Search

The phosphorylation capacity of Na+,K+-ATPase preparations in common use is much less than expected on the basis of the molecular weight of the enzyme deduced from cDNA sequences. This has led to the popularity of half-of-the-sites or flip-flop models for the enzyme reaction mechanism. We have prepared Na+,K+-ATPase from nasal salt glands of salt-adapted ducks which has a phosphorylation capacity and specific activity near the theoretical maxima. Preparations with specific activities of >60 micromol (mg of protein)-1 min-1 at 37 degrees C had phosphorylation capacities of >60 nmol/mg of protein, and the rate of turnover of the enzyme was 9690 min-1, within the range reported for the enzyme from other sources. The fraction of the maximal specific activity of the enzyme compared well with the fraction of the protein on SDS-PAGE which was alpha and beta chains, especially at the highest specific activity which indicates that all of the alphabeta protomers are active. The gels of the most reactive preparations contained only alpha and beta chains, but less active preparations contained a number of extraneous proteins. The major contaminant was actin. The preparation did not contain any protein which migrated in the molecular weight range of the gamma subunit. The subunit composition of the enzyme was alpha1 and beta1 only. This is the first report of a pure, homogeneous, fully active preparation of the protein. Reaction models which incorporate a half-of-the-sites or flip-flop mechanism do not apply to this enzyme.

Cation binding to Na,K-ATPase is characterized in native membranes at room temperature by solid-state NMR spectroscopy using the K(+) congener (205)Tl. It has been demonstrated that the signals from occluded Tl(+) and nonspecifically bound Tl(+) can be detected and distinguished by NMR. Effects of dipole-dipole coupling between (1)H and (205)Tl in the occlusion sites show that the ions are rigidly bound, rather than just occluded. Furthermore, a low chemical shift suggests occlusion site geometries with a relatively small contribution from carboxylate and hydroxyl groups. Nonspecific binding of Tl(+) is characterized by rapid chemical exchange, in agreement with the observed low binding affinity.

Extensive trypsinization of Na,K-ATPase from the salt gland of Squalus acanthias removes about half of the extramembranous protein mass of the alpha-subunit, while leaving the beta-subunit intact. Sequence analysis and epitope recognition of the remaining alpha-peptides show that transmembrane segments M1/M2 and M3/M4 are present when trypsinization is performed in either NaCl or RbCl. The M5/M6 segment and the intact 19-kDa peptide (M7-M10) are detected in Rb-trypsinized membranes but not in Na-trypsinized membranes. The L7/L8 loop is associated with Na-trypsinized membranes, indicating the presence of an M7/M8 or M8/M9 fragment. Freeze-fracture electron microscopy of both Rb- and Na-trypsinized membranes reveals intramembranous particles that indicate a retained cluster of peptides, even in the absence of an intact 19-kDa fragment. The rotational diffusion of covalently spin-labeled trypsinized complexes is studied in the presence of poly(ethylene glycol) or glycerol by using saturation transfer electron spin resonance. Rotational correlation times in aqueous poly(ethylene glycol) are longer than in glycerol solutions of the same viscosity and increase nonlinearly with the viscosity of the suspending medium, indicating that poly(ethylene glycol) induces aggregation of the tryptic peptides (and beta-subunit) within the membrane. The aggregates of enzyme trypsinized in the presence of NaCl are larger than those for enzyme trypsinized in RbCl, at both low and high aqueous viscosities. Similarities in mobility for native and Rb-trypsinized enzymes suggest either a change in average orientation of the spin-label upon trypsinization or that trypsinization leads to a reorganized protein structure that is more prone to aggregation.

The effects of phospholipid acyl chain length (n(c)) and cholesterol on several partial reactions of Na,K-ATPase reconstituted into liposomes of defined lipid composition are described. This regards the E(1)/E(2) equilibrium, the phosphoenzyme level, and the K(+)-deocclusion reaction. In addition, the lipid effects on some steady-state properties were investigated. Finally, the effects of cholesterol on the temperature sensitivity of the phosphorylation and spontaneous dephosphorylation reactions were investigated. The fatty acid and cholesterol composition of the native Na,K-ATPase membrane preparation showed a remarkable similarity to the lipid composition known to support maximum hydrolytic capacity as determined from in vitro experiments. The main rate-determining step of the Na,K-ATPase reaction, the E(2) --> E(1) reaction, as well as several other partial reactions were accelerated by cholesterol. This regards the phosphorylation by ATP as well as the E(1) - P --> E(2)-P reaction. Moreover, cholesterol shifted the E(1)/E(2) equilibrium toward the E(1) conformation and increased the K(+)-deocclusion rate. Finally, cholesterol significantly affected the temperature sensitivity of the spontaneous dephosphorylation reaction and the phosphorylation by ATP. The effects of cholesterol were not completely equivalent to those induced by increasing the phospholipid acyl chain length, indicating that the cholesterol effects are not entirely caused by increasing the hydrophobic bilayer thickness, which indicates an additional mechanism of action on the Na,K-ATPase.

