Cardiolipin-deficient cells have decreased levels of the iron – sulfur biogenesis protein frataxin

Cardiolipin (CL) is the signature phospholipid of mitochondrial membranes, where it is synthesized locally and plays an important role in mitochondrial bioenergetics. Previous studies in the yeast model have indicated that CL is required for optimal iron homeostasis, which is disrupted by a mechanism not yet determined in the yeast CL mutant, crd1 D . This finding has implications for the severe genetic disorder, Barth syndrome (BTHS), in which CL metabolism is perturbed because of muta-tions in the CL-remodeling enzyme, tafazzin. Here, we investi-gate the effects of tafazzin deficiency on iron homeostasis in the mouse myoblast model of BTHS tafazzin knockout (TAZ-KO) cells. Similarly to CL-deficient yeast cells, TAZ-KO cells exhibited elevated sensitivity to iron, as well as to H 2 O 2 , which was alleviated by the iron chelator deferoxamine. TAZ-KO cells exhibited increased

Cardiolipin (CL) is the signature phospholipid of mitochondrial membranes, where it is synthesized locally and plays an important role in mitochondrial bioenergetics. Previous studies in the yeast model have indicated that CL is required for optimal iron homeostasis, which is disrupted by a mechanism not yet determined in the yeast CL mutant, crd1D. This finding has implications for the severe genetic disorder, Barth syndrome (BTHS), in which CL metabolism is perturbed because of mutations in the CL-remodeling enzyme, tafazzin. Here, we investigate the effects of tafazzin deficiency on iron homeostasis in the mouse myoblast model of BTHS tafazzin knockout (TAZ-KO) cells. Similarly to CL-deficient yeast cells, TAZ-KO cells exhibited elevated sensitivity to iron, as well as to H 2 O 2 , which was alleviated by the iron chelator deferoxamine. TAZ-KO cells exhibited increased expression of the iron exporter ferroportin and decreased expression of the iron importer transferrin receptor, likely reflecting a regulatory response to elevated mitochondrial iron. Reduced activities of mitochondrial iron-sulfur cluster enzymes suggested that the mechanism underlying perturbation of iron homeostasis was defective iron-sulfur biogenesis. We observed decreased levels of Yfh1/frataxin, an essential component of the iron-sulfur biogenesis machinery, in mitochondria from TAZ-KO mouse cells and in CL-deleted yeast crd1D cells, indicating that the role of CL in iron-sulfur biogenesis is highly conserved. Yeast crd1D cells exhibited decreased processing of the Yfh1 precursor upon import, which likely contributes to the iron homeostasis defects. Implications for understanding the pathogenesis of BTHS are discussed.
As the signature lipid of the mitochondrial membrane, cardiolipin (CL) is essential for optimal mitochondrial function and bioenergetics (1)(2)(3)(4)(5)(6). The de novo synthesis of CL occurs in the inner mitochondrial membrane (7) and is followed by a remodeling process in which saturated fatty acyl chains are replaced by unsaturated fatty acids (8,9). Perturbation of CL synthesis leads to defects in mitochondrial bioenergetics, resulting in decreased respiration, reduced mitochondrial membrane potential, and decreased ATP synthesis (3,(10)(11)(12). CL is tightly associated with a wide range of proteins in the mitochondrial inner membrane (3,10) and is essential for the stability of respiratory chain supercomplexes (13)(14)(15). The severe genetic disorder Barth syndrome (BTHS) results from loss of function of tafazzin, the transacylase that remodels CL, highlighting the significance of this lipid (16). Patients with BTHS suffer from cardio-and skeletal myopathy, neutropenia, and growth retardation (17) and exhibit biochemical and bioenergetic defects at the cellular level that are similar to those observed in CL-deficient yeast cells (18)(19)(20). However, the clinical presentation of the disorder in BTHS individuals is highly variable, ranging from neonatal death to lack of clinical symptoms, suggesting that physiological modifiers strongly influence the phenotype of CL deficiency (21). The mechanism linking the pathology in BTHS to CL deficiency is not known.
A previous study indicated that yeast cells lacking CL exhibit decreased iron-sulfur biosynthesis, resulting in perturbation of iron homeostasis (22). Iron is essential for cellular functions in all eukaryotes, from yeast to humans (23). A deficiency in iron causes anemia and even death, whereas iron overload can be toxic as a result of free radical damage (24). Therefore, appropriate intracellular iron levels must be maintained by the regulation of iron transport, storage, and regulatory proteins (25). In mammalian cells, transferrin receptor 1 (TFR1) is responsible for the uptake of extracellular iron by binding to the plasma iron transport protein, transferrin (25). The transmembrane protein ferroportin-1 (FPN1) transports intracellular iron to the outside of the cell (26). Iron is utilized in the formation of iron-sulfur clusters, which are essential co-factors for many enzymatic reactions, including those of electron transfer, enzyme catalysis, and regulatory processes (27). Iron-sulfur biosynthesis is a complex and highly conserved process. The protein Isu provides the primary scaffold for iron-sulfur cluster assembly (28). Cysteine desulfurase (Nfs1) removes sulfur from cysteine and delivers it to Isu (29), a step that requires interaction with accessory protein Isd11 (30). It has been proposed that imported iron is carried by Yfh1/ frataxin to Isu (31). After the incorporation of sulfur and iron, the cluster assembly reaction requires reducing equivalents, which are provided by ferredoxin reductase (Arh1) and ferredoxin (Yah1) (32,33). The iron-sulfur cluster is then delivered to apoproteins by interacting with the Hsp70 chaperone, Ssq1, as well as several other proteins (34).
Frataxin plays a key role in iron-sulfur biogenesis (35) and is highly conserved from prokaryotes to eukaryotes (36)(37)(38). It is encoded in the nucleus, is targeted to the mitochondria (39), and acts as a gatekeeper for the formation of iron-sulfur clusters. Several functions of frataxin have been proposed, including regulation of iron homeostasis, storage of iron to decrease toxic levels of free mitochondrial iron, and control of Nfs1 activity (40). Although the exact role of frataxin is unclear, its interaction with Isu1, Nfs1, and Isd11 suggests that it is a key player in regulating iron-sulfur biogenesis (41). Frataxin deficiency in humans causes the severe genetic disorder, Friedreich ataxia (42,43), which is characterized by progressive neurodegeneration and hypertrophic cardiomyopathy (44). Interestingly, the yeast frataxin mutant yfh1D exhibits phenotypes similar to those of CL-deficient crd1D cells, including accumulation of mitochondrial iron (45), decreased activity of aconitase (46), and hypersensitivity to oxidative stress (46). In the current study, we show that CL deficiency impairs the second processing step of imported Yfh1 precursor and that CL-deficient cells have decreased levels of mature frataxin and concomitant perturbation of iron homeostasis. Elucidating the role of CL in maintaining iron homeostasis may identify specific mitochondrial defects that contribute to the pathology in BTHS.

