Mitochondrial oxysterol biosynthetic pathway gives evidence for CYP7B1 as controller of regulatory oxysterols

Journal of Steroid Biochemistry and Molecular Biology 189 (2019) 36–47

Contents lists available at ScienceDirect

Journal of Steroid Biochemistry and Molecular Biology journal homepage: www.elsevier.com/locate/jsbmb

Mitochondrial oxysterol biosynthetic pathway gives evidence for CYP7B1 as controller of regulatory oxysterols

T

Genta Kakiyamaa,d, , Dalila Marquesa,d, Hajime Takeie, Hiroshi Nittonoe, Sandra Ericksonf, Michael Fuchsa,d, Daniel Rodriguez-Agudoa,d, Gregorio Gilc, Phillip B. Hylemonb,d, Huiping Zhoub,d, Jasmohan S. Bajaja,d, William M. Pandaka,d ⁎

a

Department of Internal Medicine, Virginia Commonwealth University, United States Department of Microbiology and Immunology, Virginia Commonwealth University, United States c Department of Biochemistry & Molecular Biology, Virginia Commonwealth University, United States d Department of Veterans Affairs, Richmond, VA, United States e Junshin Clinic Bile Acid Institute, Tokyo, Japan f School of Medicine, University of California, San Francisco, United States b

ARTICLE INFO

ABSTRACT

Keywords: Bile acids and salts Cytochrome p450 Liver Mitochondria Non-alcoholic fatty liver disease Oxysterols

The aim of this paper was to more completely study the mitochondrial CYP27A1 initiated acidic pathway of cholesterol metabolism. The mitochondrial CYP27A1 initiated pathway of cholesterol metabolism (acidic pathway) is known to synthesize two well-described vital regulators of cholesterol/lipid homeostasis, (25R)-26hydroxycholesterol (26HC) and 25-hydroxycholesterol (25HC). Both 26HC and 25HC have been shown to be subsequently 7α-hydroxylated by Cyp7b1; reducing their regulatory abilities and furthering their metabolism to chenodeoxycholic acid (CDCA). Cholesterol delivery into the inner mitochondria membrane, where CYP27A1 is located, is considered the pathway’s only rate-limiting step. To further explore the pathway, we increased cholesterol transport into mitochondrial CYP27A1 by selectively increased expression of the gene encoding the steroidogenic acute transport protein (StarD1). StarD1 overexpression led to an unanticipated marked downregulation of oxysterol 7α-hydroxylase (Cyp7b1), a marked increase in 26HC, and the formation of a third vital regulatory oxysterol, 24(S)-hydroxycholesterol (24HC), in B6/129 mice livers. To explore the further metabolism of 24HC, as well as, 25HC and 26HC, characterizations of oxysterols and bile acids using three murine models (StarD1 overexpression, Cyp7b1−/−, Cyp27a1−/−) and human Hep G2 cells were conducted. This report describes the discovery of a new mitochondrial-initiated pathway of oxysterol/bile acid biosynthesis. Just as importantly, it provides evidence for CYP7B1 as a key regulator of three vital intracellular regulatory oxysterol levels.

1. Introduction The aim of this paper was to more completely study the mitochondrial CYP27A1 initiated acidic pathway of cholesterol metabolism. characterizing the oxysterols and bile acids formed using three murine

models (StarD1 overexpression, Cyp7b1−/−, Cyp27a1−/−) and human Hep G2 cells. Mammalian liver has two major pathways of bile acid synthesis: the neutral (or classical) pathway and the acidic (or alternative) pathway [1–3]. Under normal physiological conditions, the neutral pathway accounts for up to 90% of total bile acid production. The

Abbreviations: ApoE, apolipoprotein E; CYP7A1, cholesterol 7α-hydroxylase (Cyp7a1 denotes murine); CA, cholic acid 3α,7α,12α-trihydroxy-5β-cholanoic acid); CDCA, chenodeoxycholic acid 3α,7α-dihydroxy-5β-cholanoic acid); CYP7B1, oxysterol 7α-hydroxylase (Cyp7b1 denotes murine); CYP8B1, sterol 12α-hydroxylase; CYP27A1, sterol 27-hydroxylase (Cyp27a1 denotes murine); CYP39A, oxysterol 7α-hydroxylase 2; diHC, dihydroxycholesterol; FXR, farnesoid X receptor; HMG-CoA, hydroxy-methylglutaryl-CoA reductase; 24HC, 24(S)-hydroxycholesterol; 25HC, 25-hydroxycholesterol; 26HC, (25R)-26-hydroxycholestetol; HDCA, hyodeoxycholic acid 3α,6α-dihydroxy-5β-cholanoic acid); isoDCA, isodeoxycholic acid 3β,12α-dihydroxy-5β-cholanoic acid); LCA, lithocholic acid (3α-hydroxy-5β-cholanoic acid); LXR, liver X receptor; βMCA, β-muricholic acid 3α,6β,7β-trihydroxy-5β-cholanoic acid); MDCA, murideoxycholic acid 3α,6β-dihydroxy-5β-cholanoic acid); NAFLD, nonalcoholic fatty liver disease; NASH, nonalcoholic steatohepatitis; NTCP, Na+-taurocholate co-transporting polypeptide; RTqPCR, quantitative real-time PCR; SHP, short heterodimer partner; SREBP-1c, sterol regulatory element binding protein-1c; StarD1, steroidogenic acute regulatory protein ⁎ Corresponding author at: Department of Internal Medicine, Virginia Commonwealth University, Richmond, VA 23298, United States.. E-mail address: [email protected] (G. Kakiyama). https://doi.org/10.1016/j.jsbmb.2019.01.011 Received 8 August 2018; Received in revised form 17 January 2019; Accepted 22 January 2019 Available online 30 January 2019 0960-0760/ © 2019 Elsevier Ltd. All rights reserved.

Journal of Steroid Biochemistry and Molecular Biology 189 (2019) 36–47

G. Kakiyama et al.

neutral pathway is initiated by a highly regulated microsomal cytochrome P-450, cholesterol 7α-hydroxylase (CYP7A1). The 7α-hydroxylation of cholesterol is not only the initiating but considered the ratedetermining step in this biochemical pathway; leading to the production of two primary bile acids, cholic acid (CA) and chenodeoxycholic acid (CDCA). The acidic pathway contributes less to total bile acid production. Sterol 27-hydroxylase (CYP27A1) is located in the inner mitochondrial membrane and has been shown to hydroxylate cholesterol to form (25R)-26-hydroxycholesteol (26HC) and 25-hydroxycholesterol (25HC), two vital regulators of cholesterol/lipid homeostasis [4,5]. Unlike liver specific CYP7A1, CYP27A1 is widely expressed in various tissues in the body where it produces these regulatory oxysterols [6,7]. Although CYP27A1 catalyzes the first reaction of the acidic pathway, it is not the rate-limiting step. It appears to be more constitutively expressed as compared to the more highly regulated feed-back response seen with CYP7A1 to feeding of bile acids or cholesterol [8]. Our laboratories have previously shown that a rate-limiting step in this bile acid synthesis pathway is cholesterol transport into the inner mitochondria as mediated by the LXR induced steroidogenic acute regulatory protein (StarD1) [9,10]; creating a feed-forward pathway driven by increasing LXR activating oxysterol levels. 25HC and 26HC are rapidly 7α-hydroxylated by microsomal oxysterol 7α-hydroxylase (CYP7B1), reducing their regulatory abilities and cytotoxicity until ultimately being metabolized to chenodeoxycholic acid (CDCA) [7,11]. Of interest, 24(S)-hydroxycholesterol (24HC) which is predominantly produced in the brain, is carried to the liver by apolipoprotein E (ApoE) and is metabolized to bile acids by the same pathway [12,13]. Interestingly, previous work in purified pig microsomal preparations was not shown CYP7B1 to have 7αhydroxylation activity toward this oxysterol [14]. Instead, two other cytochrome P-450 enzymes, CYP7A1 and oxysterol 7α-hydroxylase 2 (CYP39A1), were reported to have 7α-hydroxylation activity for 24HC in the liver [14,15]. Despite important historical observations, the acidic pathway of cholesterol metabolism has become relegated to one of lessor importance after it was observed to play a more minor role in the rates of bile acid synthesis. However, Javitt et al. have shown that the acidic pathway is present early in human development [7,16]. From a developmental perspective, the foremost role of the acidic pathway is to generate and control the levels of life sustaining regulatory oxysterols that help control cellular cholesterol and lipid homeostasis. The later evolution in the acidic pathway to synthesize bile acids is believed to serve as a means to aid developing organisms solubilize and absorb dietary lipids; lessening the need for these sterols to be internally synthesized. Thus, the highly regulated CYP7A1-initated pathway likely developed later as a means to more tightly control the levels of bile acids synthesized without affecting vital regulatory oxysterol synthesis. Despite extensive descriptions of the cholesterol and lipid pathways controlled by acidic pathway derived oxysterols, it is not clear what controls the levels of these important regulatory oxysterols. However, the absence of or excess of oxysterols as found in the absence of CYP27A1 or of CYP7B1, respectively, leads to devastating clinical human phenotypes; cerebrotendineous xanthomatosis (CTX) [17] and end stage fibrotic liver disease [18–21]. CTX is a well-described cholesterol excess phenotype which develops in the absence of a functional acidic pathway which generates oxysterols [17,22]. In CTX, cholesterol synthesis continues in the face of tissue cholesterol excess; evidence supportive of oxysterols as key physiologic regulators of de novo cholesterol biosynthesis [17,22,23]. As originally described by Setchell and Russell [18], and confirmed by other groups [19,20,21], the complete absence of CYP7B1 leads to aggressive inflammation and fibrosis in association with a marked increase in 24HC, 25HC, 26HC, 3β-hydroxy5-cholestenoic, and 3β-hydroxy-5-cholenoic acids. The current research project was initially begun to determine if increased catabolism of cholesterol through the alternative pathway could represent a means to reduce cholesterol and fat within a nonalcoholic fatty liver disease (NAFLD) mouse model. Our laboratory has

