Calcein as a fluorescent probe for ferric iron. Application to iron nutrition in plant cells.

The recent use of calcein (CA) as a fluorescent probe for cellular iron has been shown to reflect the nutritional status of iron in mammalian cells (Breuer, W., Epsztejn, S., and Cabantchik, Z. I. (1995) J. Biol. Chem. 270, 24209-24215). CA was claimed to be a chemosensor for iron(II), to measure the labile iron pool and the concentration of cellular free iron(II). We first study here the thermodynamic and kinetic properties of iron binding by CA. Chelation of a first iron(III) involves one aminodiacetic arm and a phenol. The overall stability constant log beta111 of FeIIICAH is 33. 9. The free metal ion concentration is pFeIII = 20.3. A (FeIII)2 CA complex can be formed. A reversible iron(III) exchange from FeIIICAH to citrate and nitrilotriacetic acid is evidenced when these ligands are present in large excess. The kinetics of iron(III) exchange by CA is compatible with metabolic studies. The low reduction potential of FeIIICAH shows that the ferric form is highly stabilized. CA fluorescence is quenched by 85% after FeIII chelation but by only 20% using FeII. Real time iron nutrition by Arabidopsis thaliana cells has been measured by fluorimetry, and the iron buffer FeIIICAH + CA was used as source of iron. As a siderophore, FeIIICAH promotes cell growth and regreening of iron-deficient cells more rapidly than FeIIIEDTA. We conclude that CA is a good chemosensor for iron(III) in cells and biological fluids, but not for Fe(II). We discuss the interest of quantifying iron buffers in biochemical studies of iron, in vitro as well as in cells.

Reactivity studies of the tyrosyl radical in ribonucleotide reductase from Mycobacterium tuberculosis and Arabidopsis thaliana--comparison with Escherichia coli and mouse.

Ribonucleotide reductase (RNR) is a key enzyme for DNA synthesis since it provides cells with deoxyribonucleotides, the DNA precursors. Class I alpha2beta2 RNRs contain a dinuclear iron center and an essential tyrosyl radical in the beta2 component (protein R2). This is also true for the purified protein R2 of Mycobacterium tuberculosis RNR, as shown by iron analysis, light absorption and EPR spectroscopy. EPR spectroscopy at 286 GHz revealed a high g(x) value, suggesting that the radical is not hydrogen bonded, as in other prokaryotic R2s and in contrast with eukaryotic R2s (from Arabidopsis thaliana and mouse). Furthermore, it proved to be very resistant to scavenging by a variety of phenols and thiols and by hydroxyurea, similar to the Escherichia coli radical. By comparison, the plant and mouse radicals are very sensitive to drugs such as resveratrol and 2-thiophenthiol. The radical from M. tuberculosis RNR does not seem to be an appropriate target for new antituberculous agents.

Dynamic equilibria in iron uptake and release by ferritin.

The function of ferritins is to store and release ferrous iron. During oxidative iron uptake, ferritin tends to lower Fe2+ concentration, thus competing with Fenton reactions and limiting hydroxy radical generation. When ferritin functions as a releasing iron agent, the oxidative damage is stimulated. The antioxidant versus pro-oxidant functions of ferritin are studied here in the presence of Fe2+, oxygen and reducing agents. The Fe(2+)-dependent radical damage is measured using supercoiled DNA as a target molecule. The relaxation of supercoiled DNA is quantitatively correlated to the concentration of exogenous Fe2+, providing an indirect assay for free Fe2+. After addition of ferrous iron to ferritin, Fe2+ is actively taken up and asymptotically reaches a stable concentration of 1-5 microM. Comparable equilibrium concentrations are found with plant or horse spleen ferritins, or their apoferritins. After addition of ascorbate, iron release is observed using ferrozine as an iron scavenger. Rates of iron release are dependent on ascorbate concentration. They are about 10 times larger with pea ferritin than with horse ferritin. In the absence of ferrozine, the reaction of ascorbate with ferritins produces a wave of radical damage; its amplitude increases with increased ascorbate concentrations with plant ferritin; the damage is weaker with horse ferritin and less dependent on ascorbate concentrations.

