Heat Shock Interferes with Steroidogenesis by Reducing Transcription of the Steroidogenic Acute Regulatory Protein Gene

Abstract

A key regulatory point in fine tuning of steroidogenesis is the synthesis of steroidogenic acute regulatory protein, which transfers cholesterol into mitochondria. Heat shock and toxic insults reduce steroidogenic acute regulatory protein, severely compromising steroid synthesis. As the molecular mechanisms for this reduction remain elusive, we tested the hypothesis that heat shock directly interferes with transcription of the steroidogenic acute regulatory protein gene. We show that, in mouse MA-10 Leydig tumor cells, heat shock caused drastic declines in (Bu)₂cAMP-induced progesterone accumulation and steroidogenic acute regulatory protein transcript abundance. A proximal steroidogenic acute regulatory protein promoter fragment (−85 to +39) is sufficient to direct both cAMP inducibility and heat shock inhibition. Nuclear extracts from MA-10 cells displayed binding to this proximal promoter fragment as a low mobility complex in gel shift experiments. This complex disappeared in nuclear extracts taken at 5 and 10 min after initiation of heat shock and reappeared in extracts taken at 2 and 8 h. Similar low- mobility complexes formed on oligonucleotides representing the overlapping subfragments of the minimal steroidogenic acute regulatory protein promoter fragment sensitive to the heat shock effect. Extracts from heat-shocked MA-10 cells displayed reduced complex formation to each of the subfragments. We conclude that heat shock reduces progesterone synthesis, steroidogenic acute regulatory protein mRNA abundance, and steroidogenic acute regulatory protein promoter activity and disrupts binding of nuclear proteins to the proximal region of the steroidogenic acute regulatory protein promoter. Together these observations provide strong evidence for a mechanism of transcriptional inhibition in the down-regulation of steroidogenic acute regulatory protein expression by heat shock.

CELLS RESPOND TO elevations in temperature, to organic toxins, to mediators of inflammation such as eicosanoids and cytokines, and to a number of environmental contaminants, including heavy metals, by the activation of heat shock factors (reviewed in Ref. 1). This, in turn, results in rapid and transient transcription of a family of genes encoding what are known as heat shock proteins. Many of these proteins have constitutive functions in cells, including those of the heat shock protein 90 (HSP-90) family, which are components of steroid hormone receptor complexes (2) and heat shock protein 70 (HSP-70), which appears to be involved in intracellular protein translocation (3). Most significantly, heat shock proteins ameliorate cellular perturbation by reversing the deleterious effects of temperature elevation on the synthesis, folding, and translocation of nascent proteins (4).

Steroidogenic cells respond to heat shock (5, 6) as well as to inhibitors of steroid synthesis such as PGF2α, ionomycin, and TNFα (5) and to other cytokines (7) by rapid elevation in HSP-70 synthesis, accompanied by reduction in progesterone production. Reduction in HSP-70 expression by cognate antisense oligonucleotides reverses both thermal and PGF2α-induced inhibition of progesterone synthesis in rat luteal cells (5), suggesting that HSP-70 mediates heat shock effects in these cells. Heat shock impairment of steroid synthesis in ovarian granulosa cells is also fully reversed by 5-cholestane-3β, 22[R]-diol, which can permeate both cells and mitochondria (8). This finding implicates the steroidogenic acute regulatory protein (StAR), the protein responsible for transport of cholesterol across mitochondrial membranes (Ref. 9 and reviewed in Ref. 10) as the target of heat shock and other insults (11) to steroidogenic cells. Indeed, the cAMP induction of both StAR protein and progesterone synthesis in mouse MA-10 Leydig tumor cells is gravely compromised by brief elevation of temperature [10 min at 45 C (6)]. The synthesis of other proteins essential to progesterone synthesis, cytochrome P450 side chain cleavage (P450scc), and 3β-hydroxysteroid dehydrogenase (3βHSD) is only slightly altered by temperature elevation (6), further indicating that StAR is a principal target of thermal elevation.

