Structural analysis and chromosomal localization of the mouse Psmb5 gene coding for the constitutively expressed β-type proteasome subunit

 The proteasome is a multi-subunit protease responsible for the production of peptides presented by major histocompatibility complex class I molecules. Accumulated evidence indicates that, upon stimulation with interferon-γ (IFN-γ), three β-type subunits, designated LMP2, LMP7, and PSMB10, are incorporated into the 20S proteasome by displacing the housekeeping β-type subunits designated PSMB6, PSMB5, and PSMB7, respectively. These changes in the subunit composition appear to facilitate class I-mediated antigen presentation, presumably by altering the cleavage specificities of the proteasome. In the present study, we determined the organization of the mouse gene Psmb5, coding for the PSMB5 subunit. Psmb5 is made up of three exons, spanning ∼5 kilobases. Its exon-intron organization differs radically from those of the other IFN-γ-regulated, β-type subunit genes including Lmp7 with which Psmb5 is believed to share an immediate common ancestor. The structure of the mouse Psmb5 gene is identical to that of its recently characterized human counterpart. Thus, the unique organization of the gene coding for the PSMB5 subunit appears to have been established before mammalian radiation. As well as the Psmb5 gene, the mouse genome contains a processed pseudogene designated Psmb5-ps. Interspecific backcross mapping showed that Psmb5 maps close to the Gtrgal2 locus on chromosome 14 and that Psmb5-ps is located in the vicinity of the Psme3 locus on chromosome 11. These results were confirmed by fluorescent in situ hybridization analysis that localized Psmb5 to band C2 to proximal D1 of chromosome 14 and Psmb5-ps to band D of chromosome 11. Received: 29 May 1997 / Revised: 4 June 1997     

Probing functional interfaces of rod PDE γ-subunit using scanning fluorescent labeling

In the dark, the activity of the rod cGMP phosphodiesterase (PDE) catalytic α- and β-subunits (Pαβ) is inhibited by two γ-subunits (Pγ). On light stimulation of the photoreceptor cells, the GTP-bound α-subunit of visual G-protein transducin (G_taGTP) displaces the Pγ-subunits from their inhibitory sites on Pαβ, leading to the effector enzyme activation. We designed a number of Pγ mutants, each with a single cysteine residue evenly distributed at a different position along the Pγ polypeptide chain. These cysteine residues served as sites for the introduction of the environmentally sensitive fluorescent probe, 3-(bromoacetyl)-7-diethyl aminocoumarin (BC). Analysis of the interactions of Pαβ and G_ta with the fluorescently labeled Pγ mutants suggests two distinct functional interfaces of Pγ. The Pαβ/Pγ interface is formed essentially by the C-terminus of Pγ and by the N-terminal portion of the Pγ polycationic region, Pγ-24-45, whereas the Pγ/G_ta interface includes the C-terminal portion of Pγ-24-45 and the region surrounding Pγ Cys68. Such functional organization of Pγ may represent an important element for the PDE activation mechanism during transduction of visual signals.     

Contribution of the C-terminal region to the thermostability of the archaeal group II chaperonin from Thermococcus sp. strain KS-1

Chaperonin is a double ring-shaped oligomeric protein complex, which captures a protein in the folding intermediate state and assists its folding in an ATP-dependent manner. The chaperonin from a hyperthermophilic archaeum, Thermococcus sp. strain KS-1, is a group II chaperonin and is composed of two distinct subunits, α and β. Although these subunits are highly homologous in sequence, the homo-oligomer of the β-subunit is more thermostable than that of the α-subunit. To identify the region responsible for this difference in thermostability, we constructed domain-exchange mutants. The mutants containing the equatorial domain of the β-subunit were more resistant to thermal dissociation than the mutants with that of the α-subunit. Thermostability of a β-subunit mutant whose C-terminal 22 residues were replaced with those of the α-subunit decreased to the comparable level of that of the α-subunit homo-oligomer. These results indicate that the difference in thermostability between α- and β-subunits mainly originates in the C-terminal residues in the equatorial domain, only where they exhibit substantial sequence difference.Takao Yoshida, Taro Kanzaki, Ryo Iizuka and Toshihiro Komada contributed equally to this paper.     

