Three-Dimensional Electrical Control of the Excitonic Fine Structure for a Quantum Dot in a Cavity.

The excitonic fine structure plays a key role for the quantum light generated by semiconductor quantum dots, both for entangled photon pairs and single photons. Controlling the excitonic fine structure has been demonstrated using electric, magnetic, or strain fields, but not for quantum dots in optical cavities, a key requirement to obtain high source efficiency and near-unity photon indistinguishability. Here, we demonstrate the control of the fine structure splitting for quantum dots embedded in micropillar cavities. We propose and implement a scheme based on remote electrical contacts connected to the pillar cavity through narrow ridges. Numerical simulations show that such a geometry allows for a three-dimensional control of the electrical field. We experimentally demonstrate tuning and reproducible canceling of the fine structure, a crucial step for the reproducibility of quantum light source technology.

Fiber-based angular filtering for high-resolution Brillouin spectroscopy in the 20-300†GHz frequency range.

Brillouin spectroscopy emerges as a promising non-invasive tool for nanoscale imaging and sensing. One-dimensional semiconductor superlattice structures are eminently used for selectively enhancing the generation or detection of phonons at few GHz. While commercially available Brillouin spectrometers provide high-resolution spectra, they consist of complex experimental techniques and are not suitable for semiconductor cavities operating at a wide range of optical wavelengths. We develop a pragmatic experimental approach for conventional Brillouin spectroscopy that can integrate a widely tunable excitation-source. Our setup combines a fibered-based angular filtering and a spectral filtering based on a rotating single etalon and a double grating spectrometer for sequential reconstruction of Brillouin spectra. This configuration allows probing confined acoustic phonon modes in the 20-300 GHz frequency range with excellent laser rejection and high spectral resolution. Remarkably, our scheme based on the excitation and collection of the enhanced Brillouin scattering signals through the optical cavity allows for better angular filtering with decreasing phonon frequency. It can be implemented for the study of cavity optomechanics and stimulated Brillouin scattering over broadband optical and acoustic frequency ranges.

Optomechanical properties of GaAs/AlAs micropillar resonators operating in the 18 GHz range.

Recent experiments demonstrated that GaAs/AlAs based micropillar cavities are promising systems for quantum optomechanics, allowing the simultaneous three-dimensional confinement of near-infrared photons and acoustic phonons in the 18-100 GHz range. Here, we investigate through numerical simulations the optomechanical properties of this new platform. We evidence how the Poisson's ratio and semiconductor/vacuum boundary conditions lead to very distinct features in the mechanical and optical three-dimensional confinement. We find a strong dependence of the mechanical quality factor and strain distribution on the micropillar radius, in great contrast to what is predicted and observed in the optical domain. The derived optomechanical coupling constants g0 reach ultra-large values in the 106 rad/s range.

Probing gigahertz coherent acoustic phonons in TiO2 mesoporous thin films.

Ultrahigh-frequency acoustic-phonon resonators usually require atomically flat interfaces to avoid phonon scattering and dephasing, leading to expensive fabrication processes, such as molecular beam epitaxy. Mesoporous thin films are based on inexpensive wet chemical fabrication techniques that lead to relatively flat interfaces regardless the presence of nanopores. Here, we report mesoporous titanium dioxide-based acoustic resonators with resonances up to 90 GHz, and quality factors from 3 to 7. Numerical simulations show a good agreement with the picosecond ultrasonics experiments. We also numerically study the effect of changes in the speed of sound on the performance of the resonator. This change could be induced by liquid infiltration into the mesopores. Our findings constitute the first step towards the engineering of building blocks based on mesoporous thin films for reconfigurable optoacoustic sensors.

Sequential generation of linear cluster states from a single photon emitter.

Light states composed of multiple entangled photons-such as cluster states-are essential for developing and scaling-up quantum computing networks. Photonic cluster states can be obtained from single-photon sources and entangling gates, but so far this has only been done with probabilistic sources constrained to intrinsically low efficiencies, and an increasing hardware overhead. Here, we report the resource-efficient generation of polarization-encoded, individually-addressable photons in linear cluster states occupying a single spatial mode. We employ a single entangling-gate in a fiber loop configuration to sequentially entangle an ever-growing stream of photons originating from the currently most efficient single-photon source technology-a semiconductor quantum dot. With this apparatus, we demonstrate the generation of linear cluster states up to four photons in a single-mode fiber. The reported architecture can be programmed for linear-cluster states of any number of photons, that are required for photonic one-way quantum computing schemes.

