Cloning of the cDNA encoding an RNA regulatory protein--the human iron-responsive element-binding protein

Down-regulation of iron regulatory protein 1 gene expression by nitric oxide

Iron regulatory protein 1 (IRP1) is an RNA binding protein that posttranscriptionally modulates the expression of mRNAs coding for proteins involved in iron metabolism. It has long been held that its RNA binding activity is regulated posttranslationally by the insertion/extrusion of a 4Fe-4S cluster, without changes in IRP1 levels. However, the question of a possible regulation of the expression of this protein has remained open. In the present study we analyzed the modulation of IRP1 expression in murine macrophages. We showed that activation by IFN-gamma and/or lipopolysaccharide, which induces IRP1 RNA binding activity via nitric oxide (NO), results simultaneously in a reduction in IRP1 protein levels, as determined by Western blot analyses. IRP1 expression decreased time-dependently to about 40% of control levels after 16 h. Down-regulation of IRP1 protein levels was correlated with the amount of NO produced and was partially abolished by the NO synthase (NOS) inhibitor N-monomethyl-l-arginine. No changes in IRP1 levels could be detected in stimulated peritoneal macrophages from NOS2 knockout (NOS2(-/-)) mice, unlike wild-type mice. Converse modulation of IRP1 RNA binding activity and IRP1 levels could be reproduced by exogenous NO and also was observed in nonmacrophage cells cocultured with NO-producing macrophages. We also analyzed IRP1 mRNA levels by Northern blotting and found a decrease in IRP1 mRNA expression after stimulation with IFN-gamma plus lipopolysaccharide, which was abrogated in the presence of N-monomethyl-l-arginine. This is evidence that IRP1 is regulated by a physiological stimulus other than posttranslationally.

Characterization of fibroblast cytoplasmic proteins that bind to the 3' UTR of human catalase mRNA

The excessive expression of catalase protein and its activity in cultured skin fibroblast from Zellweger Syndrome (ZS), a disorder of peroxisomal biogenesis, was found to be regulated at the translational level (J. Neurochem. 67: 2373-2378, 1996). Overall there is a considerable increase in the association of catalase mRNA with polysomes in ZS cell lines as compared to control indicating translational upregulation. To investigate the possibility that RNA-protein interactions are involved in the mediation of this increase in translation, the interaction between 3' untranslated region of human catalase mRNA and human fibroblast cytoplasmic proteins were investigated by RNA gel shift assay technique. Competition experiments demonstrated that all the 600 bases of 3' UTR (of human catalase gene) was required for efficient binding. Catalase RNA- protein interaction was sensitive to the altered redox state in these in vitro assays and this RNA-protein interaction could be enhanced by the addition of beta-mercaptoethanol in cytoplasm from control fibroblast but not in cytoplasm from ZS fibroblast. UV cross linked RNA-protein complexes on SDS polyacrylamide gel electrophoresis revealed the presence of at least four protein bands with approximate molecular masses of 38 kDa, 50 kDa, 66 kDa and 80 kDa. The potential role of these mRNA binding proteins in the regulation of catalase gene expression is discussed.

Ferriporphyrins and endothelium: a 2-edged sword—promotion of oxidation and induction of cytoprotectants

Heme arginate infusions blunt the symptoms of patients with acute intermittent porphyria without evidence of the vascular or thrombotic side effects reported for hematin. To provide a rationale for heme arginate's safety, the present study examined the effects of various ferriporphyrins to sensitize human endothelial cells to free radical injury and to induce heme oxygenase and ferritin expression. Heme arginate, unlike hematin, did not amplify oxidant-induced cytotoxicity mediated by hydrogen peroxide (5.3 ± 2.4 versus 62.3 ± 5.3% 51Cr release,P < .0001) or by activated neutrophils (14.4 ± 2.9 versus 41.1 ± 6.0%, P < .0001). Nevertheless, heme arginate efficiently entered endothelial cells similarly to hematin, since both markedly induced heme oxygenase mRNA (more than 20-fold increase) and enzyme activity. Even with efficient permeation, endothelial cell ferritin content was only minimally increased by heme arginate compared with a 10-fold induction by hematin; presumably less free iron was derived from heme arginate despite up-regulation of heme oxygenase. Hematin is potentially vasculopathic by its marked catalysis of oxidation of low-density lipoprotein (LDL) to endothelial-toxic moieties. Heme arginate was significantly less catalytic. Heme arginate–conditioned LDL was less than half as cytotoxic to endothelial cells as hematin-conditioned LDL (P < .004). It is concluded that heme arginate may be less vasculotoxic than hematin since it is an effective heme oxygenase gene regulator but a less efficient free-radical catalyst.

Ferriporphyrins and endothelium: a 2-edged sword—promotion of oxidation and induction of cytoprotectants

Heme arginate infusions blunt the symptoms of patients with acute intermittent porphyria without evidence of the vascular or thrombotic side effects reported for hematin. To provide a rationale for heme arginate's safety, the present study examined the effects of various ferriporphyrins to sensitize human endothelial cells to free radical injury and to induce heme oxygenase and ferritin expression. Heme arginate, unlike hematin, did not amplify oxidant-induced cytotoxicity mediated by hydrogen peroxide (5.3 ± 2.4 versus 62.3 ± 5.3% 51Cr release,P &lt; .0001) or by activated neutrophils (14.4 ± 2.9 versus 41.1 ± 6.0%, P &lt; .0001). Nevertheless, heme arginate efficiently entered endothelial cells similarly to hematin, since both markedly induced heme oxygenase mRNA (more than 20-fold increase) and enzyme activity. Even with efficient permeation, endothelial cell ferritin content was only minimally increased by heme arginate compared with a 10-fold induction by hematin; presumably less free iron was derived from heme arginate despite up-regulation of heme oxygenase. Hematin is potentially vasculopathic by its marked catalysis of oxidation of low-density lipoprotein (LDL) to endothelial-toxic moieties. Heme arginate was significantly less catalytic. Heme arginate–conditioned LDL was less than half as cytotoxic to endothelial cells as hematin-conditioned LDL (P &lt; .004). It is concluded that heme arginate may be less vasculotoxic than hematin since it is an effective heme oxygenase gene regulator but a less efficient free-radical catalyst.

