The crystal structure of a PNA duplex reveals both a right- and a left-handed helix in the unit cell. The helices are wide (28A), large pitched (18bp) with the base pairs perpendicular to the helix axis, thereby demonstrating that PNA besides adapting to oligonucleotide partners also has a unique structure by itself.

The sequence Y771TLTSNIPEIT781P in the fifth transmembrane segment of the alpha-subunit of Na,K-ATPase is unique among cation pump proteins. Here, in search of the molecular basis for Na,K specificity, alanine and conservative substitutions were directed to six oxygen-carrying residues in this segment. The contribution of the residues to cation binding was estimated from direct binding of Tl+ [Nielsen, et al. (1998) Biochemistry 37, 1961-1968], K+ displacement of ATP binding at equilibrium, and Na+-dependent phosphorylation from ATP in the presence of oligomycin. As an intrinsic control, substitution of Thr781 had no effect on Tl+(K+) or Na+ binding. There are several novel observations from this work. First, the carboxamide group of Asn776 is equally important for binding Tl+(K+) or Na+, whereas a shift of the position of the carboxamide of Asn776 (Asn776Gln) causes a large depression of Na+ binding without affecting the binding of Tl+(K+). Second, Thr774 is important for Na+ selectivity because removal of the hydroxyl group reduces the binding of Na+ with no effect on binding of Tl+(K+). Removal of the methyl groups of Thr774 or Thr772 reduces binding of both Tl+(K+) and Na+, whereas the hydroxyl group of Thr772 does not contribute to cation binding. Furthermore, the hydroxyl groups of Ser775 and Tyr771 are important for binding both Tl+(K+) and Na+. The data suggest that rotating or tilting of the cytoplasmic part of the fifth transmembrane segment may adapt distances between coordinating groups and contribute to the distinctive Na+/K+ selectivity of the pump.

Site-directed mutagenesis and assay of Rb+ and Tl+ occlusion in recombinant Na,K-ATPase from yeast were combined to establish structure-function relationships of amino acid side chains involved in high-affinity occlusion of K+ in the E2[2K] form. The wild-type yeast enzyme was capable of occluding 2 Rb+ or Tl+ ions/ouabain binding site or alpha 1 beta 1 unit with high apparent affinity (Kd(Tl+) = 7 +/- 2 microM), like the purified Na,K-ATPase from pig kidney. Mutations of Glu327(Gln,Asp), Asp804(Asn, Glu), Asp808(Asn, Glu) and Glu779(Asp) abolished high-affinity occlusion of Rb+ or Tl+ ions. The substitution of Glu779 for Gln reduced the occlusion capacity to 1 Tl+ ion/alpha 1 beta 1-unit with a 3-fold decrease of the apparent affinity for the ion (Kd(Tl+) = 24 +/- 8 microM). These effects on occlusion were closely correlated to effects of the mutations on K0.5(K+) for K+ displacement of ATP binding. Each of the four carboxylate residues Glu327, Glu779, and Asp804 or Asp808 in transmembrane segments 4, 5, and 6 is therefore essential for high-affinity occlusion of K+ in the E2[2K] form. These residues either may engage directly in cation coordination or they may be important for formation or stability of the occlusion cavity.

No abstract

For understanding the regulation of sodium reabsorption, it is important to know whether the alpha 2- or alpha 3-isoform of Na-K-adenosinetriphosphatase (Na-K-ATPase) is expressed in mammalian kidney in addition to the predominant alpha 1 beta 1-isozyme. Here we applied competitive polymerase chain reaction (PCR) for estimation of mRNA in parenchymal zones of rat kidney for comparison to high-affinity [3H]ouabain binding. The alpha 3-isoform mRNA was demonstrated to form 0.04-0.05% of the amount of alpha 1-isoform mRNA in the cortex, medulla, and papilla of rat kidney. The alpha 2-mRNA was demonstrated in all three zones and constituted 0.03% of the amount of alpha 1-mRNA in cortex. The abundance of the alpha 1 truncated mRNA was 0.1-0.8% of that of the alpha 1-mRNA. The upper limit for expression of Na-K-ATPase isozyme with high ouabain affinity (dissociation constant, 69-141 nM) was 0.096-0.14% of the concentration of alpha 1 beta 1-Na-K-ATPase. Thus a small but well-defined pool of alpha 2- and alpha 3-isoforms constitutes < or = 0.1% of the amount of alpha 1-isoform at both the mRNA and protein level in rat kidney.