Transient kinetics and thermodynamics of anthroylouabain binding to Na/K-ATPase.

The Na/K-ATPase is an integral membrane protein enzyme which uses energy derived from hydrolysis of ATP to pump Na+ out of and K+ into the cell. Ouabain belongs to a class of drugs known as cardiac glycosides, which are useful for treating congestive heart failure. Therapeutic value is achieved when these drugs bind to and inhibit the Na/K-ATPase of cardiac muscle. We gain insight into this important interaction by measuring the thermodynamics of the interaction of anthroylouabain (AO), a fluorescent derivative of ouabain, with the Na/K-ATPase. AO has the useful property that its fluorescence intensity is greatly enhanced (approximately 10x) when it binds to the enzyme. Using this enhancement, we measure temperature dependence of transient kinetics for the association and dissociation of AO interacting with membrane fragments of Na/K-ATPase purified from dog kidney. Using a standard Eyring analysis, we find that the overall association of AO with the enzyme is driven by substantial contributions from both enthalpy and entropy, and that in an energy diagram for the association pathway, the free energy change is quite similar to that of ouabain deduced from previously published results [E. Erdmann, W. Schoner, BBA 307 (1973) 386]. However, in the transition state, there are substantial differences for the enthalpy and entropy, presumably due to the presence of the anthracene moiety.

Nucleotide binding to IAF-labelled Na+/K(+)-ATPase measured by steady state fluorescence quenching by TNP-ADP.

Nucleotide binding to 5-iodoacetamidofluorescein (IAF) labelled Na+/K(+)-ATPase was measured by steady state fluorescence quenching of the fluorescein label via energy transfer to trinitrophenyl (TNP) labelled nucleotide. TNP-nucleotides are valuable probes of nucleotide binding to ATPases. Interpretation of these and other experiments in our laboratory using TNP-nucleotides with the Na+/K(+)-ATPase rely on having a good model for the interaction of TNP-nucleotide with the enzyme. Sets of fluorescence quenching curves obtained by titrating the enzyme with TNP-ADP in the presence of various concentrations of ADP could not be adequately modelled using a simple model with a single nucleotide binding site. Therefore, we compare various models which allow for additional TNP-nucleotide binding to the enzyme. In the two-site model, the additional binding is to a second specific site for which TNP-nucleotide and unlabelled nucleotide compete. In two other models, the additional binding (in one case saturable, and in the other case non-saturable) of TNP-nucleotide is not blocked by or affected by unlabelled nucleotide, and is, therefore, referred to as non-specific binding of the TNP-nucleotide. The goal of this work is to determine which of the distinctly different physical pictures associated with these models most accurately describes the interaction of TNP-nucleotide with the enzyme. We find that the interaction of TNP-ADP with IAF-labelled Na+/K(+)-ATPase is best described by a model in which there are two classes of binding: TNP-ADP and ADP compete for a specific binding site with dissociation binding constants of 0.13 microM for TNP-ADP and 2.0 microM for ADP; and non-saturable non-specific binding of TNP-ADP.

Transient kinetics of substrate binding to Na+/K(+)-ATPase measured by fluorescence quenching.

This paper examines the transient kinetics of substrate binding to the Na+/K(+)-ATPase labelled with iodoacetamidofluorescein (IAF) using fluorescence quenching by trinitrophenyl-ATP (TNP-ATP). Earlier work (E.H. Hellen, P.R. Pratap, 1996, Fluorescence quenching of IAF-Na+/K(+)-ATPase via energy transfer to TNP-labelled nucleotide, Proceedings of the VIIIth International Conference on the Na+/K(+)-ATPase, in press) has shown that TNP-nucleotide binds to specific sites (from which unlabelled nucleotide can displace it) and nonspecific sites (from which unlabelled nucleotide cannot displace it). Under stopped-flow conditions, quenching of IAF-enzyme fluorescence was well described by a stretched exponential (F(t) = F infinity + delta F exp[-Bt alpha]). Physically, this function may be interpreted in terms of its inverse Laplace transform phi (k), which describes a distribution of rate-constants; alpha reflects the width of this distribution. As TNP-ATP concentration increased, alpha decreased, reflecting TNP-ATP binding to sites with higher energy barriers. alpha decreased by about the same amount with increasing [TNP-ATP] in the presence of saturating ATP, indicating that the distribution of rate-constants is largely associated with the nonspecific binding sites. However, alpha was significantly less than 1 for ATP-induced fluorescence recovery in the presence of TNP-ATP, indicating that rate-constants associated with specific binding site are also distributed. The distribution of rate-constants for binding to the specific site indicates a distribution in the energy of the transition state for substrate binding. These results suggest that the specific binding site (in either the empty or the full state) may exist in a series of conformations separated by small energy barriers. However, the energy barriers for binding associated with these conformations are significantly distributed.