Changes in gene expression with iron loading and chelation in cardiac myocytes and non-myocytic fibroblasts

Iron overload is associated with long-term cardiac iron accumulation and tissue changes such as fibrosis. To determine short-term iron-dependent changes in expression of genes associated with iron homeostasis and fibrosis we measured mRNA on Northern blots prepared from cultured rat neonatal cardiomyocytes and non-myocytes (fibroblasts) as a function of iron loading and chelation. Transferrin receptor mRNA was reduced in myocytes exposed to various concentrations of iron for 3 days and this decline was associated with a 63% decline in iron-response element (IRE) binding of iron regulatory protein-1, indicating that myocytes utilize IRE-dependent mechanisms to modulate gene expression. In myocytes iron caused a dose-dependent decline in mRNAs coding for transforming growth factor- beta(1)(TGF- beta(1)), biglycan, and collagen type I while plasminogen activator inhibitor-1 mRNA was unaffected by iron loading and decorin mRNA doubled. Total TGF- beta bioactivity was also decreased by iron loading. Thus, the effects of iron loading on genes related to cardiac fibrosis are gene-specific. Addition of deferoxamine for 1 day did not have any significant effect on any of these genes. Parallel changes in gene expression were exhibited by non-myocytes (fibroblasts), where chelation also decreased TGF- beta(1)mRNA and activity, and mRNA for collagen type I and biglycan, and collagen synthesis. In addition to these changes in transcripts associated with matrix formation the mRNA of the metabolic enzyme glyceraldehyde-3-phosphate dehydrogenase was unaffected by iron loading but doubled in both cell types upon treatment with deferoxamine. These findings suggest that in both cardiac myocytes and non-myocyte fibroblasts gene expression is coupled to intracellular iron pools by gene-specific and IRE-dependent and idependent mechanisms. This linkage may influence matrix deposition, a significant component of cardiac injury.

A developmentally regulated aconitase related to iron-regulatory protein-1 is localized in the cytoplasm and in the mitochondrion of Trypanosoma brucei

Mitochondrial energy metabolism and Krebs cycle activities are developmentally regulated in the life cycle of the protozoan parasite Trypanosoma brucei. Here we report cloning of a T. brucei aconitase gene that is closely related to mammalian iron-regulatory protein 1 (IRP-1) and plant aconitases. Kinetic analysis of purified recombinant TbACO expressed in Escherichia coli resulted in a K(m) (isocitrate) of 3 +/- 0.4 mM, similar to aconitases of other organisms. This was unexpected since an arginine conserved in the aconitase protein family and crucial for substrate positioning in the catalytic center and for activity of pig mitochondrial aconitase (Zheng, L., Kennedy, M. C., Beinert, H., and Zalkin, H. (1992) J. Biol. Chem. 267, 7895-7903) is substituted by leucine in the TbACO sequence. Expression of the 98-kDa TbACO was shown to be lowest in the slender bloodstream stage of the parasite, 8-fold elevated in the stumpy stage, and increased a further 4-fold in the procyclic stage. The differential expression of TbACO protein contrasted with only minor changes in TbACO mRNA, indicating translational or post-translational mechanisms of regulation. Whereas animal cells express two distinct compartmentalized aconitases, mitochondrial aconitase and cytoplasmic aconitase/IRP-1, TbACO accounts for total aconitase activity in trypanosomes. By cell fractionation and immunofluorescence microscopy, we show that native as well as a transfected epitope-tagged TbACO localizes in both the mitochondrion (30%) and in the cytoplasm (70%). Together with phylogenetic reconstructions of the aconitase family, this suggests that animal IRPs have evolved from a multicompartmentalized ancestral aconitase. The possible functions of a cytoplasmic aconitase in trypanosomes are discussed.

Control of transferrin receptor expression via nitric oxide-mediated modulation of iron-regulatory protein 2

Cellular iron storage and uptake are coordinately regulated post-transcriptionally by cytoplasmic factors, iron-regulatory proteins 1 and 2 (IRP-1 and IRP-2). When iron in the intracellular transit pool is scarce, IRPs bind to iron-responsive elements (IREs) in the 5'-untranslated region of the ferritin mRNA and 3'-untranslated region of the transferrin receptor (TfR) mRNA. Such binding inhibits translation of ferritin mRNA and stabilizes the mRNA for TfR, whereas the opposite scenario develops when iron in the transit pool is plentiful. However, we (Richardson, D. R., Neumannova, V., Nagy, E., and Ponka, P. (1995) Blood 86, 3211-3219) and others reported that the binding of IRPs to IREs can also be modulated by nitric oxide (NO). In this study, we showed that a short exposure of RAW 264.7 cells (a murine macrophage cell line) to the NO(+) donor, sodium nitroprusside (SNP), caused a significant decrease in IRP-2 binding to the IREs followed by IRP-2 degradation and that these changes occurred without affecting IRP-1 binding. The SNP-mediated degradation of IRP-2 in RAW 264.7 cells could be prevented by MG-132 or lactacystin, known inhibitors of proteasome-dependent protein degradation. A SNP-mediated decrease in IRP-2 binding and levels was associated with a dramatic decrease in TfR mRNA levels and an increase in ferritin synthesis. Importantly, the proteasome inhibitor MG-132 prevented the SNP-mediated decrease in TfR mRNA levels. These observations suggest that IRP-2 can play an important role in controlling transferrin receptor expression.

Regulation of genes of iron metabolism by the iron-response proteins

Iron is an essential nutrient, yet excess iron can be toxic to cells. The uptake of iron by mammalian cells is post-transcriptionally regulated by the interaction of iron-response proteins (IRP1 and IRP2) with iron-response elements (IREs) found in the mRNAs of genes of iron metabolism, such as ferritin, the transferrin receptor, erythroid aminolevulinic acid synthase, and mitochondrial aconitase. The IRPs are RNA binding proteins that bind to the IRE (found in the mRNAs of the regulated genes) in an iron- dependent manner. Binding of IRPs to the IREs leads to changes in the expression of the regulated genes and subsequent changes in the uptake, utilization, or storage of intracellular iron. Recent work has demonstrated that the binding of the IRPs to the IREs can also be modulated by changes in the redox state or oxidative stress level of the cell. These findings provide an important link between iron metabolism and states of oxidative stress.

