INTRODUCTION

Top Abstract Introduction Procedures Results Discussion References

Diazepam-binding inhibitor (DBI)¹ is a highly conserved 10-kDa polypeptide that is expressed in a wide variety of species ranging from yeast to mammals. It is found in various tissues and organs and was identified independently in various different experimental settings. DBI was first isolated from rat brain based on its ability to displace diazepam (Valium) from the -aminobutyric acid A receptor (1). When injected intraventricularly in rats, it produces conflict behavior, an action that can be blocked by the benzodiazepine antagonist flumazenil (2, 3). Independently, a 10-kDa peptide was isolated from bovine liver by virtue of its ability to bind and induce the synthesis of medium-chain acyl-CoA esters. This peptide was designated acyl-CoA-binding protein (ACBP) (4). Sequencing revealed that it was the bovine homologue of rat DBI (5, 6). ACBP has been shown to bind acyl-CoA esters with high affinity, to protect them from hydrolysis and to attenuate acyl-CoA inhibition of enzymatic activities such as acetyl-CoA carboxylation (7, 8). In addition, ACBP has been shown to extract acyl-CoA from phosphatidylcholine membranes and donate it for  oxidation or glycerolipid synthesis (8). Based on these findings, together with the observation that overexpression of recombinant bovine ACBP in yeast leads to a substantial increase in cellular acyl-CoA content (9), it has been proposed that ACBP functions as a pool former and transporter of acyl-CoA. In another experimental setting, a protein able to stimulate mitochondrial steroid synthesis was isolated from adrenal cortex and found to be identical to DBI (10-12). Furthermore, in Leydig and in glial cells, DBI/ACBP has been shown to stimulate steroidogenesis by facilitating cholesterol translocation to the inner mitochondrial membrane. This process is mediated by a peripheral-type benzodiazepine receptor (see Refs. 13 and 14 for review). A protein with sequences identical to DBI was also isolated from porcine intestine and found to inhibit both early and late phases of glucose-induced insulin release from isolated perfused rat pancreas (15). This finding has been confirmed by several other investigators and may be of physiological significance (16-19). In fact, DBI-like immunoreactivity was found in cells of pancreatic islets (16-17). In yet another study, a trypsin-sensitive peptide that is secreted intraduodenally and that functions as a potent cholecystokinin-releasing peptide in the intestine was found to be identical to DBI (20). Furthermore, peptides with sequences identical to DBI have been shown to have antibacterial properties (21) and to function as paracrine or autocrine modulators of cell proliferation and cell function (22, 23).

In agreement with its expression in various tissues and cell types and its postulated roles in many different biological processes, the promoter of the DBI/ACBP gene displays all the hallmarks of a typical housekeeping gene (24, 25) but may also allow controlled activation related to specific regulatory pathways, including hormonal stimulation (26). Hormones that have been shown to stimulate the expression of DBI/ACBP include insulin and androgens. Insulin regulation of DBI/ACBP expression has been observed in 3T3-L1 preadipocytes (27). Regulation by androgens has been found in various male accessory sex organs (28) and in the human prostate cancer cell line LNCaP (25, 29). Interestingly, both conditions promote lipogenesis and involve the activation of sterol regulatory element-binding proteins (SREBPs) (30-36). SREBPs are cholesterol-regulated transcription factors that are synthesized as inactive membrane-bound precursors (see Ref. 36 for review). Proteolytic activation results in release and translocation to the nucleus. In the nucleus, SREBPs bind to specific sterol-responsive elements (SREs) and in cooperation with the generic transcription factors SP-1 or NF-Y (37-46), they coordinately modulate the transcription of a wide array of genes involved in cholesterol and fatty acid metabolism (36).

In view of the postulated role of DBI/ACBP in fatty acid and cholesterol metabolism and its induction under conditions that involve SREBP activation, we explored whether DBI/ACBP is directly controlled by SREBPs. Both LNCaP and HepG2 cells were used as experimental paradigm.