Removal of extramembranous portions of the integral membrane protein Na,K-ATPase from shark salt glands by trypsin in the presence of Rb+ (a K+ congener) preserves the intramembranous association of the remaining membrane-spanning tryptic peptides. This is evidenced from comparison of the rotational mobility of native and trypsinized Na,K-ATPase using saturation transfer electron spin resonance spectroscopy (ESR) and from study of the lipid-protein interactions using conventional ESR spectroscopy. The interface between the lipids and the intramembranous domains is conserved on removal of the extramembranous parts of the protein, since the population of motionally restricted boundary lipids remains essentially the same in the native and trypsinized preparations. The ability to occlude Rb+ is also retained by the trypsinized membranes, as previously observed with pig kidney Na,K-ATPase. A 19-kDa fragment remaining when Na,K-ATPase is trypsinized in the presence of Rb+ is degraded further when the trypsinization is carried out in the presence of Na+ instead of Rb+. The rotational mobility of the tryptic fragments in the Na(+)-trypsinized membranes is lower than for the Rb(+)-trypsinized membranes, indicating rearrangement of the peptides. In addition, occlusion capacity is lost when trypsinization is carried out in Na+, suggesting a correlation between structure and function in the trypsinized membranes. The sequences of four membrane-spanning tryptic fragments of shark Na,K-ATPase are found to be almost identical to corresponding sequences in pig kidney Na,K-ATPase.

Prior studies identified phosphoenzyme intermediates in the turnover of sodium- and potassium-activated adenosinetriphosphatase [(Na,K)ATPase] from several sources and of the calcium-activated adenosinetriphosphatase [(Ca)-ATPase] of skeletal muscle sarcoplasmic reticulum. In both cases, the transphosphorylation is to a beta-aspartyl carboxyl group at the active site. We now report observation of a K+-sensitive phosphorylated intermediate of purified (Na,-K)ATPase from the salt gland of the duck using high-field 31P nuclear magnetic resonance. Addition of ATP to a suspension of this enzyme in the presence of Mg2+ and Na+ produced a resonance at about +17 ppm relative to 85% phosphoric acid. Addition of inorganic phosphate and Mg2+ to (Na,K)ATPase also produced a resonance at about +17 ppm which was enhanced in the presence of a saturating concentration of the inhibitor, ouabain; again, addition of K+ made this resonance disappear. These findings are consistent with earlier kinetic characterization of an acid-stable (Na,K)ATPase phosphoenzyme intermediate by 32P-labeled phosphate incorporation into a denatured precipitate of the enzyme. We attribute the +17-ppm resonance to formation of an acyl phosphate at an aspartyl residue of the catalytic site of (Na,K)ATPase. This is supported by our finding of a similar resonance at +17 ppm after phosphorylation of another membrane-bound cation transport enzyme, sarcoplasmic reticulum (Ca)ATPase, as well as by a similar resonance at about +17 ppm after phosphorylation of the model dipeptide L-seryl-L-aspartate.

Lipid-protein interactions in (Na+,K+)-ATPase-rich membranes from the rectal gland of Squalus acanthias have been studied by using spin-labeled lipids in conjunction with electron spin resonance (ESR) spectroscopy. Lipid-protein associations are revealed by the presence of a second component in the ESR spectra of the membranes in addition to a component which corresponds very closely to the ESR spectra obtained from dispersions of the extracted membrane lipids. This second component corresponds to spin-labeled lipids whose motion is very significantly restricted relative to that of the fluid lipids in the membrane or the lipid extract. A stoichiometry of approximately 66 lipids per 265 000-dalton protein is found for the motionally restricted component of those spin-labeled lipids (e.g., phosphatidylcholine) which show least specificity for the protein. This corresponds approximately to the number of lipids which may be accommodated within the first shell around the alpha 2 beta 2 protein dimer. A selectivity of the various spin-labeled lipids for the motionally restricted component associated with the protein is found in the following order: cardiolipin greater than phosphatidylserine approximately stearic acid greater than or equal to phosphatidic acid greater than phosphatidylglycerol approximately phosphatidylcholine approximately phosphatidylethanolamine approximately androstanol.

The (Na+,k+)ATPase from the rectal gland of Carcharhinus obscurus has been solubilized in Lubrol WX as an active complex containing 379,900 g of protein and 61 mol of phospholipid. This detergent-lipid-protein complex contains two catalytic subunits of molecular weight 106,400 and four glycopeptide subunits of protein molecular weight 36,600. The latter subunit has a total molecular weight of 51,700 when the carbohydrate is included. Attempts to dissociate this active enzyme complex to smaller size by increasing the detergent concentration led to inactivation. Thus, the smallest active particle in the presence of Lubrol WX contains the two polypeptide subunits in a mole ratio of 2:4 under conditions where the micellar form of detergent is present at a 70:1 molar ratio. This large excess of Lubrol WX eliminates any possibility of artificial togetherness as the result of statistical considerations.

The pH dependence and salt dependence of the lipid-protein interactions of phosphatidic acid, phosphatidylserine, and stearic acid with Na+,K+-ATPase membranes from Squalus acanthias have been studied with spin-label electron spin resonance spectroscopy, using lipids with nitroxide labels on the 14-position C atom of the sn-2 chain. For phosphatidic acid and stearic acid, the fraction of motionally restricted spin-label increases with increasing pH, with pKa's of 6.6 and 8.0, respectively. In contrast, the pKa of stearic acid in the bulk lipid environment of the membrane is estimated from spin-label spectroscopy to be approximately equal to 6.6. The fraction of motionally restricted phosphatidylserine spin-label remains constant over the pH range 4.7-9.2. In the fully dissociated state the fractions of motionally restricted spin-labeled phosphatidic and stearic acids decrease with increasing salt concentration, reaching an approximately constant value at [NaCl] = 0.5-1.0 M. For stearic acid the net decrease is comparable to that obtained on protonation, but for phosphatidic acid the decrease is considerably smaller (by approximately 55%) than that obtained on protonating the lipid. The fraction of motionally restricted phosphatidylserine spin-label varies relatively little with salt concentration up to 1 M NaCl. Direct electrostatic effects alone cannot account for the whole of the observed specificity of interaction of the two phospholipids with Na+,K+-ATPase membranes.