Mutant L-chain ferritins that cause neuroferritinopathy alter ferritin functionality and iron permeability

Justin R. McNally a, Matthew R. Mehlenbacher a, Sara Luscieti b, Gideon L. Smith a, Aliaksandra A. Reutovich a, Poli Maura b, Paolo Arosio b and Fadi Bou-Abdallah *a
aDepartment of Chemistry, State University of New York, Potsdam, New York 13676, USA. E-mail:
bDepartment of Molecular and Translational Medicine, University of Brescia, viale Europa 11, 25123 Brescia, Italy

Received 16th June 2019 , Accepted 14th August 2019

First published on 14th August 2019

In mammals, the iron storage and detoxification protein ferritin is composed of two functionally and genetically distinct subunit types, H (heavy) and L (light). The two subunits co-assemble in various ratios, with a tissue specific distribution, to form shell-like protein structures of 24 subunits within which a mineralized iron core is stored. The H-subunits possess ferroxidase centers that catalyze the rapid oxidation of ferrous ions, whereas the L-subunit does not have such centers and is believed to play an important role in electron transfer reactions that occur during the uptake and release of iron. Pathogenic mutations on the L-chain lead to neuroferritinopathy, a neurodegenerative disease characterized by abnormal accumulation of ferritin inclusion bodies and iron in the central nervous system. Here, we have characterized the thermal stability, iron loading capacity, iron uptake, and iron release properties of ferritin heteropolymers carrying the three pathogenic L-ferritin mutants (L154fs, L167fs, and L148fs, which for simplicity we named Ln1, Ln2 and Ln3, respectively), and a non-pathogenic variant (L135P) bearing a single substitution on the 3-fold axes of L-subunits. The UV-Vis data show a similar iron loading capacity (ranging between 1800 to 2400 Fe(III)/shell) for all ferritin samples examined in this study, with Ln2 holding the least amount of iron (i.e. 1800 Fe(III)/shell). The three pathogenic L-ferritin mutants revealed higher rates of iron oxidation and iron release, suggesting that a few mutated L-chains on the heteropolymer have a significant effect on iron permeability through the ferritin shell. DSC thermograms showed a strong destabilization effect, the severity of which depends on the location of the frameshift mutations (i.e. wt heteropolymer ferritin ≅ homopolymer H-chain > L135P > Ln2 > Ln1 > Ln3). Variant L135P had only minor effects on the protein functionality and stability, suggesting that local melting of the 3-fold axes in this variant may not be responsible for neuroferritinopathy-like disorders. The data support the hypothesis that hereditary neuroferritinopathies are due to alterations of ferritin functionality and lower physical stability which correlate with the frameshifts introduced at the C-terminal sequence and explain the dominant transmission of the disorder.

Significance to metallomics

Ferritins are ubiquitous and well-characterized iron storage proteins that play essential roles in cellular iron homeostasis. Neuroferritinopathy is caused by modification of the C-terminal region of ferritin L-subunit and is characterized by ferritin inclusion bodies and brain iron deposits. Our results suggest that disruption of ferritin structure significantly alters the protein functionality, leading to lower thermal stability and enhanced iron permeability. We propose that neuroferritinopathy pathogenesis may be due to structural and functional impairment of ferritin resulting in iron mis-management, iron-induced oxidative damage, and that protein aggregation may be a secondary phenomenon and not a causative effect.


Cellular iron availability depends largely on ferritin, a ubiquitous and exceptionally stable iron storage and detoxification protein that plays an essential role in iron homeostasis. Ferritin belongs to a family of highly conserved supramolecular protein nanostructures composed of 24 subunits that are functionally and genetically different (H-subunit of ∼21[thin space (1/6-em)]000 Da and L-subunit of ∼19[thin space (1/6-em)]000 Da).1–5 The H-subunit is enzymatically active and contains a ferroxidase center that catalyzes the oxidation of Fe(II) to Fe(III) by molecular oxygen (or hydrogen peroxide) followed by Fe(III) hydrolysis and iron core formation.1–5 The L-subunit is enzymatically inactive and lacks such a center but instead has high density of carboxyl groups on the protein's inner surface that are shown to provide efficient sites for iron nucleation and mineralization. More recent work have suggested a possible role for L-chain in electron transfer reactions.6 In humans, the H and L subunits co-assemble in various ratios with a tissue specific distribution to form heteropolymeric 24-mer proteins (i.e. up to 90% H-subunits or H-rich ferritins are found in hearts and brains whereas spleens and livers are primarily composed of L-subunits and thus are L-rich ferritins).5,7–9 A couple of notable exceptions include human mitochondrial ferritin, a homopolymer composed of 100% H-like subunits,10 and human serum ferritin, a homopolymer composed of ∼100% L subunits.11 Structurally, human H and L subunits share a 55% sequence identity and have a remarkably similar 3D structure consisting of a pair of anti-parallel α-helices (A–B and C–D) with a long loop connecting helix B to helix C, and a fifth shorter E-helix that forms a 60° angle with respect to the other helices. The E-helix is directed towards the center of the cavity and appears to play an important role in stabilizing the 24-mer protein.12

Neuroferritinopathy (NF) is a rare autosomal-dominant genetic disease caused by mutations of the ferritin light chain gene (FTL).12,13 It belongs to a group of movement disorders referred to “Neurodegeneration with Brain Iron Accumulation” characterized by focal iron accumulation in specific regions of the brain.14 Clinically, NF is characterized by an abnormal involuntary movement disorder and cognitive decline that start to manifest in the 3rd decade of life.15,16 Neuropathologically, the central nervous system and other organs of NF patients show the presence of intracellular ferritin inclusion bodies and iron accumulation in glia and neurons.12,15 To date, nine causative mutations have been identified, eight of them are frameshift mutations determined by nucleotide(s) insertion in the exon 4 of the L-ferritin gene,13,15,17–25 and one missense mutation in exon 3 of FTL, found in only one subject.25 A detailed description of the L-chain ferritin mutants (i.e. neuroferritinopathy variants known to date) is reviewed elsewhere.12 Given the extensive inter-chain contacts of the C-termini of L-subunits, all frameshift mutations strongly affect the proper folding and assembly of the L-subunit C-terminus at the 4-fold symmetry axis.15 Notably, no pathogenic mutations on the H-chain have been reported to date. In this study, three pathogenic L-ferritin mutants were employed (L148fs or Ln3, L154fs or Ln1, and L167fs or Ln2) and consist of different nucleotides insertions at the C-terminus of L-chain that introduce frameshifts with a fully penetrant effect making the L-subunit longer than the undisturbed native L-subunit. For instance, insertion of one nucleotide (i.e. 442InsC or 460InsA, representing L148fs (Ln3) and L154fs (Ln1), respectively) introduces 4-amino acids extension at the C-terminus, whereas insertion of two nucleotides (i.e. 498InsTC representing L167fs) introduces a 16-amino acid extension. These insertions modify the L-chain sequence (comprising of 175 amino acids) starting at residues 148 for Ln3, 154 for Ln1, and 167 for Ln2.12,15

Previous mutational studies on human H-chain ferritin showed that amino acid extension at the C-terminus, or fusion of a short peptide, had no major effects on the protein stability or functionality.26–29 However, amino acid residues truncation or deletion around the C-terminus-region produced ferritin molecules that are either unable to retain an iron core30 or promote protein assembly when expressed in E. coli.31 Interestingly, nucleotide insertions in H-ferritin at sites equivalent to those found in neuroferritinopathy variants led to a decrease in protein solubility and functionality, the severity of which depended on the location of the nucleotide insertions.27 A recent site-directed mutagenesis study on the effect of amino acid modification around the C-terminal region of Chlorobium tepidum ferritin revealed an important role of the C-terminus for the protein ferroxidase activity and stability,32 consistent with the above observations.