Kinetics of conformational changes associated with potassium binding to and release from Na+/K(+)-ATPase.

The Na+/K(+)-ATPase functions in cells to couple energy from the hydrolysis of ATP to the transport Na+ out and K+ in. The fluorescent probe IAF (iodoacetamidofluorescein) covalently binds to this enzyme, reporting conformational changes without inhibiting enzyme activity. This paper describes experiments using dog kidney enzyme labeled with IAF to examine kinetics of conformational changes resulting from added Na+ and K+, measured in terms of steady-state and stopped-flow fluorescence changes. Kinetics of these fluorescence changes were examined as a function of temperature from two initial conditions: (a) enzyme in the high-fluorescence form (E(high)) was rapidly mixed with varying [K+]; and (b) enzyme in the low-fluorescence form (E(low)) was rapidly mixed with varying [ATP]. These experiments showed: (1) The rate constant for the fluorescence change from E(high) to E(low) was much larger than that for the opposite transition, E(low) to E(high); (2) the apparent free energy of activation (Ea(app)) for the two transitions were different (as estimated from Arrhenius plots); (3) under steady-state conditions, IAF fluorescence did not change when ATP was added to E(low)(K+) in the absence of Na+; (4) the apparent free energy of activation was independent of [K+] for the E(high) to E(low) transition (at 16.4 kcal/mol) but increased with [ATP] for the E(low) to E(high) transition; (5) Ea(app) for the E(low) to E(high) transition with 1 mM ATP was approximately the same as that in the absence of ATP (34 kcal/mol). These results can be interpreted as: (i) in the transition from E(low) to E(high), IAF reported a conformational change that occurred after K+ release to the intracellular side and which is involved in Na+ binding; (ii) Ea(app) increased with [ATP], while increasing the entropy of the transition state. Thus, ATP appeared to destabilize the enzyme during the transition from E(low) to E(high).

Voltage dependence of DIDS-insensitive chloride conductance in human red blood cells treated with valinomycin or gramicidin.

Net K and Cl effluxes induced by valinomycin or by gramicidin have been determined directly at varied external K, denoted by [K]o, in the presence and absence of the anion transport inhibitors DIDS (4,4'-diiso-thiocyano-2,2'-disulfonic acid stilbene), and its less potent analogue SITS (4-acetamido-4'-isothiocyanostilbene-2,2'-disulfonic acid). The results confirm that pretreatment with 10 microM DIDS, or 100 microM SITS, for 30 min at 23 degrees C inhibits conductive Cl efflux, measured in the continued presence of the inhibitors at 1 mM [K]o, by only 59-67%. This partial inhibition by 10 microM DIDS at 1 mM [K]o remains constant when the concentration of DIDS, or when the temperature or pH during pretreatment with DIDS, are increased. Observations of such partial inhibition previously prompted the postulation of two Cl conductance pathways in human red blood cells: a DIDS-sensitive pathway mediated by capnophorin (band 3 protein), and a DIDS-insensitive pathway. The present experiments demonstrate that at [K]o corresponding to values of EK between -35 and 0 mV the DIDS-insensitive component of net Cl efflux is negligible, being < or = 0.1 muMol/g Hb/min, both with valinomycin (1 microM) and with gramicidin (0.06 microgram/ml). At lower [K]o, where EK is below approximately -35 mV, the DIDS-insensitive fraction of net Cl efflux increases to 2.6 muMol/g Hb/min with valinomycin (1 microM), and to 4.8 muMol/g Hb/min with gramicidin (0.06 microgram/ml). With net fluxes determined from changes in mean cell volume, and with membrane potentials measured from changes in the external pH of unbuffered red cell suspensions, a current-voltage curve for DIDS-insensitive Cl conductance has been deduced. While specific effects of varied [K]o on net Cl efflux are unlikely but cannot strictly be ruled out, the results are consistent with the hypothesis that DIDS-insensitive Cl conductance turns on at an Em of approximately -40 mV.

