Synthetic High-Density Lipoprotein (sHDL) Inhibits Steroid Production in HAC15 Adrenal Cells

Abstract

High density lipoprotein (HDL) transported cholesterol represents one of the sources of substrate for adrenal steroid production. Synthetic HDL (sHDL) particles represent a new therapeutic option to reduce atherosclerotic plaque burden by increasing cholesterol efflux from macrophage cells. The effects of the sHDL particles on steroidogenic cells have not been explored. sHDL, specifically ETC-642, was studied in HAC15 adrenocortical cells. Cells were treated with sHDL, forskolin, 22R-hydroxycholesterol, or pregnenolone. Experiments included time and concentration response curves, followed by steroid assay. Quantitative real-time RT-PCR was used to study mRNA of 3-hydroxy-3-methyl-glutaryl-coenzyme A reductase, lanosterol 14-α-methylase, cholesterol side-chain cleavage enzyme, and steroid acute regulatory protein. Cholesterol assay was performed using cell culture media and cell lipid extracts from a dose response experiment. sHDL significantly inhibited production of cortisol. Inhibition occurred in a concentration- and time-dependent manner and in a concentration range of 3μM–50μM. Forskolin (10μM) stimulated cortisol production was also inhibited. Incubation with 22R-hydroxycholesterol (10μM) and pregnenolone (10μM) increased cortisol production, which was unaffected by sHDL treatment. sHDL increased transcript levels for the rate-limiting cholesterol biosynthetic enzyme, 3-hydroxy-3-methyl-glutaryl-coenzyme A reductase. Extracellular cholesterol assayed in culture media showed a positive correlation with increasing concentration of sHDL, whereas intracellular cholesterol decreased after treatment with sHDL. The current study suggests that sHDL inhibits HAC15 adrenal cell steroid production by efflux of cholesterol, leading to an overall decrease in steroid production and an adaptive rise in adrenal cholesterol biosynthesis.

Low density lipoprotein (LDL) and high density lipoprotein (HDL) are extracellular carriers that deliver cholesterol to steroidogenic tissues such as the ovaries, testes, and adrenal glands. LDL receptor and scavenger receptor BI (SR-BI, the HDL receptor) are highly expressed in adrenal tissue, allowing the adrenal gland to efficiently take up cholesterol from lipoprotein particles (1–5). Both LDL and HDL have been shown to provide cholesterol for steroidogenesis (1, 6–8). In addition, the lipoprotein cholesterol levels (LDL-cholesterol and HDL-cholesterol [HDL-C]) are indicators of cardiovascular health and ischemic event risk (9). HDL inhibits atherosclerotic development through reverse cholesterol transport, a mechanism by which nascent HDL effluxes cholesterol from peripheral tissue and delivers it to the liver for elimination (10).

Strong epidemiological evidence for the cardioprotective properties of HDL prompted development of several therapeutic agents to either increase HDL-C or to stimulate reverse lipid transport. Although niacin is Federal Drug Administration approved to elevate HDL levels, the clinical efficacy is somewhat controversial. Additionally, inhibiting cholesteryl-ester transfer protein with torcetrapib or anacetrapib increased HDL-C levels clinically, but did not show either plaque or event reduction (11). Of interest, torcetrapib increased adrenal aldosterone production through nonspecific activity of this cholesteryl-ester transfer protein's chemical structure, and not through the resultant increase in HDL-C (12).

Infusion therapy with nascent, cholesterol-free synthetic HDL (sHDL) represents an alternative strategy to stimulate reverse cholesterol transport through targeted increase of efflux capacity, rapid mobilization, and elimination of cholesterol (13). There is strong in vitro evidence that sHDL enhances cholesterol efflux from foam macrophages (14). Several of these products are clinically tested, including ETC-216 (15), CSL-111 (16), ETC-642 (17), CSL-112, and CER-001 (18). Although ETC-216 and CSL-111 were shown to reduce atheroma burden (15, 16), CER-001 failed to show efficacy (18). CSL-112 was shown to be effective in initiating cholesterol efflux in patients (19) and a larger safety and efficacy phase 2 trial is ongoing (20).