believed that the excess fat within the liver in NAFLD is in large part a function of hepatic cholesterol dysregulation. And, that increasing cholesterol catabolism through the alternative pathway would also lead to a marked and rapid reduction of liver triglycerides. This hypothesis was premised on our previous observations where we showed StarD1 overexpression could reverse aortic endothelial lipid accumulation in ApoE−/− Western diet fed mouse model [24]. Preliminary observations in a StarD1 overexpression in a fatty liver mouse model showed a marked accumulation of 25HC and 26HC, and an unanticipated increase in 24HC. In the current study we chose to further characterize oxysterol and bile acid metabolism using a StarD1 overexpression mouse model, and Cyp7b1 and Cyp27a1 knockout models. Based on our results, we provide evidence for a novel bile acid synthesis pathway initiated by cholesterol 24(S)-hydroxylation in liver mitochondria. We also provide evidence for CYP7B1 as a key regulator of the intracellular oxysterol levels for 26HC, 25HC, and 24HC, and, as an additional ratedetermining step in the synthesis of bile acids via the acidic pathway. 2. Materials and methods 2.1. General reagents and materials All chemicals were of the highest purity commercially available. HPLC grade solvents were purchased from Fisher Scientific (New Lawn, NJ), and all other reagents were from Sigma-Aldrich, Inc. (St. Louis, MO), unless indicated otherwise: The radiolabeled cholesterol, 1,2-[3H (N)]-Cholesterol (50 mCi/mmol) and 4-[14C]-cholesterol (59.4 mCi/ mmol) were obtained from New England Nuclear (Boston, MA). 22,23[3H]-24(S)-Hydroxycholesterol was purchased from American Radiolabeled Chemicals, Inc. (St. Louis, MO), and was purified by HPLC prior to use. Un-labeled 24(S)-Hydroxycholesterol and 7α,24(S)-dihydroxycholesterol was purchased from Avanti Polar Lipids, Inc. (Alabaster, AL). 1,2-[3H(N)]-(25R)-26-hydroxycholesterol and 1,2-[3H (N)]-25-hydroxycholesterol were prepared from 1,2-[3H(N)]Cholesterol as described previously [25]. All other oxysterol standards and bile acids were from Steraloids, Inc. (Newport, RI). Recombinant CYP7B1 protein was available from Novus Biologicals, LLC (Centennial, CO). Cell culture reagents and supplies were purchased from GIBCO BRL (Grand Island, NY). SV Total RNA Isolation System was obtained from Promega Co. (Madison, WI). Regular Chow and Western Diet (TD.88137) were purchased from Envigo (Frederick, MD). 2.2. Animals The animal protocols used were reviewed and approved annually by the Institutional Animal Care and Use Committees (IACUC) of the McGuire VA Medical Center and the Virginia Commonwealth University; both are accredited by the Association for Assessment and Accreditation of Laboratory Animal Care International and comply with the “Guide for Care and Use of Laboratory Animals” published by the National Institute of Health. Cyp7b1−/− mice were recovered from cryo preserved zygotes at Jackson Labs (Coatesville, PA). Cyp27a1−/− mice were obtained from Dr. Sandra Erickson. An F2 intercross of the mouse strains, 129S1/SvlmJ and C57Bl/6 J (B6/129) that develop classic fatty liver with progression to nonalcoholic steatohepatitis (NASH) on a supplemental Western diet chow were obtained through the generous co-operation of Dr. Sandra Erickson. Adult male mice, 13–15 weeks of age, bred at our facility, kept on a 12 h light cycle (6 a.m. to 6 pm), were fed a restricted Western diet (TD.88137) with 42% of calories from fat and 43% from carbohydrates. All mice received the same amounts of calories and had free access to water. 2.3. Adenovirus preparation The adenovirus constructs Ad-β-Gal, Ad-StarD1, or Ad-CYP27A1 were prepared as previously described [26,27]. Briefly, recombinant 37

Journal of Steroid Biochemistry and Molecular Biology 189 (2019) 36–47

G. Kakiyama et al.

Fig. 1. Overexpression of StarD1 increases oxysterol levels in mouse livers. Shown is the effect of StarD1 overexpression on 3 key oxysterol levels in B6/129 (wild type), Cyp7b1−/− and Cyp27a1−/− mice liver and brain. Ad-StarD1 recombinant adenovirus was injected to from the tail vain of 13–14 week old mice to overexpress the selected gene as described. One week after adenovirus injection, mice were briefly anesthetized, and liver and brain were harvested. Oxysterol levels were quantified by HPLC as described in Method. 1) Liver oxysterol levels (Panel A–C) (Control, n = 5; StarD1, n = 14) and brain 24HC levels (Panel D) (Control, n = 3; StarD1, n = 6) of B6/129 mice; 2) Liver oxysterol levels (Panel E–G) (Control, n = 6; StarD1, n = 14) and brain 24HC levels (Panel H) (Control, n = 8; StarD1, n = 4) of Cyp7b1−/− mice; 3) Liver oxysterol levels of Cyp27a1−/− mice (Panel I) (n = 9). By two tails unpaired t-test (homoscedastic), *P ≤ 0.05, and **P ≤ 0.01, and ***P ≤ 0.001 vs Control.　†There were no significant differences in the results between equal variances t-test and unequal variances t-test. ns, not significant; nd, not detected.

virus were amplified by infecting confluent monolayers of human embryonic kidney 293 cells (American Type Culture Collection, Manassas, VA) grown in 150 mm tissue culture dishes in Dulbecco’s Modified Eagle Medium (DMEM) with 5% fetal bovine serum (FBS). For in vivo studies, the virus was purified using the Adeno-X Purification kit from Clontech (Mountain View, CA) according to manufacturer’s directions.

for another week. On day 14, mice were briefly anesthetized, and liver and brain were harvested, which were then quickly frozen in liquid nitrogen and stored at −78 °C until analysis. 2.5. In vivo cell culture experiments Primary hepatocytes were isolated from the livers of B6/129, Cyp7b1−/−, or Cyp27a1−/− mice using the collagenase-perfusion technique as previously described [28]. The cells (9.0 × 106) were plated on 150 mm tissue culture dishes in Williams’ E media (20 ml) containing insulin (0.25 units/ml), penicillin (100 units/ml) and Dexamethasone (0.1 μM). The cells were maintained in the absence of thyroid hormone in 5% CO2 at 37 °C. Twenty-four hours after plating, the spent medium was replaced with fresh medium (4 ml) and recombinant adenovirus encoding

2.4. Animal tissue collecting for oxysterol analysis Mice were fed a Western diet (TD.88137) for 2 weeks. On day 7, Adβ-Gal (control) or Ad-StarD1 recombinant adenovirus was injected from the tail vain to overexpress the selected gene. Previously, we have shown that StarD1 overexpression in mice led to marked increases in hepatic StarD1 protein levels [24]. Western diet feeding was continued 38

Journal of Steroid Biochemistry and Molecular Biology 189 (2019) 36–47

G. Kakiyama et al.

Fig. 2. Effect of overexpression of StarD1 on mRNA levels of 4 key regulators of bile acid homeostasis. Shown are the relative hepatic mRNA expressions of bile acid synthesis related genes in StarD1 overexpressed B6/129 mice: Gene expression was quantified by real-time RT-qPCR as described in the Method section. The mRNA expression levels obtained for each gene were normalized to the expression of the GAPDH housekeeping gene. n = 6 each group. By paired t-test, *P ≤ 0.05, and **P ≤ 0.01, and ***P ≤ 0.001 vs Control.

either Ad-StarD1, Ad-CYP7A1, or Ad-β-Gal (as control) was added at a multiplicity of 10 pfu/cell. The culture was maintained at 37 °C for 2 h. After unbound virus was removed, fresh medium (20 ml) containing either 2.5 μCi of 4-[14C]-cholesterol, 1,2-[3H(N)]-cholesterol, 22,23-[3H]-24(S)hydroxycholesterol, 1,2-[3H(N)]-25-hydroxycholesterol or 1,2-[3H(N)](25R)-26-hydroxycholesterol was added depending on the purpose of the experiment. The resulting culture was incubated at 37 °C in 5% CO2 for 48 h to form oxysterols and bile acids. The cells were harvested with cold PBS (1 ml) and oxysterols and bile acids were analyzed as described below. HepG2 cells (9.0 × 106 cells) were plated on 150 mm tissue culture dishes in MEM media (20 ml) with FBS. They were infected with unpurified recombinant adenovirus encoding either Ad-StarD1, or Ad-βGal (as control) as same manner above. After a 48-hour incubation with 1,2-[3H(N)]-cholesterol, the cells were harvested in cold PBS (1 ml) and formed oxysterols were analyzed as described below.