Ligand-induced structural alterations in human iron regulatory protein-1 revealed by protein footprinting

Human iron regulatory protein-1 (IRP-1) is a bifunctional protein that regulates iron metabolism by binding to mRNAs encoding proteins involved in iron uptake, storage, and utilization. Intracellular iron accumulation regulates IRP-1 function by promoting the assembly of an iron-sulfur cluster, conferring aconitase activity to IRP-1, and hindering RNA binding. Using protein footprinting, we have studied the structure of the two functional forms of IRP-1 and have mapped the surface of the iron-responsive element (IRE) binding site. Binding of the ferritin IRE or of the minimal regulatory region of transferrin receptor mRNA induced strong protections against proteolysis in the region spanning amino acids 80 to 187, which are located in the putative cleft thought to be involved in RNA binding. In addition, IRE-induced protections were also found in the C-terminal domain at Arg-721 and Arg-728. These data implicate a bipartite IRE binding site located in the putative cleft of IRP-1. The aconitase form of IRP-1 adopts a more compact structure because strong reductions of cleavage were detected in two defined areas encompassing residues 149 to 187 and 721 to 735. Thus both ligands of apo-IRP-1, the IRE and the 4Fe-4S cluster, induce distinct but overlapping alterations in protease accessibility. These data provide evidences for structural changes in IRP-1 upon cluster formation that affect the accessibility of residues constituting the RNA binding site.

Iron regulatory protein as an endogenous sensor of iron in rat intestinal mucosa. Possible implications for the regulation of iron absorption

Duodenal enterocytes adjust intestinal iron absorption to the body's state of iron repletion. Here we tested how iron supply from the blood modulates the RNA-binding activity of iron regulatory proteins (IRP-1 and IRP-2) in immature duodenal rat enterocytes, and whether the modulation is compatible with the hypothesis that IRPs, in turn, may regulate the expression of iron transport proteins in maturating enterocytes during migration to the villus tips. Tissue uptake of parenterally applied 59Fe along the duodenal crypt-villus axis was compared to local IRP-1 and IRP-2 activity and to duodenal 59Fe transport capacity 12 h, 48 h, and 72 h after intravenous iron administration to iron-deficient rats. IRP-1 and IRP-2 activity was significantly increased in iron-deficiency. 59Fe administrated from the blood side was almost exclusively taken up by crypt enterocytes. Accordingly, the activity of IRP-1 decreased at this site 12 h after parenteral iron administration, but remained high at the villus tips. After 48 h the bulk of 59Fe containing enterocytes had migrated to the villus tips. Correspondingly, IRP-1 activity was decreased at duodenal villus tips after 48 h. IRP-2 activity also tended to decrease, though the change was statistically not significant. IRP-2 activity remained significantly higher at duodenal villus tips than in crypts, even after 72 h. Intestinal iron absorption capacity decreased with the same delay as IRP-1 activity after intravenous iron administration. In the ileum 59Fe uptake from the blood and IRP activity showed no significant difference between crypt and villus region. Luminal administration of iron decreased duodenal IRP-1 and IRP-2 activity at tips and crypts within 2 h. Thus, recently absorbed iron becomes available to cytosolic IRP during its passage through the enterocyte. Our results are compatible with a role of IRPs in gearing the expression of intestinal iron transporters in the duodenal brushborder to the body's state of iron repletion.

Desferioxamine increases iron depletion and apoptosis induced by ara-C of human myeloid leukaemic cells

We investigated whether changes in iron metabolism and the transferrin receptor (TRF-R) expression were involved in the antileukaemic effects of arabinoside cytosine (ara-C). Treatment with 100 nM ara-C for 48h reduced thymidine uptake and increased the surface expression of the TRF-R on leukaemic blasts derived from 13/16 (81%) patients and on the HL-60 and U-937 cell lines. Whereas intracellular non-haem iron was strongly depleted 24 h after ara-C addition, TRF-R up-regulation and recovery of intracellular non-haem iron concentration occurred together after a longer exposure of the cultured cells to the drug. Since iron is an essential regulator of cell proliferation we have evaluated the effects of the combination between ara-C and the iron chelator desferioxamine (DSF) on the growth of HL-60 and U-937 cells. We found that desferioxamine strongly potentiated the effects of ara-C on leukaemic cell growth inhibition and apoptosis. This is the first report of a positive interaction between ara-C and an iron chelator in terms of antileukaemic effects.

Lack of coordinate control of ferritin and transferrin receptor expression during rat liver regeneration

Transferrin receptor (TfR) and ferritin, key proteins of cellular iron metabolism, are coordinately and divergently controlled by cytoplasmic proteins (iron regulatory proteins, IRP-1 and IRP-2) that bind to conserved mRNA motifs called iron-responsive elements (IRE). IRP, in response to specific stimuli (low iron levels, growth and stress signals) are activated and prevent TfR mRNA degradation and ferritin mRNA translation by hindering ferritin mRNA binding to polysomes. We previously found that, in regenerating liver, IRP activation was accompanied by increased TfR mRNA levels, but not by reduced ferritin expression. The basis for this unexpected behavior was investigated in the present study. Liver regeneration triggered by carbon tetrachloride (CCl4) stimulated by four- to fivefold the synthesis of both L and H ferritin chains. This increase was accompanied with a transcriptionally regulated twofold rise in the amount of ferritin mRNAs. Moreover, polysome-associated ferritin transcripts were fourfold higher in CCl4-treated animals than in control animals. Because RNA bandshift assays showed a fourfold increase in IRP-2 binding activity after CCl4 administration, activated IRP in regenerating liver seemed unable to prevent ferritin mRNAs binding to polysomes. This was confirmed by direct demonstration in the wheat germ translation system that the efficiency of IRP as a translational repressor of a mRNA bearing an IRE motif in front of a reporter transcript is impaired in CCl4-treated rats in spite of an enhanced IRE-binding capacity. In conclusion, we show for the first time that the paradigm of coordinate and opposite control of ferritin and TfR by IRP is contradicted in liver regeneration. Under these circumstances, growth-dependent signals may activate ferritin gene transcription and at the same time hamper the ability of activated IRP-2 to repress translation of ferritin mRNAs, thus preserving for growing liver cells an essential iron-storage compartment.