Developmental changes in β-subunit composition of Na,K-ATPase in the Drosophila eye

The Drosophila genome contains at least three loci for the Na,K-ATPase β-subunit; however, only the protein products of nrv1 and nrv2 have been characterized hitherto. Here, we provide evidence that nrv3 also encodes for a functional Na,K-ATPase β-subunit, as its protein product co-precipitates with the Na,K-ATPase α-subunit. Nrv3 expression in adult flies is restricted to the nervous system in which Nrv3 is enriched in selective types of sensory cells. Because Nrv3 expression is especially prominent in the compound eye, we have analyzed the subcellular and developmental distribution of Nrv3 within the visual cells and related this distribution to those of the α-subunit and of the β-subunits Nrv1 and Nrv2. Prospective visual cells express Nrv2 in the third larval instar stage and during the first half of pupal development. During the last third of pupal life, Nrv3 gradually replaces Nrv2 as the Na,K-ATPase β-subunit in the photoreceptor cells. Adult photoreceptors express Nrv3 as their major β-subunit; the visual cells R1–R6 co-express Nrv2 at a low level, whereas R7 and R8 co-express Nrv1. Notably, β-subunits do not co-distribute exactly with the α-subunit at some developmental stages, supporting the concept that the α-subunit and β-subunit can exist in the plasma membrane without being engaged in α/β heterodimers. The non-visual cells within the compound eye express almost exclusively Nrv2, which segregates together with the α-subunit to septate junctions throughout development.     

Expression of the Avian Na,K-ATPase Subunits in Dictyostelium discoideum

This study explored whether Dictyostelium discoideum can be used to express the avian Na,K-ATPase, a heterodimeric membrane protein. Dictyostelium was able to express mRNAs encoding the avian Na,K-ATPase subunits. However, Dictyostelium expressed avian Na,K-ATPase protein when only when a Dictyostelium consensus ribosomal binding sequence, AAAATAAA, was inserted in front of the open reading frames of the α₁- and β₁-subunit cDNAs and the first eight codons following the start-translation codons were changed to Dictyostelium preferred codons. These modified mRNAs appeared to be much less stable than the forms that were not readily translated. Dictyostelium could express the avian β-subunit alone but only expressed the α₁-subunit when the β₁-subunit was co-expressed. Subunit assembly occurred in cells expressing both α₁- and β₁-subunits. The bulk of the exogenously expressed sodium pump subunits remained in an intracellular compartment, presumed to be the endoplasmic reticulum. Dictyostelium exported little or no Na,K-ATPase or free β-subunit to the plasma membrane. Received: 7 July 1998/Revised: 8 October 1998     

A PCR-based method to test for the presence or absence of β-conglycinin α′- and α-subunits in soybean

The soybean cultivar Yumeminori, which lacks the α′- and α-subunits of β-conglycinin, carries both naturally occurring and induced mutations. While the cause of the natural mutation resulting in the α′-subunit deficiency has been determined, the induced mutation in the CG-2 gene encoding the α-subunit has not been characterized at the molecular level. In this study, we identified a four base pair insertion in the first exon of CG-2, which introduced a premature stop codon. The insertion co-segregated with the lack of α-subunit, indicating that this mutation is the cause of the α-subunit deficiency. A multiplex PCR method of testing for the presence or absence of α′- and α-subunits was developed based on the sequences of mutated and wild-type alleles. This PCR-based test was also capable of detecting the presence of wild-type genes when Yumeminori DNA samples were contaminated with wild-type DNA at levels of 0.2% or greater. Thus, this method will be useful both for marker-assisted selection in soybean breeding programs, and for seed purity tests in food industries.     