Here, we have characterized the thermodynamic stability, iron loading capacity, iron uptake and iron release properties of three pathogenic L-ferritin mutants (L154fs, L167fs, and L148fs, also known as Ln1, Ln2 and Ln3, respectively), and one non-pathogenic variant (L135P) that has a single substitution at the 3-fold axes of the L-subunits. L135P is the human homologue of L134P in frog ferritin that was shown to cause a local unfolding around the 3-fold axes and to strongly accelerate the reductive release of iron from frog H-subunit ferritin.33 Our results show an iron loading capacity for the three pathogenic ferritin samples ranging between 1800 to 2400 Fe(III)/shell, with Ln2 holding the least amount of iron (i.e. 1800 FeIII/shell). Compared to wild type ferritin, the relatively faster rates of iron oxidation and iron release (particularly with Ln3) suggest that a few mutated L-chains within the pathogenic heteropolymer ferritin shell exhibit a significant effect on iron permeability. The non-pathogenic 3-fold channel mutant L135P showed no significant changes in the protein thermal denaturation profile or in the rates of iron uptake and release, suggesting that modification of the hydrophilic 3-fold axes do not exhibit a major negative-dominant effect as observed with the pathogenic variants. The DSC thermograms displayed a strong destabilization effect induced by the introduction of frameshift mutations, the severity of which depended on the location of the frame shift mutation (i.e. Wt > L135P > Ln2 > Ln1 > Ln3). Compared to wt H/L (main Tm = 101.8 °C at the highest heat capacity change), mutant Ln3 was shown to be the least stable with a main unfolding temperature of Tm = 85.6 °C, followed by Ln1 (Tm = 89.4 °C) and then Ln2 (Tm = 93.9 °C). Overall, our data support the hypothesis that hereditary ferritinopathies are due to alterations of ferritin functionality including lower thermal stability and increased rates of iron uptake and mobilization.

Materials and methods

Plasmid construction, expression, and purification of recombinant heteropolymer WT and mutant ferritins

The construction of the plasmids, expression and purification of the heteropolymers is described in details elsewhere.15 Briefly, the cDNAs coding for the L-subunit or its mutants Ln1, Ln2 and Ln3, preceded by the ribosome binding site, were cloned immediately downstream the stop codon of the ferritin H subunit in the pET-Hwt vector that expresses the H ferritin under the T7 promoter. We thus obtained the vectors pET-H/L, pET-H/Ln1, pET-H/Ln2 and pET-H/Ln3 which expressed about 90% of the H subunit together with about 10% of the L subunit (where the mutations occurred). The plasmids were used to transform E. coli BL21-pLYS and the expression induced by IPTG (0.4 mM). The proteins were purified as described before.34 Ferritin purity and preliminary subunit composition was analyzed on 7.5% native PAGE and 15% SDS-PAGE with Coomassie Blue and Prussian blue staining (Fig. 1). The schematic below show the alignment of the C-terminal sequence of the wt H/L heteropolymer and the pathogenic L-ferritin mutants of this study. The amino acid numbering above includes the N-terminal methionine. L148fs, L154fs and L167fs represent ferritin L-mutants 442InsC (or Ln3), 460InsA (or Ln1) and 498InsTC (or Ln2), respectively, as described in ref. 15. The D and E helices on the ferritin subunit are indicated by boxes.
image file: c9mt00154a-f1.tif
Fig. 1 Characterization of wt and mutant ferritin and their subunit composition by electrophoresis (non-denaturing PAGE, SDS-PAGE and capillary gel electrophoresis). (A) Non-denaturing and SDS-PAGE stained with Coomassie Blue (Co Blue or protein stain), or Western blotted and stained with polyclonal anti FtL antibodies for L-ferritin (anti-FtL Ab), and rH02 or anti-FtH Ab. (B) Non-denaturing PAGE stained for protein (Co Blue) or iron (Prussian blue, Pr. Blue or iron stain). Each lane of the native and SDS gels was loaded with 1 μg protein and each ferritin sample was incubated with 4000 Fe(II) atoms/protein. (C) SDS-CGE electropherograms of human heteropolymers ferritin H/L and various mutants. The concentration of the protein samples ranged between 0.5–2 mg ml−1 and were prepared from different ferritin samples having different stock concentrations. The instrument conditions are specified under Materials and methods.
Iron incorporation. Apoferritins (1 μM, 0.5 mg ml−1) were incubated for 2 h at room temperature with 0.5–4.0 mM freshly prepared ferrous ammonium sulfate in 0.1 M Hepes Buffer, pH 7. The samples were run on non-denaturing 7.5% polyacrylamide gels and stained for protein (Coomassie Brilliant Blue) or iron (Prussian Blue).
Immunological methods. In Western blotting of non-denaturing PAGE, the monoclonal antibody rH02 (5 μg ml−1) was used for H-ferritin detection in the heteropolymer. For the detection of L-ferritin, SDS-PAGE was used with the polyclonal rabbit anti-ferritin antibody (Sigma) that preferentially recognizes the L-chains. Antibody binding was revealed using horseradish peroxidase-labeled secondary antibody and ECL detection. Coomassie Brilliant Blue staining was used to verify the loading of both non-denaturing and SDS-PAGE.
image file: c9mt00154a-u1.tif

Proteins and chemicals

Holoferritins (i.e. iron containing ferritin) were rendered iron-free by dialysis against sodium hydrosulfite (dithionite), Na2S2O4, and complexation with 2,2′-bipyridyl at pH 6.0. Protein concentrations were determined spectrophotometrically using a molar absorptivity of 24[thin space (1/6-em)]000 cm−1 M−1 at 280 nm for the 24-mer apoprotein (iron-free protein).35 All chemicals were reagent grade and used without further purification. Mops (3-(N-morpholino)-propanesulfonic acid) buffer was purchased from Research Organics (Cleveland, OH) and FeSO4·7H2O from J. T. Baker (Phillipsburg, NJ). Sodium dithionite, Na2S2O4 ferrozine and 2,2′-bipyridyl were purchased from Sigma-Aldrich (St. Louis, MO). Ferrous sulfate stock solutions were freshly prepared immediately before each experiment in a dilute HCl solution at pH 2.0.

Capillary gel electrophoresis (CGE): chemicals and experimental conditions

All chemicals, reagents, and supplies were used as received and included a Sciex CE-SDS analysis kit, 2-mercaptoethanol and bovine serum albumin (BSA) from Sigma Aldrich (St. Louis, MO, USA). The reagents in the kit include SDS-MW gel buffer (proprietary formulation, pH 8, 0.2% SDS), CE-SDS sample buffer (100 mM Tris–HCl pH 9.0, 1% SDS), acidic wash solution (0.1 N HCl), and basic wash solution (0.1 N NaOH). An Agilent Technologies 50 μm ID bare fused silica capillary, with a total length of 33 cm and an effective length 24.5 cm, was used in the CGE experiments. The SDS-CGE capillary was pre-conditioned under a high pressure of 2.0 bar using 0.1 N NaOH (10 min), 0.1 N HCl (5 min), water (2 min), and finally a high pressure flush (4.0 bar) of the SDS gel buffer for 10 min.

