Hormone-Sensitive Lipase Is Required for High-Density Lipoprotein Cholesteryl Ester-Supported Adrenal Steroidogenesis

Abstract

Steroid hormones are synthesized using cholesterol as precursor, with a substantial portion supplied by the selective uptake of lipoprotein-derived cholesteryl esters. Adrenals express a high level of neutral cholesteryl ester hydrolase activity, and recently hormone-sensitive lipase (HSL) was shown to be responsible for most adrenal neutral cholesteryl ester hydrolase activity. To determine the functional importance of HSL in adrenal steroidogenesis, adrenal cells were isolated from control and HSL−/− mice, and the in vitro production of corticosterone was quantified. Results show that, even though adrenal cholesteryl ester content was substantially elevated in both male and female HSL−/− mice, basal corticosterone production was reduced approximately 50%. The maximum corticosterone production induced by dibutyryl cAMP, and lipoproteins was approximately 75–85% lower in adrenal cells from HSL−/− mice compared with control. There is no intrinsic defect in the conversion of cholesterol into steroids in HSL−/− mice. Dibutyryl cAMP-stimulated conversion of high-density lipoprotein cholesteryl esters into corticosterone was reduced 97% in HSL−/− mice. An increase in low-density lipoprotein receptor expression appears to be one of the compensatory mechanisms for cholesterol delivery in HSL−/− mice. These findings suggest that HSL is functionally linked to the selective pathway and is critically involved in the intracellular processing and availability of cholesterol for adrenal steroidogenesis.

ACTH STIMULATION EXERTS both acute and chronic effects on the regulation of steroid hormone production. The acute response is characterized by a rapid increase in the rate of steroid hormone biosynthesis as a result of the induction of steroidogenic acute regulatory protein (StAR) (1) and delivery of cholesterol to mitochondrial cholesterol side chain cleavage enzyme (CYP11A1) (2). Additionally, there is an increase in the expression of enzymes of the steroidogenic pathway (3). The cholesterol used for steroidogenesis can potentially be derived from a combination of sources (4): 1) de novo cellular cholesterol synthesis, 2) lipoprotein-derived cholesteryl esters obtained by either receptor-mediated endocytic or selective cellular uptake, and 3) the mobilization of stored cholesteryl esters via the actions of neutral cholesteryl ester hydrolase (CEH).

Although cellular cholesterol synthesis can supply adequate amounts of cholesterol substrate to support steroidogenesis, it would not appear to be optimal because ACTH-stimulated steroid production is markedly augmented in the presence of lipoproteins (5). Lipoprotein-derived cholesteryl esters can be delivered by receptor-mediated endocytic uptake in which low-density lipoprotein (LDL), or other apolipoprotein B- or apolipoprotein E-containing lipoproteins, are processed via the LDL receptor (6). Whereas LDL receptor-mediated uptake of cholesterol allows for efficient delivery, steroidogenic cells are able to process exceptionally large amounts of lipoprotein-derived cholesteryl esters through the use of a specialized pathway known as the selective cholesteryl ester uptake pathway. The term selective cholesterol uptake is used when surface-associated cholesterol-rich lipoproteins [high-density lipoprotein (HDL) or LDL, regardless of apolipoprotein composition] release cholesteryl esters directly into cells without internalizing the lipoprotein particle itself (7–10). Cells of the trophic hormone-stimulated rodent adrenal rely heavily on selective cholesteryl ester uptake to fulfill their cholesterol needs (7–13). Although the molecular mechanism(s) by which cells can selectively internalize large quantities of a neutral lipid are not yet clear, it appears that the initial, or extracellular phase of the process, involves lipoprotein binding to the scavenger receptor, class B type I (SR-BI) found on the surface of a variety of cholesterol-requiring cells, especially those involved in steroid hormone production (9, 10, 14–16). SR-BI is associated with tissues and/or cells that utilize large quantities of cholesterol; hormone treatment or other factors which increase the demand for cholesterol also increase the expression of SR-BI and the influx of lipoprotein cholesteryl esters (9, 10, 14–16). Deletion of the SR-BI gene in mice resulted in increased circulating levels of HDL cholesterol, substantially reduced stored tissue cholesterol (17), and inhibited the selective uptake, storage, and utilization of cholesterol by steroid hormone-producing cells. The mechanism by which cholesteryl esters are selectively transferred to cells by SR-BI is not fully understood, but it has been proposed that SR-BI forms a hydrophobic channel through which the cholesteryl esters in SR-BI-bound HDL move down a concentration gradient into the plasma membrane (18). Once selectively delivered to the cell, the cholesteryl esters must be hydrolyzed by nonlysosomal cholesteryl ester hydrolases before the free cholesterol can be transferred to mitochondria for steroid production (10). The identity of and the mechanisms of the participation of the nonlysosomal cholesteryl ester hydrolase(s) that processes cholesteryl esters derived by selective uptake have heretofore been unclear.

