Steroidogenic Acute Regulatory Protein (StAR) Is A Sterol Transfer Protein*

Steroidogenic acute regulatory protein (StAR) plays a critical role in steroidogenesis by enhancing the delivery of substrate cholesterol from the outer mitochondrial membrane to the cholesterol side chain cleavage enzyme system on the inner membrane. A recombinant StAR protein lacking the first N-terminal 62 amino acid residues that includes the mitochondrial targeting sequence was shown to stimulate the transfer of cholesterol and β-sitosterol from liposomes to heat-treated mitochondria in a dose-, time-, and temperature-dependent manner. A recombinant mutant StAR protein that cannot stimulate steroidogenesis by isolated mitochondria did not promote sterol transfer. Unlike the more promiscuous lipid transfer protein, sterol carrier protein 2 (SCP2), StAR did not stimulate phosphatidylcholine transfer in our assay system. The recombinant StAR protein increased cholesterol transfer to heat-treated microsomes as well as to heat- and trypsin-treated mitochondria. These observations demonstrate that StAR has sterol transfer activity, which may reflect an ability to enhance desorption of cholesterol from sterol-rich donor membranes. We suggest that the ability of StAR to promote sterol transfer explains its steroidogenic activity.

Steroidogenic acute regulatory protein (StAR) plays a significant role in steroidogenesis at the initial step in hormone biosynthesis, the conversion of cholesterol into pregnenolone (1)(2)(3)(4)(5). This reaction is catalyzed by the cholesterol side chain cleavage system, which is localized to the mitochondrial inner membranes. StAR appears to increase the delivery of sterol substrate from the outer mitochondrial membrane to the cholesterol side chain cleavage enzyme (P450scc) because individ-uals carrying mutations that inactivate the StAR protein (a disease named congenital lipoid adrenal hyperplasia) have markedly impaired gonadal and adrenal steroidogenesis associated with massive accumulation of cholesterol in cytoplasmic lipid droplets (6 -8). Targeting of the murine StAR gene by homologous recombination yields an identical phenotype of impaired steroidogenesis and adrenal lipid accumulation in nullizygous animals (9).
The StAR protein has an N-terminal sequence that directs the preprotein into the mitochondria where it is processed to yield the mature protein (1). The N terminus of the StAR preprotein is not essential for the steroidogenic activity of StAR because a StAR mutant lacking the first 62 N-terminal amino acid residues (N-62 StAR) is as steroidogenically active as full-length protein when expressed in COS-1 cells in conjunction with the human cholesterol side chain cleavage system (10,11). Moreover, recombinant N-62 StAR protein, despite the fact that it is incapable of being imported into mitochondria, stimulates pregnenolone synthesis by mitochondria isolated from bovine corpus luteum (11). These observations have been interpreted as indicating that StAR acts on the outer mitochondrial membrane to promote cholesterol delivery to cytochrome P450scc. The mechanism by which StAR acts to enhance substrate delivery to the cholesterol side chain cleavage enzyme has eluded investigators, and no mitochondrial receptor has yet been identified for the StAR protein. The present experiments were carried out to test the hypothesis that StAR is a sterol transfer protein. To this end, we examined the ability of recombinant N-62 StAR to stimulate the movement of lipids from donor vesicles to acceptor membranes. Our findings demonstrate that StAR promotes sterol but not phosphatidylcholine transfer, distinguishing its activity from other known lipid transfer proteins.

Recombinant StAR and SCP 2 Proteins-N-62
StAR containing a C-terminal or N-terminal 6-histidine tag (His 6 tag), and a N-62 StAR C-terminal His 6 tag mutant with an A218V amino acid replacement were produced in Escherichia coli and purified as described previously (11). The A218V mutation was selected because it inactivates the protein and causes congenital lipoid adrenal hyperplasia (7). Our previous studies demonstrated that N-62 C-His 6 tag StAR stimulates pregnenolone synthesis by isolated bovine corpus luteum mitochondria, whereas the N-62 C-His 6 tag A218V mutant is inactive (11).
Recombinant human mature sterol carrier protein 2 (SCP 2 ) was produced in E. coli and purified as described by Matsuura et al. (12). The N-terminal amino acid residue of the recombinant SCP 2 was an alanine rather than serine, a change engineered in the SCP 2 cDNA sequence to enhance translation. The recombinant SCP 2 employed has previously been shown to have lipid transfer activity (12).
Lipid Transfer Assays-The lipid transfer activities of recombinant StAR and SCP 2 were examined using methods previously described (12)(13)(14) that entail the measurement of transfer of 14 C-radiolabeled lipids (1.0 mCi/mol) incorporated into egg yolk phosphatidylcholine liposomes to heat-treated mitochondria or microsomes. The liposomes contained tracer amounts of [ 3 H]triolein to monitor liposome-acceptor fusion or liposome trapping. Briefly, liposomes were prepared by sonication as described by Bloj and Zilversmit (13). Sterol-containing liposomes had a molar ratio of phosphatidylcholine/sterol of 1:0.875. In each assay, donor liposomes contained 16 nmol of phosphatidylcholine and 14 nmol of cholesterol or ␤-sitosterol (sterol transfer) or 30 nmol of phosphatidylcholine (phospholipid transfer). The liposomes produced by this method have a mean diameter between 30 and 45 nm (14).
