Humans possess two mitochondrial ferredoxins, Fdx1 and Fdx2, with distinct roles in steroidogenesis, heme, and Fe/S cluster biosynthesis (V体育2025版)

Fdx2 Is a Mitochondrial [2Fe-2S] Protein.

To examine whether Fdx2 is expressed in human cells and imported into mitochondria, we raised antibodies against recombinant, His-tagged Fdx2. We then analyzed subcellular fractions generated from human HeLa cells by immunoblotting VSports app下载. A protein cofractionating with a mitochondrial marker (α/β subunits of F1-ATPase) and with an apparent molecular mass of 22 kDa was recognized by our anti-Fdx2 antibody (Fig.  1A).

To compare the relative expression levels of the two ferredoxins in HeLa cells, we used a dilution series of recombinant purified proteins for immunoblotting. Fdx1 and Fdx2 were present in HeLa cells at 90 and 40 ng/mg cell protein, respectively. We then analyzed the relative abundance of these proteins in various human tissues. Fdx1 protein was undetectable in most tissue lysates, but was highly expressed in the adrenal gland, and present in kidney and testes (Fig. S1). This expression pattern fits perfectly with the function of Fdx1 in steroidogenesis, but is inconsistent with its predicted function in Fe/S protein biogenesis. In contrast, Fdx2 was detectable in virtually all tissues analyzed with the highest amounts present in testis, kidney, and brain.

We next asked whether Fdx2 is an Fe/S protein. Recombinant Fdx2 purified from Escherichia coli was red-brown in color and exhibited a UV-Vis spectrum characteristic of a [2Fe-2S] protein and remarkably similar to that of purified Fdx1, with an undulating, reduction-sensitive absorption shoulder between 400 and 500 nm (21) (Fig. 1B). The presence of a [2Fe-2S] cluster on Fdx2 was supported by EPR spectroscopy of dithionite-reduced protein (Fig. 1C). The axial EPR spectra of Fdx1 and Fdx2 had similar g values (2.02, 1.94) and were both detectable up to 80 K. We also determined the reduction potentials of Fdx1 and Fdx2 to be -267 ± 5 mV and -342 ± 5 mV, respectively (22). In summary, these data indicate that Fdx2 is a mitochondrial [2Fe-2S] protein ubiquitously expressed in human tissues.

Fdx2, and Not Fdx1, Is a Member of the Mitochondrial Fe/S Protein Assembly Machinery.

To investigate the function of Fdx2, the protein was depleted in HeLa cells using RNAi technology. To this end, cells were transfected with a pool of modified siRNAs that targeted three distinct regions of the mRNA. The transfections were repeated thrice, once every 3 d, and a final harvest of the cells was performed 3 d after the third transfection (i.e., day 9; cf. ref. 23). Fdx1 was analyzed in parallel. Immunoblotting against Fdx1 or Fdx2 verified an extensive depletion of the individual proteins over the course of the experiment (Fig. 2A).

Depletion of Fdx2 by RNAi caused a strong decrease in cell proliferation assigning an important function to Fdx2 in HeLa cells (Fig. 2B). In contrast, no significant effect was observed upon Fdx1 depletion. We next performed a series of enzyme assays to evaluate the results of Fdx1 or Fdx2 knockdowns on the activities of several Fe/S proteins. As depicted in Fig. 2C, the catalytic activities of mitochondrial and cytosolic aconitases, complex I, COX, and succinate dehydrogenase (SDH) relative to control proteins were all severely (65–90%) decreased upon Fdx2 depletion. This effect escalated with each subsequent transfection. In contrast, Fdx1 depletion had no effect on either aconitase isoform or complex I activity, and elicited a relatively weak (10–30%) diminishment of COX and SDH activities. More cycles of transfection did not augment these effects of Fdx1 deficiency. By transfecting HeLa cells separately with the individual siRNAs against Fdx1, we found that all three individual siRNA sequences elicited the same response as the siRNA pool, suggesting that the slightly dimished activities of COX and SDH were indeed the result of lowered Fdx1 levels. Mitochondria of Fdx2-deficient, but not of Fdx1-deficient cells exhibited a substantial loss of cristae membranes (Fig. S2), consistent with the severe effects on respiratory chain function upon Fdx2 depletion (23, 24).