Characterization of the human mitochondrial aconitase gene (ACO2)

We have cloned and characterized the ACO2 gene on human chromosome 22q13, which encodes the essential iron-dependent metabolic enzyme mitochondrial aconitase. We determined that the ACO2 gene comprises 18 translated exons distributed over approximately 35 kilobasepairs (kbp) of DNA. We have shown that the ACO2 mRNA is 2.7kb in length and is expressed ubiquitously, and we can detect multiple isoforms of the ACO2 protein. As others had reported the existence of biochemically active electrophoretic variants of mitochondrial aconitase, we wished to find common ACO2 gene allozymes, functional polymorphisms that might be associated with susceptibility to human genetic diseases. We looked for ACO2 allozymes by DNA sequencing and genotyping in a population of 217 subjects, many of which had idiopathic Parkinson's disease (IPD). We studied patients with IPD because this movement disorder is thought to arise from defects in neuronal iron and energy metabolism, two properties with which aconitase is involved. Furthermore, reports of associations between alleles of the CYP2D6 locus (nearby on 22q13) and IPD, although inconsistent, indicated that an IPD susceptibility locus might be in strong linkage disequilibrium with CYP2D6. We found three functionally silent single nucleotide polymorphisms (SNPs) present in transcribed sequences that exist in similar frequencies in IPD patients and healthy controls. These ACO2 SNPs are in linkage disequilibrium with each other, providing evidence for distinct ACO2 haplotypes. We have, as yet, not detected polymorphisms that would lead to ACO2 allozymes, nor have we observed differences in ACO2 isoform prevalence or distribution in our population of IPD patients and controls. We conclude it is unlikely that polymorphism in the ACO2 gene or post-translational modification of the enzyme predispose to IPD.

Abnormal localization of iron regulatory protein in Alzheimer's disease

A role for altered iron metabolism in the pathogenesis of Alzheimer's disease has been suggested by several reports associating the cardinal neuropathologic lesions with markers of free radical-induced damage and redox-active iron. We hypothesized that the abnormal distribution of iron in Alzheimer brain might result from alterations in iron regulatory proteins (IRP) such as IRP-1 and IRP-2, the main control elements of cellular iron homeostasis. Here, we report that while IRP-1 is present at similar levels in both Alzheimer and control brain tissue, IRP-2 shows striking differences and is associated with intraneuronal lesions, including neurofibrillary tangles, senile plaque neurites and neuropil threads. Since IRP-2 colocalizes with redox-active iron, our results suggest that alterations in IRP-2 might be directly linked to impaired iron homeostasis in Alzheimer's disease.

Identification of proteins binding specifically to the 3'-untranslated region of granulocyte/macrophage-colony stimulating factor mRNA

The 3'-untranslated region of granulocyte/macrophage colony-stimulating factor (GM-CSF) mRNA contributes to the post-transcriptional regulation of gene expression. Degradation is partly mediated by adenosine- uridine-rich sequence elements (ARE), which serve as binding sites for specific proteins. Stabilization of RNA by phytohemagglutinin and concanavalin A treatment is dependent on regulatory sequence elements upstream of ARE. We have performed northwestern blot and filter binding assays using cell extracts and RNA sequences containing or lacking ARE. Murine and human T cell extracts (EL-4 and Jurkat) yielded two specific proteins of 93 and 94 kDa, respectively, that were binding to sequences upstream of ARE. Within this region, the human and murine RNA do not share any obvious sequence identity, yet both are target sites for the binding proteins. The smallest RNA fragments protected by the proteins from RNase A digestion, were 44 in the murine, and 38 ribonucleotides long in the human sequence. The binding activity of the 94 kDa protein derived from human Jurkat cells could be enhanced by phytohemagglutinin. The interaction with regulatory mRNA sequences and the responsiveness to phytohemagglutinin suggests that the proteins are involved in controlling GM-CSF mRNA turnover.

Regulation of antioxidant enzyme expression by NGF

The rapid decreases in viability seen in H2O2-treated PC12 cells reflect enhanced susceptibility of neural cell types to oxidant injury. The dose-response relationship between NGF concentration and survival after H2O2 treatment resembles that for NGF effects on PC12 survival in serumless medium. Previously we have shown that NGF treatment enhances the activity of GSH-Px and catalase which catalyze the degradation of H2O2. Here in order to ascertain whether NGF stimulates transcription, affects mRNA stability, or acts post-transcriptionally, we measured catalase and GSH-Px mRNA half-lives. While both catalase and GSH-Px transcripts are stable with a relatively long half life and a gradual decay in mRNA levels, NGF had different effects on their stability. NGF had marked effects on catalase mRNA stability. The catalase gene has a 3' flanking region with T-rich clusters and CA repeats known to be susceptible to regulation by destabilization or ubiquination. NGF maintained catalase mRNA levels of actinomycin D (ACT-D) treated PC12 cells at twice that of cells exposed to ACT-D alone, delaying the rate of decay for catalase mRNA for 24 h. The NGF induction of GSH-Px and catalase mRNA was inhibited by cycloheximide (CHX) treatment with a slight decrease in their mRNA levels due to prolonged exposure to CHX. When the CHX treatment was delayed relative to the NGF treatment there was no effect on NGF effects on catalase and GSH-Px. The GSH-Px gene has conserved sequences in the open reading frame and 3' untranslated region which forms a stem-loop structure necessary for the incorporation of Se into this selenoprotein. While Se is important in stabilizing GSH-Px transcripts, it did not affect transcription rates or mRNA stability. These results are consistent with the hypothesis that NGF regulates catalase and GSH-Px expression via a primary effect on transcription factor pathways.

Interaction between iron-regulatory proteins and their RNA target sequences, iron-responsive elements

In this chapter, we have focused on the biochemistry of IRP-1 and the features which distinguish it from the related RNA-binding protein, IRP-2. IRP-1 is the cytoplasmic isoform of the enzyme aconitase, and, depending on iron status, may switch between enzymatic and RNA-binding activities. IRP-1 and IRP-2 are trans-acting regulators of mRNAs involved in iron uptake, storage and utilisation. The finding of an IRE in the citric acid cycle enzymes, mitochondrial aconitase and succinate dehydrogenase, suggests that the IRPs may also influence cellular energy production. These two proteins appear to bind RNAs with different but overlapping specificity, suggesting that they may regulate the stability or translation of as yet undefined mRNA targets, possibly extending their regulatory function beyond that of iron homeostasis. The interaction between the IRPs and the IRE represents one of the best characterised model systems for posttranscriptional gene control, and given that each IRP can also recognise its own unique set of RNAs, the search for new in vivo mRNA targets is expected to provide yet more surprises and insights into the fate of cytoplasmic mRNAs.

