Effects of CYP7A1 overexpression on cholesterol and bile acid homeostasis

Transcription

1 Am J Physiol Gastrointest Liver Physiol 281: G878 G889, Effects of CYP7A1 overexpression on cholesterol and bile acid homeostasis W. M. PANDAK, 1 C. SCHWARZ, 1 P. B. HYLEMON, 2 D. MALLONEE, 2 K. VALERIE, 1 D. M. HEUMAN, 1 R. A. FISHER, 3 KAYE REDFORD, 1 AND Z. R. VLAHCEVIC 1 Departments of 1 Medicine, 2 Microbiology, and 3 Surgery, Veterans Affairs Medical Center and Virginia Commonwealth University, Richmond, Virginia Received 21 July 2000; accepted in final form 25 June 2001 Pandak, W. M., C. Schwarz, P. B. Hylemon, D. Mallonee, K. Valerie, D. M. Heuman, R. A. Fisher, Kaye Redford, and Z. R. Vlahcevic. Effects of CYP7A1 overexpression on cholesterol and bile acid homeostasis. Am J Physiol Gastrointest Liver Physiol 281: G878 G889, The initial and rate-limiting step in the classic pathway of bile acid biosynthesis is 7 -hydroxylation of cholesterol, a reaction catalyzed by cholesterol 7 -hydroxylase (CYP7A1). The effect of CYP7A1 overexpression on cholesterol homeostasis in human liver cells has not been examined. The specific aim of this study was to determine the effects of overexpression of CYP7A1 on key regulatory steps involved in hepatocellular cholesterol homeostasis, using primary human hepatocytes (PHH) and HepG2 cells. Overexpression of CYP7A1 in HepG2 cells and PHH was accomplished by using a recombinant adenovirus encoding a CYP7A1 cdna (AdCMV-CYP7A1). CYP7A1 overexpression resulted in a marked activation of the classic pathway of bile acid biosynthesis in both PHH and HepG2 cells. In response, there was decreased HMG-CoA-reductase (HMGR) activity, decreased acyl CoA:cholesterol acyltransferase (ACAT) activity, increased cholesteryl ester hydrolase (CEH) activity, and increased low-density lipoprotein receptor (LDLR) mrna expression. Changes observed in HMGR, ACAT, and CEH mrna levels paralleled changes in enzyme specific activities. More specifically, LDLR expression, ACAT activity, and CEH activity appeared responsive to an increase in cholesterol degradation after increased CYP7A1 expression. Conversely, accumulation of the oxysterol 7 -hydroxycholesterol in the microsomes after CYP7A1 overexpression was correlated with a decrease in HMGR activity. acyl-coenzyme A:cholesterol acyltransferase; cholesterol 7 hydroxylase; 3-hydroxy-3-methyglutaryl-CoA reductase; lowdensity lipoprotein receptor; liver; neutral cholesterol ester hydrolase CHOLESTEROL IS SYNTHESIZED in essentially all cells in the human body. It serves as a structural component of cell membranes and as a precursor for steroid hormones and bile acids. Cholesterol also plays a role as a regulatory molecule for several enzymes in the pathways of cholesterol metabolism in the liver (21). Bile acid synthesis from cholesterol is a major pathway for elimination of cholesterol from the body, occurring either via Address for reprint requests and other correspondence: W. M. Pandak, Veterans Affairs Medical Center, Division of Gastroenterology 111-N, 1201 Broad Rock Rd., Richmond, VA G878 the classic (also called neutral ) or alternative (also called acidic ) bile acid biosynthetic pathways (54). Cholesterol 7 -hydroxylase (CYP7A1) and sterol 27- hydroxylase (CYP27) are initial and rate-determining enzymes in the classic and alternative pathways, respectively. CYP7A1 and CYP27 are subject to regulation by bile acids and certain hormones (54). The contribution of the alternative pathway to bile acid synthesis is not precisely known and may be species dependent. In the rat, the estimated contribution may be as high as 50% (36, 45, 55), whereas in humans, the contribution may be 10% (10, 15, 51). Excess cholesterol accumulation in the body is associated with two major diseases of Western civilization, i.e., atherosclerosis and cholesterol gallstones. Cholesterol accumulation reflects excess cholesterol input into the body, reduced cholesterol elimination, or both. The effects of excessive cholesterol input via de novo cholesterol synthesis or via an increase in dietary cholesterol on serum cholesterol levels are well known (21). The importance of cholesterol degradative pathways, i.e., bile acid synthesis, on overall cholesterol homeostasis and serum cholesterol levels is less well understood. Different species differ widely in their response to dietary cholesterol load. The rat responds to cholesterol feeding with induction of CYP7A1 and bile acid synthesis via the classic pathway (33) and is relatively resistant to development of hypercholesterolemia. In contrast, increased dietary cholesterol fails to induce CYP7A1 in rabbits (57), Green monkeys (37), and hamsters (19, 32). In these species, excess cholesterol leads to the development of hypercholesterolemia and atherosclerosis. The ability of most humans to respond to a cholesterol load is probably more akin to hamsters, but some humans can process excessive dietary cholesterol by increasing their bile acid synthesis rate (8, 25). A seminal study by Spady et al. (42) demonstrated that overexpression of CYP7A1 in hamsters fed a diet high in cholesterol resulted in enhanced bile acid synthesis and prevented an increase in low-density lipoprotein (LDL) cholesterol levels. The results of this The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

2 G879 study firmly established the importance of bile acid biosynthetic pathways in the pathogenesis of hypercholesterolemia and atherosclerosis. Similar data were obtained in mice lacking the LDL receptor gene in which increased expression of hepatic CYP7A1 also lead to a 50% reduction in plasma LDL cholesterol concentrations (43). No similar data are available in humans. Cholesterol homeostasis in the liver is governed by several key regulatory enzymes and receptors, including 1) HMG-CoA reductase (HMGR), the rate-determining step in the cholesterol biosynthetic pathway; 2) CYP7A1 and CYP27, the initial enzymes in the classic and alternative pathways of bile acid biosynthesis, respectively; 3) acyl-coenzyme A:cholesterol acyltransferase (ACAT) and neutral cholesterol ester hydrolase (CEH), the two enzymes that regulate the sizes of free and esterified cholesterol compartments in the liver; and 4) cholesterol uptake via LDL and other lipoprotein receptors (21). The ability of each of these to respond to changes in a metabolically active sterol pool within the liver is the basis for how cholesterol homeostasis is maintained. In the present study, we determined the effects of overexpression of CYP7A1 on several parameters of bile acid and hepatic cholesterol metabolism in primary human hepatocytes (PHH) and in HepG2 cells, a wellcharacterized human hepatoblastoma cell line (34). The data show that overexpression of CYP7A1 in cultured cells derived from humans resulted in a marked increase of bile acid synthesis via the classic pathway. Alterations of hepatic cholesterol metabolism that have not been previously