Search

PEX genes encode peroxins, which are proteins required for peroxisome assembly. The PEX19 gene of the yeast Yarrowia lipolytica was isolated by functional complementation of the oleic acid-nonutilizing strain pex19-1 and encodes Pex19p, a protein of 324 amino acids (34,822 Da). Subcellular fractionation and immunofluorescence microscopy showed Pex19p to be localized primarily to peroxisomes. Pex19p is detected in cells grown in glucose-containing medium, and its levels are not increased by incubation of cells in oleic acid-containing medium, the metabolism of which requires intact peroxisomes. pex19 cells preferentially mislocalize peroxisomal matrix proteins and the peripheral intraperoxisomal membrane peroxin Pex16p to the cytosol, although small amounts of these proteins could be reproducibly localized to a subcellular fraction enriched for peroxisomes. In contrast, the peroxisomal integral membrane protein Pex2p exhibits greatly reduced levels in pex19 cells compared with its levels in wild-type cells. Importantly, pex19 cells were shown by electron microscopy to contain structures that resemble wild-type peroxisomes in regards to size, shape, number, and electron density. Subcellular fractionation and isopycnic density gradient centrifugation confirmed the presence of vesicular structures in pex19 mutant strains that were similar in density to wild-type peroxisomes and that contained profiles of peroxisomal matrix and membrane proteins that are similar to, yet distinct from, those of wild-type peroxisomes. Because peroxisomal structures form in pex19 cells, Pex19p apparently does not function as a peroxisomal membrane protein receptor in Y. lipolytica. Our results are consistent with a role for Y. lipolytica Pex19p in stabilizing the peroxisomal membrane.

The genes encoding the first two enzymes of the peroxisomal beta-oxidation pathway, acyl-CoA oxidase (AOx) and enoyl-CoA hydratase/3-hydroxyacyl-CoA dehydrogenase (HD), contain upstream cis-acting regulatory regions termed peroxisome proliferator response elements (PPRE). Transcription of these genes is mediated through the binding of peroxisome proliferator-activated receptor alpha (PPARalpha), which binds to a PPRE as a heterodimer with the 9-cis-retinoic acid receptor (RXRalpha). Here we demonstrate that the HD-PPRE is also a target for the constitutive androstane receptor beta (CARbeta). In vitro binding analysis showed that CARbeta bound the HD-PPRE, but not the AOx-PPRE, as a heterodimer with RXRalpha. Binding of CARbeta/RXRalpha to the HD-PPRE occurred via determinants that overlap partially with those required for PPARalpha/RXRalpha binding. In vivo, CARbeta/RXRalpha activated transcription from an HD-PPRE luciferase reporter construct. Interestingly, CARbeta was shown to also modulate PPARalpha/RXRalpha-mediated transactivation in a response element-specific manner. In the presence of the peroxisome proliferator, Wy-14,643, CARbeta had no effect on PPARalpha/RXRalpha-mediated transactivation from the HD-PPRE but antagonized transactivation from the AOx-PPRE in both the presence and the absence of proliferator. Our results illustrate that transcription of the AOx and HD genes is differentially regulated by CARbeta and that the HD gene is a specific target for regulation by CARbeta. Overall, this study proposes a novel role for CARbeta in the regulation of peroxisomal beta-oxidation.

pex mutants are defective in peroxisome assembly. The mutant strain pex23-1 of the yeast Yarrowia lipolytica lacks morphologically recognizable peroxisomes and mislocalizes all peroxisomal matrix proteins investigated preferentially to the cytosol. pex23 strains accumulate vesicular structures containing both peroxisomal matrix and membrane proteins. The PEX23 gene was isolated by functional complementation of the pex23-1 strain and encodes a protein, Pex23p, of 418 amino acids (47,588 Da). Pex23p exhibits high sequence similarity to two hypothetical proteins of the yeast Saccharomyces cerevisiae. Pex23p is an integral membrane protein of peroxisomes that is completely, or nearly completely, sequestered from the cytosol. Pex23p is detected at low levels in cells grown in medium containing glucose, and its levels are significantly increased by growth in medium containing oleic acid, the metabolism of which requires intact peroxisomes.

