted lipids decreased significantly with increasing temperature. The effect of melittin was similar to that of decreased temperature, i.e., each spectrum obtained in the presence of melittin (10:1) was nearly identical to the spectrum obtained without melittin at a temperature approximately 5 degrees C lower. The results suggest that the principal effect of melittin on SR membranes is to induce protein aggregation and this in turn, augmented by direct binding of melittin to the lipid, is responsible for the observed decreases in lipid mobility. Protein aggregation is concluded to be the main cause of inactivation of the Ca-ATPase by melittin, with possible modulation also by the decrease in mobility of the boundary layer lipids.

Protein sulfhydryls are protected from irreversible oxidation by conversion to mixed disulfides

Protein mixed thioselenides formed by reaction of sarcoplasmic reticulum (SR) with diselenide biradical spin labels were quantified by ESR. Whereas the reaction of SR membranes with the diselenide spin label led to a large ESR signal of the unbound monoselenide at equilibrium, treatment of the reaction mixture with a few millimolar hydrogen peroxide converted all of the nitroxides to protein-bound thioselenides. This technique of spin-labeling protein thiols avoids the need to remove unreacted spin labels. The bound spin labels were removable by reduction with excess mercaptoethanol, indicating a specific and reversible labeling of protein thiols. SR that had been extensively labeled with the diselenide spin label was resistant to ATPase inactivation by potent oxidants that arise when myoglobin reacts with hydroperoxides. Unmodified SR lost all activity within 10 min of exposure to either 1 mM tert-butyl hydroperoxide in the presence of 200 microM equine myoglobin or to 100 mM hydrogen peroxide in the absence of myoglobin. In both cases the loss of activity could not be reversed by subsequent treatment with mercaptoethanol. On the other hand, membranes that had been extensively treated with the diselenide spin label and were then subjected to these peroxide treatments were fully active after mercaptoethanol-mediated cleavage of the thioselenides. ESR analysis of spin-labeled SR showed no detectable oxidative cleavage of the thioselenide bonds. Sodium dodecyl sulfate gel electrophoresis showed that peroxide-mediated crosslinking of ATPase observed in unmodified SR membranes did not occur in the diselenide-modified SR membranes. Only limited protection was observed when SR pretreated with glutathione disulfide was incubated with hydroperoxides. In this case, however, the degree of protection was greatly increased when the reaction with glutathione disulfide was carried out in the presence of the supernatant of centrifuged rat liver homogenate, consistent with an acceleration of mixed disulfide formation by a factor tentatively identified as thiol transferase. It is concluded that conversion of protein thiol residues to either thioselenides or mixed disulfides confers protection against irreversible peroxide-dependent oxidation. We suggest that mixed disulfide formation by thiol transferase activity may help protect protein thiols from irreversible oxidation by heme-activated hydroperoxides.

Effects of ouabain on the rotational dynamics of renal Na,K-ATPase studied by saturation-transfer EPR

The interaction of a nitroxide spin-labeled derivative of ouabain with sheep kidney Na,K-ATPase and the motional behavior of the ouabain spin label-Na,K-ATPase complex have been studied by means of electron paramagnetic resonance (EPR) and saturation-transfer EPR (ST-EPR). Spin-labeled ouabain binds with high affinity to the Na,K-ATPase with concurrent inhibition of ATPase activity. Enzyme preparations retain 0.61 +/- 0.1 mol of bound ouabain spin label per mole of ATP-dependent phosphorylation sites, even after repeated centrifugation and resuspension of the purified ATPase-containing membrane fragments. The conventional EPR spectrum of the ouabain spin label bound to the ATPase consists almost entirely (greater than 99%) of a broad resonance at 0 degrees C, characteristic of a tightly bound spin label which is strongly immobilized by the protein backbone. Saturation-transfer EPR measurements of the spin-labeled ATPase preparations yield effective correlation times for the bound labels significantly longer than 100 microseconds at 0 degrees C. Since the conventional EPR measurements of the ouabain spin-labeled Na,K-ATPase indicated the label was strongly immobilized, these rotational correlation times most likely represent the motion of the protein itself rather than the independent motion of mobile spin probes relative to a slower moving protein. Additional ST-EPR measurements of ouabain spin-labeled Na,K-ATPase (a) cross-linked with glutaraldehyde and (b) crystallized in two-dimensional arrays indicated that the observed rotational correlation times predominantly represented the motion of large Na,K-ATPase-containing membrane fragments, as opposed to the motion of individual monomeric or dimeric polypeptides within the membrane fragment. The results suggest that the binding of spin-labeled ouabain to the ATPase induces the protein to form large aggregates, implying that cardiac glycoside induced enzyme aggregation may play a role in the mechanism of action of the cardiac glycosides in inhibiting the Na,K-ATPase.