Prior to each protein run, the capillary was conditioned under a high pressure flush of 4 bar using 0.1 N NaOH (3 min), 0.1 N HCl (1 min), water (1 min) and finally SDS gel buffer (10 min). The protein samples were injected electro-kinetically by applying a negative voltage of −5 kV for 20 s. Protein separation was followed under a negative applied voltage of −16.5 kV for 30 min. A 2.0 bar pressure was applied to both inlet and outlet vials for the duration of the experiment with the capillary temperature maintained at 25 °C. The detection wavelength was set at 220 nm (10 nm bandwidth) with a reference wavelength of 350 nm (10 nm bandwidth) and a response time of 1 s.

CGE sample preparation

Typically, ferritin solutions (100 μl at 1–2 mg ml−1) were prepared in an SDS sample buffer (>60% by volume) in the presence of 5 μl 2-mercaptoethanol (5% v/v). The ferritin solution was mixed thoroughly, tightly capped and heated in a 100 °C water bath for at least 10 min. The protein solution was then cooled to room temperature prior to running on the 7100 model capillary electrophoresis instrument from Agilent Technologies.

Oxygen electrode (oximetry)

The oximetry experiments were performed with an OM-4 oxygen meter (Microelectrodes, Inc., Bedford, NH) equipped with an MI-730 micro-oxygen electrode. The electrode oximetry apparatus and standardization reactions have been described in detail elsewhere.36 Typical experimental conditions were 1 μM protein in 100 mM Mops, 50 mM NaCl, pH 7.0 and 25 °C, with multiple 42 Fe(II)/protein added from a freshly prepared 10.0 mM reagent-grade ferrous sulfate (Baker Scientific Inc.) in pH 3 water.

UV-vis spectroscopy: kinetics of iron oxidation and release

Conventional UV-vis spectroscopy was performed on a Varian Cary 50 Bio spectrophotometer from Agilent Technologies. All experiments were conducted at 25.00 °C, in 100 mM MOPS buffer and 50 mM NaCl, pH 7.0. Experimental conditions are provided in the figure captions. All kinetic experiments were repeated two to four times using freshly prepared samples and/or independent protein preparations to ensure reproducibility. The kinetic traces shown in the figures are representative results from multiple runs.
Iron oxidation kinetics. The kinetics of iron oxidation in ferritin were followed at 305 nm where the Fe(III) oxo(hydroxo) species absorbs. The instrument was zeroed using the iron-free ferritin solution, prepared in buffer as the blank. Typically two or three μl of a ferrous sulfate solution prepared in deionized H2O (pH ∼ 2) were injected into a 1.0 ml protein solution, with rapid spin bar stirring under the conditions stated in the figure captions.
Iron release kinetics. The iron release kinetics were performed aerobically in a 1 ml UV-vis quartz cuvette (1 cm path light) and a screw-on cap to prevent atmospheric oxygen diffusion inside the cell. To exclude any interference from trapped oxygen bubbles in the cuvette head space, and to further minimize atmospheric oxygen diffusion inside the cell, the experiments were repeated using a UV cell filled to the brim with the protein solution. In either case, similar results of iron release kinetics were observed. In each iron release experiment, solutions of iron-loaded human recombinant ferritin (0.5 or 0.05 μM) were prepared in 100 mM MOPS, 50 mM NaCl, pH 7.0 and mixed with NADH (2.5 mM), ferrozine (0.4 mM), and FMN (2.5 mM). Soon after the addition of all reagents, the solution was rapidly inverted several times for thorough mixing. The lag phase that preceded the rise in absorbance at 560 nm varied in length (or time), and depended on the amount of dissolved oxygen present in solution.37–39 All iron release experiments were performed two to three times to ensure reproducibility. Time-dependent absorbance kinetic traces were collected at 25 °C and the data analyzed with OriginLab version 8.0 (OriginLab Corp.). Initial rates were obtained from the linear A1 term of a second-order polynomial curve fitted to the initial part of the experimental data, namely Y = A0 + A1t + A2t2 and dY/dt = A1 + 2A2t (at t = 0, dY/dt = A1), where t is time and Y is the change in O2 uptake, or in absorbance at 305 nm.

Differential scanning calorimetry: ferritin thermal stability

DSC experiments were performed on a NanoDSC instrument from TA Instruments under a constant pressure of 3 atm, in the temperature range of 25–120 °C, using a scan rate of 1 °C min−1. All ferritin samples (1 mg ml−1) were prepared in 50 mM phosphate buffer, 50 mM NaCl, pH 7.0 and degassed for 20 minutes prior to analysis using the TA Instrument degassing station. The DSC raw data was corrected by subtracting a buffer–buffer background scan and analyzed using the TA Instruments DSC NanoAnalyze software. Integration of the heat capacity of the sample vs. temperature yields the enthalpy of the unfolding process. Model-independent calorimetric enthalpy (ΔH) values were obtained from the area under the unfolding curves. The transition midpoint (Tm, or the melting temperature) is the temperature at which half the protein molecules are folded and half are unfolded and is determined at maximum heat capacity at the top of the melting curve. The entropy (ΔS) is obtained from the area under the curve of Cp/T vs. T.

Results and discussion

The presence and subunit composition of ferritin heteropolymers (denoted H/L in this study) have been confirmed by western blotting with non-denaturing-PAGE and H-ferritin specific antibody (Fig. 1A). Specifically, the H-rich ferritin heteropolymers containing the L-chain mutants Ln1, Ln2, Ln3 and L135P were run on non-denaturing PAGE, and protein samples stained with Coomassie blue showed similar pattern and mobility analogous to that of wild-type ferritin heteropolymers (H/L) and H- and L-homopolymers. The presence of the H-chain was confirmed by western blotting decorated with a monoclonal antibody specific for the human H-chain, named rH02 (Fig. 1A, upper gel). In all samples tested in this study (except in homopolymer L-ferritin), H-subunit was present, as expected. Ferritin L-subunit was detected by western blotting decorated with a polyclonal antibody specific for the L-chain. Of note is the slower mobility of Ln1 and Ln2 due to their larger size (Fig. 1A, bottom gel). The ferritin samples were analyzed on 7.5% polyacrylamide non-denaturing-PAGE, in the absence of SDS and beta-mercaptoethanol, in which the H-chain and L-chain homopolymers have different mobility. A preliminary analysis of the ferritin subunits was performed on samples denatured by heating at 100 °C in 2% SDS, 1% beta mercaptoethanol and then run on 15% polyacrylamide SDS-PAGE, which separates the H and L subunits. The ferritin samples were then analyzed for their capacity to incorporate iron (Fig. 1B). Apo-ferritin samples (1 μg) prepared in 0.1 M Hepes buffer pH 7, were incubated for 2 h with ferrous ammonium sulfate (i.e. 4000 Fe(II)/shell) and then separated on non-denaturing PAGE. Protein stain confirmed the equal load of the samples, with iron stain by Prussian blue showing that mutants Ln1 and Ln3 incorporated less iron than the wt H/L, Ln2 and L135P. As expected, the iron content of L-ferritin was less than that of homopolymer H-ferritin and heteropolymer H/L ferritin (Fig. 1B). To more accurately and quantitatively determine the ferritin subunit composition (wt and mutants), several freshly prepared heteropolymer H-rich ferritin samples were run on CE and showed a similar subunit composition (i.e. 20.5H[thin space (1/6-em)]:[thin space (1/6-em)]3.5L or 85.5% H and 14.5% L) (Fig. 1C), in agreement with earlier measurements.35 To further confirm the subunit composition of the H-rich ferritin, an aerobic spectrophotometric titration35 (or protein fluorescence quenching experiments) of apo H/L ferritin with Fe(II) showed a discontinuity in absorbance at ∼42 Fe(II) atoms/shell, corresponding to 2 Fe(III) atoms at each of the dinuclear ferroxidase center of the ∼20 H-subunits of the ferritin shell (data not shown).

