Available online at www.sciencedirect.com

ScienceDirect

Journal of Nutritional Biochemistry 39 (2017) 59–67

Deletion of liver-specific STAT5 gene alters the expression of bile acid metabolismgenes and reduces liver damage in lithogenic diet-fed mice☆

Myunggi Baika,b,⁎, 1, Jangseon Kimc,1, Min Yu Piaoa, Hyeok Joong Kanga, Seung Ju Parka, Sang Weon Naa,Sung-Hoon Ahnd, Jae-Hyuk Leee

aDepartment of Agricultural Biotechnology and Research Institute of Agriculture and Life Sciences, College of Agriculture and Life Sciences, Seoul National University, 1 Gwanak-ro, Gwanak-gu,Seoul 151-921, Republic of Korea

bInstitute of Green Bio Science Technology, Gangwon-do, Pyeungchang-gun 232-916, Republic of KoreacBioneer Corporation, 8-11 Munpyengseoro, Daedeok-gu, Daejeon 306-220, Republic of Korea

dCollege of Pharmacy, Kangwon National University, Chuncheo, Republic of KoreaeDepartment of Pathology, Chonnam National University Medical School, Gwangju 501-746, Republic of Korea

Received 17 February 2016; received in revised form 6 September 2016; accepted 6 September 2016

Abstract

Signal transducers and activators of transcription 5 (STAT5) mediates growth hormone signals, which may control hepatic cholesterol uptake and bile acidmetabolism. Deregulation of liver cholesterol homeostasis and bile acid metabolism may cause liver damage and cholesterol gallstone development. The purposeof this study was to understand the role of local STAT5 signaling in cholesterol and bile acid metabolism using liver-specific STAT5 knock-out (STAT5 LKO) miceon a normal diet and a cholesterol- and bile acid-containing lithogenic diet. STAT5 LKO mice showed significant down-regulation of STAT5 and insulin-likegrowth factor-1 genes. STAT5 gene deletion had a minor effect on cholesterol metabolism, as evidenced by a minor change in circulating cholesterol levels and nochanges in expression of hepatic low-density lipoprotein receptor and cholesterol synthesis genes in STAT5 LKO mice. In contrast, bile acid synthesis and uptakegenes were profoundly down-regulated and bile acid detoxification genes were up-regulated in STAT5 LKO mice. In STAT5 fl/fl mice, a lithogenic diet inducedliver damage, as evidenced by moderate increases in liver ballooning, inflammation and fibrosis. However, STAT5 deletion ameliorated the degree of liverdamage induced by the lithogenic diet. In STAT5 LKO mice, a lithogenic diet did not alter the incidence or severity of cholesterol gallstones. In conclusion, localSTAT5 signaling does not have a significant role in cholesterol metabolism. In contrast, hepatic STAT5 signaling has significant roles in regulating transcription ofgenes for synthesis, transport and detoxification of bile acids, but it has only a minor role in bile acid metabolism.© 2016 Elsevier Inc. All rights reserved.

Keywords: Liver damage; Growth hormone; STAT5; Lithogenic diet; Bile acids; Gallstones

