Androgenic Suppression of ATP-binding Cassette Transporter A1 Expression in LNCaP Human Prostate Cancer Cells

Prostate cancer is a significant source of morbidity and mortality, especially in many Western societies. More than 60 years ago, Charles Huggins showed that androgens regulate the growth of advanced prostate tumors, and this finding ushered in hormonal therapy for treatment of this disease (1). However, in most cases, androgen ablation is only a temporary respite from this disease, and soon, the cancer becomes androgen-independent and continues to grow. Other targets are urgently needed to control the growth and progression of prostate cancer. Recently, Ettinger et al. (2) reported that sterol response element-binding proteins (SREBPs), the transcriptional regulators that control the metabolic pathway of lipogenesis and cholesterol, are activated in androgen-independent tumors. The normal prostate produces and secretes cholesterol in prostatic fluid, and the level of both intracellular and secreted cholesterol are 2- to 10-fold higher in benign prostatic hyperplasia and carcinoma of the prostate (3, 4). Therefore, exploring the mechanism of alteration of lipid metabolism in prostate cancer may be helpful in elucidating the molecular mechanism responsible for prostate cancer cell progression. To mimic prostate cancer progression, we established a cell culture system with an androgen-dependent LNCaP clonal cell line, LNCaP 104-S. After long-term culture in androgen-depleted medium, LNCaP 104-S cells progress to androgen-independent LNCaP 104-R1 and 104-R2 cells (5, 6). Using our progression model, we analyzed the expression levels of genes that play important roles in cholesterol and lipid homeostasis in the androgen-dependent 104-S and androgen-independent 104-R1 and 104-R2 cells. We report that the expression levels of ATP-binding cassette transporter A1 (ABCA1), which functions as a cholesterol exporter (7), is altered during prostate cancer progression, is regulated by androgens, and can regulate cell proliferation.

The synthetic androgen 17β-hydroxy,17α-methylestratrien-3-one (R1881) was obtained from New England Nuclear (Boston, MA). The antiandrogen Casodex [ICI 176,334; (2RS)-4′-cyano-3-(4-fluorophenylsulfonyl)-2-hydroxy-2-methyl-3′-(trifluoromethyl)-propionanilide; bicalutamide] was from AstraZeneca Pharmaceuticals (Wilmington, DE). A polyclonal antibody against ABCA1 was prepared as described previously (8).

LNCaP 104-S, 104-R1, and 104-R2 cells were maintained and cultured as described previously (5, 6). To observe androgenic effects, cells were grown in DMEM supplemented with 10% dextran-coated charcoal-stripped fetal bovine serum.

Total RNA was isolated with the TRIZOL Reagent (Invitrogen, Carlsbad, CA) and was treated with DNase I (DNA-free, Ambion, Austin, TX). Reverse transcription was performed with random hexamers and Moloney murine leukemia virus reverse transcriptase (Omniscript, Qiagen, Valencia, CA). The TaqMan primer/probe was designed using Primer Express (Applied Biosystems, Foster City, CA). The sequences of primers and probes are as follows: ABCA1 primers, 5′-TGTCCAGTCCAGTAATGGTTCTGT-3′ and 5′-AAGCGAGATATGGTCCGGATT-3′, and ABCA1 probe, 5′-ACACCTGGAGAGAAGCTTTCAACGAGACTAACC-3′; ABCG1 primers, 5′-CCAGAAGTCGGAGGCCATC-3′ and 5′-AAGTCCAGGTACAGCTTGGCA-3′, and ABCG1 probe 5′-TGCGGGAGCTGGACGTGGAAA-3′; LXRα primers, 5′-GCCCTGCATGCCTACGTCT-3′ and 5′-TCATTAGCATCCGTGGGAACA-3′, and LXRα probe, 5′-CATCCACCATCCCCATGACCGACT-3′; LXRβ primers, 5′-GGAGCTGGCCATCATCTCA-3′ and 5′-GTCTCTAGCAGCATGATCTCGATAGT-3′, and LXRβ probe, 5′-CAGGAGATCGTGGACTTCGCTAAGCAA-3′; SREBP-1a primers, 5′-TGCTGACCGACATCGAAGAC-3′ and 5′-CTGTCTTGGTTGTTGATAAGCTGAA-3′, and SREBP-1a probe, 5′-ATCGCGGAGCCATGGATTGCACT-3′; SREBP-1c primers, 5′-GGTAGGGCCAACGGCCT-3′ and 5′-CTGTCTTGGTTGTTGATAAGCTGAA-3′, and SREBP-1c probe, 5′-ATCGCGGAGCCATGGATTGCACT; The 5′-end of the probe was labeled with the reporter-fluorescent dye, carboxyfluorescein. The 3′-end of probe was labeled with the quencher dye, 6-carboxytetramethylrhodamine. Real-time PCR was performed on an ABI PRISM 7700 system (Applied Biosystems) with the QuantiTect Probe PCR protocol (Qiagen). The rRNA Control Kit (Applied Biosystems) was used to normalize transcript levels between samples.