Although sHDL may lower intracellular cholesterol and its clinical use is being expanded, the effects of sHDL on steroidogenic tissues, which use cholesterol as a substrate for steroid hormone synthesis, have not been investigated. Therefore, the current study defined the effects of sHDL on adrenal cell steroidogenesis using the forskolin-responsive human adrenocortical cell line (HAC15) as a model (21). HAC15 cells are a clonal cell line derived from the nonclonal H295R human adrenocortical cancer cells. HAC15 cells have a modest response to ACTH which is not found in the original H295R line (21, 22). Here, the HAC15 cell line was used to define the effects of sHDL on steroidogenesis. Our results demonstrate a broad inhibition of steroid production that is associated with sHDL depletion of cellular cholesterol.

Materials and Methods

Materials

Apolipoprotein A-I mimic peptide was synthesized by Genscript using solid-phase fluorenylmethyloxycarbonyl chemistry. Peptide purity was more than 95% as determined by HPLC. Egg sphingomyelin and 1,2-dipalmitoyl-sn-glycero-3-phosphocholine were generously donated by Nippon Oil and Fat. All other materials were obtained from commercial sources.

Preparation of sHDL

ETC-642, a model sHDL nanoparticle, was prepared using homogenization method as follows: the composition of sHDL was apolipoprotein A-I mimetic peptide (ESP24218) combined with sphingomyelin and 1,2-dipalmitoyl-sn-glycero-3-phosphocholine at 1M:3.75M:3.75M ratio (23). Phospholipids were dispersed in PBS by sonication. Peptide was dissolved in PBS mixed with lipid suspension. The mixture was heated to 50°C and incubated for 15 minutes until the solution became clear indicating sHDL formation. The resulting sHDL solution concentration was 7.5 mg/mL of ESP24218 peptide corresponding to approximately 3mM sHDL concentration. The solution was sterile filtered and stored frozen at −20°C until use. The purity and size of sHDL remained unchanged after 3 freeze-thaw cycles.

Characterization of sHDL

The size distribution of sHDL was assessed by dynamic light scattering (DLS) using a Zetasizer Nano, Malvern Instruments. The number and intensity average values were reported. The purity of sHDL was analyzed by gel permeation chromatography with UV detection at 220 nm using a Tosoh TSK gel G3000SWxl column (Tosoh Bioscience) on a Waters Breeze Dual Pump system. The samples were diluted to a 200μM concentration and a 50-μL injection volume was used.

Transmission electron microscopy (EM) images were obtained using a Tecnai T12 EM (JEOL USA) equipped with a Gatan US4000 CCD camera. Images were acquired at 120 kV in low-dose mode with a defocus of approximately −0.96 μm. The samples were negatively stained with uranyl formate solution.

Cell culture and treatment

HAC15 cells were plated at a density of 100 000 cells per well in a 48-well plate (Figures 1 and 2; see also figures 4 and 5 below), 200 000 cells per well in a 24-well plate (see figure 5 below), or 400 000 cells per well in a 12-well plate (Figure 3; see figure 6 below). Cells were plated in DMEM/F12 media supplemented with 10% cosmic calf serum (HyClone), 1% insulin/transferrin/selenium premix (ITS; BD Biosciences), 1% penicillin/streptomycin, and 0.01% Gentamicin and were given 48 hours to adhere to the plate. Before treatment with sHDL, the cells were cultured in experimental media (0.1% cosmic calf serum and antibiotics) for 18 hours. Cells were then treated with varying concentrations and times with sHDL as well as steroidogenesis agonists (22R-hydroxycholesterol [22OHC], pregnenolone, or forskolin). For cholesterol flux experiments, cells were treated in phenol-free DMEM/F12 media in order to not interfere with the colorimetric cholesterol assay.

Figure 1.

Concentration-dependent effects of sHDL (ETC-642) on adrenal cell cortisol production. HAC15 adrenocortical cells were incubated for 24 hours with increasing amounts of ETC-642. Cortisol from the media was then measured and normalized to cell protein. Values shown are the mean ± SD from 3 independent experiments; *, P < .05 compared with basal. NS, not significant.