Abbreviations: STAT5, signal transducer and activator of transcription 5; GH, growth hormone; IGF-1, insulin-like growth factor-1; LDLR, low-density lipoproteinreceptor; STAT5 LKO, liver-specific STAT5 knock-out; STAT5 fl/fl, STAT5 flox/flox; TG, triacylglycerol; ChMCs, cholesterol monohydrate crystals; H&E, hematoxylinand eosin; ORO, Oil-Red O; CA, cholic acid; CDCA, chenodeoxycholic acid; DCA, deoxycholic acid; TCA, taurocholic acid; TCDCA, tauro chenodeoxycholic acid; TDCA,taurodeoxycholic acid; GCA, glycocholic acid; GCDCA, glycochenodeoxycholic acid;αMCA,α-muricholic acid; βMCA, β-muricholic acid; γMCA, γ-muricholic acid;TβMCA, tauro-β-muricholic acid; PCR, polymerase chain reaction; ωMCA,ω-muricholic acid; TαMCA, tauro-α-muricholic acid; Lrp1, lipoprotein receptor-relatedprotein 1; Scarb1, scavenger receptor class B, member 1; Hmgcr, 3-hydroxy-3-methylglutaryl-Coenzyme A reductase; Slco10a1, solute carrier family 10 (sodium/bile acid co-transporter family), member 1; Slco1a1, solute carrier organic anion transporter family, member 1a1; Slco1a4, solute carrier organic anion transporterfamily,member 1a4; Slco1b2, solute carrier organic anion transporter family,member 1b2; FXR, farnesoid X receptor; Cyp7a1, cytochrome P450, family 7, subfamilya, polypeptide 1; Hsd3b5, hydroxy-delta-5-steroid dehydrogenase, 3 beta- and steroid delta-isomerase 5; Cyp8b1, cytochrome P450, family 8, subfamily b,polypeptide 1; Cyp7b1, cytochrome P450, family 7, subfamily b, polypeptide 1; Cyp2b10, cytochrome P450, family 2, subfamily b, polypeptide 10; Cyp2b13,cytochrome P450, family 2, subfamily b, polypeptide 13; Sult2a1, sulfotransferase family 2 A, dehydroepiandrosterone (DHEA)-preferring, member 1; Abcb11, ATP-binding cassette, sub-family B, member 11; Abcg5, ATP-binding cassette, sub-family G (WHITE), member 5; Abcg8, ATP-binding cassette, sub-family G (WHITE),member 8; Abcc2, ATP-binding cassette, sub-family C, member 2 (mrp2); Abcb4, ATP-binding cassette, sub-family B, member 4 (mdr2).

☆ This work was supported by grants fromWCU Project (R33-10059), the National Research Laboratory Program (ROA-2007-0056702) through the NRF fundedby the Ministry of Education, Science and Technology and the Next-Generation BioGreen 21 Program (No. PJ01114001), Rural Development Administration,Republic of Korea.⁎ Corresponding author at: Department of Agricultural Biotechnology, Seoul National University, 1 Gwanak-ro, Gwanak-gu, Seoul 151-921, Republic of Korea.

Tel.: +82 2 880 4809; fax: +82 2 873 2271.E-mail addresses: mgbaik@snu.ac.kr, mgbaik2164@naver.com (M. Baik).

1 Equal contribution.

http://dx.doi.org/10.1016/j.jnutbio.2016.09.0120955-2863/© 2016 Elsevier Inc. All rights reserved.

60 M. Baik et al. / Journal of Nutritional Biochemistry 39 (2017) 59–67

1. Introduction

Signal transducers and activators of transcription 5 (STAT5)mediates growth hormone (GH) signaling [1,2] and activatesinsulin-like growth factor-1 (IGF-1) gene transcription. The GH–STAT5–IGF-1 axis regulates body growth and lipid and glucosemetabolism in the liver [3,4]. The liver is a major organ that regulatescholesterol and bile acid metabolism. Several studies have shown thatGH has important functions related to the regulation cholesteroluptake and bile acid metabolism. For example, in humans, GHtreatment increased levels of hepatic low-density lipoprotein recep-tors (LDLR) and reduced levels of plasma LDL cholesterol [5].Hypophysectomized LDLR knock-out mice had increased LDL choles-terol levels that were attenuated by GH treatment [6]. Studies haveshown that GH–STAT5 signaling regulates the expression of genesinvolved in the uptake, synthesis, detoxification and excretion of bileacids in the liver [7,8].

Clinical studies have reported controversial results on the effect ofGH on cholesterol crystal and gallstone development. For example, anincreased prevalence of gallstone development was found in patientswith acromegaly, a hormonal disorder that results from excess GHproduction from the anterior pituitary gland [9]. In contrast, otherclinical studies found that GH administration to healthy subjects orGH-deficient children did not alter biliary lipid composition orincrease the risk of cholesterol gallstone development [5,10].Therefore, studies are needed to clarify the effect of GH on cholesteroland bile acid metabolism and cholesterol crystal and gallstonedevelopment. Little information is available for direct link betweenGH and STAT5 signaling for regulation of cholesterol and bile acidmetabolism.

A lithogenic diet, which consists of high cholesterol and 0.5% cholicacid (CA), induces cholesterol crystallization and gallstone formationin gallbladder bile from inbredmice, such as C57L [11]. However, littleis known about how STAT5 gene deletion in lithogenic diet-fed miceaffects cholesterol and bile acid metabolism, liver function andcholesterol gallstone development.