nitrogen gas stream. Obtained residue was then re-suspended in methanol-acetonitrile-water 50:50:1, v/v/v (120 μl) and an aliquot (100 μl) was applied to the preparative HPLC to purify desired oxysterol fraction as follows: A Waters Nova Pak C18 column (300 mm × 3.9 mm inner diameter, 4 μm particle size) was employed at 32 °C. The mobile phase used was a mixture of methanol-acetonitrile (1:1, v/v) (90%) and water (10%), and the flow rate was kept constant at 1.0 ml/min. Under this condition, the effluent between 9 min and 20 min was collected (total 11 ml) in a glass test tube. After solvent was evaporated under a nitrogen gas stream, oxysterols were treated with cholesterol oxidase to form 3-Oxo-Δ4 sterols according to the method of Zhang et al [29]. Thus prepared samples were dissolved in the initial mobile phase (120 μl) and an aliquot (100 μl) was subjected to the normal phase HPLC analysis as follows: A Waters alliance® series 2695 separation module equipped with a 2487 dual λ absorbance detector, which was controlled by Empower Pro software, was used. A Waters Sun Fire Silica column (250 mm × 4.6 mm inner diameter, particle size 5 μm) was kept at 32 °C during analysis. The mobile phase was initially n-hexaneisopropanol (100:2, v/v) and then changed to n-hexane-isopropanol (100:3, v/v) at the time point of 35 min. The flow rate was kept at 1.2 ml/min throughout the analysis. The 3-Oxo-Δ4 sterols were monitored by their 233 nm absorption. In case of tracing radioactive metabolites, the effluents were fractionated every 20 s (0.4 ml/fr.) and radioactivity in each fraction was measured by Beckmann LS 6000IC liquid scintillation counter.

2.6. Oxysterols analysis in the mouse tissue and cultured cells Frozen liver (500 mg) was homogenized using cold PBS (1.5 ml). For the frozen cells, they were re-suspended in PBS (1.5 ml). To the liver homogenate or cell suspension, 20 mg/ml Proteinase K solution (150 μl) was added, and the mixture was incubated at 55 °C for 16 h. Methanol (2 ml) was added, and the mixture was ultra-sonicated in a Branson type B-220 ultra-sonic bath (Danbury, CT) for 30 min. After adding 24Ketocholesterol (250 pmol) as an internal standard, oxysterols were extracted with n-hexane (2 ml x 3 times). After n-hexane was evaporated under a nitrogen stream, the residue was re-dissolved in methanol-acetonitrile-water, 50:50:1, v/v (2 ml) and loaded onto a Waters Sep-Pak® tC18 cartridge (500 mg) that had been primed with above solvents (10 ml). The column was continuously eluted with additional 5 ml of the same solvents. The collected effluent was dried under a

2.7. Bile acid analysis in primary mouse hepatocytes The spent culture media (20 ml) was alkalized with 2 N NaOH (0.5 ml), and loaded onto a Waters Sep Pak tC18 cartridge (10 g) which had been primed with methanol and water. After washing the column 39

Journal of Steroid Biochemistry and Molecular Biology 189 (2019) 36–47

G. Kakiyama et al.

Fig. 3. Overexpression of StarD1 in isolated primary mouse hepatocytes increases levels of 24HC, 25HC, and 26HC. Shown are the HPLC profiles of oxysterols generated in the primary mouse hepatocytes. Radiolabeled cholesterol was incubated with primary hepatocytes isolated from either B6/129 (wild type), Cyp7b1−/−, or Cyp27a1−/− mouse as described in Methods. Each condition was repeated three times (n = 3) and representative profile is shown. (A) In B6/129 (wild type), radioactive oxysterols were undetectable under basal condition; (B) StarD1 overexpression in B6/129 increased the levels of 24HC, 25HC and 26HC to detectable levels; (C) In Cyp7b1−/−, 26HC was detectable under basal condition, but 25HC and 24HC were below detectable levels. (D) StarD1 overexpression in Cyp7b1−/ − significantly increased the 26HC peak as well as 24HC and 25HC. (E) Cyp27a1−/− did not produce 24HC, 25HC, or 26HC. (F) StarD1 overexpression in Cyp27a1−/− did not increase levels of 24HC, 25HC, or 26HC. Two radioactive peaks appeared at 19 min (##) and 22.5 min (#) were unidentified.

Fig. 4. Overexpression of StarD1 in HepG2 cells increases levels of 24HC, 25HC, and 26HC. Generation of 24HC, 25HC, and 26HC in the HepG2 cell culture. (A) Representative HPLC profile of [3H]-oxysterols generated by incubation of 1,2-[3H(N)]-cholesterol with HepG2 cell culture for 48 h. Upper panel presents StarD1 overexpressed HepG2 cell culture; Lower panel presents Control HepG2 cell culture. (B) Quantifications of 24HC, 25HC, and 26HC presented in the HepG2 cells with/ without StarD1 overexpression after 48 h incubation with 1,2-[3H(N)]-cholesterol. Each oxysterol level was total of endogenous oxysterol and exogenous radioactive oxysterol, and was expressed as pmol/106 cells, n = 3.

40

Journal of Steroid Biochemistry and Molecular Biology 189 (2019) 36–47

G. Kakiyama et al.

with water (120 ml), desired bile acids were eluted with methanol (70 ml). Methanol was evaporated under a nitrogen stream at 40 °C. The dried residue was re-suspended in 50 mM sodium acetate buffer, pH 5.6 (1 ml) containing cholylglycine hydrolase (15 units) and sulfatase (Type H) (15 units), and the mixture was incubated at 37 °C for 16 h. After adding norDCA (50 nmol) as an internal standard, the mixture was diluted with 0.1 N NaOH (3 ml). The alkalized mixture was applied to a Waters Sep Pak tC18 cartridge (500 mg sorbent), which had been primed with methanol (10 ml) and water (10 ml). The cartridge was successively washed with water (5 ml), 10% acetone (4 ml), and water (5 ml). Retained bile acids were eluted with methanol (6 ml), which was then evaporated under a nitrogen gas stream at 40 °C. Obtained de-conjugated bile acids were derivatized to 24-phenacyl esters as described in the previously [26]. A Waters Nova Pak C18 column (300 mm × 3.9 mm inner diameter, particle size 4 μm) was kept at 32 °C during analysis. The initial mobile phase was a mixture of acetonitrile-methanol (1:1, v/v) (65%) & water (35%) which was kept for 30 min. From 30 min to 70 min, the composition of acetonitrile-methanol (1:1, v/v) was linearly increased to 75% (25% water). The composition of acetonitrile-methanol (75%) & water (25%) was kept for another 20 min (total run time 90 min). The flow rate was kept at 0.72 ml/min throughout the analysis. The effluents were fractionated every 20 s (0.24 ml/fr.) and radioactivity in each fraction was measured in a Beckmann LS 6000IC liquid scintillation counter. The 24-phenacyl bile acids were also monitored by their absorption at 254 nm. 2.8. Mitochondria isolation and conversion of labeled cholesterol to oxysterols Mitochondria were isolated from primary mouse hepatocytes as described previously [30,31]. Two cultures were prepared: uninfected and infected with Ad-CYP27A1. Purified mitochondria were incubated with 20 μCi of 1,2-[3H(N)]-Cholesterol in the following solution: 20 mM β-Cyclodextrin (10 μl), 1 M Potassium phosphate buffer, pH7.4 (50 μl), 100 mM DTT (5 μl), 20 mM EDTA (5 μl), 10 mM β-NADPH (60 μl), Trisodium isocyanate (50 μl) and water (305 μl) [32]. After 3.5 h incubation, chloroform-methanol, 1:1 (4 ml) was added, and the mixture was ultrasonicated in a Branson type B-220 ultra-sonic bath (Danbury, CT) for 15 min. The solvents were evaporated under a nitrogen gas stream. The residue was re-suspended in methanol-acetonitrile-water 100:100:0.5, v/v/v (1.5 ml), and applied to a Waters Sep Pak tC18 cartridge (500 mg sorbent) which had been primed with same eluent. After loading the sample, the column was continuously eluted with additional 5.5 ml of same solvent, and total of 7 ml of the effluent was collected. The eluent was evaporated under a nitrogen gas stream. Obtained residue was then re-suspended in methanol-acetonitrile-water 50:50:1, v/v (120 μl) and an aliquot (100 μl) was applied to the preparative HPLC to purify the oxysterol fraction by the same manner described in tissue oxysterol analysis above (See 2.6). 2.9. mRNA analysis Total RNA was isolated from 30 mg of liver tissue with the Promega SV Total RNA Isolation System, which includes a DNase step. Reverse transcription (RT) was performed with 2 μg of RNA and real time qPCR (RTqPCR) was performed with 20 ng of cDNA from the RT step using the Affimetrix (Santa Clara, CA) very Quest SYBR Green PCR Master Mix in an Applied Biosystems 7500 PCR System (Foster City, CA). All data were produced in triplicates for each mRNA. Primers used for the PCR are listed in the Supplementary Table 1. The comparative CT method was used to calculate relative quantification of gene expression. The mRNA expression levels obtained for each gene were normalized to the expression of the GAPDH housekeeping gene by using the following equation: relative mRNA expression = 2−(Ct of each gene−Ct of GAPDH) (where Ct is the threshold cycle).

