Frataxin-mediated Iron Delivery to Ferrochelatase in the Final Step of Heme Biosynthesis*

Human ferrochelatase, a mitochondrial membrane-associated protein, catalyzes the terminal step of heme biosynthesis by insertion of ferrous iron into protoporphyrin IX. The recently solved x-ray structure of human ferrochelatase identifies a potential binding site for an iron donor protein on the matrix side of the homodimer. Herein we demonstrate Hs holofrataxin to be a high affinity iron binding partner for Hs ferrochelatase that is capable of both delivering iron to ferrochelatase and mediating the terminal step in mitochondrial heme biosynthesis. A general regulatory mechanism for mitochondrial iron metabolism is described that defines frataxin involvement in both heme and iron-sulfur cluster biosyntheses. In essence, the distinct binding affinities of holofrataxin to the target proteins, ferrochelatase (heme synthesis) and ISU (iron-sulfur cluster synthesis), allows discrimination between the two major iron-dependent pathways and facilitates targeted heme biosynthesis following down-regulation of frataxin.

Human ferrochelatase, a mitochondrial membrane-associated protein, catalyzes the terminal step of heme biosynthesis by insertion of ferrous iron into protoporphyrin IX. The recently solved x-ray structure of human ferrochelatase identifies a potential binding site for an iron donor protein on the matrix side of the homodimer. Herein we demonstrate Hs holofrataxin to be a high affinity iron binding partner for Hs ferrochelatase that is capable of both delivering iron to ferrochelatase and mediating the terminal step in mitochondrial heme biosynthesis. A general regulatory mechanism for mitochondrial iron metabolism is described that defines frataxin involvement in both heme and iron-sulfur cluster biosyntheses. In essence, the distinct binding affinities of holofrataxin to the target proteins, ferrochelatase (heme synthesis) and ISU (iron-sulfur cluster synthesis), allows discrimination between the two major iron-dependent pathways and facilitates targeted heme biosynthesis following down-regulation of frataxin.
Frataxin is a nuclear-encoded protein that is targeted to the mitochondrial matrix. Reduced frataxin expression, a causative agent of the neurological disorder Friedreich ataxia, results in mitochondrial iron accumulation. Recent evidence has pointed to a functional role for frataxin in mitochondrial iron metabolism, including iron-sulfur cluster (1)(2)(3)(4)(5) and heme (6 -8) biosynthesis. We have reported earlier that frataxin serves as an iron donor to ISU, the iron-sulfur cluster scaffold protein (1). Isothermal titration calorimetry and fluorescence quenching experiments demonstrated human frataxin to bind 6 or 7 iron ions with K D ϳ10 -50 M for the isolated protein (1). Holofrataxin was further shown to bind to ISU with a K D ϳ0.15 M, and the functional viability of frataxin as an iron donor for assembly of the [2Fe-2S] cluster of ISU in the presence of a sulfur donor was demonstrated through kinetic and spectroscopic studies (1). Iron release by frataxin appeared to be the rate-limiting step. Overall these results correlate well with other published observations concerning a possible role for frataxin in iron-sulfur cluster biosynthesis (2)(3)(4)(5).
To further characterize potential roles for frataxin as a mitochondrial iron donor, we have investigated the involvement of Hs frataxin in cellular heme biosynthesis as an iron donor to Hs ferrochelatase. Although the identity of the iron donor protein in heme biosynthesis has not been established, involvement by frataxin has been suggested on the basis of yeast studies that demonstrated mitochondrial iron to be unavailable for heme biosynthesis in cells lacking frataxin (6 -8). Dancis and co-workers (6) have recently reported genetics experiments that implicate the involvement of yeast frataxin in heme biosynthesis and have estimated a binding affinity (K D ) for frataxin to ferrochelatase of ϳ40 nM by surface plasmon resonance, although no evidence for frataxin-mediated iron delivery in heme biosynthesis was presented. Herein we characterize the interaction of human frataxin and ferrochelatase and demonstrate holofrataxin to serve as a potential iron donor to ferrochelatase for insertion into the protoporphyrin ring during heme synthesis.

