Turnover of Mitochondrial Steroidogenic Acute Regulatory (StAR) Protein by Lon Protease: The Unexpected Effect of Proteasome Inhibitors

Abstract

Steroidogenic acute regulatory protein (StAR) is a vital mitochondrial protein promoting transfer of cholesterol into steroid making mitochondria in specialized cells of the adrenal cortex and gonads. Our previous work has demonstrated that StAR is rapidly degraded upon import into the mitochondrial matrix. To identify the protease(s) responsible for this rapid turnover, murine StAR was expressed in wild-type Escherichia coli or in mutant strains lacking one of the four ATP-dependent proteolytic systems, three of which are conserved in mammalian mitochondria—ClpP, FtsH, and Lon. StAR was rapidly degraded in wild-type bacteria and stabilized only in lon⁻mutants; in such cells, StAR turnover was fully restored upon coexpression of human mitochondrial Lon. In mammalian cells, the rate of StAR turnover was proportional to the cell content of Lon protease after expression of a Lon-targeted small interfering RNA, or overexpression of the protein. In vitro assays using purified proteins showed that Lon-mediated degradation of StAR was ATP-dependent and blocked by the proteasome inhibitors MG132 (IC₅₀ = 20 μm) and clasto-lactacystin β-lactone (cLβL, IC₅₀ = 3 μm); by contrast, epoxomicin, representing a different class of proteasome inhibitors, had no effect. Such inhibition is consistent with results in cultured rat ovarian granulosa cells demonstrating that degradation of StAR in the mitochondrial matrix is blocked by MG132 and cLβL but not by epoxomicin. Both inhibitors also blocked Lon-mediated cleavage of the model substrate fluorescein isothiocyanate-casein. Taken together, our former studies and the present results suggest that Lon is the primary ATP-dependent protease responsible for StAR turnover in mitochondria of steroidogenic cells.

STEROIDS ARE SYNTHESIZED from cholesterol substrate in mitochondria of specialized cells of the adrenal cortex, the gonads and the placenta. Cholesterol supply into the mitochondria is promoted by steroidogenic acute regulatory protein (StAR) (1–5), which is a vital nuclear-encoded mitochondrial protein that is acutely expressed upon stimulation of the steroidogenic tissues by their respective trophic hormones (6–8). StAR transfers cholesterol from the outer to the inner mitochondrial membranes, where the enzyme complex of cholesterol side chain cleavage cytochrome P450 (P450scc) converts it to the first steroid, pregnenolone (9–11). StAR is synthesized in the cytosol as a 37-kDa preprotein carrying an amino-terminal targeting sequence that directs its import to the mitochondrial matrix, where it is proteolytically processed to yield a mature 30-kDa form (1, 12–14). The mechanism by which StAR facilitates cholesterol supply into the inner mitochondrial membranes is not known. Recent models of StAR activity (15) propose that cholesterol mobilization from the outer to the inner membranes of mitochondria may not necessarily depend on cholesterol binding to StAR (16). Alternative hypothetical mechanisms suggest that StAR activates cholesterol transfer by virtue of its association with a macromolecular complex consisting of the outer membrane proteins TSPO (translocator protein known as PBR, peripheral benzodiazepine receptor) and VDAC (voltage-dependent anion channel), which also interacts with peripheral proteins such as PAP7 (PBR-associated protein 7) and PKARIα [protein kinase A (PKA) regulatory subunit Iα] (17). Recruitment of PKARIα and possibly the holoenzyme to this complex could have accounted for the long-suspected functional importance of cAMP-dependent phosphorylation of StAR (18, 19); however, the substantial activity retained by an artificial nonphosphorylated S195A StAR mutant is not consistent with such a PKA-dependent role. Surprisingly, data show that artificial StAR variants that remain in the cytosol due to lack of a mitochondrial targeting signal (human N-62 and rodent N-47 StARs) (5, 20), or that are covalently bound to components of the outer membrane translocase complex, have high cholesterol transfer activity (15, 21). Thus, it has been proposed that import into the mitochondrial matrix serves as a switch-off mechanism for terminating StAR activity. Consequently, it is logical to assume that after import, mature StAR in the matrix possesses no further biological role.

Regardless of StAR’s mode of action, it is evident that hormonal activation of steroidogenesis is associated with a sudden and massive accumulation of StAR in the mitochondrial matrix (13, 14). We have previously demonstrated that StAR overexpression in COS cells results in its excessive accumulation within mitochondria, leading to mitochondrial swelling and shape changes usually indicative of organelle dysfunction (22–24). Such shape changes are not observed in mitochondria of authentic steroidogenic cells, in the gonads or the placenta (25). We therefore hypothesized that under physiological conditions, detrimental accumulation of StAR in the mitochondrial matrix is limited by resident proteases (26, 27) within this compartment. Our recent studies confirmed this notion, demonstrating that StAR is rapidly degraded within the mitochondrial matrix with a half-life of 4–5 h (22, 23). Interestingly, StAR degradation in the mitochondrial matrix was arrested by two inhibitors widely used in studies of the proteasome, MG132 and clasto-lactacystin β-lactone (cLβL) (22).

To identify the mitochondrial protease(s) involved in StAR degradation, we employed a screening strategy based on the high homology found between the ATP-dependent proteases of bacteria and mitochondria (28). To date, five ATP-dependent proteolytic complexes, all considered members of the extended AAA⁺ protein superfamily, have been identified that include FtsH, ClpAP, ClpXP, ClpYQ (HslVU), and Lon (La) (29). FtsH is a membrane-bound zinc metalloprotease that is essential in bacteria (30–33). FtsH orthologs localized to the yeast and human mitochondrial inner membrane have been identified as the Yta10/12 and paraplegin/AFG3L2 complexes, respectively (34–36). ClpP holoenzymes are soluble two-component systems composed of the same proteolytic component (ClpP) and different ATPase chaperone components (ClpA or ClpX) (37–39). ClpXP is present in the human, but not yeast, mitochondrial matrix. ClpYQ is another soluble two-component complex that shares common structural features with ClpAP but has no mitochondrial ortholog; ClpYQ exhibits similarities with the eukaryotic proteasome (40, 41). Finally, Lon is a soluble homo-oligomeric protease localized to the bacterial cytoplasm and mitochondrial matrix (42, 43). Mitochondrial Lon selectively degrades misfolded, unassembled or oxidatively damaged proteins (44) and is involved in the maintenance of mitochondrial DNA integrity in yeast and humans (45–47).