Recently we reported that inactivation of hormone-sensitive lipase (HSL) resulted in the loss of more than 98% of neutral CEH activity in both male and female mouse adrenals (19). Loss of neutral CEH activity by inactivation of HSL resulted in a reduction in corticosterone response to ACTH that was highly significant in female mice. The current studies were undertaken to test the hypothesis that HSL is an essential component of the selective pathway and that it plays an indispensable role in the cellular processing of lipoprotein-derived cholesteryl esters for glucocorticoid production.

RESULTS

To explore the effects of HSL deficiency on steroid hormone production, adrenal cells were isolated from control and HSL−/− mice, and the in vitro production of corticosterone after ACTH (10 ng/ml) stimulation was examined after 3 h (Table 1). Basal (absence of ACTH) corticosterone production from adrenal cells was reduced approximately 50% in HSL−/− mice compared with control (P < 0.001), whereas female adrenals produced approximately twice as much corticosterone as males in both control and HSL−/− mice (P < 0.05). Exposure of adrenal cells to ACTH increased corticosterone production 6- to 8-fold in male and female control mice, but only approximately 3- to 5-fold in HSL−/− mice. Therefore, corticosterone production induced by ACTH was approximately 60–75% lower in adrenal cells from HSL−/− mice compared with control (P < 0.001). In the absence of lipoproteins, corticosterone production should be maintained by endogenous cholesterol synthesis and/or mobilization of cellular cholesteryl ester stores. Figure 1 displays the cholesteryl ester stores of adrenals from control and HSL−/− mice. The absence of neutral CEH activity in the adrenal of HSL−/− mice has profound effects on cellular cholesterol homeostasis. Adrenal cholesteryl ester content was increased 4-fold in male and 2-fold in female HSL−/− mice when compared with the respective control mice (P < 0.001). Adrenal cholesteryl ester content was higher in female, as compared with male mice, and this was true in control mice as well, where it is consistent with the lower neutral CEH activity observed in control female mice (19).

Fig. 1.

Adrenal Cholesteryl Ester Content in Control and HSL−/− Mice Results are the mean ± sem of n = 4 animals in each group. ^A, P < 0.001 compared with control; ^B, P < 0.01 compared with male mice.

Table 1.

Effect of ACTH on Corticosterone Secretion by Adrenocortical Cells

	Basal	ACTH
Corticosterone (ng/mg protein·3 h)
Control
Male	23 ± 3	147 ± 16
Female	48 ± 4a	306 ± 17a
HSL−/−
Male	10 ± 2b	55 ± 9b
Female	19 ± 3ab	77 ± 16b

Adrenocortical cells were isolated from control, wild-type, and HSL−/− mice as described in Materials and Methods. Cells were incubated in the absence or presence of ACTH (10 ng/ml) for 3 h and the amount of corticosterone secreted into the media measured. Results are the mean ± sem of four independent experiments.

^a

, P < 0.05 compared with male mice;

^b

, P < 0.001 compared with control mice.

Because a substantial portion of corticosterone production is derived from lipoprotein cholesterol, corticosterone production after dibutyryl cAMP (Bt₂cAMP) (2.5 mm) stimulation in the presence or absence of HDL (500 μg protein/ml) was examined in adrenal cells isolated from control and HSL−/− mice after 24 h. As shown in Table 2, basal (absence of Bt₂cAMP and HDL) corticosterone production from adrenal cells was reduced approximately 30–50% in HSL−/− mice compared with control (P < 0.001), whereas female adrenals produced approximately twice as much corticosterone as males in both control and HSL−/− mice (P < 0.01). Exposure of adrenal cells to Bt₂cAMP alone increased corticosterone production 6- to 8-fold in male and female control mice, but only approximately 2-fold in HSL−/− mice. Therefore, corticosterone production induced by Bt₂cAMP was approximately 75–80% lower in adrenal cells from HSL−/− mice compared with control (P < 0.001). Exposure of adrenal cells to HDL had no effect on corticosterone production in either control or HSL−/− mice, but the combination of Bt₂cAMP and HDL increased corticosterone production 11- to 13-fold in male and female control mice, but only approximately 4- to 5-fold in HSL−/− mice. Therefore, maximum corticosterone production induced by Bt₂cAMP and HDL was approximately 75% lower in adrenal cells from HSL−/− mice compared with control (P < 0.001). As a control to ensure that deletion of HSL did not alter the ability of cholesterol to be metabolized to corticosterone, ACTH-stimulated corticosterone production was measured in adrenocortical cells in the presence of 20α-hydroxycholesterol, which can freely diffuse into the cell and be transported into the mitochondria. Under these conditions, corticosterone production was similar in control (male 392 ± 50; female 532 ± 62 ng/mg protein·3 h) and HSL−/− mice (male 402 ± 102; female 546 ± 101 ng/mg protein·3 h). Hence, there is no intrinsic defect in the steroidogenic pathway, i.e. conversion of cholesterol into steroids, in HSL−/− mice. These findings support the notion that the actions of neutral CEH (HSL) are involved in the delivery of cholesterol for adrenal steroidogenesis, and that these effects are particularly apparent when HDL-delivered cholesterol is being used.