Metabolization of iron by plant cells using O-Trensox, a high-affinity abiotic iron-chelating agent.

A synthetic siderophore, O-Trensox (L), has been designed and synthesized to improve iron nutrition of plants. The affinity for iron of this ligand [pFe(III) = 29.5 and pFe(II) = 17.9] is very high compared with EDTA. In spite of its high and specific affinity for iron, O-Trensox was found to be able to prevent, and to reverse, iron chlorosis in several plant species grown in axenic conditions. It also allows the iron nutrition and growth of Acer pseudoplatanus L. cell suspensions. The rate of iron metabolization was monitored by 59Fe radioiron. Ferritins, the iron storage proteins, are shown to be the first iron-labelled proteins during iron metabolization and to be able to further dispatch the metal. Using Fe(III)-Trensox, the rate of iron incorporation into ferritin was found to be higher than when using Fe-EDTA, but slower than with Fe-citrate, the natural iron carrier in xylem. During a plant cell culture, the extracellular concentrations of iron complex and free ligand were measured; changes in their relative amounts showed that the iron complex is dissociated extracellularly and that only iron is internalized. This suggests a high affinity for iron of a putative carrier on the plasmalemma. In contrast with Fe-citrate and Fe-EDTA complexes, Fe(III)-Trensox is not photoreducible. Its ability to induce radical damage as a Fenton reagent was tested using supercoiled DNA as target molecule. Unlike Fe-citrate and Fe-EDTA, Fe(II)-Trensox and Fe(III)-Trensox were proven to be harmless even during ascorbate-driven reduction, while Fe-EDTA and Fe-citrate generate heavy damage to DNA.

Iron Deficiency Induced by Chrysobactin in Saintpaulia Leaves Inoculated with Erwinia chrysanthemi.

In this communication, we examine the fate of iron during soft rot pathogenesis caused by Erwinia chrysanthemi on its host, Saintpaulia ionantha. The spread of soft rot caused by this enterobacterium was previously shown to depend on a functional genetic locus encoding a high-affinity iron assimilation system involving the catechol-type siderophore chrysobactin. Leaf intercellular fluid from healthy plants was analyzed with regard to the iron content and its availability for bacterial growth. It was compared to the fluid from diseased plants for the presence of strong iron ligands, using a new approach based on the iron-binding property of an ion-exchange resin. Further characterization allowed the identification of chrysobactin in diseased tissues, thus providing the first evidence for the external release of a microbial siderophore during pathogenesis. Competition for nutritional iron was also studied through a plant-bacterial cell system: iron incorporated into plant ferritin appeared to be considerably reduced in bacteria-treated suspension soybean cells. The same effect was visualized during treatment of soybean cells with axenic leaf intercellular fluid from E. chrysanthemi-inoculated saintpaulia leaves or with chrysobactin.

Iron release and uptake by plant ferritin: effects of pH, reduction and chelation.

Ferritins are iron-storage proteins that accumulate in plastids during seed formation, and also in leaves during senescence or iron overload. Iron release from ferritins occurs during growth of seedlings and greening of plastids. Depending on the concentration of the reducing agent ascorbate, either an overall iron release or uptake by ferritins from iron(III) citrate may occur. We have designed methods to measure these simultaneous and independent uptake and release fluxes. Each individual step of the exchange was studied using different iron chelates and an excess of ligand. It is shown that: (i) the chelated form of iron, and not ionic Fe3+, is the substrate for iron reduction, which controls the subsequent uptake by ferritin; (ii) iron uptake by ferritins is faster at pH 8.4 than at pH 7 or 6 and is inhibited by an excess of strongly binding free ligands; and (iii) strongly binding free ligands are inhibitory during iron release by ascorbate. When reactions are allowed to proceed simultaneously, the iron chelating power is shown to be a key factor in the overall exchange. The interactions of iron chelating power, reducing capacity and pH are discussed with regard to their influence on the biochemical mobilization of iron.