Fig. 5. Conversion of 1,2-[3H(N)]-cholesterol to 1,2-[3H(N)]-oxysterols by isolated mouse hepatocyte mitochondria. Isolated mitochondria were incubated with 1,2-[3H(N)]-cholesterol in the presence of β-CD and NADPH as described in the methods. Each condition was repeated three times (n = 3) and representative profile is shown. (A) HPLC profile of 1,2-[3H(N)]-oxysterols generated by the mitochondria isolated from B6/129 mice hepatocytes. (B) HPLC profile of produced 1,2-[3H(N)]-oxysterols by the mitochondria isolated from CYP27 A1 overexpressing primary mouse hepatocytes. (C) HPLC profile obtained from heat inactivated mitochondria (80 °C for 10 min). 41

Journal of Steroid Biochemistry and Molecular Biology 189 (2019) 36–47

G. Kakiyama et al.

Fig. 6. Bioconversion of 24HC, 25HC, and 26HC to bile acids in primary mouse hepatocyte cultures. (A) HPLC profile obtained by incubation of 22,23-[3H]-24HC with B6/129 (wild type) hepatocytes; (B) HPLC profile obtained by incubation of 22,23-[3H]-24HC with Cyp7b1−/− hepatocytes; (C) HPLC profile obtained by incubation of 1,2-[3H(N)]-25HC with B6/129 hepatocytes; (D) HPLC profile obtained by incubation of 1,2-[3H]-26HC with B6/129 hepatocytes. A representative profile of three experiments are shown (n = 3).

2.10. Statistical analysis

encoding StarD1. These results provide evidence for that Cyp7b1 plays an important role in controlling oxysterol levels and ratios.

Data are reported as means ± SEM. Where indicated, data were subjected to Student’s t-test and/or Welch’s t-test and determined to be significantly different at P ≤ 0.05.

3.1.2. Hepatic mRNA expressions in StarD1 overexpressed mice Fig. 2 shows steady-state mRNA levels for bile acid synthesis related genes in B6/129 mouse livers fed a Western Diet. Previously, we have shown both in vitro and in vivo mouse studies an increase in rates of bile acid synthesis with StarD1 overexpression [9,10]. StarD1 overexpression led to decreases in cyp7a1 and nctp, and, increases in shp mRNA levels. Interestingly, a marked suppression of cyp7b1 mRNA (↓ 85 ± 6%, P ≤ 0.001) was observed in the StarD1 overexpressed mice.

3. Results 3.1. Generation of 24(S)-hydroxycholesterol in hepatocyte mitochondria 3.1.1. Oxysterol analysis in StarD1 overexpressed mice liver and brain Fig. 1 presents the effect of StarD1 overexpression on the levels of key oxysterols (24HC, 25HC, and 26HC) in B6/129 (as wild type), Cyp7b1−/−, and Cyp27a1−/− mice livers. For 24HC, the levels in the brain of these mice are also presented. In the StarD1 overexpressed B6/ 129 mice, 26HC level was significantly higher than that in the control uninfected group (Panel B). Unexpectedly, a high level (13-fold vs control, P ≤ 0.001) of 24HC was also found in the StarD1 overexpressed mice (Panel A). In Cyp7b1−/− mice liver, 25HC and 26HC levels were dramatically increased with StarD1 overexpression (5-7fold, P ≤ 0.001) (Panel F and G). A significant increase was also seen in 24HC (3-fold vs control, P ≤ 0.001) (Panel E). The increase in all three oxysterol levels was much more than seen in B6/129 mice. While, Cyp27a1−/− mouse liver presented with only a small amount of 25HC (Panel I). No detectable levels of 24HC and 26HC were present in the Cyp27a1−/− mouse. Since brain is the major source of 24HC in vivo, we analyzed 24HC levels in the brain of these mice. Regardless of the mice strain, no significant differences were found with/without StarD1 overexpressed mice (Panel D and H). These results suggested 24HC was produced within the liver following overexpression of the gene

3.1.3. Conversion of radiolabeled cholesterol to oxysterols in primary mouse hepatocytes Primary mouse hepatocytes (9.0 × 106 cells) cultures were prepared from B6/129, Cyp7b1−/−, and Cyp27a1−/− mice, respectively. They were infected with StarD1 or control recombinant adenovirus (Adβ-Gal), and incubated with radiolabeled cholesterol as described in Materials and methods. Fig. 3 shows HPLC profiles of radiolabeled oxysterols generated after 48 h incubation. B6/129 mice hepatocytes, which were representative of wild types of Cyp7b1−/− and Cyp27 A1-/hepatocytes, did not produce any detectable level of 4-[14C]-oxysterol (Panel A). The StarD1 overexpressed B6/129 hepatocytes produced larger [14C]-26HC peak along with smaller [14C]-24HC and [14C]-25HC peaks (Panel B). In Cyp7b1-/- primary mouse hepatocytes culture (Panel C), detectable levels of [14C]-26HC was produced but, unexpectedly, [14C]-25HC and [14C]-24HC were not detected. This unchanged levels of these two oxysterols was likely because the exogenously added radiolabeled cholesterol could not be transported into mitochondria efficiently. When StarD1 was overexpressed in the Cyp7b1-/-

42

Journal of Steroid Biochemistry and Molecular Biology 189 (2019) 36–47

G. Kakiyama et al.

hepatocytes (Panel D), [14C]-26HC, [14C]-25HC and [14C]-24HC levels were significantly elevated, and these elevation levels was roughly 2–3 times higher than those when StarD1 was overexpressed in the B6/129 hepatocyte (Panel B; note Y-axis scale change from Panel D). In Cyp27a1-/- hepatocytes (Panel E and F), no radioactive peaks corresponding to 24HC, 25HC, and 26HC were observed with or without

StarD1 overexpression. Of note, two unknown large peaks (labeled as # and ##) were produced in the Cyp27a1-/- hepatocytes. Their identification was not attempted as not considered relevant to focus of this study.

(caption on next page) 43

Journal of Steroid Biochemistry and Molecular Biology 189 (2019) 36–47

G. Kakiyama et al.

Fig. 7. (a): Alternative pathway of bile acid synthesis. This summary figure outlines the early steps in the metabolism of cholesterol to bile acids as initiated by mitochondrial CYP27A1. Hepatocyte mitochondrial CYP27A1 generated three oxysterols 26HC, 25HC, and 24HC. CYP27A1 can further oxidize part of 26HC to 3βHydroxy-5-cholestenoic acid. Similarly, CYP27A1 can oxidize part of 24HC and 25HC to give a same di-hydroxylated product, 24(S),25-Dihydroxycholesterol (24(S),25-diHC). For subsequent 7α-hydroxylation of these oxystsrols, CYP7B1 is responsible for 25HC and 26HC, and two cytochrome P-450 enzymes, CYP7A1 and CYP39A, are reported for 24HC’s 7α-hydroxylation. However, in present study, CYP7B1 also showed some 7α-hydroxylation activity towards 24HC at least in the mouse. (b): Alternative pathway of bile acid synthesis (continuation of Fig. 7A). This figure outlines in a continuation figure the key metabolites of cholesterol via the alternative pathway to bile acids initiated by mitochondrial CYP27A1. Following possible pathways to bile acids: HSD3B7 converts these dihydroxycholesterols to corresponding cholest-4-en-3-ones. These cholest-4-en-3-ones can be eventually converted to CDCA via several more steps. In mouse, CDCA can be hydroxylated at C6 position to form β-MCA. In mouse hepatocyte culture (See Fig. 6A), CA was produced from 24HC. For CA biosynthesis, 12α-hydroxylation of 7α,24(S)-Dihydroxycholest-4-en-3-one is required and CYP8B1 could catalyze this hydroxylation, but further investigation is needed. While, (25R)-24(S),25-diHC and 3β-Hydroxy5-cholestenoic acid, which can be produced by CYP27A1 under chronic down-regulation of CYP7B1, can be metabolized to LCA through 3β-Hydroxy-5-cholenoic acid creating a potential hepatotoxic pathway (arrow 4). In mouse, C6-position of LCA can be hydroxylated in microsomal hydroxylase forming MDCA and HDCA. This pathway was particularly dominant in CYP7B1−/− mouse hepatocytes (See Fig. 6B).

