Search

Insulin rapidly stimulates GLUT4 translocation and glucose transport in fat and muscle cells. Signals from the occupied insulin receptor are translated into downstream signalling changes in serine/threonine kinases within timescales of seconds, and this is followed by delivery and accumulation of the glucose transporter GLUT4 at the plasma membrane. Kinetic studies have led to realisation that there are distinct phases of this stimulation by insulin. There is a rapid initial burst of GLUT4 delivered to the cell surface from a subcellular reservoir compartment and this is followed by a steady-state level of continuing stimulation in which GLUT4 recycles through a large itinerary of subcellular locations. Here, we provide an overview of the phases of insulin stimulation of GLUT4 translocation and the molecules that are currently considered to activate these trafficking steps. Furthermore, we suggest how use of new experimental approaches together with phospho-proteomic data may help to further identify mechanisms for activation of these trafficking processes., (© 2022 The Author(s).)

The regulation of proteasome activity is essential to cellular homeostasis and defects have been implicated in various disorders including Parkinson disease. The F-box protein FBXO7 has been implicated in early-onset parkinsonism and has previously been shown to have a regulatory role in proteasome activity and assembly. Here, we report the association of the E3 ubiquitin ligase FBXO7-SCF (SKP1, cullin-1, F-box protein) with the BAG6 complex, consisting of the subunits BAG6, GET4 and UBL4A. We identify the subunit GET4 as a direct interactor of FBXO7 and we show that the subunits GET4 and UBL4A are required for proper proteasome activity. Our findings demonstrate reduced binding of FBXO7 variants to GET4 and that FBXO7 variants bring about reduced proteasome activity. In addition, we find that GET4 is a non-proteolytic substrate of FBXO7, that binding of GET4 to BAG6 is enhanced in the presence of active FBXO7-SCF and that the cytoplasmic localization of the BAG6 complex is dependent on the E3 ubiquitin ligase activity. Taken together, our study shows that the parkinsonism-associated FBXO7 cooperates with the BAG6 complex in proteasome function and determines the subcellular localization of this complex., (© 2021 The Author(s). Published by Portland Press Limited on behalf of the Biochemical Society.)

Periplasmic PPIases (peptidylprolyl cis-trans isomerases) catalyse the cis-trans isomerization of peptidyl-prolyl bonds, which is a rate-limiting step during protein folding. We demonstrate that the surA, ppiA, ppiD, fkpA and fklB alleles each encode a periplasmic PPIase in the bacterial pathogen Yersinia pseudotuberculosis. Of these, four were purified to homogeneity. Purified SurA, FkpA and FklB, but not PpiD, displayed detectable PPIase activity in vitro. Significantly, only Y. pseudotuberculosis lacking surA caused drastic alterations to the outer membrane protein profile and FA (fatty acid) composition. They also exhibited aberrant cellular morphology, leaking LPS (lipopolysaccharide) into the extracellular environment. The SurA PPIase is therefore most critical for maintaining Y. pseudotuberculosis envelope integrity during routine culturing. On the other hand, bacteria lacking either surA or all of the genes ppiA, ppiD, fkpA and fklB were sensitive to hydrogen peroxide and were attenuated in mice infections. Thus Y. pseudotuberculosis exhibits both SurA-dependent and -independent requirements for periplasmic PPIase activity to ensure in vivo survival and a full virulence effect in a mammalian host., (© The Authors Journal compilation © 2011 Biochemical Society)

Precise and dynamic measurement of intracellular metabolite levels has been hampered by difficulties in differentiating between adsorbed and imported fractions and the subcellular distribution between cytosol, endomembrane compartments and mitochondria. In the present study, genetically encoded FRET (Förster resonance energy transfer)-based sensors were deployed for dynamic measurements of free cytosolic glucose and ATP with varying external supply and in glucose-transport mutants. Moreover, by using the FRET sensors in a microfluidic platform, we were able to monitor in vivo changes of intracellular free glucose in individual yeast cells. We demonstrate the suitability of the FRET sensors for gaining physiological insight by demonstrating that free intracellular glucose and ATP levels are reduced in a hxt5Δ hexose-transporter mutant compared with wild-type and other hxtΔ strains.

ZnT2 (zinc transporter-2) expression is restricted to tissues with unique zinc requirements such as mammary and prostate glands. We previously determined that ZnT2 plays a major role in zinc export from mammary glands, as women with a mutation in the gene encoding ZnT2 (SLC30A2) had an approximately 75% reduction in milk zinc concentration. Two distinct human ZnT2 isoforms (approximately 42 and 35 kDa) are predicted to result from alternative splicing of SLC30A2. We examined the localization and function of each ZnT2 isoform, in cells generated to express ZnT2-HA (haemagglutinin) fusion proteins. The 42 kDa isoform was localized primarily to the endosomal/secretory compartment and overexpression resulted in increased zinc vesicularization. In contrast, the 35 kDa isoform is associated with the plasma membrane. Importantly, zinc transport was higher in cells over-expressing each isoform, indicating that both proteins are functional. Endogenous expression of the secretory vesicle-associated ZnT2 isoform predominates in mammary cells and expression is higher in secreting cells, whereas the smaller isoform plays a minor role in zinc export, directly reflecting the secretory function of the mammary gland. Together our data shed further light on the complex integration of cellular zinc transport mechanisms, which may be facilitated by multiple isoforms of specific zinc transporters with unique cellular functions.

Until recently, a modest number of approx. 40 lysosomal membrane proteins had been identified and even fewer were characterized in their function. In a proteomic study, using lysosomal membranes from human placenta we identified several candidate lysosomal membrane proteins and proved the lysosomal localization of two of them. In the present study, we demonstrate the lysosomal localization of the mouse orthologue of the human C1orf85 protein, which has been termed kidney-predominant protein NCU-G1 (GenBank accession number: AB027141). NCU-G1 encodes a 404 amino acid protein with a calculated molecular mass of 39 kDa. The bioinformatics analysis of its amino acid sequence suggests it is a type I transmembrane protein containing a single tyrosine-based consensus lysosomal sorting motif at position 400 within the 12-residue C-terminal tail. Its lysosomal localization was confirmed using immunofluorescence with a C-terminally His-tagged NCU-G1 and the lysosomal marker LAMP-1 (lysosome-associated membrane protein-1) as a reference, and by subcellular fractionation of mouse liver after a tyloxapol-induced density shift of the lysosomal fraction using an anti-NCU-G1 antiserum. In transiently transfected HT1080 and HeLa cells, the His-tagged NCU-G1 was detected in two molecular forms with apparent protein sizes of 70 and 80 kDa, and in mouse liver the endogenous wild-type NCU-G1 was detected as a 75 kDa protein. The remarkable difference between the apparent and the calculated molecular masses of NCU-G1 was shown, by digesting the protein with N-glycosidase F, to be due to an extensive glycosylation. The lysosomal localization was impaired by mutational replacement of an alanine residue for the tyrosine residue within the putative sorting motif.

Although bystin has been identified as a protein potentially involved in embryo implantation (a process unique to mammals) in humans, the bystin gene is evolutionarily conserved from yeast to humans. DNA microarray data indicates that bystin is overexpressed in human cancers, suggesting that it promotes cell growth. We undertook RT (reverse transcription)-PCR and immunoblotting, and confirmed that bystin mRNA and protein respectively are expressed in human cancer cell lines, including HeLa. Subcellular fractionation identified bystin protein as nuclear and cytoplasmic, and immunofluorescence showed that nuclear bystin localizes mainly in the nucleolus. Sucrose gradient ultracentrifugation of total cytoplasmic ribosomes revealed preferential association of bystin with the 40S subunit fractions. To analyse its function, bystin expression in cells was suppressed by RNAi (RNA interference). Pulse-chase analysis of ribosomal RNA processing suggested that bystin knockdown delays processing of 18S ribosomal RNA, a component of the 40S subunit. Furthermore, this knockdown significantly inhibited cell proliferation. Our findings suggest that bystin may promote cell proliferation by facilitating ribosome biogenesis, specifically in the production of the 40S subunit. Localization of bystin to the nucleolus, the site of ribosome biogenesis, was blocked by low concentrations of actinomycin D, a reagent that causes nucleolar stress. When bystin was transiently overexpressed in HeLa cells subjected to nucleolar stress, nuclear bystin was included in particles different from the nuclear stress granules induced by heat shock. In contrast, cytoplasmic bystin was barely affected by nucleolar stress. These results suggest that, while bystin may play multiple roles in mammalian cells, a conserved function is to facilitate ribosome biogenesis required for cell growth.