Rapid kinetic analyses of the Na+/K(+)-ATPase distinguish among different criteria for conformational change.

The Na+/K(+)-ATPase couples the hydrolysis of ATP to the transport of Na+ and K+ via a phosphorylated intermediate and conformational changes. In order to identify these conformational changes, we have probed the sequence of steps from EP(3Na+ in) to EP + 3Na+ out with three fluorescent probes (IAF: 5-iodoacetamidofluorescein; BIPM: N-[p-(2-benzimidazolyl)phenyl]maleimide; and RH421) and the sensitivity of their fluorescence change to oligomycin and divalent cations (Ca2+ and Mn2+). The magnitude (% delta F) and rate constant (k(obs)) of ATP-induced fluorescence changes were measured on a fluorescence stopped-flow apparatus, and yielded the following results. (a) With RH421, k(obs) and % delta F varied with [Na+] (maximal k(obs) = 100 s-1, K1/2 = 6 mM; % delta Fmax = 6%, K1/2 = 1 mM); these values are comparable to those previously reported using IAF-labeled enzyme (Pratap, P.R., Robinson, J.D. and Steinberg, M.I. (1991) Biochim. Biophys. Acta 1069, 288-298). (b) With BIPM-labeled enzyme k(obs) did not vary with [Na+] over the range tested, and was twice as high as the maximum k(obs) for RH421. (c) Treatment with oligomycin reduced k(obs) for all three probes to about the same level (approximately 1-2 s-1) while % delta Fmax was largely unaffected. (d) Replacing Mg2+ with Ca2+ had similar effects to treatment with oligomycin. (e) RH421 fluorescence change was completely abolished in the presence of oligomycin and Ca2+. (f) Replacing Mg2+ with Mn2+ decreased IAF fluorescence, i.e., put the enzyme in an E2-like form, reduced k(obs), and rendered oligomycin less effective in reducing k(obs). From these results, we conclude: (a) the release of the second/third Na+ is the rate-limiting step for the conformational change measured by IAF and charge transfer measured with RH421; (b) BIPM indicates an earlier step, either the deocclusion of Na+ and/or the release of the first Na+; (c) oligomycin blocks Na+ deocclusion, and this step is sensitive to the divalent cation used to activate enzyme phosphorylation; and (d) Ca2+ slows the step reported by IAF as well. These experiments indicate that a simple model with two conformations (E1 and E2) is insufficient to explain transient kinetic data.

Proton (or hydroxide) fluxes and the biphasic osmotic response of human red blood cells.