Table 2.

Effect of cAMP or HDL on Corticosterone Secretion by Adrenocortical Cells

	Basal	Bt2cAMP	HDL	Bt2cAMP + HDL
	Corticosterone (ng/mg protein·24 h)
Control
Male	107 ± 6	854 ± 110	96 ± 3	1160 ± 87
Female	271 ± 9a	1572 ± 180a	243 ± 36a	3648 ± 245a
HSL−/−
Male	76 ± 9b	174 ± 28b	73 ± 3	315 ± 42b
Female	145 ± 8ab	349 ± 49ab	129 ± 10ab	824 ± 80ab

Adrenocortical cells were isolated from control, wild-type, and HSL−/− mice as described in Materials and Methods. Cells were incubated in the absence or presence of Bt₂cAMP (2.5 mm), HDL₃ (500 μg protein/ml), or both Bt₂cAMP (2.5 mm) and HDL₃ (500 μg protein/ml) for 24 h and the amount of corticosterone secreted into the media measured. Results are the mean ± sem of four independent experiments.

^a

, P < 0.01 compared with male mice;

^b

, P < 0.001 compared with control mice.

To assess whether HSL is required for the processing (i.e. conversion) of HDL-derived cholesteryl esters into corticosterone, we have examined the conversion of HDL₃-derived [³H]cholesteryl oleate into [³H]corticosterone by adrenocortical cells from control and HSL−/− mice. As shown in Fig. 2, the basal conversion of HDL₃-derived [³H]cholesteryl oleate into [³H]corticosterone was reduced by at least 90% in both male and female HSL−/− mice (P < 0.01). Moreover, treatment with Bt₂cAMP increased the conversion of HDL₃-derived [³H]cholesteryl oleate into [³H]corticosterone by more than 3-fold in both male and female controls (P < 0.02), whereas in HSL−/− mice the conversion of HDL₃-derived [³H]cholesteryl oleate into [³H]corticosterone remained very low and unchanged. Therefore, in the presence of Bt₂cAMP stimulation, the conversion of HDL₃-derived [³H]cholesteryl oleate into [³H]corticosterone was 97% lower in the HSL−/− mice compared with control (P < 0.01). These findings suggest that HSL is critically involved in mediating the utilization of HDL-derived cholesteryl esters in the adrenal.

Fig. 2.

Conversion of HDL₃-Derived [³H]Cholesteryl Oleate into [³H]Corticosterone by Cultured Mouse Adrenocortical Cells Adrenocortical cells were isolated from control, wild-type, and HSL−/− mice as described in Materials and Methods. Cells were incubated with HDL₃ (500 μg protein/ml), in which the cholesteryl ester core was reconstituted with [³H]cholesteryl oleate, in the absence or presence of Bt₂-cAMP (2.5 mm) for 24 h and the amount of [³H]corticosterone secreted into the media measured. Results are the mean ± sem of four independent experiments. ^A, P < 0.02 compared with basal; ^B, P < 0.01 compared with control mice.

Because murine adrenals utilize LDL-derived cholesteryl esters, as well as HDL-derived cholesteryl esters (20), we compared the effect of HSL deficiency on the ability of LDL to supply cholesterol for steroidogenesis. Adrenal cells isolated from control and HSL−/− male mice were incubated with HDL or LDL containing equivalent amounts of lipoprotein cholesterol (180 μg/ml) in the absence or presence of Bt₂cAMP, and corticosterone production was examined (Table 3). As observed before, basal and Bt₂cAMP-induced corticosterone production were reduced in adrenal cells of HSL−/− mice (P < 0.001), and neither HDL nor LDL affected corticosterone production in the absence of Bt₂cAMP. In the presence of Bt₂cAMP, LDL increased corticosterone production to similar levels as seen with HDL; however, corticosterone production in adrenals from HSL−/− mice was reduced 80–85% in the presence of either LDL (P < 0.01) or HDL (P < 0.001) when compared with control. These findings are consistent with the fact that the majority of LDL cholesterol-supported steroidogenesis in murine adrenals is mediated via the selective pathway (21, 22). Therefore, HSL appears to be essential for lipoprotein-derived cholesterol (both LDL and HDL) utilization for steroidogenesis by the mouse adrenal.

Table 3.