Peroxisome proliferator-activated receptor alpha (PPARalpha) heterodimerizes with the 9-cis-retinoic acid receptor (RXRalpha) to bind to peroxisome proliferator-response elements (PPRE) present in the upstream regions of a number of genes involved in metabolic homeostasis. Among these genes are those encoding fatty acyl-CoA oxidase (AOx) and enoyl-CoA hydratase/3-hydroxyacyl-CoA dehydrogenase (HD), the first two enzymes of the peroxisomal beta-oxidation pathway. Here we demonstrate that the orphan nuclear hormone receptor, RevErbalpha, modulates PPARalpha/RXRalpha- dependent transactivation in a response element-specific manner. In vitro binding analysis showed that RevErbalpha bound the HD-PPRE but not the AOx-PPRE. Determinants within the HD-PPRE required for RevErbalpha binding were distinct from those required for PPARalpha/RXRalpha binding. In transient transfections, RevErbalpha antagonized transactivation by PPARalpha/RXRalpha from an HD-PPRE luciferase reporter construct, whereas no effects were observed with an AOx-PPRE reporter construct. These data identify the HD gene as a target for RevErbalpha and illustrate cross-talk between the RevErbalpha and PPARalpha signaling pathways on the HD-PPRE. Our results suggest a novel role for RevErbalpha in regulating peroxisomal beta-oxidation.

p62 is a recently identified ubiquitin-binding, cytosolic phosphoprotein that interacts with several signal transduction molecules including the tyrosine kinase p56(lck) and the protein kinase C-zeta. p62 is therefore suggested to serve an important role in signal transduction in the cell, although the physiological function of p62 remains undefined. Here we demonstrate by transient transfection assays that p62 stimulates the transcription of reporter genes linked to the simian virus 40 (SV40) enhancer. A putative p62-responsive element was localized to the B domain of the distal 72-base pair repeat of the SV40 enhancer. p62 was unable to bind this element in vitro, nor was it able to activate transcription when directly tethered to a promoter, suggesting that p62 stimulates transcription via an indirect mechanism. Stimulation of transcription mediated by p62 was dependent on its amino-terminal region, which is also necessary for interaction with cell surface signaling molecules. These findings indicate that p62 may link extracellular signals directly to transcriptional responses, and identify the SV40 enhancer as a downstream target for signal transduction pathways in which p62 participates.

Immunoblot analysis showed that the 47 kDa platelet substrate of protein kinase C (P47) was expressed at low levels in undifferentiated HL-60 leukaemia cells. Treatment of these cells with dimethyl sulphoxide, 1 alpha,25-dihydroxycholecalciferol or retinoic acid caused progressive increases in P47 content. Retinoic acid (1 microM) elicited the largest response, a 4-fold increase in P47 protein after 7 days that was accompanied by an increase in translatable P47 mRNA. The induction of P47 by retinoic acid preceded cessation of cell proliferation and development of the capacity to reduce Nitro Blue Tetrazolium, indicating that its expression is an early event in the myeloid differentiation of HL-60 cells.

A tissue-screening survey employing quantitative radioautography was carried out at 2 min after the intravascular injection of 125I-insulin into laboratory rats. The results revealed a substantial binding of insulin to cells forming the proximal convoluted tubule in kidney, hepatocytes of liver, acinar cells of the pancreas, parenchymal cells of the adrenal cortex and medulla, and epithelial cells of the gastrointestinal tract. Control experiments indicated that this binding was due to a specific interaction with the insulin receptor, except in the case of kidney where the binding was shown to be nonspecific. Although the major target for insulin action (liver) clearly demonstrated specific insulin binding, several other classical targets (adipocytes, skeletal, cardiac, and smooth muscle cells) showed no specific 125I-insulin binding and therefore indicated the limits of sensitivity of the in vivo radioautographic method. Nevertheless, the working hypothesis of a direct correlation of insulin receptor density with insulin action points to the hitherto unemphasized targets of pancreas, adrenal gland, and gastrointestinal tract as major sites of insulin action in the body.