EXPERIMENTAL PROCEDURES
Protein Production and Purification-Human ferrochelatase was expressed and purified as described previously (9). Human frataxin was also obtained as described previously (1), with the following modifications. The expressed protein was purified by nickel-nitrilotriacetic acid chromatography as described (1), and the mature truncated form of frataxin, encompassing residues 81-210, was obtained by an autocleavage reaction following incubation at 4°C for 2 weeks (as observed previously (10 -12)) and subsequent fast protein liquid chromatography gel filtration using 50 mM Tris (pH 7.5), 100 mM NaCl. The purity of the truncated protein was confirmed by SDS-PAGE, which demonstrated a single protein band with a molecular mass of ϳ14 kDa, consistent with electrospray ionization mass spectrometric analysis that yielded a mass of 14,661 Da. Samples were either used immediately following purification or stored in their final elution buffers at Ϫ80°C. Holofrataxin was prepared under strictly anaerobic conditions as described previously (9) with ferrous iron in 50 mM HEPES, 100 mM KCl, pH 8.0 buffer.
ITC Measurements of Binding-ITC 1 measurements were carried out at 25°C using a MicroCal Omega ultrasensitive titration calorimeter. The sample solutions were made from the stock buffer solution (50 mM HEPES, 100 mM KCl, 1% sodium cholate, pH 8), and the titrant solution was made up in 50 mM HEPES, 100 mM KCl, pH 8. Both experimental solutions were argon-purged and then thoroughly degassed before each titration. The solution in the cell (an 0.06 mM solution of ferrochelatase) was stirred at 300 rpm by syringe to ensure rapid mixing. Typically, from 3 to 7 l of titrant (a 1.3 mM solution of holofrataxin) was delivered over 10 s with an adequate interval (8 min) between injections to allow complete equilibration. Excess titrant was added to ensure saturation binding. At the end of the experiment, added iron was found to remain in the ferrous state. A background titration, consisting of the identical titrant solution but with only the buffer solution in the sample cell, was subtracted from each experimental titration to account for heat of dilution. The data were collected automatically and subsequently analyzed with a one-site binding model by the Windows-based Origin software package supplied by MicroCal. Origin software uses a nonlinear least-squares algorithm (minimization of 2 ) and the concentrations of the titrant and the sample to fit the heat flow per injection to an equilibrium binding equation, providing best fit values of the stoichiometry (n), change in enthalpy (⌬H), and binding constant (K). A set of titration experiments was carried out for apofrataxin binding to apoISU using similar buffer conditions (50 mM HEPES, 100 mM KCl, 1% sodium cholate, pH 8), with 30 M ISU (the D37A mutant was used as described previously (1,9) but behaves similarly to native ISU) in the reaction cell and titration with 10-l aliquots of 0.4 mM holofrataxin.