Acceptor organelles were prepared by differential centrifugation (14). Mitochondria were isolated from mouse and rat liver, bovine corpus luteum, and yeast. Microsomes were prepared from the post-mitochondrial supernatant of rat liver homogenates. 3 mg of the organelle protein was added as acceptor in each assay in a final volume of 500 l 20 mM Tris buffer containing 250 mM sucrose, 1 mM EDTA, and 1 mg of bovine serum albumin, pH 7.4, with the indicated concentration of recombinant N-62 His 6 tag StAR or SCP 2 . The organelle preparations were pre-heated at 60°C for 30 min to denature endogenous lipases as described previously (13). In selected experiments, the heat-treated mitochondria were also treated with trypsin (200 g/ml) for 40 min at 4°C. The trypsin was then neutralized with 1 mM phenylmethylsulfonyl fluoride and 1.25 mg/ml soybean trypsin inhibitor for 10 min at 4°C prior to the transfer assay. The data presented in Figs. 1, 2, and 4 represent the stimulation of lipid transfer over that occurring in the absence of added recombinant protein. In addition, all results presented in these figures were corrected for the small amount of liposome-acceptor fusion/trapping detected by the transfer of [ 3 H]triolein, which always amounted to Ͻ10%, and usually Յ3%, of N-62 StAR-or SCP 2stimulated lipid transfer. Experiments were reproduced on at least three separate occasions except where noted. Values presented are the means Ϯ S.E.
Western Blotting-Western blot analysis of mitochondrial and microsomal protein markers was carried out as described by Towbin et al. (15). Endosomal transferrin receptors were detected with a mouse antitransferrin receptor antibody purchased from Zymed Laboratories Inc. (South San Francisco, CA), whereas mitochondrial heat shock protein-70 (HSP70) was detected with a mouse anti-HSP70 antibody obtained from Affinity BioReagents, Inc. (Golden, CO). Antibody detection was performed using horseradish peroxidase-conjugated anti-mouse antibodies purchased from Cappel (Durham, NC) and an enhanced chemiluminescence kit from Amersham Pharmacia Biotech. was reached at 1-4 M StAR protein, resulting in 60% cholesterol transfer from donor vesicles to the heat-treated mitochondria. The incubation medium in all cases contained 2 mg/ml bovine serum albumin, and the reported lipid transfer activities in Fig. 1 reflect sterol transfer above that occurring in the absence of recombinant protein. C-His 6 tag N-62 StAR stimulated sterol transfer to heat-treated mouse and rat liver, bovine corpus luteum, and yeast mitochondria to a similar extent (data not shown). Hence, rat liver mitochondria were used in most of the studies.
The action of C-His 6 tag N-62 StAR on cholesterol transfer was rapid, with the majority of transfer occurring within 5 min of incubation at 37°C. The cholesterol transfer activity of the C-His 6 tag recombinant N-62 StAR protein was temperaturedependent because cholesterol transfer was minimal at 4°C (Fig. 2). N-His 6 tag N-62 StAR also displayed cholesterol transfer activity (Fig. 2), as did recombinant human SCP 2 at a 1 M concentration. The former finding demonstrates that the position of the His 6 tag does not substantively influence sterol transfer activity of N-62 StAR. In contrast, heat-denatured C-His 6 tag N-62 StAR and the C-His 6 tag N-62 A218V mutant displayed minimal cholesterol transfer activity (Ͻ2.5 and Ͻ1% sterol transfer, respectively), even when added at concentra-tions greater than the wild-type recombinant protein (Fig. 2). Importantly, C-His 6 tag N-62 StAR increased cholesterol transfer to heat-treated mitochondria that had been pre-treated with trypsin to digest outer mitochondrial membrane proteins. The efficacy of the trypsin treatment was documented by SDSpolyacrylamide gel electrophoresis analysis, which revealed substantial degradation of mitochondrial proteins (Fig. 3).
C-His 6 tag N-62 StAR increased the transfer of cholesterol to heat-treated microsomes, as did SCP 2 (Fig. 4). In contrast to SCP 2 , also called nonspecific lipid transfer protein because it promotes the transfer of a variety of lipids including various sterols, phospholipids, and glycolipids, N-62 C-His 6 tag StAR did not stimulate phosphatidylcholine transfer to mitochondria or to microsomes (Fig. 4). The microsomal preparation was shown to have modest mitochondrial contamination by Western blot analysis of mitochondrial heat shock protein-70 (Fig.  5). The mitochondrial preparation did not contain detectable levels of the transferrin receptor, a microsomal/endosomal marker (Fig. 5).