Upon exposure of human red blood cells to hypertonic sucrose, the fluorescence of the potentiometric indicator 3,3'-dipropylthiadicarbocyanine iodide, denoted diS-C3(5), displays a biphasic time course indicating the rapid development of an inside-positive transmembrane voltage, followed by a slow DIDS (4,4'-diisothiocyano-2,2'-disulfonic acid stilbene)-sensitive decline of the voltage. In addition to monitoring membrane potential, proton (or hydroxide) fluxes were measured by a pH stat method, cell volume was monitored by light scattering, and cell electrolytes were measured directly when red cells were shrunken either with hypertonic NaCl or sucrose. Shrinkage by sucrose induced an initial proton efflux (or OH- influx) of 5.5 mu eq/g Hb.min and a Cl shift of 21-31 mu eq/g Hb in 15 min. Upon shrinkage with hypertonic NaCl, the cells are initially close to Donnan equilibrium and exhibit no detectable shift of Cl or protons. Experiments with the carbonic anhydrase inhibitor ethoxzolamide demonstrate that for red cell suspensions exposed to air and shrunken with sucrose, proton fluxes mediated by the Jacobs-Stewart cycle contribute to dissipation of the increased outward Cl concentration gradient. With maximally inhibitory concentrations of ethoxzolamide, a residual proton efflux of 2 mu eq/g Hb.min is insensitive to manipulation of the membrane potential with valinomycin, but is completely inhibited by DIDS. The ethoxzolamide-insensitive apparent proton efflux may be driven against the electrochemical gradient, and is thus consistent with HCl cotransport (or Cl/OH exchange). The data are consistent with predictions of equations describing nonideal osmotic and ionic equilibria of human red blood cells. Thus osmotic equilibration after shrinkage of human red blood cells by hypertonic sucrose occurs in two time-resolved steps: rapid equilibration of water followed by slower equilibration of chloride and protons (or hydroxide). Under our experimental conditions, about two-thirds of the osmotically induced apparent proton efflux is mediated by the Jacobs-Stewart cycle, with the remainder being consistent with mediation via DIDS-sensitive HCl cotransport (or Cl/OH exchange).

Effects of choline on Na(+)- and K(+)-interactions with the Na+/K(+)-ATPase.

Choline chloride, 100 mM, stimulates Na+/K(+)-ATPase activity of a purified dog kidney enzyme preparation when Na+ is suboptimal (9 mM Na+ and 10 mM K+) and inhibits when K+ is suboptimal (90 mM Na+ and 1 mM K+), but has a negligible effect at optimal concentrations of both (90 mM Na+ and 10 mM K+). Stimulation occurs at low Na+ to K+ ratios, but not at those same ratios when the actual Na+ concentration is high (90 mM). Stimulation decreases or disappears when incubation pH or temperature is increased or when Li+ is substituted for K+ or Rb+. Choline+ also reduces the Km for MgATP at the low ratio of Na+ to K+ but not at the optimal ratio. In the absence of K+, however, choline+ does not stimulate at low Na+ concentrations: either in the Na(+)-ATPase reaction or in the E1 to E2P conformational transition. Together, these observations indicate that choline+ accelerates the rate-limiting step in the Na+/K(+)-ATPase reaction cycle, K(+)-deocclusion; consequently, optimal Na+ concentrations reflect Na+ accelerating that step also. Thus, the observed K0.5 for Na+ includes high-affinity activation of enzyme phosphorylation and low-affinity acceleration of K(+)-deocclusion. Inhibition of Na+/K(+)-ATPase and K(+)-nitrophenylphosphatase reactions by choline+ increases as the K(+)-concentration is decreased; the competition between choline+ and K+ may represent a similar antagonism between conformations selected by choline+ and by K+.

The reaction sequence of the Na+/K(+)-ATPase: rapid kinetic measurements distinguish between alternative schemes.

Conformational changes between E1 and E2 enzyme forms of a dog kidney Na+/K(+)-ATPase preparation labeled with 5-iodoacetamidofluorescein were followed with a stopped-flow fluorimeter, in terms of the rate constant, kobs, and the steady-state magnitude, % delta F of fluorescence change. On rapid mixing of enzyme plus Mg2+ plus Na+ with saturating (0.5 mM) ATP in the absence of K+, kobs varied with Na+ concentration in the range 0-155 mM, with a K1/2 of 10 mM, while % delta F was relatively insensitive to Na+, with a K1/2 of 0.5 mM. Oligomycin reduced kobs by 98-99% for Na+ greater than or equal to 10 mM, but only by 50% for Na+ = 1 mM; % delta F was reduced at most by 20%. At 155 mM Na+, both kobs and % delta F changed if K+ was present with the enzyme. kobs decreased by 50% when K+ was increased from 0 to 0.2 mM, but increased when K+ was varied in the range 0.2-5 mM. K+ increased % delta F by a factor of 3 with a K1/2 of 0.3-0.5 mM as measured in both stopped-flow and steady-state experiments. These data are considered in terms of the derived presteady-state equations for two alternate schemes for the enzyme, with the E1P to E2P conformational change either preceding (Albers-Post) or following (Nørby-Yoda-Skou) Na+ transport and release. The analysis indicates that: (i) Na+ must be released before the conformational transition, from an E1 form; (ii) the step in which the second and/or third Na+ is released is rate-limiting, but this release is accelerated by Na+; and (iii) the release is also accelerated by K+ acting with low affinity (possibly at extracellular sites).

