Redox reactivity of animal apoferritins and apoheteropolymers assembled from recombinant heavy and light human chain ferritins

The redox reactivities of air-oxidized apo horse spleen ferritin (HoSF) and apo rat liver ferritin (RaF) were examined by microcoulometry and reductive optical titrations. Microcoulometry on several independent lots of commercial HoSF revealed two distinct types of redox activity: one requiring 3-4 electrons and one requiring 6-7 electrons for full reduction of the protein shell. ApoRaF required 8-9 electrons to fully reduce the oxidized form. Reductive optical titrations confirmed the microcoulometric reduction stoichiometry and, in addition, showed that the spectra of both oxidized and reduced apoHoSF were distinct and possessed absorbances tailing into the visible region. The redox reactivity of both apoRaF and apoHoSF correlated with their H-subunit composition. Identical microcoulometric and optical experiments were conducted with recombinant apo human liver heavy (rHuHF) and light (rHuLF) ferritins, but neither was redox-active. These results suggest that the redox reactivity of native ferritins is due to their heteropolymeric nature. This was confirmed by mixing various proportions of rHuHF and rHuLF, dissociating the 24-mers into individual subunits with guanidine hydrochloride at pH 3.5, and renaturing to form heteropolymeric 24-mers. Microcoulometric measurements of these apoheteropolymers reassembled in vitro showed that they were redox-active like their native apoheteropolymer counterparts. The redox activity of these apoheteropolymers increased with H-subunit composition, reached a maximum near 12 H- and 12 L-subunits, and then declined to zero with increasing L-subunit composition. The decline in redox reactivity at high L-subunit concentrations indicates that both H- and L-subunits are involved in forming the observed redox centers. Apoheteropolymers formed from rHuLF and W93F (an H-chain mutant) were redox-inactive, suggesting that the conserved tryptophan is necessary for redox center formation.

An ELISA for the H-subunit of human ferritin which employs a combination of rabbit poly- and mice monoclonal antibodies and an enzyme labeled anti-mouse-IgG

We describe a sensitive ELISA for measuring the H-type subunit of human ferritin. A high detection sensitivity was attained by the use of antibodies from different species and an enzyme-conjugated secondary antibody. It consisted of a sandwich assay using a solid phase coated with a rabbit polyclonal antibody for human ferritin from term placenta and a soluble monoclonal antibody for human H-ferritin, followed by a secondary anti-mouse immunoglobulin (Ig)G conjugated to beta-galactosidase. The assay was calibrated with purified recombinant human H-ferritin from E. coli. The colorigenic chlorophenol red beta-D-galactopyranoside and the fluorogenic 4-methyl-umbelliferyl-beta-D-galactopyranoside substrates were used with similar outcome. The described method permits the measurement of human H-ferritin at a concentration ranging from 0.1 to 100 micrograms/l (or 20-20,000 pg per 200 microliters sample) and is accurate at a concentration as low as 0.3 microgram/l. The coefficient of variation of the assay was 6.05-10.3% and the recovery of H-ferritin added to cell lysates was 105.8 +/- 7.52%. Depending on the H-ferritin content of the cell line tested, only 600 to 60,000 cells of different human cell lines were needed to measure their H-ferritin content.

Characterization and distribution of ferritin binding sites in the adult mouse brain

Studies on iron uptake into the brain have traditionally focused on transport by transferrin. However, transferrin receptors are not found in all brain regions and are especially low in white matter tracts where high iron concentrations have been reported. Several lines of research suggest that a receptor for ferritin, the intracellular storage protein for iron, may exist. We present, herein, evidence for ferritin binding sites in the brains of adult mice. Autoradiographic studies using 125I-recombinant human ferritin demonstrate that ferritin binding sites in brain are predominantly in white matter. Saturation binding analyses revealed a single class of binding sites with a dissociation constant (K(D)) of 4.65 x 10(-9) M and a binding site density (Bmax of 17.9 fmol bound/microg of protein. Binding of radiolabeled ferritin can be competitively displaced by an excess of ferritin but not transferrin. Ferritin has previously been shown to affect cellular proliferation, protect cells from oxidative damage, and deliver iron. The significance of a cellular ferritin receptor is that ferritin is capable of delivering 2,000 times more iron per mole of protein than transferrin. The distribution of ferritin binding sites in brain vis-à-vis transferrin receptor distribution suggests distinct methods for iron delivery between gray and white matter.

Antiferritin single-chain antibody: a functional protein with incomplete folding?