Conformational heterogeneity and spin-labeled -SH groups: pulsed EPR of Na,K-ATPase.

Membranous Na,K-ATPase from shark salt gland and from pig kidney was spin-labeled on class I -SH groups in the presence of glycerol, or on class II -SH groups in the absence of glycerol. The class I-labeled preparations retain full enzymatic activity, whereas the class II-labeled preparations are at least partially inactivated. This provides an excellent testbed on which to demonstrate how advanced electron paramagnetic resonance (EPR) can provide novel information on specific residues in unique environments in a complex, membrane-bound transport system. The polarity of the environment, and the librational dynamics and conformational exchange, of the spin-labeled groups were studied with pulsed EPR by using electron spin echo envelope modulation (ESEEM) spectroscopy and spin-echo detected (ED) EPR spectroscopy, respectively. 2H-ESEEM spectra of membranes dispersed in D2O reveal that class I groups of the shark enzyme are more exposed to water than are those of the pig enzyme or class II groups of either species, consistent with the more superficial membrane location in the former case. Spin-echo decay curves indicate conformational heterogeneity at low temperatures (<150 K), but a more homogeneous conformational state at higher temperatures that is characterized by a single phase-memory T2M relaxation time. Conventional EPR lineshapes also demonstrate conformational microheterogeneity at low temperatures: the inhomogeneously broadened lines narrow progressively with increasing temperature reaching an almost pure Lorentzian line shape at temperatures of ca. 220 K and above. The inhomogeneous broadening at low temperature is well described by a Gaussian distribution of Lorentzian lines. ED spectra as a function of echo-delay time demonstrate the onset of rapid librational motions of appreciable amplitude, and slower conformational exchange, at temperatures above 220 K. These motions could drive transitions between the different conformational substates, which are frozen in at lower temperatures but contribute to the pathways between the principal enzymatic intermediates at higher temperatures.

Precipitation of solubilized Na+/K+-ATPase by divalent cations.

A method for preparation of membranous fragments of pure and highly active shark rectal gland Na+/K+-ATPase by Mn2+ precipitation of C12E8-solubilized enzyme is described. The method is rapid and inexpensive, and yields enzyme with a specific Na+/K+-ATPase activity of up to 1800 mumol/mg per h at 37 degrees C. The influence of the detergent/protein and lipid/protein ratios on the yield of membrane bound enzyme is described.

Sulphydryl groups of (Na+ + K+)-ATPase from rectal glands of Squalus acanthias. Detection of ligand-induced conformational changes.

1. Modification of the Class II sulphydryl groups on the (Na+ + K+)-ATPase from rectal glands of Squalus acanthias with N-ethylmaleimide has been used to detect conformational changes in the protein. The rates of inactivation of the enzyme and the incorporation of N-ethylmaleimide depend on the ligands present in the incubation medium. With 150 mM K+ the rate of inactivation is largest (k1 = 1.73 mM-1 . min-1) and four SH groups per alpha-subunit are modified. The rate of inactivation in the presence of 150 mM Na+ is smaller (k1 = 1.08 mM-1 . min-1) but the incorporation of N-ethylmaleimide is the same as with K+. 2. ATP in micromolar concentrations protects the Class II groups in the presence of Na+ (k1 = 0.08 mM-1 . min-1 at saturating ATP) and the incorporation is drastically reduced. ATP in millimolar concentrations protects the Class II groups partially in the presence of K+ (k1 = 1.08 mM-1 . min-1) and three SH groups are labelled per alpha subunit. 3. The K+-dependent phosphatase is inhibited in parallel to the (Na+ + K+)-ATPase under all conditions, and the ligand-dependent incorporation of N-ethylmaleimide was on the alpha-subunit only. 4. It is shown that the difference between the Na+ and K+ conformations sensed with N-ethylmaleimide depends on the pH of the incubation medium. At pH 6 there is a very small difference between the rates of inactivation in the presence of Na+ and K+, but at higher pH the difference increases. It is also shown that the rate of inactivation has a minimum at pH 6.9, which suggests that the conformation of the enzyme changes with pH. 5. Modification of the Class III groups with N-ethylmaleimide--whereby the enzyme activity is reduced from about 16% to zero--shows that these groups are also sensitive to conformational changes. As with the Class II groups, ATP in micromolar concentrations protects in the presence of Na+ relative to Na+ or K+ alone. ATP in millimolar concentrations with K+ present increases the rate of inactivation relative to K+ alone, in contrast to the effect on the Class II groups. 6. Modification of the Class II groups with a maleimide spin label shows a difference between Class II groups labelled in the presence of Na+ (or K+) and Class II groups labelled in the presence of K + ATP, in agreement with the difference in incorporation of N-ethylmaleimide. The spectra suggest that the SH group protected by ATP in the presence of K+ is buried in the protein. 7. The results suggest that at least four different conformations of the (Na+ + K+)-ATPase can be sensed with N-ethylmaleimide: (i) a Na+ form of the enzyme with ATP bound to a high-affinity site (E1-Na-ATP); (ii) a Na+ form without ATP bound (E1-Na); (iii) a K+ form without ATP bound (E2-K); and (iv) an enzyme form with ATP bound to a low-affinity site in the presence of K+, probably and E1-K-ATP form.