A cellular membrane fraction was prepared with Mem-PER (PIERCE, Rockford, IL), according to the manufacturer’s instructions. Protein concentration was determined with the Bradford reagent (Bio-Rad Laboratories, Hercules, CA) with a BSA standard. Proteins were separated on 6% polyacrylamide gels containing SDS. Electrophoresis and blotting were performed as described previously (5).

A DNA fragment containing a 1.7-kb portion of the human ABCA1 gene (representing nucleotides −211 to +1439 relative to the transcriptional start site) was amplified by PCR from 104-S genomic DNA. The primers were designed from the published ABCA1 gene sequence and consisted of 5′-CTGGAGATCCTGTTGACTGTAGCAT-3′ and 5′-CACTTTTGTGTTTGCGTCTCTT-3′. The 1.7-kb amplification product was cloned into the pSTBlue-1 cloning vector (Novagen, Madison, WI). After the sequence was verified, the ABCA1 promoter fragment was transferred from pSTBlue-1 to the pGL3 firefly luciferase reporter plasmid (Promega, Madison, WI). The c-fos-Ren plasmid consists of nucleotides −56 to +109 of the human c-fos minimal promoter (9) inserted into pGL3 and substitution of the firefly luciferase gene with a Renilla luciferase gene from phRL-SV40 (Promega).

Cells were seeded at 1 × 10⁵ cells per well on a 12-well plate, and then, the indicated compounds were added. Forty-eight hours later, cells were transiently transfected with 0.2 μg/well of the ABCA1 promoter-reporter plasmid and 0.1 μg/well of the c-fos-Ren plasmid (normalization reporter plasmid) with Effectene (Qiagen). After 12 hours, the medium was changed, and the indicated compounds added again. After an additional 24 hours, cells were harvested and luciferase activity was measured with a Dual-Luciferase kit (Promega) with a Monolight luminometer (PharMingen, San Diego, CA).

The RNA interference sequence was designed by using the AA Scanning program from OligoEngine (Seattle, WA). DNA coding for a RNAi for human ABCA1 was prepared with the following oligonucleotides: 5′-GATCCCCTGAGTTTAGGTATGGCGGCTTCAAGAGAGCCGCCATACCTAAACTCATTTTTGGAAA-3′ and 5′-AGCTTTTCCAAAAATGAGTTTAGGTATGGCGGCTCTCTTGAAGCCGCCATACCTAAACTCAGGG-3′. These 64-mer oligonucleotides were annealed and ligated into the pH1RP vector (10), and the construct was verified by sequencing. The ABCA1-RNAi expression plasmid was stably transfected into 104-S cells with Effectene and selection for G418 resistance.

Cell number was analyzed by measuring DNA content with fluorescent dye Hoechst 33258 as described previously (10).

Data are presented as the mean +/− SD or SE of three experiments or are representative of experiments repeated at least three times.

To investigate the relationship between prostate cancer progression and the expression levels of the genes playing important roles in cholesterol and lipid homeostasis, we analyzed gene expression levels by using real-time quantitative PCR analysis. In androgen-dependent LNCaP 104-S cells, treatment with the synthetic androgen, R1881, drastically suppressed ABCA1 gene expression (15- to 20-fold; Fig. 1,A). ABCG1 also has been proposed to play a role in cholesterol and phospholipid efflux (7). ABCG1 mRNA levels in 104-R1 and 104-R2 cells were 2.5-fold higher than in 104-S cells (Fig. 1,B). As with ABCA1, androgen down-regulated ABCG1 mRNA levels in 104-S cells by ∼50% compared with vehicle-treated cells. Neither the expression levels of mRNA for SREBP-1a, which preferentially regulates cholesterol synthesis, nor the mRNA expression of SREBP-1c, which controls expression of various lipogenic enzymes (11), were affected by androgens (Fig. 1, C and D). Consistent with the studies by Ettinger et al. (2), the expression of SREBP-1c was increased in androgen-independent 104-R1 and 104-R2 cells. We also measured the mRNA levels of liver X receptor (LXR) α and β because these nuclear receptors are activated by oxysterols and directly up-regulate the transcription of the mRNA for ABCA1, ABCG1, SREBP-1c, and LXRα itself (12). LXRα is preferentially expressed in the liver, intestine, adipose tissue, and macrophages (12), whereas LXRβ, which was discovered in our laboratory, is expressed ubiquitously and was named ubiquitous receptor (13). LXRα mRNA expression was 2-fold lower in androgen-independent 104-R1 and 104-R2 cells than in 104-S cells (Fig. 1,E), and LXRβ mRNA expression was similar in 104-S cells and 104-R1 cells but was much lower in 104-R2 cells (Fig. 1 F). Treatment of cells with androgen did not significantly alter the mRNA levels for LXRs.