Figure 2.

Time-dependent effects of sHDL (ETC-642) on adrenal cell cortisol production. HAC15 adrenal cells were incubated for the indicated times (h) with or without ETC-642 (30μM). Cortisol was measured in medium and normalized to cell protein. Results represent mean ± SD from 3 independent experiments; *, P < .001 compared with basal.

Figure 3.

Time-dependent effect of sHDL (ETC-642) on agonist-stimulated adrenal cell cortisol production. HAC15 adrenal cells were incubated for the indicated times (h) with or without forskolin (10μM), as well as forskolin + ETC-642 (50μM). Cortisol production was measured in the experimental media and normalized to protein. Results represent mean ± SD from 3 independent experiments; ‡, P < .05 for forskolin compared with basal; *, P < .05 for forskolin + sHDL compared with forskolin.

Protein extraction and protein assay

Cells were lysed in mammalian protein extraction buffer (Pierce Chemic Co). Protein content in wells was determined by bicinchoninic acid protein assay using the micro-BCA protocol (Pierce Chemical Co). Samples were read on an Epoch Microtiter Spectrophotometer at 562 nm absorbance.

mRNA isolation and quantitative real-time RT-PCR analysis

RNA was extracted from cells using the RNA isolation kit (QIAGEN Sciences) according to manufacturer recommendations. Briefly, RNA was extracted using proprietary RLT lysis buffer and isolated using spin columns provided in the kit. Total RNA was resuspended in nuclease-free water (QIAGEN Sciences) and reverse transcribed using random primers (Life Sciences). Quantitative real-time PCRs were performed with TaqMan primer probes for 3-hydroxy-3-methyl-glutaryl-coenzyme A reductase (HMGCR) (Hs00168352_m1; Life Technologies) and lanosterol 14-α-methylase (CYP51A1) (Hs00426415_m1; Life Technologies). Primer probes for steroid acute regulatory protein (StAR) and cholesterol side-chain cleavage enzyme (CYP11A1) were designed by the Rainey laboratory and purchased through IDT. The following primer sequences were used: StAR forward, 5′-ATGAGTAAAGTGGTCCCAGATG-3′, reverse 5′-ACCTTGATCTCCTTGACATTGG-3′, and probe 5′-/56-FAM/ATCCGGCTGGAGGTCGTGGTGGA-3′; CYP11A1 forward 5′-GAGATGGCACGCAACCTGAAG-3′, reverse 5′-CTTAGTGTCTCCTTGATGCTGGC-3′, and probe 5′-/56-FAM/CGATCTGCCGCGCAGCCAAGACC-3′.

Steroid assays

Cortisol immunoassay was performed according to the manufacturer's recommendations (Alpco). The standard curve was prepared using cortisol standards dissolved in the cell culture media. Briefly, 20 μL of experimental media were incubated on antibody-coated plates, in proprietary assay buffer, for 45 minutes at room temperature. After incubation, the cells were washed with 1× wash buffer. Tetramethylbenzidine substrate was added followed by 20 minutes incubation with shaking. Stop buffer was then added and absorbance was measured at 450 nm on a microtiter spectrophotometer.

Aldosterone and androstenedione RIAs (Coat-a-Count) were performed according to the manufacturer's recommendations (Siemens). Steroid standard curves were prepared using aldosterone and androstenedione dissolved in cell culture media. Radioactivity was measured on a Wallac Wizard 1470 multicrystal γ-counter (PerkinElmer).

Intracellular cholesterol extraction and colorimetric cholesterol assay

Intracellular cholesterol was extracted from cells using a 2:1 mixture of chloroform:methanol. Purified water was then added, and upon centrifugation, the organic, bottom phase was taken and dried by vacuum centrifugation. The resulting lipid pellet was resuspended in 1× cholesterol assay buffer from the kit described below.