The pET(scF11) plasmid was constructed comprising the gene of a single-chain antibody against human ferritin. This plasmid encodes the leader peptide pelB followed by the heavy chain variable V(H) domain, (Gly4Ser)3 linker peptide, and light chain variable V(L) domain. The correctly processed scF11 antibody was expressed in Escherichia coli as an insoluble protein without the leader peptide. Purified soluble scF11 was obtained after solubilization in 6 M GdnHCl followed by a sequential dialysis against decreasing urea concentrations and ion-exchange chromatography. ScF11 demonstrated only a approximately 8-fold decrease in the affinity (Ka = 5.1 x 10(8) M(-1) in RIA and 1.8 x 10(8) M(-1) in ELISA) vs. the parent IgG2a/kappa monoclonal antibody F11. The emission maximum of intrinsic fluorescence strongly suggests a compact conformation with tryptophanyl fluorophores buried in the protein interior, consistent with the functionality of the protein. However, scF11 demonstrated (i) the lack of denaturant-induced fluorescence 'dequenching' effect characteristic of the completely folded parent antibody, and (ii) prominent binding, under physiological conditions, of a hydrophobic probe 8-anilino-1-naphthalenesulfonate (ANS) recognizing partially structured states of a protein. These findings are indicative of an incomplete tertiary fold that gives ANS access to the protein hydrophobic core. This work provides the first indication that the functional single-chain antibody scF11 displays some properties of a partially structured state and therefore may possess incomplete folding.

Production of a soluble and functional recombinant streptavidin in Escherichia coli

The cDNA for streptavidin (residues 15-159) was subcloned into an expression vector in fusion at the N-terminus with the T7-tag (12 residues). Conditions were found to express the protein in Escherichia coli in a soluble, assembled, and active form. The protein was purified in two simple steps which involved heating at 75 degreesC and affinity chromatography on iminobiotin agarose. The purified protein was obtained in yields of 70 mg per liter of bacterial culture. Electron spray mass spectrometry analysis showed that the recombinant streptavidin had the expected molecular mass without covalent modifications. ELISA and surface plasmon resonance analyses showed it to be functionally analogous to the natural streptavidin. This appears to be an improvement over the reported methods of recombinant streptavidin production which involve protein renaturation or the use of eukaryotic expression systems.

Biochemical characterization and crystal structure of a recombinant hen avidin and its acidic mutant expressed in Escherichia coli

The mature hen avidin encoded by a synthetic cDNA was expressed in Escherichia coli in an insoluble form. After resolubilization, renaturation and purification, a recovery of about 20 mg/l cell culture was obtained. ELISA assays indicated no apparent differences in biotin binding between the natural and recombinant avidins. In addition, an acidic avidin mutant, bearing the substitutions Lys3-->Glu, Lys9--> Glu, Arg26-->Asp and Arg124-->Leu of four exposed basic residues, was produced. The protein, expressed and renatured as wild-type avidin, showed unaltered biotin-binding activity. The acidic pI (approximately 5.5) and lack of aggregation of the mutant allowed easy electrophoretic analysis under non-denaturing conditions of the protein alone and of its complexes with biotin, biotinylated transferrin or peroxidase. Analysis of the sera from sensitized subjects revealed that the avidin mutant has altered antigenicity. Both recombinant avidins were crystallized and the three-dimensional structures solved by molecular replacement and refined to 0.22 nm resolution. The three-dimensional structures of the two recombinant molecules, in the absence of biotin and of glycosylation, are fully comparable with those of the natural hen avidin previously reported.

Reaction paths of iron oxidation and hydrolysis in horse spleen and recombinant human ferritins

UV-visible spectroscopy, electrode oximetry, and pH stat were used to study Fe(II) oxidation and hydrolysis in horse spleen ferritin (HoSF) and recombinant human H-chain and L-chain ferritins (HuHF and HuLF). Appropriate test reactions and electrode responses were measured, establishing the reliability of oxygen electrode/pH stat for kinetics studies of iron uptake by ferritin. Stoichiometric ratios, Fe(II)/O2 and H+/Fe(II), and rates of oxygen uptake and proton production were simultaneously measured as a function of iron loading of the protein. The data show a clear distinction between the diiron ferroxidase site and mineral surface catalyzed oxidation of Fe(II). The oxidation/hydrolysis reaction attributed to the ferroxidase site has been determined for the first time and is given by 2Fe2+ + O2 + 3H2O --> [Fe2O(OH)2]2+ + H2O2 + 2H+ where [Fe2O(OH)2]2+ represents the hydrolyzed dinuclear iron(III) center postulated to be a mu-oxo-bridged species from UV spectrometric titration data and absorption band maxima. The transfer of iron from the ferroxidase site to the mineral core has been now established to be [Fe2O(OH)2]2+ + H2O --> 2FeOOH(core) + 2H+. Regeneration of protein ferroxidase activity with time is observed for both HoSF and HuHF, consistent with their having enzymatic properties, and is facilitated by higher pH (7.0) and temperature (37 degreesC) and by the presence of L-subunit and is complete within 10 min. In accord with previous studies, the mineral surface reaction is given by 4Fe2+ + O2 + 6H2O --> 4FeOOH(core) + 8H+. As the protein progressively acquires iron, oxidation/hydrolysis increasingly shifts from a ferroxidase site to a mineral surface based mechanism, decreasing the production of H2O2.