The purpose of this study was to understand local STAT5 signalingin hepatic cholesterol and bile acid metabolism using liver-specificSTAT5 knock-out (STAT5 LKO)mice. To this end, STAT5 LKOmicewerefed either a normal diet or a lithogenic diet to amplify possible changesobserved with the normal diet. We hypothesized that STAT5 genedeletion would deregulate the expression of cholesterol and bile acidmetabolism genes, affecting liver damage and cholesterol crystalformation in the gallbladder of STAT5 LKO mice.

2. Materials and methods

2.1. Mice and diets

All experimental procedures involving mice were approved by the ChonnamNational University Institutional Animal Use and Care Committee. STAT5 flox/flox (fl/fl)albumin-Cre mice and STAT5 fl/fl mice were kindly provided by Dr. LotharHennighausen at the National Institutes of Health (Bethesda, MD, USA). STAT5 LKOmice were generated by breeding STAT5 fl/fl albumin-Cre mice with STAT5 fl/fl mice.Mice genotyping was done as previously described [4].

Mice were grown as previously described [4]. Mice were individually housed incarbonate cages and maintained under pathogen-free conditions at Chonnam NationalUniversity in a temperature- and humidity-controlled room on a 12-h light/12-h darkcycle, with access to conventional rodent chow (control diet) ad libitum. At 15 weeks ofage, male STAT5 LKO and STAT5 fl/flmice were randomly assigned to either the controldiet group or the lithogenic diet group and fed for 12 weeks ad libitum. The normal dietwas low fat and low cholesterol (Research Diets, Inc., New Brunswick, NJ, USA; Product#D12337), and the lithogenic diet contained cholesterol and bile acid (Research DietsProduct #D12336). Control and lithogenic diet chow were purchased through theCentral Laboratory of Animals Corporation (Seoul, Republic of Korea). Control andlithogenic chow compositions are shown in Supporting Table S1. During theexperimental period, food intake and body weight were recorded twice per week andonce per week, respectively.

2.2. Collection and analysis of blood and tissues

At the end of the experimental period, bloodwas collected by cardiac puncturewitha 1-ml syringe. The blood was maintained for 30 min at room temperature and thencentrifuged at 1000g for 20 min at 4°C to separate the serum. Animals were killed bydecapitation after being anesthetized with CO2. Tissue samples were weighed, rapidlyfrozen in liquid nitrogen and stored at −80°C.

Phospholipid levels were determined by an enzymatic method using L-Type PL(Wako, Japan). Bile acids levels were determined by an enzymatic method using a TBAkit (Randox, UK). Levels of LDL cholesterol, high-density lipoprotein (HDL) cholesteroland total cholesterol were determined by enzymatic colorimetric assays using LDL-Cplus 2nd generation, HDL-C plus 3rd generation, and CHOL kits (Roche, Germany),respectively. Serum triacylglycerol (TG) levels were determined by enzymaticcolorimetry using a TG kit (Roche, Germany). Serum IGF-1 levels were determined bya Octeia Rat/Mouse IGF-1 kit (IBT, Germany). Serum alkaline phosphatase (ALP) andalanine aminotransferase (ALT) levels were determined by enzymatic colorimetryusing ALP and ALT kits (Roche), respectively. Intra-assay and inter-assay coefficients ofvariation for IGF-1, TG, cholesterol, phospholipid, bile acid, LDL cholesterol, HDLcholesterol, ALP and ALT were 7.3% and 8.8%, 1.5% and 2.4%, 1.0% and 2.7%, 5% and 5%,4.5% and 4.4%, 0.8% and 1.2%, 0.8% and 1.3%, 0.67% and 0.67%, and 1.8% and 3.2%,respectively.

2.3. Gallbladder bile cholesterol crystal and gallstone analysis

Animals were fed a lithogenic diet for 2, 4, 6 or 12 weeks (as indicated in thesupporting figure legend). Gallbladder bile was collected from intact gallbladders afteran overnight fast. One microliter of fresh gallbladder bile was immediately analyzed bypolarizing microscopy (Carl Zeiss Microscopy GmBH, Jena, Germany). A gallstone scorewas assigned using a minor modification of the method of Xie et al. [12] with thefollowing criteria: 0=absence of cholesterol monohydrate crystals (ChMCs), 1=smallnumber of aggregated ChMCs (b3/50× power field of view), 2=many aggregatedChMCs (≥3/50× power field of view), 3=small number of “sandy” light-translucentstones (b3/50× power field of view), and 4=many “sandy” light-translucent stones(≥3/50× power field of view) or presence of “solid” light-opaque stones.