reported and that differ from the changes observed in other species, were also noted. MATERIALS AND METHODS Materials. All cell culture materials were obtained from GIBCO BRL (Grand Island, NY) unless otherwise specified. William s E medium was purchased from GIBCO BRL. The CsCl, agarose, and RNA ladder used to size CYP7A1 mrna were also purchased from GIBCO BRL. Taurocholate and cholesterol oxidase were purchased from Calbiochem (San Diego, CA). All other chemicals used were obtained from Sigma Chemical (St. Louis, MO) or Bio-Rad Laboratories (Hercules, CA) unless otherwise specified. -Cyclodextrin was purchased from Cylcodextrin Technologies Development (Gainesville, FL). All solvents were obtained from Fisher Scientific (Fair Lawn, NJ) unless otherwise indicated. All radionucleotides, aquasol solution, and enhanced chemiluminescence reagents were purchased from DuPont NEN (Boston, MA). Poly Attract mrna isolation system II was obtained from Promega (Madison, WI). The nick translation kit was obtained from Bethesda Research Laboratories (Grand Island, NY), and Tri-Reagent was purchased from Molecular Research Center (Cincinnati, OH). Nylon membranes were purchased from Micron Separation (Westborough, MA). Silica gel thin-layer chromatography plates (LK6 D) were from Whatman (Clifton, NJ). Generation of recombinant adenovirus encoding rat CYP7A1. The AdCMV-CYP7A1 and AdCMV- -gal adenovirus clones were constructed in an AD5dl309 adenovirus strain (24) essentially as previously described (53). Briefly, for AdCMV-CYP7A1, a 2.6-kb rat cdna encoding CYP7A1 was obtained from Dr. John Chiang (7) and was cloned into the EcoR I site of padcmv. Plasmid padcmv (7.7 kb) was constructed from a pgem4z (Promega) backbone and contained AD5dl309 sequence on either side of the CMV promoter/multiple cloning site region of pcdna3 (Invitrogen, Carlsbad, CA). The resulting padcmv/cyp7a1 recombinant plasmid was cotransfected with the right arm (XbaI digest) of Ad5dl309 into 293 cells (16). Adenovirus DNA isolated from resulting plaques were screened by Southern blot for the presence of the CYP7A1 insert. Propagation of AdCMV-CYP7A1. Large-scale production of recombinant virus was performed by infecting confluent monolayers of human embryonic kidney 293 cells (American Type Culture Collection, Rockville, MD) grown in 15-cm tissue culture dishes with stock adenovirus at a multiplicity of 1 plaque-forming units/cell. After 2 h of infection, the virus was removed and replaced with DMEM with 2% fetal bovine serum (FBS). The infected monolayers were harvested by scraping when 90% of the cells showed cytopathic changes and were then centrifuged at 2,700 g, 4 C, for 10 min. The cellular pellet was suspended in DMEM/2% FBS and subjected to five cycles of freeze-thaw lysis to release the virus. Cell debris was removed by centrifugation at 7,700 g, 4 C, for 5 min. To purify, the crude supernatant fluid was carefully layered over a twostep gradient containing 3 ml of CsCl (density 1.4 g/ml) in Tris dialysis (TD) buffer (0.14 M NaCl, 5 mm KCl, 19 mm Tris ph 7.4, 0.7 mm Na 2 HPO 4) layered over 3 ml of CsCl (d 1.25 g/ml) in TD buffer, and centrifuged at 155,000 g, 20 C, 1 h. The viral band was removed, layered over 8 ml of CsCl (d 1.33 g/ml) in TD buffer and centrifuged at 155,000 g, 20 C, 18 h. The pure viral opalescent band was removed and dialyzed against 10 mm Tris HCE ph 7.4, 1 mm MgCl 2, 10% glycerol, overnight at 4 C. The virus was aliquoted and stored at 70 C. The virus titer (pfu) was determined by plaque assay, and viral particles were determined by optical density using spectrophotometry ( 260). Infection of HepG2 cells with AdCMV-CYP7A1. HepG2 cells were grown in 162-cm 2 (25-ml) tissue culture flasks until they were 80 90% confluent as described previously (34) in MEM containing nonessential amino acids, 0.03 M NaCO 3, 10% FBS, 1 mm L-glutamine, 1 mm sodium pyruvate, and 1% Pen/Strep and were incubated at 37 C in 5% CO 2. Before infection, 15 ml of culture medium was removed. Control flasks (i.e., no addition of adenovirus) were treated in a similar manner. HepG2 cells were then infected with Ad- CMV-CYP7A1 with a multiplicity of infection of 1 using virus in a volume of ml. The virus was allowed to dwell for 3 h. After an additional 15 ml of fresh medium was added back to the flasks, they were allowed to incubate at 37 C, 5% CO 2 for 72 h. Toxicity to cells was assessed as previously described (34). Limited additional studies were done in the absence of FBS (i.e., serum-free media). Of note is that each value designated as N is the mean of duplicate-to-triplicate cultures for each condition measured during the same setting. Cells were divided in half and harvested in two distinct buffers as previously described (34, 35). One-half of the cells was used to isolate microsomes for measurement of HMGR and CYP7A1 specific activities. The remainder was used to isolate microsomes and cytosol for measurement of ACAT and CEH specific activities, respectively (35). In brief, the cells were scraped and suspended in the appropriate buffer. Cell disruption was performed by using a Sonifier cell disruptor 350 (Branson Sonic Power) for 1 min at an output of 3. Microsomes were isolated from cellular extracts by centrifugation at 660 g. The supernatant fluid was centrifuged at 105,000 g. The resulting microsomal pellets were suspended

3 G880 in the appropriate buffer by using a hand-driven homogenizer and were stored at 70 C until assayed for activity. The microsomal protein concentration was determined by the Bradford procedure (2), using BSA as a standard. Isolation and culture of primary human hepatocytes. Primary human hepatocytes were isolated according to the method of Strom et al. (46, 47). Before plating, cells were judged to be 90% viable by use of trypan blue exclusion. Cells were then plated onto 150-mm plastic Petri dishes previously coated with rat tail collagen and were incubated in 20 ml of William s E medium supplemented with insulin (0.25 units/ml) and penicillin (100 units/ml) in a 5% CO 2 atmosphere at 37 C. Cells were routinely harvested at 72 h of culture as previously described by our laboratory (45). Unless otherwise indicated, culture medium contained 0.1 M dexamethasone and 1.0 M L-thyroxine. Toxicity was assessed as previously described in the experiments using primary rat hepatocytes (45). Primary human hepatocytes were plated in P150-cm 2 plates to confluency ( cells). Twenty-four hours after plating, culture medium was removed, and 2.5 ml of fresh medium were added. Cells were then infected with a replication defective adenovirus containing CMV-CYP7A1 or control virus (replicative defective adenovirus without CMV- CYP7A1) with an MOI of 10. The virus was allowed to dwell for at least 2 h in minimal culture medium by shaking the plates gently every 15 min, after which the medium containing unbound virus was removed and 20 ml of fresh medium were added back to the plates