Extracellular cholesterol assay (Cell Biolabs) was performed, according to manufacturer's directions with phenol-free experimental cell culture media. The intracellular assay was performed with the suspended pellet described above. In short, 50 μL of sample were added to 50 μL of cholesterol reaction reagent containing cholesterol esterase, cholesterol oxidase, colorimetric probe, and horseradish peroxidase diluted in assay buffer. After an incubation step, absorbance was read at 562 nm. Cholesterol standards were prepared in phenol-free experimental cell culture media for extracellular cholesterol and in 1× assay buffer for intracellular cholesterol.

Statistical analysis

Studies were replicated in a minimum of 3 independent experiments. Results are expressed as mean ± SD. Statistics were calculated using a one-way ANOVA with pairwise analysis, when necessary (Sigma Plot 12.5). P < .05 was considered statistically significant.

Results

sHDL characterization

Analysis of sHDL nanoparticles by gel permeation chromatography revealed formation of a monodispersed nanoparticle and absence of unbound apolipoprotein A-I peptide or liposomes (Supplemental Figure 1A). The size distribution by DLS confirmed the presence of monodispersed particles of 8.3 nm in average diameter (number averaged distribution was used for DLS data fitting) (Supplemental Figure 1B). EM showed a discoidal shape of sHDL that is characteristic of nascent or cholesterol-free HDL (Supplemental Figure 1C).

Steroid production

Treatment of HAC15 adrenal cells with increasing concentrations of sHDL for 24 hours caused a decrease in aldosterone, cortisol, and androstenedione (Supplemental Figure 2). Compared with basal steroid levels, sHDL profoundly inhibited cortisol production in a concentration dependent manner. Inhibition of cortisol production plateaued at 80% with a 30μM concentration with no additional effects seen at 50μM sHDL (Figure 1). Importantly, 48 hours of sHDL treatments as high as 100μM did not have cytotoxic effects on the HAC15 cells (data not shown). To establish the time dependence of sHDL effect on cortisol production, cells were incubated in the presence of 30μM sHDL for up to 24 hours (Figure 2). Cortisol production dropped to approximately 20% of basal levels after 6 hours of treatment, and steroidogenesis was consistently inhibited through all time points examined.

Cortisol production was also examined in cells that were untreated, treated with forskolin (10μM), an activator of cAMP production, alone, and forskolin with sHDL (50μM) for 3, 6, 12, and 24 hours (Figure 3). Compared with basal levels, forskolin significantly increased cortisol production at each time point (ranging from ∼1.7-fold above basal at 3 h to 2.6-fold at 24 h) (Figure 3). However, in the presence of sHDL, the forskolin effect was significantly abrogated, because the production of cortisol was decreased to near basal levels for each time point (1.1-fold at 6 h, 0.8-fold at 12 h, and 0.7-fold at 24 h compared with basal levels) (Figure 3). This highlights the ability of sHDL to inhibit both basal and agonist stimulated cortisol production.

The rate-limiting step in adrenal cell steroid production can be bypassed by incubation of cells with 22OHC or pregnenolone. Addition of 22OHC (10μM) or pregnenolone (10μM) increased basal cortisol production significantly (∼2.5- and 3-fold, respectively). As noted above, treatment with sHDL (30μM) alone inhibited cortisol production to 20% of that seen in control cells. Coincubation of sHDL with 22OHC or pregnenolone showed no significant difference when compared with treatment groups without sHDL (Figure 4). The lack of sHDL inhibition of 22OHC metabolism to cortisol suggests that sHDL is not cytotoxic and likely acts before mitochondrial pregnenolone production.

Figure 4.

Effect of sHDL (ETC-642) on 22OHC and pregnenolone stimulated cortisol production. HAC15 cells were treated for 24 hours with and without ETC-642 (30μM). Samples were then incubated for 6 hours with and without 22OHC (10μM) and pregnenolone (10μM). Cortisol production was measured in media and normalized to protein. Results represent mean ± SD from 3 independent experiments; *, P < .001, pairwise analysis comparing treatment group with and without sHDL.