Using cell-based assays, we provide evidence that Lon is the only bacterial protease that degrades StAR overexpressed in Escherichia coli. In cultured human cells, StAR turnover is substantially abrogated by RNA interference-mediated knockdown of Lon, and accelerated by Lon overexpression. Finally, in support of our previous in vivo studies, in vitro degradation assays show now that Lon-mediated degradation of StAR is efficiently inhibited by the threonine protease inhibitors MG132 and cLβL, generally used to block the proteasome. We discuss these findings in the context of StAR function in steroid hormone biogenesis.

RESULTS

Visualized Degradation of StAR in Granulosa Cell Mitochondria

Using pulse-chase studies of metabolically labeled cells in culture, we previously showed that the StAR preprotein in the cytosol (37 kDa) is rapidly degraded by a proteolytic activity that was effectively inhibited by three inhibitors of the proteasome—MG132, cLβL, and epoxomicin (22). However, studies showed that degradation of the mature processed form of StAR within the mitochondrial matrix (30 kDa) was inhibited only by MG132 and cLβL, but not by epoxomicin, (22). One possibility that was not examined was whether mature processed form of StAR was released from mitochondria back into the cytoplasm and subjected to degradation by the proteasome. To address this possibility we used confocal microscopy to examine the level of StAR protein and its localization in the presence and absence of proteasome inhibitors. Granulosa cells from prepubertal rat ovaries were placed into culture while having no steroidogenic apparatus and lacking StAR expression (Fig. 1A). After 3 h of FSH induction, a marked accumulation of StAR in mitochondria was observed as represented by the semiquantitative pseudo-colored fluorescence (Fig. 1B, FSH visual pulse; red-yellow indicates intense fluorescence). Termination of StAR synthesis within the cell monolayer was achieved by extensive washing and replacement with hormone-free medium. Five hours later in the absence of StAR synthesis, a visual chase demonstrated that the mitochondrial StAR signal had decreased to less than 50% of its initial levels (Fig. 1C, mostly green fluorescence). The decline of StAR fluorescence was fully inhibited by MG132 or cLβL treatments, and strong mitochondrial fluorescence was maintained with no detectable fluorescence signal within the cytosol. These results provided direct evidence that MG132 or cLβL inhibited StAR degradation within mitochondria, ruling out the possibility of a retrograde movement of StAR from the mitochondrial matrix back to the cytosol. The inability of the most selective proteasome inhibitor, epoxomicin, to prevent the loss of StAR signal in the mitochondria (Fig. 1F), further ruled out the possibility of a retrograde release of StAR to the cytoplasm and its degradation there.

Fig. 1.

Immunofluorescent Visual Pulse-Chase Analysis of StAR Degradation in FSH-Treated Granulosa Cells Degradation of immunofluorescently labeled StAR was studied in primary rat ovarian granulosa cells in the presence and absence of three inhibitors used widely to block the 20S proteasome, MG132, clasto-lactacystin β-lactone (cLβL) and epoxomicin (epoxo). A, Untreated control cells. B, Cells treated with FSH for 3 h to induce StAR synthesis (FSH pulse). StAR-loaded mitochondria (arrows) were documented by confocal microscopy. C, StAR synthesis was terminated by the removal of FSH from cell cultures and after 5 h the StAR signal in the absence of inhibitor was documented (chase). D–F, The 5-h chase was performed in the presence of 20 μm of either MG132 (D, chase + MG132); or cLβL (E, chase + cLβL); or epoxomicin (F, chase + epoxo). The relative fluorescence level is color-coded; red-yellow coloration indicates high emission, green-blue colors represent lower intensities of fluorescence (color code bar). n, Nuclei.

StAR Expression and Degradation in E. coli

To identify the mitochondrial protease(s) responsible for the in vivo degradation of StAR, we expressed murine StAR in genetically manipulated bacterial strains lacking expression of endogenous ATP-dependent proteases. The mature form of StAR lacking its N-terminal presequence (N-47 StAR) was subcloned into a pBAD33 arabinose inducible vector. A 30-kDa StAR was readily expressed within 5 min of induction in both W3110 and JM101 E. coli strains (Figs. 2 and 3). Pulse-chase experiments showed that murine StAR was efficiently degraded in the wild-type W3110 bacterial strain with a t_1/2 of 7.7 ± 0.2 min, demonstrating that it was degraded by bacterial protease(s). To reveal the specific protease involved, we tested four different mutant strains, each incapable of expressing either ClpP, FtsH, ClpQ, or Lon. Because FtsH is an essential protease in E. coli, unlike the other protease systems, a temperature sensitive mutant was used (ftsH1; see Materials and Methods). The rates of StAR degradation within the clpP, ftsH1, and clpQ mutant strains, as compared with wild type, were not significantly affected (P = 0.1); the half-life of StAR in the clpP null mutant was 9.2 ± 2.5 min, in the clpQ null mutant 8.1± 0.2 min and in ftsH1 cells was t_1/2 = 6.2 ± 0.7 min.

Fig. 2.

Turnover Rates of Murine StAR Expressed in Wild-Type and E. coli Mutant Strains Lacking ClpP, ClpQ, and FstH^t Determined by Pulse-Chase Analysis A, StAR expression was induced by the addition of 0.2% arabinose added to the indicated culture medium of wild-type E. coli [W3110 (wild type; WT)] or mutant strains harboring a N-47 StAR expression vector and IPTG inducible empty vector (Materials and Methods). After a 5-min induction (time = 0), translation was stopped by addition of 1 mg/ml spectinomycin (Spec.) and aliquots were taken at various time points (5–20 min) and processed for Western blot analysis to determine StAR levels as shown. Representative ECL blots are shown in the upper panels, whereas quantitative analyses (mean ± sd, n = 3) of time-dependent StAR levels are presented in the lower panel. StAR half-life values were 7.7 ± 0.2 min in WT bacteria 6.2± 0.7 min in temperature sensitive FtsH-deficient bacteria (FtsH^ts), 9.2 ± 2.0 min in ClpP-deficient cells (ClpP⁻) and 8.1 ± 0.2 min in ClpQ⁻ cells. Differences in StAR degradation rates were found to be statistically insignificant (P = 0.1). B, StAR degradation in WT E. coli was repeated as described above but for 20 μm MG132 added 10 min before addition of arabinose.

Fig. 3.