2.4. Liver histology

Liver histology was performed as previously described [4]. Liver specimens werefixed in 10% buffered formalin (Sigma-Aldrich, St. Louis, MO, USA), equilibrated in a70%–100% ethanol-xylene series, and embedded in paraffin. Paraffin sections werestained with hematoxylin and eosin (H&E) and then examined for ballooning andinflammatory cells under a light microscope. Sections were also stained with Masson'strichrome (MT) and evaluated for fibrosis. Degrees of ballooning, inflammation andfibrosis were determined using Kleiner scores [13,14].

Liver specimens fixed in 10% buffered formalin were also equilibrated in 20% sucrosewith gum Arabic from the acacia tree (Sigma-Aldrich, G9752) and embedded usingoptimal cutting temperature compound. Tissueswere sectioned at−20°C into 7-μm-thicksections using a cryostat. Cryosections were mounted by touching the glass slide to thetissue section. To detect fat deposition in the liver, frozen sectionswerefixed for 10min inneutral-buffered 10% formalin and then rinsed with distilled water. Sections were thenstained with 0.5% Oil Red O (ORO; Sigma-Aldrich, O0625) in 60% 2-propanol (Sigma-Aldrich, 190764) for 10 min and then rinsed with distilled water. Stained sections wereexamined and photographed using an Olympus 1X70 light microscope (Olympusinstruments Inc., Tokyo, Japan). The degree of ORO staining was determined by referenceORO staining (from none to the highest grade: 0, 1, 2, 3, 4 and 5).

2.5. Bile acid analysis using liquid chromatography–tandem mass spectrometry

Individual bile acids were analyzed by liquid chromatography–tandem massspectrometry (LC–MS/MS). All chemicals and reagents, such as CA, chenodeoxycholicacid (CDCA), deoxycholic acid (DCA), taurocholic acid (TCA), tauro chenodeoxycholicacid (TCDCA), taurodeoxycholic acid (TDCA), glycocholic acid (GCA) and glycocheno-deoxycholic acid (GCDCA), were purchased from Sigma-Aldrich. α-Muricholic acid(αMCA),β-muricholic acid (βMCA),γ-muricholic acid (γMCA) and tauro-β-muricholicacid (TβMCA) were purchased from Steraloids, Inc. (Newport, RI, USA). Individual bileacids were extracted from liver and gallbladder bile, as previously described [15,16].Liver tissues were weighed and homogenized with three volumes of 50% methanol onice using an IKA homogenizer (Staufen, Germany). The mixture was vortexedvigorously and then centrifuged at 3000g for 10 min. The supernatant was filteredusing a syringe filter (Phenomenex RC membrane 0.2 μm, Torrance, CA, USA). Filteredsupernatants were frozen at −20°C until bile acid analysis. For bile acid analysis, threevolumes of 50% methanol were added to bile for protein precipitation and bile acidextraction. The mixture was vortexed and then centrifuged at 12,000g for 5 min. Thesupernatant was filtered and stored at −20°C until analysis.

Bile acid standard stock solutions were dissolved in dimethyl sulfoxide (10 mM),and working solutions were diluted with acetonitrile at concentrations from 0.01 to 20μM. For the standard solution, 100 μl working solution, 20 μl internal standard (1 μMcilomilast in acetonitrile) and 900 μl of blank-filtered tissue homogenate were mixedand dried under a centrifugal evaporator (Eyela, Tokyo, Japan). The residue was

61M. Baik et al. / Journal of Nutritional Biochemistry 39 (2017) 59–67

reconstituted in 100 μl mobile phase and injected into a LC–MS/MS system. For thetissue and fluid samples, 1000 μl filtered tissue homogenate and 20 μl internal standard(1 μM cilomilast in acetonitrile) were mixed and pretreated with the same methoddescribed for standard samples and injected into a LC–MS/MS system.