Comparison of HDL and LDL on Corticosterone Secretion by Adrenocortical Cells

	Control	HSL−/−
	Corticosterone (ng/mg protein·24 h)
Basal	119 ± 5	32 ± 3a
HDL	99 ± 6	54 ± 3b
LDL	94 ± 3	40 ± 2a
Bt2cAMP	524 ± 55	70 ± 7a
Bt2cAMP + HDL	1375 ± 84	204 ± 7a
Bt2cAMP + LDL	1110 ± 84	231 ± 28b

Adrenocortical cells were isolated from control, wild-type, and HSL−/− male mice as described in Materials and Methods. Cells were incubated in the absence or presence of Bt₂cAMP (2.5 mm), HDL₃ (500 μg protein/ml), LDL (100 μg protein/ml), or both Bt₂cAMP (2.5 mm) and HDL₃ (500 μg protein/ml) or both Bt₂cAMP (2.5 mm) and LDL (100 μg protein/ml) for 24 h and the amount of corticosterone secreted into the media measured. Results are the mean ± sem of three independent experiments.

^a

, P < 0.001 compared with control mice;

^b

, P < 0.01 compared with control mice.

Because adrenal cells from HSL−/− mice still produce substantial amounts of corticosterone, we have assessed potential changes in pathways other than HDL cholesteryl ester delivery that might have compensated for the defect in cholesterol delivery from HDL. First, circulating ACTH values were not apparently different in control and HSL−/− mice (83.3 ± 8.6 vs. 75.7 ± 8.0 pg/ml, respectively). Second, when the incorporation of [¹⁴C]acetate into cellular lipids was measured in adrenals under basal conditions from control and HSL−/− mice, there were no differences in cholesterol, cholesteryl ester or phospholipid synthesis between control and HSL−/− mice; however, female mice displayed higher rates of cholesterol (P < 0.005) and phospholipid (P < 0.02) synthesis than males (Table 4). Third, to explore whether there was any change in lipoprotein receptor expression, SR-BI and LDL receptor expression were assessed by immunoblotting adrenals obtained before and after administration of ACTH (Fig. 3). Under basal conditions, the expression of SR-BI receptors was increased approximately 50% (P < 0.05) and LDL receptors were increased approximately 3-fold (P < 0.01) in HSL−/− mice compared with controls. SR-BI expression tended to increase after ACTH administration in both controls and HSL−/− mice; however, LDL receptor expression increased after ACTH administration only in controls (P < 0.05).

Fig. 3.

Immunoblot of SR-BI (Panel A) and LDL Receptors (LDLRs) (Panel B) in Adrenals from Control and HSL−/− Mice before (Basal) and 24 h after ACTH (25 μg) Samples were prepared from pools of adrenals from control, wild-type, or HSL−/− mice, extracted, electrophoresed, and immunoblotted with either antipeptide rat SR-BI or antirat LDLR/fusion protein IgG as described in Materials and Methods. Each bar represents the scanned results from four samples of adrenal pools from male mice. ^A, P < 0.05 compared with basal control; ^B, P < 0.05 compared with basal HSL−/−; ^C, P < 0.01 compared with ACTH control; ^D, P < 0.01 compared with basal control.

Table 4.

Incorporation of [¹⁴C]Acetate into Cholesterol, Cholesteryl Esters, and Phospholipid of Mouse Adrenocortical Cells

	Cholesterol	Cholesteryl Esters	Phospholipids
	dpm/mg cell protein
Control
Male	1482 ± 186	1500 ± 160	1710 ± 158
Female	2966 ± 34a	1203 ± 90	2160 ± 52b
HSL−/−
Male	1492 ± 63	1152 ± 156	1274 ± 141
Female	3432 ± 275a	1169 ± 73	2376 ± 240a

Adrenocortical cells were isolated from control, wild-type, and HSL−/− mice as described in Materials and Methods. Cells were incubated in the presence of [¹⁴C]acetate for 24 h and the amounts of cellular ¹⁴C-labeled cholesterol, cholesteryl esters, and phospholipids measured as described in Materials and Methods. Results are the mean ± sem of three independent experiments.

^a

, P < 0.005 compared with male mice;

^b

, P < 0.02 compared with male mice.

To determine whether the changes in lipoprotein receptor expression led to any alterations in lipoprotein uptake, we measured both receptor-mediated and selective uptake using double-labeled HDL₃ ([¹²⁵I]-dilactitol-[³H]-cholesteryl oleoyl ether-human HDL₃) in isolated adrenal cells. As shown in Table 5, endocytic uptake of HDL₃-derived cholesteryl esters was extremely low (<3%) compared with selective uptake, consistent with the vast majority of HDL cholesteryl ester delivery occurring via SR-BI; however, there were no differences in either receptor-mediated endocytic or selective uptake of HDL cholesteryl esters between control and HSL−/− mice. For comparison, we measured both receptor-mediated and selective uptake of double-labeled LDL ([¹²⁵I]-dilactitol-[³H]-cholesteryl oleoyl ether-human LDL). As opposed to HDL₃, substantial amounts of LDL cholesteryl esters were derived from endocytic uptake; yet selective uptake still accounted for the majority of cholesteryl ester delivery. There were no detectable differences in either receptor-mediated endocytic or selective uptake of LDL cholesteryl esters between control and HSL−/− mice.