Palytoxin Acts on Na+,K+-ATPase but not Nongastric H+,K+-ATPase

Palytoxin (PTX) opens a pathway for ions to pass through Na,K-ATPase. We investigate here whether PTX also acts on nongastric H,K-ATPases. The following combinations of cRNA were expressed in Xenopus laevis oocytes: Bufo marinus bladder H,K-ATPase α₂- and Na,K-ATPase β₂-subunits; Bufo Na,K-ATPase α₁- and Na,K-ATPase β₂-subunits; and Bufo Na,K-ATPase β₂-subunit alone. The response to PTX was measured after blocking endogenous Xenopus Na,K-ATPase with 10 μm ouabain. Functional expression was confirmed by measuring ⁸⁶Rb uptake. PTX (5 nm) produced a large increase of membrane conductance in oocytes expressing Bufo Na,K-ATPase, but no significant increase occurred in oocytes expressing Bufo H,K-ATPase or in those injected with Bufo β₂-subunit alone. Expression of the following combinations of cDNA was investigated in HeLa cells: rat colonic H,K-ATPase α₁-subunit and Na,K-ATPase β₁-subunit; rat Na,K-ATPase α₂-subunit and Na,K-ATPase β₂-subunit; and rat Na,K-ATPase β₁- or Na,K-ATPase β₂-subunit alone. Measurement of increases in ⁸⁶Rb uptake confirmed that both rat Na,K and H,K pumps were functional in HeLa cells expressing rat colonic HKα₁/NKβ₁ and NKα₂/NKβ₂. Whole-cell patch-clamp measurements in HeLa cells expressing rat colonic HKα₁/NKβ₁ exposed to 100 nm PTX showed no significant increase of membrane current, and there was no membrane conductance increase in HeLa cells transfected with rat NKβ₁- or rat NKβ₂-subunit alone. However, in HeLa cells expressing rat NKα₂/NKβ₂, outward current was observed after pump activation by 20 mm K⁺ and a large membrane conductance increase occurred after 100 nm PTX. We conclude that nongastric H,K-ATPases are not sensitive to PTX when expressed in these cells, whereas PTX does act on Na,K-ATPase.     

Treatment with TNF-α and IFN-γ alters the activation of SER/THR protein kinases and the metabolic response to IGF-I in mouse c2c12 myogenic cells

The aim of this study was to compare the effects of TNF-α, IL-1β and IFN-γ on the activation of protein kinase B (PKB), p70^S6k, mitogen-activated protein kinase (MAPK) and p90 ^rsk , and on IGF-I-stimulated glucose uptake and protein synthesis in mouse C2C12 myotubes. 100 nmol/l IGF-I stimulated glucose uptake in C2C12 myotubes by 198.1% and 10 ng/ml TNF-α abolished this effect. Glucose uptake in cells differentiated in the presence of 10 ng/ml IFN-γ increased by 167.2% but did not undergo significant further modification upon the addition of IGF-I. IGF-I increased the rate of protein synthesis by 249.8%. Neither TNF-α nor IFN-γ influenced basal protein synthesis, but both cytokines prevented the IGF-I effect. 10 ng/ml IL-1β did not modify either the basal or IGF-I-dependent glucose uptake and protein synthesis. With the exception of TNF-α causing an 18% decrease in the level of PKB protein, the cellular levels of PKB, p70^S6k, p42^MAPK, p44^MAPK and p90 ^rsk were not affected by the cytokines. IGF-I caused the phosphorylation of PKB (an approximate 8-fold increase above the basal value after 40 min of IGF-I treatment), p42^MAPK (a 2.81-fold increase after 50 min), and the activation of p70^S6k and p90 ^rsk , manifesting as gel mobility retardation. In cells differentiated in the presence of TNF-α or IFN-γ, this IGF-I-mediated PKB and p70^S6k phosphorylation was significantly diminished, and the increase in p42^MAPK and p90 ^rsk phosphorylation was prevented. The basal p42^MAPK phosphorylation in C2C12 cells treated with IFN-γ was high and comparable with the activation of this kinase by IGF-I. Pretreatment of myogenic cells with IL-1β did not modify the IGF-I-stimulated phosphorylation of PKB, p70^S6k, p42^MAPK and p90 ^rsk . In conclusion: i) TNF-α and IFN-γ, but not IL-1β, if present in the extracellular environment during C2C12 myoblast differentiation, prevent the stimulatory action of IGF-I on protein synthesis. ii) TNF-α- and IFN-γ-induced IGF-I resistance of protein synthesis could be associated with the decreased phosphorylation of PKB and p70^S6k. iii) The activation of glucose uptake in C2C12 myogenic cells treated with IFN-γ is PKB independent. iv) The similar effects of TNF-α and IFN-γ on the signalling and action of IGF-I on protein synthesis in myogenic cells could suggest the involvement of both of these cytokines in protein loss in skeletal muscle.     

