Recombinant iron-regulatory factor functions as an iron-responsive-element-binding protein, a translational repressor and an aconitase. A functional assay for translational repression and direct demonstration of the iron switch.

The translation of ferritin and erythroid 5-aminolevulinate synthase mRNAs is regulated via a specific high-affinity interaction between an iron-responsive element in the 5' untranslated region of ferritin and erythroid 5-aminolevulinate synthase mRNAs and a 98-kDa cytoplasmic protein, the iron-regulatory factor. Iron-regulatory factor was expressed in vaccinia-virus-infected HeLa cells (hIRFvac) and in Escherichia coli (hIRFeco). An N-terminal histidine tag allowed a rapid one-step purification of large quantities of soluble recombinant protein. Both hIRFvac and hIRFeco bound specifically to iron-responsive elements and were immunoprecipitated by iron-regulatory-factor antibodies. Using in-vitro-transcribed chloramphenicol-acetyltransferase mRNAs bearing an iron-responsive element in the 5' untranslated region, specific repression of chloramphenicol-acetyltransferase translation by hIRFvac and hIRFeco was demonstrated in wheat-germ extract. In addition, hIRFvac and hIRFeco were shown to display aconitase activity. Treatment of hIRFvac and hIRFeco with FeSO4 resulted in a drastic reduction in iron-responsive-element-binding of iron-regulatory factor, but caused a strong stimulation of its aconitase activity. The results establish that recombinant iron-regulatory factor is a bifunctional protein; after purification, it binds to iron-responsive elements and represses translation in vitro. Following iron treatment, iron-responsive-element binding is lost and aconitase activity is gained. No eukaryotic co-factor seems to be required for the conversion of the iron-responsive-element binding to the aconitase form of the protein.

Translational repression by the human iron-regulatory factor (IRF) in Saccharomyces cerevisiae.

The regulation of the synthesis of ferritin and erythroid 5-aminolevulinate synthase in mammalian cells is mediated by the interaction of the iron regulatory factor (IRF) with a specific recognition site, the iron responsive element (IRE), in the 5' untranslated regions (UTRs) of the respective mRNAs. A new modular expression system was designed to allow reconstruction of this regulatory system in Saccharomyces cerevisiae. This comprised two components: a constitutively expressed reporter gene (luc; encoding luciferase) preceded by a 5' UTR including an IRE sequence, and an inducibly expressed cDNA encoding human IRF. Induction of the latter led to the in vivo synthesis of IRF, which in turn showed IRE-binding activity and also repressed translation of the luc mRNA bearing an IRE-containing 5' UTR. The upper stem-loop region of an IRE, with no further IRE-specific flanking sequences, sufficed for recognition and repression by IRF. Translational regulation of IRE-bearing mRNAs could also be demonstrated in cell-free yeast extracts. This work defines a minimal system for IRF/IRE translational regulation in yeast that requires no additional mammalian-specific components, thus providing direct proof that IRF functions as a translational repressor in vivo. It should be a useful tool as the basis for more detailed studies of eukaryotic translational regulation.

Translational regulation via iron-responsive elements by the nitric oxide/NO-synthase pathway.

Nitric oxide (NO) produced from L-arginine by NO synthases (NOS) is a transmitter known to be involved in diverse biological processes, including immunomodulation, neurotransmission and blood vessel dilatation. We describe a novel role of NO as a signaling molecule in post-transcriptional gene regulation. We demonstrate that induction of NOS in macrophage and non-macrophage cell lines activates RNA binding by iron regulatory factor (IRFs), the central trans regulator of mRNAs involved in cellular iron metabolism. NO-induced binding of IRF to iron-responsive elements (IRE) specifically represses the translation of transfected IRE-containing indicator mRNAs as well as the biosynthesis of the cellular iron storage protein ferritin. These findings define a new biological function of NO and identify a regulatory connection between the NO/NOS pathway and cellular iron metabolism.

Translational control of 5-aminolevulinate synthase mRNA by iron-responsive elements in erythroid cells.

Hemoglobin synthesis in red cells is the major iron utilization pathway in the human body and accounts for > 80% of systemic iron turnover. The first step in erythroid heme biosynthesis is catalyzed by a tissue-specific isoform of 5-aminolevulinate synthase (ALAS). The previous identification of iron-responsive elements in the 5'-untranslated region of human and murine erythroid ALAS mRNA raised the intriguing possibility that eALAS expression might be under iron-dependent translational control. As a consequence, a single post-transcriptional regulatory system could coordinate cellular iron acquisition via the transferrin receptor, storage via ferritin, and utilization via eALAS. We directly demonstrate iron-dependent translational regulation of eALAS mRNA in murine erythroleukemia (MEL) cells. The iron-responsive element motif contained in eALAS mRNA is shown to be sufficient to confer translational control to a reporter mRNA both in transfected MEL cells and in vitro.

Position is the critical determinant for function of iron-responsive elements as translational regulators.