Capillary electrophoresis: subunit composition and integrity of wt ferritins and its mutants

To quantify the integrity and the H- and L-subunit composition of recombinant mammalian heteropolymers ferritins and its mutants, capillary gel electrophoresis (CGE) was used under denaturing conditions (SDS-CGE). The SDS-CGE electropherograms of all ferritin samples showed well-resolved peaks for the L-subunit and the H-subunit with the exception of Ln2 and L135P which showed an additional feature in their H-subunit peaks suggesting some degree of heterogeneity within the H-subunit (Fig. 1C). Interestingly, the thermal denaturation profiles of these mutants (Fig. 2), which looks at the whole ferritin shell, exhibited multiple unfolding events attributed to the sequential thermal unfolding of different protein domains within the ferritin shell or to structurally different protein populations. From the area under the CE peaks, the wt and mutant heteropolymer ferritin samples were found to contain ∼10–15% L-subunits and 85–90% H-subunits, a result in good agreement with previously published SDS-PAGE data.15
image file: c9mt00154a-f2.tif
Fig. 2 DSC thermograms of different ferritin samples. The dotted lines represent the fits to the experimental solid curves (i.e. model sum of the blue, magenta, brown, and green lines). Conditions: 1 μM protein solutions, 50 mM sodium phosphate, 50 mM NaCl, pH 7.0. The experimental pressure was 3.00 ± 0.03 atm. For direct comparison, the last panel of this figure represents an overlay of the DSC thermograms of all ferritin samples tested in this study.

Thermal stability of ferritin samples

To obtain the denaturation or stability profiles of the wild type and mutant ferritins, differential scanning calorimetry (DSC) experiments were performed (Fig. 2). All ferritin samples exhibit complex thermal denaturation profiles involving multiple transitions, as indicated in Fig. 2 and could be attributed to ferritin shell heterogeneity and/or the thermal unfolding of different protein domains due to structural modifications introduced by the various frameshift mutations. The DSC profiles for wild type apo-heteropolymer H/L ferritin, homopolymers H-chain ferritin (HuHF) and L135P were relatively sharp (compared to the pathogenic variants) with main Tm values of 101.8 °C, 102 °C and 98.7 °C, respectively (Table 1). In contrast, deconvolution of the experimental DSC calorimetric curves of the three pathogenic ferritin variants showed much less stable proteins with the main Tm values ranging between 85 and 97 °C compared 102 °C for wt H/L (Fig. 2 and Table 1). The results of Fig. 2 revealed a protein stability profile in the order of Ln3 < Ln1 < Ln2 < L135P < HuHF ≈ wt H/L, consistent with the location of the frameshift mutation at the C-terminus of the L chain. While all frameshift mutations in the pathogenic variants introduced a partial or complete sequence change of the fifth ferritin E-helix (Fig. 3), Ln3 and Ln1 frameshift location is upstream of the loop connecting the D and E helices, whereas that of Ln2 is located downstream of the loop. The DSC results of Fig. 2 indicate that the more upstream the frameshift mutation, the stronger the effect and the less stable the protein, with Ln3 being the least stable and Ln2 the most stable. Due to its higher stability, several in vitro and in vivo studies have been performed on variant Ln2 in order to understand the pathogenic mechanism of the disease.12 When holo-ferritin samples (containing 200–300 Fe(III)/shell) were run, similar DSC patterns were observed (i.e. the Tm values for holo-ferritin were within 5% of those of apo-ferritin), suggesting no major differences in the thermal profiles between apo and holo-ferritins.
Table 1 DSC results for the thermal unfolding of 0.5 mg ml−1 ferritin samples. Subscripts 1, 2, 3 and 4 for the melting temperatures (Tm) and the enthalpy change of unfolding (ΔH) represent the different components of the deconvoluted DSC endotherms. The reported values of Tm and ΔH are average values from two runs that typically differ by 10% or less from run to run
Protein Enthalpy change of unfolding ΔH1 (kJ mol−1) Melting temp. Tm1 (°C) Enthalpy change of unfolding ΔH2 (kJ mol−1) Melting temp. Tm2 (°C) Enthalpy change of unfolding ΔH3 (kJ mol−1) Melting temp. Tm3 (°C) Enthalpy change of unfolding ΔH4 (kJ mol−1) Melting temp. Tm4 (°C) Entropy change of unfolding ΔS (kJ mol−1 K−1)
a Main Tm with the highest heat capacity change.
H/L 662.2 100.7 1755 101.8 1633 104.6 18.5
HuHF 925 97.9 1144 100.7 1852 102 2397 104.7 23.2
L135P 900.8 96.4 1239 98.7 1721 100.9 14.5
Ln1 485 84.8 523 89.4 469 94.2 26.7
Ln2 619 91.8 1093 93.9 906 97.5 1222 102.3 30.5
Ln3 473 78.5 866 82.1 808 85.6 29.8

image file: c9mt00154a-f3.tif
Fig. 3 Left panel: Schematic representation of the three pathogenic variants whereby the red segments at the C-terminus represent the mutated E-helix peptide region on the L-subunit as a result of the various frameshift mutations. Right panel: Full ferritin shell of the four-fold channels with one subunit (in cyan) showing the 5 helices encompassing the ferritin subunit (note the proper folding of the E-helix forming the 4-fold axes). Frameshift mutations up the D and E helices of the L-subunit cause neuroferritinopathy. The PDB for the ferritin structure is 2FHA and the program used to generate that image is PyMOL.