	EXPERIMENTAL PROCEDURES

Top Abstract Introduction Procedures Results Discussion References

Cell Culture-- LNCaP and HepG2 cells were obtained from the American Type Culture Collection (ATCC, Rockville, MD) and were cultured as described previously (32, 47). To assess the impact of cholesterol on gene expression, cells were incubated in media supplemented with 5% lipoprotein-deficient serum (LPDS) (Perimmune, Rockville, MD) in the absence or presence of 10 µg/ml cholesterol and 1 µg/ml 25-hydroxycholesterol added from stock solutions in ethanol. Control cultures received similar amounts of ethanol only. In experiments assessing the effects of steroids, fetal calf serum was pretreated with dextran-coated charcoal (CT-FCS) to reduce the background levels of steroids. The synthetic androgen R1881 (methyltrienolone), purchased from DuPont New England Nuclear (Dreiech, Germany), was dissolved in absolute ethanol and added to the cultures. Final ethanol concentrations did not exceed 0.2%. All experiments involving LNCaP cells were carried out with cells of passages 30 to 75.

Electrophoretic Mobility Shift Assay-- Recombinant 6×His-tagged SREBP-la (amino acids 1-490) was expressed in Escherichia coli BL21(DE3)pLysS (Stratagene, La Jolla, CA) from plasmid pRSA (kindly provided by Dr. T. Osborne, Department of Molecular Biology and Biochemistry, University of California, Irvine, USA) (37) and was purified by means of TALON metal affinity chromatography (CLONTECH). Complementary single-stranded oligonucleotides corresponding to nucleotides 137 to 113 of the published human DBI promoter sequence (25) were end-labeled with [-³²P]ATP and T4 polynucleotide kinase and annealed. The probe (50,000 cpm) was incubated with recombinant SREBP-1a in a solution containing 10 mM Tris-HCl, pH 7.6, 50 mM NaCl, 0.05 mM EDTA, 2.5 mM MgCl₂, 8.5% glycerol, l mM dithiothreitol, 0.5 µg/ml poly(dI-dC), 0.1% Triton X-100, and 2.5 mg/ml nonfat milk for 20 min on ice. Unlabeled wild type or mutated double-stranded oligonucleotides were added as indicated in the figure legends. In the oligonucleotides designated "mutSRE," the SRE-like sequence CTCGCCCGAG was replaced by CTACAAAATG. Oligonucleotides designated "delSRE" encompassed bases 145 to 105, but lacked the SRE-like sequence. Where indicated, antibody K-10 against SREBP-la (Santa Cruz Biotechnology, Santa Cruz, CA) was added and incubation on ice was continued for another 20 min. The DNA-protein complexes were resolved on a 5% nondenaturing polyacrylamide gel in 0.25× TBE and 0.02% Triton X-100 at room temperature. Gels were vacuum-dried and exposed to Kodak X-Omat AR film.

Plasmids-- A cDNA fragment encoding amino acids 1-460 of human SREBP-la was generated by polymerase chain reaction and inserted into the eukaryotic expression vector pIRES1neo (CLONTECH). The resulting plasmid was designated pSREBP-la_1-460. Plasmid pDBI-264luc encompassing bases 264 to 13 of the human DBI/ACBP gene linked to a luciferase reporter gene has been described previously (25) and is here referred to as "wt." Mutations or deletions within the promoter were created using the Quikchange mutagenesis kit (Stratagene, La Jolla, CA). The constructs mutSRE and delSRE contain the same mutation and deletion, respectively, as the oligonucleotides described under "Electrophoretic Mobility Shift Assay." In construct "mutNF-Y," the A at position 142, which is part of a putative NF-Y site, was replaced by a C. In construct +4A, 4 A residues were inserted at position 131, which is located between the NF-Y and the SRE-like sequence. Plasmid pFASluc, a fatty acid synthase promoter-reporter construct, and plasmid pSV-AR₀ expressing the androgen receptor, have been described previously (32, 48). Plasmid pPA-7, a luciferase-based plasmid with the prostate-specific antigen (PSA) promoter (49), was kindly provided by Dr. J. Trapman (Erasmus University, Rotterdam, The Netherlands).

Transient Transfections and Reporter Gene Assays-- LNCaP cells were seeded in 6-cm dishes in Dulbecco's modified Eagle's medium containing 10% fetal calf serum at a density of 7 × 10⁵ cells. On the next day, the medium was replaced with Dulbecco's modified Eagle's medium with 2% CT-FCS. Cells were transfected with 5 µg of the indicated luciferase reporter constructs, the indicated amounts of pIRES1neo or pSREBP-1a_1-460, and a plasmid encoding -galactosidase. After 4 h of exposure to DNA, cells were subjected to a glycerol shock and washed with phosphate-buffered saline. To explore potential effects of androgens, cells were cotransfected with an androgen receptor expression vector (pSV-ARo) (48) and incubated with 10⁸ M R1881 or with ethanol vehicle in medium containing 5% CT-FCS. In experiments assessing the effects of sterols, cells were incubated in medium containing 5% LPDS in the absence or in the presence of 10 µg/ml cholesterol and 1 µg/ml 25-hydroxycholesterol, or ethanol vehicle. One day after treatment, cells were washed with phosphate-buffered saline and harvested in 500 µl of reporter lysis buffer (Promega, Madison, WI). Aliquots of 10 µl of cleared lysate were assayed for luciferase activity using a luciferase reporter assay kit from Promega and a Berthold Microlumat LB 96P luminometer. The activity of -galactosidase was used to normalize for transfection efficiencies.

