Uremic solutes modulate hepatic bile acid handling and induce mitochondrial toxicity.

Chronic kidney disease (CKD) is accompanied by accumulating levels of uremic solutes in the circulation. Changes in the size and composition of the bile acid pool have also been observed. We investigated via which mechanisms uremic solutes may interfere with hepatocyte function and thus contribute to altered bile acid handling. We studied interference on the level of bile acid synthesis by cytochrome P450 7A1 (CYP7A1), explored effects on hepatic bile acid transporters, and investigated effects on mitochondrial function. In HEK293 cells overexpressing bile salt transporters, we observed that p-cresyl sulfate inhibited Na+-taurocholate cotransporting polypeptide (NTCP)-mediated uptake of taurocholic acid (TCA), whereas organic anion-transporting polypeptide 1B1 (OATP1B1)-mediated TCA uptake was increased. Assays in transporter-overexpressing membrane vesicles revealed that kynurenic acid inhibited TCA transport via the bile salt efflux pump (BSEP), whereas p-cresyl glucuronide and hippuric acid increased TCA efflux via multidrug resistance-associated protein 3 (MRP3). Moreover, indoxyl sulfate decreased mRNA expression of NTCP, OATP1B3 and CYP7A1 in primary human hepatocytes. Transport studies confirmed a decreased TCA uptake in indoxyl sulfate-exposed hepatocytes. Decreased hepatocyte viability was found for all seven uremic solutes tested, whereas five out of seven also decreased intracellular ATP levels and mitochondrial membrane potential. In conclusion, uremic solutes affect hepatic bile acid transport and mitochondrial function. This can contribute to the altered bile acid homeostasis observed in CKD patients.

Development of a mechanistic biokinetic model for hepatic bile acid handling to predict possible cholestatic effects of drugs.

Drug-induced liver injury (DILI) is a common reason for drug withdrawal from the market. An important cause of DILI is drug-induced cholestasis. One of the major players involved in drug-induced cholestasis is the bile salt efflux pump (BSEP; ABCB11). Inhibition of BSEP by drugs potentially leads to cholestasis due to increased (toxic) intrahepatic concentrations of bile acids with subsequent cell injury. In order to investigate the possibilities for in silico prediction of cholestatic effects of drugs, we developed a mechanistic biokinetic model for human liver bile acid handling populated with human in vitro data. For this purpose we considered nine groups of bile acids in the human bile acid pool, i.e. chenodeoxycholic acid, deoxycholic acid, the remaining unconjugated bile acids and the glycine and taurine conjugates of each of the three groups. Michaelis-Menten kinetics of the human uptake transporter Na+-taurocholate cotransporting polypeptide (NTCP; SLC10A1) and BSEP were measured using NTCP-transduced HEK293 cells and membrane vesicles from BSEP-overexpressing HEK293 cells. For in vitro-in vivo scaling, transporter abundance was determined by LC-MS/MS in these HEK293 cells and vesicles as well as in human liver tissue. Other relevant human kinetic parameters were collected from literature, such as portal bile acid levels and composition, bile acid synthesis and amidation rate. Additional empirical scaling was applied by increasing the excretion rate with a factor 2.4 to reach near physiological steady-state intracellular bile acid concentrations (80µM) after exposure to portal vein bile acid levels. Simulations showed that intracellular bile acid concentrations increase 1.7 fold in the presence of the BSEP inhibitors and cholestatic drugs cyclosporin A or glibenclamide, at intrahepatic concentrations of 6.6 and 20µM, respectively. This simplified model provides a tool for a first indication whether drugs at therapeutic concentrations might cause cholestasis by inhibiting BSEP.

A missense variant of the ATP1A2 gene is associated with a novel phenotype of progressive sensorineural hearing loss associated with migraine.