Resolution of phospholipid conformational heterogeneity in model membranes by spin-label EPR and frequency-domain fluorescence spectroscopy

We have utilized both fluorescent and nitroxide derivatives of stearic acid as probes of membrane structural heterogeneity in phospholipid vesicles under physiological conditions, as well as conditions of varying ionic strengths and temperatures where spectral heterogeneity has been previously observed and attributed to multiple ionization states of the probes. To identify the source of this spectral heterogeneity, we have utilized complimentary measurements of the relaxation properties (lifetimes) and motion of both (a) spin labeled and anthroyloxy derivatives of stearic acid (i.e., SASL and AS) and (b) a diphenylhexatriene derivative of phosphatidylcholine (DPH-PC) in single component membranes containing dimyristoylphosphatidylcholine (DMPC). We use an 15N stearic-acid spin label for optimal sensitivity to membrane heterogeneity. The lifetime and dynamics of the fluorescent phospholipid analogue DPH-PC (with no ionizable groups over this pH range) were compared with those of AS, allowing us to discriminate between changes in membrane structure and the ionization of the label. The quantum yield and rotational dynamics of DPH-PC are independent of pH, indicating that changes in pH do not affect the conformation of the host phospholipids. However, both EPR spectra of SASL and the lifetime or dynamics of AS are affected profoundly by changes in solution pH. The apparent pKa's of these two probes in DMPC membranes were determined to be near pH 6.3, implying that at physiological pH and ionic strength these stearic-acid labels exist predominantly as a single ionized population in membranes. Therefore, the observed temperature- and ionic-strength-dependent alterations in the spectra of SASL as well as the lifetime or dynamics of AS in DMPC membranes at neutral pH are due to changes in membrane structure rather than the ionization of the probes. The possibility that ionic gradients across biological membranes induce alterations in phospholipid structures, thereby modulating lipid-protein interactions is discussed.

Membrane fluidity as affected by the organochlorine insecticide DDT

Fluorescence polarization of 1,6-diphenyl-1,3,5-hexatriene (DPH) was used to study the interaction of DDT with model and native membranes. DDT decreases the phase transition midpoint temperature (Tm) of liposomes reconstituted with dimyristoyl-, dipalmitoyl- and distearoylphosphatidylcholines (DMPC, DPPC and DSPC), and broadens the thermotropic profile of the transition. The effects of DDT are concentration dependent and are more pronounced in bilayers of short-chain lipids, e.g., DMPC. The insecticide fails to alter DPH polarization in the fluid phase of the above lipids. Similar effects were observed in binary mixtures of DMPC plus DPPC. Furthermore, DDT alters the single broad transition of the equimolar mixture of DMPC plus DSPC into a biphasic transition. The lower temperature component has a midpoint at 25 degrees C, i.e., a value close to the Tm of DMPC. DDT inhibits to some extent the cholesterol-induced ordering in DMPC bilayers and high cholesterol concentrations (greater than or equal to 30 mol%) do not prevent insecticide interaction, conversely to the effect observed for lindane (Antunes-Madeira, M.C. and Madeira, V.M.C. (1989) Biochim. Biophys. Acta 982, 161-166). Apparently, the bilayer order is not disturbed by DDT in fluid native membranes of mitochondria and sarcoplasmic reticulum, but moderate disordering effects are noticed in membranes enriched in cholesterol, namely, brain microsomes and erythrocytes.