Structure and composition of ferritin cores from pea seed (Pisum sativum).

Iron cores from native pea seed (Pisum sativum) ferritin have been analysed by electron microscopy and Mössbauer spectroscopy and shown to be amorphous. This correlates with their relatively high phosphate content (Fe: P = 2.83; 1800 Fe, 640 P atoms/molecule). Reconstituted cores obtained by adding iron (2000 Fe atoms/molecule) in the absence of phosphate to pea seed apoferritin were crystalline ferrihydrite. In vitro rates of formation of pea-seed ferritin iron cores were intermediate between those of recombinant human H-chain and horse spleen apoferritin and this may reflect the amino-acid residues of its ferroxidase and putative nucleation centres. The high phosphate content of pea-seed ferritin suggests that this molecule could be involved in both phosphorus and iron storage. The high phosphate concentration found within plastids, from which the molecules were isolated, is a possible source of the ferritin phosphate.

Structure, function, and evolution of ferritins.

The ferritins of animals and plants and the bacterioferritins (BFRs) have a common iron-storage function in spite of differences in cytological location and biosynthetic regulation. The plant ferritins and BFRs are more similar to the H chains of mammals than to mammalian L chains, with respect to primary structure and conservation of ferroxidase center residues. Hence they probably arose from a common H-type ancestor. The recent discovery in E. coli of a second type of iron-storage protein (FTN) resembling ferritin H chains raises the question of what the relative roles of these two proteins are in this organism. Mammalian L ferritins lack ferroxidase centers and form a distinct group. Comparison of the three-dimensional structures of mammalian and invertebrate ferritins, as well as computer modeling of plant ferritins and of BFR, indicate a well conserved molecular framework. The characterisation of numerous ferritin homopolymer variants has allowed the identification of some of the residues involved in iron uptake and an investigation of some of the functional differences between mammalian H and L chains.

Purification, characterization and function of bacterioferritin from the cyanobacterium Synechocystis P.C.C. 6803.

Storage and buffering of iron is achieved by a class of proteins, the ferritins, widely distributed throughout the living kingdoms. All ferritins have in common their three-dimensional structure and their ability to store large amounts of iron in their central cavity. However, eukaryotic ferritins from plants and animals and bacterioferritins have no sequence similarity, and besides non-haem iron bacterioferritins contain haem residues whereas eukaryotic ferritins do not. In this paper we report the first purification and characterization of a bacterioferritin from a cyanobacterium. It has a molecular mass of 400 kDa and is built up from 19 kDa subunits. Its N-terminal sequence shows 73% identity with that of the Escherichia coli bacterioferritin subunit. It contains 2300 atoms of iron and 1500 molecules of phosphate per ferritin molecule and 0.25 haem residue per subunit; the alpha-peak of the cytochrome has its maximum at 559 nm. In contrast with what is known for eukaryotic ferritins, we found that bacterioferritin from Synechocystis is not inducible by iron under the conditions that we have tested and that it has a constant concentration whatever the iron status of the cells, even at very low iron concentration. Bacterioferritin from Synechocystis P.C.C. 6803 is fully assembled in vivo and it is shown by labelling with 59Fe that it is able to load iron in vitro as well as in vivo. Bacterioferritin from Synechocystis is shown to have an iron-buffering function while the bulk of cellular iron is found associated with a pool of low-molecular-mass electronegative molecules. The role of Synechocystis bacterioferritin in iron metabolism is discussed.

Evidence for conservation of ferritin sequences among plants and animals and for a transit peptide in soybean.