Sulphydryl groups of (Na+ + K+)-ATPase from rectal glands of Squalus acanthias. Titrations and classification.

1. (Na+ + K+)-ATPase from rectal glands of Squalus acanthias contains 34 SH groups per mol (Mr 265000). 15 are located on the alpha subunit (Mr 106000) and two on the beta subunit (Mr 40000). The beta subunit also contains one disulphide bridge. 2. The reaction of (Na+ + K+)-ATPase with N-ethylmaleimide shows the existence of at least three classes of SH groups. Class I contains two SH groups on each alpha subunit and one on each beta subunit. Reaction of these groups with N-ethylmaleimide in the presence of 40% glycerol or sucrose does not alter the enzyme activity. Class II contains four SH groups on each alpha subunit, and the reaction of these groups with 0.1 mM N-ethylmaleimide in the presence of 150 mM K+ leads to an enzyme species with about 16% activity. The remaining enzyme activity can be completely abolished by reaction with 5-10 mM N-ethylmaleimide, indicating a third class of SH groups (Class III). This pattern of inactivation is different from that of the kidney enzyme, where only one class of SH groups essential to activity is observed. 3. It is also shown that N-ethylmaleimide and DTNB inactivate by reacting with the same Class II SH groups. 4. Spin-labelling of the (Na+ + K+)-ATPase with a maleimide derivative shows that Class II groups are mostly buried in the membrane, whereas Class I groups are more exposed. It is also shown that spin label bound to the Class I groups can monitor the difference between the Na+- and K+-forms of the enzyme.

Occlusion of Rb+ by detergent-solubilized (Na+ + K+)-ATPase from shark salt glands.

Occlusion of Rb+ by C12E8-solubilized (Na+ + K+)-ATPase from shark salt glands has been measured. The rate of de-occlusion at room temperature is about 1 s-1, which is the same as for the membrane-bound enzyme. The amount of Rb+ occluded is 3 moles Rb+ per mole membrane-bound shark enzyme, whereas only about 2 moles Rb+ are occluded by the C12E8-solubilized enzyme.

Oligomycin interaction with Na,K-ATPase: oligomycin binding and dissociation are slow processes.

Oligomycin interacts with the Na,K-ATPase by increasing the apparent Na+ affinity in the non-phosphorylated state of the enzyme. This property is used to estimate rate constants attributed to oligomycin binding and dissociation reactions with Na,K-ATPase. The rate constants are determined indirectly, employing stop-flow fluorimetry of eosin, the fluorescence of which is a marker for the E1 state of the enzyme, i.e. for Na+ binding. The second-order rate constants derived for oligomycin binding are in the range (6-12).10(4) M-1 s-1 at 6 degrees C for both shark rectal gland and pig kidney enzyme. Rate constants for dissociation of the enzyme-oligomycin complex are about 0.05 s-1 at 6 degrees C. The slow rates of binding and dissociation suggest that oligomycin acts from within the membrane lipid phase rather than from the aqueous phase. The dissociation constant at 6 degrees C for the enzyme-oligomycin complex can be calculated to be about 1 microM for shark enzyme and about 2 microM for kidney enzyme, at pH 7.0 in 2 mM NaCl.

Determination of rate constants for nucleotide dissociation from Na,K-ATPase.

A method for determining individual rate constants for nucleotide binding to and dissociation from membrane bound pig kidney Na,K-ATPase is presented. The method involves determination of the rate of relaxation when Na,K-ATPase in the presence of eosin is mixed with ADP or ATP in a stopped-flow fluorescence apparatus. It is shown that the nucleotide dependence of this rate of relaxation--taken together with measured equilibrium binding values for eosin and ADP--makes possible a reasonably reliable determination of the rate constant for dissociation of nucleotide, i.e., determination of the rate constant k-1 in the following model (where E denotes Na,K-ATPase): [formula: see text] All experiments are carried out at about 4 degrees C in a buffer containing 200 mM sucrose, 10 mM EDTA, 25 mM Tris and 73 mM NaCl (pH 7.4). Values obtained for the rate constants for dissociation are about 6 s-1 for ADP and 2-3 s-1 for ATP.