Impaired iron homeostasis and iron-sulfur deficiencies in TAZ-KO cells
We reported previously that CL-deficient yeast cells exhibit defective iron-sulfur biogenesis, as reflected in decreased activities of mitochondrial and cytoplasmic enzymes requiring iron-sulfur co-factors and in altered iron homeostasis (22). To determine whether the role of CL in these essential processes is conserved, iron-sulfur biogenesis and iron homeostasis were analyzed in the CL-deficient mouse cell line, TAZ-KO, in which the tafazzin gene is disrupted (47). TAZ-KO cells exhibit the prototypical biochemical phenotypes of BTHS, including decreased levels of total CL and unsaturated CL species and increased monolysocardiolipin (47). Iron-sulfur-requiring enzymes were assayed in TAZ-KO cells, as shown in Table   1. Activities of aconitase, NADH dehydrogenase, succinate dehydrogenase, and ubiquinol-cytochrome c reductase were decreased in TAZ-KO mitochondria to 42-57% of WT. Levels of these mitochondrial proteins were not decreased in the mutant mitochondria ( Fig. 1), suggesting that the defect was in catalytic activity. Although previous studies indicated that CL is required for activity of cytochrome c oxidase (48,49), which does not require an ironsulfur co-factor, complex IV activity was not affected in TAZ-KO cells.
A decrease in iron-sulfur biogenesis is expected to lead to impaired iron homeostasis, which would be reflected in altered transport of iron (50). As discussed above, TFR1 and FPN1 mediate iron transport in mammalian cells (51). TFR1 transfers extracellular iron into cells by endocytosis (25), and FPN1 exports iron (26). In TAZ-KO cells, the TFR1 mRNA and protein levels are decreased, whereas the levels of FPN1 are increased (Fig. 2), as expected if iron levels are elevated. In agreement with this, an 18% increase in mitochondrial iron was observed in TAZ-KO cells relative to WT (data not shown), similar to what was observed in yeast crd1D cells, which lack CL (22), and in cells with a deficiency in iron-sulfur cluster synthesis or export of iron-sulfur co-factors (52)(53)(54)(55)(56)(57). Fibroblasts obtained from Friedreich ataxia patients exhibit a greater (40%) increase in mitochondrial iron (58). Increased mitochondrial iron levels lead to hypersensitivity to oxidative stress, which results in sensitivity to H 2 O 2 (28,45,53) or iron supplementation (46,(59)(60)(61). Consistent with this, TAZ-KO cells exhibited significantly greater sensitivity than WT cells to H 2 O 2 (Fig. 3A) and to iron (Fig. 3B). To test whether increased sensitivity of TAZ-KO cells to H 2 O 2 was due to elevated iron, the effect on viability of the iron chelator deferoxamine (DFO) was Table 1 Decreased iron-sulfur enzyme activities in TAZ-KO mitochondria. The enzyme activities were assayed as described under "Experimental procedures." The data shown are means 6 S.D. (n = 3) determined. Growth of TAZ-KO cells treated with DFO was less sensitive to H 2 O 2 -induced death than untreated TAZ-KO cells (Fig. 3C). Taken together, these findings are consistent with impaired iron homeostasis and defective iron-sulfur biogenesis in TAZ-KO cells and indicate that the requirement for CL in these processes is conserved from yeast to mammals.