Transcript levels

The enzymes required for cholesterol biosynthesis are increased in response to experimental manipulations that cause depletion of cellular cholesterol levels. HMGCR and CYP51A1 represent 2 cholesterol biosynthetic enzyme transcripts that increase after depletion of cellular cholesterol levels. Adrenal cells responded to sHDL with a time-dependent increase in HMGCR, which peaked with a 2.6-fold increase at 6 hours (Figure 5A). HMGCR transcripts remained significantly elevated at 12 and 24 hours. CYP51A1 encodes a cholesterol biosynthetic enzyme responsible for removing the 14-α-methyl group from lanosterol. CYP51A1 mRNA was also significantly increased at 3, 6, an 12 hours and peaked at 24 hours (∼2.5-fold), further illustrating the induction of the cholesterol biosynthetic pathway. In contrast, the transcripts for both CYP11A1 and StAR, 2 important proteins in the early steroidogenesis pathway, were not significantly changed after 6 hours of incubation with sHDL (Figure 5C).

Figure 5.

A, Time-dependent effects of sHDL on HMGCR gene expression. B, Time-dependent effects of sHDL on the cholesterol biosynthetic gene, CYP51A1. C, Effect of sHDL on the expression of key steroidogenesis genes, CYP11A1, and StAR. For all experiments, HAC15 cells were incubated with or without ETC-642 (50μM, for the times indicated in A and B, and 6 h, the peak of HMGCR expression, in C). For A and B, results represent mean ± SD from 2 independent experiments with 3 replicates within each experiment. C, Results represent mean ± SD from 3 independent experiments; *, P < .05 compared with basal.

Cholesterol efflux

Extracellular cholesterol was measured in HAC15 cell experimental media after 24 hours of incubation of untreated cells, as well as from cells treatment with 10μM and 50μM sHDL (Figure 6). Intracellular cholesterol was measured in the same cells of the respective experiments after cell lysis and lipid extraction. Although basal levels of extracellular cholesterol were undetectable, the cells incubated with 50μM sHDL effluxed approximately 100 nmol/mg cell protein after 24 hours of incubation (Figure 6). Intracellular cholesterol was significantly depleted compared with basal at the 50μM concentration of sHDL (∼40 nmol cholesterol/mg protein compared with the 65 nmol/mg protein at basal), whereas no significant effect was observed at the 10μM concentration (Figure 6).

Figure 6.

Increasing concentration of sHDL leads to cholesterol efflux from adrenocortical cells. HAC15 adrenocortical cells were incubated for 24 hours with or without ETC-642 at the indicated concentration. Extracellular cholesterol was measured in the media. Intracellular cholesterol was measured from the cell lysate as described in Materials and Methods. Both were normalized to total cell protein. Results represent mean ± SD from 3 independent experiments; *, P < .05 compared with extracellular basal; ‡, P < .05 compared with intracellular basal.

Discussion

The adrenal gland produces considerable amounts of steroid hormones using cholesterol as a precursor. Sources for adrenal cholesterol include de novo synthesis, LDL, and HDL (3, 6–8, 24). As an indication of the important roles for each of these cholesterol sources, the adrenal expresses high levels of the enzymes needed for cholesterol synthesis, as well as the receptors for both HDL (SR-BI) and LDL (LDL receptor) (1, 25–27). Because of the beneficial cardio-protective effects of HDL, several groups have developed synthetic versions of HDL as potential therapeutics. Despite the fact that the adrenal cortex has the highest tissue expression of SR-BI, the effects of sHDL on adrenal steroid hormone production have not been previously reported.

There have been numerous studies directed at defining the role of native HDL and LDL in the adrenal using human, bovine, rat and mouse adrenal models (6, 7, 28–31). Evidence for the steroidogenic role of LDL is fairly consistent. LDL has been demonstrated to increase steroidogenesis in murine Y1 cells in the presence of agonist ACTH (28). In the 1980s, several studies demonstrated the ability of LDL to increase steroidogenesis in the human adrenal cell (4, 32, 33). On the other hand, the role of HDL in human steroidogenesis has not been as clearly defined. Although a stimulatory effect on cortisol and aldosterone production has been documented on adult adrenal cells (6, 7), HDL had little effect on fetal adrenal cell steroidogenesis (8, 24). So far, there have been no studies indicating an inhibitory role for either native LDL or HDL. Our findings clearly demonstrated that sHDL inhibited cortisol, aldosterone, and androstenedione. This effect may be attributed to the difference in composition between synthetic and mature native HDL. Unlike native HDL, sHDL does not have a cholesterol component. The lack of cholesterol in sHDL and its ability to efficiently take up cholesterol may explain its inhibitory effects on steroid production.