Ferritin is a large multisubunit protein that stores iron in plants, animals, and bacteria. In animals, the protein is mainly cytoplasmic and is highly conserved, while in plants ferritin is found in chloroplasts and other plastids. Ferritin is synthesized in plants as a larger precursor of the mature subunit. There is no sequence information for ferritin from plants, except an NH2-terminal peptide of 35 residues which shows little similarity to any known ferritin sequences or transit peptides (Laulhere, J. P., Laboure, A. M., and Briat, J. F. (1989) J. Biol. Chem. 264, 3629-3635). To understand the genetic origin and the location of ferritin synthesis in plant cells, as well as the structure of ferritin from plants, we have sequenced both CNBr peptides from pea seed ferritin and nucleotides of a soybean hypocotyl ferritin cDNA, identified using a frog ferritin cDNA as a probe. Comparison of pea and soybean sequences showed an identity of 89%. Alignment of the plant ferritin sequences with animal ferritins showed 55-65% sequence identity in the common regions. However, a peptide of 28 amino acids extended the NH2 terminus of the plant ferritins. Furthermore, the cDNA encoded additional amino acids which appear to be a transit peptide. None of the sequences in soybean ferritin were found in the tobacco chloroplast genome, suggesting, as does the transit peptide, a nuclear location of ferritin gene(s) in plants. Plant ferritin mRNA is 400-500 nucleotides longer than animal ferritin mRNAs, a difference accounted for in part by the extra peptides encoded. The size of soybean ferritin mRNA was constant in different tissues but expression varied in different tissues (leaf greater than hypocotyl). Thus, higher plants and animal ferritins display sequence homology and differential tissue expression. An ancient, common progenitor apparently gave rise to contemporary eukaryotic ferritins after specific modifications, e.g. transport to plasmids.

Photoreduction and incorporation of iron into ferritins.

Pea seed ferritin is able to incorporate ferrous iron into the mineral core. Fe2+ may be formed by reduction of exogenous Fe3+ with ascorbate or by photoreduction by ferritin and by ferric citrate. In our experimental conditions the bulk of the photoreduction is carried out by ferritin, which is able to photoreduce its endogenous iron. Citrate does not enhance the photoreduction capacity of ferritin, and exogenous ferric citrate improves the yield of the reaction by about 30%. The mineral core of the ferritin is shown to photoreduce actively, and the protein shell does not participate directly in the photoreduction. Low light intensities and low concentration of reducing agents do not allow a release of iron from ferritins, but induce a 'redox mill' of photoreduction and simultaneous ferroxidase-mediated incorporation. High ascorbate concentrations induce the release of ferritin iron. These reactions are accompanied by the correlated occurrence of damage caused by radicals arising from Fenton reactions, leading to specific cleavages in the 28 kDa phytoferritin subunit. This damage caused by radicals occurs during the oxidative incorporation into the mineral core and is prevented by o-phenanthroline or by keeping the samples in the dark.

A jack bean protein similar to pea seed ferritin and unrelated to concanavalin A is the only iron storage protein in jack bean seeds, although concanavalin A can form ferritin-like iron cores in vitro.

Iron storage in living organisms is performed by ferritins. These proteins are built up from 24 subunits organized in a spherical shell forming a coat to a core of bioavailable iron. Recently, it was observed that concanavalin A from jack bean can form in vitro iron cores similar to those of animal ferritins (Yariv, J., Kalb, A. J., Helliwell, J. R., Papiz, M. Z., Bauminger, E. R., and Nowik, I. (1988) J. Biol. Chem. 263, 13508-13510). From this observation and from the comparison of the three-dimensional structures of horse spleen ferritin and of the form of concanavalin A which forms soluble polynuclear iron complexes, these authors suggested that concanavalin A is the apoferritin of jack bean seeds. On the basis of immunological and biochemical results, we report here that a protein purified from jack bean seeds, unrelated to concanavalin A and similar to plant seed ferritins, is responsible for iron storage in jack bean seeds. Furthermore, concanavalin A does not contain iron in vivo. Therefore we conclude that a unique protein, ferritin, stores iron in vivo in jack bean seeds and that concanavalin A provides an unusual model for studying the formation of iron cores inside a protein shell.