Impaired processing of imported Yfh1 precursor in yeast crd1D cells
Previous studies in yeast cells indicated that CL is required for the optimal import of some mitochondrial proteins (11,62). Therefore, we speculated that altered iron-sulfur biogenesis in CL-deficient cells may result from defective import of proteins required for iron-sulfur biogenesis. In addition, Yfh1 maturation may be perturbed, because the biogenesis of Yfh1 involves two processing steps of the Yfh1 precursor by the mitochondrial processing peptidase upon import into mitochondria (63,64). To analyze whether biogenesis of Yfh1 depends on CL, we imported radiolabeled Yfh1 precursor into isolated WT and crd1D mitochondria. The intermediate and mature forms were only detected in the presence of a membrane potential, confirming that the Yfh1 precursor was imported into WT and crd1D mitochondria (Fig. 4A). Interestingly, the second processing step of the Yfh1 precursor was impaired, whereas the first processing was only mildly decreased (Fig. 4A). In contrast, the import of other proteins (Isu1p and Yah1p) involved in iron-sulfur biosynthesis was largely unaffected or only mildly reduced (Nfs1p) (Fig. 4B). The decreased import rates of Yfh1p and Nfs1p are likely due to the reduced membrane potential (Figs. 4C) (10, 60), which drives protein translocation into the matrix and inner membrane. Consistent with the impaired processing of Yfh1, crd1D exhibited decreased Yfh1 protein levels but unchanged Nfs1 and Isu1 levels (Fig. 5A). Decreased Yfh1 in crd1D was not the result of reduced gene expression, because YFH1 transcript levels were comparable in WT and CL mutant cells (Fig. 5B).
Yfh1 was previously shown to associate with the mitochondrial inner membrane (39), which is enriched in CL. Therefore, the possibility was tested that CL physically interacts with Yfh1. As shown in the protein lipid overlay assay (Fig. 6), purified recombinant Yfh1 displayed a high affinity for CL. Yfh1 did not bind to phosphatidylcholine, phosphatidylethanolamine, phosphatidylinositol 4,5-bisphosphate, or phosphatidylinositol (3,4,5)-trisphosphate. Comparatively weaker binding to phosphatidic acid was also observed. These findings suggest that binding to CL may be required for optimal processing of Yfh1.

Decreased frataxin in TAZ-KO cells
Because perturbation of iron homeostasis and iron-sulfur biogenesis is conserved in CL-deficient mammalian cells, we considered the possibility that in mammalian cells, as in yeast, the underlying defect is in mitochondrial frataxin. In agreement with this, mitochondria from TAZ-KO cells exhibited decreased levels of mature frataxin, consistent with defective mitochondrial processing of the protein (Fig. 7A). Taken together, these experiments suggest that CL is required for optimal processing of Yfh1/frataxin and that perturbation of CL synthesis leads to defects in iron-sulfur biogenesis and iron homeostasis as a result of Yfh1/frataxin deficiency.