The distribution of C12E8-solubilized oligomers of the (Na+ + K+)-ATPase.

Gel filtration of (Na+ + K+)-ATPase (ATP phosphohydrolase, EC 3.6.1.8) solubilized in octaethyleneglycol dodecylmonother ( C12E8 ) has been performed under conditions where active (alpha beta)2 dimers (Mr 265000) are obtained, and under conditions where dissociation into alpha beta monomers occurs without appreciable loss of activity. It is shown that the alpha beta monomers aggregate with time to form (alpha beta)2 dimers at low detergent concentrations with no change in enzymatic activity. At high detergent concentrations the aggregation is much slower, but the enzymatic activity is lost rapidly. Polyacrylamide gel electrophoresis in the presence of C12E8 also suggest that high concentrations of detergent dissociate the (alpha beta)2 dimer into smaller particles, and conditions for gel electrophoresis are described. The inactivating effect of C12E8 at high C12E8 /protein ratios can be related to a delipidation of the enzyme, with about 0.19 mg phospholipid required per mg protein for optimal activity. The experiments suggest that the solubilized (Na+ + K+)-ATPase can be disrupted into particles containing only one alpha-chain and one or two beta-chains without irreversible loss of activity, and that the stable form of the enzyme is an (alpha beta)2 dimer.

Solubilized (Na+ + K+)-ATPase from shark rectal gland and ox kidney--an inactivation study.

The bi-exponential time-course of detergent inactivation at 37 degrees C of C12E8-solubilized (Na+ + K+)-ATPase from shark rectal glands and ox kidney was investigated. The data for shark enzyme, obtained at detergent/protein weight ratios between 2 and 16, are interpreted in terms of a simple model where the membrane bound enzyme is solubilized predominantly as (alpha-beta)2 diprotomers at low detergent concentrations and as alpha-beta protomers at high C12E8 (octaethyleneglycoldodecylmonoether) concentrations. It is observed that the protomers are inactivated 15-fold more rapidly than the diprotomers, and that the rate of inactivation of both oligomers is proportional to the detergent/protein ratio. Inactivation of kidney enzyme was biexponential with a very rapid inactivation of up to 40% of the enzyme activity. The observed rate of inactivation of the slower phase varied with the detergent/protein ratio, but the inactivation pattern for the kidney enzyme could not readily be accommodated within the model for inactivation of the shark enzyme. The rates of inactivation at 37 degrees C were about the same in KCl and NaCl, i.e., in the E2(K) and E1 X Na forms, for both enzymes.

Properties of oligomycin-induced occlusion of Na+ by detergent-solubilized Na,K-ATPase from pig kidney or shark rectal gland.

Oligomycin induces occlusion of Na+ in membrane-bound Na,K-ATPase. Here it is shown that Na,K-ATPase from pig kidney or shark rectal gland solubilized in the nonionic detergent C12E8 is capable of occluding Na+ in the presence of oligomycin. The apparent affinity for Na+ is reduced for both enzymes upon solubilization, and there is an increase in the sigmoidicity of binding curves, which indicates a change in the cooperativity between the occluded ions. A high detergent/protein ratio leads to a decreased occlusion capacity. De-occlusion of Na+ by addition of K+ is slow for solubilized Na,K-ATPase, with a rate constant of about 0.1 s-1 at 6 degrees C. Stopped-flow fluorescence experiments with 6-carboxyeosin, which can be used to monitor the E1Na-form in detergent solution, show that the K(+)-induced de-occlusion of Na+ correlates well with the fluorescence decrease which follows the transition from the E1Na-form to the E2-form. There is a marked increase in the rate of fluorescence change at high detergent/protein ratios, indicating that the properties of solubilized enzyme are subject to modification by detergent in other respects than mere solubilization of the membrane-bound enzyme. The temperature dependence of the rate of de-occlusion in the range 2 degrees C to 12 degrees C is changed slightly upon solubilization, with activation energies in the range 20-23 kcal/mol for membrane-bound enzyme, increasing to 26-30 kcal/mol for solubilized enzyme. Titrations of the rate of transition from E1Na to E2K with oligomycin can be interpreted in a model with oligomycin having an apparent dissociation constant of about 2.5 microM for C12E8-solubilized shark Na,K-ATPase and 0.2 microM for solubilized pig kidney Na,K-ATPase.