Regulation of iron metabolism in eukaryotes

Iron metabolism is regulated in cells to ensure that iron supplies are adequate and nontoxic. The expression of iron metabolism is regulated primarily by posttranscriptional mechanisms. Ferritin, eALAS, SDHb of Drosophila, and mammalian mitochondrial aconitase are translationally regulated. The TfR is regulated at the level of mRNA stability. Iron regulatory proteins are regulated either by assembly or by disassembly of an iron-sulfur cluster (IRP1) or by rapid degradation in the presence of iron (IRP2). The list of targets for IRP-mediated regulation is growing longer, and a range of possibilities for versatile regulation exists, as each IRP can bind to unique targets that differ from the consensus IRE. The reactivity of iron with oxygen and the creation of toxic by-products may be the evolutionary stimulus that produced this system of tight posttranscriptional gene regulation.

Iron regulatory proteins 1 and 2

Iron uptake and storage in mammalian cells is at least partly regulated at a post-transcriptional level by the iron regulatory proteins (IRP-1 and IRP-2). These cytoplasmic regulators share 79% similarity in protein sequence and bind tightly to conserved mRNA stem-loops, named iron-responsive elements (IREs). The IRP:IRE interaction underlies the regulation of translation and stability of several mRNAs central to iron metabolism. The question of why the cell requires two such closely related regulatory proteins may be resolved as we learn more about the expression and regulation of these proteins. It is evident so far that, despite similarities, the IRPs differ in several important respects. They are coordinately regulated by cellular iron, but whereas IRP-1 is inactivated by high iron levels, IRP-2 is rapidly degraded. Further differences arise in their expression and RNA-binding specificity. The two proteins each recognise a large repertoire of IRE-like sequences, including a small group of exclusive RNA targets. These findings hint that IRP-1 and IRP-2 may bind preferentially to certain mRNAs in vivo, possibly extending their known functions beyond the regulation of intracellular iron homeostasis.

Molecular control of vertebrate iron metabolism: mRNA-based regulatory circuits operated by iron, nitric oxide, and oxidative stress

As an essential nutrient and a potential toxin, iron poses an exquisite regulatory problem in biology and medicine. At the cellular level, the basic molecular framework for the regulation of iron uptake, storage, and utilization has been defined. Two cytoplasmic RNA-binding proteins, iron-regulatory protein-1 (IRP-1) and IRP-2, respond to changes in cellular iron availability and coordinate the expression of mRNAs that harbor IRP-binding sites, iron-responsive elements (IREs). Nitric oxide (NO) and oxidative stress in the form of H2O2 also signal to IRPs and thereby influence cellular iron metabolism. The recent discovery of two IRE-regulated mRNAs encoding enzymes of the mitochondrial citric acid cycle may represent the beginnings of elucidating regulatory coupling between iron and energy metabolism. In addition to providing insights into the regulation of iron metabolism and its connections with other cellular pathways, the IRE/IRP system has emerged as a prime example for the understanding of translational regulation and mRNA stability control. Finally, IRP-1 has highlighted an unexpected role for iron sulfur clusters as post-translational regulatory switches.

Interpreting cDNA sequences: some insights from studies on translation

This review discusses some rules for assessing the completeness of a cDNA sequence and identifying the start site for translation. Features commonly invoked-such as an ATG codon in a favorable context for initiation, or the presence of an upstream in-frame terminator codon, or the prediction of a signal peptide-like sequence at the amino terminus-have some validity; but examples drawn from the literature illustrate limitations to each of these criteria. The best advice is to inspect a cDNA sequence not only for these positive features but also for the absence of certain negative indicators. Three specific warning signs are discussed and documented: (i) The presence of numerous ATG codons upstream from the presumptive start site for translation often indicates an aberration (sometimes a retained intron) at the 5' end of the cDNA. (ii) Even one strong, upstream, out-of-frame ATG codon poses a problem if the reading frame set by the upstream ATG overlaps the presumptive start of the major open reading frame. Many cDNAs that display this arrangement turn out to be incomplete; that is, the out-of-frame ATG codon is within, rather than upstream from, the protein coding domain. (iii) A very weak context at the putative start site for translation often means that the cDNA lacks the authentic initiator codon. In addition to presenting some criteria that may aid in recognizing incomplete cDNA sequences, the review includes some advice for using in vitro translation systems for the expression of cDNAs. Some unresolved questions about translational regulation are discussed by way of illustrating the importance of verifying mRNA structures before making deductions about translation.

The cis-acting elements involved in endonucleolytic cleavage of the 3' UTR of human IGF-II mRNAs bind a 50 kDa protein

Site-specific cleavage of human insulin-like growth factor II mRNAs requires two cis-acting elements, I and II, that are both located in the 3' untranslated region and separated by almost 2 kb. These elements can interact and form a stable RNA-RNA stem structure. In this study we have initiated the investigation of transacting factors involved in the cleavage of IGF-II mRNAs. The products of the cleavage reaction accumulate in the cytoplasm, suggesting that cleavage occurs in this cellular compartment. By electrophoretic mobility shift assays, we have identified a cytoplasmic protein with an apparent molecular weight of 48-50 kDa, IGF-II cleavage unit binding protein (ICU-BP), that binds to the stem structure formed by interaction of parts of the cis-acting elements I and II. The binding is resistant to high K+ concentrations and is dependent on Mg2+. In addition, ICU-BP binding is dependent on the cell density and correlates inversely with the IGM-II mRNA levels. In vivo cross-linking data show that this protein is associated with IGF-II mRNAs in vivo.