3.1.4. Conversion of radiolabeled cholesterol to oxysterols in HepG2 cells In an attempt to correlate human to murine hepatocyte findings, HepG2 cells were infected with StarD1 recombinant adenovirus. Fig. 4A shows the HPLC profiles of [3H]-oxysterols produced after 48 h incubation with [3H]-cholesterol. As with the primary mouse hepatocyte culture, StarD1 overexpression elevated [14C]-26HC, [14C]-25HC and [14C]-24HC productions. Fig. 4B shows the quantifications of the oxysterol levels in the HepG2 cells. In StarD1 overexpressed culture, 24HC was about 6-fold increased from untreated culture (1.1 pmol/106 cells → 6.4 pmol/106 cells). The increase in 25HC was also about 6-fold (4.6 pmol/106 cells → 26 pmol/106 cells). As seen in mouse hepatocytes, the largest increase was seen in 26HC which was dramatically increased over control (37 pmol/106 cells) with StarD1 overexpression (400 pmol/106 cells).

appeared at 3 min and 10 min; likely conjugated bile acids that were not hydrolyzed in the deconjugation process. Panel B presents the metabolites of [3H]-24HC in the Cyp7b1-/- hepatocytes. [3H]-CDCA and [3H]-CA were not detected, evidence Cyp7b1 is required for their synthesis. Furthermore, the level of [3H]-β-MCA was much smaller than that detected in B6/129 hepatocytes. Instead, significantly increased [3H]-MDCA and [3H]-HDCA were detected. Of note, HDCA and MDCA were likely derived from potentially lipotoxic 3β-hydroxy-5-cholenoic acid and lithocholic acid (LCA) [19,33] (See Fig. 7); creating a potential hepatotoxic pathway in the presence of chronic Cyp7b1 down-regulation. Previously, using a biliary diverted mouse model, StarD1 overexpression led to a 2-fold increase in rates of total bile acid synthesis [10]. Panel C shows the [3H]-bile acid profile from incubation of [3H]25HC with the B6/129 hepatocytes. [3H]-β-MCA and [3H]-MDCA were identified as major bile acid products. A considerably large peak appeared at 21 min (*) along with some smaller peaks between 20 min and 28 min. These peaks were not produced from other two oxysterols. Unfortunately, the retention times of these peaks were not compatible with any available bile acid standards, and further identification was not attempted. Finally, Panel D presents the [3H]-bile acid profile from incubation of [3H]-26HC with the wild type hepatocytes. This oxysterol metabolized mostly to [3H]-β-MCA and [3H]-MDCA. The retention times of these authentic bile acids for the references are presented in the Supplemental Figure S1.

3.1.5. Conversion of labeled cholesterol to oxysterols in isolated mitochondria Mitochondria were isolated from B6/129 mouse hepatocytes, and were assayed with [3H]-labeled cholesterol as described in the Methods. Note: the use of β-cyclodextrin in vitro in the described isolated mitochondrial assay system delivers a surplus of cholesterol substrate to inner mitochondrial membrane [30–32]; bypassing need for StarD1 overexpression needed in isolated primary hepatocytes or in in vivo models. Fig. 5 shows HPLC profiles of produced [3H]-oxysterols. The untreated mitochondria gave smaller radioactive peaks corresponding to 26HC, 25HC, and 24HC (Panel A). Mitochondria isolated from hepatocytes overexpressing Cyp27a1 prior to their isolation gave larger radioactive 24HC, 25HC and 26HC (Panel B). None of these radioactive peaks were detected from incubation with heat inactivated mitochondria (Panel C).

4. Discussion In hepatocyte mitochondria, CYP27A1 converts cholesterol into 26HC and 25HC. Both oxysterols can then be 7α-hydroxylated and ultimately metabolized to CDCA and its muricholic metabolites (Fig. 6 C and D). The present study shows that mitochondrial CYP27A1 is also able to convert cholesterol to a third oxysterol, 24HC; and, that this oxysterol, like 26HC and 25HC, can be further metabolized to bile acids (Fig. 4A). Interestingly, CYP27A1 is a bifunctional enzyme involved in both bile acid and vitamin D metabolism [5]. In vitamin D metabolism, CYP27 A1 catalyzes 24-hydroxylation of Vitamin D2, 1α-OH vitamin D2 [34,35], and 1α-OH vitamin D3 [36]. In addition, it has been reported that CYP27A1 is capable of catalyzing 24(S)-hydroxylation of artificial analogues of vitamin D in vitro [37]. In addition, primary mouse hepatocyte mitochondria converted [3H]-cholesterol to [3H]-24HC; conversion that is increased when CYP27A1 is overexpressed in cells prior to mitochondria isolation (Fig. 5). Conversely, Cyp27a1−/− hepatocytes were incapable of conversion of [3H]-cholesterol to [3H]-24HC (Fig. 3 E and F); and that Cyp27a1−/− mouse liver did not contain detectable levels of 24HC or 26HC in vivo (Fig. 1 I). These results supported two earlier reports in rat [27] and in pig [38] liver mitochondria which without certainty hypothesized the possibility of mitochondrial 24(S)-hydroxylation activity [27,39]. Previous reports have shown that oral administration

3.2. Metabolism of 24(S)-hydroxycholesterol to bile acids 3.2.1. Bioconversion of labeled 24(S)-, 25-, and (25R)-26hydroxycholesterol to bile acids in primary mouse hepatocytes Primary hepatocyte cultures were prepared from B6/129 and Cyp7b1−/− mice, respectively. These were incubated with 22,23-[3H]24HC, 1,2-[3H(N)]-25HC, or 1,2-[3H(N)]-26HC. The produced [3H]bile acids were extracted from the culture media as described in Methods. Since the majority of bile acids were conjugated with taurine or glycine, the extracts were treated with cholylglycine hydrolase for deconjugation prior to HPLC analysis [26]. Fig. 6 shows the radioactive HPLC profiles of [3H]-bile acids generated after 48 h incubation. Incubation of [3H]-24HC with B6/129 mouse (wild type) hepatocytes (Panel A) produced [3H]-β-muricholic acid (β-MCA), [3H]-chenodeoxycholic acid (CDCA), and [3H]-cholic acid (CA). Additionally, [3H]-24HC was metabolized to murideoxycholic acid (MDCA: 3α,6βdihydroxy-5β-cholanoic acid) and hyoxdeoxycholic acid (HDCA: 3α,6α-dihydroxy-5β-cholanoic acid). Two relatively large peaks

44

Journal of Steroid Biochemistry and Molecular Biology 189 (2019) 36–47

G. Kakiyama et al.

25HC nor [3H]-26HC was converted to [3H]-CA in the primary hepatocyte culture (Fig. 6 C and D). While, 24HC can be also converted to (25R)-24(S),26-diHC by Cyp27a1 [12] (not shown in Fig. 7). A possibility of conversion of this dihydroxy cholesterol intermediate to CA cannot be excluded. Oxysterols represent key regulators of cholesterol and lipid homeostasis [11]; down-regulating intracellular cholesterol biosynthesis by inhibiting processing of SREBP-2, while serving as ligands of LXR to up-regulate SREBP-1c transcription and lipogenesis [11,48–51]. Consistent with findings of present study, previous studies also provide evidence that CYP7B1 plays a pivotal role in regulating oxysterol levels [11,18]. However, it was not the intent of this study to determine the quantitative relevance of the amounts of oxysterols or bile acids synthesized, or to further our understanding of what regulates CYP7B1. To more clearly define the exact role of the acidic pathway of hepatic cholesterol/bile acid metabolism, a better understanding of the regulation of CYP7B1 is necessary. The downregulation of Cyp7b1 elicited by StarD1 overexpression suggests Cyp7b1 to be a highly regulated gene [8]. Although regulation of CYP7B1 has been observed in many studies, its role and reason for its regulation have not been clearly defined. Our laboratory has demonstrated a circadian rhythm [52]. Additionally, we and others have shown a direct correlation between Cyp7b1 mRNA levels and its specific activity [8,52,53]; suggesting the regulation to be primarily at the transcriptional level. There exist investigations both supportive of and against the role of FXR [54,55] in bile acid feeding mediated negative feedback of CYP7B1 [8,52,53]. However, given its lessor role in determining the rates of bile acid synthesis, there does not appear to be the physiological need for CYP7B1 to be feedback regulated by FXR like CYP7A1 to control rates of bile acid synthesis. [56,57]. Furthermore, the absence of an FXR mediated promoter regulation of Cyp7b1−/− would suggest that bile acids are likely not eliciting major feedback regulation through FXR [56]. Whether bile acids can elicit a direct feedback regulation, or the increasing levels of regulatory oxysterols can elicit feedback [58], is not clearly defined. However, Cyp7b1 appears to be a much more highly regulated gene than previously believed, and one which may be able to respond ‘acutely’ to cell cholesterol or lipid levels to modulate SREBPs and LXR attempting to rapidly regulate oxysterols levels. Our previous studies in rat hepatocytes support this concept [53]. In addition to cholesterol and lipid metabolism, CYP7B1 may regulate pregnenolone [59,60], dehydroepiandrosterone (DHEA) [61,62], and ultimately cortisol levels. CYP7B1 can also hydroxylate the potent estrogen receptor-β agonist, 5α-androstane-3β,17β-diol [63,64]. Furthermore, 26HC has been shown to possess estrogen-like properties that have been suggested to have a role in breast cancer [65], and 25HC is an innate immune modulator through its effects on IgA [66,67]. Reduced Cyp7b1 expression was also observed in models of insulin resistance [68]. Most recently Cyp7b1 was shown to be associated with adaptive thermogenesis [69]. Furthermore, down-regulation of Cyp7b1 has been proposed to increase oxysterols as a homeostatic response for recovering liver from stress or insult, which could explain the presence of large amount of conjugated 24HC in the plasma of infants with sever intrahepatic cholestasis [70]. The absence of CYP7B1 in the first year of life leads first to a marked elevation in hepatic oxysterols and oxysterol driven activation of LXR with fat accumulation in the liver; a process which is then rapidly followed by inflammation and fibrosis [18]. Whether in delivery states of low cholesterol substrate supply such as in the human CYP7B1 deficiency just outlined or with high substrate supply as with the StarD1 overexpression model used in this study, these findings provide strong evidence it is ultimately CYP7B1 which controls the levels and ratios of the key regulatory oxysterols. These concepts are summarized in Fig. 8.