Lack of StAR Degradation in Lon-Deficient E. coli Is Reconstituted by Expression of Human Lon The degradation rate of mouse N-47 StAR was assessed in wild-type JM101 E. coli (WT) or in Lon-null JM101 cells (Lon). After a 5-min induction of StAR synthesis by arabinose and termination of translation by spectinomycin (Spec,), aliquots were taken at various time points (15–20 min) and processed for Western blot analysis to determine StAR levels as described in Fig. 2. StAR was stabilized in the absence of endogenous bacterial Lon (Lon⁻). StAR degradation in Lon⁻ cells was reconstituted by the coexpression of human Lon (Lon⁻ + hLon) and StAR; arabinose was used to induce StAR and IPTG to induce human Lon expression. Representative StAR blots are shown in the upper panels, whereas quantitative analyses (mean ± sd, n = 4) of the data are shown in the lower panel. The half-life of StAR in WT cells was t1/2 = 7.9 ± 1.1 min, whereas that in Lon-null bacteria expressing human Lon was t1/2 = 5 ± 0.9 min, respectively. There was no significant difference (P = 0.8) between the degradation rates of StAR in the WT strains W3110 and JM101 used in Figs. 2 and 3.

By contrast, in a mutant E. coli strain lacking the Lon protease, StAR turnover was blocked and stable as compared with the wild-type JM101 strain. Figure 3 shows that the half-life of StAR in JM101 (t_1/2 = 7.9 ± 1.1 min, P = 0.8) was similar to that observed for the W3110 strain; StAR was significantly stabilized in the Lon-deficient bacteria throughout the entire duration of the chase period. This result suggested that, in bacteria, the degradation of StAR was mediated mainly by Lon. To ascertain that lack of StAR degradation was directly related to loss of the enzyme activity in the mutant bacteria, we coexpressed the human Lon protease and murine N-47 in the lon mutant strain. Pulse-chase experiments showed that coexpression of human Lon protease efficiently restored StAR turnover (t_1/2 = 5 ± 0.9 min). Interestingly, execution of StAR degradation by the mammalian protease expressed in the bacteria was 60 times faster than the rate of StAR proteolysis in its authentic location—the mitochondrial matrix [t_1/2 = 5 h, (22)].

Although StAR degradation was blocked in a lon⁻ strain, the somewhat slower degradation of StAR in ClpP⁻ cells made us question whether ClpP can recognize StAR as a potential substrate after all. We addressed this question by two approaches: first, StAR turnover was examined in lon⁻ cells expressing human ClpP (hClpP). Previous work has shown that the E. coli ClpX chaperone ATPase (eClpX) can activate the hClpP peptidase, whereas E. coli ClpA does not associate with hClpP to form an active protease (48). Figure 4 shows that coexpression of human ClpP (hClpP) and StAR in the lon-mutant strain failed to reconstitute StAR degradation. These results suggest that the heterologous eClpX-hClpP complex, which is expected to form in the transformed lon⁻ cells, does not degrade StAR. Seeking support to this notion, we performed cell-free assays using purified murine StAR incubated in vitro in the presence of ATP and human ClpX, ClpP, or Lon protease. Clearly, StAR was rapidly degraded by Lon but not by ClpXP (Fig. 5).

Fig. 4.

Human ClpP Does Not Rescue the Lack of StAR Degradation in Lon-Deficient E. coli Human ClpP and StAR were over-coexpressed in Lon-deficient E. coli cells (JM101) using an arabinose inducible N-47 StAR vector and an IPTG inducible human ClpP vector. Induction of StAR synthesis was terminated by spectinomycin and StAR degradation was determined as described in Fig. 2. Representative ECL blots of StAR levels in wild-type cells (WT), Lon-deficient cells (Lon⁻) and Lon-deficient cells expressing human ClpP (Lon⁻ + hClpP) are depicted in the upper panels and quantitative analyses are shown in the lower panel (mean ± sd, n = 3). The half-life of StAR in WT bacteria was 7.9 ± 1.1 min, whereas in Lon-null bacteria and Lon-null bacteria cells expressing hClpP, StAR was completely stabilized.

Fig. 5.

ClpXP Does Not Degrade StAR in Cell-Free Assay Purified StAR was incubated with purified ClpXP (A) or Lon (B) in the presence of ATP at 37 C for the indicated time periods as described in Materials and Methods. The levels of StAR, Lon, ClpX, and ClpP were detected by immunoblotting using ECL. Shown are representative ECL blots. Similar results were obtained in two independent experiments.

Inhibitor-Sensitive Cell Free Degradation of StAR by Lon

Unlike the impact of the proteasome inhibitors on StAR degradation in cultured mammalian cells, when tested in bacterial cells expressing StAR, MG132 and cLβL had no detectable effect on StAR turnover, as shown in Fig. 2B. Noteworthy, cLβL has been used as efficient inhibitor of the 20S proteasome activity in intact haloarchaeon Haloferax volcanii (49), a cell lacking the typical gram-negative outer envelops of E. coli, which might render the latter impermeable to the proteasome inhibitors. Therefore, to examine whether MG132 and cLβL interfered directly with Lon activity, in vitro degradation assays were performed using purified murine StAR and human Lon protease incubated in the presence and absence of ATP and each of the proteasome inhibitors. In the presence of ATP, StAR was rapidly degraded by Lon with a half-life of less than 5 min (Fig. 6, upper panels). Upon addition, both MG132 and cLβL (50 μm) efficiently inhibited StAR turnover (Fig. 6, middle panels). The inhibition of StAR degradation by cLβL was nearly complete throughout the entire time course, whereas the effect of MG132 was less dramatic, nevertheless increasing the half-life of StAR 3-fold or more. By contrast, epoxomicin had no effect on the Lon mediated degradation of StAR (Fig. 6, lower right panel). This profile of inhibition is consistent with previous results demonstrating that the degradation of StAR within the mitochondrial matrix of intact cells was inhibited by MG132 or cLβL but not by epoxomicin. The inhibition of Lon-mediated degradation of StAR was blocked by MG132 or cLβL at concentrations well below that of the serine protease inhibitor phenylmethylsulfonyl fluoride (PMSF, Fig. 6, lower panel); millimolar concentrations of PMSF were required to achieve an effect equivalent to that of 3 μmcLβL.

Fig. 6.