Table 5.

Endocytic and Selective Uptake of Lipoprotein-Derived Cholesteryl Esters by Mouse Adrenocortical Cells

	Endocytic Uptake	Selective Uptake
Cholesteryl esters internalized (ng/mg cell protein·24 h)
A. HDL3
Control
Male	158 ± 16	6908 ± 853
Female	186 ± 31	7931 ± 1016
HSL−/−
Male	143 ± 6	5869 ± 748
Female	167 ± 42	8223 ± 930
B. LDL
Control
Male	5357 ± 386	8651 ± 998
Female	4937 ± 510	7867 ± 1153
HSL−/−
Male	4623 ± 345	9395 ± 1188
Female	4089 ± 635	8516 ± 1201

Adrenocortical cells were isolated from control, wild-type, and HSL−/− mice as described in Materials and Methods. Cells were incubated with A: ¹²⁵I-labeled dilactitol-[³H]cholesteryl oleoyl ether-hHDL₃ or B: ¹²⁵I-labeled dilactitol-[³H]cholesteryl oleoyl ether-hLDL for 24 h at 37 C. At the end of incubation, the cell samples were washed and then solubilized in 2 ml of 0.1 n NaOH. One-milliliter aliquots were precipitated with an equal volume of 20% trichloroacetic acid to determine insoluble and soluble ¹²⁵I radioactivity or extracted with organic solvents to determine ³H-radioactivity. Endocytic uptake is calculated from trichloroacetic acid-soluble ¹²⁵I label only. The difference between total and trichloroacetic acid-soluble activity is taken as the surface-associated ¹²⁵I radioactivity. Because both ¹²⁵I and ³H labels are on the same particle, the surface bound ¹²⁵I is also equal to the surface bound ³H. Thus, total ³H minus surface bound ³H equals the total amount of ³H internalized. To calculate selective uptake of cholesteryl ester, soluble ¹²⁵I radioactivity is subtracted from soluble ³H-radioactivity. The results are the mean ± sem of four separate experiments.

DISCUSSION

The mobilization of stored cholesteryl esters via the actions of neutral CEH is an important source of cholesterol for adrenal steroidogenesis. The fact that adrenal cholesteryl ester stores are rapidly depleted after ACTH treatment supports the notion that mobilization of stored cholesteryl esters via the actions of neutral CEH is required for the acute stimulation of steroidogenesis (23). However, to sustain steroidogenesis for an extended period of time and to ensure a constant supply of cholesterol substrate, this pool is continuously replenished from lipoprotein-derived cholesteryl esters (5). Stored cholesteryl esters are derived either from LDL receptor-mediated uptake of LDL with subsequent lysosomal hydrolysis and cytoplasmic reesterification by acyl-coenzyme A:cholesterol acyltransferase, or directly from SR-BI-mediated selective cholesteryl ester uptake, or from endogenous cholesterol synthesis and esterification. Whichever the source of the cholesteryl esters, the cholesteryl esters must be hydrolyzed to unesterified cholesterol via the actions of neutral CEH before being transported into mitochondria by StAR and other essential factors and subsequently delivered to CYP11A1 (2).

The neutral CEH activity purified from the adrenal has previously been shown to be identical with HSL purified from adipose tissue (24). Recently, we reported that both male and female HSL−/− mice have less than 2% of the neutral CEH activity observed in control wild-type mice (19). Thus, it appears that HSL is responsible for the vast majority, if not all, of the neutral CEH activity in the adrenal. Indeed, it is likely that the small amount of neutral CEH activity observed in adrenals of HSL−/− mice is due to residual activity of lysosomal acid lipase that can still be seen at neutral pH (25). Interestingly, loss of neutral CEH activity by inactivation of HSL resulted in a reduction in corticosterone response to ACTH in vivo, but normal basal serum corticosterone levels (19, 26). Indeed, morphological examination of the adrenals of HSL−/− mice has revealed the appearance of hypertrophied adrenals with extensive accumulation of lipid droplets throughout the cortex (26).

In the current studies, we have provided evidence to support an important role of HSL in adrenal cholesterol homeostasis and, subsequently, steroidogenesis. Consistent with HSL accounting for the vast majority of the neutral CEH activity in the adrenal, we have documented that adrenal cholesteryl ester content was substantially elevated in both male and female HSL−/− mice compared with control, wild-type animals, confirming that the lipid droplets previously identified throughout the adrenal cortex of HSL−/− mice (26) represent abnormally increased stores of cholesteryl esters.