Discussion
In this study, we demonstrate for the first time that CL-deficient cells exhibit defective processing of imported Yfh1 precursor. Yfh1/frataxin is a unique protein with respect to processing in both yeast and mammalian cells, because the precursor undergoes two cleavages by the mitochondrial processing peptidase (MPP) to form the mature protein (63). Impaired second-step processing (Fig. 4) is the likely mechanism underlying reduced levels of mature Yfh1/frataxin in CLdeficient yeast and mouse mitochondria (Figs. 5 and 7). These findings suggest the model shown in Fig. 8. Accordingly, following import into mitochondria, processing of the Yfh1/frataxin precursor to the intermediate by the MPP occurs similarly in CL mutant and WT cells. However, optimal processing of the intermediate to the mature form, which is deficient in the mutant (Fig. 4), requires CL. A possible explanation for this defect is that CL is required to maintain the proper conformation of Yfh1/frataxin to expose its second cleavage site to MPP to facilitate the processing step. This is supported by the observation that Yfh1/frataxin binds to CL in vitro (Fig. 6) and has been shown to associate with the inner membrane (37). Gordon et al. (65) reported that cells deficient in the mature form of Yfh1 failed to exhibit any defects, suggesting that the intermediate can functionally compensate for the deficiencies in the yfh1D mutant. However, iron-sulfur defects were clearly observed in both crd1D yeast and TAZ-KO mouse cells, which have abundant levels of the intermediate form of Yfh1/frataxin. Therefore, it is likely that Yfh1 does not fold properly in CL-deficient mitochondria, leading to a defect in its processing and potentially contributing to the iron homeostasis defects.
In the current study, the Yfh1/frataxin deficiency in crd1D could account for defective iron-sulfur biogenesis, resulting in decreased activity of enzymes requiring iron-sulfur co-factors ( Table 1) and perturbation of iron homeostasis (Figs. 2 and 3). The phenotypes observed in the crd1D cells are similar to those of the yfh1D mutant, including accumulation of mitochondrial iron (45), decreased activities of iron-sulfur enzymes (46), and hypersensitivity to oxidative stress (46). These findings are consistent with defective iron-sulfur biogenesis in both mutants. Yfh1/frataxin, which is essential for iron-sulfur biogenesis, functions as a chaperone that delivers iron to the scaffold assembly protein, Isu1 (66). Direct interaction between CL and frataxin, as suggested by the experiment in Fig. 6, may facilitate processing of the protein by the MPP. The current studies do not distinguish between decreased total and unsaturated CL and increased monolyso-CL as the cause of frataxin deficiency in TAZ-KO cells. However, the observation that mature Yfh1 is decreased in yeast cells that completely lack CL suggests that CL interaction promotes frataxin maturation.
The findings in this report may have implications for understanding the pathology in BTHS, which is characterized by wide disparities in clinical phenotypes among patients, strongly suggesting that physiological modifiers exacerbate the outcome of altered CL metabolism (21). In this context, iron homeostasis and the robustness of iron-sulfur biogenesis may be factors that modify the clinical phenotype of BTHS. Interestingly, the iron chelator DFO, which has been proposed as a treatment to counteract oxidant sensitivity in Friedreich ataxia (58), partially rescued decreased growth of TAZ-KO cells (Fig. 3C).
Elegant studies in other laboratories have established that many of the proteins required for iron-sulfur biogenesis are highly conserved from bacteria to humans (67). In this light, it is not surprising that model organisms have been invaluable in elucidating the complex mechanisms underlying this process. The striking finding in the current study, utilizing yeast and mammalian models of CL deficiency, is that the signature mitochondrial lipid CL plays a pivotal and highly conserved role in the process, linking lipid biosynthesis to iron-sulfur biogenesis for the first time.
MTT cell proliferation assay 3000 cells were suspended in 100 ml of growth medium and seeded into 96-well plates. Viable cells were measured in triplicate using 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT; Fisher) after 3, 24, and 48 h. In brief, 10 ml of 5 mg/ml MTT were added to each well. As a negative control, MTT was applied to wells lacking cells. The plates were incubated for the indicated times at 37°C, after which the medium was carefully removed, and 150 ml of DMSO was added to dissolve the MTT product. The plates were covered with foil and incubated for 10 min at 37°C. Samples from each well were mixed well with a pipette, and absorbance was read at 570 nm.