At least two groups of eukaryotic mRNAs (ferritin and erythroid 5-aminolevulinate synthase) are translationally regulated via iron-responsive elements (IREs) located in a conserved position within the 5' untranslated regions of their mRNAs. We establish that the spacing between the 5' terminus of an mRNA and the IRE determines the potential of the IRE to mediate iron-dependent translational repression. The length of the RNA spacer rather than its nucleotide sequence or predicted secondary structure is shown to be the primary determinant of IRE function. When the position of the IRE is preserved, sequences flanking the IRE in natural ferritin mRNA can be replaced by altered flanking sequences without affecting the regulatory function of the IRE in vivo. These results define position as a critical cis requirement for IRE function in vivo and imply the potential to utilize transcription start site selection to modulate the function of this translational regulator.

Identification of a novel iron-responsive element in murine and human erythroid delta-aminolevulinic acid synthase mRNA.

Iron-responsive elements (IREs) are regulatory RNA elements which are characterized by a phylogenetically defined sequence-structure motif. Their biological function is to provide a specific binding site for the IRE-binding protein (IRE-BP). Iron starvation of cells induces high affinity binding of the cytoplasmic IRE-BP to an IRE which has at least two different known biological consequences, repression of ferritin mRNA translation and stabilization of the transferrin receptor transcript. We report the identification of a novel, evolutionarily conserved IRE motif in the 5' UTR of murine and human erythroid-specific delta-aminolevulinic acid synthase (eALAS) mRNA which encodes the first, and possibly rate limiting, enzyme of the heme biosynthetic pathway. We demonstrate the function of the eALAS IRE as a specific binding site for the IRE-BP by gel retardation analyses and by in vitro translation experiments. In addition, we show that the 5' UTR of eALAS mRNA is sufficient to mediate iron-dependent translational regulation in vivo. These findings strongly suggest involvement of the IRE-IRE-BP system in the control of heme biosynthesis during erythroid differentiation.

Effects of beta-xylosides on proteoglycan biosynthesis and morphology of PC12 pheochromocytoma cells and primary cultures of rat cerebellum.

We have examined the effects of beta-xylosides, which act as exogenous acceptors for glycosaminoglycan chain initiation, on the morphology and proteoglycan biosynthesis of PC12 pheochromocytoma cells, and on monolayer, aggregate and explant cultures of early postnatal rat cerebellum. PC12 cells cultured for 13 days in the presence of nerve growth factor (NGF) and beta-xyloside, and labeled during days 11-13 with sodium [35S]sulfate, showed an 8- to 11-fold increase in [35S]sulfate-labeled macromolecules released into the culture medium. Most of the increase was accounted for by chondroitin sulfate, which was in the form of free glycosaminoglycan chains, which were not acid-precipitable. The presence of beta-xyloside also led to a 65-115% increase in [35S]sulfate incorporation into cell-associated glycosaminoglycans and glycoproteins of untreated and NGF-treated PC12 cells, respectively. beta-Xyloside treatment reduced the size of the chondroitin sulfate chains in both the cells and medium from approximately 34,000 to 10,000 Mr, but had much less effect on heparan sulfate, which decreased in size from 16,000 to 13,000-14,500 Mr (in the medium and cells, respectively). beta-Xyloside inhibition of proteoglycan biosynthesis was accompanied by significant morphological effects in NGF-treated PC12 cells, consisting of an increase in length and decrease in the branching, diameter and adhesion to the collagen substratum of the PC12 cell processes. p-Nitrophenyl- and 4-methylumbelliferyl-beta-D-xylosides produced similar effects, which were not seen with p-nitrophenyl-beta-D-galactoside. beta-Xylosides also produced distinct alterations in the adhesion and morphology of monolayer, aggregate, and explant cultures of early postnatal rat cerebellum, which occurred together with inhibition of chondroitin sulfate proteoglycan biosynthesis and a decrease in glycosaminoglycan chain size. These studies indicate that chondroitin sulfate (and probably also heparan sulfate) proteoglycans play a significant role in modulating cell-cell and cell-matrix interactions in nervous tissue development and differentiation.

Translational repression by a complex between the iron-responsive element of ferritin mRNA and its specific cytoplasmic binding protein is position-dependent in vivo.

The interaction of ferritin mRNA is regulated by iron via the interaction of a cytoplasmic binding protein (IRE-BP) with a specific stem-loop structure in the 5' untranslated region (UTR), referred to as the iron-responsive element (IRE). A high affinity RNA-protein complex between the IRE and the IRE-BP functions as a repressor of translation in vivo. Translational repression appears to depend upon the position of the IRE in the 5' UTR of the mRNA. IREs located in the 5' untranslated region 67 nucleotides or more downstream of the 5' terminus of the mRNA fail to mediate iron-dependent translational regulation and give rise to constitutively derepressed transcripts. A model is proposed in which translational regulation of protein biosynthesis involves a position-dependent interference of the IRE/IRE-BP complex with one of the initial steps in translation initiation.