The LC–MS/MS system consisted of an Agilent 1200 series HPLC system with CTLPAL autosampler (CTC Analytics, Zwingen, Switzerland) coupled to an API 4000 Qtrapmass spectrometer (AB SCIEX, Foster City, CA, USA). Chromatographic separation wasperformed on a Hypersil Gold C18 column (100 mm×2.1 mm i.d., 3-μm particle size;Thermo, Waltham, MA, USA) with a flow rate of 0.2 ml/min at a column oventemperature of 40°C [17,18]. For bile acid separation, a gradient system of mobile phaseA (10 mM ammonium formate in HPLC-grade water) and mobile phase B (HPLC-gradepure acetonitrile) was used as follows: 80% mobile phase A (0–0.5 min), 80%–50% A(0.5–6.5min), 50%–20% A (6.5–7min), 20% A (7–8.5min), and 80% A (8.5–10min)witha flow rate of 0.3 ml/min. The injected sample volume was 5 μl under a constanttemperature of 4°C. The eluent was introduced directly into the tandem quadrupolemass spectrometer through a nebulizer (GS1), 50 psi; heating (GS2), 50 psi;declustering potential (DP, V) of −150 for CA and αMCA, −160 for βMCA, −165 forγMCA, −170 for CDCA and DCA, −165 for UDCA and HDCA, −180 for TCA, TβMCA,TCDCA, and TDCA,−140 for GCA and GCDCA, and−100 for IS; collision energy (CE, V),−48 for CA andαMCA,−46 for βMCA and γMCA,−44 for CDCA and DCA,−74 for TCA,TβMCA, TCDCA, and TDCA,−48 for GCA, −46 for GCDCA, and−40 for IS, respectively.Molecular ions were fragmented with collision-activated dissociation with nitrogen asthe collision gas. Multiple reaction monitoring mode based on most abundant productions was at m/z 407.3→343.2 for CA, m/z 407.3→387.3 for αMCA, m/z 407.3→371.3 forβMCA, m/z 407.3→389.2 for γMCA, m/z 391.3→373.3 for CDCA, m/z 391.3→345.2 forDCA,m/z 514.3→123.9 for TCA and TβMCA,m/z 498.3→123.9 for TCDCA and TDCA,m/z464.3→402.2 for GCA, m/z 448.3→386.3 for GCDCA, and m/z 342.2→214.0 for IS. Peakareas for all components were automatically integrated using Analyst software (ABSCIEX, version 1.4.2). The standard curve was linear between 0.01 and 20 μM. Theconcentrations of standard calibrations were 0.01, 0.05, 0.1, 0.5, 1, 5, 10 and 20 μM, andthe R2 values for linear calibration ranged from 0.9977 to 0.9999.

2.6. RNA extraction, cDNA synthesis and real-time polymerase chain reaction

RNA was extracted as previously described [4]. Total RNA was isolated from tissuesamples using TRIzol Reagent (Invitrogen, Carlsbad, CA, USA) according to themanufacturer's instructions. RNA quantity and quality were assessed using a NanoDropND-1000 UV–Vis Spectrophotometer (NanoDrop Technologies, Wilmington, DE, USA).Total RNA (3 μg) was reverse-transcribed into cDNA using Accupower RT premix(Bioneer, Daejeon, Republic of Korea).

Real-time polymerase chain reaction (PCR) was performed using QuantiTect SYBRgreen RT-PCR Master Mix (Qiagen, Valencia, CA, USA) and an Opticon sequencedetection system (Rotorgen Qiagen). Briefly, the PCR was performed in a total reactionvolume of 25 μl containing 200 ng cDNA, 12.5 μl SYBR Green RT-PCR Master Mix(Qiagen) and 1.25 μl primers (10 μM) (Bioneer). Thermal cycling parameters were asfollows: 95°C for 15min, followed by 40 cycles at 94°C for 15 s, 60°C for 30 s, and72°C for30 s. All primerswere designed using IntegratedDNA Technology and Primer 3 softwarebased on published sequences in GenBank and synthesized by Bioneer. Primerinformation is shown in Supporting Table S2. The ΔΔCTmethod was used to determinerelative fold changes, and all data were normalized to the housekeeping β-actin gene.

2.7. Statistical analyses

Data are presented as the mean±standard error of the mean (S.E.M.). Data wereanalyzed using the general linear model procedure of SAS (SAS Inst. Inc., Cary, NC, USA).The model included genotype (STAT5 fl/fl or STAT5 LKO), diet (control or lithogenic) andgenotype×diet interaction. Thenumber of animals per groupwasfiveor six formost of thedata. The PDIFF option of LSMEANS in SASwas used to separatemeans. A P value b.05wasconsidered to indicate statistical significance.