Because Fdx1 and Fdx2 are highly homologous, we investigated whether they may have coinciding functions by testing if Fdx1 can replace Fdx2 function. Cells were thrice cotransfected with the Fdx2 siRNAs along with a vector encoding EGFP or the fusion proteins Fdx1-GFP (pFDX1) and Fdx2-GFP (psmFDX2). Silent mutations (sm) were introduced into the latter construct to avoid RNA interference. The strong effects of Fdx2 depletion on the two aconitases, SDH, COX, complex I, and mitochondrial ultrastructure were almost fully restored by cotransfection with psmFDX2 indicating the specificity of the RNAi approach (Fig. 2D, Fig. S2, and Fig. S3). No restoration of the enzyme activities was seen upon overexpression of Fdx1, indicating that Fdx1 cannot generally assume the role of Fdx2 in Fe/S protein biogenesis.

Estimating the protein amounts of several Fe/S proteins is also informative with respect to a cell’s capacity to form the metallo-cofactor (25, 26). In line with the aforementioned results, the levels of mitochondrial aconitase, cytosolic aconitase (IRP1) and SDH were all decreased under Fdx2 depletion (Fig. 2E). Again, this defect was restored by cotransfection with psmFDX2, but not with pFDX1. In addition, an integral subunit (NDUFS3) of complex I, which harbors eight Fe/S clusters, ferrochelatase (FC; one [2Fe-2S] cluster) and glutamate phosphoribosylpyrophosphate amidotransferase (GPAT; one [4Fe-4S] cluster) were all specifically diminished by Fdx2 depletion relative to tubulin (loading control; Fig. 2E). As an additional strategy to detect alterations in respiratory complexes, we measured the radiolabelling of these complexes in cells that were treated with ⁵⁵Fe-transferrin for the 3 d following the third transfection. Analysis by blue native-polyacrylamide gel electrophoresis (BN-AGE) and autoradiography (24, 27) revealed that Fdx2 deficiency led to decreased ⁵⁵Fe incorporation into complexes I and II, which was restored by expression from plasmid psmFDX2 but not pFDX1 (Fig. 2F). The lack of an effect on complex III may be explained by dissociation of the supercomplex of complexes I, III, and COX upon diminishment of complex I (24) and the fact that half of the iron present in CIII is in heme. Taken together, the data presented in Fig. 2 strongly suggest a major role of Fdx2, but not of Fdx1, in Fe/S protein biogenesis.

V体育官网入口 - Human Fdx2 Can Perform the Biochemical Functions of Yeast Yah1.

We asked whether human Fdx2 can functionally replace yeast Yah1 in its tasks in Fe/S protein and heme A biosynthesis. We used a yeast strain (Gal-YAH1), in which a GAL1-10 promoter has been inserted upstream of the YAH1 coding region, thus permitting regulated expression of the YAH1 gene (12). Gal-YAH1 cells grow at wild-type rates in galactose-containing media, yet exhibit a severe growth defect upon substituting glucose or glycerol for galactose. Expression and localization of the human ferredoxins, to which fungal mitochondrial targeting sequences were fused, was verified by immunoblotting (Fig. 3A). Human Fdx2 was able to restore growth of Gal-YAH1 cells on media containing either glucose or the nonfermentable carbon source glycerol (Fig. 3B). Consistent with a previous study, Fdx1 expression failed to replace Yah1 (28).