The impact of oxidative stress on eukaryotic iron metabolism

The processes of iron uptake and distribution are highly regulated in mammalian cells. Expression of the transferrin receptor is increased when cells are iron-depleted, while expression of the iron sequestration protein ferritin is increased in cells that are iron-replete. Regulation of expression of proteins of iron uptake (transferrin receptor) and iron sequestration (ferritin) presumably ensures that levels of reactive free iron are not high in cells. Formation of reactive oxygen species occurs when free iron reacts with oxygen, and tight regulation of iron metabolism may enable cells to avoid engaging in destructive chemical reactions. Levels of intracellular iron are directly sensed by two iron sensing proteins. Iron regulatory protein 1 (IRP1) is a bifunctional protein; in cells that are iron-replete, IRP1 contains an iron-sulfur cluster and functions as cytosolic aconitase. In cells that are iron-depleted, IRP1 binds stem-loop structures in RNA transcripts known as iron responsive elements (IREs). Iron regulatory protein 2 (IRP2) binds similar stem-loop structures, but the mode of regulation of IRP2 is different in that IRP2 is rapidly degraded in iron-replete cells. The post-transcriptional regulation of genes of iron metabolism in mammalian cells ensures that cells have an adequate supply of iron, and also ensures that cells do not generate excess reactive oxygen species through the interaction of free iron and oxygen.

Iron regulates the intracellular degradation of iron regulatory protein 2 by the proteasome

Iron regulatory proteins (IRP1 and IRP2) are RNA-binding proteins that bind to specific structures, termed iron-responsive elements (IREs), that are located in the 5'- or 3'-untranslated regions of mRNAs that encode proteins involved in iron homeostasis. IRP1 and IRP2 RNA binding activities are regulated by iron; IRP1 and IRP2 bind IREs with high affinity in iron-depleted cells and with low affinity in iron-repleted cells. The decrease in IRP1 RNA binding activity occurs by a switch between apoprotein and 4Fe-4S forms, without changes in IRP1 levels, whereas the decrease in IRP2 RNA binding activity reflects a reduction in IRP2 levels. To determine the mechanism by which iron decreases IRP2 levels, we studied IRP2 regulation by iron in rat hepatoma and human HeLa cells. The iron-dependent decrease in IRP2 levels was not due to a decrease in the amount of IRP2 mRNA or to a decrease in the rate of IRP2 synthesis. Pulse-chase experiments demonstrated that iron resulted in a 3-fold increase in the degradation rate of IRP2. IRP2 degradation depends on protein synthesis, but not transcription, suggesting a requirement for a labile protein. IRP2 degradation is not prevented by lysosomal inhibitors or calpain II inhibitors, but is prevented by inhibitors that block proteasome function. These data suggest the involvement of the proteasome in iron-mediated IRP2 proteolysis.

Characterization and expression of iron regulatory protein 2 (IRP2). Presence of multiple IRP2 transcripts regulated by intracellular iron levels

Iron regulatory proteins (IRP1 and IRP2) are RNA-binding proteins that bind to stem-loop structures, termed iron-responsive elements (IREs), present in either the 5'- or 3'-untranslated regions of specific mRNAs. The binding of IRPs to 5'-IREs inhibits translation of mRNA, whereas the binding of IRPs to 3'-IREs stabilizes mRNA. To study the structure and regulation of IRP2, we isolated cDNAs for rat and human IRP2. The derived amino acid sequence of rat IPR2 is 93% identical with that of human IRP2 and is present in lower eukaryotes, indicating that IRP2 is highly conserved. IRP1 and IRP2 share 61% overall amino acid identity. IRP2 is ubiquitously expressed in rat tissues, the highest amounts present in skeletal muscle and heart. IRP2 is encoded by multiple mRNAs of 6.4, 4.0, and 3.7 kilobases. The 3'-untranslated region of rat IRP2 contains multiple polyadenylation signals, two of which could account for the 4.0-kb and 3.7-kb mRNAs. The 3.7-kb mRNA is increased in iron-depleted cells and occurs with a reciprocal decrease in the 6.4-kb transcript. These data suggest that the 3.7-kb mRNA is produced by alternative poly(A) site utilization in iron-depleted cells.

Characterization of the nuclear gene encoding mitochondrial aconitase in the marine red alga Gracilaria verrucosa

We have cloned a nuclear gene from the marine red alga Gracilaria verrucosa that encodes the complete 779 amino-acid mitochondrial aconitase (m-ACN), the first characterized from a photosynthetic organism. The N-terminal 28 deduced amino acids are predicted to constitute the mitochondrial transit peptide, the first described from a red alga. Putative transcriptional cis-acting elements were identified in the upstream untranslated region. The G. verrucosa m-ACN gene (m-ACN) is present in a single copy and is located ca. 1.5 kb upstream from the single-copy polyubiquitin gene. The single spliceosomal intron is located near the 5' end of the region encoding the mature m-ACN in precisely the same location and phase as intron 2 in Caenorhabditis elegans m-ACN; sequences at its 3' and 5' splice junctions and at the predicted lariat branch point conform well to the eukaryote consensus sequences. Multiple protein-sequence alignment of m-ACN, bacterial aconitase (b-ACN) and iron-responsive element-binding protein (IRE-BP), and phylogenetic analyses, revealed that m-ACN does not share a recent common ancestry with either b-ACN or IRE-BP.

Translational repressor activity is equivalent and is quantitatively predicted by in vitro RNA binding for two iron-responsive element-binding proteins, IRP1 and IRP2

Iron regulatory proteins (IRPs) bind to specific RNA stem-loop structures known as iron-responsive elements (IREs) which mediate the post-transcriptional regulation of many genes of iron metabolism. Most studies have focused on the role of IRP1, which has previously been shown to bind with high affinity to IREs and mediate repression of in vitro translation of ferritin mRNAs. More recently, a second IRP has been identified that is expressed in all tissues and that binds IREs (Rouault, T. A., Haile, D. H., Downey, W. E., Philpott, C. C., Tang, C., Samaniego, F., Chin, J., Paul, I., Orloff, D., Harford, J. B., and Klausner, R. D. (1992) BioMetals 5, 131-140; Henderson, B. R., Seiser, C., and Kuhn, L. C. (1993) J. Biol. Chem. 268, 27327-27334; Guo, B., Yu, Y., and Leibold, E. A. (1994) J. Biol. Chem. 269, 24252-24260; Samaniego, F., Chin, J., Iwai, K., Rouault, T. A., and Klausner, R. D. (1994) J. Biol. Chem. 269, 30904-30910). Here we report that purified recombinant IRP2 inhibits translation of ferritin mRNAs with a molar efficacy equal to that of recombinant IRP1. There is a quantitative correlation between binding to isolated RNA target motifs, as judged by gel retardation assays, and translational repressor function as assayed in an in vitro translation system. In contrast to IRP1, IRP2 is not inactivated for RNA binding by alkylation with N-ethylmaleimide or phenylmaleimide, and as we would therefore predict, IRP2 treated with N-ethylmaleimide remains an effective repressor of ferritin translation. As IRP1 and IRP2 clearly have equal capability of mediating translational repression in vitro, the contributions of both IRPs to overall regulation must be considered in describing the pathways of iron regulated gene expression in individual cells.