and allowed to incubate at 37 C, 5% CO 2 for an additional 48 h. Cells were then harvested for the isolation of microsomes and RNA. Viral toxicity was determined by trypan blue exclusion and lactic dehydrogenase release as previously described in primary rat hepatocytes in culture (27, 45). Proper approval for these studies was obtained through the university institutional review board. More specifically, PHH cultures were obtained from the following livers: 58-yr-old Hispanic man involved in a motor vehicle accident (cells used for RNA and bile acid synthesis), 45-yr-old Caucasian man with an intracranial hemorrhage (cells used for bile acid synthesis and CYP7A1 specific activity), and 15-yr-old African-American woman with drug intoxication (cells used for RNA). Analysis of conjugated bile acids. Conjugated bile acids were extracted according to the method of Folch et al. (11). Conjugated bile acids in the culture medium aspirated from HepG2 cells were analyzed by gas liquid chromatography (GLC) by the methods of Setchell and Worthington (40). The conversion of [4-14 C]cholesterol into MeOH: H 2O-soluble materials was used as an indication of bile acid synthesis. Determinations of enzyme specific activities. The specific activity of CYP7A1 was determined in microsomes by using a HPLC assay described previously (22). HMGR specific activity was assayed as described by Whitehead et al. (56). Neutral cytosolic CEH specific activity was determined according to the methods of Ghosh et al. (13). ACAT specific activity was determined by the method of Burrier et al. (4) and Pape et al. (35) with the following modifications. ACAT reactions were carried out using 10 g microsomal protein with buffer conditions identical to those of Burrier et al. Cholesterol was added as uniform vesicles in a cholesterol-phosphatidylcholine (0.5 molar) ratio using an extruder. After a 15-min incubation, 1-[ 14 C]oleoyl-CoA was added to a final concentration of (0.072 nm) (20 nci). Fifteen minutes after addition of 1-[ 14 C]oleoyl-CoA, reactions were terminated by direct application to silica gel TLC plates (Fisher Gel G cm). Dried plates were developed in hexane-diethyl ether-acetic acid 85:15:1 by volume. 4-[ 14 C]cholesteryl oleate formation was detected using a PhosphorImager (Molecular Dynamics). Microsomal free and total cholesterol were determined by HPLC analysis. Free cholesterol was converted to cholest- 4-ene-3-one by using cholesterol oxidase as previously described (22). Total cholesterol was measured as free cholesterol after saponification. Experiments involving the exogenous addition of 7 -hydroxycholesterol. In an attempt to further define the effects of 7 -hydroxycholesterol on the specific activity of HMGR, 7 hydroxycholesterol was added to cell culture medium in molecusol (3 mg 7 -hydroxycholesterol to 1 ml -cyclodextrin) to obtain final concentrations of 1, 25, and 50 M. Microsomes were harvested 24 h after the addition of the 7 -hydroxycholesterol, and endogenous microsomal 7 -hydroxycholesterol levels and HMG-R specific activities were then determined. As reference, 7 -hydroxycholesterol serum levels have been reported to be in the range of M in healthy humans (3). In other studies, the addition of 6 12 M concentrations of 7 -hydroxycholesterol have not been shown to downregulate HMG-R activity in HepG2 cells (23). 7 -Hydroxycholesterol-induced cell toxicity was assessed as previously described (34). Quantitation of mrna levels. Methods for the isolation of RNA and the determination of mrna levels by Northern blotting have been previously described (33). Rat cyclophilin cdna was used as the internal loading standard. The cdna probes used were as follows: CYP7A1 (28); HMGR (26); LDL receptor (LDLR; pldlr3 was obtained from American Type Culture Collection); ACAT (6); CEH (13); CYP27 (48); and cyclophilin (9). RESULTS Studies in HepG2 cells. Shown in Fig. 1 is a representative comparison of CYP7A1 mrna levels in HepG2 cells infected with AdCMV-CYP7A1 vs. control cells. Infection with this recombinant adenovirus (AdCMV-CYP7A1) led to a marked increase in CYP7A1 mrna levels (11, %; P 0.01) compared with control cells. To more easily detect basal CYP7A1 mrna levels in control cells, it was necessary to isolate and probe poly A mrna (34). However, after infection with AdCMV-CYP7A1, CYP7A1 mrna levels were easily detectable by use of total RNA. Analysis of RNA isolated from AdCMV- CYP7A1 infected cells showed the 3.6-kb CYP7A1 band was not detectable, whereas when poly A was probed from control cells, the usual three bands at 3.6, 2.5, and 1.2 kb were detected (34). To ensure that the adenovirus itself was having no effect on CYP7A1 mrna levels, HepG2 cells were also infected with a recombinant adenovirus encoding -galactosidase (AdCMV- Gal). No increase in CYP7A1 mrna was observed in those cells (see Fig. 8). These data demonstrate that highly effective overexpression of CYP7A1 could be accomplished in HepG2 cells. Fig. 2 shows a marked increase in CYP7A1 specific activity in HepG2 cells after infection with AdCMV- CYP7A1 compared with control, noninfected cells. Activity in HepG2 cells increased from undetectable levels in uninfected cells to nmol h 1 mg 1 for virus-infected cells. Preformed 7 -hydroxycholesterol was also undetectable in uninfected cells, whereas sig-

4 G881 nificant levels were detected in the infected cells (Table 1). To ensure that the recombinant adenovirus itself was not having a stimulatory effect on CYP7A1 specific activity, specific activity was determined in cells infected with AdCMV- Gal, a representative control recombinant adenovirus. Shown in Fig. 3 are representative HPLC tracings demonstrating levels of 7 -hydroxycholesterol in cells infected with AdCMV- CYP7A1 (Fig. 3A) vs. AdCMV- Gal (Fig. 3B). A large amount of 7 -hydroxycholesterol was observed after infection with AdCMV-CYP7A1. No 7 -hydroxycholesterol was detected after AdCMV- Gal (i.e., no specific activity). Control uninfected HepG2 cells also had no detectable CYP7A1 activity (see Fig. 2). These representative tracings demonstrate that control recombinant adenovirus had no stimulatory effect on endogenous CYP7A1 expression. Not shown are tracings demonstrating no stimulatory effect of control virus (AdCMV- Gal) on CYP7A1 activity compared with uninfected cells. An approximately twofold increase in total bile acids synthesized by HepG2 cells was observed in the culture Fig. 1. Representative Northern blot of cholesterol 7 -hydroxylase (CYP7A1) mrna levels in HepG2 cells after infection with AdCMV- CYP7A1. HepG2 cells were grown until 80 90% confluent under optimal culture conditions. Cells infected with AdCMV-CYP7A1 as described under MATERIALS AND METHODS are compared with control (CTRL) cells (no viral infection). Cells were harvested 72 h after infection, and mrna levels for CYP7A1 and cyclophilin (loading standard) were determined. Fig. 2. Effect of infection with AdCMV-CYP7A1 