StAR expression is dramatically reduced during both natural and PGF2α-induced luteal regression (11), as well as by signals as diverse as lipopolysaccharide, TNFα, atrial natriuretic peptide, TGFβ, and interferon-γ (Ref. 12 for review). As luteal regression and heat shock appear to function by common mechanisms (13), considerable physiological information can be derived from understanding their mode of action. Recent investigations addressing the means by which stimulatory ligands such as LH and its intracellular signal, cAMP, direct StAR gene transcription have shown that the region in the first 105 bp upstream of the transcription start site confers cAMP sensitivity to the StAR gene (14, 15). In the current investigation, we demonstrate that heat shock interferes with StAR expression at the transcriptional level. The heat shock effect acts on the minimal region of the promoter responsible for basal and cAMP-inducible expression of StAR.

RESULTS

Heat Shock Reduces Steroidogenesis and StAR mRNA Abundance

MA-10 cells subjected for 10 min to 45 C heat shock had compromised steroidogenic capability, as indicated by the drastic reduction (>20-fold, P < 0.001) in progesterone accumulation, following 3 or 6 h of (Bu)₂cAMP stimulation (Fig. 1). As expected, treatment of MA-10 cells with 1 mm (Bu)₂cAMP increased the accumulation of StAR, P450, and 3βHSD transcripts at all times tested (1, 3, and 6 h, P < 0.05) (Figs. 1 and 2). Treatment at 45 C for 10 min strikingly reduced StAR transcript induction in response to (Bu)₂cAMP at 1 and 3 h after the 2 h recovery period (Figs. 1 and 2). The capacity to accumulate StAR mRNA appeared to be reconstituted at later times, as indicated by detectable signals at 6 h after the 2 h postheat shock recovery period. P450scc and 3βHSD were reduced by heat shock, but displayed more robust recoveries by 3 h. Transcript levels for these two genes in control and heat-shocked cultures were similar after 6 h of (Bu)₂cAMP stimulation. Heat shock did not reduce the abundance of transcripts for either of the mouse Niemann Pick C-1 (NPC-1) and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) genes (Fig. 2).

Figure 1.

Effect of Heat Shock Treatment on Accumulation of Progesterone and mRNA for StAR, P450scc, and 3βHSD Cultures of MA-10 cells were exposed to 45 C for 10 min, and then allowed to recover for 2 h, at which time the medium was replaced and cultures stimulated with 1 mm (Bu)₂cAMP. Incubations were terminated at 0, 3, and 6 h after the end of the recovery period. Control cultures received no heat shock, but were likewise treated with cAMP for 0, 3, or 6 h. Progesterone values represent the mean ± sem of five replicate experiments. The Northern blots shown are from a single experiment repeated five times with similar results.

Figure 2.

Specificity of Heat Shock Inhibition of StAR, but Not Niemann-Pick C1 (NPC-1) nor GAPDH mRNA Abundance MA-10 cells were subjected to heat shock and (Bu)₂cAMP stimulation as described in the legend to Fig. 1. Consequent Northern blots were probed with the homologous mouse StAR cDNA probe, as well as the mouse NPC-1 and GAPDH probes.

Heat Shock Reduces Transcription by the StAR Promoter

To ascertain whether heat shock produces down-regulation of StAR at the transcriptional level, we employed a series of five StAR promoter constructs linked to the luciferase reporter gene in transient transfection experiments. All of these constructs, when transfected into MA-10 cells, displayed a detectable signal in the absence of (Bu)₂cAMP stimulation and showed significant increases in response to incubation with 1 mm (Bu)₂cAMP (Fig. 3). ANOVA revealed no significant variation in mean fold increase in promoter activity induced by (Bu)₂cAMP among the five constructs. Importantly, the 10 min treatment at 45 C reduced both constitutive and (Bu)₂cAMP induction of the luciferase signal for each of the five constructs (P < 0.01). The mean inhibition of (Bu)₂cAMP induction of luciferase transcription ranged from 95 to 99% for the five constructs. The promoter-less reporter plasmid, pGL2-basic, displayed neither constitutive nor cAMP-inducible activity (not shown).

Figure 3.

Heat Shock Interferes with Basal and (Bu)₂cAMP-Induced StAR Promoter Expression MA-10 cells were transfected with five deletion constructs of the human StAR promoter luciferase, beginning from −1.3 kb to −85 bp. Control represents the luciferase activity of each StAR deletion construct in nonstimulated cells. Cultures designated HS were subjected to the heat shock experimental paradigm, 10 min at 45 C. Following 2 h recovery, cultures designated dbcAMP were incubated with 1 mm (Bu)₂cAMP for 6 h followed by determination of luciferase activity. Results were corrected forβ -galactosidase activity. Average control luciferase activity values for each of five experiments were set to 1.0, and fold increase values were calculated and presented as mean ± sem of the five replicates.