Localization of a regulatory region on the 5'-untranslated region of human hepatoma HepG2 gamma-glutamyltransferase mRNA and response to dexamethasone and antisense oligonucleotide treatment

We are reporting the functional analysis of the 5'-untranslated region (5'UTR) of human hepatoma HepG2 gamma-glutamyltransferase (GGT) mRNA. Transient expression of hybrid GGT-luciferase reporter gene mutants in HepG2 shows that this 5'UTR acts as a tissue-specific translational enhancer. A domain of 173 bases containing a steroid hormone response element (HRE) is responsible for the enhancing effect, which can be amplified by addition of dexamethasone at 10(-6) M. Furthermore, the regulatory role of the 5'UTR is demonstrated by interaction with sense and antisense oligonucleotides.

Iron regulatory elements (IREs): a family of mRNA non-coding sequences

Images

Exclusion of ferritins and iron-responsive element (IRE)-binding proteins as candidates for the hemochromatosis gene

We have looked for genes for ferritin and its translational control protein that could account for anomalies in the expression of ferritin (FT) and the transferrin receptor in the duodenum of individuals with hemochromatosis (HC). We show that there are probably only two FTH-like sequences near the HC locus on the short arm of chromosome 6 and no FTL-like sequences. We report the cloning of the previously uncharacterized FTH sequence from 6p (FTHL15) and show that it is probably a processed pseudogene. This gene has been mapped with a panel of radiation hybrid cells to near 6p12. Additionally, we show that there are no sequences on chromosome 6p for a protein that coordinately regulates expression of ferritin and the transferrin receptor.

The bifunctional iron-responsive element binding protein/cytosolic aconitase: the role of active-site residues in ligand binding and regulation

The iron-responsive element binding protein/cytosolic aconitase functions as either an RNA binding protein that regulates the uptake, sequestration, and utilization of iron or an enzyme that interconverts citrate and isocitrate. These mutually exclusive functions are regulated by changes in cellular iron levels. By site-directed mutagenesis we show that (i) ligation of a [4Fe-4S] cluster is necessary to inactivate RNA binding and activate enzyme function in vivo, (ii) three of four arginine residues of the aconitase active site participate in RNA binding, and (iii) aconitase activity is not required for iron-mediated regulation of RNA binding.

Localization of an RNA binding element of the iron responsive element binding protein within a proteolytic fragment containing iron coordination ligands

The iron responsive element binding protein (IRE-BP) regulates iron storage and uptake in response to iron. This control results from the interaction of the IRE-BP with the iron responsive element (IRE), a conserved sequence/structure element located near the 5' end of all ferritin mRNAs and in the 3' UTR of transferrin receptor mRNAs. Proteolysis was used to probe for functional elements of the IRE-BP. Partial chymotrypsin digestion generates a simple digestion pattern yielding fragments of 68, 56, 41, and 30 kDa. The 68 and 30 kDa fragments are derived from a single cleavage at Trp623. Further cleavages of the 68 kDa polypeptide yield the 56 and 41 kDa peptides. A combination of UV-crosslinking and chymotrypsin digestion was used to localize an RNA binding element within the C-terminus of the 68 kDa fragment, between amino acid residues 480 and 623. This region includes cysteine residues 503 and 506 which have been shown to be required for iron-sulfur cluster assembly and for iron regulation of the IRE-BP. Proteolytic fragments of the IRE-BP that contain this RNA binding region can be crosslinked to the IRE but do not bind with high affinity, suggesting that elements within the IRE-BP, in addition to those located between residues 480 and 623, are required for high affinity binding to the IRE.

A dioxygenase gene (Ids2) expressed under iron deficiency conditions in the roots of Hordeum vulgare

A lambda zapII cDNA library was constructed from mRNA isolated from Fe-deficient barley roots and screened with cDNA probes made from mRNA of Fe-deficient and Fe-sufficient (control) barley roots. Seven clones were selected. Among them a clone having the putative full-length mRNA of dioxygenase as judged by northern hybridization was selected and named Ids2 (iron deficiency-specific clone 2). Using a cDNA fragment as probe, two clones from the genomic library (lambda EMBL-III) were isolated and one was sequenced. The predicted amino acid sequence of Ids2 resembled that of 2-oxoglutarate-dependent dioxygenase. Ids2 is expressed in the Fe-deficient barley roots but is not in the leaves. The expression is repressed by the availability of Fe. Ids2 was also strongly expressed under Mn deficiency and weakly under Zn deficiency or excess NaCl (0.5%). The upstream 5'-flanking region of Ids2 has a root-specific cis element of the CaMV 35S promoter and a nodule-specific element of leghemoglobin, a metal regulatory element (MRE) and several Cu regulatory elements (UAS) of yeast metallothionein (CUP1).

Assembly of mitochondrial ribonucleoprotein complexes involves specific guide RNA (gRNA)-binding proteins and gRNA domains but does not require preedited mRNA

RNA editing in kinetoplastids probably employs a macromolecular complex, the editosome, that is likely to include the guide RNAs (gRNAs) which specify the edited sequence. Specific ribonucleoprotein (RNP) complexes which form in vitro with gRNAs (H. U. Göringer, D. J. Koslowsky, T. H. Morales, and K. D. Stuart, Proc. Natl. Acad. Sci. USA, in press) are potential editosomes or their precursors. We find that several factors are important for in vitro formation of these RNP complexes and identify specific gRNA-binding proteins present in the complexes. Preedited mRNA promotes the in vitro formation of the four major gRNA-containing RNP complexes under some conditions but is required for the formation of only a subcomponent of one complex. The 5' gRNA sequence encompassing the RYAYA and anchor regions and the 3' gRNA oligo(U) tail are both important in complex formation, since their deletion results in a dramatic decrease of some complexes and the absence of others. UV cross-linking experiments identify several proteins which are in contact with gRNA and preedited mRNA in mitochondrial extracts. Proteins of 25 and 90 kDa are highly specific for gRNAs, and the 90-kDa protein binds specifically to gRNA oligo(U) tails. The gRNA-binding proteins exhibit a differential distribution between the four in vitro-formed complexes. These experiments reveal several proteins potentially involved in RNA editing and indicate that multiple recognition elements in gRNAs are used for complex formation.