Analysis of ferritins in lymphoblastoid cell lines and in the lens of subjects with hereditary hyperferritinemia-cataract syndrome

Hereditary hyperferritinemia-cataract syndrome (HHCS) is an autosomal and dominant disease caused by heterogeneous mutations in the iron responsive element (IRE) of the 5' untranslated flanking region of ferritin L-chain mRNA, which reduce the binding to the trans iron regulatory proteins and make L-chain synthesis constitutively upregulated. In the several families identified so far, the serum and tissue L-ferritin levels are fivefold to 20-fold higher than in nonaffected control subjects, iron metabolism is apparently normal, and the only relevant clinical symptom is early onset, bilateral cataract. Some pathogenetic aspects of HHCS remain obscure, with particular reference to the isoferritins produced by HHCS cells, as well as the mechanism of cataract formation. We analyzed lymphoblastoid cell lines obtained from two nonaffected control subjects and from HHCS patients carrying the substitution A40G (Paris-1), G41C (Verona-1), and the deletion of the residues 10-38 (Verona-2) in the IRE structure. Enzyme-linked immunosorbent assays specific for the H- and L-type ferritins showed that L-ferritin levels were up to 20-fold higher in HHCS than in control cells and were not affected by iron supplementation or chelation. Sequential immunoprecipitation experiments of metabolically-labeled cells with specific antibodies indicated that in HHCS cells about half of the L-chain was assembled in L-chain homopolymers, which did not incorporate iron, and the other half was assembled in isoferritins with a high proportion of L-chain. In control cells, all ferritin was assembled in functional heteropolymers with equivalent proportion of H- and L-chains. Cellular and ferritin iron uptake was slightly higher in HHCS than control cells. In addition, we analyzed the lens recovered from cataract surgery of a HHCS patient. We found it to contain about 10-fold more L-ferritin than control lens. The ferritin was fully soluble with a low iron content. It was purified and partially characterized. Our data indicate that: (1) in HHCS cells a large proportion of L-ferritin accumulates as nonfunctional L-chain 24 homopolymers; (2) the concomitant fivefold to 10-fold expansion of ferritin heteropolymers, with a shift to L-chain-rich isoferritins, does not have major effects on cellular iron metabolism; (3) L-chain accumulation occurs also in the lens, where it may induce cataract formation by altering the delicate equilibrium between other water-soluble proteins (ie, crystallins) and/or the antioxidant properties.

Distinct stability of recombinant L and H subunits of human ferritin: calorimetric and ANS binding studies

Thermodynamic and pH stability of recombinant human L- and H-ferritins were probed by differential scanning calorimetry and 8-anilino-1-naphthalenesulfonate (ANS) binding in the pH range 2-7. At pH 2.0-2.8 they were dissociated into subunit monomers and in this pH interval the H-subunit displayed a single calorimetrically-revealed domain with properties of a molten globule-like state: low enthalpy (6.3-8.0 J/g or 169-172 kJ/mol) and Tm of thermal unfolding (approximately 50 degrees C), a wide transition range (approximately 20 degrees C) and high ANS binding. In contrast, at pH 2 the L-ferritin subunit showed two calorimetric domains with Tm of 35 and 40 degrees C with similar unfolding enthalpies and with moderate extent of interactions, as indicated by the ratio of calorimetric enthalpy (293.9 kJ/mol) and van't Hoff enthalpy (174.2 kJ/mol) for the thermal transition. A pH increase from 2.0 to 2.8 determined the coupling of the two domains into a single cooperative folding unit and drastic increase of the transition temperature (from 37 to 80 degrees C). The contacts between the two domains in the L-subunit appeared to contribute to about 30% of the total stabilization free energy. The unfolding enthalpies, heat capacity changes and pronounced ANS binding of the L-subunit at pH 2.0-2.8 indicated that part of the structure lacked 'meltable' tertiary interactions. The results indicate that H- and L-subunits are stabilized by largely different intra-chain interactions with a critical contribution to L-subunit stability of embedded salt bridge(s) absent in the H-subunit.