Heat Shock Interferes with Binding of Nuclear Proteins to the StAR Promoter

To study the molecular events that are coupled to the transcriptional inhibition elicited by heat shock, we undertook an investigation of nuclear protein binding to the minimal promoter fragment responsive to the treatment. EMSAs were performed on the promoter element −85 to +39. Nuclear extracts from untreated MA-10 cells bound to this StAR promoter sequence, generating a low-mobility complex (Fig. 4). Complex formation was drastically reduced in extracts from cells treated for 5 min at 45 C, a reduction that was more pronounced in cells treated for 10 min. By 2 h, appearance of the low-mobility complex had partially recovered, and recovery continued through 8 h after heat shock. Conversely, nuclear extracts from heat-shocked MA-10 cells conserved their capacity to bind a canonical cAMP response element (CRE) sequence at all times after heat shock (Fig. 4).

Figure 4.

EMSA Showing the Effects of Temperature Elevation to 45 C on Binding of Nuclear Extracts to the ³²P-Labeled StAR Promoter Fragment Comprising Nucleotides from −85 to +39 of the Human StAR Promoter (left) The times at which cells were harvested after initiation of the 10-min heat shock treatment are indicated. A low-mobility complex formed with the CRE oligonucleotide (right), which served as a control for binding activity of the nuclear extracts. In the lower panel, results are expressed as the ratio of the optical density between binding complexes for the two probes at each time point.

We then wished to investigate in more detail the binding of nuclear factors from heat-shocked MA-10 cells to the minimal StAR promoter fragment sensitive to heat shock. To this purpose, the StAR promoter sequence was subdivided into four overlapping oligonucleotide sequences, designated O1–O4 (Fig. 5). These oligonucleotides displayed similar formation of low-mobility complexes with nuclear extracts from untreated MA-10 cells (Fig. 5). Nuclear extracts also formed a series of higher mobility complexes with all of the oligonucleotides (Fig. 5). As seen with the −85 to +39 fragment (Fig. 4), formation of the lowest mobility complex in MA-10 nuclear extracts with each of the oligonucleotides was reduced in the extracts of cells treated for 5 min at 45 C. Complex formation recovered over the next 8 h (Fig. 5). Inspection of autoradiograms suggested that loss of the low-mobility complex was accompanied by increases in binding in higher mobility bands, particularly for oligonucleotide sequences O1 and O2. The oligonucleotide sequence designated O3, which contains the TATA box and the proximal steroidogenic factor 1 (SF-1) site (16), consistently displayed the least binding to the low-mobility complex (Figs. 5 and 6), and this was extinguished at 5 and 10 min after initiation of heat shock. Partial recovery of binding to all four oligonucleotides in higher mobility complexes was observed for extracts of cells at 8 h after heat shock (Fig. 5).

Figure 5.

EMSA Showing Binding of Nuclear Extracts from MA-10 Cells to Four Overlapping Double-Stranded Oligonucleotides (O1–O4) from the Mouse StAR Promoter (−85/+14) The alignment of these to the human StAR promoter fragment is depicted in the lower panel. As above, 0 designates nuclear extracts made from cells before heat shock, 5 and 10 min indicate the periods during heat shock, and 2 h and 8 h are the recovery period times following 10 min of heat shock.

Figure 6.

EMSA Showing Specificity of Formation of Complexes with MA-10 Nuclear Extracts and Four Overlapping Double- Stranded Oligonucleotides (O1–O4, Fig. 4) from the Mouse StAR Promoter (−85/+14) Extracts were incubated in the presence of each labeled probe and previously titrated 50- fold excess concentrations of each unlabeled oligonucleotide, as well as with 50-fold excess of the irrelevant oligonucleotide, HST. It can be seen that formation of the low- mobility complexes (arrows) is suppressed by the corresponding unlabeled nucleotide.