The specificity of C-His 6 tag N-62 StAR sterol transfer activity was tested by examining the ability of the protein to promote transfer of the plant sterol, ␤-sitosterol. The C-His N-62 StAR protein enhanced ␤-sitosterol transfer from sterol-rich liposomes to heat-treated mitochondria to the same extent as it promotes cholesterol transfer (Table I).

DISCUSSION
Our observations demonstrate that N-62 StAR has sterol transfer activity with specificity that can be distinguished from SCP 2, a protein that stimulates the transfer of several different lipids from donor vesicles to acceptor membranes. The fact that recombinant N-62 C-His 6 tag StAR effectively stimulates sterol transfer from sterol-rich liposomes to heat-treated mitochondria isolated from various animal sources, to mitochondria that were trypsin-treated, and to microsomes demonstrates that StAR protein lacking the N-terminal mitochondrial targeting sequence does not have acceptor organelle specificity and that it does not require heat-and/or trypsin-sensitive proteins on the acceptor for efficient sterol transfer. Indeed, these observations suggest that the specificity of action of the StAR protein is determined by the N-terminal mitochondrial targeting sequence. These conclusions support a bipartite model of functional domains in StAR: an organelle-specifying N terminus and a sterol transferring activity encoded by the remaining C terminus of the protein. StAR, like SCP 2, does not distinguish between cholesterol and the plant sterol, ␤-sitosterol, a cholesterol structure with an ethyl group at carbon 24. However, further experiments are needed to determine whether the sterol transfer activity of StAR differentiates other features of the cholesterol molecule, including the 3-oxo function and the ⌬ 5 double bond.
StAR is a phosphoprotein (4,5), and the phosphorylation state of StAR is directly correlated with steroidogenesis (16,17). We have previously shown that the ability of StAR to be phosphorylated at serine residue 195, which is in the context of a consensus protein kinase A phosphorylation motif that is conserved in all species of StAR identified to date, is associated with increased steroidogenic activity (18). Because a StAR mutant that cannot be phosphorylated on serine 195 (S195A) had approximately 40% of the steroidogenic activity of the wild-type protein in transfected COS-1 cells, phosphorylation of StAR is evidently not obligatory for activity of StAR. The recombinant StAR protein used in the present studies is not phosphorylated. Thus, we can conclude that StAR does not need to be posttranslationally modified to promote sterol transfer. It should also be noted that at present we do not know what percentage of the recombinant C-His 6 tag N-62 StAR is properly folded and biologically active. Thus, the native protein may be active in promoting sterol transfer at much lower concentrations than those employed in the experiments described here.
Because StAR is not found in large quantities in steroidogenic cells, a 1:1 carrier mechanism for sterol delivery to the inner mitochondrial membrane is unattractive. The promotion of membrane fusion by StAR is excluded as a mechanism because the amount of lipid transfer in our assays attributable to a fusion process was minimal. This is not consistent with models of StAR action based on a fusogenic role for the protein. A mechanism invoking formation of bridges between donor and acceptor membranes, as has been suggested previously for SCP 2 action (14), is also unlikely because StAR evidently can act on the outer mitochondrial membrane and thus could not easily be envisioned to form bridges between the outer and inner mitochondrial membranes. Therefore, we favor a "catalytic" process by which StAR promotes multiple molecules of sterol to move from the outer to the inner mitochondrial membranes. StAR perturbation of donor membrane structure, resulting in sterol desorption from the relatively cholesterol-rich outer mitochondrial membrane (19) is a possible mechanism for substrate transfer to the cholesterol-poor inner mitochondrial membrane. This mechanism predicts that the direction of sterol flux is influenced by the abundance of sterol in the two membranes with net sterol transfer slowing or ceasing as the sterol contents of the donor and acceptor membranes equilibrate. In mitochondria, cholesterol provoked to exit the outer membrane by StAR might travel to inner the membrane via pre-existing contact sites.
Enhanced sterol desorption has also been proposed as the explanation for the ability of SCP 2 to increase sterol transfer in cell-free assays (20). However, because StAR displays greater specificity for sterol transfer than SCP 2 , it is likely that the two proteins act by different molecular mechanisms. It has been attractive to envision StAR interacting with a protein "receptor" or recognition protein on the outer mitochondrial membrane. However, the fact that StAR can promote steroidogenesis by mitochondria of cells that do not normally produce steroid hormones (i.e. COS-1 cells transfected with the cholesterol side chain cleavage enzyme system) and that StAR can stimulate cholesterol transfer to mitochondrial as well as microsomal membranes argues against the existence of a receptor on mitochondria of steroidogenic cells for StAR-mediated steroidogenesis. In summary, the present studies demonstrate that StAR has sterol transfer activity that may account for its steroidogenic properties.