Biogenesis and recycling of synaptic vesicles are accompanied by sorting processes that preserve the molecular composition of the compartments involved. In the present study, we have addressed the targeting of synaptobrevin 2/VAMP2 (vesicle-associated membrane protein 2), a critical component of the synaptic vesicle--fusion machinery, in a heterotypic context where its sorting is not confounded by the presence of other neuron-specific molecules. Ectopically expressed synaptophysin I interacts with VAMP2 and alters its default surface targeting to a prominent vesicular distribution, with no effect on the targeting of other membrane proteins. Protein-protein interaction is not sufficient for the control of VAMP2 sorting, which is mediated by the C-terminal domain of synaptophysin I. Synaptophysin I directs the sorting of VAMP2 to vesicles before surface delivery, without influencing VAMP2 endocytosis. Consistent with this, dynamin and alpha-SNAP (soluble N-ethylmaleimide-sensitive fusion protein-attachment protein) mutants which block trafficking at the plasma membrane do not abrogate the effect of synaptophysin I on VAMP2 sorting. These results indicate that the sorting determinants of synaptic vesicle proteins can operate independently of a neuronal context and implicate the association of VAMP2 with synaptophysin I in the specification of the pathway of synaptic vesicle biogenesis.

The NET [noradrenaline (norepinephrine) transporter], an Na+/Cl--dependent neurotransmitter transporter, has several isoforms produced by alternative splicing in the C-terminal region, each differing in expression and function. We characterized the two major isoforms of human NET, hNET1, which has seven C-terminal amino acids encoded by exon 15, and hNET2, which has 18 amino acids encoded by exon 16, by site-directed mutagenesis in combination with NE (noradrenaline) uptake assays and cell surface biotinylation. Mutants lacking one third or more of the 24 amino acids encoded by exon 14 exhibited neither cell surface expression nor NE uptake activity, with the exception of the mutant lacking the last eight amino acids of hNET2, whose expression and uptake resembled that of the WT (wild-type). A triple alanine replacement of a candidate motif (ENE) in this region mimicked the influences of the truncation. Deletion of either the last three or another four amino acids of the C-terminus encoded by exon 15 in hNET1 reduced the cell surface expression and NE uptake, whereas deletion of all seven residues reduced the transport activity but did not affect the cell surface expression. Replacement of RRR, an endoplasmic reticulum retention motif, by alanine residues in the C-terminus of hNET2 resulted in a similar expression and function compared with the WT, while partly recovering the effects of the mutation of ENE. These findings suggest that in addition to the function of the C-terminus, the common proximal region encoded by exon 14 regulates the functional expression of splice variants, such as hNET1 and hNET2.

Our previous studies have demonstrated de novo haem biosynthesis in the malarial parasite (Plasmodium falciparum and P. berghei). It has also been shown that the first enzyme of the pathway is the parasite genome-coded ALA (delta-aminolaevulinate) synthase localized in the parasite mitochondrion, whereas the second enzyme, ALAD (ALA dehydratase), is accounted for by two species: one species imported from the host red blood cell into the parasite cytosol and another parasite genome-coded species in the apicoplast. In the present study, specific antibodies have been raised to PfFC (parasite genome-coded ferrochelatase), the terminal enzyme of the haem-biosynthetic pathway, using recombinant truncated protein. With the use of these antibodies as well as those against the hFC (host red cell ferrochelatase) and other marker proteins, immunofluorescence studies were performed. The results reveal that P. falciparum in culture manifests a broad distribution of hFC and a localized distribution of PfFC in the parasite. However, PfFC is not localized to the parasite mitochondrion. Immunoelectron-microscopy studies reveal that PfFC is indeed localized to the apicoplast, whereas hFC is distributed in the parasite cytoplasm. These results on the localization of PfFC are unexpected and are at variance with theoretical predictions based on leader sequence analysis. Biochemical studies using the parasite cytosolic and organellar fractions reveal that the cytosol containing hFC accounts for 80% of FC enzymic activity, whereas the organellar fraction containing PfFC accounts for the remaining 20%. Interestingly, both the isolated cytosolic and organellar fractions are capable of independent haem synthesis in vitro from [4-14C]ALA, with the cytosol being three times more efficient compared with the organellar fraction. With [2-14C]glycine, most of the haem is synthesized in the organellar fraction. Thus haem is synthesized in two independent compartments: in the cytosol, using the imported host enzymes, and in the organellar fractions, using the parasite genome-coded enzymes.

EEN (extra eleven nineteen), also known as EA2 (endophilin A2), a fusion partner of the MLL (mixed-lineage leukaemia) gene in human acute leukaemia, is a member of the endophilin A family, involved in the formation of endocytic vesicles. We present evidence to show that EEN/EA2 is localized predominantly in nuclei of various cell lines of haemopoietic, fibroblast and epithelial origin, in contrast with its reported cytoplasmic localization in neurons and osteoclasts, and that EEN/EA2 exhibits nucleocytoplasmic shuttling. During the cell cycle, EEN/EA2 shows dynamic localization: it is perichromosomal in prometaphase, co-localizes with the bipolar spindle in metaphase and anaphase and redistributes to the midzone and midbody in telophase. This pattern of distribution coincides with changes in protein levels of EEN/EA2, with the highest levels being observed in G2/M-phase. Our results suggest that distinct subcellular localization of the endophilin A family members probably underpins their diverse cellular functions and indicates a role for EEN/EA2 in the cell cycle.

Research on metabolism of nucleotides and their derivatives has gained increasing interest in the recent past. This includes de novo synthesis, analysis of salvage pathways, breakdown and transport of nucleotides, nucleosides and nucleobases. To perform a further step towards the analysis of nucleoside transport in Arabidopsis, we incubated leaf discs with various radioactively labelled nucleosides. Leaf cells imported labelled nucleosides and incorporated these compounds into RNA, but not into DNA. Furthermore, we report on the biochemical properties of three so far uncharacterized members of the Arabidopsis ENT (equilibrative nucleoside transporter) family (AtENT4, AtENT6 and AtENT7). After heterologous expression in yeast, all three proteins exhibited broad substrate specificity and transported the purine nucleosides adenosine and guanosine, as well as the pyrimidine nucleosides cytidine and uridine. The apparent K(m) values were in the range 3-94 microM, and transport was inhibited most strongly by deoxynucleosides, and to a smaller extent by nucleobases. Typical inhibitors of mammalian ENT proteins, such as dilazep and NBMPR (nitrobenzylmercaptopurine ribonucleoside, also known as nitrobenzylthioinosine) surprisingly exerted almost no effect on Arabidopsis ENT proteins. Transport mediated by the AtENT isoforms differed in pH-dependency, e.g. AtENT7 was not affected by changes in pH, AtENT3, 4 and 6 exhibited a less pronounced pH-dependency, and AtENT1 activity was clearly pH-dependent. Using a GFP (green fluorescent protein)-fusion protein transiently expressed in tobacco leaf protoplasts, a localization of AtENT6 in the plant plasma membrane has been revealed.

Using EPAC1 (exchange protein directly activated by cAMP 1) as bait in two-hybrid screens of foetal and adult human brain libraries, we identified the LC2 (light chain 2) of MAP1A (microtubule-associated protein 1A) as a protein capable of interaction with EPAC1. We applied an immunoprecipitation assay to demonstrate protein interaction between EPAC1 and LC2 in co-transfected human embryonic kidney 293 cells. EPAC2 also co-immunoprecipitated with LC2 from extracts of rat cerebellum. Immunolocalization in co-transfected human embryonic kidney 293 cells revealed that EPAC1 co-localizes with LC2 throughout the cell body. We found that endogenous EPAC2 is also immunolocalized with LC2 in PC12 cells. Immunolocalization of EPAC1 in transfected COS1 cells showed that EPAC1 is associated with the perinuclear region surrounding the nucleus and filamentous structures throughout the cell. Removal of the cAMP-binding domain of EPAC1 (DeltacAMP-EPAC1) appeared to disrupt targeting of EPAC1 in cells resulting in a more dispersed staining pattern. Using two-hybrid assay, we tested the ability of LC2 to interact with DeltacAMP-EPAC1 and DeltaDEP-EPAC1, which lacks a DEP domain (dishevelled, Egl-10 and pleckstrin homology domain). We found that deletion of the cAMP-binding domain inhibited interaction between EPAC1 and LC2 in a two-hybrid assay, but removal of the DEP domain had little effect. LC2 was found to interact with a glutathione-S-transferase-fusion protein of the cAMP-binding domain of EPAC1 in a pull-down assay, but not the DEP, REM (Ras exchange motif) or CAT (catalytic) domains. Together with our two-hybrid results, this suggests that the cAMP-binding domain of EPAC1 mediates interaction with LC2.