Mechanism of the transition from plant ferritin to phytosiderin.

Soluble and insoluble forms of ferritins have been purified from dry pea seeds by gel filtration. The insoluble form is called phytosiderin by analogy with animal hemosiderin. Native gel electrophoresis of these two forms have shown that the soluble one (ferritin) is homogenous in size and more compact than the insoluble one (phytosiderin) which is heterogenous in size. However, when iron is removed from these two classes of molecules (apoferritin), they have the same mobility in isopycnic centrifugations. Polyacrylamide-sodium dodecyl sulfate gel electrophoresis revealed a difference in their subunit composition: ferritin molecules are built up from a 28-kDa subunit and phytosiderin from a 26.5-kDa subunit. Partial proteolysis using a Staphylococcus aureus protease indicates a strong relationship between these two polypeptides. Intermediates between these two forms have also been characterized and are composed of both subunits in various amounts. Ferritin and phytosiderin are both able to incorporate iron in vitro into their mineral core. It is also shown that in vitro iron exchange induces ferritin degradation. This degradation is prevented by inhibitors of the Fenton cycle (iron chelates like o-phenanthroline and desferrioxamine B) and reduced by Tris, a radical scavenger. Under in vitro conditions of controlled radical damage the 28-kDa subunit is converted into the 26.5-kDa subunit. Purification of the 28-kDa subunit has allowed us to determine the NH2-terminal sequence. The NH2 extremity of the 26.5-kDa subunit is heterogenous, but the sequence of its main component is identical to the sequence of the 28-kDa subunit downstream residue Leu-21. These data indicate that the 26.5-kDa subunit is produced by radical mediated damage leading to a series of cleavages in the NH2 terminal part of the 28-kDa subunit.

Purification and characterization of ferritins from maize, pea, and soya bean seeds. Distribution in various pea organs.

Ferritins from maize, pea, and soya bean seeds were purified. They contain two polypeptides of 28 and 26.5 kDa. The molecular weight of native pea seed ferritin has been estimated to be 540,000. Pea and maize seed ferritins were compared by reverse phase high performance liquid chromatography, amino acid composition, and two-dimensional gel electrophoresis. They are very similar, although four isoforms of the 28-kDa polypeptide from the pea were observed in contrast to a unique polypeptide in maize. No isoforms of the 26.5-kDa polypeptide were detected. Rabbit antibodies were produced in response to pea seed ferritin. It was shown by Western blot analysis that ferritins of the three plants analyzed share immunological determinants. However, horse spleen ferritin was not recognized by the phytoferritin antibodies. Antibodies were also used to demonstrate that ferritins are not uniformly distributed in different pea organs from 30-day-old iron-unloaded plants. The protein was more abundant in flowers than in fruits and roots, and was not detected in leaves.

Characterization and properties of the spinach chloroplast transcriptionally active chromosome isolated at high ionic strength.

The transcriptionally active chromosome (TAC) of spinach (Spinacia oleracea L.) chloroplasts has been isolated at a high ionic strength, with low mechanical shearing, by glycerol gradient centrifugation. The properties of the TAC differ from those previously reported for the TAC isolated either from Euglena chloroplasts or from spinach using a low-ionic-strength solubilization medium and gel filtration. The high-salt-isolated TAC is homogenous in density but not in size and contains fewer weakly bound proteins than its lowsalt-isolated homologue. In vitro, it promotes elongation of the RNA chains previously initiated in vivo. Transcription is not limited to the ribosomal DNA. The transcriptional pattern is not strongly affected by the high-salt preparation. Ribonuclease pretreatment of the TAC, prior to the in-vitro transcription, leads to a more than tenfold increase of the transcription activity. These properties are discussed in relation to the structure of the spinach chloroplast chromosome.