Effect of mutations on ferritin iron uptake capacity

To determine the maximum amount of iron that the proteins (wt H/L and its variants) can accommodate with O2 as the oxidant, a UV-vis spectrophotometric experiment was carried out. Multiple 300 Fe(II)/shell injections were added to the same protein sample, and the iron oxidation kinetics followed at 305 nm, where the oxo/hydroxo Fe(III)–protein species absorbs. As shown in Fig. 4, precipitation is quite evident once a certain ratio of Fe(III)/shell is reached (i.e. ∼1800 Fe(III)/shell for Ln2, ∼2100 Fe(III)/shell for Ln1, ∼2400 Fe(III)/shell for Ln3, ∼2100 Fe(III)/shell for wt H/L, and ∼2400 Fe(III)/shell for L135P). The deviation observed at the end of each kinetic curve is due to FeOOH(s) precipitation, and the noise at longer times in the wt H/L and L135P curves is due to light scattering from the formation of FeOOH(s) particles. In a different set of experiments, preparation and then centrifugation of several protein samples (wt H/L and H/L variants) containing between 1500 and 2000 Fe(III)/shell suggested that iron remained soluble in the presence of the protein, otherwise a brown precipitate of ferric hydroxide would have formed at the bottom of the centrifugation tube. In a series of follow up experiments, iron-content analysis of these iron-loaded samples gave values between (1430 ± 50) and (1920 ± 80) Fe(III)/shell, indicating that the majority of the added iron (>95%) is sequestered by the protein (data not shown). Overall, our data suggest that the somewhat faster iron oxidation kinetics observed with the L-chain mutants (Fig. 4 and 6) are due to a disturbance around the E-helix region where the mutations occurred. Molecular surface representation of ferritin subunits forming the 4-fold axis hydrophobic pore showed very narrow channels in the case of the wt H/L protein (due to tightly packed hydrophobic residues), but an enlarged pore size for the variants.12 The altered conformation of the C-terminus and the wider and loose four-fold tetramer assembly in the pathogenic variants is probably disrupting the normal functioning of the protein thus affecting iron permeability. Our data is consistent with earlier studies15,40 showing a reduced capacity of the L-chain variant Ln2 to incorporate iron and a reduced physical stability (for the three pathogenic variants Ln1, Ln2 and Ln3), even though there is only a handful of L-chains present in these variants (i.e. ∼3 L-chain on average per ferritin shell). However, we note that the ability of variants Ln1 and Ln3 to load iron and form a sizeable iron core was as good as the wild type protein, under our experimental conditions, indicating that not all NF variants have a reduced iron loading capacity. Additionally, we observed that the size of the iron core (i.e. total amount of iron added to ferritin before precipitation occurs) depends on the size of the injection (i.e. the ratio of Fe(II)/protein) added at once to the ferritin sample.
image file: c9mt00154a-f4.tif
Fig. 4 Iron oxidation kinetics in ferritin at 305 nm using light absorption spectroscopy. The kinetic curves represent the formation of ferritin Fe(III) core following multiple additions of 300 Fe(II) per protein. Conditions: 0.1 μM ferritin, 30 μM FeSO4, 100 mM Mops, 50 mM NaCl, pH 7.0, 25 °C.

Oxygen electrode kinetics and iron oxidation stoichiometry

To characterize and compare the iron oxidation chemistry between the different ferritin samples, oxygen uptake kinetics were performed using an electrode oximetry apparatus encompassing an OM-4 oxygen meter (Microelectrodes, Inc., Bedford, NH) and an MI-730 micro-oxygen electrode. The oxygen uptake curves for ten consecutive additions of 42 Fe(II) atoms/protein (i.e. ∼2 Fe(II) atoms/H-subunit) to wt H/L ferritin and its H-rich variants, and the corresponding Fe(II)[thin space (1/6-em)]:[thin space (1/6-em)]O2 stoichiometric ratios are shown in Fig. 5. The first injection of 42 Fe(II) atoms/protein shell made to wt H/L, Ln2 and L135P produced an Fe(II)/O2 stoichiometry of 1.9 ± 0.2. When a second injection of 42 Fe(II)/shell is made to the same protein samples (Fig. 5, left panel), the Fe(II)/O2 ratio increased to approximately 2.3 ± 0.2 Fe(II) atoms/O2 and remained at that value up to a total of 420 Fe(II/shell). However, with Ln3 and Ln1, the Fe(II)/O2 ratios for the first injection of 42 Fe(II)/shell are a bit different (i.e. ∼1.7 and ∼2.4 Fe(II)/O2, respectively), suggesting that the frameshift mutations in these variants have a different but significant effect on the stoichiometry of iron oxidation. Furthermore, the Fe(II)/O2 ratios in these variants remained either lower (at 2.1 ± 0.1 for Ln3) or higher (at 2.8 ± 0.1 for Ln1) for subsequent Fe(II) additions, compared to the other samples.
image file: c9mt00154a-f5.tif
Fig. 5 Iron oxidation in ferritin followed by oximetry. Fe(II)/O2 stoichiometry (left panel) and initial rates of oxygen uptake (right panel) as a function of Fe(II) loadings into the proteins for 10 sequential additions of 42 Fe(II) atoms/protein. Conditions: 1 μM protein, 42 μM Fe(II) per addition, 100 mM Mops, 50 mM NaCl, pH 7.0, 25.0 °C. Each experiment was run at least twice with the error bars corresponding to the standard deviation between the trials.

Mechanistically speaking, these observations are similar to previously reported data for recombinant human H-subunit homopolymer ferritin and other ferritins as well4,35 and have been attributed to the involvement of three main pathways, the ferroxidase center oxidation reaction (i.e. 2 Fe(II) atoms/O2), the detoxification reaction (i.e. 2 Fe(II) atoms/H2O2, with the H2O2 being a by-product of the ferroxidase reaction), and the mineral surface reaction (i.e. 4 Fe(II) atoms/O2 at high iron loadings). The details of these oxidation reactions and the measured Fe(II):O2 stoichiometry have been described in details elsewhere,4,41 and most recently in a study from our lab involving heteropolymer H-rich ferritin.35 Although somewhat different, the constant Fe(II)[thin space (1/6-em)]:[thin space (1/6-em)]O2 stoichiometric ratios (i.e. ∼2.1 Fe(II)/O2 for Ln3, ∼2.8 for Ln1 and ∼2.3 for wt H/L, Ln2 and L135P) at higher iron loadings (Fig. 5) suggest a role for the L-subunit in facilitating iron turnover at the ferroxidase center,35 and that the different frameshift mutations affect in different ways the iron oxidation mechanism in these variants.

Similarly, when the initial rates of iron oxidation, as measured by electrode oximetry, is plotted as a function of added Fe(II)/shell, a similar pattern evolves whereby the addition of the first 42 Fe(II)/shell exhibited a higher Fe(II) oxidation rate of (1.9 ± 0.1 μM s−1) (Fig. 5, right panel), corresponding to the rapid and fast oxidation of Fe(II) ions at the ferroxidase centers of all protein samples (except for Ln1 and Ln3). Slower Fe(II) oxidation rates are observed (1.2 ± 0.2 μM s−1) with subsequent Fe(II)/shell additions (Fig. 5). Here again, Ln1 and Ln3 showed different iron oxidation rates (starting at ∼1.0 ± 0.1 μM s−1 for Ln1 and ∼1.6 ± 0.2 μM s−1 for Ln3, and slowly dropping to ∼0.6 ± 0.1 μM s−1 and ∼1.4 ± 0.1 μM s−1, respectively) (Fig. 5). The higher oxidation rates in Ln3 is consistent with a lower Fe(II)/O2 stoichiometry (i.e. 2 Fe(II)/O2 ratio corresponds to a rapid oxidation reaction at the dinuclear center of the H-subunits). It is unclear why an opposite pattern is observed with Ln1 (i.e. lower oxidation rate and higher Fe(II)/O2 ratio); however, since the oximetry experiments follow the uptake of the dissolved oxygen in the bulk of the solution, it is possible that oxygen diffusion through the ferritin shell in Ln1 is hindered by disturbances around the four-fold axes, which in mammals are hydrophobic and thought to be involved in diffusion of O2.4,42,43 Altogether, the oximetry results are consistent with the DSC data and suggest that the different mutations introduce unique structural changes in each variant that affect the ability of the protein to process iron oxidation.