Solute carrier family 11 member 1 (Slc11a1, formerly Nramp1) is a highly glycosylated, 12 transmembrane domain protein expressed in macrophages. It resides in the membrane of late endosomes and lysosomes, where it functions as a bivalent cation transporter. Mice susceptible to infection by various intracellular pathogens including Leishmania donovani and Salmonella typhimurium carry a glycine to aspartic acid substitution at position 169 (G169D, Gly(169)-->Asp), within transmembrane domain 4 of Slc11a1. To investigate the molecular pathogenesis of infectious disease susceptibility, we compared the behaviour of heterologously and endogenously expressed wild-type and mutant Slc11a1 by immunofluorescence, immunoelectron microscopy and Western-blot analysis. We found occasional late endosome/lysosome staining of mutant protein using immunoelectron microscopy, but most of the mutant Slc11a1 was retained within the ER (endoplasmic reticulum). Using glycosylation as a marker for protein maturation in two independent heterologous expression systems, we found that most mutant Slc11a1 existed as an ER-dependent, partially glycosylated intermediate species. Correct endosomal targeting of wild-type Slc11a1 continued despite disruption of N-glycosylation sites, indicating that glycosylation did not influence folding or sorting. We propose that the G169D mutation causes localized misfolding of Slc11a1, resulting in its retention in the ER and manifestation of the loss of function phenotype.

We examined the function of GFP-IP(3)R3 (green fluorescent protein-tagged inositol 1,4,5-trisphosphate receptor type 3) in Ca(2+) release and entry using a mutant DT40 cell line (IP(3)R-KO) in which all three IP(3)R genes had been disrupted. GFP-IP(3)R3 fluorescence largely overlapped with the distribution of endoplasmic reticulum, whereas a portion of GFP-IP(3)R3 apparently co-localized with the plasma membrane. The application of IP(3) to permeabilized WT (wild-type) DT40 cells induced Ca(2+) release from internal stores. Although this did not occur in IP(3)R-KO cells it was restored by expression of GFP-IP(3)R3. In intact cells, application of anti-IgM, an activator of the BCR (B-cell receptor), or trypsin, a protease-activated receptor 2 agonist, did not cause any Ca(2+) response in IP(3)R-KO cells, whereas these treatments induced oscillatory or transient Ca(2+) responses in GFP-IP(3)R3-expressing IP(3)R-KO cells, as well as in WT cells. In addition, BCR activation elicited Ca(2+) entry in WT and GFP-IP(3)R3-expressing IP(3)R-KO cells but not in IP(3)R-KO cells. This BCR-mediated Ca(2+) entry was observed in the presence of La(3+), which blocks capacitative Ca(2+) entry. Thapsigargin depleted Ca(2+) stores and led to Ca(2+) entry in IP(3)R-KO cells irrespective of GFP-IP(3)R3 expression. In contrast with BCR stimulation, thapsigargin-induced Ca(2+) entry was completely blocked by La(3+), suggesting that the BCR-mediated Ca(2+) entry pathway is distinct from the capacitative Ca(2+) entry pathway. The present study demonstrates that GFP-IP(3)R3 could compensate for native IP(3)R in both IP(3)-induced Ca(2+) release and BCR-mediated Ca(2+) entry.

MARCKS (myristoylated alanine-rich C kinase substrate) is a major cytoskeletal protein substrate of PKC (protein kinase C) whose cellular functions are still unclear. However numerous studies have implicated MARCKS in the stabilization of cytoskeletal structures during cell differentiation. The present study was performed to investigate the potential role of Ca(2+)-dependent proteinases (calpains) during myogenesis via proteolysis of MARCKS. It was first demonstrated that MARCKS is a calpain substrate in vitro. Then, the subcellular expression of MARCKS was examined during the myogenesis process. Under such conditions, there was a significant decrease in MARCKS expression associated with the appearance of a 55 kDa proteolytic fragment at the time of intense fusion. The addition of calpastatin peptide, a specific calpain inhibitor, induced a significant decrease in the appearance of this fragment. Interestingly, MARCKS proteolysis was dependent of its phosphorylation by the conventional PKCalpha. Finally, ectopic expression of MARCKS significantly decreased the myoblast fusion process, while reduced expression of the protein with antisense oligonucleotides increased the fusion. Altogether, these data demonstrate that MARCKS proteolysis is necessary for the fusion of myoblasts and that cleavage of the protein by calpains is involved in this regulation.

The location of sarco/endoplasmic-reticulum calcium ATPase (SERCA) retention/retrieval motifs in the sequence of the SERCA1 has been investigated by examining the subcellular location in COS-7 cells of enhanced-green-fluorescent-protein-tagged calcium-pump chimaeras. These chimaeras have been constructed from the fast-twitch SERCA1 and the plasma-membrane calcium ATPase PMCA3. The N-terminal, central and C-terminal segments of these calcium pumps were exchanged between SERCA1 and PMCA3. The segments exchanged correspond to residues 1-211, 212-711 and 712-994 of SERCA1, and residues 1-264, 265-788 and 789-1159 of PMCA3 respectively. Only chimaeras containing the N-terminal segment of SERCA1 were located in the endoplasmic reticulum (ER), whereas chimaeras containing the N-terminal segment from PMCA3 were able to escape from the ER and enter the endomembrane pathway en route for the plasma membrane. Co-localization of SERCA1 in COS-7 cells with the ER/Golgi-intermediate compartment marker ERGIC53 indicates that SERCA1 is maintained in the ER by a process of retrieval. These results indicate that the N-terminal region of SERCA1, containing transmembrane helices M1 and M2, contains an ER-retrieval signal.

The ryanodine receptor, a Ca(2+)-releasing channel in sarcoplasmic reticulum (SR), plays an important role in the excitation-contraction coupling of skeletal muscle. In a previous study [Hirata, Nakahata and Ohizumi (2000) Mol. Pharmacol. 57, 1235-1242], we reported that mastoparan caused Ca(2+) release through ryanodine receptor from the heavy fraction of SR (HSR) isolated from rabbit skeletal muscle, and that it specifically bound to a 97 kDa protein which was distinct from Ca(2+)-pump or triadin. The present study was undertaken to identify and characterize the 97 kDa mastoparan-binding protein. The 97 kDa protein was purified from solubilized HSR by DEAE-Sepharose column chromatography and preparative SDS/PAGE. The partial amino acid sequence of the purified 97 kDa protein was matched with that of glycogen phosphorylase (GP). The proteolytic cleavage pattern of the 97 kDa protein was identical with that of GP. Furthermore, [(125)I-Tyr(3)]mastoparan specifically bound to GP. Interestingly, mastoparan-induced Ca(2+) release was inhibited by exogenous addition of GP-a, and mastoparan dissociated GP from HSR. These results indicate that the 97 kDa mastoparan-binding protein is GP, which negatively regulates Ca(2+) release from HSR. There may be a functional cross-talk between Ca(2+) release from HSR and glycogenolysis for energy supply mediated through GP in skeletal muscles.

DNA fragmentation is one of the critical steps in apoptosis, which is induced by DNA fragmentation factor (DFF). DFF is composed of two subunits, a 40 kDa caspase-activated nuclease (DFF40) and a 45 kDa inhibitor (DFF45). Recently a novel family of cell-death-inducing DFF45-like effectors (CIDEs) has been identified. Among CIDEs, two from human (CIDE-A and CIDE-B) and three from mouse (CIDE-A, CIDE-B and FSP27) have been reported. In this study human CIDE-3, a novel member of CIDEs, was identified upon sequence analysis of a previously unidentified cDNA that encoded a protein of 238 amino acids. It was shown to be a human homologue of mouse FSP27, and shared homology with the CIDE-N and CIDE-C domains of CIDEs. Apoptosis-inducing activity was clearly shown by DNA-fragmentation assay of the nuclear DNA of CIDE-3 transfected 293T cells. The expression pattern of CIDE-3 was different from that of CIDE-B. As shown by Northern-blot analysis, CIDE-3 was expressed mainly in human small intestine, heart, colon and stomach, while CIDE-B showed strong expression in liver and small intestine and at a lower level in colon, kidney and spleen. Green-fluorescent-protein-tagged CIDE-3 was revealed in some cytosolic corpuscles. Alternative splicing of the CIDE-3 gene was also identified by reverse transcription PCR, revealing that two transcripts, CIDE-3 and CIDE-3alpha, were present in HepG2 and A375 cells. CIDE-3 comprised a full-length open reading frame with 238 amino acids; in CIDE-3alpha exon 3 was deleted and it encoded a protein of 164 amino acids. Interestingly the CIDE-3alpha isoform still kept the apoptosis-inducing activity and showed the same pattern of subcellular localization as CIDE-3. Consistent with its chromosome localization at 3p25, a region associated with high frequency loss of heterozygosity in many tumours, CIDE-3 may play an important role in prevention of tumorigenesis.