Transient overexpression of human H- and L-ferritin chains in COS cells

The understanding of the in vitro mechanisms of ferritin iron incorporation has greatly increased in recent years with the studies of recombinant and mutant ferritins. However, little is known about how this protein functions in vivo, mainly because of the lack of cellular models in which ferritin expression can be modulated independently from iron. To this aim, primate fibroblastoid COS-7 cells were transiently transfected with cDNAs for human ferritin H- and L-chains under simian virus 40 promoter and analysed within 66 h. Ferritin accumulation reached levels 300-500-fold higher than background, with about 40% of the cells being transfected. Thus ferritin concentration in individual cells was increased up to 1000-fold over controls with no evident signs of toxicity. The exogenous ferritin subunits were correctly assembled into homopolymers, but did not affect either the size or the subunit composition of the endogenous heteropolymeric fraction of ferritin, which remained essentially unchanged in the transfected and non-transfected cells. After 18 h of incubation with [59Fe]ferric-nitrilotriacetate, cellular iron incorporation was similar in the transfected and non-transfected cells and most of the protein-bound radioactivity was associated with ferritin heteropolymers, while H- and L-homopolymers remained iron-free. Cell co-transfection with cDNAs for H- and L-chains produced ferritin heteropolymers that also did not increase cellular iron incorporation. It is concluded that transient transfection of COS cells induces a high level of expression of ferritin subunits that do not co-assemble with the endogenous ferritins and have no evident activity in iron incorporation/metabolism.

Biochemical and immunological characterization of recombinant allergen Lol p 1

Pollen from perennial rye grass (Lolium perenne), a major cause of type-I allergy worldwide, contains a complex mixture of allergenic proteins among which Lol p 1 is one of the most important. We describe the expression, purification and characterization of a recombinant Lol p 1 overproduced in Escherichia coli. The recombinant allergen, expressed in high yields and purified in milligram amounts, bound to specific IgE antibodies from human sera, induced histamine release from sensitized human basophils, and elicited rabbit antisera that recognize specifically recombinant Lol p 1 and natural Lol p 1 of pollen extract. Recombinant Lol p 1 was used to develop ImmunoCAP assays for analysis of 150 sera that were Radioallergosorbent test positive to L. perenne pollen. In 130 of them (87%) the assay detected a significant level of IgE antibodies to Lol p 1, reaching on average 37% of the level obtained with a test for IgE to the whole grass pollen extract. To map epitopes on Lol p 1, we produced three deletion mutants [des-(116-240)-Lol p 1, des-(1-88)-Lol p 1 and des-(133-189)-Lol p 1], which were efficiently expressed in bacteria. These all showed a strong reactivity with the specific rabbit IgG antibodies, but lacked most or all the allergenic properties of recombinant Lol p 1. A study of the antigenic structure of Lol p 1 was performed using the three deletion mutants and a set of 17-18-residue overlapping synthetic peptides covering the whole allergen sequence. The results indicate that human IgE and rabbit IgG antibodies bind to distinct regions of Lol p 1, and that at least some important IgE epitopes are mainly conformational. The findings suggest that recombinant allergens constitute useful reagents for further development of serological diagnosis of allergy, and that it should be possible to produce immunogenic fragments of allergenic proteins without allergenic properties.

International collaborative study to evaluate a recombinant L ferritin preparation as an International Standard

A recombinant L ferritin preparation, lyophilized in ampoules and designated 94/572, was evaluated by 18 laboratories in 9 countries for its suitability as an International Standard (IS). The preparation was assayed in a wide range of in-house and commercial immunoassays against the 2nd IS for ferritin (of spleen origin; 80/578). The immunological reactivity of the recombinant material was similar to that of the 2nd IS for ferritin in the majority of assays and demonstrated adequate stability in accelerated degradation studies. On the basis of the results presented here, the WHO Expert Committee on Biological Standardization established 94/572 as the 3rd IS for ferritin, recombinant.

Hereditary hyperferritinemia-cataract syndrome: relationship between phenotypes and specific mutations in the iron-responsive element of ferritin light-chain mRNA