Lon-Mediated Degradation of StAR in Vitro Is Blocked by MG132 and cLβL But Not Epoxomicin Purified human Lon (0.2 μg monomer) was incubated with purified mouse StAR (0.2 μg) in the presence or absence of ATP at 37 C (upper panel). The effect of proteasome inhibitors on StAR degradation was tested by preincubating Lon with either 50 μm of MG132, or cLβL (middle panel), or epoxomicin (lower panel), or PMSF (1 mm, lower panel) for 5 min before the addition of StAR. Aliquots were withdrawn at the indicated time points and StAR levels were examined by Western blot analysis. Shown are representative ECL blots. Similar results were obtained in three independent experiments.

Effects of Proteasome Inhibitors on Lon-Mediated Proteolysis of Fluorescein Isothiocyanate (FITC)-Casein

Further characterization of inhibitor-sensitive Lon activity was carried out in vitro using FITC-casein as substrate. Historically, casein has been used as a substrate to assay the activities of the various bacterial as well as eukaryotic ATP-dependent proteases (43, 50). The cleavage of FITC-casein resulting in fluorescently labeled peptides was monitored by on-line fluorimetry. As can be seen in Fig. 7, A–C, incubation of FITC-casein with purified Lon in the presence of ATP resulted in a marked and linearly increasing fluorescence release, indicating FITC-casein degradation. As expected, no degradation was observed in the absence of ATP. Furthermore, results show that when the degradation assay was performed in the presence of 1–25 μm MG132 (Fig. 7A), the reaction was efficiently inhibited in a dose-dependent manner. The effect of cLβL (Fig. 7B) appeared to be even more potent than that of MG132 (IC₅₀ of 3 μm and 20 μm, respectively, Fig. 7D). Epoxomicin did not inhibit the Lon activity at any of the concentrations tested (Fig. 7C), which is consistent with results obtained in vitro as well as in intact cells (Fig. 1 and Refs. 22 and 23).

Fig. 7.

Lon-Mediated Degradation of the Reporter Substrate FITC-Casein in Vitro Is Blocked by MG132 and cLβL But Not by Epoxomicin Human Lon (1 μg) in a reaction cocktail containing ATP was preincubated for 5 min (37 C) with different concentrations of MG132 (A), cLβL (B), or epoxomicin (C) before addition of FITC-casein substrate (2 μg). Control reactions were performed in the absence of ATP; because the inhibitors were solubilized in DMSO, all reactions contained DMSO (1% final). Panel D depicts the relative rate constants in the presence of the inhibitor compared with the maximal rate in the absence of the inhibitor. The results depict averaged triplicate measurements; two independent experiments generated identical results.

To define the biochemical nature of Lon inhibition, the initial rate of FITC-casein degradation by Lon (1 μg, 10 pmol monomer) was measured as a function of increasing substrate concentrations (0.4–4 μm) in the absence or presence of MG132 (20 μm) or cLβL (3 μm). Figure 8 shows that the substrate dose response curve was sigmoidal, indicative of cooperative enzyme kinetics. This result is consistent with observations showing positive cooperativity of substrate binding to Lon as well as two-component ATP-dependent proteases such as the proteasome and ClpXP (51–53). Figure 8 also shows that addition of MG132 and cLβL resulted in a profound inhibition of Lon activity (>95%) that was overcome, in part, at elevated substrate concentrations.

Fig. 8.

MG132 and cLβL Act as Noncompetitive Inhibitors of Lon Purified Lon (1 μg monomer) was preincubated with MG132 (20 μm) or cLβL (3 μm) for 5 min at 37 C before addition of different concentrations of FITC-casein (0.4–4 μm). Reactions with or without inhibitor contained 1% DMSO. Depicted is a V/[S] plot of Lon activity in the presence or absence of the proteasome inhibitors. The data (mean ± sd) represent three independent experiments.

In Vivo Effects of Lon Small Interfering RNA (siRNA) and Protein Overexpression

To examine the role of Lon in mitochondrial degradation of StAR, we studied the rate of StAR turnover in COS cells expressing various amount of Lon protein. To this end, we used cells transfected with either siRNA to reduce Lon protein level, or cells enriched for Lon content after Lon overexpression. Figure 9A shows a typical 85% reduction of the Lon protein level in cells undergoing a two-round transfection cycles with Lon siRNA (siLon). Lon knockdown also resulted in a 2-fold reduction of ³⁵S-StAR turnover rate observed within the first 3 h of a chase experiment (Fig. 2B). Transfection with irrelevant GAPDH siRNA had no effect on the levels of Lon protein or StAR degradation. The reverse approach attempting to overexpress Lon led to a 2.5-fold increase of mitochondrial Lon protease (Fig. 9A), which was associated with a significantly faster degradation of StAR (Fig. 9B). These results suggest that indeed the rate of StAR degradation is directly correlated with the mitochondrial content of the Lon protease.

Fig. 9.

Effect of Lon siRNA and Lon Overexpression on Lon Protein and Activity A, Lon protein: COS cells were transfected (Materials and Methods) using either empty pCMV-5 plasmid DNA (Empty), GAPDH siRNA (siGAPDH), Lon siRNA (siLon) or human Lon expression plasmid (Lon O/Ex). All cells were also cotransfected with murine StAR plasmid added during the second round of transfection. Twenty hours later, cells were extracted for Western blot analysis (7.5 μg/lane). This experiment was repeated four times with similar results and shown is a typical chemiluminescence result of the assay. B, Effects of Lon levels on StAR degradation: COS cells were transfected with the indicated plasmids as described in A and pulse-chase experiments were conducted 20 h after the second transfection. Shown are ³⁵S-StAR levels recovered by StAR immunoprecipitation conducted at time 0 and 3 h after onset of chase. Results of three independent transfections are presented as percent of StAR level at the end of the pulse period. P < 0.05 for siLon and Lon O/Ex when compared with Empty plasmid transfection; StAR degradation in cells transfected with empty or siGAPDH plasmids did not differ significantly.

DISCUSSION

The study described here sought to reveal the mechanism of StAR turnover in mitochondria of steroidogenic cells. To this aim, we used bacteria mutant strains as a model system to identify the protease(s) involved in StAR turnover, taking advantage of the fact that the prokaryotic and mitochondrial ATP-dependent proteases are structurally and functionally conserved. This experimental series revealed that in the E. coli, Lon/La is the prime, probably the only, bacterial protease that degrades StAR. Experiments demonstrating that StAR was rapidly degraded in wild-type strains as well as in mutant strains deficient for ClpP, ClpQ, or FtsH; by contrast, StAR was almost completely stabilized in a Lon mutant strain. The exclusive role of Lon in StAR turnover was further demonstrated by expressing human ClpP in lon-cells, which did not rescue this phenotype. This result suggested that ClpP does not recognize StAR as a proteolytic substrate, a notion also corroborated by our in vitro assays of StAR degradation. Collectively, these observations suggest that Lon is solely responsible for degrading StAR in the bacterial cytoplasm.