Chondroitin sulfate and heparan sulfate proteoglycans of PC12 pheochromocytoma cells.

We have isolated and characterized the cell-associated and secreted proteoglycans synthesized by a clonal line of rat adrenal medullary PC12 pheochromocytoma cells, which have been extensively employed for the study of a wide variety of neurobiological processes. Chondroitin sulfate accounts for 70-80% of the [35S] sulfate-labeled proteoglycans present in PC12 cells and secreted into the medium. Two major chondroitin sulfate proteoglycans were detected with molecular sizes of 45,000-100,000 and 120,000-190,000, comprising 14- and 105-kDa core proteins and one or two chondroitin sulfate chains with an average molecular size of 34 kDa. In contrast to the chondroitin sulfate proteoglycans, one major heparan sulfate proteoglycan accounts for most of the remaining 20-30% of the [35S] sulfate-labeled proteoglycans present in the PC12 cells and medium. It has a molecular size of 95,000-170,000, comprising a 65-kDa core protein and two to six 16-kDa heparan sulfate chains. Both the chondroitin sulfate and heparan sulfate proteoglycans also contain O-glycosidically linked oligosaccharides (25-28% of the total oligosaccharides) and predominantly tri- and tetraantennary N-glycosidic oligosaccharides. Proteoglycans produced by the original clone of PC12 cells were compared with those of two other PC12 cell lines (B2 and F3) that differ from the original clone in morphology, adhesive properties, and response to nerve growth factor. Although the F3 cells (a mutant line derived from B2 and reported to lack a cell surface heparan sulfate proteoglycan) do not contain a large molecular size heparan sulfate proteoglycan species, there was no significant difference between the B2 and F3 cells in the percentage of total heparan sulfate released by mild trypsinization, and both the B2 and F3 cells synthesized cell-associated and secreted chondroitin sulfate and heparan sulfate proteoglycans having properties very similar to those of the original PC12 cell line but with a reversed ratio (35:65) of chondroitin sulfate to heparan sulfate.

Phosphatidylinositol-anchored glycoproteins of PC12 pheochromocytoma cells and brain.

PC12 pheochromocytoma cells and cultures of early postnatal rat cerebellum were labeled with [3H]glucosamine, [3H]fucose, [3H]leucine, [3H]ethanolamine, or sodium [35S]sulfate and treated with a phosphatidylinositol-specific phospholipase C. Enzyme treatment of [3H]glucosamine- or [3H]fucose-labeled PC12 cells led to a 15-fold increase in released glycoproteins. On sodium dodecyl sulfate-polyacrylamide gel electrophoresis, most of the released material migrated as a broad band with an apparent molecular size of 32,000 daltons (Da), which was specifically immunoprecipitated by a monoclonal antibody to the Thy-1 glycoprotein. A second glycoprotein, with an apparent molecular size of 158,000 Da, was also released. After treatment with endo-beta-galactosidase, 40-45% of the [3H]glucosamine or [3H]fucose radioactivity in the phospholipase-released glycoproteins was converted to products of disaccharide size, and the molecular size of the 158-kDa glycoprotein decreased to 145 kDa, demonstrating that it contains fucosylated poly-(N-acetyllactosaminyl) oligosaccharides. The phospholipase also released labeled Thy-1 and the 158-kDa glycoprotein from PC12 cells cultured in the presence of [3H]ethanolamine, which specifically labels this component of the phosphatidylinositol membrane-anchoring sequence, while in the lipid-free protein residue of cells not treated with phospholipase, Thy-1 and a doublet at 46/48 kDa were the only labeled proteins. At least eight early postnatal rat brain glycoproteins also appear to be anchored to the membrane by phosphatidylinositol. Sulfated glycoproteins of 155, 132/134, 61, and 21 kDa are the predominant species released by phospholipase, which does not affect a major 44-kDa protein seen in [3H]ethanolamine-labeled brain cultures. The 44-48- and 155/158-kDa proteins may be common to both PC12 cells and brain.

Occurrence of the HNK-1 epitope (3-sulfoglucuronic acid) in PC12 pheochromocytoma cells, chromaffin granule membranes, and chondroitin sulfate proteoglycans.

After biosynthetic labeling of sulfated glycoproteins in rat and goldfish brain and PC12 pheochromocytoma cells with sodium [35S]sulfate, it was observed that all of the bands reactive with the HNK-1 antibody on immunoblots of sodium dodecyl sulfate-polyacrylamide gels corresponded with sulfate-labeled proteins detected by fluorography. These results support data from other studies, which indicate that the HNK-1 epitope is a 3-sulfo-glucuronic acid residue. In addition to its presence in a wide range of nervous tissue glycoproteins, the HNK-1 epitope was also detected in chromaffin granule membranes, chondroitinase ABC, and in chondroitin sulfate proteoglycans of brain, cartilage, and chondrosarcoma. However, it is not present in the heparan sulfate proteoglycan of brain, or in either of two chondroitin sulfate/dermatan sulfate proteoglycans in the chromaffin granule matrix.