Recent reports have described families in whom a combination of elevated serum ferritin not related to iron overload and congenital nuclear cataract is transmitted as an autosomal dominant trait. We have studied the molecular pathogenesis of hyperferritinemia in two families showing different phenotypic expression of this new genetic disorder. Serum ferritin levels ranged from 950 to 1,890 microg/L in affected individuals from family 1, and from 366 to 635 microg/L in those from family 2. Cataract was clinically manifested in family 1 and asymptomatic in family 2. By using monoclonal antibodies specific for the H and L ferritin subunits, serum ferritin was found to be essentially L type in both normal and affected individuals. The latter also showed normal amounts of H-type ferritin in circulating mononuclear cells; on the contrary, L-type ferritin contents were 13 times normal in family 1 and five times normal in family 2 on average. Serum ferritin was glycosylated in both normal and affected individuals. There was a close relationship between mononuclear cell L-type ferritin content and serum ferritin concentration (r = 0.95, P < .00001), suggesting that the excess production of ferritin in cells was directly responsible for the hyperferritinemia. The dysregulated L-subunit synthesis was found to result from different point mutations in a noncoding sequence of genomic L-subunit DNA, which behaves as an mRNA cis-acting element known as iron regulatory element (IRE). Affected individuals from family 1 were heterozygous for a point mutation (a single G to A change) in the highly conserved, three-nucleotide motif forming the IRE bulge. Affected members from family 2 were heterozygous for a double point mutation in the IRE lower stem. Using a gel retardation assay, the observed molecular lesions were shown to variably reduce the IRE affinity for an iron regulatory protein (IRP), which inhibits ferritin mRNA translation. The direct relationship between the degree of hyperferritinemia and severity of cataract suggests that this latter is the consequence of excessive ferritin production within the lens fibers. These findings provide strong evidence that serum ferritin is a byproduct of intracellular ferritin synthesis and that the L-subunit gene on chromosome 19 is the source of glycosylated serum ferritin. From a practical standpoint, this new genetic disorder should be taken into account by clinicians when facing a high serum ferritin in an apparently healthy person.

Amnioinfusion for prevention of pulmonary hypoplasia in second-trimester rupture of membranes

We conducted a study to evaluate the feasibility and benefits of transabdominal amnioinfusion in preterm premature rupture of membranes with persistent oligohydramnios for the prevention of pulmonary hypoplasia. To this purpose, we designed a cohort study in which the pregnancy outcome of women with rupture of membranes at < or = 25 weeks and persistent (> or = 4 days) oligohydramnios managed with serial amnioinfusions (n = 18) was compared with that of a historic cohort group (controls) with similar characteristics but managed expectantly (n = 16). Pulmonary hypoplasia was diagnosed at birth in the presence of strict radiological and pathological criteria. No amnioinfusion-related complications occurred. The prevalence of pulmonary hypoplasia was significantly lower among the amnioinfused cases compared with the controls (46% [6 of 13] vs 86% [12 of 14], odds ratio [OR] = 0.4, 95% confidence interval [CI] 0.2-0.9), despite a lower gestational age at rupture of membranes in the treated group. Within the group undergoing amnioinfusions, those in which the infused solution was rapidly lost had a higher rate of pulmonary hypoplasia compared with those in which amnioinfusion alleviated oligohydramnios for > 48 hours (considered successful) (0 of 4 vs. 6 of 9, OR = 2.3, 95% CI 1-5.5). Cases of successful amnioinfusion had a longer interval between membrane rupture and appearance of oligohydramnios than those in which the procedure failed to correct oligohydramnios, even though both groups had similar gestational age at appearance of oligohydramnios. This suggests that the rate of loss of amniotic fluid after membrane rupture may predict the rate of loss of the infused solution, and therefore identify a subset of patients who may benefit from the procedure.

Effects of modifications near the 2-, 3- and 4-fold symmetry axes on human ferritin renaturation

Ferritin is a protein of 24 subunits which assemble into a shell with 432 point symmetry. It can be denatured reversibly in acidic guanidine hydrochloride, with the formation of poorly populated renaturation intermediates. In order to increase the accumulation of intermediates and to study the mechanism of ferritin renaturation, we analysed variants of the human ferritin H-chain altered at the N-terminus (delta(1-13)), near the 4-fold axis (Leu-169 --> Arg), the 3-fold axis (Asp-131 --> Ile + Glu-134 --> Phe) or the 2-fold axis (Ile-85 --> Cys). We also carried out specific chemical modifications of Cys-130 (near the 3-fold axis) and Cys-85 (near the 2-fold axis). Renaturation of the modified ferritins yielded assembly intermediates that differed in size and physical properties. Alterations of residues around the 2-, 4- and 3-fold axes produced subunit monomers, dimers and higher oligomers respectively. All these intermediates could be induced to assemble into ferritin 24-mers by concentrating them or by co-renaturing them with wild-type H-ferritin. The results support the hypothesis that the symmetric subunit dimers are the building blocks of ferritin assembly, and are consistent with a reassembly pathway involving the coalescence of dimers, probably around the 4-fold axis, followed by stepwise addition of dimers until the 24-mer cage is completed. In addition they show that assembly interactions are responsible for the large hysteresis of folding and unfolding plots. The implications of the studies for in vivo heteropolymer formation in vertebrates, which have two types of ferritin chain (H and L), are discussed.