To provide a more direct evidence in support of the role of Lon in mitochondrial StAR degradation, we tested StAR turnover in COS cells expressing reduced or enhanced Lon contents. Optimizing conditions for siRNA-mediated knockdown, or overexpression of Lon coupled with StAR overexpression, was difficult for several reasons. Lon protease is a relatively long-lived protein with a half-life greater than 24 h, as shown in murine cells (54). Thus, substantial reduction in Lon protein levels requires both breakdown of the protease and sustained siRNA expression. Because transiently transfected StAR is expressed at a high level, the period of Lon depletion was carefully optimized such that mitochondrial and cellular function were not compromised. Our previous results have shown that StAR overexpression leads to mitochondrial damage (23), and other have shown that complete Lon depletion leads to cell death (55). In the experiments performed in this study, siRNA knockdown led to a substantial (85%), but not complete, reduction of Lon protein in reasonable time frame without damaging the cell function. In addition, we have not been able to increase Lon expression above 2- to 2.5-fold in transfected cells (Orly, J., M. R. Maurizi, and C. K. Suzuki, unpublished data), suggesting that the content of Lon in the mitochondria is tightly controlled. Our combined results clearly show that the rates of StAR degradation in situ are proportionally related to the cellular content of mitochondrial Lon and that Lon-mediated proteolysis appears to be the primary mechanism for the turnover of StAR in mammalian mitochondria. However, it is possible that other mitochondrial proteases may also be involved in determining StAR levels (56). It should be noted that, in light of a potential role of phosphorylation in StAR activity (18, 19), we are currently testing whether phosphorylation of StAR by PKA may, or may not, affect the rate of StAR degradation under similar circumstances.

Consistent with the cell-based studies presented here and previously, in vitro experiments demonstrate that degradation of purified StAR by purified Lon is inhibited by MG132 and clasto-lactacystin β-lactone but not by epoxomicin. Our previous in vitro results have shown that mitochondrial Lon degrades StAR by a unique mechanism of ATP-dependent proteolysis that does not require protein substrate unfolding; Lon initiates StAR cleavage at specific solvent exposed sites that consist of hydrophobic amino acids surrounded by a highly charged environment (57). Noteworthy is the fact that the rate of the in vitro degradation of StAR by Lon was not affected by pH changes of the reaction buffer examined within the pH range of 7–9 (not shown); such control experiments were conducted because our former in vivo findings demonstrated effects of matrix pH on StAR degradation (22, 56). Therefore, it is highly likely that changes of matrix pH values (22) probably did not affect Lon-mediated catalysis directly, but rather manipulated the availability of StAR protein for Lon degradation in the matrix microenvironment (56).

By use of both cell-based and cell-free assays, the present study reveals for the first time novel inhibition of Lon by MG132 and cLβL, which are principally inhibitors of the proteasome. At 50 μm, these compounds inhibited StAR degradation more effectively than did PMSF at 1 mm; cLβL was six times more potent than MG132; epoxomicin, which is considered to be a more selective inhibitor of the proteasome (58), had no inhibitory effect on Lon-mediated StAR turnover in vitro. Selective inhibition of Lon-mediated degradation of the reporter substrate FITC-casein by MG132 and cLβL was also observed. These findings are consistent with our in vivo observations in intact steroidogenic cells, demonstrating that MG132 and cLβL, but not epoxomicin, blocked degradation of mitochondrial StAR.

In addition to this study, previous data show that peptide aldehydes such as MG132 inhibit not only the proteasome; lysosomal and Ca²⁺-activated proteases are blocked as well (59, 60). Similarly, lactacystin inhibits proteases other than the proteasome such as cathepsin A and tripeptidyl peptidase II (61). Furthermore, lactacystin has been shown to inhibit an archaeal proteasome mutant in which the catalytic threonine residue is replaced by a serine (T1S); T1S exhibits the same rate of hydrolysis of short fluorogenic peptides such as Suc-LLVY-amc as the wild-type proteasome, although a slower rate of hydrolysis is observed for larger proteins (62). Thus, it appears that the hydroxyl group of the active site serine residue within either Lon or T1S is capable of attacking the carbonyl moiety of MG132 and cLβL, which is required for protease inhibition. By contrast, epoxomicin has greater specificity for the proteasome and does not react with serine proteases such as chymotrypsin (63). The hydroxyl group of the active site serine within Lon may not be positioned properly to attack the epoxide ring in epoxomicin, which forms an ether adduct required for protease inhibition.

Our cell-free assays also showed that increasing concentrations of the casein substrate do not revert the inhibitory action of MG132 and cLβL, or do so only partially. Therefore, an irreversible binding mechanism could underlie the observed data. This conclusion is likely to hold for cLβL, which is an irreversible inhibitor of the proteasome that binds as a pseudo-substrate, becomes covalently linked to the hydroxyl groups of the active site threonine of the proteasome β-subunit, and, thereby, irreversibly inhibits both tryptic and chymotryptic activities of the protease (64). The noncompeting nature of MG132 inhibition is less obvious to explain because this peptide aldehyde (Cbz-leu-leu-leucinal) is a reversible substrate analog and transition-state inhibitor of the proteasome (64). However, it is possible that MG132 inhibits Lon by a mechanism involving an irreversible inhibition of the enzyme due to chemical interaction with the active residues such as the catalytic lysine of the Lon Ser-Lys dyad at the active site.

The rate of FITC-casein degradation by Lon in vitro showed a sigmoidal curve as a function of substrate concentration, indicating positive cooperativity. These data are consistent with observations showing that the ATPase and peptidase activities of E. coli Lon exhibit positive cooperativity in the binding and hydrolysis of substrate (51, 65). Two-component ATP-dependent proteases such as the proteasome and ClpP have also been shown to exhibit cooperative substrate binding and peptide bond hydrolysis (52, 53).