Association of α-subunits with nucleotide-modified β-subunits induces asymmetry in the catalytic sites of the F1-ATPase α3β3

The photoaffinity spin-labeled ATP analog, 2-N₃-SL-adenosine triphosphate (ATP), was used to covalently modify isolated β-subunits from F₁-ATPase of the thermophilic bacterium PS3. Approximately 1.2 mol of the nucleotide analog bound to the isolated subunit in the dark. Irradiation leads to covalent incorporation of the nucleotide into the binding site. ESR spectra of the complex show a signal that is typical for protein-immobilized radicals. Addition of isolated α-subunits to the modified β-subunits results in ESR spectra with two new signals indicative of two distinctly different environments of the spin-label, e.g., two distinctly different conformations of the catalytic sites. The relative ratio of the signals is approx 2∶1 in favor of the more closed conformation. The data show for the first time that when nucleotides are bound to isolated β-subunits, binding of α-subunits induces asymmetry in the catalytic sites even in the absence of the γ-subunit. This work was supported by a grant from the Deutsche Forschungsgemeinschaft to PDV.     

Characterization of glycerol dehydratase expressed by fusing its α- and β-subunits

The gdh and gdhr genes, encoding B₁₂-dependent glycerol dehydratase (GDH) and glycerol dehydratase reactivase (GDHR), respectively, in Klebsiella pneumoniae, were cloned and expressed in E. coli. Part of the β-subunit was lost during GDH purification when co-expressing α, β and γ subunit. This was overcome by fusing the β-subunit to α- or γ-subunit with/without the insertion of a linker peptide between the fusion moieties. The kinetic properties of the fusion enzymes were characterized and compared with wild type enzyme. The results demonstrated that the fusion protein GDHALB/C, constructed by linking the N-terminal of β-subunit to the C-terminal of α subunit through a (Gly₄Ser)₄ linker peptide, had the greatest catalytic activity. Similar to the wild-type enzyme, GDHALB/C underwent mechanism-based inactivation by glycerol during catalysis and could be reactivated by GDHR. Electronic supplementary material  The online version of this article (doi:) contains supplementary material, which is available to authorized users.     

Initial analysis of the molecular image of pamlin, a sea urchin cell adhesion protein, by transmission electron microscopy

Pamlin, an important extracellular protein required early for sea urchin embryogenesis, is readily isolated from the embryos of Hemicentrotus pulcherrimus . A molecular image analysis of pamlin was conducted using immuno-electron microscopy, rotary shadowing and negative staining technique-applied electron microscopy. The electron microscopy showed that a monoclonal antibody to the pamlin α-subunit bound to a position 13.5 nm from one end of a purified 255 kDa pamlin molecule, which is a 132 nm long and 6.8 nm wide linear structure. The pamlin structure is composed of three subunits, a 47 nm long 52 kDa α-subunit that attaches to one end of a 105 nm long 180 kDa β-subunit, and a 15.6 nm diameter globular 23 kDa γ-subunit that binds to the middle of the β-subunit. The α- and β-subunits together form a 125–140 nm linear structure. Intermolecular aggregation frequently occurred between the free end of two β-subunits of the αβγ pamlin molecule, leaving the entire α-subunit surface free. Occasionally associations between the ends of α-subunits, or between an α-subunit and the middle of a β-subunit also occurred, but no aggregations of pamlin formed through the γ-subunit. These homophilic molecular aggregations of pamlin formed a large supramolecular network. In addition, the single pamlin molecule rounded at one end under high calcium ion concentration to form a 'loop', suggesting the presence of a calcium sensitive region in the molecule.     

Polarized membrane distribution of potassium-dependent ion pumps in epithelial cells: Different roles of the <Emphasis Type="Italic">N</Emphasis>-glycans of their <Emphasis Type="Italic">β</Emphasis> subunits