Assembly of mitochondrial ribonucleoprotein complexes involves specific guide RNA (gRNA)-binding proteins and gRNA domains but does not require preedited mRNA

RNA editing in kinetoplastids probably employs a macromolecular complex, the editosome, that is likely to include the guide RNAs (gRNAs) which specify the edited sequence. Specific ribonucleoprotein (RNP) complexes which form in vitro with gRNAs (H. U. Göringer, D. J. Koslowsky, T. H. Morales, and K. D. Stuart, Proc. Natl. Acad. Sci. USA, in press) are potential editosomes or their precursors. We find that several factors are important for in vitro formation of these RNP complexes and identify specific gRNA-binding proteins present in the complexes. Preedited mRNA promotes the in vitro formation of the four major gRNA-containing RNP complexes under some conditions but is required for the formation of only a subcomponent of one complex. The 5' gRNA sequence encompassing the RYAYA and anchor regions and the 3' gRNA oligo(U) tail are both important in complex formation, since their deletion results in a dramatic decrease of some complexes and the absence of others. UV cross-linking experiments identify several proteins which are in contact with gRNA and preedited mRNA in mitochondrial extracts. Proteins of 25 and 90 kDa are highly specific for gRNAs, and the 90-kDa protein binds specifically to gRNA oligo(U) tails. The gRNA-binding proteins exhibit a differential distribution between the four in vitro-formed complexes. These experiments reveal several proteins potentially involved in RNA editing and indicate that multiple recognition elements in gRNAs are used for complex formation.

The iron-responsive element-binding protein: localization of the RNA-binding site to the aconitase active-site cleft

The iron-responsive element-binding protein (IRE-BP) binds to specific stem-loop RNA structures known as iron-responsive elements (IREs) present in a variety of cellular mRNAs (e.g., those encoding ferritin, erythroid 5-aminolevulinate synthase, and transferrin receptor). Expression of these genes is regulated by interaction with the IRE-BP. The IRE-BP is identical in sequence to cytosolic aconitase, and the function of the protein is determined by the presence or absence of an Fe-S cluster. The protein either functions as an active aconitase when the Fe-S cluster is present or as an RNA-binding protein when the protein lacks this cluster. Aconitase activity and IRE-binding activity are mutually exclusive, and interconversion between the two activities is determined by intracellular Fe concentrations. Mapping of the RNA-binding site of the IRE-BP by UV cross-linking studies defines a major contact site between IRE and protein in the active-site region. Modeling based on probable structural similarities between the previously crystallized mitochondrial aconitase and the IRE-BP predicts that these residues would be accessible to the IRE only were there a major change in the predicted conformation of the protein when cells are iron-depleted.

Modulation of iron metabolism in monocyte cell line U937 by inflammatory cytokines: changes in transferrin uptake, iron handling and ferritin mRNA

We have investigated the effects of the pro-inflammatory cytokines interleukin 1 beta (IL-1 beta), tumour necrosis factor alpha (TNF alpha) and interferon gamma (IFN gamma) on the iron metabolism of the human monocytic cell line U937. Cells were treated with each cytokine for up to 24 h, and then iron uptake from diferric transferrin was determined. The intracellular distribution of this iron, the expression of the transferrin receptor and levels of mRNA for the two ferritin subunits were also studied. IL-1 beta, TNF alpha and IFN gamma all decreased transferrin-iron uptake into cells, and all three cytokines had effects on the proportion of iron associated with ferritin. With TNF alpha there was a marked enhancement of the fraction incorporated into ferritin. Transferrin-receptor expression was diminished by TNF alpha and IL-1 beta, but not IFN gamma, suggesting different effector mechanisms. Both TNF alpha and IFN gamma increased the amount of cellular mRNA for ferritin H-chain, but not the L-chain; IL-1 beta affected mRNA for neither ferritin. These data demonstrate that cytokines, which can be present at high concentrations in inflammation, have the capacity to affect macrophage iron uptake, transferrin receptor expression, intracellular iron handling and the relative abundance of ferritin-subunit mRNA, and may therefore be important mediators in the observed perturbations of iron metabolism in inflammatory diseases.

A fibroblast protein binds the 3'-untranslated region of pro-alpha 1(I) collagen mRNA

Post-transcriptional regulation of the expression of the pro alpha 1(I) chain of type I collagen (COL1A1) was studied by analysing cytoplasmic RNA-binding proteins and by transient transfections with collagen minigene plasmids. In this paper we present evidence for a factor from NIH 3T3 cells and human skin fibroblasts that interacts with the conserved 3'-untranslated region (UTR) of the shorter 4.8 kb mRNA species of the COL1A1 gene. The specificity of the interaction was confirmed by using (i) unlabelled specific and non-specific competitor RNAs and (ii) oligodeoxyribonucleotides annealed to the probe or used as single-stranded competitors. An antisense oligonucleotide annealed to the RNA probe near its 3'-terminus [20-42 nucleotides upstream of the first polyadenylation signal of the alpha 1(I) collagen mRNA] inhibited the binding, whereas other sense or antisense oligonucleotides had no effect on the interaction. The binding was sensitive to alkylation of free SH groups but not to phosphatase treatment of the extracts. In u.v. cross-linking analysis this factor migrated as a single polypeptide chain of about 67 kDa, and was named alpha 1-RBF67 (type I collagen alpha 1 chain RNA-binding factor). Dexamethasone treatment of fibroblasts, which is known to accelerate the turnover of COL1A1 mRNA, decreased the alpha 1-RBF67 activity markedly as evaluated by gel-retardation and u.v. cross-linking assays. Transient transfections with plasmids carrying the alpha 1(I) collagen promoter and 3'-UTR sequences demonstrated that the 3'-UTR participates in the response to dexamethasone. Thus the loss of alpha 1-RBF67 activity might be associated with decreased alpha 1(I) collagen mRNA levels after dexamethasone treatment.