of labeled cholesterol to rats led to accumulation of 24HC in the liver as well as in the circulation, but failed to accumulate in the brain [40]. A similar observation was found in the present study. The adenoviral StarD1 overexpressing mice developed a markedly elevated liver 24HC level, but it is not incorporated to the brain through circulation (Fig. 1). It should also be noted in a comparative study, a trend was found in the liver 24HC levels in Cyp7b1−/− mice versus their wild type controls [41]. In two separate studies [42,43] increases in liver 24HC levels were found to be associated with dietary influences (2% cholesterol diet; 20% peanut oil diet). In this study, liver oxysterol levels were determined in Western diet fed mice overexpressing StarD1 in the presence and absence of Cyp7b1; conditions attempting to more clearly outline acidic pathway metabolic products. Based upon present in vitro and these in vivo findings, we concluded that mitochondrial CYP27A1 is responsible for producing not only 25HC and 26HC, but 24HC in the mouse. Furthermore, HepG2 cell culture was also shown to have the ability to metabolize [3H]-cholesterol to [3H]-24HC. Therefore, it is likely human liver is also able to metabolize cholesterol to 24HC. Of note, trace amounts of 25HC were detected in the Cyp27a1−/− mice liver (Fig. 1I). This oxysterol was likely derived from microsomal cholesterol 25-hydroxylase [44] and/or CYP3A [45]. The suggested metabolic pathways of these oxysterols to bile acids are summarized in Fig. 7. In human hepatocytes, brain derived 24HC, is reported to be metabolized to two primary bile acids, CDCA and CA [12]. We confirmed this bioconversion in the primary mouse hepatocytes as well (Fig. 6A). In the mouse liver, 24HC was also metabolized to murideoxycholic acid (MDCA) and, subsequently, to hyodeoxycholic acid (HDCA). These two 7-dehydroxylated bile acids were likely derived from 24(S),25-dihydroxycholesterol (24(S),25-diHC) which was a C-25 oxidation product of 24HC by additional Cyp27a1. It is known that 24(S),25-diHC can be metabolized to potentially hepatotoxic 3βhydroxy-5-cholenoic acid and lithocholic acid (LCA) in the absence of 7α-hydroxylation [2,33]. In mice liver, LCA can be hydroxylated at C-6 position to form MDCA and HDCA. In this context, in the Cyp7b1−/− mice hepatocytes, 24HC was converted mostly to MDCA and HDCA and little was converted to 7α-hydroxylated bile acids (ie βMCA, CDCA and CA) (Fig. 6B). Although Cyp39a1 and Cyp7a1 expression levels in these hepatocytes are not known, these results suggested Cyp7b1 also has 7αhydroxylation activity towards 24HC at least in the mouse. A fact that in vivo Cyp7b1−/− mouse liver accumulated not only 25HC and 26HC, but also 24HC, could be explained by Cyp7b1′s 7α-hydroxylation activity towards 24HC. A significantly higher serum 24HC level in the patients with CYP7B1 deficiency [18,19] also supports the present results. We have confirmed conversion of [3H]-24HC to [3H]-7α,24HC by the direct CYP7B1 enzymatic assay (Supplemental Figure S2). The inability of CA to be made in Cyp7b1−/− primary hepatocyte cultures is an evidence that once formed by Cyp27a1, Cyp7b1 activity is needed for 24HC to ultimately be converted to CA. These findings also suggest that under the culture conditions employed, isolated primary hepatocytes lack Cyp7a1 and Cyp39a1 activity. Albeit small as compared to Cyp39a1 and there being a clear preference for 25HC and 26HC, Li-Hawkins et al provided suggestive evidence for 7α-hydroxylation of 24HC in CHOP cells transfected with murine Cyp7b1 cDNAs [15]. Like 7α-hydroxylated 25HC and 26HC, the 7α-hydroxylated 24HC can be converted to corresponding 4-cholesten-3-one form by HSD3B7 (see Fig. 7B). However, for CA generation, subsequent 12α-hydroxylation is essential. Although, in the neutral pathway, Cyp8b1 catalyzes 12α-hydroxylation of 7α-hydroxy-4-cholesten-3-one, it is uncertain whether Cyp8b1 catalyzes for 7α,24(S)-dihydroxy-4-cholesten-3-one. To the best of our knowledge, Cyp8b1 does not catalyze side chain oxidized sterols, 7α,25-dihydroxy-4-cholesten-3-one and (25R)-7α,26dihydroxyly-4-cholesten-3-one [46,47]. In this regard, neither [3H]-

45

Journal of Steroid Biochemistry and Molecular Biology 189 (2019) 36–47

G. Kakiyama et al.

1191–1212. [4] X. Li, W.M. Pandak, S.K. Erickson, Y. Ma, L. Yin, P. Hylemon, S. Ren, Biosynthesis of the regulatory oxysterol, 5-cholesten-3beta,25-diol 3-sulfate, in hepatocytes, J. Lipid Res. 48 (2007) 2587–2596. [5] J.J. Cali, D.W. Russell, Characterization of human sterol 27-hydroxylase. A mitochondrial cytochrome P-450 that catalyzes multiple oxidation reaction in bile acid biosynthesis, J. Biol. Chem. 266 (1991) 7774–7778. [6] I. Bjorkhem, Cerebrotendinous xanthomatosis, Curr. Opin. Lipidol. 24 (2013) 283–287. [7] N.B. Javitt, 25R,26-Hydroxycholesterol revisited: synthesis, metabolism, and biologic roles, J. Lipid Res. 43 (2002) 665–670. [8] M. Schwarz, E.G. Lund, R. Lathe, I. Bjorkhem, D.W. Russell, Identification and characterization of a mouse oxysterol 7alpha-hydroxylase cDNA, J. Biol. Chem. 272 (1997) 23995–24001. [9] W.M. Pandak, S. Ren, D. Marques, E. Hall, K. Redford, D. Mallonee, P. Bohdan, D. Heuman, G. Gil, P. Hylemon, Transport of cholesterol into mitochondria is ratelimiting for bile acid synthesis via the alternative pathway in primary rat hepatocytes, J. Biol. Chem. 277 (2002) 48158–48164. [10] S. Ren, P.B. Hylemon, D. Marques, E. Gurley, P. Bodhan, E. Hall, K. Redford, G. Gil, W.M. Pandak, Overexpression of cholesterol transporter StAR increases in vivo rates of bile acid synthesis in the rat and mouse, Hepatology 40 (2004) 910–917. [11] Z. Wu, J.Y. Chiang, Transcriptional regulation of human oxysterol 7 alpha-hydroxylase gene (CYP7B1) by Sp1, Gene 272 (2001) 191–197. [12] I. Bjorkhem, U. Andersson, E. Ellis, G. Alvelius, L. Ellegard, U. Diczfalusy, J. Sjovall, C. Einarsson, From brain to bile. Evidence that conjugation and omega-hydroxylation are important for elimination of 24S-hydroxycholesterol (cerebrosterol) in humans, J. Biol. Chem. 276 (2001) 37004–37010. [13] I. Bjorkhem, Five decades with oxysterols, Biochimie 95 (2013) 448–454. [14] M. Norlin, A. Toll, I. Bjorkhem, K. Wikvall, 24-hydroxycholesterol is a substrate for hepatic cholesterol 7alpha-hydroxylase (CYP7A), J. Lipid Res. 41 (2000) 1629–1639. [15] J. Li-Hawkins, E.G. Lund, A.D. Bronson, D.W. Russell, Expression cloning of an oxysterol 7alpha-hydroxylase selective for 24-hydroxycholesterol, J. Biol. Chem. 275 (2000) 16543–16549. [16] R. Arnon, T. Yoshimura, A. Reiss, K. Budai, J.H. Lefkowitch, N.B. Javitt, Cholesterol 7-hydroxylase knockout mouse: a model for monohydroxy bile acid-related neonatal cholestasis, Gastroenterology 115 (1998) 1223–1228. [17] N.B. Javitt, 26-Hydroxycholesterol: synthesis, metabolism, and biologic activities, J. Lipid Res. 31 (1990) 1527–1533. [18] K.D. Setchell, M. Schwarz, N.C. O’Connell, E.G. Lund, D.L. Davis, R. Lathe, H.R. Thompson, R. Weslie Tyson, R.J. Sokol, D.W. Russell, Identification of a new inborn error in bile acid synthesis: mutation of the oxysterol 7alpha-hydroxylase gene causes severe neonatal liver disease, J. Clin. Invest. 102 (1998) 1690–1703. [19] D. Dai, P.B. Mills, E. Footitt, P. Gissen, P. McClean, J. Stahlschmidt, I. Coupry, J. Lavie, F. Mochel, C. Goizet, T. Mizuochi, A. Kimura, H. Nittono, K. Schwarz, P.J. Crick, Y. Wang, W.J. Griffiths, P.T. Clayton, Liver disease in infancy caused by oxysterol 7 alpha-hydroxylase deficiency: successful treatment with chenodeoxycholic acid, J. Inherit. Metab. Dis. 37 (2014) 851–861. [20] I. Ueki, A. Kimura, A. Nishiyori, H.L. Chen, H. Takei, H. Nittono, T. Kurosawa, Neonatal cholestatic liver disease in an Asian patient with a homozygous mutation in the oxysterol 7alpha-hydroxylase gene, J. Pediatr. Gastroenterol. Nutr. 46 (2008) 465–469. [21] T. Mizuochi, A. Kimura, M. Suzuki, I. Ueki, H. Takei, H. Nittono, T. Kakiuchi, T. Shigeta, S. Sakamoto, A. Fukuda, A. Nakazawa, T. Shimizu, T. Kurosawa, M. Kasahara, Successful heterozygous living donor liver transplantation for an oxysterol 7alpha-hydroxylase deficiency in a Japanese patient, Liver Transpl. 17 (2011) 1059–1065. [22] G. Salen, S.M. Grundy, The metabolism of cholestanol, cholesterol, and bile acids in cerebrotendinous xanthomatosis, J. Clin. Invest. 52 (1973) 2822–2835. [23] A.L. Esterman, H. Baum, N.B. Javitt, G.J. Darlington, 26-hydroxycholesterol: regulation of hydroxymethylglutaryl-CoA reductase activity in Chinese hamster ovary cell culture, J. Lipid Res. 24 (1983) 1304–1309. [24] Y. Ning, L. Xu, S. Ren, W.M. Pandak, S. Chen, L. Yin, StAR overexpression decreases serum and tissue lipids in apolipoprotein E-deficient mice, Lipids 44 (2009) 511–519. [25] D. Rodriguez-Agudo, S. Ren, P.B. Hylemon, K. Redford, R. Natarajan, A. Del Castillo, G. Gil, W.M. Pandak, Human StarD5, A cytosolic StAR-related lipid binding protein, J. Lipid Res. 46 (2005) 1615–1623. [26] G. Kakiyama, A. Muto, H. Takei, H. Nittono, T. Murai, T. Kurosawa, A.F. Hofmann, W.M. Pandak, J.S. Bajaj, A simple and accurate HPLC method for fecal bile acid profile in healthy and cirrhotic subjects: validation by GC-MS and LC-MS, J. Lipid Res. 55 (2014) 978–990. [27] L. Aringer, P. Eneroth, L. Nordstrom, Side chain hydroxylation of cholesterol, campesterol and beta-sitosterol in rat liver mitochondria, J. Lipid Res. 17 (1976) 263–272. [28] W.M. Pandak, P. Bohdan, C. Franklund, D.H. Mallonee, G. Eggertsen, I. Bjorkhem, G. Gil, Z.R. Vlahcevic, P.B. Hylemon, Expression of sterol 12alpha-hydroxylase alters bile acid pool composition in primary rat hepatocytes and in vivo, Gastroenterology 120 (2001) 1801–1809. [29] Z. Zhang, D. Li, D.E. Blanchard, S.R. Lear, S.K. Erickson, T.A. Spencer, Key regulatory oxysterols in liver: analysis as delta4-3-ketone derivatives by HPLC and response to physiological perturbations, J. Lipid Res. 42 (2001) 649–658. [30] R.T. Stravitz, Z.R. Vlahcevic, T.L. Russell, M.L. Heizer, N.G. Avadhani, P.B. Hylemon, Regulation of sterol 27-hydroxylase and an alternative pathway of bile acid biosynthesis in primary cultures of rat hepatocytes, J. Steroid Biochem. Mol. Biol. 57 (1996) 337–347.