The effect of androgen treatment or withdrawal on the kinetics of ABCA1 mRNA expression in 104-S cells was studied. The maximum level of ABCA1 mRNA was achieved at 72 hours after the onset of androgen depletion (Fig. 2,A). Addition of androgen to the androgen-depleted culture decreased the maximum level of ABCA1 expression with kinetics similar to induction (solid triangle with dashed line in Fig. 2,A). The androgenic suppression of ABCA1 expression by 0.1 nmol/L R1881 was reversed by 5 μmol/L Casodex (Fig. 2,B). Casodex was unable to block the suppressive effects of 10 nmol/L R1881. This is most likely because Casodex has a 50- to 100-fold lower affinity for androgen receptor than R1881. The effect of Casodex suggests that androgenic suppression is mediated by the androgen receptor. In androgen-independent 104-R1 and 104-R2 cells, ABCA1 gene expression was low relative to 104-S cells and was not significantly affected by R1881 or Casodex (Fig. 1 and data not shown). Consistent with mRNA expression studies, Western blot analysis showed that ABCA1 was expressed in androgen-dependent 104-S cells in the absence of R1881 or in the presence of Casodex (Fig. 3 A). Expression of ABCA1 protein, with or without androgen, was not detected in androgen-independent 104-R1 and 104-R2 cells.

We investigated the effect of androgen on ABCA1 promoter activity with transient transfection of 104-S cells with a luciferase reporter gene linked to a 1.7-kb fragment of the ABCA1 gene promoter adjacent to the transcriptional start. R1881 suppressed ABCA1 promoter activity in 104-S cells (Fig. 3,B) but not in 104-R1 cells (Fig. 3,C). This suppression was reversed by 10 μmol/L Casodex in 104-S cells (Fig. 3,B). No obvious androgen-response element was present in the 1.7-kb promoter fragment. Because strong androgenic suppression of ABCA1 promoter activity was observed in androgen-dependent 104-S cells but not in androgen-independent 104-R1 cells (Fig. 3), a 104-S cell-specific trans-acting factor may be involved in androgenic regulation of ABCA1 in 104-S cells. Although the induction of ABCA1 expression by cyclic AMP is well characterized (14), the intracellular levels of cyclic AMP were not altered by androgen (data not shown), and there is no cyclic AMP response element in this 1.7-kb ABCA1 promoter. This region contains several binding sites for trans-acting factors. Among these binding sites, the Oct1 site is a potentially interesting candidate because Oct1 has been reported to be related to androgenic suppression of the rat dehydroepiandrosterone sulfotransferase gene (15). Additional studies will be required to identify trans-acting factors that may be involved in prostate cancer progression and how they mediate the androgenic suppression of ABCA1 expression.

Because the expression of ABCA1 mRNA was drastically suppressed by androgen in 104-S cells and was decreased during progression from104-S to 104-R1 cells, we examined the role of ABCA1 in 104-S cell proliferation with RNAi to decrease the ABCA1 expression level. An RNAi sequence for human ABCA1 was expressed in 104-S cells with the RNAi-expression vector pH1RP. Cells transformed with pH1RP-ABCA1-RNAi (clone 1 and 2) showed significant decreases in ABCA1 mRNA expression measured by real-time quantitative PCR, compared with control cells transformed with pH1RP empty vector (Fig. 4,A). Decreases in ABCA1 expression in these ABCA1-RNAi transformants were confirmed by Western blot analysis (Fig. 4,B). ABCA1-RNAi transformants had increased growth rates compared with control transformants (Fig. 4,C). Although androgen still stimulated the proliferation of these transformants, the growth rate of the ABCA1-knockdown cells was faster than control cells. In the absence of androgen, these transformants grew as fast as control cells treated with R1881, although cell cycle distribution in ABCA1-knockdown cells was not significantly changed (data not shown). We observed that in the absence of androgen, 104-S cells express a high level of ABCA1 (Fig. 1). The observations that 104-S cells grow faster when the expression of the ABCA1 gene is down-regulated (Fig. 4 C) is consistent with ABCA1 expression inhibiting LNCaP growth. The molecular mechanism by which ABCA1 affects proliferation is not clear at present. One possibility is that a decrease in ABCA1 expression may modulate cholesterol and/or phospholipid levels in lipid raft domains, which function as membrane platforms for cellular signal transduction (16). Additional exploration of ABCA1 function in our system may also enable us to define how cholesterol and/or lipids directly contribute to proliferation. Another possibility is that the interaction of the ABCA1 with Fas-associated death domain protein (17) may affect LNCaP 104-S cell proliferation because Fas-associated death domain protein has been reported to be involved in proliferation, as well as apoptosis (18). Recently, several ABCA1-interacting proteins involved in multiple cellular functions have been identified (19, 20). Some of these interactions may contribute to ABCA1-mediated regulation of proliferation in LNCaP 104-S cells.

Our findings point to an important role for ABCA1 in human androgen-dependent prostate cancer LNCaP cells, and androgen regulation of LNCaP cell proliferation may involve transcriptional suppression of ABCA1. Our results also suggest that it may be possible to modulate prostate cancer growth by induction of ABCA1, for example, through use of LXR agonists because the ABCA1 gene is a direct transcriptional target of LXR.

We thank Dr. Itsuko Ishii of Chiba University for valuable advice.