Northern Blot Analysis-- Total RNA was prepared using a modified guanidinium/CsCl ultracentrifugation method as described previously (29). Equal aliquots of total RNA (20 µg) were denatured and subjected to electrophoresis in a 1% agarose gel containing formaldehyde. The RNA was transferred to Biotrans + membranes (ICN Pharmaceuticals, Inc., Costa Mesa, CA), prehybridized and hybridized with DBI, FAS, PSA, and 18 S probes as described before (29, 31). Blots were autoradiographed by exposure to Amersham Hyperfilm-MP or to Kodak Biomax film (Amersham International, Buckinghamshire, UK). Hybridization signals were quantitated using PhosphorImager screens (Molecular Dynamics, Sunnyvale, CA) and normalized for differences in RNA loading.

	RESULTS

Top Abstract Introduction Procedures Results Discussion References

Identification of a SREBP-binding Site in the DBI/ACBP Promoter-- Examination of the nucleotide sequences of the human DBI/ACBP promoter for potential SREBP sites revealed a sequence CTCGCCCGAG at positions 127 to 118 of the published sequence (25), resembling the 10-base pair SREs found in other SREBP-regulated genes (Fig. 1). This sequence is most homologous to the SRE of the farnesyl-diphosphate synthase gene (57) and differs only at two positions. In agreement with the requirement for a coregulator for SREBP function (SP-1 or NF-Y) (37-46), the putative SRE is closely positioned to a reverse CCAAT box, which is a potential binding site for the generic heterotrimeric transcription factor NF-Y. Interestingly, both SRE and NF-Y sites are perfectly conserved in the otherwise less homologous rat DBI/ACBP promoter (Fig. 1A).

View larger version (23K): [in this window] [in a new window]   Fig. 1.   Nucleotide sequence alignment of the promoter region of the human and rat DBI/ACBP gene (A) and comparison of SREBP sites from known sterol-responsive genes with the putative SRE site of the DBI/ACBP gene (B). The nucleotide sequences of the human (h) and rat (r) DBI/ACBP genes, as reported previously (24, 25) are aligned in panel A. The sequences corresponding to the putative binding sites for NF-Y, SREBP and TATA-like sequences are boxed. B, the sequences of the SREBP sites of the hamster (ham), rat (r), frog (f), human (h), and mouse (m) LDL receptor, of the rat fatty acid synthase gene (rFAS) at 150, 72, and 62, human SREBP-2, hamster HMG-CoA synthase promoter (hamSyn) elements 1 and 2, hamster HMG-CoA reductase (hamRed) at 150 and 165, mouse glycerol-3-phosphate acyltransferase (mGPAT), farnesyl-diphosphate synthase (FPPsyn), and DBI/ACBP are aligned. The underlined residue represents a base that separates two direct repeats (arrows). Asterisks indicate residues that are conserved in all aligned sequences. SREBP sites from the acetyl-CoA carboxylase gene (58) and the caveolin gene (59), which show a limited degree of sequence similarity or contain only a half-site homology, were not included in the comparison.

View larger version (60K): [in this window] [in a new window]   Fig. 2.   SREBP-1a binding to the SRE-like sequence in the human DBI/ACBP promoter. Oligonucleotides corresponding to a 25-base pair DBI/ACBP promoter fragment encompassing the SRE-like sequence were radiolabeled (probe wt) and incubated with recombinant SREBP-1a1-490 in the absence or in the presence of a 50-fold molar excess of unlabeled competitor. The competitors used were: wt oligonucleotides, oligonucleotides in which the SRE-like sequence CTCGCCCGAG was mutated to CTACAAAATG (mutSRE), oligonucleotides with a deletion of the SRE (delSRE), or a DNA fragment corresponding to a well characterized SRE-containing fragment of the low density lipoprotein receptor (LDLR). Where indicated, a polyclonal antibody against SREBP1a (anti-SREBP) was added. After incubation on ice, the binding reaction mixtures were subjected to electrophoretic mobility shift assay on a nondenaturing polyacrylamide gel as described under "Experimental Procedures." SS, supershift.