on CYP7A1 specific activity in HepG2 cells. HepG2 cells were grown until 80 90% confluent under optimal culture conditions and then were infected with AdCMV-CYP7A1 as described under MATERIALS AND METHODS. Cells were harvested 72 h after infection, and specific activity for CYP7A1 was determined. Data are expressed as means SE. CYP7A1 specific activity was not detectable in control cells. Preformed 7 -hydroxycholesterol (endogenous) was also not found in control cells, whereas endogenous microsomal 7 -hydroxycholesterol was easily detected in infected cells before incubation (Table 1). medium after CYP7A1 overexpression. [ 14 C]cholesterol conversion to CH 3 OH/H 2 O extractable counts increased % (P 0.05) in HepG2 cells infected with AdCMV-CYP7A1 compared with uninfected controls (Fig. 4A). The percent increase in bile acid synthesis was in agreement with the increase in primary bile acids (cholic and chenodeoxycholic acid) measured in the medium as determined by GLC (Fig. 4B). Using GLC analysis, we found an increase of % (P 0.05) over paired controls in bile acid concentration (i.e., synthesis) in the medium in the cells infected with AdCMV-CYP7A1 (Fig. 4B). The composition of the primary bile acids in HepG2 cell culture medium was largely chenodeoxycholic acid in control cells, with approximately equal proportions of cholic and chenodeoxycholic acid after CYP7A1 overexpression (Fig. 5). In limited studies, HepG2 cells were grown in serum-free medium compared with the serum-containing conditions employed above. As in HepG2 cells grown in serum-containing medium, cells grown in serum-free medium increased bile acid synthesis after CYP7A1 overexpression, but to a lesser degree (130%; P 0.05). This lower rate of bile acid synthesis after CYP7A1 overexpression in a growing cell line in the absence of cholesterol-containing serum is most likely a function of a decreased cholesterol pool availability for bile acid synthesis (i.e., absence of medium cholesterol uptake). The impact of overexpression of CYP7A1 on cellular cholesterol homeostasis is reflected in changes in HMGR, ACAT, and CEH specific activities. A decrease in HMGR specific activity (234 7%; P 0.01) was observed in AdCMV-CYP7A1-infected cells compared with uninfected control cells (Fig. 6). Overexpression of CYP7A1 in HepG2 cells also perturbed hepatocellular cholesterol esterification and storage, as can be seen by the changes in cytosolic CEH and microsomal ACAT specific activi-

5 G882 Fig. 3. Effect in HepG2 cells of infection with AdCMV-CYP7A1 vs. infection with AdCMV- Gal on CYP7A1 specific activity. Representative HPLC tracing (see MATERIALS AND METHODS) shows a large amount of detectable 7 -hydroxycholesterol after infection with recombinant adenovirus encoding CMV- CYP7A1 (A). No 7 -hydroxycholesterol was detectable after control virus infection (B). Control uninfected HepG2 cells also showed no CYP7A1 specific activity (see Fig. 2). ties. Specific activity of neutral CEH increased 38 5% (P 0.01) after infection with AdCYP7A1 compared with uninfected control cells (Fig. 6). A 56 7% (P 0.01) decrease in ACAT specific activity occurred in HepG2 cells infected with AdCMV-CYP7A1. A significant increase in LDLR mrna levels (41 16%; P 0.01) was seen in HepG2 cells after infection with AdCMV-CYP7A1 compared with uninfected controls (Fig. 7). LDLR mrna levels were not affected in cells infected with AdCMV- Gal adenovirus lacking CYP7A1. HMGR and ACAT mrna levels were decreased 28 (n 2; range %) and 25% (n 2; range %), respectively. CEH mrna levels were increased 19% (n 2; range %). A representative Northern blot comparing control HepG2 cell mrna levels with recombinant control virus (AdCMV- Gal) and AdCMV-CYP7A1 shown in Fig. 8 supports these cumulative results. With the exception of LDLR, for which receptor activity was not measured, changes in HMGR, ACAT, and CEH mrna levels correlated with changes observed in their respective specific activities. In limited studies, HepG2 cells were also grown in serum-free medium as opposed to the serum-containing conditions employed above. As in HepG2 cells grown in serum-containing medium, LDLR mrna increased (154 27%; P 0.05) in cells grown in serum-free medium after CYP7A1 overexpression. Overexpression of CYP7A1 also led to changes in microsomal free and total cholesterol (Fig. 9). A 27

6 G883 Fig. 6. Response in enzymes involved in heptatic cholesterol homeostatis after overexpession of CYP7A1 in HepG2 cell cultures. HMGR, HMG-CoA reductase; ACAT, acyl coenzyme A/cholesterol acyltransferase; CEH, cholesterol ester hydrolase. Data are expressed as percent of paired control (means SE). Basal specific activities: HMGR nmol h 1 mg 1 microsomal protein; CEH pmol h 1 mg 1 microsomal protein; ACAT pmol min 1 mg 1 microsomal protein. Fig. 4. Effect of overexpression of CYP7A1 on bile acid synthesis in HepG2 cell cultures. HepG2 cells were grown until 80 90% confluent under optimal culture conditions. Cells were infected with AdCMV- CYP7A1 as described under MATERIALS AND METHODS. A: bile acid synthesis as measured by conversion of [ 14 C]cholesterol to 14 C- labeled bile acids. B: bile acid synthesis as determined by GLC analysis of culture medium. 1% (P 0.01) decrease in microsomal free cholesterol was seen after infection with AdCMV-CYP7A1 in HepG2 cells (control cell microsomal free cholesterol levels, 200 nmol/mg microsomal protein). A corresponding decrease (2 45%; n 2) in microsomal total cholesterol was also noted with increased expression of CYP7A1. After overexpression of CYP7A1, increased endogenous microsomal levels of 7 -hydroxycholesterol were observed. As demonstrated (Fig. 6; Table 1), the increase in these levels was associated with lower HMGR specific activities. Previously, it has been suggested that 7 -hydroxycholesterol may be one of several oxysterols capable of downregulating HMGR. To further substantiate this hypothesis, additional studies were performed in which 7 -hydroxycholesterol was added directly to culture medium (Table 1). Concentrations attempting to approximate and exceed previously documented human serum values were chosen. A range of M has previously been demonstrated in the serum of healthy humans (3). A concentration of 6 12 Fig. 5. Effect of overexpression of CYP7A1 on relative concentrations of cholic and chenodeoxycholic acids (Cholic and Cheno, respectively) in HepG2 cell cultures. The relative concentrations of these 2 primary bile acids secreted into HepG2 cell culture medium as determined by GLC analysis. Data are expressed as means SE. Fig. 7. Effect on steady-state mrna levels in HepG2 cells after overexpression of CYP7A1 mediated through infection with AdCMV- CYP7A1. LDL-R, LDL receptor; ACAT-1, ACAT. Data expressed as a percent of paired controls (means SE). By using laser densitometry, mrna levels were calculated as a ratio of mrna of interest to that of cyclophilin, employed as an internal loading standard.