Basal corticosterone production from isolated adrenocortical cells was reduced approximately 50% in both male and female HSL−/− mice. This finding is in contrast to the in vivo condition where basal corticosterone values were similar for controls and HSL−/− mice (26). Thus, in vivo measurements appear to underestimate the severity of the defect in steroid production in the HSL−/− mice under unstimulated conditions. The discrepancy between the larger defect observed in vitro and the defect in corticosterone production observed in vivo can probably be attributed to the adrenal hypertrophy/hyperplasia, characteristic of a chronically stressed state, seen in the HSL−/− mice, thus partially compensating for what is apparently a severe defect in steroidogenesis. Moreover, by examining an equivalent number of cells, the in vitro data highlight the severity of the defect in steroidogenesis in HSL−/− mice. Because lipoproteins were absent in the in vitro experiments, corticosterone production under these basal conditions is dependent on either endogenously synthesized cholesterol, unesterified cholesterol present in cell membranes, or the mobilization and utilization of stored cholesteryl esters. Therefore, the finding of a reduction in basal corticosterone production is consistent with an important role for HSL in one or more of these cholesterol pathways. Moreover, because HSL most likely functions in the mobilization and utilization of stored cholesteryl esters, these results provide evidence for the importance of cholesteryl ester stores for maintaining normal, basal steroid production.

Similar to the defects seen in basal corticosterone production, stimulation of corticosterone production from isolated adrenocortical cells by ACTH or Bt₂cAMP was markedly reduced 75–85% in HSL−/− mice, whether in the absence or presence of HDL or LDL. Therefore, as seen in vivo (19, 26), stimulated corticosterone production is defective in HSL−/− mice; however, similar to the observation of basal corticosterone production, the in vivo measurements appear to underestimate the severity of the defect in steroid production in the HSL−/− mice under conditions where corticosterone production should be maximal. Furthermore, neither increased cellular stores of cholesteryl esters nor delivery of lipoprotein cholesteryl esters appears to be able to supply sufficient cholesterol for adrenal steroidogenesis in the absence of HSL, supporting an important role for HSL in processing lipoprotein-derived cholesteryl esters. Fourth, we documented that there are no intrinsic defects in adrenal steroidogenesis in HSL−/− mice by showing that the ACTH-stimulated production of corticosterone was normal in the presence of 20α-hydroxycholesterol, a source of cholesterol that can freely diffuse into the cell and mitochondria. Thus, HSL appears to participate in the delivery of cholesterol for adrenal steroidogenesis and these effects are particularly apparent when the selective uptake of lipoprotein-delivered cholesterol is involved. The application of these findings to the relative importance of HSL in human adrenals should be interpreted carefully because, even though human adrenals selectively utilize lipoprotein-derived cholesteryl esters (27, 28), receptor-mediated endocytic uptake of LDL appears to be a more important source of cholesterol for steroidogenesis in humans (29).

Direct assessment of HDL cholesteryl ester conversion into corticosterone showed a more than 90% reduction in HSL−/− mice under basal conditions where corticosterone production is low. Whereas there was a 3-fold increase in the conversion of HDL cholesteryl esters into corticosterone in control wild-type adrenocortical cells after stimulation by Bt₂cAMP, there was no change in the conversion of HDL cholesteryl esters into corticosterone in HSL−/− mice, resulting in a 97% lower conversion of HDL cholesteryl esters into corticosterone in HSL−/− mice compared with controls. Thus, HSL appears to be critical for processing HDL-derived cholesteryl esters for adrenal steroidogenesis.

Although it is clear that there is a marked defect in Bt₂cAMP- and lipoprotein-stimulated corticosterone production in isolated adrenocortical cells from HSL−/− mice, the severity of this defect is attenuated in vivo even though defective corticosterone production is observed. Moreover, corticosterone production is increased to some extent in HSL−/− adrenocortical cells by the presence of HDL or LDL above that observed with Bt₂cAMP alone. It is likely that a small fraction of cholesteryl esters may be hydrolyzed by some other nonspecific esterases. Also, a portion of the attenuation of the defect in corticosterone production in vivo is probably due to the adrenal hypertrophy/hyperplasia, characteristic of a chronically stressed state, seen in the HSL−/− mice, thus partially compensating. In addition, under most conditions, the supply of cholesterol is not rate limiting for steroidogenesis, due to the fact that there are multiple pathways that can satisfy the cholesterol needs of the cell under changing physiological conditions. Examination of some of the additional compensatory mechanisms that might be operative showed that adrenal cholesterol synthesis was not altered in isolated adrenocortical cells from HSL−/− mice. Although this suggests that there were no increases in cholesterol synthesis to compensate for the apparent loss of cholesterol supplied from cholesteryl ester stores or HDL, it is possible that elevated ACTH increases adrenal cholesterol synthesis in vivo. Thus, the more pronounced defect in basal adrenal steroid production observed in vitro might have been due in part to the absence of ACTH, whereas small elevations in ACTH in vivo might have allowed partial compensation. However, we were unable to document differences in circulating ACTH values in control and HSL−/− mice.