Na+/K(+)-ATPase: modes of inhibition by Mg2+.

Adding 15 mM free Mg2+ decreased Vmax of the Na+/K(+)-ATPase reaction. Mg2+ also decreased the K0.5 for K+ activation, as a mixed inhibitor, but the increased inhibition at higher K+ concentrations diminished as the Na+ concentration was raised. Inhibition was greater with Rb+ but less with Li+ when these cations substituted for K+ at pH 7.5, while at pH 8.5 inhibition was generally less and essentially the same with all three cations: implying an association between inhibition and ion occlusion. On the other hand, Mg2+ increased the K0.5 for Na(+)-activation of the Na+/K(+)-ATPase and Na(+)-ATPase reactions, as a mixed inhibitor. Changing incubation pH or temperature, or adding dimethylsulfoxide affected inhibition by Mg2+ and K0.5 for Na+ diversely. Presteady-state kinetic studies on enzyme phosphorylation, however, showed competition between Mg2+ and Na+. In the K(+)-phosphatase reaction catalyzed by this enzyme Mg2+ was a (near) competitor toward K+. Adding Na+ with K+ inhibited phosphatase activity, but under these conditions 15 mM Mg2+ stimulated rather than inhibited; still higher Mg2+ concentrations then inhibited with K+ plus Na+. Similar stimulation and inhibition occurred when Mn2+ was substituted for Mg2+, although the concentrations required were an order of magnitude less. In all these experiments no ionic substitutions were made to maintain ionic strength, since alternative cations, such as choline, produced various specific effects themselves. Kinetic analyses, in terms of product inhibition by Mg2+, require Mg2+ release at multiple steps. The data are accommodated by a scheme for the Na+/K(+)-ATPase with three alternative points for release: before MgATP binding, before K+ release and before Na+ binding. The latter alternatives necessitate two Mg2+ ions bound simultaneously to the enzyme, presumably to divalent cation-sites associated with the phosphate and the nucleotide domains of the active site.

Two mechanisms by which fluorescent oxonols indicate membrane potential in human red blood cells.

Optical potentiometric indicators have been used to monitor the transmembrane electrical potential (Em) of many cells and organelles. A better understanding of the mechanisms of dye response is needed for the design of dyes with improved responses and for unambiguous interpretation of experimental results. This paper describes the responses to delta Em of 20 impermeant oxonols in human red blood cells. Most of the oxonols interacted with valinomycin, but not with gramicidin. The fluorescence of 15 oxonols decreased with hyperpolarization, consistent with an "on-off" mechanism, whereas five oxonols unexpectedly showed potential-dependent increases in fluorescence at less than 2 microM [dye]. Binding curves were determined for two dyes (WW781, negative response and RGA451, positive response) at 1 mM [K]o (membrane hyperpolarized with gramicidin) and at 90 mM [K]o (delta Em = 0 with gramicidin). Both dyes showed potential-dependent decreases in binding. Changes in the fluorescence of cell suspensions correlated with changes in [dye]bound for WW781, in accordance with the "on-off" mechanism, but not for RGA451. Large positive fluorescence changes (greater than 30%) dependent on Em were observed between 0.1 and 1.0 microM RGA451. A model is suggested in which RGA451 moves between two states of different quantum efficiencies within the membrane.

System analysis of Phycomyces light-growth response: madC, madG, and madH mutants.