Human recombinant antibody fragments specific for a rye-grass pollen allergen: characterization and potential applications

One of the major allergens from the pollen of perennial rye grass (Lolium perenne), Lol pII, was used to isolate specific antibody fragments from a random combinatorial library displaying a large repertoire of human Fab on filamentous phages. After five panning cycles on recombinant Lol pII immunotubes, phage binders were isolated and the antibody fragments expressed as soluble Fab molecules in the Escherichia coli periplasm. The DNA sequencing of the clones producing antibodies with the highest binding activity showed three of them to be identical, while one differed by two amino acid substitutions in the heavy chain. The antibody fragments were produced in milligram amounts, affinity-purified and further characterized. They bound the natural allergen as well as the recombinant one, with no cross-reactivity with other allergens contained in the pollen extract of L. perenne. One antibody bound the allergen with Kd = 2.63 x 10(-9) M, as demonstrated by the surface plasmon resonance technique, and was able to compete with a fraction of serum IgE. Epitope mapping using synthetic peptides revealed that antigenic domains, located between amino acids 39 and 51 of Lol pII, are recognized by Fab and polyclonal IgE from sera of allergic donors. The Fab fragments inhibited the binding of serum IgE to the allergen. In vitro experiments on whole blood from allergic subjects showed that recombinant Fab fragments had a blocking activity on histamine release from cells challenged with recombinant Lol pII allergen. Thus, serum IgE and recombinant Fab fragments recognize common epitopes, although they represent the outcome of different maturation and/or selection processes. Our molecular and functional findings altogether indicate that allergen-specific human antibodies may be useful for the characterization of the antigenic structure of allergens. We conclude that a phage library is a powerful source of anti-allergen human antibodies with high affinity and high specificity. Moreover, these molecules may be potentially innovative reagents for the treatment of atopic allergy.

The ferritins: molecular properties, iron storage function and cellular regulation

The iron storage protein, ferritin, plays a key role in iron metabolism. Its ability to sequester the element gives ferritin the dual functions of iron detoxification and iron reserve. The importance of these functions is emphasised by ferritin's ubiquitous distribution among living species. Ferritin's three-dimensional structure is highly conserved. All ferritins have 24 protein subunits arranged in 432 symmetry to give a hollow shell with an 80 A diameter cavity capable of storing up to 4500 Fe(III) atoms as an inorganic complex. Subunits are folded as 4-helix bundles each having a fifth short helix at roughly 60 degrees to the bundle axis. Structural features of ferritins from humans, horse, bullfrog and bacteria are described: all have essentially the same architecture in spite of large variations in primary structure (amino acid sequence identities can be as low as 14%) and the presence in some bacterial ferritins of haem groups. Ferritin molecules isolated from vertebrates are composed of two types of subunit (H and L), whereas those from plants and bacteria contain only H-type chains, where 'H-type' is associated with the presence of centres catalysing the oxidation of two Fe(II) atoms. The similarity between the dinuclear iron centres of ferritin H-chains and those of ribonucleotide reductase and other proteins suggests a possible wider evolutionary linkage. A great deal of research effort is now concentrated on two aspects of ferritin: its functional mechanisms and its regulation. These form the major part of the review. Steps in iron storage within ferritin molecules consist of Fe(II) oxidation, Fe(III) migration and the nucleation and growth of the iron core mineral. H-chains are important for Fe(II) oxidation and L-chains assist in core formation. Iron mobilisation, relevant to ferritin's role as iron reserve, is also discussed. Translational regulation of mammalian ferritin synthesis in response to iron and the apparent links between iron and citrate metabolism through a single molecule with dual function are described. The molecule, when binding a [4Fe-4S] cluster, is a functioning (cytoplasmic) aconitase. When cellular iron is low, loss of the [4Fe-4S] cluster allows the molecule to bind to the 5'-untranslated region (5'-UTR) of the ferritin m-RNA and thus to repress translation. In this form it is known as the iron regulatory protein (IRP) and the stem-loop RNA structure to which it binds is the iron regulatory element (IRE). IREs are found in the 3'-UTR of the transferrin receptor and in the 5'-UTR of erythroid aminolaevulinic acid synthase, enabling tight co-ordination between cellular iron uptake and the synthesis of ferritin and haem. Degradation of ferritin could potentially lead to an increase in toxicity due to uncontrolled release of iron. Degradation within membrane-encapsulated "secondary lysosomes' may avoid this problem and this seems to be the origin of another form of storage iron known as haemosiderin. However, in certain pathological states, massive deposits of "haemosiderin' are found which do not arise directly from ferritin breakdown. Understanding the numerous inter-relationships between the various intracellular iron complexes presents a major challenge.