To assess the specificity of oligonucleotide binding to MA-10 nuclear complexes, we incubated each of the O1–O4 double-stranded sequences with cold oligonucleotides in an amount (50-fold) known to be in excess by previous titration. The results at time 0 (before heat shock, Fig. 6) indicate that each oligonucleotide interacts with at least two complexes, as in Fig. 5. Further, binding of each labeled sequence to the low-mobility complex was specific, in that it was not altered by an excess of an unrelated oligonucleotide, heat shock transcription factor (HST). The low-mobility complex formed by nuclear extract binding to O2 was displaced by an excess of cold O2, but not by any of the other nucleotide sequences, indicating specificity of this binding. Labeled O1 was readily displaced on the low-mobility band by an excess of cognate oligonucleotide, and binding was reduced to a lesser degree by incubation with O4. The low-mobility complex formed by nuclear extract binding to O3 was the most readily reduced by heat shock (Fig. 5) and appeared to be displaced, not only by an excess of unlabeled O3, but also by other oligonucleotides, particularly O1 and O4. The binding of O4 to the low-mobility complex was entirely displaced by excess O4, and reduced by O1–O3. Our principal focus was the low-mobility complex, as it is similar to that which bound to the −85 to +39 promoter fragment (Fig. 4), and complex formation was reversibly altered by the heat shock treatment (Fig. 4) in a manner reminiscent of the effects on StAR message abundance (Fig. 1). Fig. 6 reveals that the higher mobility complexes that formed between the nuclear extract and oligonucleotides O1–O4 were displaced by HST. We further observed high-mobility complex cross-displacement by O1–O4, indicating that oligonucleotide binding to the proteins in these bands was not specific to the oligonucleotide sequence in question.

The same analysis was then performed employing labeled and unlabeled probes and nuclear extracts taken at the same intervals after initiation of heat shock as shown in Figs. 4 and 5. As in Figs. 4 and 5, binding to the lowest mobility complexes was compromised by heat shock, and inhibition of complex formation by unlabeled probes recapitulated the results at time 0 (data not shown). The most profound suppression of complex formation and the least conspicuous recovery of low-mobility complexes after heat shock occurred with sequence O3 (data not shown). As above, partial recovery of low-mobility complex formation occurred at 8 h for all four of the oligonucleotide fragments. As with the CRE oligonucleotide described above, high-mobility complex formation with labeled HST was not perturbed by heat shock treatment (data not shown).

Heat Shock Effects Are Temperature and Cell Type Specific

Incubation of MA-10 cells for 10 min at 43 C in lieu of 45 C did not reduce constitutive nor cAMP-induced activity of the 1.3-kb StAR promoter-luciferase construct (Table 1). In addition, it had no effect on progesterone accumulation in response to cAMP (Table 1). Treatment of Y-1 mouse adrenal tumor cells for 6 h with (Bu)₂cAMP resulted in modest StAR promoter stimulation (Table 1). Nonetheless, incubation of cells at 45 C for 10 min had no effect on the promoter activity of the 1.3-kb construct and on the accumulation of 20α-dihydroprogesterone by cultures (Table 1) and produced only a mild reduction in StAR protein at 6 h postrecovery (Fig. 7).

Figure 7.

Western Blot Showing StAR Protein Abundance in Y-1 Mouse Adrenal Tumor Cells and MA-10 Leydig Cells Subjected to No Temperature Elevations (Control) or 10 min at 45 C (Heat Shock) Mitochondrial protein extracts were purified from cells harvested at 3 and 6 h after heat shock treatment and analyzed by Western blot.

Table 1.

Absence of Response of MA-10 Cells to 43 C Heat Shock and Mouse Adrenal Y-1 Cells to 45 C Heat Shock

	MA-10 37 C	MA-10 43 C	MA-10 45 C	Y-1 37 C	Y-1 45 C
Constitutive RLU	6,634 ± 1,082	5,936 ± 1,884	4,146 ± 554	23,455 ± 3,364	21,822 ± 2,281
Fold change in promoter activity in response to 1 mm (Bu)2cAMP	6.9 ± 4.3	8.1 ± 3.0	0.83 ± 0.1	1.6 ± 0.4	1.7 ± 0.3
Steroid accumulation	598 ± 38	522 ± 145	213 ± 3	6,214 ± 137	6,333 ± 108
	(ng/ml) medium	(ng/ml) medium	(ng/ml) medium	(ng/mg protein)	(ng/mg protein)
	P4	P4	P4	DHP	DHP

Concentrations of progesterone (P4) and dihydoprogesterone (DHP) were measured by RIA. RLUs are from replicates within a typical experiment. Fold increases are mean ± sd of two replicate experiments.