The light-growth response of Phycomyces has been studied further with the sum-of-sinusoids method in the framework of the Wiener theory of nonlinear system identification. The response was treated as a black box with the logarithm of light intensity as the input and elongation rate as the output. The nonlinear input-output relation of the light-growth response can be represented mathematically by a set of weighting functions called kernels, which appear in the Wiener intergral series. The linear (first-order) kernels of wild type, and of single and double mutants affected in genes madA to madG were determined previously with Gaussian white noise test stimuli, and were used to investigate the interactions among the products of these genes (R.C. Poe, P. Pratap, and E.D. Lipson. 1986. Biol. Cybern. 55:105.). We have used the more precise sum-of-sinusoids method to extend the interaction studies, including both the first- and second-order kernels. Specifically, we have investigated interactions of the madH ("hypertropic") gene product with the madC ("night blind") and madG ("stiff") gene products. Experiments were performed on the Phycomyces tracking machine. The log-mean intensity of the stimulus was 6 x 10(-2) W m-2 and the wavelength was 477 nm. The first- and second-order kernels were analyzed in terms of nonlinear kinetic models. The madH gene product was found to interact with those of madC and madG. This result extends previous findings that themadH gene product is associated with the input and the ouput of the sensory transduction complex for the lightgrowth response.

Impermeant potential-sensitive oxonol dyes: II. The dependence of the absorption signal on the length of alkyl substituents attached to the dye.

We have measured the potential-dependent light absorption changes of 43 impermeant oxonol dyes with an oxidized cholesterol bilayer lipid membrane system. The size of the signal is strongly dependent on the chain length of alkyl groups attached to the chromophore. Dye molecules with intermediate chain lengths give the largest signals. To better understand the dependence of the absorbance signal on alkyl chain length, a simple equilibrium thermodynamic analysis has been derived. The analysis uses the free energy of dye binding to the membrane and the "on-off" model (E.B. George et al., J. Membrane Biol., 103:245-253, 1988a) for the potential-sensing mechanism. In this model, a population of dye molecules in nonpolar membrane binding sites is in a potential-dependent equilibrium with a second population of dye that resides in an unstirred layer adjacent to the membrane. Dye in the unstirred layer is in a separate equilibrium with dye in the bulk bathing solution. The equilibrium binding theory predicts a "sigmoidally shaped" increase in signal with increasing alkyl chain length, even for very nonpolar dyes. We suggest that aggregation of the more hydrophobic dyes in the membrane bathing solution may be responsible for their low signals, which are not predicted by the theory.

Impermeant potential-sensitive oxonol dyes: III. The dependence of the absorption signal on membrane potential.

We have measured potential-dependent changes in the absorption of light by oxidized cholesterol bilayer lipid membranes in the presence of impermeant oxonol dyes. The magnitude of the absorption signal increased linearly with the size of potential steps over a range of 500 mV. The signal also increased when the offset voltage of the pulse train was increased from -150 to +150 mV. The data are consistent with the "on-off" mechanism proposed by E. B. George et al. (J. Membrane Biol. 103:245-253, 1988) in which the probe undergoes potential-dependent movement between a binding site in the membrane and an aqueous region just off the surface of the membrane. An equilibrium thermodynamic analysis of the experimental data indicates that the negatively charged oxonol chromophore senses only 5-10% of the total membrane potential difference across the membrane when it is driven into a nonpolar binding site on the membrane.

Impermeant potential-sensitive oxonol dyes: I. Evidence for an "on-off" mechanism.

This series of papers addresses the mechanism by which certain impermeant oxonol dyes respond to membrane-potential changes, denoted delta Em. Hemispherical oxidized cholesterol bilayer membranes provided a controlled model membrane system for determining the dependence of the light absorption signal from the dye on parameters such as the wavelength and polarization of the light illuminating the membrane, the structure of the dye, and delta Em. This paper is concerned with the determination and analysis of absorption spectral changes of the dye RGA461 during trains of step changes of Em. The wavelength dependence of the absorption signal is consistent with an "on-off" mechanism in which dye molecules are driven by potential changes between an aqueous region just off the membrane and a relatively nonpolar binding site on the membrane. Polarization data indicate that dye molecules in the membrane site tend to orient with the long axis of the chromophore perpendicular to the surface of the membrane. Experiments with hyperpolarized human red blood cells confirmed that the impermeant oxonols undergo a potential-dependent partition between the membrane and the bathing medium.