Total RNA, extracted with guanidinium thiocyanate from liver of clofibrate-treated rats, was translated in vitro. Heating the RNA at 60 degrees C for 5 min before translation increased the synthesis of three peroxisomal polypeptides 10-100-fold. Preproalbumin synthesis increased 10-fold. Total incorporation of [35S]methionine into proteins merely doubled. Heating is essential for reproducible and adequate translation of mRNAs coding for peroxisomal and some other proteins.

The yeast two-hybrid system was used to isolate novel cellular factors that interact with the mouse peroxisome proliferator-activated receptor alpha (PPARalpha). One of the interacting clones isolated encoded LXRalpha, a recently described human orphan nuclear hormone receptor. LXRalpha bound directly to PPARalpha, as well as to the common heterodimerization partner 9-cis-retinoic acid receptor (RXRalpha). LXRalpha did not form a DNA binding complex with PPARalpha on synthetic hormone response elements composed of direct repeats of the TGACCT consensus half-site or on naturally occurring peroxisome proliferator response elements (PPREs) or LXRalpha response elements. However, LXRalpha inhibited binding of PPARalpha/RXRalpha heterodimers to PPREs, and coexpression of LXRalpha in mammalian cells antagonized peroxisome proliferator signaling mediated by PPARalpha/RXRalpha in vivo. These findings identify a novel partner for PPARalpha and suggest that LXRalpha plays a role in modulating PPAR-signaling pathways in the cell.

The manner of synthesis and assembly of the peroxisomal membrane proteins is unknown. Understanding these processes is essential to an understanding of the formation of the organelle. We have investigated the biogenesis of the previously identified major 21.7-kDa integral peroxisomal membrane polypeptide [Fujiki, Y., Fowler, S., Shio, H., Hubbard, A. L. & Lazarow, P. B. (1982) J. Cell Biol. 93, 103-110]. This protein was purified to apparent homogeneity and used to elicit a rabbit antiserum. In immunoblotting analysis, antibody bound only to the 22-kDa membrane polypeptide present exclusively in peroxisomal membranes. Total rat liver RNA was translated in a nuclease-treated rabbit reticulocyte cell-free protein-synthesizing system. The in vitro translation product, isolated by means of the antibody and Staphylococcus aureus cells, comigrated with the mature 22-kDa polypeptide in NaDodSO4/PAGE. Analysis of the translation products of RNAs from free and membrane-bound polysomes indicated that the mRNA for the 22-kDa membrane polypeptide is found predominantly in free polysomes. The results imply post-translational insertion of the membrane polypeptide into the peroxisomal membrane without proteolytic processing and suggest that peroxisomes, like mitochondria and chloroplasts, form by fission from preexisting organelles.

The intracellular migration of G protein in vesicular stomatitis virus-infected cells was visualized by light and electron microscope radioautography after a 2-min pulse with [3H]mannose followed by nonradioactive chase for various intervals. The radioactivity initially (at 5-10 min) appeared predominantly in the endoplasmic reticulum, and the [3H]mannose-labeled G protein produced was sensitive to endoglycosidase H. Silver grains were subsequently (at 30-40 min) observed over the Golgi apparatus, and the [3H]mannose-labeled G protein became resistant to endoglycosidase H digestion. Our data directly demonstrate the intracellular transport of a plasmalemma-destined transmembrane glycoprotein through the Golgi apparatus.