3. Results

3.1. Expression of STAT5 signaling genes, body growth and food intake

STAT5 LKO mice showed a 74%–87% reduction (Pb.05) in hepaticmRNA levels of STAT5a/b genes comparedwith STAT5 fl/flmice (Fig. 1).The hepatic mRNA levels of the STAT5 target gene IGF-1 were alsodown-regulated (Pb.05) by 66%–85% in STAT5 LKO mice, regardless ofdiet. Circulating IGF-1 levelswere 52%–79% lower (Pb.05) in STAT5 LKOmice than in STAT5 fl/fl mice.

The average daily food intake over the 12-week feeding periodfrom 15 to 27 weeks did not significantly differ between STAT5 fl/fland STAT5 LKOmice (STAT5 fl/flmice: 3.60 g/day vs. STAT5 LKOmice:3.50 g/day) or according to diet type (mice fed control diet: 3.55 g/dayvs. mice fed lithogenic diet: 3.55 g/day). The average initial body

weight of STAT5 fl/fl mice (29.4 g) at 15 weeks of age did not differfrom that of STAT5 LKO mice (28.3 g). At 27 weeks of age, after 12weeks of feeding, the average body weight did not differ according togenotype (STAT5 fl/fl mice: 30.7 g vs. STAT5 LKO mice: 29.4 g) and diettype (mice fed control diet: 29.3 g vs. mice fed lithogenic diet: 30.5 g).Lithogenic diet-fed mice had higher (Pb.001) liver weight (1.33 g/mousevs. 2.29 g/mouse) than control diet-fed mice. However, the liver weightsin STAT5 LKO and STAT5 fl/flmice did not significantly differ.

3.2. Serum and liver lipid metabolites

Lipid metabolite levels were determined in serum and liver tissuefrom STAT5 LKO and STAT5 fl/fl mice. STAT5 LKO mice had increased(Pb.05) blood phospholipid and TG levelswhen data from the two dietgroups were combined, but diet type did not affect either parameter(Table 1). STAT5 LKO mice had increased (Pb.05) serum HDLcholesterol levels in the control but not in the lithogenic diet group.Compared with STAT5 fl/fl mice, STAT5 LKO mice did not havesignificantly altered levels of other lipid metabolites, including bileacid, total cholesterol and LDL cholesterol, regardless of diet type,although levels were numerically increased for all parameters. Thelithogenic diet markedly increased (Pb.05) serum levels of bile acid,total cholesterol and LDL cholesterol in both STAT5 LKO and STAT5 fl/flmice.

In the liver, cholesterol, phospholipid and TG levels were notaffected in STAT5 LKO mice, regardless of diet type (Table 1). Thelithogenic diet profoundly increased (Pb.05) cholesterol concentrationin the liver. Gallbladder bile volume in the lithogenic diet group wasincreased (Pb.05) in STAT5 LKO mice compared with STAT5 fl/flmice,but this was not the case in the control diet group.

Using LC–MS/MS, the concentrations of individual bile acids fromserum, liver and gallbladder bile were determined. Serum concentra-tions of all bile acids examined were not changed by STAT LKO,regardless of diet type (Fig. 2A; Table S3). The lithogenic diet increasedserum concentrations of CDCA, DCA, TDCA and GCA.

Increased (Pb.05) hepatic concentrations of CA and GCA werefound in STAT5 LKOmice in the lithogenic diet group but not in controldiet group (Fig. 2B). Decreased (Pb.05) hepatic concentrations ofTβMCA were found in STAT5 LKO mice in the control but not in thelithogenic diet group. Hepatic concentrations of other bile acids,includingβMCA,ω-muricholic acid (ωMCA), DCA, tauro-α-muricholicacid (TαMCA), TCDCA and TDCA, were the same in STAT5 LKO andSTAT5 fl/flmice, regardless of diet type (Table S4). The lithogenic dietincreased hepatic concentrations of CA, oMCA, TCA, DCA, TDCA andGCA but decreased the concentration of TCDCA.

In the gallbladder, there was an increase (Pb.05) in the concentra-tions of DCA and TDCA but a decrease (Pb.05) in the concentration ofGCA in STAT5 LKO mice in the lithogenic but not in the control group(Fig. 2C). The lithogenic diet increased (Pb.05) concentrations of CA,DCA, TDCA and GCA but decreased (Pb.05) concentrations of αMCA,βMCA, oMCA, TαMCA, TβMCA and TCDCA (Table S5).