Mitochondria isolation
The cell pellets were washed with cold PBS and suspended in mitochondrial isolation buffer (280 mM sucrose, 0.25 mM EDTA, 20 mM Tris-HCl, pH 7.2). Mouse cells were collected and manually homogenized with a glass homogenizer. Yeast cell extracts were collected by disrupting the cell wall with glass beads. Mitochondria were isolated by differential centrifugation. Cell debris was removed by centrifugation at 800 rpm for 5 min. Mitochondria were subsequently collected by  35 S-labeled Isu1, Yah1, or Nfs1 (B) in the presence or absence of a membrane potential (Dc). Following import, the samples were treated with proteinase K, which removes portions of the nonimported precursor proteins. The import reactions were analyzed by SDS-PAGE and autoradiography. A, lower panels, quantification of the intermediate and mature forms of Yfh1 in four independent import experiments with the corresponding standard error of the mean. As control, the longest import time point of Yfh1 into WT mitochondria was set to 100%. C, the membrane potential of WT and crd1D mitochondria was measured by quenching of the fluorescence of DiSC 3 . The reaction was stopped by addition of the AVO mix. centrifugation at 11,500 rpm for 10 min. Protein concentration was determined using a DC protein assay kit (Bio-Rad). For protein import studies, mitochondria were isolated by differential centrifugation (68). In brief, yeast cells were grown to an early logarithmic growth. Subsequently, the cells were harvested by centrifugation (5,500 3 g, 8 min, 20°C), incubated with DTT buffer (10 mM DTT, 100 mM Tris, H 2 SO 4 pH 9.4) and treated with Zymolyase 20T from Arthrobacter luteus (Nacalai Tesque, Kyoto, Japan) to digest the cell wall. We opened the plasma membrane with a glass potter in homogeni-zation buffer (0.6 M sorbitol, 10 mM Tris-HCl, pH 7.4, 1 mM EDTA, 1 mM PMSF) at 4°C. After removal of cell debris (2,500 3 g, 5 min, 4°C), mitochondria were isolated (17,000 3 g, 15 min, 4°C), resuspended in S.E. buffer (250 mM sucrose, 1 mM EDTA, 10 mM MOPS/KOH, pH 7.2) in 10 mg/ml protein concentration and stored in aliquots at 280°C until use.

Immunoblotting
Mammalian or yeast cell extracts corresponding to 30 mg of protein were analyzed by SDS-PAGE using a 10% gel. Immunoblotting was performed using primary antibodies as indicated and corresponding secondary antibodies conjugated to horseradish peroxidase. Immunoreactivity was visualized using enhanced chemiluminescence substrate (Thermo).

Enzyme assays
NADH dehydrogenase activity was assayed by measuring electron transfer from NADH to coenzyme Q 1 . The absorbance of each sample was measured at 340 nm to account for rotenone-insensitive NADH oxidation. Succinate dehydrogenase activity was measured spectrophotometrically at 600 nm as described, with some modifications (68). The 1-ml incubation volume contained 1 g/liter BSA, 80 mM potassium phosphate, 2 mM EDTA, 10 mM succinate, 0.2 mM ATP, 0.3 mM potassium cyanide, 80 mM dichloroindophenol, 1 mM antimycin A, 50 mM decylubiquinone, and 3 mM rotenone, pH 7.8. Antimycin A and rotenone were dissolved in ethanol. Decylubiquinone was dissolved in DMSO. Succinate dehydrogenase from purified mitochondria was measured in the incubation buffer at 1-min intervals for 5 min with absorptivity at 600 nm. Ubiquinolcytochrome c oxidoreductase activity of isolated mitochondria was measured spectrophotometrically at 550 nm by supplying  CL is required for frataxin maturation decylubiquinol and cytochrome c and following the rate of reduction of cytochrome c, while inhibiting cytochrome c oxidase with 2 mM KCN. Aconitase activity was assayed by the aconitase-isocitrate dehydrogenase-coupled assay in which NADPH formation was monitored at A 340 (69). Cytochrome c oxidase was assayed by monitoring oxygen consumption. The cells were solubilized in buffer (10 mM HEPES, pH 7.4, 40 mM KCl, 1% Tween 20, 2 mM EGTA, 1 mM sodium vanadate, 10 mM KF, 1 mM PMSF, 1 mM oligomycin) and sonicated for 5 s. Total protein concentration was determined using the DC protein assay kit (Bio-Rad) with BSA as the standard. Oxygen consumption was analyzed in a closed 200-ml chamber equipped with a micro Clark-type oxygen electrode type (Oxygraph Plus System, Hansatech Instruments) at 25°C. 30 mM cytochrome c (from bovine heart, Sigma) was added after a baseline measurement. Oxygen consumption was recorded on a computer and analyzed with Oxygraph software. The turnover number was defined as nmol of oxygen consumed per minute per milligram of protein.