DISCUSSION

It has been shown that gonadal cells subjected to heat shock have impaired steroidogenic capability in response to stimulatory ligands and second messengers (6–8). In the present study, we confirm that brief incubation at 45 C reduced the capacity of MA-10 cells to accumulate their principal steroid product, progesterone, in response to cAMP stimulation. It has been shown that heat shock effects on steroidogenesis can be entirely reversed by supplying cells with soluble hydroxylated cholesterol, implicating StAR and the mitochondrial transport of cholesterol as the key step impaired by heat treatment (6–8). Herein we present evidence indicating that heat shock impairs expression of the StAR protein by interference with the transcription of the StAR gene.

Incubation of MA-10 cells at 45 C for 10 min reduced the abundance of StAR mRNA in a pattern consistent with the known decrease in StAR protein levels in these cells (6). In other cell systems, it has been shown that the half-life of StAR transcripts is brief (17). Likewise, the activity of StAR protein is rapid and transient (reviewed in Ref. 10). The present investigation shows that heat shock induces a transient reduction in the abundance of transcripts for 3βHSD and P450scc, in keeping with the modest declines in enzyme protein abundance associated with this treatment (6). As the half-life of enzyme activity for 3βHSD and P450scc is in the order of days (18, 19), it is unlikely that the minor and transitory reduction in their abundance following heat shock contributes to the loss of steroidogenic capacity. The specificity of the heat shock is further demonstrated by the absence of an effect on a constitutive protein, GAPDH, and on a protein believed to be involved in intracellular organelle-related cholesterol transfer, the NPC-1 protein (20).

Evidence for a transcriptional effect of heat shock on StAR expression is provided by the sensitivity of basal and (Bu)₂cAMP-induced transcription of StAR-luciferase to brief heat shock. All StAR promoter deletion constructs tested here displayed increases in luciferase signal in MA-10 cells in response to (Bu)₂cAMP. The magnitude of response was similar to their activity in BeWo cells cotransfected with an SF-1 expression plasmid (21) and in primary cultures of human theca and granulosa cells (15). Similar results have been seen in homologous systems, i.e. mouse StAR reporter deletion constructs in MA-10 cells (14) and of bovine StAR constructs in primary cultures of bovine theca cells (22). Suppression of cAMP effects by heat shock was observed in all of the deletion constructs employed, including the sequence that contains the 85 nucleotides (nt) upstream of the transcription start site. This region is highly conserved in all StAR genes known to date (12). MA-10 nuclear extracts formed a low-mobility complex with this conserved region of the StAR promoter in gel retardation experiments. Analysis of this region by means of overlapping oligonucleotide fragments revealed that MA-10 nuclear extracts formed similarly migrating low-mobility complexes with each of the fragments, in addition to the formation of one or more higher mobility complexes. Heat shock disrupts the formation of low-mobility complexes between MA-10 nuclear extracts and the StAR promoter −85/+39 sequence. Changes occurred quickly, as reduction in binding was seen as early as 5 min after the initiation of elevation of temperature. Binding changes were transient and reversible, as significant binding capability was recovered in extracts at 8 h after heat shock. The mechanisms of alteration of nuclear protein binding to DNA by heat shock are currently undefined. It has previously been reported that transient heat shock denatures and aggregates nuclear proteins (Ref. 23 and references therein). Indeed, redistribution of nuclear kinases associated with transcription follows heat shock in HeLa cells (24). It is therefore possible that reversible redistribution of nuclear proteins acting on the StAR promoter can explain the present observations.

The nature of the MA-10 nuclear proteins that bind to the −85/+39 fragment of the StAR promoter is not currently known. Nonetheless, this region bears consensus sites that have been previously shown to bind a number of protein factors. Oligonucleotide O3, used in the present study, contains a sequence for binding to the orphan nuclear receptor SF-1 (16), and has a TATA box sequence, known to bind the transcription factors found in the multisubunit basal transcription complexes (reviewed in Ref. 25). Nonetheless, this reasoning does not account for the nearly identical binding patterns of the remaining three oligonucleotide fragments (O1, O2, and O4). An alternative explanation for these findings is that the low-mobility complexes comprise a series of proteins distributed along the entire− 85 to +39 region of the promoter. Heat shock may then cause the disassembly of the complex that binds this region, as well as of the more mobile complexes that bind to the individual oligonucleotide fragments that constitute this region of the promoter.