The chimaeric protein interleukin-2 (IL-2)-Bax was designed to target and kill specific cell populations expressing the IL-2 receptor. However, it is not well understood how IL-2-Bax causes target cells to die. In the present study, we investigated the pathway of apoptosis evoked by IL-2-Bax and the possible involvement of endogenous Bax in this process. We report here that, upon internalization of IL-2-Bax into target cells, it is localized first mainly in the nucleus, and only later is it translocated to the mitochondria. Similarly, endogenous Bax is also partially localized in the nucleus, and accumulates mainly in this compartment soon after physiological triggering of apoptosis. Despite the fact that Bax has no nuclear localization sequence, our data suggest that Bax has one or more physiological roles and/or substrates within the nucleus. Indeed, a dramatic repression of nuclear Tax protein expression was induced following treatment of HUT-102 cells with IL-2-Bax, similar to what occurs following serum deprivation of these cells. Unexpectedly, induction of apoptosis using IL-2-Bax was preceded by enhanced expression of newly synthesized Bax protein and suppression of Bcl-2. This imbalance between the pro- and anti-apoptotic genes was associated with p53 induction, although IL-2-Bax activity was also evident in cells lacking p53 expression. By studying the mechanism of action of IL-2-Bax, we were able to follow the intrinsic events and their cascade that culminates in cell death. We have shown that the ability of IL-2-Bax to affect the intracellular apoptotic machinery within the target cells, and to cause the cells to die, uses a mechanism similar to that induced following a normal apoptotic signal.

Recently, we established that cellular contractions increase long-chain fatty-acid (FA) uptake by cardiac myocytes. This increase is dependent on the transport function of an 88 kDa membrane FA transporter, FA translocase (FAT/CD36), and, in analogy to skeletal muscle, is likely to involve its translocation from an intracellular pool to the sarcolemma. In the present study, we investigated whether cAMP-dependent signalling is involved in this translocation process. Isoproterenol, dibutyryl-cAMP and the phosphodiesterase (PDE) inhibitor, amrinone, which markedly raised the intracellular cAMP level, did not affect cellular FA uptake, but influenced the fate of intracellular FAs by directing these to mitochondrial oxidation in electrostimulated cardiac myocytes. The PDE inhibitors 3-isobutyl-1-methylxanthine, milrinone and dipyridamole each significantly stimulated FA uptake as well as intracellular cAMP levels, but these effects were quantitatively unrelated. The stimulatory effects of these PDE inhibitors were antagonized by sulpho- N -succinimidylpalmitate, indicating the involvement of FAT/CD36, albeit that the different PDE inhibitors use different molecular mechanisms to stimulate FAT/CD36-mediated FA uptake. Notably, 3-isobutyl-1-methylxanthine and milrinone increased the intrinsic activity of FAT/CD36, possibly through its covalent modification, and dipyridamole induces translocation of FAT/CD36 to the sarcolemma. Elevation of intracellular cGMP, but not of cAMP, by the PDE inhibitor zaprinast did not have any effect on FA uptake and metabolism by cardiac myocytes. The stimulatory effects of PDE inhibitors on cardiac FA uptake should be considered when applying these agents in clinical medicine.

The AU-rich element (ARE) is an important instability determinant for a large number of early-response-gene mRNAs. AREs also mediate the stabilization of certain pro-inflammatory mRNAs, such as tumour necrosis factor (TNF)-alpha and cyclo-oxygenase-2 (COX-2), in response to inflammatory stimuli. To understand how AREs control mRNA stability, it is necessary to identify trans-acting factors. We have purified a new ARE-binding protein and identified it as CArG box-binding factor-A (CBF-A). The amino acid sequence of CBF-A is highly similar to that of the ARE-binding protein AUF1. Recombinant CBF-A bound the COX-2 and TNF-alpha AREs, but not a non-specific control RNA. In contrast, in an electrophoretic-mobility-shift assay (EMSA) of crude RAW 264.7 macrophage-like cell extracts, an antiserum that recognizes both AUF1 and CBF-A failed to supershift complexes formed on the TNF-alpha ARE, but did supershift a complex specific for the COX-2 ARE. CBF-A exists as two isoforms, p37 and p42, that differ by a 47-amino-acid insertion close to the C-terminus. By expressing epitope-tagged isoforms of CBF-A it was shown that the p42 isoform binds the COX-2 ARE in EMSA of crude cell extracts. In a HeLa-cell tetracycline-regulated reporter system, overexpression of the p42 CBF-A isoform resulted in stabilization of a COX-2 ARE reporter mRNA. Epitope-tagged p42 CBF-A expressed in HeLa cells co-immunoprecipitated with endogenous COX-2 mRNA, but not glyceraldehyde-3-phosphate dehydrogenase mRNA, as shown by reverse-transcription PCR. The similarity between CBF-A and AUF1 suggests that CBF-A could be re-named AUF2.

The aldo-keto reductase (AKR) 7 family is composed of the dimeric aflatoxin B(1) aldehyde reductase (AFAR) isoenzymes. In the rat, two AFAR subunits exist, designated rAFAR1 and rAFAR2. Herein, we report the molecular cloning of rAFAR2, showing that it shares 76% sequence identity with rAFAR1. By contrast with rAFAR1, which comprises 327 amino acids, rAFAR2 contains 367 amino acids. The 40 extra residues in rAFAR2 are located at the N-terminus of the polypeptide as an Arg-rich domain that may form an amphipathic alpha-helical structure. Protein purification and Western blotting have shown that the two AFAR subunits are found in rat liver extracts as both homodimers and as a heterodimer. Reductase activity in rat liver towards 2-carboxybenzaldehyde (CBA) was resolved by anion-exchange chromatography into three peaks containing rAFAR1-1, rAFAR1-2 and rAFAR2-2 dimers. These isoenzymes are functionally distinct; with NADPH as cofactor, rAFAR1-1 has a low K(m) and high activity with CBA, whereas rAFAR2-2 exhibits a low K(m) and high activity towards succinic semialdehyde. These data suggest that rAFAR1-1 is a detoxication enzyme, while rAFAR2-2 serves to synthesize the endogenous neuromodulator gamma-hydroxybutyrate (GHB). Subcellular fractionation of liver extracts showed that rAFAR1-1 was recovered in the cytosol whereas rAFAR2-2 was associated with the Golgi apparatus. The distinct subcellular localization of the rAFAR1 and rAFAR2 subunits was confirmed by immunocytochemistry in H4IIE cells. Association of rAFAR2-2 with the Golgi apparatus presumably facilitates secretion of GHB, and the novel N-terminal domain may either determine the targeting of the enzyme to the Golgi or regulate the secretory process. A murine AKR protein of 367 residues has been identified in expressed sequence tag databases that shares 91% sequence identity with rAFAR2 and contains the Arg-rich extended N-terminus of 40 amino acids. Further bioinformatic evidence is presented that full-length human AKR7A2 is composed of 359 amino acids and also possesses an additional N-terminal domain. On the basis of these observations, we conclude that AKR7 proteins can be divided into two subfamilies, one of which is a Golgi-associated GHB synthase with a unique, previously unrecognized, N-terminal domain that is absent from other AKR proteins.

Exogenous application of beta-amyloid (Abeta(25-35), a fragment of Abeta(1-42)) significantly elevated levels of reactive oxygen species (ROS) in C6 astroglioma cells, as measured by confocal microscopic analysis of H(2)O(2)-sensitive 2',7'-dichlorofluorescin fluorescence. Subsequent characterization of the signalling pathway revealed that expression of RacN17, a dominant-negative Rac1 mutant, completely blocked Abeta(25-35)-induced generation of ROS, which is indicative of the crucial role played by Rac GTPase in this process. To better understand the downstream mediators affected by Rac, we assessed the degree to which inhibition of cytosolic phospholipase A(2) (cPLA(2)) and 5-lipoxygenase (5-LO) contributed to the response and found that inhibition of either enzyme completely blocked Abeta(25-35)-induced ROS generation, indicating its dependence on arachidonic acid synthesis and metabolism to leukotrienes (e.g. leukotriene B(4)). Consistent with those findings, Abeta(25-35) Rac-dependently stimulated translocation of 5-LO to the nuclear envelope and increased intracellular levels of leukotriene B(4), while exogenous application of leukotriene B(4) increased intracellular H(2)O(2) via BLT, its cell-surface receptor. In addition to the aforementioned downstream mediators, inhibition of phosphoinositide 3-kinase (PI 3-kinase), an enzyme situated upstream of Rac, also completely blocked Abeta(25-35)-induced H(2)O(2) generation. Our findings thus demonstrate that PI 3-kinase, Rac, cPLA(2) and 5-LO are all essential components of the beta-amyloid signaling cascade leading to generation of ROS.