Transcriptional Regulation of DBI/ACBP Promoter-Reporter Genes by Coexpressed SREBP-- In order to determine whether SREBP binding is functional, we transiently transfected LNCaP cells with a DBI/ACBP promoter-reporter construct harboring the SRE site (pDBI-264luc, here referred to as wt), together with a plasmid encoding -galactosidase and increasing amounts of pSREBP-1a_1-460, a plasmid encoding transcriptionally active SREBP-1a. Two days after the transfection, the luciferase activity was measured and the values were corrected for any differences in transfection efficiency, as determined from the -galactosidase assay. As Fig. 3A shows, the transcriptional activity of the DBI/ACBP promoter was elevated with increasing amounts of co-transfected pSREBP-1a_1-460. Maximal effects were reached at 20 ng of pSREBP-1a_1-460. In order to demonstrate that the stimulation of transcriptional activity by SREBP-la overexpression is mediated by the SRE-like site, we generated DBI/ACBP promoter-reporter constructs in which the SRE site is mutated (mutSRE) or deleted (delSRE) (Fig. 3B). Transfection experiments were carried out as described above with maximally effective amounts of pSREBP-1a_1-460 (20-50 ng). In support of the involvement of the SRE site in SREBP-induced transcriptional activation, stimulation of luciferase activity was severely decreased when the SRE was deleted or mutated (Fig. 3C). Additionally, the basal transcriptional activity of the mut and del constructs was 5-10-fold lower than that of the wild type construct, demonstrating the importance of the SRE site for the transcriptional activity of the DBI/ACBP gene (Fig. 3D).

View larger version (15K): [in this window] [in a new window]   Fig. 3.   Transcriptional activation of DBI/ACBP promoter-reporter constructs by coexpression of SREBP-1a. A, triplicate dishes of LNCaP cells were transiently cotransfected with a DBI/ACBP promoter luciferase construct harboring the SRE-like site (pDBI-264luc, here referred to as wt), together with a plasmid encoding -galactosidase and increasing amounts of pSREBP-1a1-460 plasmid expressing a transcriptionally active form of SREBP-1a. Analogous control transfections were carried out with increasing amounts of pIRES1neo, the empty expression vector. Two days after transfection, the luciferase activity was measured, corrected for any differences in transfection efficiency as determined from the -galactosidase assay, and expressed relative to the control transfections with the empty expression vector (pIRES1neo). The results shown are representative of two independent experiments. B, schematic representation of the wt DBI/ACBP promoter-reporter construct and those with a mutated (mutSRE) or deleted SRE (delSRE). The putative binding site for NFY is indicated. The sequences of the wild type and mutated SRE-like site are shown. C, LNCaP cells were transiently cotransfected with the DBI/ACBP promoter-reporter constructs shown in panel B, together with a -galactosidase-encoding plasmid and a maximally effective amount of pSREBP-la1-460 (20-50 ng) or with pIRES1neo. Luciferase activity was measured, corrected for any differences in transfection efficiency, and expressed as -fold stimulation relative to the control transfection with pIRES1neo as described in A. Values represent the means ± S.E. of triplicate dishes. The results are representative of three separate experiments. D, basal promoter activities of the DBI/ACBP promoter-reporter constructs shown in panel B were measured as luciferase activities and were compared relative to the normalized values of the wt construct. Values represent the means ± S.E. of triplicate dishes. The results are representative of three experiments.

Requirement of NF-Y Binding for SREBP Activation of DBI/ACBP Transcription-- SREBPs are weak activators of transcription in isolation and are known to function more efficiently when a co-regulatory factor (SP-1 or NF-Y) binds to a neighboring site (37-46). The SRE in the DBI/ACBP promoter is preceded by a potential NF-Y site. To test whether this latter site is important for SREBP-induced activation of DBI/ACBP transcription, we generated a DBI/ACBP promoter-reporter construct in which the NF-Y site is mutated (Fig. 4A). In another construct, we increased the distance between the NF-Y and the SRE sites by insertion of four deoxyadenosine residues. Transient co-transfection of these constructs with pSREBPla_1-460 revealed that the stimulatory effect of SREBP overexpression as observed with the wild type construct is severely diminished when the NF-Y site is mutated or when the distance between the NF-Y and the SRE sites is modified (Fig. 4B). Basal transcriptional activities, however, remained high. Mutation of the NF-Y site even led to a 3-4-fold increase in basal transcription. Insertion of four bases between the NF-Y and the SRE site caused a 2-fold decrease in luciferase activity (Fig. 4C).