It has previously been shown that the StAR promoter is preferentially active in MA-10 and Y-1 cells, compared with nonsteroidogenic cell lines (14). In the current investigation, steroid and StAR protein accumulation, as well as luciferase transcription driven by the StAR promoter, appeared unaffected by heat shock treatment of Y-1 cells. This is in stark contrast to the results in MA-10 cells, and the molecular reasons for this are currently obscure. It is known that heat shock, as employed in the current experiment, causes elevation in HSP-70 in MA-10 cells (6), but not in Y-1 cells (Liu, Z., and D. M. Stocco, unpublished observations). Further, HSP-70 may be the mediator of heat shock effects on steroidogenesis in luteal cells (5). Thus, the absence of heat shock protein response may explain the insensitivity of Y-1 cells to thermal elevation.

In summary, we have shown that heat shock interference with steroidogenesis in MA-10 cells is due, at least in part, to interference with transcription of StAR. The evidence for this conclusion includes heat shock-induced reduction in StAR mRNA expression, reduction in StAR promoter activity, and disruption of binding of nuclear proteins to the proximal region of the StAR promoter.

MATERIALS AND METHODS

Cell Lines, Cell Culture, and Pharmacological Treatments

The mouse MA-10 Leydig tumor cell line, a gift of Dr. Mario Ascoli (26) was cultured in Waymouth’s MB/752 medium supplemented with 15% horse serum according to procedures previously described (6). The Y-1 mouse tumor adrenocortical cell line (ATCC, Manassas, VA) was maintained in DMEM/Ham’s F-10 medium (Sigma, St. Louis MO), supplemented with 15% horse serum and 2.5% FCS as previously described (27).

The experimental paradigm comprised a heat shock treatment for 10 min in a water bath at 45 C, followed by a 2-h recovery period at 37 C and stimulation for 6 h with (Bu)₂cAMP (1 mm, Sigma, St. Louis MO). This treatment has previously been shown to reduce the abundance of StAR protein (6). The temperature dependence of the heat shock effect was examined by employing 43 C instead of 45 C. In further experiments, cultures were terminated at 0, 5, and 10 min and 2 and 8 h after initiation of the 45 C heat treatment.

Plasmids and Transient Transfections

Reporter plasmids containing deletion constructs of the human StAR promoter region comprising the 1.3-kb (−1,300 to +39) HindIII fragment, and fragments spanning the regions from− 885 to +39, −235 to +39, −150 to +39, and −85 to + 39 (21) were the generous gift of Dr. J. F. Strauss III. The promoter-less plasmid pGL2 basic (Promega Corp., Charbonnieres, France) was employed as negative control, and plasmid pCH110 (Pharmacia Biotech, Lyon, France), which constitutively expresses β-galactosidase, was employed to determine efficiency of transfection. MA-10 cells were seeded into six-well culture plates and the lipofectamine reagent (Life Technologies, Inc., Pointoise, France) was employed for transfection as previously described (9). In a typical experiment, 4 μg pGL2-based StAR or promoter-less plasmid and 1 μg pCH110 were transfected with 10 μl per well lipofectamine for 6 h under serum- and antibiotic-free conditions, and then in medium containing serum and antibiotics overnight. Treatments were initiated on the following morning. The calcium phosphate transfection technique was employed for overnight transfection of Y-1 cells, as described previously (27). Luciferase and β-galactosidase were assessed by standard techniques (27), and results were normalized for β-galactosidase activity.

Northern Analysis

Total RNA was isolated from cells using the Trizol reagent (Life Technologies, Inc.) according to the manufacturer’s specifications. Aliquots of 15 μg total RNA were separated on 1% agarose-formaldehyde gels, and then blotted overnight onto nylon membranes, baked, and hybridized with cDNA probes. The bovine P450scc cDNA (1.3 kb) was obtained from Dr. Michael Waterman (Vanderbilt University, Nashville, TN). The mouse 3β-HSD I cDNA was provided by Dr. Anita Payne (Stanford University, Stanford, CA). Probes for the Niemann-Pick C1 (0.8 kb) and GAPDH (0.8 kb) were generated from the mouse liver by RT-PCR using primers designed from sequences deposited in GenBank. The human 28S rRNA probe, obtained from Dr. G. Schultz (University of Calgary, Calgary, Canada), was employed as a control for loading and transfer. Estimates of mRNA abundance were made by PhosphorImager (Amersham Pharmacia, Baie d\\'Urf|$$|Aae, Canada) quantification as the dimensionless ratio between the mRNA of interest and 28S rRNA.