SR-BI and LDL receptor expression were increased in adrenocortical cells of HSL−/− mice. Although no changes in selective cholesteryl ester uptake or in the receptor-mediated endocytic uptake of LDL or HDL could be documented, the increase in receptor expression suggests that these compensatory pathways for cholesterol delivery were up-regulated and perhaps that the uptake of apolipoprotein E-containing lipoproteins might have been increased. It should be noted that receptor expression was analyzed under in vivo conditions, whereas direct measurements of cholesteryl ester delivery were conducted on isolated cells in vitro, perhaps contributing to the discrepancy between the findings. Nonetheless, the ability of cholesteryl esters to be delivered to and taken up by adrenocortical cells does not appear to be diminished by the absence of HSL, but the ability of HSL−/− cells to process cellular cholesteryl esters was severely impaired. It is possible that compensatory increases in other components of the cholesterol homeostatic and steroidogenic machinery, such as StAR, Niemann-Pick C1 protein, and various steroid biosynthetic enzymes, occur in HSL−/− mice to attenuate the severity of the defect in corticosterone production in vivo, but evidence for these possibilities awaits experimental proof.

In conclusion, using several different experimental approaches, we have provided direct evidence that HSL plays a critical role in maintaining cholesterol homeostasis in the adrenal gland. Moreover, evidence is presented to suggest that HSL is functionally linked to the selective pathway and is directly responsible for hydrolyzing cholesteryl esters derived from the selective delivery of cholesteryl esters from HDL and, probably, other lipoproteins. Because the selective delivery of cholesteryl esters from lipoproteins and stores of cellular cholesteryl esters constitute the major sources of cholesterol for steroid production, HSL occupies a vital role in adrenal steroidogenesis.

MATERIALS AND METHODS

Chemicals

Reagents were obtained from the following sources: BSA (fraction V) (Intergen Co., Purchase, NY); cholesterol oleate (Sigma Chemical Co., St. Louis, MO); [³H]cholesteryl oleate, [¹⁴C]acetate, Na¹²⁵I, [³H]cholesteryl oleoyl ether (Perkin-Elmer Life Sciences, Inc., Boston, MA); chloroform, methanol, heptane (J. T. Baker, Inc., Phillipsburg, NJ). All other chemicals were obtained from standard commercial sources.

Animals and Isolation and Culture of Mouse Adrenocortical Cells

HSL−/− mice were generated as described previously (30). All animal experimentation was conducted in accord with accepted standards of humane animal care and was approved by the Veterans Affairs Palo Alto institutional committee on animal care. To measure plasma ACTH values, mice were quickly decapitated (within 30 sec after handling) and bled into microtainer-EDTA tubes (Becton Dickinson, Franklin Lakes, NJ). The resulting plasma samples were stored at −80 C until analyzed. Mouse adrenocortical cells were isolated from control (HSL+/+) and HSL−/− mice by a slight modification of the procedure described previously for rats (5). In brief, mice were killed by cervical dislocation, and the adrenal glands aseptically removed and dissected free of fat. A group of 20 adrenals (from 10 mice) was finely minced with scissors, and tissue fragments suspended in sterilized Medium 199 containing 40 mg/ml of BSA, 3.7 mg/ml of collagenase, and 5 μg/ml of deoxyribonuclease and incubated with shaking for 50 min in an atmosphere of 95% air -5% CO₂. The tissue suspension was then dissociated by repeated pipetting with a tuberculin syringe, the resulting suspension filtered through a nylon mesh, washed by centrifugation, and the final pellet resuspended in DMEM:F12 (1:1). These cell preparations were either used immediately for various measurements or cultured overnight in serum-free culture medium (DMEM:F12 containing 1 mg/ml BSA, 2 μg/ml insulin, 5 μg/ml transferrin, and 2 μg/cm² fibronectin), plus other test substances as specified under each figure and table.

Preparation of Lipoproteins

Human (h) LDL and apolipoprotein E-free HDLs (hHDL₃) were isolated and characterized as previously described (31). For lipoprotein cholesteryl ester utilization studies, hHDL was reconstituted with [³H]cholesteryl oleate as described previously (5). For uptake and internalization studies, lipoproteins were equipped with two nonreleasable labels, i.e. [¹²⁵I]-labeled dilactitol tyramine to mark lipoprotein proteins and [³H]cholesteryl oleoyl ether to mark lipoprotein cholesteryl esters, as described previously (13).

Measurement of Corticosterone Secretion

To assay steroidogenesis, triplicate samples of freshly isolated cells were incubated without (basal) or with ACTH (10 ng/ml), Bt₂cAMP (2.5 mm) or 20α-hydroxycholesterol (10 μm) for 3 h and subsequently samples of incubation medium were frozen and stored until assayed for corticosterone. To examine lipoprotein-supported corticosterone production, isolated adrenocortical cells were cultured for 24 h with ± Bt₂cAMP (2.5 mm) or ± hHDL₃ (500 μg protein/ml) or ± hLDL (100 μg protein/ml), and after incubation the incubation media were collected, frozen and stored frozen until analyzed. Results are expressed as nanograms corticosterone produced per milligram cell protein and represent the mean ± se of duplicate determination of three different samples.