View larger version (18K): [in this window] [in a new window]   Fig. 4.   Requirement of NF-Y binding for SREBP activation of DBI/ACBP transcription. A, schematic representation of the wt DBI/ACBP promoter-reporter construct, mutNF-Y, in which the A in the NF-Y site is mutated to a C, and construct +4A, in which 4A residues were inserted in between the NF-Y and the SRE site. B, triplicate dishes of LNCaP cells were transiently cotransfected with the DBI/ACBP promoter-reporter constructs shown in panel A together with a plasmid encoding -galactosidase and 50 ng of pSREBP-1a1-460 or the empty pIRES1neo vector. Two days after the transfection, luciferase activity was measured, normalized for -galactosidase activity and expressed as -fold stimulation relative to control transfections with pIRES1neo. Values represent the means ± S.E. The results are representative of three independent experiments. C, basal promoter activities of the DBI/ACBP promoter-reporter constructs shown in panel A were measured as luciferase activities and were compared relative to the normalized luciferase activity of the wt construct. Values represent the means ± S.E. The results are representative of two separate experiments.

Regulation of DBI/ACBP Expression by Sterols-- Having demonstrated that overexpression of SREBPs leads to SRE-dependent activation of the transcriptional activity of DBI/ACBP promoter-reporter genes, we examined whether these constructs are also responsive to physiological changes in endogenous SREBP levels. One of the main and universal physiological signals that trigger SREBP processing resulting in increased nuclear levels of SREBPs is cholesterol depletion. Fig. 5A shows that the luciferase activity of LNCaP cells that were transiently transfected with DBI/ACBP promoter-reporter constructs was 2-fold higher in cells that were deprived of cholesterol as compared with sterol-treated cells. Similar results were obtained with a luciferase reporter construct harboring the promoter of the fatty acid synthase gene (31), a well known sterol-regulated gene (38, 54, 60). The promoter activity of a reporter construct containing a promoter fragment of the PSA gene (a gene encoding a prostate-secreted protein that is not directly related to cholesterol and fatty acid metabolism) (49), was not affected by sterols, indicating that the effects of cholesterol were specific. In support of the involvement of SREBPs in the effects of cholesterol depletion on DBI/ACBP gene transcription, no effects of sterols were observed when LNCaP cells were transfected with DBI/ACBP constructs in which the SRE-site was mutated or deleted (Fig. 5A).

View larger version (31K): [in this window] [in a new window]   Fig. 5.   Regulation of DBI/ACBP gene expression by sterols. A, LNCaP cells were transiently cotransfected with the indicated promoter-reporter constructs and then incubated in medium supplemented with 5% lipoprotein-deficient serum in the absence () or in the presence (+) of sterols (10 µg/ml cholesterol and 1 µg/ml 25-hydroxycholesterol). One day after treatment, luciferase activity was measured. Values represent the means ± S.E. of three individual experiments and are expressed relative to the values obtained in the absence of sterols, which were arbitrarily set at 1.0. B, LNCaP cells were incubated in medium supplemented with 5% lipoprotein-deficient serum in the absence () or in the presence (+) of sterols (10 µg/ml cholesterol and 1 µg/ml 25-hydroxycholesterol). One day after treatment, total RNA was prepared and subjected to Northern blot analysis with the indicated probes. Hybridization signals were quantitated using PhosphorImager screens, normalized for differences in RNA loading and expressed relative to the values obtained in the absence of sterols. C, HepG2 cells were incubated in medium supplemented with 5% lipoprotein-deficient serum and 20 µM mevastatin in the absence () or in the presence (+) of sterols (10 µg/ml cholesterol and 1 µg/ml 25-hydroxycholesterol). Two days after treatment, total RNA was prepared and subjected to Northern blot analysis as described in B.