UV-vis spectrophotometry

Iron oxidation kinetics. Spectrophotometric kinetic measurements were conducted to evaluate the ability of H/L ferritin and its variants to oxidize Fe(II) and form a mineral core at low and high iron loadings. Fig. 6 shows an increase in absorbance at 305 nm (due to the formation of oxo/hydroxo Fe(III) species) following multiple iron injections to the same protein sample. When the initial rates of iron oxidation is plotted against the ratio of Fe(II)/protein, a similar trend emerges independent of iron loading (whether 42 Fe(II)/shell, 300 Fe(II)/shell, or 1000 Fe(II)/shell). In all cases, Ln3 showed a faster rate of iron oxidation compared to wt H/L (i.e. ∼2 to 3 times faster at low and intermediate iron loadings), followed by Ln1, Ln2, and L135P, consistent with the location of the frameshift insertion (i.e. the more upstream the frameshift mutation, the faster the iron oxidation kinetics). Overall, the DSC and kinetics results suggest that the integrity of the protein shell is key to its stability and proper iron oxidation kinetics, and that local unfolding around the protein C-terminus alters ferritin functionality.
image file: c9mt00154a-f6.tif
Fig. 6 Fe(II) oxidation kinetics in wt heteropolymer H/L, L135P and the three pathogenic mutants Ln1, Ln2, and Ln3, in the presence of 42, 300 or 1000 Fe(II) atoms/shell (upper panels) and the corresponding initial rates of Fe(II) oxidation (lower panels). All protein solutions (0.5 μM for the low iron loading of 42 Fe(II)/shell and 0.05 μM for the high iron loadings of 300 and 1000 Fe(II)/shell) were prepared in 0.1 M Mops and 50 mM NaCl, pH 7.0 and 25.0 °C. The error bars represent the standard deviation between the various runs (two to three runs per experiment).
Iron release kinetics. To examine the influence of the frame shift mutations on the reductive mobilization of iron from wt and pathogenic ferritins, we studied the kinetics of iron reduction by the FMN–NADH system according to previously established protocols.37–39 In brief, ferritin samples loaded with different amounts of iron (i.e. 84 Fe(III)/protein from 2 injections of 42 Fe(II)/protein, 1200 Fe(III)/protein from 4 injections of 300 Fe(II)/protein, and 1000 Fe(III)/protein from 1 injection of 1000 Fe(II)/protein) were subjected to 2.5 mM NADH, 2.5 mM FMN, and 0.4 mM ferrozine, in 100 mM MOPS, 50 mM NaCl, pH 7.0 and 25 °C and the kinetics of iron release monitored by the absorption of the Fe(II)–ferrozine complex at 562 nm.

Fig. 7 (top panel) shows the kinetic profiles of the iron release experiments for different ferritin samples loaded with different amounts of iron. In all cases, the iron release kinetics exhibited biphasic behavior following an initial lag phase lasting about 5 min during which the kinetics of iron release from ferritin is very slow, as evidenced by the slow and gradual increase in absorption at 562 nm. The initial lag phase was immediately followed by biphasic kinetics featuring an initial rapid phase, followed by a second slower phase (Fig. 7 bottom panel). The duration of the lag phase depends on the concentrations of dissolved oxygen, NADH and FMN present in solution. Most of the reduced flavins (i.e. FMNH2) produced from the reaction between FMN and NADH is rapidly re-oxidized back to FMN by dissolved molecular oxygen without any appreciable release of iron, suggesting that the rate of FMNH2 oxidation by oxygen is much faster than that of iron reduction.37–39

image file: c9mt00154a-f7.tif
Fig. 7 Reductive mobilization of iron from heteropolymer ferritins containing 84, 1000 and 1200 Fe(III)/shell, prepared as per Fig. 6. Conditions: 0.05 μM ferritin for the high iron loading and 0.5 μM ferritin for the low iron loading, 2.5 mM NADH, 2.5 mM FMN, 0.4 mM ferrozine, in 0.1 M Mops and 50 mM NaCl, pH 7.0 and 25.0 °C. (top panel) Absorbance change of [Fe(ferrozine)3]4− as a function of time. (bottom panel) Change of the iron(II)–ferrozine release rate versus time for the different Fe(III)/shell ferritin samples.

In all ferritin samples employed in this experiment, the iron release kinetics achieved a maximal rate in about 0.4–0.5 min (i.e. 24–30 seconds after the end of the lag phase). At high iron loadings (1000 and 1200 Fe(III)/shell), the maximum iron release rate was ∼26–31 μM iron(II) cations per minute for all samples tested; however, at low iron loadings (84 Fe(III)/shell), the iron release rates for Ln1, Ln3 and L135P were 1.5 times higher than wt and Ln2 (i.e. ∼20 μM vs. ∼14 μM Fe(II) ions per minutes, respectively). In all samples tested, iron release was essentially complete in about 10 min after the lag phase. Interestingly, while a similar initial rapid phase of iron release was observed in all ferritin samples, the second slower phase showed significant differences in the iron release rates, with Ln1 and Ln3 exhibiting higher rates of iron mobilization (3 to 5 fold faster compared to wt, Ln2 and L135P, Fig. 7 and 8). Our data do not support the controlled gated-pore hypothesis showing a thirty-fold faster iron release rate for L134P (the frog homologue of human L135P)44 compared to the wild type protein. We note that the gradual decrease in the rates of iron mobilization from ferritin (i.e. the 2nd slow phase of the kinetic) is due to the depletion of the iron(III) hydroxide core, such as the reduction of the ferritin iron core becomes the rate limiting step.37–39

image file: c9mt00154a-f8.tif
Fig. 8 Initial rates of the first and second phases of iron mobilization from heteropolymer ferritins containing 84 and 1200 Fe(III)/shell, prepared as per Fig. 6. Conditions are the same as Fig. 7. The error bars represent the standard deviation between the various runs (two to three runs per experiment).


Neuroferritinopathy is caused by modification of the C-terminal region of L-subunit through nucleotide insertions in the last exon of the ferritin light subunit. The disease is characterized by iron deposition in the brain and ferritin inclusion bodies with a slow progression that leads to late-onset movement disorder including dystonia, chorea, and neurodegeneration.1,45,46 Earlier studies employing mutant Ln2 have shown an enhanced propensity of ferritin L-mutants towards iron-induced oxidative damage and overall iron mishandling,47 and suggested that hereditary ferritinopathy pathogenesis is due to a reduction in iron storage function and an enhanced toxicity associated with iron-induced ferritin aggregates.48,49 Our data show that reduction in iron storage and iron-induced ferritin aggregation are not observed in all NF variants and thus should not be the only hallmarks of neuroferritinopathy; in fact, pathogenic variants Ln1 and Ln3 accumulated as much iron as the wild type and do not have a propensity to aggregate any differently than wt. Our results indicate that disruption of ferritin structure and functionality (such as enhanced iron permeability and lower thermal stability) are additional important factors that contribute to the pathogenesis of the disease. The strongest destabilization effect was observed with Ln3, followed by Ln1 and then Ln2, a result in accord with the location of nucleotide insertions (i.e. the more upstream the insertions, the stronger the effect). Furthermore, our data indicate that alteration of the hydrophilic 3-fold channels, as in variant L135P, does not exhibit a major negative-dominant effect, and thus may not be responsible for neuroferritinopathy-like disorders. Unlike NF variants, the lack of such effect in L135P may be due to the fact that mutations of a few L-subunits encompassing the eight 3-fold channels have only minor effects on the protein's functionality. Taken together, we propose that the pathogenesis of neuroferritinopathy is the result of a structural and functional impairment of ferritin which result in iron mis-management that could lead to iron-induced oxidative damage, and that protein aggregation may be a secondary phenomenon and not a causative effect.

Conflicts of interest

There are no conflicts to declare.