Fig. 8. Formation and control of intracellular oxysterol levels. The acute downregulation of CYP7B1 increases oxysterols which activates LXR. LXR, in a feed forward manner, facilitates expression of StarD1; increasing cholesterol transport into mitochondria. 24HC can be metabolized like 25HC and 26HC to bile acids or independent of CYP7B1. *The role and reason for down-regulation of CYP7B1 have not been clearly defined. Several studies have observed that bile acids (BA) [8,52], FXR [55], LXR [58] and diabetes mellitus (DM) [68] could be potential down-regulators.

In summary, the current study reports the discovery of a new pathway of bile acid synthesis initiated in liver mitochondria. The pathway begins with formation of 24(S)-hydroxycholesterol by CYP27A1. The results of this study also demonstrate CYP7B1 to be capable of being highly regulated and the major enzyme ultimately controlling cellular levels of vital regulatory oxysterols. These findings bring attention and relevance to a pathway whose physiologic importance is only beginning to be uncovered1 . Competing financial interest The authors declare no competing financial interest. Acknowledgements The authors acknowledge excellent technical help from Ms. Emily Gurley, Ms. Xuan Wang and Ms. Patsy Cooper. This work was supported by Gilead Sciences Liver Research Award 2016 and Virginia Commonwealth University DOIM Pilot Project Grant (4111393) to GK; and Veterans Administration VA Merit Award I01 BX000197-07 to WP. Appendix A. Supplementary data Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.jsbmb.2019.01.011. References [1] H. Zhou, P.B. Hylemon, Bile acids are nutrient signaling hormones, Steroids 86 (2014) 62–68. [2] J.E. Heubi, K.D. Setchell, K.E. Bove, Inborn errors of bile acid metabolism, Semin. Liver Dis. 27 (2007) 282–294. [3] J.Y. Chiang, Bile acid metabolism and signaling, Compr. Physiol. 3 (2013)

1 (25R)-26-Hydroxycholesterol often referred to as 27-hydroxycholesterol (27HC). However, here we preferred to use formal name of this compound, (25R)-26-hydroxycholestetol (26HC) according to the guidance by Fakheri and Javitt [71].

46

Journal of Steroid Biochemistry and Molecular Biology 189 (2019) 36–47

G. Kakiyama et al.

[52] S. Ren, D. Marques, K. Redford, P.B. Hylemon, G. Gil, Z.R. Vlahcevic, W.M. Pandak, Regulation of oxysterol 7alpha-hydroxylase (CYP7B1) in the rat, Metabolism 52 (2003) 636–642. [53] W.M. Pandak, P.B. Hylemon, S. Ren, D. Marques, G. Gil, K. Redford, D. Mallonee, Z.R. Vlahcevic, Regulation of oxysterol 7alpha-hydroxylase (CYP7B1) in primary cultures of rat hepatocytes, Hepatology 35 (2002) 1400–1408. [54] Y. Inoue, A.M. Yu, S.H. Yim, X. Ma, K.W. Krausz, J. Inoue, C.C. Xiang, M.J. Brownstein, G. Eggertsen, I. Bjorkhem, F.J. Gonzalez, Regulation of bile acid biosynthesis by hepatocyte nuclear factor 4alpha, J. Lipid Res. 47 (2006) 215–227. [55] T.Q. de Aguiar Vallim, E.J. Tarling, H. Ahn, L.R. Hagey, C.E. Romanoski, R.G. Lee, M.J. Graham, H. Motohashi, M. Yamamoto, P.A. Edwards, MAFG is a transcriptional repressor of bile acid synthesis and metabolism, Cell Metab. 21 (2015) 298–311. [56] Z. Wu, K.O. Martin, N.B. Javitt, J.Y. Chiang, Structure and functions of human oxysterol 7alpha-hydroxylase cDNAs and gene CYP7B1, J. Lipid Res. 40 (1999) 2195–2203. [57] Y. Yamamoto, R. Moore, H.A. Hess, G.L. Guo, F.J. Gonzalez, K.S. Korach, R.R. Maronpot, M. Negishi, Estrogen receptor alpha mediates 17alpha-ethynylestradiol causing hepatotoxicity, J. Biol. Chem. 281 (2006) 16625–16631. [58] H. Uppal, S.P. Saini, A. Moschetta, Y. Mu, J. Zhou, H. Gong, Y. Zhai, S. Ren, G.K. Michalopoulos, D.J. Mangelsdorf, W. Xie, Activation of LXRs prevents bile acid toxicity and cholestasis in female mice, Hepatology 45 (2007) 422–432. [59] K.A. Rose, G. Stapleton, K. Dott, M.P. Kieny, R. Best, M. Schwarz, D.W. Russell, I. Bjorkhem, J. Seckl, R. Lathe, Cyp7b, a novel brain cytochrome P450, catalyzes the synthesis of neurosteroids 7alpha-hydroxy dehydroepiandrosterone and 7alphahydroxy pregnenolone, Proc. Natl. Acad. Sci. U. S. A. 94 (1997) 4925–4930. [60] K. Rose, A. Allan, S. Gauldie, G. Stapleton, L. Dobbie, K. Dott, C. Martin, L. Wang, E. Hedlund, J.R. Seckl, J.A. Gustafsson, R. Lathe, Neurosteroid hydroxylase CYP7B: vivid reporter activity in dentate gyrus of gene-targeted mice and abolition of a widespread pathway of steroid and oxysterol hydroxylation, J. Biol. Chem. 276 (2001) 23937–23944. [61] C. Martin, M. Ross, K.E. Chapman, R. Andrew, P. Bollina, J.R. Seckl, F.K. Habib, CYP7B generates a selective estrogen receptor beta agonist in human prostate, J. Clin. Endocrinol. Metab. 89 (2004) 2928–2935. [62] H. Pettersson, L. Holmberg, M. Axelson, M. Norlin, CYP7B1-mediated metabolism of dehydroepiandrosterone and 5alpha-androstane-3beta,17beta-diol–potential role (s) for estrogen signaling, FEBS J. 275 (2008) 1778–1789. [63] Z. Weihua, R. Lathe, M. Warner, J.A. Gustafsson, An endocrine pathway in the prostate, ERbeta, AR, 5alpha-androstane-3beta,17beta-diol, and CYP7B1, regulates prostate growth, Proc. Natl. Acad. Sci. U. S. A. 99 (2002) 13589–13594. [64] Y. Omoto, R. Lathe, M. Warner, J.A. Gustafsson, Early onset of puberty and early ovarian failure in CYP7B1 knockout mice, Proc. Natl. Acad. Sci. U. S. A. 102 (2005) 2814–2819. [65] E.R. Nelson, S.E. Wardell, J.S. Jasper, S. Park, S. Suchindran, M.K. Howe, N.J. Carver, R.V. Pillai, P.M. Sullivan, V. Sondhi, M. Umetani, J. Geradts, D.P. McDonnell, 27-Hydroxycholesterol links hypercholesterolemia and breast cancer pathophysiology, Science 342 (2013) 1094–1098. [66] U. Diczfalusy, K.E. Olofsson, A.M. Carlsson, M. Gong, D.T. Golenbock, O. Rooyackers, U. Flaring, H. Bjorkbacka, Marked upregulation of cholesterol 25hydroxylase expression by lipopolysaccharide, J. Lipid Res. 50 (2009) 2258–2264. [67] D.R. Bauman, A.D. Bitmansour, J.G. McDonald, B.M. Thompson, G. Liang, D.W. Russell, 25-Hydroxycholesterol secreted by macrophages in response to Tolllike receptor activation suppresses immunoglobulin A production, Proc. Natl. Acad. Sci. U. S. A. 106 (2009) 16764–16769. [68] S.B. Biddinger, J.T. Haas, B.B. Yu, O. Bezy, E. Jing, W. Zhang, T.G. Unterman, M.C. Carey, C.R. Kahn, Hepatic insulin resistance directly promotes formation of cholesterol gallstones, Nat. Med. 14 (2008) 778–782. [69] B.M. Worthmann, H.C. Song, D.R. Dowling, Adaptive frequency-difference matched field processing for high frequency source localization in a noisy shallow ocean, J. Acoust. Soc. Am. 141 (2017) 543. [70] L.J. Meng, W.J. Griffiths, H. Nazer, Y. Yang, J. Sjovall, High levels of (24S)-24hydroxycholesterol 3-sulfate, 24-glucuronide in the serum and urine of children with severe cholestatic liver disease, J. Lipid Res. 38 (1997) 926–934. [71] R.J. Fakheri, N.B. Javitt, 27-Hydroxycholesterol, does it exist? On the nomenclature and stereochemistry of 26-hydroxylated sterols, Steroids 77 (2012) 575–577.