Stoichiometric Titration by Fluorescence Quenching-The tryptophan fluorescence of human ferrochelatase solution was measured in * This work was supported by Grant CHE-0111161 from the National Science Foundation (to J. A. C.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1-ml quartz cuvettes at room temperature with a Luminescence Spectrometer LS50B (PerkinElmer Life Sciences). The excitation and monitoring wavelengths were 280 and 341 nm, respectively. Corrections for the inner filter effect were considered but were negligible as a result of the weak absorbance at the emission wavelength. Intensity data from a control titration, consisting of a similar titrant solution but containing only the buffer solution in the sample cell, were subtracted from the experimental data obtained from the ferrochelatase titration to account for the minimal background fluorescence from the titrant solution. For holofrataxin binding, spectra were obtained on a 6 M solution of ferrochelatase before and after titration with a 0.15 mM stock solution of holofrataxin. Both proteins were prepared in argon-purged 50 mM Tris-HCl, pH 8.0, 100 mM KCl, 1% sodium cholate. All experiments were performed under strictly anaerobic conditions. Kinetics Measurements of Ferrochelatase Activity-The disodium salt of protoporphyrin IX was obtained from Aldrich. A protoporphyrin stock solution was prepared in 2 M NH 4 OH (0.1 ml) followed by the addition of 1.9 ml of a buffered solution containing 50 mM Tris-HCl, 100 mM KCl, 1% sodium cholate, pH 8.0. Stock solutions were diluted with the same Tris-HCl buffer to make a desirable concentration of protoporphyrin before each experiment. Ferrochelatase activity was measured anaerobically at 20°C by monitoring the rate of disappearance of the protoporphyrin IX band at 506 nm. A 2-ml solution containing 11.8 M protoporphyrin, 200 nM ferrochelatase, 2.5 mM dithiothreitol, and various concentrations of apofrataxin in 50 mM Tris-HCl, 100 mM KCl, 1% sodium cholate, pH 8.0, was degassed and argon-purged for at least 20 min. To trigger the activity of ferrochelatase, buffered solutions of ferrous ion and citrate were added by injection to the reaction mixture in the cuvette to achieve a final ferrous concentration of 5.8 M and a citrate concentration of 2 mM.

RESULTS
To demonstrate the potential involvement of frataxin in heme biosynthesis, we have quantitatively investigated complex formation between human holofrataxin and human ferrochelatase. The cloned frataxin gene incorporated a His 6 tag and residues following the second mitochondrial processing peptidase cleavage site (11) (residues 47-210 were included). Cloned His 6 -tagged frataxin was expressed and purified as described (1); however, over a period of several days this autocleaves to a truncated protein (residues 81-210) that lacks the N-terminal His 6 tag. The formation of a truncated protein has been documented previously (10 -12) and is the form of the protein that has been structurally characterized in the apo form (12). Earlier studies have demonstrated this truncated human holofrataxin to be monomeric (1).
Isothermal titration calorimetry provided a measured K D ϳ17 nM, with binding both enthalpically (⌬H Ӎ Ϫ3.7 kcal/mol) and entropically (⌬S Ӎ 23 cal/K⅐mol) favorable (Fig. 1A), and parameters are defined per mol of either frataxin or ferrochelatase dimer. No detectable binding response was measured in the absence of iron ion, consistent with observations made in binding studies of apofrataxin and apoISU (1). For the latter study it was deduced that iron ion might mediate a bridging contact between the two proteins rather than a structural transition, because iron binding does not influence frataxin structure. Certainly the proposed frataxin (13) and ferrochelatase (14,15) binding sites show a number of acidic residues that might coordinate to bridging iron ions. High affinity binding of holofrataxin and ferrochelatase was also confirmed by fluorescence quenching experiments (Fig. 1C). Although binding was too tight for accurate quantitation, both data sets are consistent with the nM affinity of one frataxin to a ferrochelatase dimer.
Holofrataxin was previously shown to bind to the iron-sulfur scaffold protein, ISU, with a K D ϳ0.15 M. To allow a comparison with holofrataxin binding to ferrochelatase, the former was determined under similar solution conditions to that reported here for frataxin binding to ferrochelatase, including the presence of 1% sodium cholate. Under similar solution conditions to those used for ferrochelatase binding (Fig. 1A), ITC experiments provided a measured K D ϳ0.48 M for frataxin binding to human apoISU, with a binding stoichiometry of 0.95, ⌬H Ӎ Ϫ2.1 kcal/mol, and ⌬S Ӎ 22 cal/K⅐mol (Fig. 1B).
To understand the role of frataxin in heme biosynthesis, we performed additional experiments to measure ferrochelatase activity as a function of frataxin concentration at fixed ferrous ion concentration. Ferrochelatase activity was measured by following the absorbance change at 506 nm that results from heme production. Activity was found to increase with frataxin concentration and was optimal at a ratio of 1 frataxin molecule/ ferrochelatase dimer (Fig. 2). These results support the hypothesis that frataxin recruits iron ion and mediates delivery to ferrochelatase. Free ferrous iron can also be delivered to ferrochelatase by nonspecific collision; however, free iron is not a bioavailable species. The sharply defined optimal frataxin/ferrochelatase ratio reflects the high affinity interaction and stoichiometric complex formation.