The Na,K-ATPases and the H,K-ATPases are two potassium-dependent homologous heterodimeric P₂-type pumps that catalyze active transport of Na⁺ in exchange for K⁺ (Na,K-ATPase) or H⁺ in exchange for K⁺ (H,K-ATPase). The ubiquitous Na,K-ATPase maintains intracellular ion balance and membrane potential. The gastric H,K-ATPase is responsible for acid secretion by the parietal cell of the stomach. Both pumps consist of a catalytic α-subunit and a glycosylated β-subunit that is obligatory for normal pump maturation and trafficking. Individual N-glycans linked to the β-subunits of the Na,K-ATPase and H,K-ATPase are important for stable membrane integration of their respective α subunits, folding, stability, subunit assembly, and enzymatic activity of the pumps. They are also essential for the quality control of unassembled β-subunits that results in either the exit of the subunits from the ER or their ER retention and subsequent degradation. Overall, the importance of N-glycans for the␣maturation and quality control of the H,K-ATPase is greater than that of the Na,K-ATPase. The roles of individual N-glycans of the β-subunits in the post-ER trafficking, membrane targeting and plasma membrane retention of the Na,K-ATPase and H,K-ATPase are different. The Na,K-ATPase β ₁-subunit is the major β-subunit isoform in cells with lateral location of the pump. All three N-glycans of the Na,K-ATPase β ₁-subunit are important for the lateral membrane retention of the pump due to glycan-mediated interaction between the β ₁-subunits of the two neighboring cells in the cell monolayer and cytosolic linkage of the α-subunit to the cytoskeleton. This intercellular β ₁–β ₁ interaction is also important for formation of cell–cell contacts. In contrast, the N-glycans unique to the Na,K-ATPase β ₂-subunit,which has up to eight N-glycosylation sites, contain apical sorting information. This is consistent with the apical location of the Na,K-ATPase in normal and malignant epithelial cells with high abundance of the β ₂-subunit. Similarly, all seven N-glycans of the gastric H,K-ATPase β-subunit determine apical sorting of this subunit. Supported in part by NIH grants DK46917, DK58333, D53642, and USVA     

On the prosthetic groups of the NiFe sulfhydrogenase from Pyrococcus furiosus: topology, structure, and temperature-dependent redox chemistry

 The sulfhydrogenase complex of Pyrococcus furiosus is an αβγδ heterotetramer with both hydrogenase activity (borne by the αδ subunits) and sulfur reductase activity (carried by the βγ subunits). The β-subunit contains at least two [4Fe-4S] cubanes and the γ-subunit contains one [2Fe-2S] cluster and one FAD molecule. The δ-subunit contains three [4Fe-4S] cubanes and the α-subunit carries the NiFe dinuclear center. Only three Fe/S signals are observed in EPR-monitored reduction by dithionite, NADPH, or internal substrate upon heating. All other clusters presumably have reduction potentials well below that of the H⁺/H₂ couple. Heat-induced reduction by internal substrate allows, for the first time, EPR monitoring of the NiFe center in a hyperthermophilic hydrogenase, which passes through a number of states, some of which are similar to states previously defined for mesophilic hydrogenases. The complexity of the observed transitions reflects a combination of temperature-dependent activation and temperature-dependent reduction potentials. Received: 10 December 1998 / Accepted: 16 February 1999     

Irregular activity oscillations of a rotary molecular motor: A simple kinetic model of F<Subscript>1</Subscript>-ATPase

F₁-ATPase is a catalytic part of the F₁F_o-ATP synthase molecular motor. The cooperative hydrolysis of ATP at three catalytic sites of F₁-ATPase is accompanied by the rotation of the central γ-subunit inside a cylinder formed by three α-subunits and three β-subunits. Experimental works of different authors have shown that the γ-subunit rotates with irregular dwells. A simple kinetic model suggested in this article provides an explanation as to why dwells occur during the rotation of F₁-ATPase. According to this model, rotation dwells happen as a result of deterministic chaos, which in turn occurs at rate constants that are close to those demonstrated experimentally. The time duration of dwells in the model is in agreement with that observed experimentally. Our model explains the known irregular occupancy of catalytic sites of F₁-ATPase by nucleotides.     

Tyrosine Kinase Phosphorylation of GABAA Receptor α1, β2 and γ2 Subunits Following Chronic Intermittent Ethanol (CIE) Exposure of Cultured Cortical Neurons of Mice

There is evidence that many of the GABA_A receptor subunits contain consensus sequence for tyrosine kinase, and phosphorylation may play a key role in ethanol’s regulation of GABA_A receptors. Recently, we investigated the effect of chronic exposure of ethanol (CE) on tyrosine kinase phosphorylation and reported that there was an up-regulation in tyrosine kinase phosphorylation of the β₂- and γ₂- subunits and no effect on α₁-subunit of the GABA_A receptor in the cultured cortical neurons of mice. In the present study, we have further investigated the effect of chronic intermittent administration of ethanol (CIE) on tyrosine kinase phosphorylation of the GABA_A receptor subunits (α₁, β₂, and γ₂) in the mouse cultured cortical neurons by immunoprecipitation and Western blot techniques. We observed that there was an up-regulation in the tyrosine kinase phosphorylation of the GABA_A receptor β₂- and γ₂-subunits following CIE exposure, and no effect on α₁-subunit in the cultured cortical neurons of mice. These CIE changes, unlike CE, were not reverted back to the control level following ethanol withdrawal even after 7 days. Acute exposure of ethanol did not cause any change in the tyrosine kinase regulation of the GABA_A receptor subunits. In conclusion, the CIE exposure, unlike chronic/acute ethanol exposure, regulates the tyrosine kinase phosphorylation of the selective population of GABA_A receptors in a long lasting manner.     