[31] S. Ren, P. Hylemon, Z.P. Zhang, D. Rodriguez-Agudo, D. Marques, X. Li, H. Zhou, G. Gil, W.M. Pandak, Identification of a novel sulfonated oxysterol, 5-cholesten3beta,25-diol 3-sulfonate, in hepatocyte nuclei and mitochondria, J. Lipid Res. 47 (2006) 1081–1090. [32] B. Petrack, B.J. Latario, Synthesis of 27-hydroxycholesterol in rat liver mitochondria: HPLC assay and marked activation by exogenous cholesterol, J. Lipid Res. 34 (1993) 643–649. [33] I. Bjorkhem, Mechanism of degradation of the steroid side chain in the formation of bile acids, J. Lipid Res. 33 (1992) 455–471. [34] Y.D. Guo, S. Strugnell, D.W. Back, G. Jones, Transfected human liver cytochrome P450 hydroxylates vitamin D analogs at different side-chain positions, Proc. Natl. Acad. Sci. U. S. A. 90 (1993) 8668–8672. [35] S. Strugnell, V. Byford, H.L. Makin, R.M. Moriarty, R. Gilardi, L.W. LeVan, J.C. Knutson, C.W. Bishop, G. Jones, 1 alpha,24(S)-dihydroxyvitamin D2: a biologically active product of 1 alpha-hydroxyvitamin D2 made in the human hepatoma, Hep3B, Biochem J 310 (Pt 1) (1995) 233–241. [36] G. Jones, D.E. Prosser, M. Kaufmann, Cytochrome P450-mediated metabolism of vitamin d, J. Lipid Res. 55 (2014) 13–31. [37] V. Byford, S. Strugnell, R. Coldwell, N. Schroeder, H.L. Makin, J.C. Knutson, C.W. Bishop, G. Jones, Use of vitamin D(4) analogs to investigate differences in hepatic and target cell metabolism of vitamins D(2) and D(3), Biochim. Biophys. Acta 1583 (2002) 151–166. [38] E. Lund, I. Bjorkhem, C. Furster, K. Wikvall, 24-, 25- and 27-hydroxylation of cholesterol by a purified preparation of 27-hydroxylase from pig liver, Biochim. Biophys. Acta 1166 (1993) 177–182. [39] S. Meaney, M. Hassan, A. Sakinis, D. Lutjohann, K. von Bergmann, A. Wennmalm, U. Diczfalusy, I. Bjorkhem, Evidence that the major oxysterols in human circulation originate from distinct pools of cholesterol: a stable isotope study, J. Lipid Res. 42 (2001) 70–78. [40] S. Meaney, D. Lutjohann, U. Diczfalusy, I. Bjorkhem, Formation of oxysterols from different pools of cholesterol as studied by stable isotope technique: cerebral origin of most circulating 24S-hydroxycholesterol in rats, but not in mice, Biochim. Biophys. Acta 1486 (2000) 293–298. [41] J. Li-Hawkins, E.G. Lund, S.D. Turley, D.W. Russell, Disruption of the oxysterol 7alpha-hydroxylase gene in mice, J. Biol. Chem. 275 (2000) 16536–16542. [42] E. Lund, O. Breuer, I. Bjorkhem, Evidence that 24- and 27-hydroxylation are not involved in the cholesterol-induced down-regulation of hydroxymethylglutaryl-CoA reductase in mouse liver, J. Biol. Chem. 267 (1992) 25092–25097. [43] S.E. Saucier, A.A. Kandutsch, A.K. Gayen, D.K. Swahn, T.A. Spencer, Oxysterol regulators of 3-hydroxy-3-methylglutaryl-CoA reductase in liver. Effect of dietary cholesterol, J. Biol. Chem. 264 (1989) 6863–6869. [44] E.G. Lund, T.A. Kerr, J. Sakai, W.P. Li, D.W. Russell, cDNA cloning of mouse and human cholesterol 25-hydroxylases, polytopic membrane proteins that synthesize a potent oxysterol regulator of lipid metabolism, J. Biol. Chem. 273 (1998) 34316–34327. [45] A. Honda, T. Miyazaki, T. Ikegami, J. Iwamoto, T. Maeda, T. Hirayama, Y. Saito, T. Teramoto, Y. Matsuzaki, Cholesterol 25-hydroxylation activity of CYP3A, J. Lipid Res. 52 (2011) 1509–1516. [46] W.J. Griffiths, J. Abdel-Khalik, P.J. Crick, E. Yutuc, Y. Wang, New methods for analysis of oxysterols and related compounds by LC-MS, J. Steroid Biochem. Mol. Biol. 162 (2016) 4–26. [47] A. Honda, T. Yoshida, G. Xu, Y. Matsuzaki, S. Fukushima, N. Tanaka, M. Doy, S. Shefer, G. Salen, Significance of plasma 7alpha-hydroxy-4-cholesten-3-one and 27-hydroxycholesterol concentrations as markers for hepatic bile acid synthesis in cholesterol-fed rabbits, Metabolism 53 (2004) 42–48. [48] M.S. Brown, A. Radhakrishnan, J.L. Goldstein, Retrospective on cholesterol homeostasis: the central role of scap, Annu. Rev. Biochem. 87 (2018) 783–807. [49] D.J. Peet, B.A. Janowski, D.J. Mangelsdorf, The LXRs: a new class of oxysterol receptors, Curr. Opin. Genet. Dev. 8 (1998) 571–575. [50] J.J. Repa, G. Liang, J. Ou, Y. Bashmakov, J.M. Lobaccaro, I. Shimomura, B. Shan, M.S. Brown, J.L. Goldstein, D.J. Mangelsdorf, Regulation of mouse sterol regulatory element-binding protein-1c gene (SREBP-1c) by oxysterol receptors, LXRalpha and LXRbeta, Genes Dev. 14 (2000) 2819–2830. [51] B.A. Janowski, M.J. Grogan, S.A. Jones, G.B. Wisely, S.A. Kliewer, E.J. Corey, D.J. Mangelsdorf, Structural requirements of ligands for the oxysterol liver X receptors LXRalpha and LXRbeta, Proc. Natl. Acad. Sci. U. S. A. 96 (1999) 266–271.

47