We investigated the site of synthesis of two abundant proteins in clofibrate-induced rat hepatic peroxisomes. RNA was extracted from free and membrane-bound polysomes, heated to improve translational efficiency, and translated in the mRNA-dependent, reticulocyte-lysate-cell-free, protein-synthesizing system. The peroxisomal acyl-CoA oxidase and enoyl-CoA hydratase-beta-hydroxyacyl-CoA dehydrogenase 35S-translation products were isolated immunochemically, analyzed by SDS PAGE and fluorography, and quantitated by densitometric scanning. The RNAs coding for these two peroxisomal proteins were found predominantly on free polysomes, and the translation products co-migrated with the mature proteins. As in normal rat liver, preproalbumin and catalase were synthesized mainly by membrane-bound and by free polysomes, respectively. mRNAs for a number of minor 35S-translation products also retained by the anti-peroxisomal immunoadsorbent were similarly found on free polysomes. These results, together with previous data, allow the generalization that the content proteins of rat liver peroxisomes are synthesized on free polysomes, and the data imply a posttranslational packaging mechanism for these major content proteins.

Candida tropicalis pK233 cells were grown with n-alkanes as carbon source to induce the synthesis of peroxisomal proteins and the proliferation of peroxisomes. Poly-(A)+ RNA was isolated and used to construct a cDNA library by insertion of double-stranded reverse transcripts into the Pst I site of pBR322 followed by cloning in Escherichia coli. Clones complementary to mRNAs induced by growth on alkanes were selected by differential DNA dot-blot analysis using [32P]cDNA reverse-transcribed from poly(A)+ RNA of glucose-grown cells (which contain few peroxisomes) or of alkane-grown cells. Among these clones, one containing a 1.7-kilobase insert coding for acyl-CoA oxidase (the first enzyme in the peroxisomal Beta-oxidation pathway) was identified by hybridization-selection translation and immunoprecipitation. By RNA blot analysis, the acyl-CoA oxidase mRNA was estimated to be approximately equal to 2.2 kilobases long, of which 2.1 kilobases are required to code for the approximately equal to 76-kDa protein. Since the mRNA is polyadenylylated, there appears to be little additional nontranslated region. Cell-free mRNA translation and RNA dot-blot hybridization analyses demonstrated that, whereas glucose-grown C. tropicalis contained little or no acyl-CoA oxidase mRNA, alkane-grown cells contained so much of this mRNA as to make acyl-CoA oxidase one of the major in vitro translation products.

Peroxisomes have been purified from Candida tropicalis grown on oleic acid and shown to be nearly pure by marker enzyme analysis, electron microscopy, and comparative sodium dodecyl sulfate-polyacrylamide gel electrophoresis. They contain approximately 20 polypeptides, among which acyl-CoA oxidase, trifunctional hydratase-dehydrogenase-epimerase, and catalase have been identified. Rabbit antisera have been elicited that react with these three proteins. When C. tropicalis is grown on alkanes, a dozen mRNAs are strikingly induced. Nine of the 12 induced mRNAs code for polypeptides that comigrate in sodium dodecyl sulfate-polyacrylamide gel electrophoresis with peroxisomal proteins, among which three have been identified immunochemically as the acyl-CoA oxidase, the trifunctional protein, and catalase. These results indicate that some genes coding for peroxisomal proteins are strongly expressed during growth of C. tropicalis on alkanes. The data are consistent with evidence in other species that peroxisomes form by the post-translational incorporation of newly made proteins into pre-existing peroxisomes, generally without proteolytic processing, followed by peroxisome division.

A cDNA clone complementary to mRNA encoding the precursor (Mr = 165,000) to the rat liver mitochondrial matrix enzyme carbamyl phosphate synthetase I (Mr = 160,000) was employed to compare relative amounts of the messenger in adult and fetal liver and in Morris hepatoma 5123D and 3924A cells. Northern blot analysis gave a size estimate for the messenger of 6,500-6,700 nucleotides. Carbamyl phosphate synthetase mRNA levels in 15-day-old fetal liver were less than 10% of adult levels; 5123D cells expressed the messenger at levels about 2-fold higher than normal adult liver, but the messenger was undetectable in 3924A cells. Albumin mRNA was also expressed in the former but not in the latter. Maintaining rats for 5 days on a diet containing 60% casein augmented the relative amount of carbamyl phosphate synthetase mRNA by about 2-fold, while a protein-free diet resulted in reduced levels of the mRNA (about 50% compared to animals on a normal diet). Finally, the pattern of hybridization of carbamyl phosphate synthetase cDNA to HindIII-digested genomic DNA showed no differences between normal liver and its corresponding hepatoma; however, a HindIII site polymorphism was observed between Buffalo and ACI rats.