qPCR analysis
The cells were harvested, and total RNA was isolated using the RNeasy plus mini kit (Qiagen). cDNAs were synthesized with transcriptor first-strand cDNA synthesis kit (Roche Diagnostics), and quantitative PCRs were performed in a 25-ml volume using Brilliant TM SYBR® Green qPCR Master Mix (Stratagene) in 96-well plates. Duplicates were included for each reaction. The primers used for qPCR are listed in Table 3. The RNA level of the gene of interest was normalized to levels of GAPDH, which served as an internal control. PCRs were initiated at 95°C for 10 min for denaturation followed by 40 cycles consisting of 30 s at 95°C and 60 s at 55°C.

Import of precursor proteins into mitochondria
Precursor proteins were synthesized in rabbit reticulocyte lysates in the presence of [ 35 S]methionine (70). Import into isolated mitochondria was performed in import buffer (3% [w/v] BSA, 250 mM sucrose, 80 mM KCl, 5 mM methionine, 5 mM MgCl 2 , 2 mM KH 2 PO 4 , 10 mM MOPS-KPH (pH 7.2), 2 mM NADH, 4 mM ATP, 20 mM creatine phosphate, 0.1 mg/ml creatine kinase) at 24°C. The import reaction was stopped by the addition of AVO (8 mM antimycin A, 1 mM valinomycin, 20 mM oligomycin). In control reactions, AVO was added before the import experiment to dissipate DC. After the import reaction, mitochondria were treated for 15 min on ice with proteinase K (15-30 mg/ml) to remove the nonimported precursor proteins. Subsequently, the protease was inactivated by incubation with PMSF for 10 min on ice (final concentration, 4 mM).

Membrane potential measurement
The mitochondrial membrane potential was determined by quenching of the fluorescence dye DiSC 3 (15). DiSC 3 is taken  Frataxin precursor is synthesized in the cytoplasm and imported into mitochondria, where it is processed twice by the MPP. The model proposes that CL is required for maintaining an optimal conformation of Yfh1/frataxin for processing from the intermediate to the mature protein. Improperly folded Yfh1/frataxin in CL-deficient mitochondria results in defective processing, potentially contributing to the iron homeostasis defects.
up by mitochondria in a membrane potential-dependent manner, which leads to quenching of the fluorescence signal in the reaction mix. For this assay, isolated mitochondria were incubated with DiSC 3 in membrane potential buffer (0.6 M sorbit, 0.1% (w/v) BSA, 10 mM MgCl 2 , 0.5 mM EDTA, 20 mM KP i , pH 7.2) supplemented with 5 mM succinate and 5 mM malate for the indicated time points. The reaction was terminated by the addition of the AVO mix.

Protein lipid overlay assay
PET28a-Yfh1 was constructed by inserting the YFH1 gene carrying a polyhistidine tag (His tag) at the C terminus into a PET28a plasmid. PET28a-Yfh1 was transformed into Escherichia coli BL21 (DE3) pLysS cells for isopropyl b-D-thiogalactopyranoside-induced overexpression. The cells were lysed with a French press, and Yfh1-His protein was purified from extracts using a His-Bind purification kit (Millipore). Protein lipid overlay assays were performed according to the protocol of Dowler et al. (71) with modification. In brief, lyophilized lipids were dissolved in a 1:1 solution of methanol and chloroform to make 1 mM stocks. Lipids were diluted to concentrations ranging from 500 to 5 mM in a 2:1:0.8 solution of methanol:chloroform:water. Then 1-ml lipid samples from each dilution were spotted on nitrocellulose membranes. After drying for 1 h at room temperature, the membranes were incubated in blocking buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl 2 , 0.1% Tween 20 and 2 mg/ml BSA) with gentle rocking for 1 h at room temperature. The membranes were incubated overnight at 4°C with gentle rocking in 10 ml of blocking buffer containing 25 mg of purified Yfh1-His protein. Yfh1-His protein bound to lipid on the membrane was detected by immunoblotting using primary antibodies against the His tag and corresponding secondary antibodies.

Data availability
All data are contained within the article. Conflict of interest-The authors declare that they have no conflicts of interest with the contents of this article.