The observed decrease in ferrochelatase activity with increasing frataxin concentration may be explained in several ways. Excess frataxin might bind elsewhere on ferrochelatase and inhibit its function. However, our previous ITC and fluorescence experiments show no evidence for binding of additional frataxin molecules in the concentration range used. A more likely explanation stems from the iron binding properties of frataxin that removes "free" iron from solution, lowering the available iron concentration and inhibiting ferrochelatase activity.

DISCUSSION
The three-dimensional structure of human ferrochelatase (14,15) demonstrates the active form of human ferrochelatase to be a homodimer, similar to that for Saccharomyces cerevisiae ferrochelatase (16). These studies show that the binding and release pathways for protoporphyrin IX and heme, respectively, are located on the inner membrane side of a homodimeric form of ferrochelatase. However, the putative ironbinding site is found on the opposite side of the protein, which is exposed to the mitochondrial matrix (15), and forms an appropriate domain for docking of an iron donor protein. The reaction stoichiometry observed by ITC and fluorescence quenching experiments is consistent with this hypothesis. The possibility that iron delivery from holofrataxin to ferrochelatase occurs via a specific intermolecular interaction was discussed, because heme levels from ferrochelatase activity did not significantly change with excess of the iron chelator, citrate. These results are consistent with findings for copper delivery from the carrier protein (Atx1) to a target protein (Ccc2) (17), a process that was not influenced by the presence of the copper chelator, glutathione.
Finally, although deletion of frataxin does not eliminate cluster or heme synthesis (low molecular weight cellular iron species or other donor proteins might serve) (18), the evidence reported here does support a direct role for frataxin in the biosynthesis of hemes. It is significant that holofrataxin functions as an iron donor in both heme and iron-sulfur cluster biosynthesis, since this allows for a simple control mechanism for utilization of holofrataxin. Thus, in the case of a hemedeficient condition, holofrataxin should be used mainly for heme synthesis, although in the case of iron-sulfur cluster deficiency the majority of holofrataxin should be utilized for iron-sulfur cluster biosynthesis. Data from a recently published study of erythroid differentiation (7), which requires elevated levels of heme production, show the levels of frataxin mRNA and protein product to be down-regulated. Similar observations were made in the case of Friend cells at the same developmental stage with increasing concentration of protoporphyrin (7). The published data can be explained in terms of a general model for iron utilization in mitochondria (Fig. 3). The binding affinity between holofrataxin and ferrochelatase (K D ϳ17 nM) is ϳ28 times greater than the affinity between holofrataxin and ISU (K D ϳ0.48 M), and so reduced frataxin levels will have a Top, time course for the disappearance of protoporphyrin IX absorbance in the ferrochelatase catalyzed reaction at various concentrations of frataxin. Concentrations of frataxin for each data set are as follows: 1, 0 nM; 2, 50 nM; 3, 100 nM; 4, 200 nM; 5, 400 nM; 6, 600 nM; 7, 800 nM; 8, 1000 nM. The control reaction was carried out as described for the other solutions, but the solution lacked both frataxin and iron. Bottom, dependence of ferrochelatase activity on frataxin concentration. Each data point represents the average of five independent experiments. more significant impact on iron-sulfur cluster biosynthesis. Although the complete details of the molecular mechanism of iron regulation in mitochondria remain unknown, Fig. 3 shows a plausible model from the perspective of iron utilization. is used as an iron donor for both heme and iron-sulfur cluster biosynthetic pathways. Under normal cell growth conditions the frataxin concentration is sufficient for both heme and iron-sulfur cluster syntheses. The level of frataxin is down-regulated in erythroid differentiation, as is the iron-sulfur cluster biosynthesis pathway. However, heme biosynthesis remains essentially normal as a consequence of the distinct binding affinities of frataxin to ISU and ferrochelatase.