Temperature-dependencies of various catalytic activities of membrane-bound Na+/K+-ATPase from ox brain, ox kidney and shark rectal gland and of C12E8-solubilized shark Na+/K+-ATPase.

The temperature dependence of ouabain-sensitive ATPase and phosphatase activities of membrane fragments containing the Na+/K+-ATPase were investigated in tissue from ox kidney, ox brain and from shark rectal glands. The shark enzyme was also tested in solubilized form. Arrhenius plots of the Na+/K+-ATPase activity seem to be linear up to about 20 degrees C, and non-linear above this temperature. The Arrhenius plots of mammalian enzyme (ox brain and kidney) were steeper, especially at temperatures below 20-30 degrees C, than that of shark enzyme. The Na+-ATPase activity showed a weaker temperature-dependence than the Na+/K+-ATPase activity. The phosphatase reactions measured, K+-stimulated, Na+/K+-stimulated and Na+/K+/ATP-stimulated, also showed a weaker temperature-dependence than the overall Na+/K+-ATPase activity. Among the phosphatase reactions, the largest change in slope of the Arrhenius plot was observed with the Na+/K+/ATP)-stimulated phosphatase reaction. The Arrhenius plots of the partial reactions were all non-linear. Solubilization of shark enzyme in C12E8 did not change the curvature of Arrhenius plots of the Na+/K+-ATPase activity or the K+-phosphatase activity. Since solubilization involves a disruption of the membrane and an 80% delipidation, the observed curvature of the Arrhenius plot can not be attributed to a property of the membrane as such.

The effects of Na+ and K+ on the conformational transitions of (Na+ + K+)-ATPase.

The equilibrium between the E1 form and the E2 form of the (Na+ + K+)-ATPase is poised towards E2 in the absence of cations when monitored by the fluorescence of eosin. Na+ converts the enzyme to the E1 form with a K0.5 of 1.9 mM (pH 7.4 and 2 microM eosin). The titration curves indicate that more than one Na+ is necessary. K0.5 for K+ for reversal of the Na+ effect is 0.8 mM with 10 mM Na+, 5.8 mM K+ with 30 mM Na+ and 37.7 mM with 100 mM Na+. In the absence of K+ the rate of transition from E2 to E1 is rapid when Na+ is added, t1/2 is about 50 ms at 6 degrees C (pH 7.4). K+ in low concentrations decreases the rate with a K0.5 for K+ of about 10 microM, and t1/2 is about 2 s at 1 mM K+. K+ in concentrations from 1 to 50 mM (the highest tested) increases the rate of transition from E2 to E1. E2 thus consists of a non-K+ form (E2), a K+ both with K+ bound with apparent high affinity, the 'occluded' form (KoccE2), and a form with K+ bound both at high and low affinity sites (KE2). K+ bound to the low-affinity sites facilitates the transition from KoocE2 to E1. So does an increase in the Na+ concentration, but Na+ is less effective than K+. The rate of transition from E1 to E2 is rapid when K+ is added to enzyme in 20 mM Na+. The rate increases with the K+ concentration, and the effect of K+ does not saturate at the concentrations of K+ sufficient to displace the equilibrium fully to E2. Na+ decreases the rate of transition from E1 to E2. A model is suggested for the effect of the cations on the transition between the different forms of the enzyme. The physiological implications of the effect of K+ at a low-affinity site are discussed.

Effect of magnesium ions on the high-affinity binding of eosin to the (Na+ + K+)-ATPase.

