Inactivation of the rabbit parotid Na/K/Cl cotransporter by N-ethylmaleimide

A novel regulatory locus of phosphorylation in the C terminus of the potassium chloride cotransporter KCC2 that interferes with N-ethylmaleimide or staurosporine-mediated activation

The neuron-specific cation chloride cotransporter KCC2 plays a crucial role in hyperpolarizing synaptic inhibition. Transporter dysfunction is associated with various neurological disorders, raising interest in regulatory mechanisms. Phosphorylation has been identified as a key regulatory process. Here, we retrieved experimentally observed phosphorylation sites of KCC2 from public databases and report on the systematic analysis of six phosphorylated serines, Ser(25), Ser(26), Ser(937), Ser(1022), Ser(1025), and Ser(1026). Alanine or aspartate substitutions of these residues were analyzed in HEK-293 cells. All mutants were expressed in a pattern similar to wild-type KCC2 (KCC2(WT)). Tl(+) flux measurements demonstrated unchanged transport activity for Ser(25), Ser(26), Ser(1022), Ser(1025), and Ser(1026) mutants. In contrast, KCC2(S937D), mimicking phosphorylation, resulted in a significant up-regulation of transport activity. Aspartate substitution of Thr(934), a neighboring putative phosphorylation site, resulted in a comparable increase in KCC2 transport activity. Both KCC2(T934D) and KCC2(S937D) mutants were inhibited by the kinase inhibitor staurosporine and by N-ethylmaleimide, whereas KCC2(WT), KCC2(T934A), and KCC2(S937A) were activated. The inverse staurosporine effect on aspartate versus alanine substitutions reveals a cross-talk between different phosphorylation sites of KCC2. Immunoblot and cell surface labeling experiments detected no alterations in total abundance or surface expression of KCC2(T934D) and KCC2(S937D) compared with KCC2(WT). These data reveal kinetic regulation of transport activity by these residues. In summary, our data identify a novel key regulatory phosphorylation site of KCC2 and a functional interaction between different conformation-changing post-translational modifications. The action of pharmacological agents aimed to modulate KCC2 activity for therapeutic benefit might therefore be highly context-specific.

A C-terminal domain in KCC2 confers constitutive K+-Cl- cotransport

The neuron-specific K(+)-Cl(-) cotransporter KCC2 plays a crucial role in determining intracellular chloride activity and thus the neuronal response to gamma-aminobutyric acid and glycine. Of the four KCCs, KCC2 is unique in mediating constitutive K(+)-Cl(-) cotransport under isotonic conditions; the other three KCCs are exclusively swelling-activated, with no isotonic activity. We have utilized a series of chimeric cDNAs to localize the determinant of isotonic transport in KCC2. Two generations of chimeric KCC4-KCC2 cDNAs initially localized this characteristic to within a KCC2-specific expansion of the cytoplasmic C terminus, between residues 929 and 1043. This region of KCC2 is rich in prolines, serines, and charged residues and encompasses two predicted PEST sequences. Substitution of this region in KCC2 with the equivalent sequence of KCC4 resulted in a chimeric KCC that was devoid of isotonic activity, with intact swelling-activated transport. A third generation of chimeras demonstrated that a domain just distal to the PEST sequences confers isotonic transport on KCC4. Mutagenesis of this region revealed that residues 1021-1035 of KCC2 are sufficient for isotonic transport. Swelling-activated K(+)-Cl(-) cotransport is abrogated by calyculin A, whereas isotonic transport mediated by KCC chimeras and KCC2 is completely resistant to this serine-threonine phosphatase inhibitor. In summary, a 15-residue C-terminal domain in KCC2 is both necessary and sufficient for constitutive K(+)-Cl(-) cotransport under isotonic conditions. Furthermore, unlike swelling-activated transport, constitutive K(+)-Cl(-) cotransport mediated by KCC2 is completely independent of serine-threonine phosphatase activity, suggesting that these two modes of transport are activated by distinct mechanisms.

Detection and characterization of protein nitrosothiols

Protein S-nitrosylation is a post-translational modification of cysteine residues elicited by nitric oxide (NO). Detection and quantification of protein nitrosothiols remains a challenge because of the lability of the nitrosothiol moiety. Here, we describe approaches for labeling S-nitrosylated proteins with affinity and radioactive tags to facilitate their detection, purification, and identification.

Ion transport and ligand binding by the Na-K-Cl cotransporter, structure-function studies

The cation-Cl cotransporters (CCCs) mediate the coupled movement of Na and/or K to that of Cl across the plasmalemma of animal cells. Eight CCCs have been identified to date: two Na-K-Cl cotransporters (NKCC), four K-Cl cotransporters (KCCs), one Na-Cl cotransporter (NCC) and one CCC interacting protein (CIP). All of the NKCCs and KCCs are inhibited by loop diuretics; mercury and other modifying agents are also known to block NKCC-mediated transport. In this work, we have utilized a mutational approach to study the interaction between different substrates and the NKCCs. We relied on the strategy of exchanging domains between functionally distinct carriers (the shark NKCCl and the human NKCCl) to identify residues or group of residues that are involved in the interaction with ions, loop diuretics and Hg. Our results show that the N- and C-termini have no role in determining the species differences in ion transport and bumetanide binding. On the other hand, the interaction between Hg and the NKCCs is found to partially involve the C-terminus through residues that contain available sulfhydryl groups. Within the transmembrane segments, variant residues in the 2nd, 4th and 7th predicted alpha-helices are shown to encode the differences in ion transport between the shark and the human cotransporters. For loop diuretic binding, several regions throughout the central domain appear to be involved. Interestingly, these regions are not the same as those involved in cation or anion transport, and in Hg binding.