To analyze if Fdx2 can replace all functions of Yah1, we analyzed several biochemical parameters associated with functional loss of Yah1, including defective Fe/S protein biogenesis, heme biosynthesis, and iron regulation (29, 30). First, expression of Fdx2 restored the activity of the Fe/S protein aconitase to the same levels as yeast Yah1 (Fig. 3C). Second, Fdx2 was able to complement the defect in heme synthesis (29, 30), as indicated by measuring the activity of the heme-dependent catalase. Third, the severe defect in COX activity in Yah1-depleted Gal-YAH1 cells (15) was fully rescued by synthesis of Fdx2, indicating that human Fdx2 can participate not only in yeast mitochondrial Fe/S cluster synthesis but also replace Yah1 function in the hydroxylation of heme O to heme A. Finally, we tested whether expression of Fdx2 can normalize the activation of the yeast iron regulon observed upon functional defects of the mitochondrial iron-sulfur cluster (ISC) assembly and export machineries (31, 32). As a reporter we used the promoter of FET3, which encodes the multi-copper ferroxidase of the high affinity iron uptake system (33), fused to GFP. Green fluorescence was strongly increased in Yah1-depleted Gal-YAH1 cells carrying the FET3-GFP construct (Fig. 3C) (32). Ectopic expression of human Fdx2 in these cells normalized the GFP fluorescence to wild-type levels. For all these four assay systems, no changes of signals measured for Yah1-deficient cells were observed upon expression of Fdx1 (Fig. 3C). Taken together, Fdx2, but not Fdx1 can fully replace yeast Yah1 in all tested biochemical parameters, making Fdx2 the functional Yah1 orthologue.

Fdx2 Depletion Adversely Affects Cellular Iron Metabolism in Human Cells.

Iron regulatory protein 1 (IRP1) is a [4Fe-4S] protein that resides in the cytosol where it has aconitase activity (see Fig. 2). When cellular Fe/S cluster assembly becomes limiting (e.g., due to low iron availability or upon defects of the biogenesis machineries), IRP1 loses its Fe/S cluster thereby gaining the ability to bind to stem-loop structures of mRNAs encoding several proteins involved in iron homeostasis (9, 34). IRP1 functions together with the related IRP2 that does not bind an Fe/S cluster, yet is degraded under iron-replete conditions. Examples of IRP-regulated mRNAs are those of transferrin receptor (TfR), which is stabilized by IRP binding, and ferritin (Ft), whose translation is inhibited by IRP binding. Because Ft is a ubiquitous iron storage protein and TfR is the gateway of iron entry into most cells, the reciprocal regulation of these two proteins by iron or Fe/S protein biogenesis endows cells with the ability to rapidly adjust iron availability to meet cellular needs. In an RNA electrophoretic mobility shift assay, Fdx2-depleted HeLa cells exhibited a robust increase in IRP1 RNA binding activity compared to controls or cells cotransfected with psmFDX2 (Fig. 4A). IRP2 also showed increased mRNA binding activity. Surprisingly, the changes in the protein levels of the two IRPs were opposite under Fdx2 depletion (Fig. 4B). For IRP1 the decrease is readily explained by the common observation that apo-IRP1 lacking its Fe/S cluster is more prone to degradation (26, 35). Because IRP2 protein levels are negatively regulated by the iron availability in the cytosol (34), our results suggest that cytosolic iron was diminished. To verify this, we radiolabeled Fdx2-depleted HeLa cells with ⁵⁵Fe-transferrin for 3 d, performed cell fractionation, and evaluated the distribution of ⁵⁵Fe by native PAGE. A low molecular mass iron species (36, 37) was dramatically increased in the crude mitochondrial fraction of cells deficient in Fdx2, suggesting mitochondrial iron accumulation when Fdx2 is depleted and hence when Fe/S protein biogenesis is hindered (Fig. 4C). The low molecular mass iron species was not detectable in the cytosolic fraction, yet the amount of iron stored in ferritin was decreased. We conclude from these data that depletion of mitochondrial Fdx2 has a severe effect on cellular iron homeostasis, which is most likely initiated by an assembly defect of the Fe/S cluster of IRP1.

Fdx2 Cannot Substitute for Fdx1 in Steroidogenesis.