7 G884 Table 1. Effect of addition of 7 -hydroxycholesterol on HMGR specific activity Condition Endogenous 7 - Hydroxycholesterol in Microsomes, M HMGR Specific Activity, % of Control Infected Additions 1 M No endogenous M M Infected: HepG2 cells infected with AdCMV-CYP7A1. Additions: 7 -hydroxycholesterol was added to HepG2 cell culture medium to achieve 1, 25, and 50 M medium concentrations. Microsomes were harvested 24 h after addition of the 7 -hydroxycholesterol, and endogenous 7 -hydroxycholesterol microsomal concentrations were determined (means SE). HMGR, HMG-CoA reductase specific activity expressed as a percent of paired controls. Fig. 8. Representative Northern blot of mrna levels in HepG2 cells after infection with AdCMV-CYP7A1 compared with control virus. HepG2 cells were grown until 80 90% confluent under optimal culture conditions. Cells were infected with AdCMV-CYP7A1 or AdCMV- Gal as described under MATERIALS AND METHODS. Cells were harvested 48 h after infection, and mrna levels for LDL-R, HMGR, CYP7A1, ACAT-1, CEH, and cyclophilin (loading standard) were determined. Note: Despite the attempt to generate a blot with the same RNA loading in all columns, mrna levels in the Gal column were less as determined by use of cyclophilin as an internal loading standard. M of7 -hydroxycholesterol had not previously been shown to downregulate HMGR activity in HepG2 cells (23). In our studies, the addition of 7 -hydroxycholesterol to medium to obtain a 1 M concentration had no effect on HMGR activity. However, 25 M and 50 M concentrations of 7 -hydroxycholesterol in the medium led to a concentration-dependent decrease in HMGR activity (Table 1). Microsomal 7 -hydroxycholesterol concentrations after CYP7A1 overexpression were between that of the 25 M and 50 M 7 -hydroxycholesterol medium concentrations, with the decrease in HMGR most closely approximating the effects of 25 M (Table 1). Studies in primary human hepatocytes. Figure 10 shows a representative Northern blot of CYP7A1 mrna levels in PHH control cells (addition of control adenovirus) compared with mrna levels in cells infected with AdCMV-CYP7A1. There is evidence for abundance of mrna levels in infected (PHH) compared with control cells in which the mrna levels are practically undetectable. Cyclophilin controls showed no changes in either control or infected cells. In Fig. 11, we show quantitative data on CYP7A1 mrna levels and the specific activities in PHH. In contrast to control HepG2 cells in which CYP7A1 specific activity was undetectable (Fig. 2), PHH controls had detectable CYP7A1 activity (0.13 nmol h 1 mg 1 protein). After overexpression of CYP7A1, steady-state mrna levels increased 18-fold, whereas specific activities increased 10-fold. These data are consistent with significant augmentation of CYP7A1 after infection with AdCMV-CYP7A1 virus. Fig. 9. Effects of overexpression of CYP7A1 on microsomal free and total cholesterol in HepG2 cells. Microsomal free and total cholesterol were determined by HPLC analysis as described under MATE- RIALS AND METHODS. Data are expressed as percent of paired controls (means SE). Fig. 10. Representative Northern blot of CYP7A1 mrna levels in primary human hepatocytes after infection with AdCMV-CYP7A1. Hepatocytes were incubated for 24 h in culture medium containing optimal concentrations of thyroxine (T 4; 1.0 M) and dexamethasone (0.1 M). At 24 h, cells were infected with control virus or AdCMV- CYP7b1 as described under MATERIALS AND METHODS. Cells were harvested 48 h after infection, and mrna levels for CYP7A1 and cyclophilin (loading standard) were determined.

8 G885 Fig. 11. Effect on CYP7A1 mrna levels and specific activity in primary human hepatocytes after infection with AdCMV-CYP7A1. Hepatocytes were incubated for 24 h in culture medium containing optimal concentrations of T 4 (1.0 M) and dexamethasone (0.1 M). At 24 h cells were infected with control virus or AdCMV-CYP7A1 as described under MATERIALS AND METHODS. Cells were harvested 48 h after infection, and mrna levels and specific activity for CYP7A1 were determined. Data are expressed as percent of paired controls. Fig. 12. Effect of overexpression of CYP7A1 using AdCMV-CYP7A1 on bile acid synthesis in primary human hepatocyte cultures. Hepatocytes were incubated for 24 h in culture medium containing optimal concentrations of T 4 (1.0 M) and dexamethasone (0.1 M). At 24 h cells were infected with control virus or AdCMV-CYP7A1 as described under MATERIALS AND METHODS. Bile acid synthesis was measured 22 and 48 h after infection as conversion of [ 14 C]cholesterol to 14 C-labeled bile acids. Data are expressed as percent of paired control (mean of 2 experiments). Fig. 13. Representative Northern blot of LDLR mrna levels in primary human hepatocytes after infection with AdCMV-CYP7A1. Hepatocytes were incubated for 24 h in culture medium containing optimal concentrations of T 4 (1.0 M) and dexamethasone (0.1 M). At 24 h, cells were infected with control virus or AdCMV-CYP7A1 as described under MATERIALS AND METHODS. Cells were harvested 48 h after infection, and mrna levels for LDLR and cyclophilin (loading standard) were determined. Bile acid synthesis in primary human hepatocytes (Fig. 12) after infection with AdCMV-CYP7A1 increased 73% (n 2; range %) at 22 h and 393% (n 2; range %) at 48 h. These data show that the marked increase in CYP7A1 mrna levels and specific activities after CYP7A1 overexpression is translated into a significant increase of bile acid synthesis and fully supports the stimulation of the classic pathway of bile acid synthesis. Simultaneously, a significant increase occurred in LDLR mrna levels (Fig. 13) in PHH after infection with AdCMV-CYP7A1 compared with paired controls. Cyclophilin levels did not change, suggesting that unequal blot RNA loading could not account for increases in LDLR mrna levels. Similar to HepG2 cells, HMG- CoA reductase activity decreased 67% (mean of 3 cultures; n 1) in PHH infected with AdCMV-CYP7A1. DISCUSSION In the present paper, we provide evidence that overexpression of CYP7A1 in HepG2 and in PHH results in an increase in bile acid synthesis via the classic pathway coupled with suppression of cholesterol synthesis. These findings are particularly noteworthy because they differ in several respects from responses observed in other species. Most of the information on the mechanism by which cholesterol homeostasis is maintained has been derived from experimental animals, which showed a great deal of species variation (19, 33, 37, 57). The rat is capable of inducing CYP7A1 and bile acid synthesis in response to a diet high in cholesterol (33). The resulting increase in cholesterol degradation appears to protect this species from hypercholesterolemia and atherosclerosis. In humans, the response to dietary cholesterol appears to be heterogeneous. A trend toward increasing LDL cholesterol by increasing dietary cholesterol is well established, but in a significant number of humans, circulating LDL cholesterol does not change in response to changes in dietary cholesterol (8, 25). In an interesting paper by Kern (25), a single patient consuming 25 eggs per day for over 20 years had a marked increase in the rate of bile acid synthesis coupled with near normal serum cholesterol levels. This combination of findings suggests that, in this particular individual, CYP7A1 was inducible by a diet high in cholesterol in a manner similar to that of rat. Couture et al. (8) studied the effect of A to C substitution at position 204 of the promoter of CYP7A1 gene and its association with variations in