(1) The fluorescence of eosin Y in the presence of (Na+ + K+)-ATPase is enhanced by Mg2+. The enhancement by Mg2+ is larger than that obtained with Na+ (Skou, J.C. and Esmann, M. (1981) Biochim. Biophys. Acta 647, 232-240). Mg2+ shifts the excitation maximum from 518 to 524 nm, the emission maximum from 538 to 542 nm. Also a shoulder appears at about 490 nm on the excitation curve, as was also observed with Na+. (2) The Mg2+-dependent enhancement of fluorescence can be reversed by K+ as well as by ATP. In the presence of Mg2+ + Pi (i.e. under conditions of phosphorylation), the fluorescence enhancement can be reversed by ouabain. With Mg2+ and a low concentration of K+ (i.e. conditions for vanadate binding), the enhancement of fluorescence can be reversed by vanadate. (3) There is a low-affinity binding of eosin which increases with the Mg2+ concentration. This is observed as a slight increase in the fluorescence when the excitation wavelength is above 520 nm. The low-affinity binding is K+-, ATP-, ouabain- and vanadate-insensitive. (4) Scatchard analysis of the binding experiments suggests that there are two high-affinity eosin-binding sites per 32P-labelling site in the presence of 5 mM Mg2+ both of which are ouabain-, vanadate- and ATP-sensitive. With 5 mM Mg2+ + 0.25 Pi, the Kd values are 0.14 microM and 1.3 microM, respectively. With 5 mM Mg2+, 150 mM Na+, the Kd values are 0.45 microM and 3.2 microM, respectively. With 5 mM Mg2+, the addition of K+ gives a pronounced decrease in affinity but does not decrease the number of binding sites (which remains at two per 32P-labelling site). With 5 mM Mg2+ + 150 mM K+, the affinities of the two binding sites become identical, at a Kd of 17 microM. (5) The rate of conformational transitions was measured using the stopped-flow method. The rate of the transition from the Mg2+-form to the K+-form is high. Oligomycin has only a small (if any) effect on the rate. Addition of Na+ in the presence of Mg2+ does not appreciably change the rate of conversion to the K+-form, giving a rate constant of about 110 s-1. However, the addition of oligomycin in the presence of Mg2+ + Na+ had a profound effect: the rate of conversion to the K+-form was decreased by a factor of 2000 to about 0.063 s-1. This suggests that the conformation with Mg2+ alone is different from the conformation with Na+ alone. (6) The effects of K+, ouabain, vanadate and ATP on the high-affinity binding of eosin suggest that the two eosin molecules bound per 32P-labelling site are bound to ATP sites.

The effect of K+ on the equilibrium between the E2 and the K+-occluded E2 conformation of the (Na+ + K+)-ATPase.

The rate of the transition from the E2 form to the E1 form of (Na+ + K+)-ATPase (ATP phosphohydrolase, EC 3.6.1.3) has been monitored by the fluorescence changes of eosin. The equilibrium between E1 and E2 is poised towards E2 in the absence of added cations. A stopped-flow tracing of the transition from E2 in the presence of 2 microM K+ (contamination) to E1 (in 150 mM Na+) is multiexponential with a large, rapidly decaying component (t 1/2 about 50 ms) and a smaller component which has a t 1/2 of about 2 s. KCl in microM concentrations decreases the amplitude of the rapidly decaying component and increases the amplitude of the slow component. The stopped-flow tracings can be satisfactorily fitted by a sum of three exponentials. An apparent Kd for K+ of about 5 microM is obtained for the conversion of the rapidly decaying component to the slowly decaying component. The experiments show that the E2 form is a mixture of at least two enzyme conformations. One E2 conformation - without K+ bound, (E2) - is transferred rapidly to the E1 conformation when Na+ is added, whereas the other E2-conformation--with K+ bound with an apparent high affinity, Kocc E2--is transferred slowly to the E1 conformation.

Selectivity of lipid-protein interactions with trypsinized Na, K-ATPase studied by spin-label EPR.

The selectivity of the lipid-protein interactions in trypsinised Na, K-ATPase membranes from Squalus acanthias has been determined by using EPR spectroscopy with different lipid probes spin-labelled on the 14-C atom of the fatty acid chain. From measurements at low ionic strength and different pH values, the pattern of selectivity is: (stearic acid)->(phosphatidylserine)->(stearic acid)0>(phosphatidylcholine)+/-, where superscripts indicate the formal electrostatic charge on the lipid headgroup. This is in the same order as that determined with native Na,K-ATPase membranes [M. Esmann, D. Marsh, Biochemistry 24 (1985) 3572-3578]. The selectivity for phosphatidylserine is independent of pH, over the range pH 6.0-9. 0, as found also for native membranes. For membranes trypsinised in the presence of Rb+ ions, and in the presence of Na+ (which allows more extensive proteolysis), the relative association constants, Kr, of all lipids are the same as for control membranes, with the exception of ionised (stearic acid)- that shows the highest specificity. Therefore, both the stoichiometry and the principal determinants of the specificity of lipid-protein interaction are preserved on extensive trypsinisation of Na,K-ATPase membranes. This has implications for the location and arrangement of those amino acid side chains that determine the lipid selectivity of the native Na,K-ATPase.