Na+,K+-ATPase actively extrudes three cytoplasmic Na+ ions in exchange for two extracellular K+ ions for each ATP hydrolyzed. The atomic structure with bound Na+ identifies three Na+ sites, named I, II, and III. It has been proposed that site III is the first to be occupied and site II last, when Na+ binds from the cytoplasmic side. It is usually assumed that the occupation of all three Na+ sites is obligatory for the activation of phosphoryl transfer from ATP. To obtain more insight into the individual roles of the ion-binding sites, we have analyzed a series of seven mutants with substitution of the critical ion-binding residue Ser777, which is a shared ligand between Na+ sites I and III. Surprisingly, mutants with large and bulky substituents expected to prevent or profoundly disturb Na+ access to sites I and III retain the ability to form a phosphoenzyme from ATP, even with increased apparent Na+ affinity. This indicates that Na+ binding solely at site II is sufficient to promote phosphorylation. These mutations appear to lock the membrane sector into an E1-like configuration, allowing Na+ but not K+ to bind at site II, while the cytoplasmic sector undergoes conformational changes uncoupled from the membrane sector.
Various dominant disease phenotypes are caused by germline variants of ATP1A1, the gene for the ubiquitously expressed Na+,K+-ATPase (NKA) α1 subunit. What characteristics link specific variants to a particular disease phenotype remains unclear. We used heterologous expression in mammalian cells and Xenopus oocytes to evaluate the characteristics of α1 variants causing hypomagnesemia with treatment-resistant seizures: L302R, G303R, M859R, and W931R. Cell survival in the presence of micromolar ouabain was reduced in cells expressing ouabain-resistant versions of these variants, indicating defective function of the variants relative to WT. Fluorescently tagged WT-α1 expressed in HEK293 cells showed plasmalemma localization, while all variants displayed increased cytosolic signal. In variant-expressing oocytes under voltage clamp, each variant showed passive “leak” currents, which were unaffected by ouabain in L302R and M859R, partially inhibited by ouabain in G303R and enhanced by ouabain in W931R. Na+ and Cl‒ permeate through G303R and W931R. M859R and W931R presented small K+-activated ouabain-inhibitable NKA currents that were absent in L302R and G303R. Growth of cells in ouabain and enzymatic studies demonstrated Na+/K+-dependent active transport for W931R, albeit with reduced Na+ affinity. We solved the cryo-EM structure of W931R-α1β1 with bound ouabain at 3 Å resolution. The structure showed a distorted transmembrane segment M9, which causes a concavity in the detergent micelle, suggesting that similar distortions may provide a leak pathway at the protein–lipid interface. Our results demonstrate that besides a loss of NKA function, all four variants carry an aberrant leak current that probably underlies the seizures and the hypomagnesemia characterizing this phenotype.
Na+,K+-ATPase possesses a highly conserved glycine (G358 in the α3 isoform) that—together with a nearby isoleucine (I363 in α3)—is targeted by mutations causing some of the most severe neurological phenotypes of the clinical spectrum of α3-Na+,K+-ATPase mutations. The disease mutations α3-G358V and α3-I363N affect Na+ and K+ transport to an extent incompatible with cell growth. However, alanine replacement of the corresponding glycine G363 in the α1 isoform is compatible with cell growth, allowing the effects on Na+,K+-ATPase function to be addressed using enzymatic assays on plasma membranes isolated from transfected cells. Occlusion of Na+ appears to be defective in mutant G363A, resulting in a reduced rate of phosphorylation from ATP. Furthermore, the mutation displaces the major conformational equilibrium of Na+,K+-ATPase such that the K+-occluded state is destabilized and occluded K+ is released faster, thereby leading to accumulation of a non-productive state without bound Na+ or K+. The critical function of the glycine can be ascribed to a strategic location at the bending point between an α helix and a β strand, where it connects the catalytic ATP hydrolysis site in the cytoplasmic P domain with the ion-binding region in the membrane and coordinates important intramolecular domain movements during the Na+,K+-ATPase transport cycle.
Serious neurological disorders are caused by mutation of Na+,K+-ATPase α1–α3 isoforms. Several disease mutations reduce the affinity for Na+, which in vivo results in an increase in the intracellular Na+ concentration with a secondary rise in intracellular Ca2+. Current treatments are ineffective, as they do not address Na+,K+-ATPase function. Hence, a platform for the development of a rational therapeutic approach that restores Na+ affinity is needed. Using the COS-1 cell expression system, one of the most conspicuous reductions in Na+ affinity of Na+,K+-ATPase has been reported for the neurological disease mutation α3-D923N of the α3 isoform. We describe here second-site mutations (“rescue mutations”) that fully restore Na+ affinity of α3-D923N and the corresponding rat α1-D928N and reveal the underlying molecular mechanism. Paradoxically, this insight evolved from investigation of a rapid-onset ataxia Na+,K+-ATPase mutant, α3-G316S, that exhibits an increased affinity for Na+. When applied as second-site mutations, α3-G316S and the corresponding rat α1-G321S are super-efficient rescuers of the reduced Na+ affinity of α3-D923N and rat α1-D928N. Replacement of the glycine with serine leads to a clash between transmembrane helices in the K+-bound E2 state of Na+,K+-ATPase but not in the Na+-bound E1 state. A similarly strong rescuing power is observed for second-site mutations breaking a salt bridge present only in the E2 state. The rescue is due to a structural change resembling the rearrangement of the ion binding sites occurring during the normal Na+,K+-ATPase transport cycle as part of the E2-to-E1 conformational transition.