NMR Based Metabonomics Study of DAG Treatment in a C2C12 Mouse Skeletal Muscle Cell Line Myotube Model of Burn-Injury

After severe burn injury, proinflammatory cytokine levels are elevated in serum and skeletal muscle, which in turn increases protein breakdown and decreases protein synthesis. In this study, C2C12 mouse skeletal muscle cell line myotubes were exposed to proinflammatory cytokines tumor necrosis factor-α (TNF-α) and interferon-γ (IFN-γ) as an in vitro cell-line model of catabolic response to burn injury and then treated with des-acyl ghrelin (DAG), a 28 amino acid polypeptide hormone thought to inhibit protein breakdown and increase protein synthesis, to assess its therapeutic potential. Nuclear magnetic resonance-based metabonomics was used to monitor metabolic activity of C2C12 myotubes under four treatment conditions: (1) control, (2) TNF-α/IFN-γ (TI), (3) DAG (DA), and (4) TNF-α/IFN-γ followed by DAG (TIDA) to assess the effect of DAG treatment on cellular metabolic response during basal or catabolic conditions. Twelve metabolites showed significant changes in concentrations following treatments in the hydrophilic cell extracts. Lactate (P < 10⁻⁴) and citrulline (P < 10⁻⁹) increased with TNF-α/IFN-γ treatment, indicating increased protein degradation, and returned to control levels in the TIDA group. Adenosine nucleotide levels had decreased trends in TI myotubes that returned to baseline levels after DAG treatment (P < 10⁻⁴). Guanidinoacetate and pantothenate, metabolites involved in protein synthesis and cell proliferation, had increased concentration trends following DAG treatment in both the DA and TIDA groups. Our metabonomics analysis provides further evidence that DAG counteracts the catabolic response caused by elevated muscle TNF-α/IFN-γ cytokine levels following severe burns and can play a potential therapeutic role in treatment of burn injury.     

The mouse adducin gene family: alternative splicing and chromosomal localization

Mouse cDNA sequences encoding α, β, and γ adducins were cloned from a mouse reticulocyte cDNA library. The purified clones contain alternatively spliced exons from all three adducin genes. In the case of α and β, the inclusion of the alternatively spliced exons results in truncated polypeptide isoforms (called α-2 and β-2). The mouse predicted amino acid sequences are compared with published rat and human sequences. For completion of this comparison, cDNA encoding the rat β-1 carboxy terminus was cloned by PCR. The carboxy terminal region containing MARCKS homology, calmodulin-binding region-2, and spectrin-actin-binding site, is conserved among α-1, β-1, and γ-1 isoforms in mouse, rat, and humans. We also report here the localization of the gene encoding γ adducin (Add3) to murine Chr 19, in a region that shows conserved synteny with human Chr 10. Received: 1 June 1999 / Accepted: 25 August 1999     

Cellular localisation by immunolabelling and transmission electron microscopy of oxaloacetate decarboxylase or its individual subunits synthesised in Escherichia coli

Abstract The genes oadGAB encoding the oxaloacetate decarboxylase γ, α and β-subunits from Klebsiella pneumoniae were expressed in Escherichia coli . Using different expression vectors, the entire enzyme or its individual subunits were synthesised. The expression was evidenced immunologically in whole cells with polyclonal antibodies raised against the purified oxaloacetate decarboxylase. The expressed α-subunit or a combination of a and β-subunits were shown to reside in the cytoplasm, while the entire oxaloacetate decarboxylase or a γα-complex were located mostly in the cytoplasmic membrane. Interestingly, overexpression of the γα-complex or the entire oxaloacetate decarboxylase in E. coli led to a significant immunogold labelling in the cytoplasm, indicating that the a-subunit was not completely complexed to the membrane-bound γ or βγ-subunits.