This work is supported by the National Institute of Health, Award Number R15GM104879 (F. B. A.) and by The Camille & Henry Dreyfus Foundation, Inc., The Henry Dreyfus Teacher-Scholar Award (F. B. A., award TH-16-007). It is also partially supported by Grant PRIN10-11 of MIUR, from the Italian Ministry of Research (P. A.). J. R. M. and M. R. M. were partly supported by Kilmer Undergraduate Research Apprenticeships (SUNY Potsdam).


  1. E. C. Theil, T. Tosha and R. K. Behera, Solving biology's iron chemistry problem with ferritin protein nanocages, Acc. Chem. Res., 2016, 49, 784–791 CrossRef CAS PubMed.
  2. N. D. Chasteen and P. M. Harrison, Mineralization in ferritin: an efficient means of iron storage, J. Struct. Biol., 1999, 126, 182–194 CrossRef CAS PubMed.
  3. P. Arosio, R. Ingrassia and P. Cavadini, Ferritins: a family of molecules for iron storage, antioxidation and more, Biochim. Biophys. Acta, 2009, 1790(7), 589–599 CrossRef CAS PubMed.
  4. F. Bou-Abdallah, The iron redox and hydrolysis chemistry of the ferritins, Biochim. Biophys. Acta, 2010, 1800, 719–731 CrossRef CAS PubMed.
  5. P. M. Harrison and P. Arosio, Ferritins: molecular properties, iron storage function and cellular regulation, Biochim. Biophys. Acta, 1996, 1275, 161–203 CrossRef.
  6. U. Carmona, L. Li, L. Zhang and M. Knez, Ferritin light-chain subunits: Key elements for the electron transfer across the protein cage, Chem. Commun., 2014, 50, 15358–15361 RSC.
  7. S. Levi, P. Santambrogio, A. Cozzi, E. Rovida, B. Corsi, E. Tamborini, A. Spada, A. Albertini and P. Arosio, The role of the L-chain in ferritin iron incorporation. Studies of homo and heteropolymers, J. Mol. Biol., 1994, 38, 649–654 CrossRef PubMed.
  8. P. Santambrogio, S. Levi, A. Cozzi, E. Rovida, A. Albertini and P. Arosio, Production and characterization of recombinant heteropolymers of human ferritin H and L chains, J. Biol. Chem., 1993, 268, 12744–12748 CAS.
  9. J. W. Drysdale, T. G. Adelman, P. Arosio, D. Casareale, P. Fitzpatrick, J. T. Harzard and M. Yokota, Human isoferritins in normal and disease states, Semin. Hematol., 1977, 14, 71–88 CAS.
  10. F. Bou-Abdallah, P. Santambrogio, S. Levi, P. Arosio and N. D. Chasteen, Unique iron binding and oxidation properties of human mitochondrial ferritin: A comparative analysis with human H-chain ferritin, J. Mol. Biol., 2006, 347, 543–554 CrossRef PubMed.
  11. W. Wang, M. A. Knovich, L. G. Coffman, F. M. Torti and S. V. Torti, Serum ferritin: Past, present and future, Biochim. Biophys. Acta, 2010, 1800, 760–769 CrossRef CAS.
  12. S. Levi and E. Rovida, Neuroferritinopathy: From ferritin structure modification to pathogenetic mechanism, Neurobiol. Dis., 2015, 81, 134–143 CrossRef CAS PubMed.
  13. A. R. Curtis, C. Fey, C. M. Morris, L. A. Bindoff, P. G. Ince, P. F. Chinnery, A. Coulthard, M. J. Jackson, A. P. Jackson, D. P. McHale, D. Hay, W. A. Barker, A. F. Markham, D. Bates, A. Curtis and J. Burn, Mutation in the gene encoding ferritin light polypeptide causes dominant adult-onset basal ganglia disease, Nat. Genet., 2001, 28, 350–354 CrossRef CAS PubMed.
  14. P. Dusek and S. A. Schneider, Neurodegeneration with brain iron accumulation, Curr. Opin. Neurol., 2012, 25, 499–506 CrossRef CAS PubMed.
  15. S. Luscieti, P. Santambrogio, B. Langlois d'Estaintot, T. Granier, A. Cozzi, M. Poli, B. Gallois, D. Finazzi, A. Cattaneo, S. Levi and P. Arosio, Mutant ferritin L-chains that cause neurodegeneration act in a dominant-negative manner to reduce ferritin iron incorporation, J. Biol. Chem., 2010, 285, 11948–11957 CrossRef CAS PubMed.
  16. S. Levi and D. Finazz, Neurodegeneration with brain iron accumulation: update on pathogenic mechanisms, Front. Pharmacol., 2014, 5(99), 1–20 CAS.
  17. R. Vidal, M. B. Delisle, O. Rascol and B. Ghetti, Hereditary ferritinopathy, J. Neurol. Sci., 2003, 207, 110–111 CrossRef PubMed.
  18. M. Mancuso, G. Davidzon, R. M. Kurlan, R. Tawil, E. Bonilla, S. Di Mauro and J. M. Powers, Hereditary ferritinopathy: a novel mutation, its cellular pathology, and pathogenetic insights, J. Neuropathol. Exp. Neurol., 2005, 64, 280–294 CrossRef CAS PubMed.
  19. E. Ohta, T. Nagasaka, K. Shindo, S. Toma, K. Nagasaka, K. Ohta and Z. Shiozawa, Neuroferritinopathy in a Japanese family with a duplication in the ferritin light chain gene, Neurology, 2008, 70, 1493–1494 CrossRef CAS.
  20. A. Kubota, A. Hida, Y. Ichikawa, Y. Momose, J. Goto, Y. Igeta, H. Hashida, K. Yoshida, S. Ikeda, I. Kanazawa and S. Tsuji, A novel ferritin light chain gene mutation in a Japanese family with neuroferritinopathy: description of clinical features and implications for genotype-phenotype correlations, Mov. Disord., 2009, 24, 441–445 CrossRef PubMed.
  21. D. Devos, P. J. Tchofo, I. Vuillaume, A. Destee, S. Batey, J. Burn and P. F. Chinnery, Clinical features and natural history of neuroferritinopathy caused by the 458dupA FTL mutation, Brain, 2009, 132, e109 CrossRef PubMed.
  22. S. Moutton, P. Fergelot, J. M. Trocello, V. Plante-Bordeneuve, N. Houcinat, E. Wenisch, V. Larue, P. Brugieres, F. Clot, D. Lacombe, B. Arveiler and C. Goizet, A novel FTL mutation responsible for neuroferritinopathy with asymmetric clinical features and brain anomalies, Parkinsonism Relat. Disord., 2014, 20, 935–937 CrossRef PubMed.
  23. K. Nishida, H. J. Garringer, N. Futamura, I. Funakawa, K. Jinnai, R. Vidal and M. Takao, A novel ferritin light chain mutation in neuroferritinopathy with an atypical presentation, J. Neurol. Sci., 2014, 342, 173–177 CrossRef CAS PubMed.
  24. E. Storti, F. Cortese, R. Di Fabio, C. Fiorillo, A. Pierallini, A. Tessa, A. Valleriani, F. Pierelli, F. M. Santorelli and C. Casali, De novo FTL mutation: a clinical, neuroimaging, and molecular study, Mov. Disord., 2013, 28, 252–253 CrossRef PubMed.