Involvement of SREBPs in the Regulation of DBI/ACBP Expression by Androgen-- Based on our recent findings that androgens coordinately stimulate the expression of lipogenic genes in LNCaP cells and that these effects are (at least in part) mediated by an androgen-induced increase in nuclear SREBP levels (32), we determined whether the previously reported effects of androgens on DBI/ACBP expression (25, 29) are also mediated by SREBPs. To this end, we transfected LNCaP cells with DBI/ACBP promoter-reporter constructs together with a -galactosidase encoding plasmid and a plasmid encoding the androgen receptor. This latter plasmid was included since the wild type DBI/ACBP promoter-reporter construct is poorly responsive to androgens in standard transient transfection experiments in LNCaP cells (25). Cotransfection with an androgen receptor plasmid has been shown to improve the androgen responsiveness of promoter-reporter constructs from several other genes including prototypical androgen-responsive genes such as PSA (61, 62). Under these conditions androgens consistently stimulated the transcriptional activity of the wild type DBI/ACBP construct (Fig. 6). In support of the involvement of SREBPs in the effects of androgens, stimulatory effects of androgens were severely reduced when the SRE site was mutated or deleted.

View larger version (27K): [in this window] [in a new window]   Fig. 6.   Involvement of the SRE site in the regulation of DBI/ACBP expression by androgens. LNCaP cells were transiently cotransfected with the indicated DBI/ACBP promoter-reporter constructs, and then incubated in medium supplemented with 5% CT-FCS in the absence () or in the presence (+) of 108 M amount of the synthetic androgen R1881. One day after treatment, luciferase activity was measured. Values represent the means ± S.E. of five individual experiments and are expressed relative to the values obtained in the absence of R1881.

	DISCUSSION

Top Abstract Introduction Procedures Results Discussion References

The current experiments provide evidence that DBI/ACBP is a SREBP-responsive gene. A SRE-like sequence was found in the human DBI/ACBP promoter by visually scanning the DNA sequence. This site resembles the SREs of other SREBP-responsive genes and is present also in the promoter of the rat homologue. The SRE-like site is functional; it binds purified SREBP-1a and mediates transcriptional activation by overexpressed SREBP-1a in cotransfection experiments. Like the genes encoding farnesyl diphosphate synthase, HMG-CoA synthase, squalene synthase, SREBP-2, and glycerol-3-phosphate acyltransferase (41-46), regulation by SREBPs requires a neighboring binding site for the generic transcription factor NF-Y. Furthermore, physiological conditions that are known to activate SREBP processing and stimulate their nuclear translocation (such as cholesterol depletion or androgen treatment of LNCaP cells) increase the expression of DBI/ACBP promoter-reporter genes and of the endogenous gene. Consistent with this finding is the report by Hansen et al. (27) that DBI/ACBP expression is stimulated during insulin-induced lipogenesis in 3T3-L1 preadipocytes, another physiological condition that involves changes in SREBP levels (35). Together with our finding that a functional SRE is important also for basal DBI/ACBP gene transcription, the current experiments illustrate the importance of SREBPs in the regulation of DBI/ACBP gene expression.

Regulation of DBI/ACBP by SREBPs is consistent with the postulated role for DBI/ACBP in fatty acid metabolism and allows DBI/ACBP to be coregulated with other proteins and enzymes involved in lipid metabolism. Our finding that DBI/ACBP is under the control of cholesterol may be of special interest in view of the role of DBI/ACBP in cholesterol translocation across mitochondrial membranes, a rate-limiting step in the biochemical synthesis of steroids.

Our observation that the here identified SRE also plays a role in the androgen regulation of DBI/ACBP transcription is consistent with our previous finding that androgens enhance the nuclear accumulation of SREBP-l in LNCaP cells, and is reminiscent of the involvement of SREBPs in the androgen regulation of FAS gene expression (32). Furthermore, the involvement of SREBPs in the androgen regulation of DBI/ACBP expression may fit with our previous observation that androgen regulation of DBI/ACBP expression may be indirect (29). It is unlikely, however, that the entire effect of androgens on DBI/ACBP expression is mediated by the SRE. (i) The effects of androgens on the 264 DBI/ACBP promoter-reporter construct are observed only under conditions (cotransfection with an androgen receptor-expressing vector) that enhance androgen responsiveness. (ii) The effects of androgens on steady state mRNA for DBI/ACBP are more pronounced than the effects on transcriptional activation of the DBI/ACBP luciferase reporter constructs in transient transfection experiments. (iii) Another androgen-responsive region was found in the DBI/ACBP promoter sequence upstream of the SRE site (25).

Issues that remain to be addressed include the question whether different SREBPs (SREBP-1a, SREBP-1c, SREBP-2) are equally effective in regulating DBI/ACBP expression, and whether changes in SREBP-mediated DBI/ACBP expression may also affect other functions of DBI/ACBP not directly related to lipid metabolism.