ATP-dependent proteases are central to the quality control system within mitochondria, which ensures that only properly folded and assembled proteins are present within the organelle. For example, mutations in a bacterial FtsH-like protease within the mitochondrial inner membrane of humans, called paraplegin, lead to an autosomal recessive form of hereditary spastic paraplegia (35). The absence of paraplegin assembled with the paraplegin-like protein AFG3L2 in mitochondria results in reduced levels of respiratory chain complex I and increased sensitivity to oxidative stress (36). Human Lon has been demonstrated to selectively degrade mitochondrial aconitase that carries mild oxidative changes (44). In many age-related disorders an increase in reactive oxygen species generated by oxidative phosphorylation is well documented (66), and data show that aconitase is sensitive to oxidative damage. The Lon-dependent degradation of oxidized aconitase likely functions to prevent its toxic accumulation or aggregation within the matrix. Like oxidative phosphorylation, steroid hydroxylation within mitochondria of steroidogenic cells also generates large amounts of reactive oxygen species that can be detrimental if not carefully controlled (67, 68). Data also show that oxidative stress interferes with hormone-sensitive cholesterol transfer into mitochondria (69, 70) and reduces levels of mitochondrial StAR (71). The observed reduction in the steady state levels of mitochondrial StAR may be explained either by a block in StAR import or an increase in StAR turnover. Taken together, these observations suggest that Lon may be an important defense mechanism to protect mitochondria of steroidogenic cells against oxidative damage induced by steroid hydroxylation and to maintain the level of post-functional StAR protein compatible with mitochondrial homeostasis.

MATERIALS AND METHODS

Materials

The following reagents were obtained from Sigma (St. Louis, MO): spectinomycin, FITC-casein, arabinose, PMSF, isopropyl β-d-1-thiogalactopyranoside (IPTG), epoxomicin, protein A Sepharose and methionine/cystine-free DMEM for metabolic labeling. MG132 and clasto-lactacystin-β lactone were from Calbiochem (San Diego, CA), polyethyleneimine (average molecular weight 25,000) from Aldrich (St. Louis, MO), α³²P-deoxy-CTP was purchased from Amersham (Little Chalfont, UK), ³⁵S-methionine was purchased from PerkinElmer (Shelton, CT).

Immunoreagents included a polyclonal rabbit antiserum to recombinant mouse N-47 StAR protein (preparation described below); peroxidase-conjugated goat antirabbit and lissamine-rhodamine-conjugated affiniPure goat antirabbit IgG were from Jackson ImmunoResearch Inc. (West Grove, PA). Antibodies recognizing Lon and ClpX were described before (Refs. 54 and 37 , respectively), anti-ClpP was purchased from Abcam (Cambridge, UK; ab15708) and histidine antiserum from Santa Cruz (Austin, TX) (sc-803).

Tissue culture media used for serum-free granulosa cell culture included DMEM and Ham’s F-12 nutrient media from Invitrogn Life Technologies (Paisley, Scotland, UK). The same manufacturer supplied the fetal bovine serum. Ovine FSH (NIDDK-oFSH-20) was kindly provided by NIDDK’s National Hormone and pituitary Program and A. F. Parlow (Harbor-UCLA Medical Center, Torrance, CA).

Human Lon plasmids for expression in mammalian and bacterial cells were described before (Refs. 54 and 57 , respectively). Predesigned Lon and GAPDH siRNAs (nos. 16708 and 4605, respectively) were purchased from Ambion (Austin, TX). Lon siRNA is derived from exon 8 of the human Lon gene (NM_004793); sense sequence 5′-GGAGCAGCUAAAGAUCAUCtt and antisense 5′-GAUGAUCUUUAGCUGCUCCtg. Ambion’s Silencer GAPDH siRNA sequence is derived from the 5′ medial region of the GAPDH mRNA sequence.

Animals

Intact, immature female Sprague-Dawley rats (21 d old) were obtained from Harlan (Jerusalem, Israel) and maintained under 16-h light, 8-h dark schedule with food and water ad libitum. All animal experimentation described in this manuscript was conducted in accord with accepted standards of humane animal care. All protocols had the approval of the Institutional Committee on Animal Care and Use at the Alexander Silberman Institute of Life Sciences, The Hebrew University of Jerusalem.

Visual Pulse-Chase Demonstration of StAR Degradation

Primary rat ovarian granulosa cells devoid of any steroidogenic proteins were isolated from E₂-primed prepubertal rats (72) and seeded onto glass coverslips (13 mm) placed in a 24-well plate containing serum-free medium (73). The next day, StAR expression was induced by addition of FSH (100 ng/ml). The FSH pulse period consisted of a 3-h induction with FSH, after which the cells plated on coverslips were fixed for later immunofluorescent documentation of StAR protein levels. Using replicate coverslips, the chase period was conducted by replacing the FSH-containing culture medium with fresh medium lacking hormone in the presence or absence of proteasome inhibitors for 5 h; MG132, cLβL or epoxomicin were each used at a final concentration of 20 μm. The cells that were fixed at the end of the pulse and chase periods, were immunofluorescently stained with anti-StAR serum (1:30) and Lissamine-Rhodamine-labeled goat antirabbit IgG (1:20), as described previously (5). Mounted cell monolayers were examined by confocal microcopy using a Bio-Rad (Hercules, CA) MRC-1024 workstation attached to a Zeiss (Jena, Germany) Axiovert 135M microscope equipped with a 63×/1.4 objective. The cells were excited with the 514 nm line of an argon ion laser, and the emission was detected using a 580 ± 16-nm filter. Confocal optical serial sections (0.5 μm thick) were taken through the volume of the cell. In the images shown, a pseudo color scale was applied to the section with peak levels of mitochondrial fluorescence, which was compared with the pixel values in FSH-treated cells (visual pulse; panel B), using Image-Pro (Media Cybernetics, Bethesda, MD).

Bacterial Strains and Growth Conditions

E. coli strains were grown in Luria-Bertani medium. The bacterial strains used in this work are summarized in Table 1.

Table 1.

Bacterial Strains Used in this Work

Strain	Genotype	Source
	DH5α	J. Orly collection
	Rosetta	J. Orly collection
A8751	XL2B	A. Oppenheim collection
A9891	W3110 wild type	A. Oppenheim collection
A9942	W3110 clpP::kan	P1.SG221691 × A9891
A9862	W3110 ftsH1	P1.AR7542 × A9891
A9899	W3110 clpQ::cat	P1.SG221921 × A9891
A6911	JM101	A. Oppenheim collection
A6975	JM101 lon-146::ΔTn10	P1.SG203221 × A6911

¹

Transduction by P1 phage from strains obtained from Susan Gottesman collection.

²

Transduction by P1 phage from strain obtained from Teru Ogura collection (33 ).

