A membrane-embedded glutamate is required for ligand binding to the multidrug transporter EmrE.

EmrE is an Escherichia coli multidrug transporter that confers resistance to a variety of toxins by removing them in exchange for hydrogen ions. The detergent-solubilized protein binds tetraphenylphosphonium (TPP(+)) with a K(D) of 10 nM. One mole of ligand is bound per approximately 3 mol of EmrE, suggesting that there is one binding site per trimer. The steep pH dependence of binding suggests that one or more residues, with an apparent pK of approximately 7.5, release protons prior to ligand binding. A conservative Asp replacement (E14D) at position 14 of the only membrane-embedded charged residue shows little transport activity, but binds TPP(+) at levels similar to those of the wild-type protein. The apparent pK of the Asp shifts to <5.0. The data are consistent with a mechanism requiring Glu14 for both substrate and proton recognition. We propose a model in which two of the three Glu14s in the postulated trimeric EmrE homooligomer deprotonate upon ligand binding. The ligand is released on the other face of the membrane after binding of protons to Glu14.

Sorting of P-type ATPases in polarized epithelial cells.

The Na,K-ATPase and the H,K-ATPase are highly homologous members of the P-type family of ion transporting ATPase. Despite their structural similarity, these two pumps are sorted to different destinations in polarized epithelial cells. While the Na,K-ATPase is restricted to the basolateral surfaces of most epithelial cells types, the H,K-ATPase is concentrated at the apical plasmalemma and in a pre-apical vesicular storage compartment in the parietal cells of the stomach. We have generated molecular chimeras composed of complementary portions of these two pumps' alpha-subunits. By expressing these pump constructs in polarized epithelial cells in culture, we have been able to identify sequence domains which participate in the targetting of the holoenzyme. We find that information embedded within the sequence of the fourth transmembrane domain of the H,K-ATPase is sufficient to account for this protein's apical localization. Stimulation of gastric acid secretion results in insertion of the intracellular H,K-ATPase pool into the apical plasma membrane and inactivation of acid secretion is accompanied by the re-internalization of these pumps. We have identified a tyrosine-based signal in the cytoplasmic tail of the H,K-ATPase beta-subunit which appears to be required for this endocytosis. We have mutated the critical tyrosine residue to alanine and expressed the altered protein in transgenic mice. The H,K-ATPase remains continuously at the apical cell surface in parietal cells from these animals, and they constitutively hypersecrete gastric acid. These results demonstrate that the beta-subunit sequence mediates the internalization of the H,K-ATPase and is required for the cessation of gastric acid secretion. Thus, at least two sorting signals are required to ensure the proper targetting and regulation of the gastric H,K pump.

Identification of sorting determinants in the C-terminal cytoplasmic tails of the gamma-aminobutyric acid transporters GAT-2 and GAT-3.

In order to perform their physiologic functions, polarized epithelial cells must target ion transport proteins to the appropriate domains of their plasma membranes. Molecular signals responsible for polarized sorting have been identified for several membrane proteins which span the bilayer once. Most ion transport proteins are polytopic, however, and little is known of the signals responsible for the targeting of this class of polypeptides. Members of the gamma-aminobutyric acid (GABA) transporter family are polytopic membrane proteins found endogenously in both epithelial cells and neurons. We have identified narrowly defined sequences which are required for the proper accumulation of two members of this transporter family in Madin-Darby canine kidney cells. The highly homologous GABA transporter isoforms, GAT-2 and GAT-3, localize to the basolateral and apical surfaces, respectively, when expressed stably in Madin-Darby canine kidney cells. We have generated deletion constructs and chimeric transporters composed of complimentary portions of GAT-2 and GAT-3. We find that information which directs their differential sorting is present in the C-terminal cytoplasmic tails of these two polypeptides. A sequence of 22 amino acids at the C terminus of GAT-2 is required for the transporter's basolateral distribution and is capable of directing GAT-3 to the basolateral surface when appended to the C terminus of this normally apical polypeptide. The deletion of 32 amino acids from the C terminus of GAT-3 causes this transporter to become mislocalized to both surfaces. Moreover, removal of the final three amino acids of GAT-3 (THF) similarly disrupts its apical sorting. The GAT-3 C-terminal sequence resembles motifs which interact with PDZ domains, raising the possibility that the steady state distribution of GAT-3 at the apical plasmalemmal surface requires a protein-protein interaction mediated by its extreme C-terminal cytoplasmic tail. These data provide the first characterization of a protein-based signal required for the apical distribution of a membrane protein.

A basolateral sorting signal is encoded in the alpha-subunit of Na-K-ATPase.

Na-K-ATPase and H-K-ATPase are highly homologous ion pumps that exhibit distinct plasma membrane distributions in epithelial cells. We have studied the alpha-subunits of these heterodimeric pumps to identify the protein domains responsible for their polarized sorting. A chimeric alpha-subunit construct (N519H) was generated in which the first 519 amino acid residues correspond to the Na-K-ATPase sequence and the remaining 500 amino acids are derived from the H-K-ATPase sequence. In stably transfected LLC-PK1 cell lines, we found that the N519H chimera is restricted to the basolateral surface under steady-state conditions, suggesting that residues within the NH2-terminal 519 amino acids of the Na-K-ATPase alpha-subunit contain a basolateral sorting signal. H-K-ATPase beta-subunit expressed alone in LLC-PK1 cells accumulates at the apical surface. When coexpressed with N519H, the H-K-ATPase beta-subunit assembles with this chimera and accompanies it to the basolateral surface. Thus the NH2-terminal basolateral signal in the Na-K-ATPase alpha-subunit masks or is dominant over any apical sorting information present in the beta-polypeptide. In gastric parietal cells, the H-K-ATPase beta-subunit targets the H-K-ATPase to an intracellular vesicular compartment which fuses with the plasma membrane in response to secretagogue stimulation. To test whether the chimera-H-K-ATPase beta-subunit complex is directed to a similar compartment in LLC-PK1 cells, we treated transfected cells with drugs that raise intracellular adenosine 3',5'-cyclic monophosphate (cAMP) levels. Elevation of cytosolic cAMP increased the surface expression of both the N519H chimera and the H-K-ATPase beta-subunit. This increase in surface expression, however, appears to be the result of transcriptional upregulation and not recruitment of chimera to the surface from a cAMP-inducible compartment.