Western Blots

The abundance of StAR protein was assessed in Y-1 and MA-10 cells harvested from culture using a polyclonal antibody according to the procedures previously described (28).

Hormone Assays

Progesterone and 20α-dihydroprogesterone were assayed in the culture medium by solid-phase RIA using kits from Diagnostic Products (Los Angeles, CA), according to the manufacturer’s specifications.

EMSA

Nuclear extracts were prepared from MA-10 cells harvested before and at 5 and 10; 2 and 8 h after the initiation of heat shock. The extracts were prepared as previously described (29).

Initial EMSA was performed using the most abbreviated StAR promoter fragment described above (−85 to +39 nt), excised by restriction digestion from pGL2 −85/+39 StAR luciferase plasmid. Subsequent analyses employing double- stranded oligonucleotides (designated oligonucleotides O1–O4) comprised four overlapping stretches of the corresponding region (−85/+14) of the mouse StAR promoter. The sequences of the oligonucleotides (upper strand) used in this study were:

Oligonucleotide O1: GACCCTCTGCACAATGACTGATGACTTT

Oligonucleotide O2:

Oligonucleotide O3 : GCACAGCCTTCCACGGGAAGCATTTAAG

Oligonucleotide O4: TTAAGGCAGCGCACTTGATCTGCGCCACAGCTGCAGGACTCAGG.

A double-stranded oligonucleotide containing a canonical CRE sequence present in the inducible cAMP early repressor promoter (30) was employed to assess binding ability of nuclear extracts. In the general scheme, nuclear extract (5 μg protein) was incubated with 2 × 10⁴ cpm labeled probe in binding buffer (20 mm Tris-HCl, pH 7.5, 80 mm KCl, 1 mm EDTA, 0.1 mm dithiothreitol, 0.1 mg/ml BSA, 2 mg poly dI-dC, 10% glycerol) for 20 min at room temperature, before application to a 4% nondenaturing polyacrylamide gel in 0.25× Tris-buffered EDTA. To establish the specificity of binding of the nuclear extract to the double-stranded oligonucleotides, an excess of each unlabeled oligonucleotide was incubated with nuclear extract and with its ³²P-labeled counterpart and with nuclear extract in the presence of each of the other three oligonucleotides and an irrelevant oligonucleotide sequence [HST (31)] known to bind to MA-10 nuclear extracts. The experiments were then repeated using nuclear extracts harvested at 5 and 10 min and 2 and 8 h after initiation of 10-min heat shock as described above.

Acknowledgments

We thank Drs. J. F. Strauss III, M. Waterman, A. Payne, and G. Schultz for the gift of reagents. The technical assistance of Maryam Rastegar, Estelle Heitz, and Mira Dobias is greatly appreciated.

This work was supported by grants from Centre Nationale de la Recherche Scientifique, INSERM, Center Hospitalier Universitaire Régional, Fondation de la recherche médicale, Rh|$$|Axone-Poulenc Rorer, Inc. (Bioavenir, France) and Association pour la recherche contre le cancer and by Canadian Institutes of Health Research Grant MT-11018 to B.D.M.

These authors contributed equally to the work.

Current address: Department of Biology, Eastern New Mexico University, Portalis New Mexico.

Current address: Hormone Research Center and Department of Biology, College of Natural Sciences, Chonham National University, Kwangju 500–757 Korea.

Abbreviations

 
• CRE

  cAMP response element;

 
• GAPDH

  glyceraldehyde-3-phosphate dehydrogenase;

 
• 3βHSD

  3β-hydroxysteroid dehydrogenase;

 
• HSP-70

  heat shock protein 70;

 
• HSP-90

  heat shock protein 90;

 
• HST

  heat shock transcription factor;

 
• NPC-1

  Niemann Pick C-1;

 
• P450scc

  cytochrome P450 side chain cleavage;

 
• SF-1;

  steroidogenic factor 1;

 
• StAR

  steroidogenic acute regulatory protein.