W3110 strains and JM101 strains were grown at 30 C. Because FtsH is an essential protease in E. coli, we tested StAR degradation in a strain expressing a temperature sensitive form of the enzyme (ftsH1). To this end, all bacteria were grown at the permissive temperature (30 C) and the culture was transferred to the restrictive (42 C) temperature 30 min before the onset of the chase period (74). DH5α and XL2B strains were used for plasmid DNA propagation.

Plasmid Construction and Expression

Mouse N-47 StAR-His.

Truncated mouse StAR cDNA lacking amino acids 1–47 of the predicted mitochondrial targeting sequence and bearing a C-terminal hexahistidine tag was constructed by PCR amplification using a pCMV-mouse-StAR plasmid as template and the following primers: forward 5′-CCAGAATTCACCATGGGTCAAGTTCGACG-3′ and Reverse 5′-ccagta tgttcgaattagtggtggtggtggtggtgtgctgcgcccattgcgccacactgggcctcagaggc-3′. The resulting N-47 StAR-His PCR fragment was purified and cloned into pDrive cloning vector (QIAGEN). StAR-His fragment was then digested with KpnI and HindIII and subcloned into pT7–7 bacterial expression vector. Expression of this vector was instrumental for the purification of N-47 StAR-His protein used in cell-free degradation assays, as well as production of StAR antiserum.

Mouse N-47 StAR.

Mouse N-47 StAR cDNA was amplified by PCR using a forward primer introducing NcoI restriction site (5′-CTGAGGCCCAGTGTTAAAAGCTTCATATTCC-3′), and a reverse primer introducing a HindIII site (5′-GGAATATGAAGCTTTTAACACTGGGCCTCAG-3′); the murine wild-type StAR pCMV plasmid (above) was used as template. This PCR fragment was digested with HindIII, partially digested with NcoI and ligated into pQE-60 (QIAGEN, Valencia, CA) vector digested with the same restriction enzymes. The N-47 StAR fragment was then excised with EcoRI and HindIII and subcloned into pBAD33 (75), or into a pQE-32 (QIAGEN, Valencia, CA) vector that was modified to confer lac I^Q.

Human N115 Lon-His.

Human Lon cDNA was cloned into the NdeI and BamHI sites of proEX. The coding sequence of human Lon cDNA begins at amino acid 115 of the precursor protein. The expressed protein contains an amino-terminal hexahistidine tag and linker—MGHHHHHHDYDIPTTENLYFQGAHMTIPDV (76).

Human N56 ClpP-His.

Human ClpP was amplified by PCR using forward 5′-CGACCCGGGGATCCGGCCGCTCATTCCCATCGTGGTGGAG-3′, and Reverse 5′-GAGGCCAAGCTTTCAGGTGCTAGCTGGGACAGGTTCTGC-3′ oligonucleotides. The resulting fragment excluded the mitochondrial targeting sequence (168 bp) and has a BamHI and HindIII sites at the 5′ and 3′ ends, respectively. The amplified fragment was digested with BamHI and HindIII and cloned into pQE-32 (QIAGEN, Valencia, CA) vector that was modified to confer lac I^Q.

StAR Expression and Degradation in E. coli

All bacterial strains were cotransformed with two plasmid vectors, pBAD33-N-47 StAR (conferring a chloramphenicol resistance) and empty vehicle pQE-32 (conferring ampicillin resistance). StAR expression was induced after an overnight growth (30 C) of W3110 or JM101 cells harboring the plasmids in LB medium containing both chloramphenicol and ampicillin. The bacteria were diluted and further grown to mid-exponential growth (A₆₀₀ nm = 0.4) before shifted to 42 C for 30 min (A₆₀₀ nm = 0.6) and induced with 0.2% arabinose for 5 min. It should be noted that induction of StAR expression with higher arabinose concentrations (over 0.5%) resulted in formation of insoluble, protease-resistant StAR protein aggregates. Also, because the ClpQ-deficient cells are chloramphenicol resistant, this strain was transformed with the IPTG inducible pQE-32-N-47 StAR expression vector conferring ampicillin resistance.

Expression of hLon (proEX-hLon) or hClpP (pQE-32-hClpP) was achieved by growing the bacteria as above, followed by induction with 20 μm IPTG for 5 min. Experiments involving the coexpression of StAR and hLon or StAR and hClpP were done by a simultaneous induction with both arabinose and IPTG.

Assay of net StAR degradation in bacteria was examined by induction of StAR synthesis (together with hLon where indicated), followed by the inhibition of protein synthesis during the chase by the addition of 1 mg/ml spectinomycin. At the indicated time points, 200-μl aliquots of cell suspension were removed and centrifuged 30 sec at maximum speed in a table-top microcentrifuge. The pellet was resuspended in 100 μl SDS-PAGE sample buffer and boiled for 3 min. Samples were resolved by SDS-PAGE on 10% gels and enhanced chemiluminescence (ECL) immunoblots procedure was preformed using StAR antiserum (preparation described below). Quantification of the ECL results was performed according to the user manual of the public domain NIH Image Program developed at the National Institutes of Health (Bethesda, MD).

Purification of Hexahistidine-Tagged Proteins

Hexahistidine tagged mitochondrial murine StAR (pT7-7-N-47 StAR-His) and hexahistidine tagged human mitochondrial Lon (proEX-Lon-His) were expressed in E. coli as described above. In brief, cells were grown at 37 C to mid-log phase, induced with 1 mm IPTG for 3 h at 30 C and harvested by centrifugation. The bacterial pellet was resuspended in solubilization buffer [50 mm HEPES (pH 8.0), 150 mm NaCl, 5 mm imidazole] containing 40% and 20% glycerol for StAR (77) and Lon purification (57), respectively. Then, the cell suspension was sonicated on ice 6 × 30 sec and centrifuged at 100,000 × g for 15 min at 4 C. The supernatant was applied to a Ni²⁺-NTA column (QIAGEN) equilibrated with solubilization buffer. N-47 StAR-His and Lon-His were eluted by washing the column in a stepwise manner with solubilization buffer containing the above mentioned respective glycerol concentrations and increasing imidazole concentrations (5, 40, 60, 80, 100, 200, and 300 mm). Fractions were analyzed by SDS-PAGE and concentration of the purified proteins was determined by Bradford assay. The majority of the recombinant proteins eluted with 80–100 mm imidazole.