Sorting and trafficking of ion transport proteins in polarized epithelial cells.

Transport proteins are targeted to specific plasmalemmal domains in polarized epithelial cells. The molecular signals that govern these sorting events are just beginning to be elucidated. In many cases, the cell surface delivery of transport proteins is subjected to tight regulation. Several different mechanisms appear to participate in these trafficking processes.

Localization of the effector-specifying regions of Gi2alpha and Gqalpha.

Heterotrimeric G proteins transmit hormonal and sensory signals received by cell surface receptors to effector proteins that regulate cellular processes. Members of the highly conserved family of alpha subunits specifically modulate the activities of a diverse array of effector proteins. To investigate the determinants of alpha subunit-effector specificity, we localized the effector-specifying regions of alphai2, which inhibits adenylyl cyclase, and alphaq, which stimulates phosphoinositide phospholipase C using chimeric alpha subunits. The chimeras were generated using an in vivo recombination method in Escherichia coli. The effector-specifying regions of both alphai2 and alphaq were localized within the GTPase domain. An alphaq/alphai2/alphaq chimera containing only 78 alphai2 residues within the GTPase domain robustly inhibited adenylyl cyclase. This alphai2 segment includes regions corresponding to two of the three regions of alphas that activate adenylyl cyclase, but does not include any of the alpha subunit regions that switch conformation upon binding GTP. Replacement of the alphaq residues that comprise the helical domain with the homologous alphai2 residues resulted in a chimeric alpha subunit that activated phospholipase C. Combined with previous studies of the effector-specifying residues of alphas and alphat, our results suggest that the effector specificity of alpha subunits is generally determined by the GTPase and not the helical domain.

Polarized expression of GABA transporters in Madin-Darby canine kidney cells and cultured hippocampal neurons.

At least three high affinity Na+- and Cl--dependent gamma-aminobutyric acid (GABA) transporters are known to exist in the rat and mouse brain. These transporters share 50-65% amino acid sequence identity with the kidney betaine transporter which also transports GABA but with lower affinity. The betaine transporter (BGT) is expressed on the basolateral surface of polarized Madin-Darby canine kidney (MDCK) cells. Recent evidence suggests that the signals and mechanisms involved in membrane protein sorting share many functional characteristics in polarized neurons and epithelial cells. It was previously shown that the rat GABA transporter GAT-1 is located in the presynaptic membrane of axons where it plays a role in terminating GABAergic neurotransmission. When expressed in MDCK cells by transfection, GAT-1 was sorted to the apical membrane. In this report, we have localized the other two GABA transporters, GAT-2 and GAT-3, in transfected MDCK cells by GABA uptake, immunofluorescence, and cell surface biotinylation. GAT-3, like GAT-1, localized to the apical membrane of MDCK cells while GAT-2, like BGT, localized to the basolateral membrane. We have also expressed BGT in low density cultures of hippocampal neurons by microinjection and immunolocalized it to the dendrites. The distribution of GAT-3 in these neurons after transfection was axonal as well as somatodendritic. These results indicate that highly homologous subtypes of GABA transporters are sorted differently when expressed in epithelial cells or neurons and suggest that these two cell types share the capacity to distinguish among these isoforms and target them to distinct destinations.

Interactions of VirB9, -10, and -11 with the membrane fraction of Agrobacterium tumefaciens: solubility studies provide evidence for tight associations.

The eleven predicted gene products of the Agrobacterium tumefaciens virB operon are believed to form a transmembrane pore complex through which T-DNA export occurs. The VirB10 protein is required for virulence and is a component of an aggregate associated with the membrane fraction of A. tumefaciens. Removal of the putative membrane-spanning domain (amino acids 22 through 55) disrupts the membrane topology of VirB10 (J. E. Ward, E. M. Dale, E. W. Nester, and A. N. Binns, J. Bacteriol. 172:5200-5210, 1990). Deletion of the sequences encoding amino acids 22 to 55 abolishes the ability of plasmid-borne virB10 to complement a null mutation in the virB10 gene, suggesting that the proper topology of VirB10 in the membrane may indeed play a crucial role in T-DNA transfer to the plant cell. Western blot (immunoblot) analysis indicated that the observed loss of virulence could not be attributed to a decrease in the steady-state levels of the mutant VirB10 protein. Although the deletion of the single transmembrane domain would be expected to perturb membrane association, VirB10 delta 22-55 was found exclusively in the membrane fraction. Urea extraction studies suggested that this membrane localization might be the result of a peripheral membrane association; however, the mutant protein was found in both inner and outer membrane fractions separated by sucrose density gradient centrifugation. Both wild-type VirB10 and wild-type VirB9 were only partially removed from the membranes by extraction with 1% Triton X-100, while VirB5 and VirB8 were Triton X-100 soluble. VirB11 was stripped from the membranes by 6 M urea but not by a more mild salt extraction. The fractionation patterns of VirB9, VirB10, and VirB11 were not dependent on each other or on VirB8 or VirD4. The observed tight association of VirB9, VirB10, and VirB11 with the membrane fraction support the notion that these proteins may exist as components of multiprotein pore complexes, perhaps spanning both the inner and outer membranes of Agrobacterium cells.