9 G886 plasma LDL cholesterol concentrations. At least in humans, the C variant was associated with higher plasma LDL cholesterol concentrations and an increase in the ratio of total cholesterol to high-density lipoprotein. Although overexpression of CYP7A1 in hamsters prevents hypercholesterolemia in animals fed a Western diet (42), it is not certain whether such observations can be extrapolated from other species to humans. We have previously characterized HepG2 cells and found them a good model to study the regulation of human CYP7A1 and other enzymes of the bile acid biosynthetic pathways. HepG2 cells were shown to synthesize primary bile acids and express enzymes and receptors involved in the maintenance of cholesterol homeostasis (34). PHH have been used to study numerous aspects of cholesterol and bile acid transport and metabolism and have been shown to be a valid model to study the physiological and metabolic events in humans. To further verify the results in HepG2 cells, we also repeated some key experiments in PHH for comparison purposes. The data in the two systems are consistent and differ somewhat from those obtained in other species. In AdCMV-CYP7A1-infected HepG2 cells, CYP7A1 mrna levels and specific activities increased severalfold. Bile acid synthesis increased approximately twofold with the associated increase in CYP7A1 expression. No such increase occurred in control cells or cells infected with a control adenovirus. After infection with the AdCMV-CYP7A1, the concentration of cholic acid in the medium increased from 10% to nearly 50% of total bile acid synthesis. This increase in the proportion of cholic acid suggests that in humans the classic pathway is mainly responsible for synthesis of cholic acid. Previous in vitro and in vivo studies in the rat suggest that up to 50% of total bile acid synthesis may occur via the alternative pathway and that the predominant bile acid is chenodeoxycholic acid (36, 45, 55). How representative this finding is to other species is uncertain. In contrast to findings in the rat, Duane and Javitt (10) have recently reported that in healthy humans 10% of total bile acid synthesis appeared to be synthesized via the alternative pathway; their findings are consistent with previous reports (15, 51). In our recent experiments, overexpression of CYP27 in Chinese hamster ovary cells and HepG2 cells led to a modest increase in 27-hydroxycholesterol and bile acid synthesis, respectively (18). These data suggested that the alternative pathway may be induced in human-derived liver cell lines. In the CYP7A1 knockout mouse, in which the classic pathway of bile acid synthesis is absent, the alternative pathway can become activated and produce sufficient bile acid synthesis to permit survival without exogenous bile acid supplements (39). A marked increase of CYP7A1 specific activity and bile acid synthesis in both types of cells was coupled with compensatory changes in hepatic cholesterol metabolism, including increased LDLR mrna levels. Surprisingly, the overexpression of CYP7A1 in HepG2 cells was associated with a decrease in HMGR activity. This finding is different from in vivo observations by Spady et al. (42), and in vitro observations by Spitsen et al. (44) and Sudjana-Sugiaman et al. (50). In general, HMGR and CYP7A1 change in tandem, resulting in coordinate regulation of cholesterol and bile acid synthesis. The previous in vivo and in vitro studies (42, 44, 50) were carried out by infection of the CYP7A1 gene into hamsters or nonhuman, nonhepatic cell lines. HepG2 cells have been shown to reflect in vivo events in humans, and because a decrease in HMGR activity after overexpression of CYP7A1 was a reproducible finding, one can conclude that the behavior of HMGR in HepG2 cells (i.e., extrapolated to humans) may be different from that in the rat and other species. Our limited findings in PHH in this study support of this finding. Although the decrease in HMGR in the face of markedly upregulated CYP7A1 was not expected, a plausible explanation is available. The ability of oxysterols to repress HMGR is well documented in several studies (3, 30, 41). In our model, overexpression of CYP7A1 results in formation of a large amount of 7 -hydroxycholesterol, which in turn may repress HMGR. This was confirmed by adding 7 -hydroxycholesterol to the culture medium (Table 1). These experiments, therefore, strongly suggest that 7 -hydroxycholesterol or a derivative may be responsible for the observed feedback repression of HMGR under these experimental circumstances. Increasing CYP7A1 activity directly in hepatocytes might not only stimulate increased cholesterol catabolism, but also simultaneously repress HMGR. This approach differs from the secondary upregulation of CYP7A1 and HMGR associated with feeding the intestinal bile acid-binding resin, cholestyramine. With cholestyramine feeding, the increased intestinal loss of bile acids stimulates bile acid synthesis via a reduction in negative bile acid biofeedback with a subsequent increase in HMGR activity. Although the serum concentration of 7 -hydroxycholesterol has been shown to increase with cholestyramine feeding (1), the microsomal levels achieved and the relationship to this study are not clearly defined. One explanation for the differing findings in this study compared with those of previous studies is the level of microsomal 7 -hydroxycholesterol. The effect of 7 -hydroxycholesterol in the regulation of the LDLR is even less clear, with previous effects clearly dependent on oxysterol cellular concentration (29, 31). These results suggest a differential effect of 7 -hydroxycholesterol on HMGR and LDLR regulation. The fact that HMGR and LDLR usually also change in tandem suggests that intracellular concentrations of 7 -hydroxycholesterol can lead to a differential regulation in HMGR and LDLR. As transcriptional control of HMGR and LDLR are believed to be controlled by the same factor (22), one could postulate that 7 -hydroxycholesterol could be mediating its regulation of HMGR through acceleration of mrna degradation. Our findings suggest that sufficiently increased CYP7A1 activity with a subsequent increase in 7 hydroxycholesterol microsomal concentrations will ex-