Evidence that residues exposed on the three-fold channels have active roles in the mechanism of ferritin iron incorporation

Iron is thought to enter the ferritin cavity via the three-fold channel, which is lined in its narrowest part by the residues Asp-131 and Glu-134. We describe here variants of human ferritins with active and inactive ferroxidase centres having Asp-131 and Glu-134 substituted with Ala and Ala or with Ile and Phe respectively. The two types of substitution had similar effects on ferritin functionality: (i) they decreased the amount of iron incorporated from Fe(II) solutions and decreased ferroxidase activity by about 50%; (ii) they inhibited iron incorporation from Fe(III) citrate in the presence of ascorbate; (iii) they resulted in loss of Fe and Tb binding sites; and (iv) they resulted in a marked decrease in the inhibition of iron oxidation by Tb (but not by Zn). In addition, it was found that substitution with Ala of Cys-130 and His-118, both of which face the three-fold channel, decreased the capacity of H-ferritin to bind terbium and to incorporate iron from Fe(III) citrate in the presence of ascorbate. The results indicate that: (i) in three-fold channels are the major sites of iron transfer into the cavity of H- and L-ferritins; (ii) at least two metal binding sites are located on the channels which play an active role in capturing and transferring iron into the cavity; and (iii) the permeability of the channel is apparently not affected by the hydrophilicity of its narrowest part. In addition, it is proposed that iron incorporation from Fe(III) citrate complexes in the presence of ascorbate is a reliable, and possibly more physiological, approach to the study of ferritin functionality.

Evidence that the specificity of iron incorporation into homopolymers of human ferritin L- and H-chains is conferred by the nucleation and ferroxidase centres

Mammalian ferritins are iron-storage proteins made of 24 subunits of two types: the H- and L-chains. L-chains, in contrast with H-chains, lack detectable ferroxidase activity. When ferritins were subjected to iron loading in vitro with increments near the saturation limit of 4000 Fe atoms per molecule, the homopolymers of human H-chains formed insoluble aggregates, caused by non-specific iron hydrolysis, whereas the homopolymers of L-chains remained soluble and incorporated most of the available iron. To analyse the molecular reasons for the difference, Glu-57 and Glu-60, which are conserved and exposed on the cavity of L-chains, were substituted with His, as in H-chains. The double substitution made the L-homopolymers as sensitive as the H-homopolymers to the iron-induced aggregation, whereas the opposite substitution in the H-chain increased homopolymer resistance to the aggregation only marginally. Millimolar concentrations of citrate and phosphate increased iron incorporation in H-homopolymers by reducing non-specific iron hydrolysis, but inhibited that in L-homopolymers by sequestering available iron. The data indicate that the specific iron incorporation into L-homopolymers is mainly due to the iron-nucleation capacity of Glu-57, Glu-60 and other carboxyl groups exposed on the cavity; in contrast, the specificity of iron incorporation into H-homopolymers is related to its ferroxidase activity, which determines rapid Fe(III) accumulation inside the cavity. The finding that ferroxidase centres are essential for the incorporation of iron in the presence of likely candidates of cellular iron transport, such as phosphate and citrate, confirms their importance in ferritin function in vivo.

A quantitative analysis of isoferritins in select regions of aged, parkinsonian, and Alzheimer's diseased brains

The brain requires a ready supply of iron for normal neurological function, but free iron is toxic. Consequently, iron bioavailability must be stringently regulated. Recent evidence has suggested that the brain iron regulatory system is dysfunctional in neurological disorders such as Alzheimer's and Parkinson's diseases (AD and PD, respectively). A key component of the iron regulatory system in the brain is ferritin. Ferritin consists of 24 subunits, which are distinguished as either a heavy-chain (H) or light-chain (L) isoform. These peptide subunits are genetically and functionally distinct. Thus, the ability to investigate separately the types of ferritin in brain should provide insight into iron management at both the cellular and the molecular level. In this study, the ratio of isoferritins was determined in select regions of adult elderly AD and PD human brains. The H-rich ferritin was more abundant in the young brain, except in the globus pallidus where the ratio of H/L ferritin was 1:1. The balance of H/L isoferritins was influenced by age, brain region, and disease state. With normal aging, both H and L ferritin increased; however, the age-associated increase in isoferritins generally failed to occur in AD and PD brain tissue. The imbalance in H/L isoferritins was disease and region specific. For example, in frontal cortex, there was a dramatic (fivefold) increase in the ratio of H/L ferritin in AD brains but not in PD brains. In PD, caudate and putamen H/L ratios were higher than in AD and the elderly control group.(ABSTRACT TRUNCATED AT 250 WORDS)

Tyrosyl radical formation during the oxidative deposition of iron in human apoferritin