The regulatory role of protein kinase C (PKC) delta isoform in the stimulation of phospholipase D (PLD) by sphingosine-1-phosphate (SPP) in a human-airway epithelial cell line (CFNPE9o(-)) was revealed by using antisense oligodeoxynucleotide to PKCdelta, in combination with the specific inhibitor rottlerin. Cell treatment with antisense oligodeoxynucleotide, but not with sense oligodeoxynucleotide, completely eliminated PKCdelta expression and resulted in the strong inhibition of SPP-stimulated phosphatidic acid formation. Indeed, among the PKCalpha, beta, delta, epsilon and zeta isoforms expressed in these cells, only PKCdelta was activated on cell stimulation with SPP, as indicated by translocation into the membrane fraction. Furthermore, pertussis toxin and genistein eliminated both PKCdelta translocation and PLD activation. In particular, a significant reduction in phosphatidylbutanol formation by SPP was observed in the presence of 4-amino-5-(4-methylphenyl)-7-(t-butyl) pyrazolo [3,4-d] pyrimidine (PP1), an inhibitor of Src tyrosine kinase. Furthermore, the activity of Src kinase was slightly increased by SPP and inhibited by PP1. However, the level of PKCdelta tyrosine phosphorylation was not increased in SPP-stimulated cells, suggesting that Src did not directly phosphorylate PKCdelta. Finally, the level of serine phosphorylation of PLD1 and PLD2 isoforms was not changed, whereas the PLD1 isoform alone was threonine-phosphorylated in SPP-treated cells. PLD1 threonine phosphorylation was strongly inhibited by rottlerin, by anti-PKCdelta oligodeoxynucleotide and by PP1. In conclusion, in CFNPE9o(-) cells, SPP interacts with a membrane receptor linked to a G(i) type of G-protein, leading to activation of PLD, probably the PLD1 isoform, by a signalling pathway involving Src and PKCdelta.

Pex7p is a WD40-containing protein involved in peroxisomal import of proteins containing an N-terminal peroxisome-targeting signal (PTS2). The interaction of human recombinant Pex7p expressed in different hosts/systems with its PTS2 ligand and other peroxins was analysed using various experimental approaches. Specific binding of human Pex7p to PTS2 could be demonstrated only when Pex7p was formed in vitro by a coupled transcription/translation system or synthesized in vivo in Chinese hamster ovary K1 cells transfected with a construct coding for a Pex7p-green fluorescent protein (GFP) fusion protein. Apparently, no cofactors are required and only monomeric Pex7p binds to PTS2. The interaction is reduced upon cysteine alkylation and is impaired upon truncation of the N-terminus of Pex7p. Interaction of Pex7p with other peroxins could not be demonstrated in bacterial or yeast two-hybrid screens, or in pull-down binding assays. The GFP fusion proteins, tagged at either the N- or C-terminus, were able to restore PTS2 import in rhizomelic chondrodysplasia punctata fibroblasts, and Pex7p-GFP was located both in the lumen of peroxisomes and in the cytosol.

Synaptotagmin VII (Syt VII) has been proposed to regulate several different types of Ca2+-dependent exocytosis, but its subcellular localization (lysosome or plasma membrane) and the number of alternative splicing isoforms of Syt VII (single or multiple forms) are matters of controversy. In the present study, we show by reverse transcriptase-PCR analysis that mouse Syt VII has one major isoform (Syt VIIalpha), the original Syt VII, and two minor isoforms (Syt VIIbeta and Syt VIIgamma), which contain unique insertions (of 44 and 116 amino acids respectively) in the spacer domain between the transmembrane and C2 domains of Syt VIIalpha. Similar results were obtained with respect to rat and human Syt VII mRNA expression. An antibody against the N-terminal domain of mouse Syt VII [anti-(Syt VII-N)], which specifically recognized recombinant Syt VII but not other Syt isoforms expressed in COS-7 cells, recognized two major, closely co-migrating bands (p58 and p60) and minor bands of approx. 65 kDa in mouse brain. Immunoaffinity purification of proteins that bind the anti-(Syt VII-N) antibody, and peptide sequence analysis revealed that: (i) the major p58 and p60 bands are identified as adenylate cyclase-associated protein 2; (ii) actin-binding protein is localized at the plasma membrane; and (iii) Syt VIIalpha (65 kDa) is the major Syt VII isoform, but with a much lower expression level than previously thought. It was also shown that FLAG-Syt VII-green-fluorescence-protein fusion protein stably expressed in PC12 cells is localized in the perinuclear region (co-localization with TGN38 protein, even after brefeldin A treatment) and in the tips of neurites (co-localization with Syt I), and not in the plasma membrane.

The human copper transporter 1 gene (hCTR1) was previously identified by functional complementation in ctr1-deficient yeast. Overexpression of hCTR1 in wild-type yeast leads to increased sensitivity to copper toxicity, and mice with a homozygous disruption at the Ctr1 locus die early during embryogenesis. It is proposed that hCTR1 is responsible for high-affinity copper uptake into human cells, but the underlying molecular mechanisms are unknown. To begin to investigate the biochemical characteristics of hCTR1, a polyclonal antiserum was raised against recombinant hCTR1-fusion peptides. Biosynthetic studies using this antiserum revealed that hCTR1 was synthesized as a precursor protein of 28 kDa containing N-linked oligosaccharides, and is then converted to a mature protein of approx. 35 kDa, which is ubiquitously expressed. Immunofluorescence studies showed that subcellular hCTR1 localization differed markedly between cell types. In some cell lines, hCTR1 was located predominantly in an intracellular vesicular perinuclear compartment, and in others hCTR1 was located predominantly at the plasma membrane. In contrast with the copper export P-type ATPases mutated in Wilson disease and Menkes disease, the localization of hCTR1 was not influenced by copper concentrations. Inhibition of endocytosis by methyl-beta-cyclodextrin caused a partial redistribution of hCTR1 to the cell surface of HeLa cells. Taken together, the results in this study suggest a cell-specific control of copper uptake, which involves subcellular localization of the hCTR1 protein.

Members of the p24 family of putative cargo receptors (subdivided into p24-alpha, -beta, -gamma and -delta) are localized in the intermediate-and cis-Golgi compartments of the early secretory pathway, and are thought to play an important role in protein transport. In the present study, we wondered what effect increased biosynthetic cell activity with resulting high levels of protein transport would have on the subcellular localization of p24. We examined p24 localization in Xenopus intermediate pituitary melanotrope cells, which in black- and white-adapted animals are biosynthetically highly active and virtually inactive respectively. In addition, p24 localization was studied in Xenopus anterior pituitary cells whose activity is not changed during background adaptation. Using organelle fractionation, we found that in the inactive melanotropes and moderately active anterior pituitary cells of white-adapted animals, the p24-alpha, -beta, -gamma and -delta proteins are all located in the Golgi compartment. In the highly active melanotropes, but not in the anterior cells of black-adapted animals, the steady-state distribution of all four p24 members changed towards the intermediate compartment and subdomains of the endoplasmic reticulum (ER), most probably the ER exit sites. In the active melanotropes, the major cargo protein pro-opiomelanocortin was mostly localized to ER subdomains and partially co-localized with the p24 proteins. Furthermore, in the active cells, in vitro blocking of protein biosynthesis by cycloheximide or dispersion of the Golgi complex by brefeldin A led to a redistribution of the p24 proteins, indicating their involvement in ER-to-Golgi protein transport and extensive cycling in the early secretory pathway. We conclude that the subcellular localization of p24 proteins is dynamic and depends on the biosynthetic activity of the cell.

We have identified a mutant of the human G-protein Cdc42Hs, R66E, that fails to form a detectable complex with the GDP-dissociation inhibitor RhoGDI in cell-free systems or in intact cells. This point mutant is prenylated, binds guanine nucleotide and interacts with GTPase-activating protein in a manner indistinguishable from wild-type Cdc42Hs. Immunofluorescence localization studies revealed that this RhoGDI-binding-defective mutant is found predominantly in the Golgi apparatus, with a staining pattern similar to that of the wild-type protein. However, unlike wild-type Cdc42Hs, which is distributed in both the microsomal membrane and cytosolic fractions, studies using differential centrifugation show that prenylated R66E Cdc42Hs is found exclusively in association with lipid bilayers. Additionally, whereas the overexpression of RhoGDI results in an apparent translocation of wild-type Cdc42Hs from the Golgi apparatus into the cytosol, identical RhoGDI-overexpression conditions do not alter the Golgi localization of the R66E mutant. Furthermore, overexpression of this RhoGDI-binding-defective mutant of Cdc42Hs seems to activate redistribution of the actin cytoskeleton and filopodia formation in fibroblasts in a manner indistinguishable from the wild-type protein. Taken together, these results suggest that the interaction of Cdc42Hs with RhoGDI is not essential for proper membrane targeting of nascent prenylated Cdc42Hs in mammalian cells; neither is this interaction an essential part of the mechanism by which Cdc42Hs activates filopodia formation. However, it does seem that redistribution of Cdc42Hs to the cytosolic compartment is absolutely dependent on RhoGDI interaction.