As opposed to steroidogenesis, sHDL appeared to increase HAC15 adrenal cell cholesterol biosynthesis. To assess cellular cholesterol synthesis, we monitored HMGCR and CYP51A1, 2 key enzymes in its synthetic pathway. HMGCR is the rate controlling enzyme for cholesterol biosynthesis and acts to convert acetyl-coenzyme A to the sterol precursor melavonate. Rainey et al demonstrated that ACTH, in the absence of an extracellular source of cholesterol, significantly increased levels of HMGCR activity in primary adrenal cells (34). The induction of HMGCR appears directly related to increased cellular needs for cholesterol. Studies in the 1970s by Balasubramaniam et al (35, 36) demonstrated an inverse relationship between HMGCR activity and plasma cholesterol levels in rats. When the plasma cholesterol levels were low, the adrenal gland increased cholesterol synthesis through activity of HMGCR by almost 30-fold to maintain steroidogenesis. Here, we found that transcripts for CYP51A1 and HMGCR significantly increased in HAC15 cells after treatment with sHDL, suggesting that sHDL activates cholesterol biosynthesis. This observation was in direct contrast to the inhibitory effects on steroid production.

As a possible explanation for the activation of cholesterol synthesis but inhibition of steroid production, we tested the hypothesis that sHDL caused a loss of steroid substrate as a result of cholesterol efflux from the cells. This concept is based on previous findings in macrophage cells where formulations of sHDL cause dose- and time-dependent cholesterol efflux through interaction with ATP-binding cassette transporter subfamily A member 1 (ABCA1) (37). In addition, kidney cells transfected with SR-BI, ATP-binding cassette transporter subfamily G member 1 (ABCG1), and ABCA1 also demonstrate cholesterol efflux when treated with sHDL (38, 39). We observed that sHDL caused a dose-dependent increase in the cell culture media levels of cholesterol. Inversely, intracellular cholesterol was significantly decreased at higher doses of sHDL treatment. The activation of cholesterol efflux appeared to parallel the activation of cholesterol synthesis. Taken together, our study suggests that sHDL inhibits HAC15 cell steroidogenesis through its efflux of cholesterol, causing an increase in cholesterol synthesis in an attempt to maintain output of steroid hormones.

There are potential concerns regarding the study that we attempted to resolve. One potential concern of the study was a possible cytotoxic effect of sHDL that might inhibit steroid production and raise extracellular cholesterol. To address this issue we did not see any effect on viability of cells treated with doses as high as 100μM of sHDL over 48 hours (data not shown). It is also important to note that the concentrations of sHDL used in the current in vitro study are within the levels generated in vivo during sHDL therapy. A recent study using 10–30 mg/kg of the sHDL particle, ETC-642, caused peak plasma concentrations of 0.250 mg/mL (40). Our experiments were conducted with doses up to 0.130 mg/mL (50μM). This is significantly lower than the peak plasma concentration of the lowest dose of the ETC-642 trial. Finally, the experiments of this study were performed in the HAC15 adrenal cell line. This represents the only available human steroidogenic adrenal cell model, but this model originated from an adrenocortical tumor. Therefore additional studies using primary cultures of human adrenal cells and/or in vivo animal model studies are warranted to confirm the current findings.

In summary, sHDL inhibits HAC15 adrenal cell steroidogenesis in both a dose and time-dependent manner through its effects on the efflux of cholesterol. Decreased availability of cholesterol causes a compensatory rise in adrenal cell cholesterol biosynthesis that allows the cells to partially maintain steroid output. sHDL therapeutics have great potential for treating cardiovascular disease. However, as sHDL clinical trials move forward, it will be important to monitor the adrenal hormonal axis as a potential unanticipated target.