  25. P. Maciel, V. T. Cruz, M. Constante, I. Iniesta, M. C. Costa, S. Gallati, N. Sousa, J. Sequeiros, P. Coutinho and M. M. Santos, Neuroferritinopathy: missense mutation in FTL causing early-onset bilateral pallidal involvement, Neurology, 2005, 65, 603–605 CrossRef CAS PubMed.
  26. G. Gerardi, G. Biasiotto, P. Santambrogio, I. Zanella, R. Ingrassia, M. Corrado, P. Cavadini, M. Derosas, S. Levi and P. Arosio, Recombinant human hepcidin expressed in Escherichia coli isolates as an iron containing protein, Blood Cells, Mol., Dis., 2005, 35, 177–181 CrossRef CAS PubMed.
  27. R. Ingrassia, G. Gerardi, G. Biasiotto and P. Arosio, Mutations of Ferritin H Chain C-Terminus Produced by Nucleotide Insertions Have Altered Stability and Functional Properties, J. Biochem., 2006, 139, 881–885 CrossRef CAS PubMed.
  28. R. Jappelli, A. Luzzago, P. Tataseo, I. Pernice and G. Cesareni, Loop Mutations can cause a substantial conformational change in the carboxy terminus of the ferritin protein, J. Mol. Biol., 1992, 227, 532–543 CrossRef CAS PubMed.
  29. A. Luzzago and G. Cesareni, Isolation of point mutations that affect the folding of the H chain of human ferritin in E. coli, EMBO J., 1989, 8, 569–576 CrossRef CAS PubMed.
  30. S. Levi, A. Luzzago, G. Cesareni, A. Cozzi, F. Franceschinelli, A. Albertini and P. Arosio, Mechanism of ferritin iron uptake: activity of the H-chain and deletion mapping of the ferro-oxidase site. A study of iron uptake and ferro-oxidase activity of human liver, recombinant H-chain ferritins, and of two H-chain deletion mutants, J. Biol. Chem., 1988, 263, 18086–18092 CAS.
  31. S. Levi, A. Luzzago, F. Franceschinelli, P. Santambrogio, G. Cesareni and P. Arosio, Mutational analysis of the channel and loop sequences of human ferritin H-chain, Biochem. J., 1989, 264, 381–388 CrossRef CAS PubMed.
  32. C. Brito, C. Matias, F. D. Gonzalez-Nilo, R. K. Watt and A. Yevenes, The C-Terminal Regions Have an Important Role in the Activity of the Ferroxidase Center and the Stability of Chlorobium tepidum Ferritin, Protein J., 2014, 33(3), 211–220 CrossRef CAS PubMed.
  33. X. S. Liu, L. D. Patterson, M. J. Miller and E. C. Theil, Peptides Selected for the Protein Nanocage Pores Change the Rate of Iron Recovery from the Ferritin Mineral, J. Biol. Chem., 2007, 282(44), 31821–31825 CrossRef CAS PubMed.
  34. P. Santambrogio, A. Cozzi, S. Levi, E. Rovida, F. Magni, A. Albertini and P. Arosio, Functional and immunological analysis of recombinant mouse H- and L-ferritins from Escherichia coli, Protein Expression Purif., 2000, 9, 212–218 CrossRef PubMed.
  35. M. Mehlenbacher, M. Poli, P. Arosio, P. Santambrogio, S. Levi, N. D. Chasteen and F. Bou-Abdallah, Iron oxidation and core formation in recombinant heteropolymeric human ferritins, Biochemistry, 2017, 56, 3900–3912 CrossRef CAS PubMed.
  36. X. Yang, Y. Chen-Barrett, P. Arosio and N. D. Chasteen, Reaction paths of iron oxidation and hydrolysis in horse spleen and recombinant human ferritins, Biochemistry, 1998, 137, 9743–9750 CrossRef PubMed.
  37. G. Melman, F. Bou-Abdallah, E. Vane, P. Maura, P. Arosio and A. Melman, Iron release from ferritin by Flavin nucleotides, Biochim. Biophys. Acta, 2013, 1830, 4669–4674 CrossRef CAS PubMed.
  38. L. E. Johnson, T. Wilkinson, P. Arosio, A. Melman and F. Bou-Abdallah, Effect of chaotropes on the kinetics of iron release from ferritin by flavin nucleotides, Biochim. Biophys. Acta, 2017, 1861, 3257–3262 CrossRef CAS PubMed.
  39. F. Bou-Abdallah, J. J. Paliakkara, G. Melman and A. Melman, Reductive Mobilization of Iron from Intact Ferritin: Mechanisms and Physiological Implication, Pharmaceuticals, 2018, 11, 120,  DOI:10.3390/ph11040120.
  40. M. A. Baraibar, B. B. Muhoberac, H. J. Garringer, T. D. Hurley and R. Vidal, Unraveling of the E-helices and disruption of 4-fold pores are associated with iron mishandling in a mutant ferritin causing neurodegeneration, J. Biol. Chem., 2010, 285, 1950–1956 CrossRef CAS PubMed.
  41. G. Zhao, F. Bou-Abdallah, P. Arosio, S. Levi, C. Janus-Chandler and N. D. Chasteen, Multiple Pathways for Mineral Core Formation in Mammalian Apoferritin. The Role of Hydrogen Peroxide, Biochemistry, 2003, 42, 3142–3150 CrossRef CAS PubMed.
  42. T. Douglas and D. R. Ripoli, Calculated electrostatic gradients in recombinant human. H-chain ferritin, Protein Sci., 1998, 7, 1083–1109 CrossRef CAS PubMed.
  43. X. Liu and E. C. Theil, Ferritins: Dynamic management of biological iron and oxygen chemistry, Acc. Chem. Res., 2005, 38, 167–175 CrossRef CAS PubMed.
  44. E. C. Theil, X. S. Liu and T. Tosha, Gated pores in the ferritin protein nanocage, Inorg. Chim. Acta, 2008, 361, 868–874 CrossRef CAS PubMed.
  45. S. Capoccia, F. Maccarinelli, B. Buffoli, L. F. Rodella, O. Cremona, P. Arosio and F. Cirulli, Behavioral Characterization of Mouse Models of Neuroferritinopathy, PLoS One, 2015, 10(2), e0118990,  DOI:10.1371/journal.pone.0118990.
  46. A. Cozzi, P. Santambrogio, B. Corsi, A. Campanella, P. Arosio and S. Levi, Characterization of the L-ferritin variant 460InsA responsible of a hereditary ferritinopathy disorder, Neurobiol. Dis., 2006, 23, 644–652 CrossRef CAS PubMed.
  47. M. A. Baraibar, A. G. Barbeitoa, B. B. Muhoberac and R. Vidal, A mutant light-chain ferritin that causes neurodegeneration has enhanced propensity toward oxidative damage, Free Radical Biol. Med., 2012, 52, 1692–1697 CrossRef CAS PubMed.
  48. M. A. Baraibar, A. G. Barbeitoa, B. B. Muhoberac and R. Vidal, Iron-mediated Aggregation and a Localized Structural Change Characterize Ferritin from a Mutant Light Chain Polypeptide That Causes Neurodegeneration, J. Biol. Chem., 2008, 283(46), 31679–31689 CrossRef CAS PubMed.
  49. B. B. Muhoberac, M. A. Baraibar and R. Vidal, Iron loading-induced aggregation and reduction of iron incorporation in heteropolymeric ferritin containing a mutant light chain that causes neurodegeneration, Biochim. Biophys. Acta, 2011, 1812, 544–548 CrossRef CAS PubMed.

This journal is © The Royal Society of Chemistry 2019