Sorting nexin 1 (SNX1), a peripheral membrane protein, has previously been shown to regulate the cell-surface expression of the human epidermal growth factor receptor [Kurten, Cadena and Gill (1996) Science 272, 1008-1010]. Searches of human expressed sequence tag databases with SNX1 revealed eleven related human cDNA sequences, termed SNX2 to SNX12, eight of them novel. Analysis of SNX1-related sequences in the Saccharomyces cerevisiae genome clearly shows a greatly expanded SNX family in humans in comparison with yeast. On the basis of the predicted protein sequences, all members of this family of hydrophilic molecules contain a conserved 70-110-residue Phox homology (PX) domain, referred to as the SNX-PX domain. Within the SNX family, subgroups were identified on the basis of the sequence similarities of the SNX-PX domain and the overall domain structure of each protein. The members of one subgroup, which includes human SNX1, SNX2, SNX4, SNX5 and SNX6 and the yeast Vps5p and YJL036W, all contain coiled-coil regions within their large C-terminal domains and are found distributed in both membrane and cytosolic fractions, typical of hydrophilic peripheral membrane proteins. Localization of the human SNX1 subgroup members in HeLa cells transfected with the full-length cDNA species revealed a similar intracellular distribution that in all cases overlapped substantially with the early endosome marker, early endosome autoantigen 1. The intracellular localization of deletion mutants and fusions with green fluorescent protein showed that the C-terminal regions of SNX1 and SNX5 are responsible for their endosomal localization. On the basis of these results, the functions of these SNX molecules are likely to be unique to endosomes, mediated in part by interactions with SNX-specific C-terminal sequences and membrane-associated determinants.

Changes in the intracellular distribution of liver glycogen synthase (GS) might constitute a new regulatory mechanism for the activity of this enzyme at cellular level. Our previous studies indicated that incubation of isolated hepatocytes with glucose activated GS and resulted in its translocation from a homogeneous cytosolic distribution to the cell periphery. These studies also suggested a relationship with insoluble elements of the cytoskeleton, in particular actin. Here we show the translocation of GS in a different experimental model that allows the analysis of this phenomenon in long-term studies. We describe the reversibility of translocation of GS and its effect on glycogen distribution. Incubation of cultured rat hepatocytes with glucose activated GS and triggered its translocation to the hepatocyte periphery. The relative amount of the enzyme concentrated near the plasma membrane increased with time up to 8 h of incubation with glucose, when the glycogen stores reached their maximal value. The lithium-induced covalent activation of GS was not sufficient to cause its translocation to the cell periphery. The intracellular distribution of GS closely resembled that of glycogen. Our results showed an interaction between GS and an insoluble element of the hepatocyte matrix. Although no co-localization between actin filaments and GS was observed in any condition, disruption of actin cytoskeleton resulted in a significantly lower percentage of cells in which the enzyme translocated to the cell periphery in response to glucose. This observation suggests that the microfilament network has a role in the translocation of GS.

We have used a heterologous system of expression of N-methyl-D-aspartate (NMDA) receptors based on the use of vaccinia virus to analyse the maturation, transport, assembly and differential expression of the NR1 and NR2A subunits of the receptors. We have demonstrated that the NR1 subunit is efficiently transported to the plasma membrane in cells expressing NR1 alone, similarly to cells producing NR1 and NR2A together. In contrast, NR2A requires NR1 expression to be located at the cell surface. The stability of both receptor subunits expressed alone is similar to that obtained in cells producing NR1 and NR2A. In pulse-chase experiments, the NR1 subunit displays a biphasic decay, with a fraction of the protein having a half-life of only 1 h and the remaining presenting a turnover longer than 24 h, similar to values obtained for the NR2A subunit. Our results also show a maturation process affecting the carbohydrate moiety in the NR1 subunit, such that immature NR1 has a much shorter half-life than the mature form or the NR2A subunit. Finally, we show that only a fraction of mature NR1 interacts with NR2A to form multimeric functional complexes.

Three anti-microbial peptides were compared with respect to their killing activity against Candida albicans and their ability to disturb its cellular and internal membranes. Histatin 5 is an anti-fungal peptide occurring naturally in human saliva, while dhvar4 and dhvar5 are variants of its active domain, with increased anti-microbial activity. dhvar4 has increased amphipathicity compared with histatin 5, whereas dhvar5 has amphipathicity comparable with that of histatin 5. All three peptides caused depolarization of the cytoplasmic and/or mitochondrial membrane, indicating membranolytic activity. For the variant peptides both depolarization and killing occurred at a faster rate. With FITC-labelled peptides, no association with the cytoplasmic membrane was observed, contradicting the formation of permanent transmembrane multimeric peptide pores. Instead, the peptides were internalized and act on internal membranes, as demonstrated with mitochondrion- and vacuole-specific markers. In comparison with histatin 5, the variant peptides showed a more destructive effect on mitochondria. Entry of the peptides and subsequent killing were dependent on the metabolic state of the cells. Blocking of the mitochondrial activity led to complete protection against histatin 5 activity, whereas that of dhvar4 was hardly affected and that of dhvar5 was affected only intermediately.

Activation of phosphatidylinositol 3-kinase (PI 3-kinase) is a common event in both insulin and platelet-derived growth factor (PDGF) signalling, but only insulin activates this enzyme in the high-speed pellet (HSP), and induces GLUT4 translocation. Recently, we have demonstrated that exposure of 3T3-L1 adipocytes to oxidative stress impairs insulin-stimulated GLUT4 translocation and glucose transport, associated with impaired PI 3-kinase translocation and activation in the HSP [Tirosh, Potashnik, Bashan and Rudich (1999) J. Biol. Chem. 274, 10595-10602]. In this study the effect of a 2 h exposure to approximately 30 microM H(2)O(2) on insulin versus PDGF-BB signalling and metabolic effects was compared. PDGF-stimulated p85-associated PI 3-kinase activity in total cell lysates, as well as co-precipitation of the PDGF receptor, were unaffected by oxidative stress. Additionally, the increase in p85 association with the plasma-membrane lawns by PDGF remained intact following oxidation, whereas the insulin effect was decreased. PDGF significantly increased protein kinase B (PKB) activity in early differentiated cells, and that of p70 S6-kinase in both early and fully differentiated 3T3-L1 adipocytes. Following oxidation the effect of PDGF on PKB and p70 S6-kinase activation remained intact, whereas significant inhibition of insulin-stimulated activation of those enzymes was observed. In accordance, in both early and fully differentiated cells, oxidative stress completely blunted insulin- but not PDGF-stimulated protein synthesis. In conclusion, oxidative stress impairs insulin, but not PDGF, signalling and metabolic actions in both early and fully differentiated 3T3-L1 adipocytes. This emphasizes compartment-specific activation of PI 3-kinase as an oxidation-sensitive step specifically leading to insulin resistance.

The recently cloned amino acid transporter SAT2 is ubiquitously expressed and confers Na(+)-dependent transport of short-chain neutral amino acids, characteristics of the functionally defined System A transporter. Here we report the presence of SAT2 mRNA and protein in both skeletal muscle and adipocytes, and the characterization of polyclonal antibodies directed against this transporter. SAT2 protein was present in both plasma-membrane and internal-membrane fractions derived from rat skeletal muscle and adipose tissue, L6 myotubes and 3T3-L1 adipocytes, having a localization similar to that of the glucose transporter GLUT4. Moreover, consistent with the adaptive up-regulation of System A activity following chronic amino acid deprivation, a time-dependent increase in SAT2 protein abundance was observed in amino-acid-deprived L6 myotubes and 3T3-L1 adipocytes. These studies provide the first evidence regarding the subcellular distribution and adaptive up-regulation of SAT2 protein and the characterization of molecular probes for this physiologically important transporter, the function of which is altered in several disease states.

We have identified a novel protein of predicted molecular mass 40 kDa that contains two FYVE domains in tandem and has therefore been named TAFF1 (TAndem FYVE Fingers-1). The protein is expressed predominantly in heart and binds to PtdIns3P specifically, even though the FYVE domains in TAFF1 lacks the first Arg of the consensus sequence R(K/R)HHCR, critical for the PtdIns3P binding of other FYVE domains identified so far. The first Arg is replaced by a Thr and Ser in the N-terminal and C-terminal FYVE domains of TAFF1 respectively. Mutational analysis indicates that both FYVE domains are required for high affinity binding to PtdIns3P. Cell localization studies using a green fluorescent protein fusion show that TAFF1 is localized to the Golgi, and that the Golgi targeting sequence is located within the N-terminal 187 residues and not in either FYVE domain.

Membrane peptidases play important roles in cell activation, proliferation and communication. Human fibroblast-like synoviocytes express considerable amounts of aminopeptidase N/CD13, dipeptidyl peptidase IV/CD26, and neprilysin/CD10, transmembrane proteins previously proposed to be involved in the regulation of intra-articular levels of neuropeptides and chemotactic mediators as well as in adhesion and cell-cell interactions. Here, we report these peptidases in synoviocytes to be localized predominantly in glycolipid- and cholesterol-rich membrane microdomains known as 'rafts'. At the ultrastructural level, aminopeptidase N/CD13 and dipeptidyl peptidase IV/CD26 were found in caveolae, in particular in intracellular yet surface-connected vesicle-like structures and 'rosettes' made up of several caveolae. In addition, clusters of peptidases were seen at the cell surface in flat patches ranging in size from about 60 to 160 nm. Cholesterol depletion of synoviocytes by methyl-beta-cyclodextrin disrupted >90% of the caveolae and reduced the raft localization of aminopeptidase N/CD13 without affecting Ala-p-nitroanilide-cleaving activity of confluent cell cultures. In co-culture experiments with T-lymphocytes, cholesterol depletion of synoviocytes greatly reduced their capability to induce an early lymphocytic expression of aminopeptidase N/CD13. We propose caveolae/rafts to be peptidase-rich 'hot-spot' regions of the synoviocyte plasma membrane required for functional cell-cell interactions with lymphocytes. The peptidases may act in concert with other types of proteins such as receptors and signal transducers localized in these specialized membrane domains.