Our results so far indicate that Fdx2 performs a specific function in cellular Fe/S protein assembly. Because this protein is highly homologous to Fdx1, we examined whether it can also participate in the reduction of mitochondrial CYP for steroidogenesis. An electron transfer system was set up in vitro, using NADPH as an electron source, recombinant purified FdxR, Fdx1, Fdx2, and 11beta-hydroxylase (CYP11B1), and 11-deoxycortisol as the substrate (Fig. 5A). Using HPLC, we observed an efficient conversion of 11-deoxycortisol into cortisol by CYP11B1 that was dependent on the Fdx1 concentration (Fig. 5B, Upper). Surprisingly, Fdx2 was highly inefficient in this reaction; even at 20 μM Fdx2 barely any product formation was detected (Lower). Quantitative analysis revealed a 500-fold lower rate of conversion by Fdx2 compared to Fdx1 (Inset). UV-Vis spectrophotometry confirmed that NADPH/FdxR almost instantly reduced the [2Fe-2S] cluster of Fdx2, indicating that this part of the electron transfer chain does not throttle back the CYP11B1 activity (compare to Fig. 1B). Taken together, our findings show a high specificity of Fdx1 for steroidogenesis and of Fdx2 for heme A and Fe/S cluster biosynthesis.

Reagents and Cell Lines.

Antibodies against Fdx1 and Fdx2 were raised in rabbits using purified proteins produced in E. coli. The remaining antibodies were against: mitochondrial aconitase (L. Szweda), GPAT (H. Puccio), F₁α/β (H. Schägger), TfR (Zymed), Ft (ICN), tubulin (clone DM1α, Sigma), SDH, and NDUFS3 (MitoScience). Small interfering RNAs (siRNAs) with predesigned sequences were purchased from Applied Biosciences. For all RNAi assays performed, transfection with two separate scrambled siRNAs yielded identical results as mock transfections. Human tissue lysates were purchased from Axxora, cell culture reagents from PAA Laboratories, and all other reagents from Carl Roth GmbH, unless otherwise noted.

HeLa cells were maintained in DMEM containing 4.5 g/liter glucose and supplemented with 7.5% FCS, 1 mM glutamine, and 1% penicillin-streptomycin. The previously described Gal-YAH1 yeast strain was generated in a W303-1A (MATα, ura3-1, ade2-1, trp1-1, his3-11,15, leu2-3,112) background by exchanging the YAH1 promoter with a galactose-inducible GAL1-10 promoter (12).

Plasmids. (VSports最新版本)

All oligonucleotides used, including siRNA sequences, are listed in Table S1. Human Fdx1 and Fdx2 cDNA clones in pCMV-SPORT6 were acquired from imaGenes. Using these clones, we performed PCR amplification to clone the genes lacking their putative mitochondrial targeting sequences into pETDuet-1 (Novagen), which was then transformed into BL21 E. coli cells for recombinant ferredoxin production. This construct incorporates an N-terminal 6xHis tag, which was used for protein purification on a Ni-nitrilotriacetic acid (NTA) column (GE Healthcare Life Sciences). Additional plasmids were generated by cloning the entire coding regions of the respective genes into pEGFP-N3 (Clontech), thus generating constructs that encode Fdx1 (plasmid pFDX1) or Fdx2 (pFDX2) fused to a C-terminal EGFP. For complementation experiments, pFDX2 was mutated in the coding regions targeted by the siRNAs, using previously described methods (43) to introduce numerous silent mutations (psmFDX2).

VSports注册入口 - Miscellaneous Procedures.

Previously described methods were used for harvesting and fractionating HeLa cells with digitonin, transfection by electroporation; measurements of the catalytic activities of the aconitases, complex I, SDH, cytochrome oxidase, CS and lactate dehydrogenase (LDH), yeast aconitase, and catalase (27, 44), estimation of yeast iron regulon activation by FET3-GFP (30), in vitro hydroxylation of 11-deoxycortisol to cortisol (19), determination of IRP binding activities (23), native PAGE separation of iron-bound moieties (36), and iron incorporation into respiratory complexes (24, 27). All bar graphs represent the average of at least three independent sets of transfections.