Visualization of a Spinach Plastid Transcriptionally Active DNA-Protein Complex in a Highly Condensed Structure.

A transcriptionally active DNA-protein complex isolated from spinach Spinacia oleracea plastids is visualized by electron microscopy in different conditions. This structure, after glutaraldehyde fixation, is highly condensed. DNA is supertwisted with proteins bound to it producing a beaded substructure. When glutaraldehyde fixation is omitted this structure is less condensed and DNA fibrils come out from a proteinous central body. The DNA-protein complex can be separated into two populations by CsCl centrifugation: one with a buoyant density of 1.570 grams per cubic centimeter and the other of 1.610 grams per cubic centimeter. By visualization of these two populations, it is concluded that proteins are either firmly bound to DNA in the central body, or more loosely bound to the DNA fibrils. These latter proteins could play a role in enzymic functions and/or in the supercoiling of DNA.The DNA from the DNA-protein complex possesses all fragments that belong to pure circular chloroplast DNA hydrolyzed by two restriction enzymes: Bam HI and Eco RI. Some molecules observed in a supercondensed form with a beaded substructure probably contain entire chloroplast DNA molecules.A hydrolysis test with microccocal nuclease gives no indication of the presence of ;nucleosome-like' structures. Thirty-six polypeptides with molecular weights ranging from 12,000 to 180,000 are present in the complex, and seven of them are highly soluble in 0.4 n H(2)SO(4); their molecular weights range from 14,000 to 46,000 as shown by two-dimensional gel electrophoresis.No linolenic acid can be detected in the preparation, indicating the absence of chloroplast membranes.

Influence of the ionic environment on the in vitro transcription of the spinach plastid DNA by a selectively bound RNA-polymerase DNA complex.

The in vitro transcription of chloroplast DNA (ctDNA) is studied using a DNA-protein complex isolated from spinach plastids. The RNA products are compared to the in vivo synthesized ctRNA by competition for hybridization. At least 80% of the in vitro RNA sequences are present in vivo. Modifications of ionic strength or introduction of heparin in the reaction medium has an important effect on transcriptional activity of the complex. Furthermore, the length of the RNA chains increases ionic strength and amount of heparin. The RNA products are analysed by Southern hybridizations to EcoRI cTDNA fragments. Changes in the ionic strength or in the amount of heparin modify heterogeneously the transcription of the various DNA regions. The quantitative distribution of transcripts among the ctDNA fragments is used as evidence for the selectivity of the transcription. The activity of the DNA-protein complex is much more resistant to high ionic strength than an heterologous transcription system using Escherichia coli RNA polymerase and ctDNA. This latter system transcribes less ctDNA fragments.

Transcription activity of a DNA-protein complex isolated from spinach plastids.

A DNA . protein complex of about 150 S is isolated from purified spinach chloroplasts by Sepharose 4B gel filtration. A DNA-dependent RNA polymerase activity is found associated with the complex. This DNA protein complex is able to initiate RNA chains in vitro. The RNA synthesis is more dependent on CTP than other nucleoside triphosphates. 50% of the activity is still present with 0.6 M KCl. The temperature optimum occurs between 30 degrees C and 35 degrees C. Rifampicin and rifamycin SV have no inhibitory effect. TNA products have been characterized by gel filtration and by hybridization with chloroplast DNA (ctDNA). At the beginning of transcription DNA products are linked to the transcription complex and are later detached. The molecular weight of the product ranges between 0.07 X 10(6) and 2 X 10(6). A part of the product (3--4%) has a molecular weight higher than 2 X 10(6). No endogenous RNase activity was present during the molecular weight determinations experiments. Hybridization experiments show that at least 75% of the RNA products are hybridizable with ctDNA and that 40% of these products are composed of chloroplast ribosomal RNA, showing that rDNA is preferentially transcribed.