Toxic milk (tx) is a copper disorder of mice that causes a hepatic accumulation of copper similar to that seen in patients with Wilson disease. Both disorders are caused by a defect in the ATP7B copper-transporting ATPase. A feature of the tx phenotype is the production of copper-deficient milk by lactating dams homozygous for the tx mutation; the milk is lethal to the pups. It has not been determined whether the production of copper-deficient milk is a direct consequence of impaired expression of ATP7B protein in the mammary gland. With the use of immunohistochemistry, our study demonstrated that the ATP7B protein was mislocalized in the lactating tx mouse mammary gland, which would explain the inability of the tx mouse to secrete normal amounts of copper in milk. Confocal microscopy analysis showed that, in the lactating tx mammary gland, ATP7B was predominantly perinuclear in comparison with the diffuse, cytoplasmic localization of ATP7B in the lactating normal mammary gland. Lactating tx mice showed impaired delivery of copper from the mammary gland to the milk and this was not ameliorated by dietary copper supplementation. In contrast, the normal mouse mammary gland responded to increased dietary copper by increasing the amount of copper in milk. A change in the distribution of the ATP7B protein from perinuclear in the non-lactating gland to a diffuse, cytoplasmic localization in the lactating gland of the normal (DL) mouse suggests that the relocalization of APT7B is a physiological process that accompanies lactation. We conclude that the impaired copper transport from the mammary gland into milk in lactating tx mice is related to the mislocalization of ATP7B.

Both protein kinase C and protein tyrosine kinases have been shown to be involved in phospholipase D (PLD) activation in intact rat myometrium [Le Stunff, Dokhac and Harbon (2000) J. Pharmacol. Exp. Ther. 292, 629-637]. In this study we assessed the involvement of monomeric G-proteins in PLD activation in a cell-free system derived from myometrial tissue. Both the PLD1 and PLD2 isoforms were detected. Two forms of PLD activity, essentially membrane-bound, were found in myometrial preparations. One form was stimulated by oleate and insensitive to guanosine 5'-[gamma-thio] triphosphate (GTP[S]). The second required ammonium sulphate to be detected and was stimulated by GTP[S]. ADP-ribosylation factors (ARF1 and ARF6) and RhoA were immunodetected in myometrial preparations. ARF1 and RhoA were present in the membrane and cytosolic fractions whereas ARF6 was detected exclusively in the membrane fraction. A synthetic myristoylated peptide corresponding to the N-terminal domain of ARF6 [myrARF6((2-13))] totally abolished PLD activation in the presence of ammonium sulphate and GTP[S], whereas myrARF1((2-17)) and the inhibitory GDP/GTP-exchange factor, Rho GDI, did not. These data are consistent with a membrane-bound ARF6-regulated PLD activity. Finally, the stimulation of PLD by ARF6 was inhibited by AlF(-)(4) and this inhibition was counteracted by the fusion protein glutathione S-transferase-beta-adrenergic receptor kinase 1 (495-689) and by the QEHA peptide (from adenylate cyclase ACII), which act as G-protein betagamma-subunit scavengers. It is concluded that G-protein subunits betagamma are involved in a pathway modulating PLD activation by ARF6, illustrating cross-talk between heterotrimeric and monomeric G-proteins.

(1) Malonyl-CoA is thought to play a signalling role in fuel-selection in cardiac muscle, but the rate at which the concentration of this potential signal can be changed has not previously been investigated. (2) Rapid changes in cellular malonyl-CoA could be observed when rat cardiac myocytes were incubated in glucose-free medium followed by re-addition of 5 mM glucose, or when cells were transferred from a medium containing glucose to a glucose-free medium. On addition of glucose, malonyl-CoA increased by 62% to a new steady-state level, at a rate of at least 0.4 nmol/g dry wt. per min. The half-time of this change was less than 3 min. After removal of glucose the malonyl-CoA content was estimated to decline by 0.43-0.55 nmol/g dry wt. per min. (3) Malonyl-CoA decarboxylase (MDC) is a possible route for disposal of malonyl-CoA. No evidence was obtained for a cytosolic activity of MDC in rat heart where most of the activity was found in the mitochondrial fraction. MDC in the mitochondrial matrix was not accessible to extramitochondrial malonyl-CoA. However, approx. 16% of the MDC activity in mitochondria was overt, in a manner that could not be explained by mitochondrial leakage. It is suggested that this, as yet uncharacterized, overt MDC activity could provide a route for disposal of cytosolic malonyl-CoA in the heart. (4) No activity of the condensing enzyme for the fatty acid elongation system could be detected in any heart subcellular fraction using two assay systems. A previous suggestion [Awan and Saggerson (1993) Biochem. J. 295, 61-66] that this could provide a route for disposal of cytosolic malonyl-CoA in heart should therefore be abandoned.

The addition of PP(i) promoted the acidification of a subcellular compartment in cell homogenates of Toxoplasma gondii tachyzoites, implying the presence of a proton-translocating pyrophosphatase. The proton gradient was collapsed by addition of the K(+)/H(+) antiporter nigericin, and was also inhibited by addition of the PP(i) analogue aminomethylenediphosphonate (AMDP). Both proton transport and PP(i) hydrolysis were dependent upon K(+), but Na(+) caused partial inhibition of these activities. PP(i) hydrolysis was sensitive in a dose-dependent manner to AMDP, imidodiphosphate, NaF and to the thiol reagent N-ethylmaleimide. This activity was unaffected by common inhibitors of phosphohydrolases, except that NaO(3)V (sodium orthovanadate) stimulated the activity by 87%. Immunofluorescence microscopy, using antisera raised against conserved peptide sequences of a plant vacuolar pyrophosphatase, suggested that the pyrophosphatase in T. gondii tachyzoites was located in the plasma membrane and intracellular vacuoles of the parasite. High-field (31)P-NMR spectroscopy showed that PP(i )was more abundant than ATP in tachyzoites. Bisphosphonates (PP(i) analogues), drugs that are used in the treatment of bone diseases, inhibited proton transport and PP(i) hydrolysis in tachyzoite homogenates, and also inhibited intracellular proliferation of tachyzoites in tissue culture cells.

The cell-free activation of human neutrophil NADPH oxidase (O(2)(-)-generating enzyme) is enhanced by actin [Morimatsu, Kawagoshi, Yoshida and Tamura (1997) Biochem. Biophys. Res. Commun. 230, 206--210]. In an attempt to elucidate the mechanism, we examined the effect of actin-depolymerizing agents on the duration of NADPH oxidase in a cell-free system. The addition of DNase I, an F-actin-depolymerizing protein, caused an accelerated deactivation of the oxidase. The deactivation was also facilitated by latrunculin A, a sponge toxin that depolymerizes F-actin. Exogenously added actin prevented the deactivation by DNase I or latrunculin A, whereas EDTA accelerated a dilution-induced deactivation of the oxidase and Mg(2+) ions retarded it. The stability in dilution was found to correlate well with free Mg(2+) concentration. Estimation of F-actin in the system showed that F-actin increased during the oxidase activation and that DNase I or EDTA decreased F-actin content in parallel with the activity. Treatment of the cell-free mixture with a chemical cross-linker prevented the deactivation and F-actin decrease by EDTA. Taken together, these results suggest that actin filaments which grow during the activation of NADPH oxidase prolong the lifetime of the oxidase.

Immunohistochemical studies suggest that canalicular secretion via multidrug resistance protein 2 (Mrp2), a conjugate export pump encoded by the Mrp2 gene, is regulated by rapid transporter retrieval from/insertion into the canalicular membrane. The present study was undertaken in order to investigate this suggestion by means of immunogold electron microscopy. Therefore the effects of lipopolysaccharide (LPS) and osmolarity on Mrp2 localization were studied following immunogold labelling in the perfused rat liver by quantitative electron microscopy and morphometric analyses, and by confocal laser scanning microscopy. Mrp2 activity was assessed in the isolated perfused rat liver by measuring the excretion of dinitrophenyl-S-glutathione as a substrate of Mrp2. Both LPS and hyperosmolarity resulted in a statistically significant decrease in immunogold-labelled Mrp2 in the canalicular membrane and canalicular villi, and an increase in labelling in the pericanalicular cytoplasm. Canalicular morphometric parameters were unchanged under these conditions compared with controls. Under hyperosmolar perfusion Mrp2, but not the canalicular protein dipeptidylpeptidase IV, was found inside the cells, as shown by double immunofluorescence and confocal laser scanning microscopy. The findings suggest a selective retrieval of Mrp2 from the canalicular membrane under the influence of hyperosmolarity and LPS, whereas canalicular morphology remains unchanged.