Conversion of Lipoprotein-Derived Cholesteryl Ester into Corticosterone

To assess the utilization of HDL₃-derived cholesteryl esters for corticosterone production, adrenal cells were incubated with [³H]cholesteryl oleate-hHDL₃ (100 μg/ml) ± Bt₂cAMP (2.5 mm) for 24 h. After which, cells and media were extracted with dichloromethane, and the organic phase was assayed for corticosterone production by thin-layer chromatography as described previously (5). Results are expressed as disintegrations per minute/milligram cell protein.

Uptake and Internalization of Lipoprotein-Derived Cholesteryl Esters

For these experiments, cells were incubated with medium ± Bt₂cAMP (2.5 mm) and ¹²⁵I-labeled dilactitol-[³H]cholesteryl oleoyl ether-hHDL₃ or ¹²⁵I-labeled dilactitol-[³H]cholesteryl oleoyl ether-hLDL (100 μg protein/ml) for 24 h at 37 C. At the end of incubation, the cell samples were washed and then solubilized in 2 ml of 0.1 n NaOH. One-milliliter aliquots were precipitated with an equal volume of 20% trichloroacetic acid to determine insoluble and soluble ¹²⁵I radioactivity or extracted with organic solvents to determine ³H-radioactivity (13, 32). Endocytic uptake is calculated from trichloroacetic acid-soluble ¹²⁵I label only. The difference between total and trichloroacetic acid-soluble activity is taken as the surface-associated ¹²⁵I radioactivity. Because both ¹²⁵I and ³H labels are on the same particle, the surface bound ¹²⁵I is also equal to the surface bound ³H. Thus, total ³H minus surface bound ³H equals the total amount of ³H internalized. To calculate selective uptake of cholesteryl ester, soluble ¹²⁵I radioactivity is subtracted from soluble ³H-radioactivity.

Determination of [¹⁴C]Acetate Incorporation into Cholesterol, Cholesteryl Esters, and Phospholipids

Aliquots of adrenal cell suspension were incubated in DMEM: F12 (1:1) containing 5 mg/ml BSA and 2 μCi of [¹⁴C]acetate at 37 C for 4 h. After incubation, the cell suspensions were centrifuged and sedimented cells washed extensively with medium to remove any extracellular radioactivity, and extracted with organic solvents (33). The organic phase was dried under N₂ and [¹⁴C]acetate incorporation into [¹⁴C]cholesterol, [¹⁴C]cholesteryl esters, and [¹⁴C]phospholipids was quantified by thin-layer chromatography followed by scintillation counting (33).

Immunoblotting

Samples were suspended in 50 mm Tris (pH 8.0), 1 mm EDTA, 1 mm phenylmethylsulfonyl fluoride, 1 U/ml leupeptin, and 0.2 mg/ml aprotinin, electrophoresed on 10% polyacrylamide gels under reducing conditions, transferred to nitrocellulose, incubated with either antipeptide rat SR-BI (16) or antirat LDLR/fusion protein IgG (34), and visualized by chemiluminescence as described previously.

Analytical Procedures

Protein was measured with a bicinchoninic acid protein assay kit (Pierce Chemical Co., Rockford, IL). The procedure of Markwell et al. (35) was used to quantify protein content of hHDL₃, hLDL and doubly labeled lipoprotein preparations. Cellular cholesteryl ester mass was measured as described previously (36). Cholesterol content of hHDL₃ and hLDL and doubly labeled lipoproteins was determined according to the procedure of Tercyak (37). Serum corticosterone was measured by RIA (5). ACTH was measured with an ACTH RIA kit supplied by Nichols Institute Diagnostics (San Clemente, CA).

Statistics

Data are expressed as mean ± sem. Statistical analyses were performed by ANOVA and comparisons among groups by Fisher’s protected t tests using GraphPad Prism software (GraphPad Software, San Diego, CA) on a Power Macintosh computer.

Acknowledgments

We thank Qing Li for help breeding and screening mice.

This work was supported in part by the Research Service of the Department of Veterans Affairs and by Grants DK 46942 and DK 56339 from the NIH.

Abbreviations:

 
• Bt₂cAMP,

  Dibutyryl cAMP;

 
• CEH,

  cholesteryl ester hydrolase;

 
• CYP11A1,

  mitochondrial cholesterol side chain cleavage enzyme;

 
• HDL,

  high-density lipoprotein;

 
• h,

  human;

 
• HSL,

  hormone-sensitive lipase;

 
• LDL,

  low-density lipoprotein;

 
• SR-BI,

  scavenger receptor, class B type I;

 
• StAR,

  steroidogenic acute regulatory protein.