Hereditary sensorineural hearing loss is an extremely clinical and genetic heterogeneous disorder in humans. Especially, syndromic hearing loss is subdivided by combinations of various phenotypes, and each subtype is related to different genes. We present a new form of progressive hearing loss with migraine found to be associated with a variant in the ATP1A2 gene. The ATP1A2 gene has been reported as the major genetic cause of familial migraine by several previous studies. A Korean family presenting progressive hearing loss with migraine was ascertained. The affected members did not show any aura or other neurologic symptoms during migraine attacks, indicating on a novel phenotype of syndromic hearing loss. To identify the causative gene, linkage analysis and whole-exome sequencing were performed. A novel missense variant, c.571G>A (p.(Val191Met)), was identified in the ATP1A2 gene that showed co-segregation with the phenotype in the family. In silico studies suggest that this variant causes a change in hydrophobic interactions and thereby slightly destabilize the A-domain of Na(+)/K(+)-ATPase. However, functional studies failed to show any effect of the p.(Val191Met) substitution on the catalytic rate of this enzyme. We describe a new phenotype of progressive hearing loss with migraine associated with a variant in the ATP1A2 gene. This study suggests that a variant in Na(+)/K(+)-ATPase can be involved in both migraine and hearing loss.

Na(+),K(+)-ATPase isoform selectivity for digitalis-like compounds is determined by two amino acids in the first extracellular loop.

Digitalis-like compounds (DLCs) comprise a diverse group of molecules characterized by a cis-trans-cis ring-fused steroid core linked to a lactone. They have been used in the treatment of different medical problems including heart failure, where their inotropic effect on heart muscle is attributed to potent Na(+),K(+)-ATPase inhibition. Their application as drugs, however, has declined in recent past years due to their small safety margin. Since human Na(+),K(+)-ATPase is represented by four different isoforms expressed in a tissue-specific manner, one of the possibilities to improve the therapeutic index of DLCs is to exploit and amend their isoform selectivity. Here, we aimed to reveal the determinants of selectivity of the ubiquitously expressed α1 isoform and the more restricted α2 isoform toward several well-known DLCs and their hydrogenated forms. Using baculovirus to express various mutants of the α2 isoform, we were able to link residues Met(119) and Ser(124) to differences in affinity between the α1 and α2 isoforms to ouabain, dihydro-ouabain, digoxin, and dihydro-digoxin. We speculate that the interactions between these amino acids and DLCs affect the initial binding of these DLCs. Also, we observed isoform selectivity for DLCs containing no sugar groups.

Biochemical characterization of sporadic/familial hemiplegic migraine mutations.

Sporadic hemiplegic migraine type 2 (SHM2) and familial hemiplegic migraine type 2 (FHM2) are rare forms of hemiplegic migraine caused by mutations in the Na(+),K(+)-ATPase α2 gene. Today, more than 70 different mutations have been linked to SHM2/FHM2, randomly dispersed over the gene. For many of these mutations, functional studies have not been performed. Here, we report the functional characterization of nine SHM2/FHM2 linked mutants that were produced in Spodoptera frugiperda (Sf)9 insect cells. We determined ouabain binding characteristics, apparent Na(+) and K(+) affinities, and maximum ATPase activity. Whereas membranes containing T345A, R834Q or R879W possessed ATPase activity significantly higher than control membranes, P796S, M829R, R834X, del 935-940 ins Ile, R937P and D999H membranes showed significant loss of ATPase activity compared to wild type enzyme. Further analysis revealed that T345A and R879W showed no changes for any of the parameters tested, whereas mutant R834Q possessed significantly decreased Na(+) and increased K(+) apparent affinities as well as decreased ATPase activity and ouabain binding. We hypothesize that the majority of the mutations studied here influence interdomain interactions by affecting formation of hydrogen bond networks or interference with the C-terminal ion pathway necessary for catalytic activity of Na(+),K(+)-ATPase, resulting in decreased functionality of astrocytes at the synaptic cleft expressing these mutants.

Alternating Hemiplegia of Childhood mutations have a differential effect on Na(+),K(+)-ATPase activity and ouabain binding.