We have identified from human liver eight alpha(1A)-adrenoceptor (alpha(1A)-AR) splice variants that were also expressed in human heart, prostate and hippocampus. Three of these alpha(1A)-AR isoforms (alpha(1A-1)-AR, alpha(1A-2a)-AR and alpha(1A-3a)-AR) gave rise to receptors with seven transmembrane domains (7TMalpha(1A)-AR). The other five (alpha(1A-2b)-AR, alpha(1A-2c)-AR, alpha(1A-3c)-AR, alpha(1A-5)-AR and alpha(1A-6)-AR) led to truncated receptors lacking transmembrane domain VII (6TMalpha(1A)-AR). The 7TMalpha(1A)-AR isoforms transiently expressed in COS-7 cells bound [(3)H]prazosin with high affinity (K(d) 0.2 nM) and mediated a noradrenaline (norepinephrine)-induced increase in cytoplasmic free Ca(2+) concentration, whereas the 6TMalpha(1A)-AR isoforms were incapable of ligand binding and signal transduction. Immunocytochemical studies with N-terminal epitope-tagged alpha(1A)-AR isoforms showed that the 7TMalpha(1A)-AR isoforms were present both at the cell surface and in intracellular compartments, whereas the 6TMalpha(1A)-AR isoforms were exclusively localized within the cell. Interestingly, in co-transfected cells, each truncated alpha(1A)-AR isoform inhibited [(3)H]prazosin binding and cell-surface trafficking of the co-expressed 'original' 7TMalpha(1A-1)-AR. However, there was no modification of either the [(3)H]prazosin-binding affinity or the pharmacological properties of alpha(1A-1)-AR. Immunoblotting experiments revealed that co-expression of the alpha(1A-1)-AR with 6TMalpha(1A)-AR isoforms did not impair alpha(1A-1)-AR expression. Therefore the expression in human tissues of many truncated isoforms constitutes a new regulation pathway of biological properties of alpha(1A)-AR.

Homer-1a is a 30 kDa protein that forms part of a family of conserved Homer-related proteins that interact with the C-termini of the metabotropic glutamate receptors mGluR1alpha and mGluR5a. Analysis of HEK-293 cells by PCR showed that they contained mRNA coding for members of the Homer family with the predominant form being Homer-1b, which is consistent with the immunochemical analysis of these cells. Homer-1a could not be detected by immunochemical analysis. To examine the function of Homer-1a, HEK-293 cells were transfected with cDNA encoding mGluR1alpha or Homer-1a or co-transfected with both cDNAs. When cells were co-transfected with the cDNAs for both proteins, immunofluorescent staining and biotinylation of cell surface molecules revealed a significant increase in the amount of receptor present at the cell surface in contrast to cells transfected with mGluR1alpha cDNA alone. This finding was consistent with a concomitant increase in the production of inositol phosphates after treatment of the doubly transfected cells with agonist. Intracellular immunostaining for both proteins revealed that they were co-localized and underwent a redistribution into a large vesicular compartment when they were co-expressed.

The rat thyrotropin-releasing hormone receptor-1 (TRHR-1) was modified by the addition of green fluorescent protein (GFP) and expressed stably in HEK293 cells. Extensive overlap of plasma membrane distribution of autofluorescent TRHR-1-GFP with that of the phosphoinositidase C-linked G-proteins Gqalpha/G11alpha, identified by indirect immunofluorescence, was monitored concurrently. Addition of thyrotropin-releasing hormone resulted in rapid separation of TRHR-1-GFP and Gqalpha/G11alpha signals as the receptor was internalized. This situation persisted for more than an hour. At longer time periods a fraction of the cellular Gqalpha/G11alpha was also internalized, although much of the Gqalpha/G11alpha immunoreactivity remained associated with the plasma membrane. Parallel experiments, in which the cellular distribution of TRHR-1-GFP and Gqalpha/G11alpha immunoreactivity were monitored in sucrose-gradient fractions following cell disruption, also demonstrated a rapid, agonist-induced movement of TRHR-1-GFP away from the plasma membrane to low-density vesicular fractions. At later time points, a fraction of the cellular Gqalpha/G11alpha immunoreactivity was also redistributed to overlapping, but non-identical, low-density-vesicle-containing fractions. Pretreatment of the cells with cytochalasin D or nocodazole prevented agonist-induced redistribution of G-protein but not TRHR-1-GFP, further indicating resolution of the mechanics of these two processes. The combination of a GFP-modified receptor and immunostaining of the G-proteins activated by that receptor allows, for the first time, concurrent analysis of the varying dynamics and bases of internalization and redistribution of two elements of the same signal-transduction cascade.

The purpose of this work was to investigate the role of N-glycosylation in the expression and pharmacological properties of the the rat AT1a angiotensin II (AII) receptor. Glycosylation-site suppression was carried out by site-directed mutagenesis (Asn-->Gln) of Asn176 and Asn188 (located on the second extracellular loop) and by the removal of Asn4 at the N-terminal end combined with the replacement of the first four amino acids by a 10 amino acid peptide epitope (c-Myc). We generated seven possible N-glycosylation-site-defective mutants, all tagged at their C-terminal ends with the c-Myc epitope. This double-tagging strategy, associated with photoaffinity labelling, allowed evaluation of the molecular masses and immunocytochemical cellular localization of the various receptors transiently expressed in COS-7 cells. We showed that: (i) each of the three N-glycosylation sites are utilized in COS-7 cells; (ii) the mutant with three defective N-glycosylation sites was not (or was very inefficiently) expressed at the plasma membrane and accumulated inside the cell at the perinuclear zone; (iii) the preservation of two sites allowed normal receptor delivery to the plasma membrane, the presence of only Asn176 ensuring a behaviour similar to that of the wild-type receptor; and (iv) all expressed receptors displayed unchanged pharmacological properties (Kd for 125I-sarcosine1-AII; sarcosine1-AII-induced inositol phosphate production). These results demonstrate that N-glycosylation is required for the AT1 receptor expression. They are discussed in the light of current knowledge of membrane-protein maturation and future prospects of receptor overexpression for structural studies.

SNARE proteins are required for vesicle docking and fusion in eukaryotic cells in processes as diverse as homotypic membrane fusion and synaptic vesicle exocytosis [SNARE stands for SNAP receptor, where SNAP is soluble NSF attachment protein]. The SNARE proteins syntaxin 4 and vesicle-associated membrane protein (VAMP) 2/3 also participate in the insulin-stimulated translocation of GLUT4 from intracellular vesicles to the plasma membrane in adipose cells. We now report the molecular cloning and characterization of rat SNAP-23, a ubiquitously expressed homologue of the essential neuronal SNARE protein SNAP-25 (synaptosomal-associated protein of 25 kDa). Rat SNAP-23 is 86% and 98% identical respectively to human and mouse SNAP-23. Southern blot analysis reveals that the rat, mouse and human SNAP-23 genes encode species-specific isoforms of the same protein. Co-immunoprecipitation of syntaxin 4 and SNAP-23 shows association of these two proteins in rat adipose cell plasma membranes, and insulin stimulation does not alter the SNAP-23/syntaxin 4 complex. In addition, we demonstrate for the first time the participation of SNAP-23, along with syntaxin 4 and VAMP2/3, in the formation of 20S SNARE complexes prepared using rat adipose cell membranes and recombinant alpha-SNAP and NSF proteins. The stoichiometry of the SNARE complexes formed is essentially identical using membranes from either unstimulated or insulin-stimulated adipose cells. These data demonstrate that rat SNAP-23 associates with syntaxin 4 before insulin stimulation and is present in the SNARE complexes known to mediate the translocation of GLUT4 from intracellular vesicles to the plasma membrane of rat adipose cells.

Munc13-1, a mammalian homologue of Caenorhabditis elegans unc-13p, is a presynaptic phorbol ester receptor that enhances neurotransmitter release. In the present study we analysed the regional, cellular and subcellular expression patterns in rat of two novel Munc13 proteins, Munc13-2 and Munc13-3. We demonstrate by hybridization in situ that Munc13-1 mRNA is expressed throughout the brain, whereas Munc13-2 mRNA is preferentially present in rostral brain regions, and Munc13-3 mRNA in caudal areas. In an analysis of subcellular brain fractions with isoform-specific antibodies, we show that the novel Munc13 proteins are enriched in synapses. Immunocytochemical examination of rat cerebellar sections indicates that Munc13-3, like Munc13-1, is concentrated in presynaptic terminals. Our results characterize Munc13 proteins as a family of neuron-specific, synaptic molecules that bind to syntaxin, an essential mediator of neurotransmitter release. Munc13-2 and Munc13-3 are expressed in a complementary fashion and might act in concert with Munc13-1 to modulate neurotransmitter release.