The site of synthesis and induction by clofibrate of peroxisomal 3-ketoacyl-CoA thiolase (acetyl-CoA acyltransferase; EC 2.3.1.16) was investigated. Free and membrane-bound polyribosomal RNA species from the livers of normal rats and rats treated with clofibrate, a hypolipidaemic drug that causes marked proliferation of peroxisomes, were translated in a nuclease-treated rabbit reticulocyte-lysate cell-free protein-synthesizing system with [35S]methionine as label. The cell-free translation products were immunoprecipitated with monospecific X rabbit anti-thiolase serum and analysed by sodium dodecyl sulphate/polyacrylamide-gel electrophoresis and fluorography. Thiolase mRNA was found predominantly in free polyribosomes, in both normal and clofibrate-treated rats. Clofibrate treatment increased mRNA activity for thiolase approx. 20-fold. The translation product of clofibrate-induced thiolase mRNA migrated slightly faster in sodium dodecyl sulphate/polyacrylamide-gel electrophoresis than did the translation product of normal thiolase mRNA. Both the normal and the clofibrate-induced translation products were approx. 6000 Da larger than the 41000-Da subunit of the purified enzyme. Immunoblot analysis of liver homogenates, isolated peroxisomes and the purified enzyme indicated that the thiolase subunit was approx. 41000 Da in all samples, ruling out proteolysis during the purification of thiolase. Thiolase biogenesis thus differs from that of rat liver peroxisomal proteins studied previously in that it is synthesized as a larger precursor, implying post-translational import of thiolase into peroxisomes with proteolytic processing. Clofibrate apparently alters the size as well as the amount of the translation product.

Chicken ovalbumin upstream promoter transcription factor II (COUP-TFII), an orphan member of the nuclear hormone receptor superfamily, acts as a transcriptional repressor by antagonizing the functions of other nuclear hormone receptors and by actively silencing transcription. However, in certain contexts, COUP-TFII stimulates transcription directly. A cellular factor, isolated by interaction cloning, bound COUP-TFII in vitro and allowed COUP-TFII to function as a transcriptional activator in mammalian cells. This factor is identical to a recently described ligand of the tyrosine kinase signaling molecule p56(lck), suggesting that it mediates cross-talk between mitogenic and nuclear hormone receptor signal transduction pathways.

Multimerin is a massive, soluble protein found in platelets and in the endothelium of blood vessels. Multimerin is composed of varying sized, disulfide-linked multimers, the smallest of which is a homotrimer. Multimerin is a factor V/Va-binding protein and may function as a carrier protein for platelet factor V. The cDNA for human multimerin was isolated from lambda gt11 endothelial cell libraries using antibodies, and the isolated cDNA clones were used to obtain the full sequence. The full-length multimerin cDNA was 4.2 kilobase pairs. Northern analyses identified a 4.7-kilobase transcript in cultured endothelial cells, a megakaryocytic cell line, platelets, and highly vascular tissues. The multimerin cDNA can encode a protein of 1228 amino acids with the probable signal peptide cleavage site between amino acids 19 and 20. The protein is predicted to be hydrophilic and to contain 23 N-glycosylation sites. The adhesive motif RGDS (Arg-Gly-Asp-Ser) and an epidermal growth factor-like domain were identified. Sequence searches indicated that multimerin is a unique protein. Analyses identified probable coiled-coil structures in the central portion of the multimerin sequence. Additionally, the carboxyl-terminal region of multimerin resembles the globular, non-collagen-like, carboxyl-terminal domains of several other trimeric proteins, including complement C1q and collagens type VIII and X.