De novo mutations in ATP1A3, the gene encoding the α3-subunit of Na(+),K(+)-ATPase, are associated with the neurodevelopmental disorder Alternating Hemiplegia of Childhood (AHC). The aim of this study was to determine the functional consequences of six ATP1A3 mutations (S137Y, D220N, I274N, D801N, E815K, and G947R) associated with AHC. Wild type and mutant Na(+),K(+)-ATPases were expressed in Sf9 insect cells using the baculovirus expression system. Ouabain binding, ATPase activity, and phosphorylation were absent in mutants I274N, E815K and G947R. Mutants S137Y and D801N were able to bind ouabain, although these mutants lacked ATPase activity, phosphorylation, and the K(+)/ouabain antagonism indicative of modifications in the cation binding site. Mutant D220N showed similar ouabain binding, ATPase activity, and phosphorylation to wild type Na(+),K(+)-ATPase. Functional impairment of Na(+),K(+)-ATPase in mutants S137Y, I274N, D801N, E815K, and G947R might explain why patients having these mutations suffer from AHC. Moreover, mutant D801N is able to bind ouabain, whereas mutant E815K shows a complete loss of function, possibly explaining the different phenotypes for these mutations.

Familial hemiplegic migraine mutations affect Na,K-ATPase domain interactions.

Familial hemiplegic migraine (FHM) is a monogenic variant of migraine with aura. One of the three known causative genes, ATP1A2, which encodes the α2 isoform of Na,K-ATPase, causes FHM type 2 (FHM2). Over 50 FHM2 mutations have been reported, but most have not been characterized functionally. Here we study the molecular mechanism of Na,K-ATPase α2 missense mutations. Mutants E700K and P786L inactivate or strongly reduce enzyme activity. Glutamic acid 700 is located in the phosphorylation (P) domain and the mutation most likely disrupts the salt bridge with Lysine 35, thereby destabilizing the interaction with the actuator (A) domain. Mutants G900R and E902K are present in the extracellular loop at the interface of the α and ß subunit. Both mutants likely hamper the interaction between these subunits and thereby decrease enzyme activity. Mutants E174K, R548C and R548H reduce the Na(+) and increase the K(+) affinity. Glutamic acid 174 is present in the A domain and might form a salt bridge with Lysine 432 in the nucleotide binding (N) domain, whereas Arginine 548, which is located in the N domain, forms a salt bridge with Glutamine 219 in the A domain. In the catalytic cycle, the interactions of the A and N domains affect the K(+) and Na(+) affinities, as observed with these mutants. Functional consequences were not observed for ATP1A2 mutations found in two sporadic hemiplegic migraine cases (Y9N and R879Q) and in migraine without aura (R51H and C702Y).

Na,K-ATPase activity modulates Src activation: a role for ATP/ADP ratio.

Digitalis-like compounds (DLCs), specific inhibitors of Na,K-ATPase, are implicated in cellular signaling. Exposure of cell cultures to ouabain, a well-known DLC, leads to up- or down regulation of various processes and involves activation of Src kinase. Since Na,K-ATPase is the only known target for DLC binding an in vitro experimental setup using highly purified Na,K-ATPase from pig kidney and commercially available recombinant Src was used to investigate the mechanism of coupling between the Na,K-ATPase and Src. Digoxin was used as a representative DLC for inhibition of Na,K-ATPase. The activation of Src kinase was measured as the degree of its autophosphorylation. It was observed that in addition to digoxin, Src activation was dependent on concentrations of other specific ligands of Na,K-ATPase: Na(+), K(+), vanadate, ATP and ADP. The magnitude of the steady-state ATPase activity therefore seemed to affect Src activation. Further experiments with an ATP regenerating system showed that the ATP/ADP ratio determined the extent of Src activation. Thus, our model system which represents the proposed very proximal part of the Na,K-ATPase-Src signaling cascade, shows that Src kinase activity is regulated by both ATP and ADP concentrations and provides no evidence for a direct interaction between Na,K-ATPase and Src.