Inhibition of Na(+)-K(+)-2Cl(-) cotransport by mercury

Mercury alters the function of proteins by reacting with cysteinyl sulfhydryl (SH(-)) groups. The inorganic form (Hg(2+)) is toxic to epithelial tissues and interacts with various transport proteins including the Na(+) pump and Cl(-) channels. In this study, we determined whether the Na(+)-K(+)-Cl(-) cotransporter type 1 (NKCC1), a major ion pathway in secretory tissues, is also affected by mercurial substrates. To characterize the interaction, we measured the effect of Hg(2+) on ion transport by the secretory shark and human cotransporters expressed in HEK-293 cells. Our studies show that Hg(2+) inhibits Na(+)-K(+)-Cl(-) cotransport, with inhibitor constant (K(i)) values of 25 microM for the shark carrier (sNKCC1) and 43 microM for the human carrier. In further studies, we took advantage of species differences in Hg(2+) affinity to identify residues involved in the interaction. An analysis of human-shark chimeras and of an sNKCC1 mutant (Cys-697-->Leu) reveals that transmembrane domain 11 plays an essential role in Hg(2+) binding. We also show that modification of additional SH(-) groups by thiol-reacting compounds brings about inhibition and that the binding sites are not exposed on the extracellular face of the membrane.

Ultrastructural localization of the Na-K-Cl cotransporter in the lateral wall of the rabbit cochlear duct

Localization of the immunoreactivity in the lateral wall of the rabbit cochlear duct was examined using a post-embedding immunogold method with a polyclonal antiserum raised against the rabbit parotid Na-K-Cl cotransporter. In the stria vascularis, the labeling was significant on the basolateral membrane infolding of marginal cells, whereas no labeling was seen on the luminal membrane of these cells. Immunoreactivity was also detected on the cell membranes of various other cells. These include fibrocytes of the spiral ligament and the spiral prominence, and vascular endothelial cells in the stria vascularis and the spiral ligament. In contrast, virtually no gold particles were seen on the membrane of intermediate cells, basal cells of the stria vascularis, the epithelial cells of the spiral prominence, or Reissner's membrane. Our result on the localization of the Na-K-Cl cotransporter in marginal cells is consistent with electrophysiological studies (Wangemann et al. (1995) Hear. Res. 84, 19-29). Our result on fibrocytes is discussed in relation to K+ circulation into endolymph from perilymph (Schulte and Steel (1994) Hear. Res. 78, 65-76).

Identification, characterization and purification of a 160 kD bumetanide-binding glycoprotein from the rabbit parotid

We demonstrate the presence of a 160 kD protein in rabbit parotid basolateral membranes that can be labeled with the irreversible sulfhydryl reagent [14C]-N-ethylmaleimide in a bumetanide-protectable fashion. The specificity of this labeling, and our previous evidence for the existence of an essential sulfhydryl group closely associated with the bumetanide-binding site on the parotid Na(+)-K(+)-Cl-cotransporter (J. Membrane Biol. 112:51-58, 1989), provide strong evidence that this protein is a part or all of the parotid bumetanide-binding site. When this protein is treated with endoglycosidase F/N-glycosidase F to remove N-linked oligosaccharides, its apparent molecular weight decreases to 135 kD. The pI of this deglycosylated protein is approximately 6.4. The bumetanide-binding protein was purified using two preparative electrophoresis steps. First, a Triton X-100 extract enriched in this protein was run on preparative electrophoresis to obtain fractions containing proteins in the 160 kD range. These were then deglycosylated with endoglycosidase F/N-glycosidase F and selected fractions were pooled and rerun on preparative electrophoresis to obtain a final 135 kD fraction. The enrichment of the bumetanide-binding protein in this final 135 kD fraction estimated from [14C]-N-ethylmaleimide labeling was approximately 48 times relative to the starting membrane extract. Since the bumetanide-binding site represents approximately 2% of the total protein in this starting extract, this enrichment indicates a high degree of purity of this protein in the 135 kD fraction.

Group-selective reagent modification of the sodium- and chloride-coupled glycine transporter under native and reconstituted conditions

Glycine transporter from rat brain stem and spinal cord is inactivated by specific sulfhydryl reagents. Modification of lysine residues also promotes a decrease of the transporter activity but in a lesser extent than that promoted by thiol group reagents. Mercurials showed a more marked inhibitory effect than maleimide derivatives. SH groups display a similar reactivity for p-chloromercuribenzenesulfonate (pCMBS) and mersalyl in synaptosomal membrane vesicles and proteoliposomes reconstituted with the solubilized transporter. However, different reactivity is observed with N-ethylmaleimide (MalNEt), the greatest effect being attained in membrane vesicles. The rate of inactivation by pCMBS and MalNEt is pseudo-first-order showing time- and concentration-dependence. pCMBS and MalNEt decrease the Vmax for glycine transport and to a lesser extent act on the apparent Km. Treatment with dithiothreitol (DTT) of the transporter modified by pCMBS results in a complete restoration of transporter activity indicating that the effect exercised by the reagent is specific for cysteine residues on the protein. It is concluded that SH groups are involved in the glycine transporter function and that these critical residues are mostly located in a relatively hydrophilic environment of the protein.