The radical chemistry of ferritin is incompletely understood. The present study was undertaken to investigate the production of radicals in H-chain recombinant human ferritin (HuHF) and mixed H/L-chain horse spleen ferritin (HoSF) and the potential role of radicals in the oxidative deposition of iron in these proteins. Radical production follows distinct pathways for the two proteins; an intact H-chain ferroxidase site is required for radical generation in both of them, however. With the H-chain HuHF, an EPR spectrum characteristic of a tyrosyl radical is seen following Fe2+ oxidation by O2 and, based on measurements with site-directed variants, is suggested to arise from residue Tyr-34 located in the vicinity of the ferroxidase site. The observation of this radical correlates with the observation of a 400-600 nm absorbance seen in stopped-flow kinetics studies which seems to require the presence of Tyr-34 (Bauminger et al. (1993) Biochem. J. 296, 709-714). The data are inconsistent, however, with the Tyr-34 radical being critically important in the protein-catalyzed mechanism of iron oxidation. Unlike HuHF, the radicals observed in L-chain-rich HoSF appear to arise from hydroxyl radical damage to the protein through Fenton chemistry. These latter radicals also appear to be centered on aromatic amino acids and may be derived from histidine.

Recombinant allergen Lol p II: expression, purification and characterization

Pollen from perennial rye grass (Lolium perenne) is a major cause of type I allergies worldwide. It contains complex mixtures of proteins, among which Lol p II is a major allergen. Previously, we have reported the cloning and sequencing of Lol p II and its expression in fusion with the heavy chain of human ferritin as carrier polypeptide (Sidoli et al., 1993, J. biol. Chem. 268, 21819-21825). Here, we describe the expression, purification and characterization of a recombinant Lol p II overproduced as a non-fusion protein in the periplasm of E. coli. The recombinant allergen was expressed in high yields and was easily purified in milligram amounts. It competed with the natural Lol p II for binding to specific IgE, and it induced allergic responses in skin prick tests, indicating to be immunologically analogous to the natural protein. Biochemical analyses indicate that recombinant Lol p II is a highly stable and soluble monomeric molecule which behaves like a small globular protein.

Reconstitution of manganese oxide cores in horse spleen and recombinant ferritins

The formation of Mn(III) oxyhydroxide (MnOOH) cores within the nanoscale cavity of the iron storage protein ferritin has been investigated by electron microscopy and visible absorption spectroscopy. At pH 8.9, discrete amorphous MnOOH cores were formed within horse spleen apoferritin at a range of metal:protein ratios, as well as in ferritin molecules seeded with a small ferrihydrite nucleus. Analysis of the resultant core size distributions showed that the reconstitution of horse spleen apoferritin with Mn(II) was similar to that observed previously for Fe(II) reconstitution in recombinant human L-chain ferritin, suggesting that horse spleen apoferritin does not exhibit Mn(II) oxidase activity at pH 8.9. Reconstitution with MnOOH shows essentially "all-or-nothing" behavior in which many protein molecules remain unmineralized whilst others are loaded to maximum capacity. Kinetic studies showed no significant differences between horse spleen ferritin, recombinant H- and L-chain homopolymers, and H-chain variants containing site-directed modifications at the ferroxidase and putative Fe nucleation centers. Our results indicate that the reconstitution of ferritin with MnOOH cores proceeds by a nonspecific pathway. We propose that the outer surface of the protein inhibits the development of MnOOH nuclei in bulk solution whereas the inner surface is inactive, enabling nucleation and growth to proceed unperturbed within the cavity. One possibility is that differences in the general polyelectrolyte properties of these two surfaces, rather than site-specific charges, account for the "Janus" behavior of the molecule. A similar mechanism might also increase the specificity of iron oxide mineralization in ferritins that lack ferroxidase centers.

Construction of a ferroxidase center in human ferritin L-chain

Ferritins are 24-mer proteins which store and detoxify intracellular iron. Mammalian ferritins are made of two subunit types, the H- and L-chains, with different functional specificity. The H-chain has a metal-binding site (the ferroxidase center) which confers ferroxidase activity to the protein and accelerates iron incorporation. In the L-chain the center is substituted by a salt bridge. We performed several site-directed mutageneses in the L-chain with the aim to construct the center and confer ferroxidase activity to the protein. Most variants were insoluble and did not refold into homopolymers, probably due to electrostatic repulsion introduced by the substitutions. However, they formed hybrids when they were renatured together with the L- or H-chains. The heteropolymers made of 90% L-chain and 10% of an L-variant with all the ligand residues of the H-chain center had 25-30% of the ferroxidase activity of the H-chain homopolymer. This corresponds to the activity of an H/L heteropolymer with 7% H-chain. It is concluded that: (i) it is possible to construct a ferroxidase center in the L-chain with an activity equivalent to that of the H-chain, (ii) the residues of the center interfere with the folding/assembly of the L-, but not of the H-chain, (iii) heteropolymers can be made even between ferritin subunits with large differences of refolding rates.