Preparation of StAR Antibody

Polyclonal antiserum to StAR protein was generated by immunizing two rabbits with N-47 mouse StAR tagged with carboxy-terminal hexahistidine residues (Antibody Unit, Weizmann Institute of Science, Rehovot, Israel). Aliquots (80 μg each) of the purified protein were administered three times with 3-wk intervals; the first injection (sc) with complete Freund’s adjuvant, the second sc injection with incomplete Freund adjuvant and a third im injection of the protein in PBS. Ten days after the last boost, the rabbits were bled.

In Vitro Degradation of StAR by hLon

Mouse StAR (0.2 μg) and human Lon (0.2 μg monomer) were added to a reaction cocktail containing 25 mm Tris-HCl (pH 7.9) and 10 mm MgCl₂ in the absence or presence of 4 mm ATP. The effect of protease inhibitors was studied in the presence of 1% dimethylsulfoxide (DMSO) and 50 μm of either MG132, cLβL, epoxomicin, or 1 mm PMSF. Lon protease was preincubated in the reaction mix with the inhibitors for 5 min at 37 C before the degradation reaction commenced upon addition of StAR. At the indicated time points, aliquots (10 μl) were sampled and StAR level was determined by Western analyses as described above.

In Vitro StAR Degradation by ClpXP

ClpX and ClpP were purified as previously described (37, 57). In vitro degradation reactions (150 μl) were performed by reacting StAR (162 pmol) and Lon (60 pmol) or ClpP (252 pmol) and ClpX (222 pmol) in buffer [50 mm Tris (pH 8.0), 0.1 m KCl, 10 mm MgCl₂] containing ATP (4 mm) at 37 C. At the indicated time points, 25 μl of the reaction were removed and 5 μl of 5× SDS-PAGE reducing sample buffer was added. Samples were analyzed by immunoblotting using antibodies recognizing Lon, ClpX, ClpP, or the carboxy-terminus histidine-tagged StAR. Immunoblotting results were visualized by ECL.

In Vitro Casein Degradation

Purified human Lon (1 μg, 10 pmol monomer) was incubated for 5 min at 37 C in a 100 μl reaction cocktail [25 mm Tris-HCl (pH 8.0) and 10 mm MgCl₂] with or without 2.5 mm ATP. Then, FITC-casein substrate (2 μg) was added to initiate degradation, which results in release of free FITC molecules and increase of fluorescence (excitation 485 nm, emission 535 nm) monitored on-line (60). The effect of protease inhibitors was assessed by a 5-min preincubation (37 C) of purified human Lon with the indicated concentrations of MG132, cLβL, epoxomicin or DMSO solvent alone before addition of the FITC-casein substrate. The effect of the inhibitors on substrate-dependent Lon activity was assessed using a single inhibitor concentration (3 μm and 20 μm for cLβL and MG132, respectively) using a range of substrate concentrations (0.4–4 μm). Control cuvettes without Lon protease included a reaction cocktail with, or without ATP, 20 μm MG132 or cLβL and FITC-casein. No significant degradation of FITC-casein was observed after a 6-min incubation at 37 C.

Lon Knockdown and Overexpression

COS-1 cells were cultured in DMEM supplemented with 10% fetal bovine serum and antibiotics as described previously (78). Cells were harvested and reseeded in 24-well plates and grown to 80% confluency on the day of transfection. Two rounds of transfections were performed with a 40-h interval in between, using the polyethyleneimine transfection method as previously described (79). The first transfection included either siRNAs (100 nm) or wild-type human LON plasmid DNA (1.75 μg/ well). Control monolayers received empty pCMV DNA (2 μg/ well). The latter DNA was also used to keep the total plasmid DNA amount at 2 μg/ well. A second transfection was similarly repeated while adding wild-type murine StAR plasmid DNA (0.25 μg/ well). Twenty hours later, cells were harvested and examined by either RT-PCR (not shown), Western blot analyses, or ³⁵S-metabolic labeling and pulse-chase assays.

Metabolic Labeling and Immunoprecipitation

³⁵S-met metabolic labeling and chase experiments were performed 20 h after the second transfection with siRNAs (or Lon), followed by an immunoprecipitation procedure as previously described (22). The amount of StAR antiserum added to the radiolabeled protein extracts was 1:100 vol:vol, followed by a 2-h incubation performed in 1.5-ml Eppendorf tubes (Lumitron Electronic Instruments, Jerusalem, Israel) held on a revolving tube carousel at 4 C.

Western Blot Analysis

Expression of StAR and LON proteins was assayed by Western blot analysis using cells extracts as previously described (14). Equal amounts of total protein were separated on 10% SDS-PAGE gels and electrotransferred to polyvinylidene difluoride filters. Each blotted filter was cut in two; the upper part was analyzed with rabbit hLon antibody and the lower part was incubated with rabbit antimurine StAR serum.

Data Presentation and Statistical Analysis

StAR degradation data are presented as average percent of maximum StAR protein that was quantified by densitometric analysis. Student’s unpaired two-tailed t test was performed using Microsoft Excel 2001 statistical analysis functions. A level of P < 0.05 was accepted as statistically significant.

Acknowledgments

We thank Dr. Irene Lee (Case Western Reserve University, Cleveland, OH), Dr. Ophri Pines (The Hebrew University of Jerusalem Medical School) and Dr. Alfred L. Goldberg (Harvard Medical School, Cambridge, MA) for helpful discussions. We also thank Dr. O. Hori (Kanazawa University, Japan) who kindly provided the antirat Lon serum and Dr. Peter Bross (Århus University, Denmark) for sharing his antihuman ClpP serum. We dedicate this work in memory of Amos B. Oppenheim.

This study was supported by the Israel Science Foundation 592/03 and the United States-Israel Binational Foundation 2003/398 (to J.O.) and from the National Institutes of Health (NIH) Grant RO1 GM61095-01 and the American Heart Association (to C.K.S.). M.R.M. receives research support from the Intramural Research Program of the NIH National Cancer Institute Center for Cancer Research.

Disclosure Statement: The authors have nothing to disclose.

Abbreviations

 
• DMSO

  Dimethylsulfoxide;

 
• ECL

  enhanced chemiluminescence;

 
• IPTG

  isopropyl β-d-1-thiogalactopyranoside;

 
• P450scc

  cholesterol side chain cleavage cytochrome P450;

 
• PKA

  protein kinase A;

 
• PMSF

  phenylmethylsulfonyl fluoride;

 
• siRNA

  small interfering RNA;

 
• StAR

  steroidogenic acute regulatory.