10 G887 hibit a cholesterol homeostatic response favorable to lowering serum cholesterol. The physiological significance of the large increase in CYP7A1 activity achieved in vitro compared with what can be achieved in vivo is uncertain. However, if this observation can be confirmed in in vivo experiments, it would suggest that overexpression of CYP7A1 may be a highly effective strategy for lowering serum cholesterol in humans. The ability of ACAT to esterify cholesterol is important in maintaining intracellular concentrations of free cholesterol as well as providing cholesterol storage in the form of cholesterol esters. In the liver, ACATderived cholesterol esters are secreted as a component of very-low-density lipoproteins. ACAT activity is upregulated by LDL and free cholesterol (49, 52). In general, as free cholesterol substrate increases, so does ACAT activity, to maintain free cholesterol in the narrow range. In the presence of decreased cholesterol availability, the activity of ACAT has been shown to decrease (49, 52). In the present study, increased cholesterol catabolism coupled with decreased microsomal free cholesterol levels was associated with a decrease in ACAT specific activity and ACAT-1 mrna levels (ACAT-2 mrna levels were not determined). As with cholesterol, in vitro addition of 25-hydroxycholesterol has previously been shown to stimulate ACAT activity (49). More recently, in isolated membranes expressing ACAT activity, Cases et al. (5) found 27-hydroxycholesterol and 24(S),25-epoxycholesterol to be stimulators of ACAT activity, with 7-hydroxycholesterol a potent suppressor. On the basis of these observations, it has been suggested that oxysterols may be important regulators of ACAT. In this study, a repression in ACAT activity was associated with an increase in microsomal 7 -hydroxycholesterol, an increase also associated with downregulation of HMGR. However, it is unclear in this study whether the change in ACAT activity is in response to a decrease in the regulatory cholesterol pool size in the liver or in response to the increase in 7 -hydroxycholesterol. Transfection studies designed to assess the role of cholesterol availability in the regulation of ACAT specific activity, using the overexpression of ACAT in ACAT-deficient cell lines, have suggested that ACAT regulation is governed at both the transcriptional and posttranscriptional level (52). The lesser decrease in steady-state mrna levels compared with that observed in specific activity in response to overexpression of CYP7A1 supports this observation. Recent findings have suggested that ACAT-2, primarily involved in lipoprotein assembly and secretion, is the predominant liver ACAT (38). However, HepG2 cells contain both ACAT-1 and ACAT-2. Therefore, the activities of two different gene products are reflected in the measured ACAT activity. Whether the ACAT-1 and -2 functions are separate or can overlap in this cell line is uncertain. Only ACAT-1 mrna levels were determined in this study. However, a decrease in mrna levels was correlated with decreased ACAT activity. CEH is the key enzyme required for releasing the pool of metabolically active free cholesterol from intracellular stores of cholesterol esters (14). Thus neutral CEH is of great importance as an enzyme that replenishes the pool of free cholesterol for bile acid synthesis and biliary cholesterol secretion. This overexpression study of CYP7A1 clearly shows that enhanced bile acid synthesis is coupled with an increase in CEH specific activity, a means by which the size of the free cholesterol substrate pool is increased. The fact that an increase in CEH specific activity was coupled with an increase in mrna levels also supports a transcriptional level of regulation of CEH. These data are consistent with primary observations showing that a cholesterol-enriched diet decreases CEH activity in rats and consistent with an increase in activity in response to cholestyramine feeding (12, 17). In summary, increased expression of CYP7A1 in primary human hepatocytes and in HepG2 cells resulted in increased bile acid synthesis via the classic pathway. The cholesterol homeostatic response in HepG2 cells included downregulation of HMGR, probably due to accumulation of 7 -hydroxycholesterol, and upregulation of LDLR. This response suggests that overexpression of CYP7A1 in humans may be a useful approach for lowering serum cholesterol. Other effects of overexpression of CYP7A1 on ACAT and CEH are consistent with attempts of the liver to maintain cholesterol homeostasis in the face of a marked increase in cholesterol catabolism. The fact that the results in HepG2 cells and PHH agree with each other adds confidence that the results of this study in human cell lines may be extrapolated to an in vivo situation. We thank Tina Lucas, Li Zhao, Pat Bohdan, Melissa Thompson, and Emily Gurley for technical assistance. We also thank Drs. J. Y. L. Chiang, S. Ghosh, W. Grogan, B. Kren, and T. Y. Chiang for providing cdnas for CYP7A1, CEH, HMGR, and ACAT, respectively. This work was supported by grants from the Veterans Administration and the National Institute of Diabetes and Digestive and Kidney Diseases (P01 DK-38030). REFERENCES 1. Bascoul J, Goze C, Domergue N, and Crastes de Paulet A. Serum level of 7 -hydroxycholesterol in hypercholesterolemic patients treated with cholestyramine. Biochim Biophys Acta 1044: , Bradford MM. A rapid and sensitive method for quantitation of microgram quantities of protein utilizing the principle of proteindye binding. Anal Biochem 72: Brown AJ and Jessup W. Oxysterols and atherosclerosis. Atherosclerosis 142: 1 28, Burrier RE, Deren S, McGregor DG, Hoos LM, Smith AA, and Davis HR Jr. Demonstration of a direct effect on hepatic acyl CoA:cholesterol acyltransferase (ACAT) activity by an orally administered enzyme inhibitor in the hamster. Biochem Pharmacol 47: , Cases S, Novak S, Zheng YW, Myers HM, Lear SR, Sande E, Welch CB, Lusis AJ, Spencer TA, Krause BR, Erickson SK, and Farese RV. ACAT-2, a second mammalian acyl-coa:cholesterol acyltransferase: its cloning, expression, and characterization. J Biol Chem 273: , Chang CY, Huh YH, Cadigan KM, and Chang TY. Molecular cloning and function expression of human acyl-coenzyme A:cholesterol acyltransferase cdna in mutant Chinese hamster ovary cells. J Biol Chem 268: , 1993.

11 G Chun YL, Wang DP, and Chiang JYL. Regulation of cholesterol 7 -hydroxylase in the liver. J Biol Chem 265: , Couture P, Otvos JD, Couples LA, Wilson PW, Schafer EJ, and Ordovas JM. Association of the A-204 C polymorphism in the cholesterol 7 -hydroxylase gene with variations in plasma low density lipoprotein levels in the Framingham Offspring Study. J Lipid Res 40: , Danielson PE, Forss-Petter S, Brow A, Calavetta L, Douglas RJ, Milner RJ, and Sutcliffe JG. p1b15: a cdna clone of the rat mrna encoding cyclophilin. DNA (NY) 7: , Duane WC and Javitt NB. 27-Hydroxycholesterol: production rates in normal human subjects. J Lipid Res 40: , Folch J, Lees M, and Sloan-Stanley GH. A simple method for the isolation and purification of total lipids from animal tissues. J Biol Chem 226: , Ghosh S, Kounas ML, and Grogan WM. Separation and differential of rat liver cytosolic cholesteryl ester hydrolase triglyceride lipase and retinyl palmitate hydrolase by cholestyramine and protein kinases. Lipids 25: , Ghosh S, Mallonee DH, Hylemon PB, and Grogan WM. Molecular cloning and expression of rat hepatic neutral cholesteryl ester hydrolase. Biochim Biophys Acta 1259: , Ghosh S, Natarajan R, Pandak WM, Hylemon P, and Grogan WM. Regulation of hepatic neutral cholesterol ester hydrolase by hormones and changes in cholesterol flux. Am J Physiol Gastrointest Liver Physiol 274: G662 G668, Goldman N, Vlahcevic ZR, Schwartz CC, Gustafsson J, and Swell L. Quantitative evaluations of bile acid synthesis from [7-3 H]7 -hydroxycholesterol and [G- 3 H]26-hydroxycholesterol. Hepatology 2: 59 66, Graham FL, Smiley J, Russell WC, and Nairn R. Characteristics of a human cell line transformed by human adenovirus type 5. J Gen Virol 36: 59 74, Grogan WM, Bailey ML, Heuman DM, and Vlahcevic ZR. Effects of perturbation in hepatic free and esterified cholesterol pools on bile acid synthesis, cholesterol 7 -hydroxylase, HMG- CoA-R, Acyl-CoA:cholesterol acyl transferase and cytosolic cholesteryl ester hydrolase. J Lipid Res 26: , Hall EA, Vlahcevic ZR, Hylemon PB, Malloonee D, and Pandak WM. Sterol 27-hydroxylase is a tissue specific regulator of cholesterol and bile acid synthesis. Gastroenterology 118: A1170, Horton JD, Cuthbert JA, and Spady DK. Regulation of hepatic 7 -hydroxylase expression and response to dietary cholesterol in the rat and hamster. J Biol Chem 270: , Horton JD and Shimomura I. Sterol regulatory elementbinding proteins: activators of cholesterol and fatty acid biosynthesis. Curr Opin Lipidol 10: , Hylemon PB, Pandak WM, and Vlahcevic ZR. Regulation of hepatic cholesterol homeostasis. In: The Liver Biology and Pathobiology (4th ed.), edited by Arias IM, Boyer JL, Chisari FV, Fausto N, Schachter DA, and Shafritz DA. New York: Lippincott Williams & Wilkins, 2001, p Hylemon PB, Studer EJ, Pandak WM, Heuman DM, Vlahcevic ZR, and Chiang JYL. Simultaneous measurement of cholesterol 7 -hydroxylase activity by reverse-phase high-performance liquid chromatography using both endogenous and exogenous [4-14 C]-cholesterol as substrate. Anal Biochem 182: , Javitt NB and Budai K. Cholesterol and bile acid synthesis in HepG2 cells: Metabolic effects of 26- and 7 -hydroxycholesterol. Biochem J 262: , Jones N and Shenk T. Isolation of adenovirus type 5 host range deletion mutants defective for transformation of rat embryo cells. Cell 17: , Kern F. Normal plasma cholesterol in an 88 year old man who eats 25 eggs a day. N Engl J Med 324: , Kren BT, Rodrigues CMP, Setchell KDR, and Steer CJ. Posttranscriptional regulation of mrna levels in the liver associated with deoxycholic acid feeding. Am J Physiol Gastrointest Liver Physiol 269: G961 G973, Kubaska WM, Gurley EC, Hylemon PB, Guzelian PS, and Vlahcevic ZR. Absence of negative feedback control of bile acid biosynthesis in cultured rat hepatocytes. J Biol Chem 260: , Li YC, Wang DP, and Chiang JYL. Regulation of cholesterol 7 -hydroxylase in the liver. Cloning sequencing and regulation of cholesterol 7 -hydroxylase mrna. J Biol Chem 265: , Lorenzo JL, Allorio M, Bernini F, Corsini A, and Fumagalli R. Regulation of low density lipoprotein metabolism by 26-hydroycholesterol in human fibroblasts. FEBS Lett 218: 77 80, Lund E and Bjorkhem I. Role of oxysterols in the regulation of cholesterol homeostasis: a critical evaluation. Accounts Chem Res 28: , Miao E, Wilson SR, and Javitt NB. Cholesterol metabolism. Effect of 26-thiacholesterol and 26-aminocholesterol analogs of 26-hydroxycholesterol on cholesterol synthesis and low-densitylipoprotein-receptor binding. Biochem J 255: , Pandak WM, Doerner K, Heuman DM, Hylemon PB, Chiang JYL, and Vlahcevic ZR. Expression of cholesterol 7 -hydroxylase in response to cholesterol and bile acid feeding in the hamster and rat (Abstract). Gastroenterology 108: A1141, Pandak WM, Li YC, Chiang JYL, Studer EJ, Gurley EC, Heuman DM, Vlahcevic ZR, and Hylemon PB. Regulation of cholesterol 7 -hydroxylase mrna and transcriptional activity by taurocholate and cholesterol in the clinically biliary diverted rat. J Biol Chem 66: , Pandak WM, Stravitz RT, Lucas V, Heuman DM, and Chiang JYL. HepG2 cells: a model for studies of regulation of human cholesterol 7 -hydroxylase at the molecular level. Am J Physiol Gastrointest Liver Physiol 270: G401 G410, Pape ME, Schultz PA, Rea TJ, DeMattos RB, Kieft K, Bisgaier CL, Newton RS, and Krause BR. Tissue specific changes in acyl-coa:cholesterol acyltransferase (ACAT) mrna levels in rabbits. J Lipid Res 36: , Princen HMG, Meijer P, Wolthers BG, Vonk RJ, and Kuipers F. Cyclosporin A blocks bile acid synthesis in cultured hepatocytes by specific inhibition of chenodeoxycholic acid synthesis. Biochem J 275: , Rudel L, Deckelman C, Wilson M, Scobey M, and Anderson R. Dietary cholesterol and down-regulation of cholesterol 7 hydroxylase and cholesterol absorption in African Green monkeys. J Clin Invest 93: , Rudel L, Lee R, and Cockman T. Acyl coenzyme A:cholesterol acyltransferase types 1 and 2: structure and function in atherosclerosis. Curr Opin Lipidol 12: , Schwarz M, Lund EG, Setchell KDR, Kayden HJ, Zerwekh JE, Björkhem I, and Russell D. Disruption of cholesterol 7 -hydroxylase gene in mice. J Biol Chem 271: , Setchell KDR and Worthington J. A rapid method for the quantitative extraction of bile acids and their conjugates from serum using commercially available reverse-phase octadecylsilane bonded silica cartridges. Clin Chim Acta 125: , Smith LL and Johnson BH. Biological activities of oxysterols. Free Radic Biol Med 7: 285, Spady DK, Cuthbert JA, Williard MN, and Meidell RS. Adenovirus-mediated transfer of a gene encoding cholesterol 7 -hydroxylase into hamsters increases hepatic enzyme activity and reduces plasma total and low density lipoprotein cholesterol. J Clin Invest 96: , Spady DK, Cuthbert JA, Williard MN, and Meidell RS. Overexpression of cholesterol 7 -hydroxylase (CYP7A) in mice lacking the low density lipoprotein (LDL) receptor gene. J Biol Chem 273: , Spitsen GM, Dueland S, Krisans SK, Slattery CJ, Miyake JH, and Davis RA. In nonhepatic cells, cholesterol 7 -hydroxylase induces the expression of genes regulating cholesterol