3.3. Liver damage and cholesterol crystal and gallstone formation in thegallbladder

Liver sections were examined for ballooning and inflammationwith H&E staining, and fibrosis and fat accumulation were assessedwith MT and ORO staining, respectively. In the control diet group,STAT5 LKO mice showed a more pronounced tendency towardballooning, inflammation, fibrosis and fat accumulation in the livercompared with STAT5 fl/fl mice. In addition, all of these parametersworsened in lithogenic diet-fed STAT5 fl/fl mice (Fig. 3). However, inthe lithogenic diet group, STAT5 LKO mice showed a decreasedtendency toward ballooning, inflammation and fibrosis comparedwith the STAT5 fl/fl mice. These results suggest that STAT5 gene

Fig. 1. Hepatic STAT5 and IGF-1 mRNA levels and serum IGF-1 levels in STAT5 fl/fl and STAT5 LKO mice fed either a normal control diet (CON) or a lithogenic diet (Lith) for 12 weeks.mRNA levels were determined by real-time PCR and normalized with the housekeeping gene β-actin. Values for STAT5 fl/fl mice fed the normal control diet were normalized to 1.0.Values are expressed as mean±S.E.M. (n=5–6). a–d Means with different superscripts were different at Pb.05 among groups.

62 M. Baik et al. / Journal of Nutritional Biochemistry 39 (2017) 59–67

deletion ameliorated liver damage (ballooning, inflammation andfibrosis) induced by a lithogenic diet.

Next, biliary cholesterol crystals and gallstone formation wereexamined. In STAT5 fl/fl and STAT5 LKO mice, gallbladders from thecontrol diet group appeared transparent (not shown), but gallbladdersfrom the lithogenic diet group appeared opaque (Fig. S1A). Nocholesterol crystals were observed under a polarizing microscope inthe gallbladder bile from control diet groups, regardless of genotype.In the lithogenic diet groups, aggregated cholesterol crystals (“sandy”light-translucent stones or “solid” light-opaque stones) were detect-ed. However, the percentage of cholesterol crystals and gallstones andthe gallstone scores of the STAT5 LKO and STAT5 fl/fl mice did notsignificantly differ (Fig. S1B).

Table 1Lipid metabolite concentrations in serum and liver tissues from STAT5 fl/fl and STAT5 LKO mi

Metabolite Control diet Lithogen

STAT5 fl/fl STAT5 LKO STAT5 fl/

BloodPhospholipids (mg/dl) 249±27.0b 309±17.1ab 280±28Bile acid (μmol/L) 4.78±2.03b 7.54±1.81b 43.7±12Total cholesterol (mg/dl) 145±16.2b 194±12.7b 367± 41

Triglyceride (mg/dl) 126±13.2b 142±13.0b 80.5±14LDL cholesterol (mg/dl) 22.4±1.94b 31.2± 3.49b 238±38HDL cholesterol (mg/dl) 121±13.2b 159±11.1a 127±10

LiverPhospholipids (mg/g) 12.1±0.73b 13.9±0.66ab 13.3±1.5Total cholesterol (mg/g) 3.42±0.87b 5.87±1.29b 17.9±2.5Triglyceride (mg/g) 8.16±1.44 13.6±2.26 20.7±2.2

GallbladderBile volume (μl) 12.4±2.01b 16.2±3.37b 23.0±6.1

Values are expressed as mean±S.E.M. (n=5 ~ 6). a,b Means in the same row with different sup

3.4. Expression of cholesterol and bile acid metabolism genes in the liver

We examinedmRNA levels of cholesterol uptake-associated genes,including LDLR, low-density lipoprotein receptor-related protein 1(Lrp1) and scavenger receptor class B,member 1 (Scarb1). STAT5 genedeletion did not significantly altered mRNA levels of LDLR, Lrp1 andScarb1 genes (Fig. 4). mRNA levels of the cholesterol synthesisepidermal gene, 3-hydroxy-3-methylglutaryl-coenzyme A reductase(HMGCR), were also the same in STAT5 LKO and STAT5 fl/fl mice,regardless of diet type. The lithogenic diet caused a decrease (Pb.05) inLrp1 and HMGCR mRNA levels when data from the two genotypeswere combined. STAT5 LKO mice in the control diet group haddecreased mRNA levels of the VLDL transporter, apolipoprotein B

ce fed either a control diet or a lithogenic diet for 12 weeks

ic diet P value

fl STAT5 LKO Genotype Diet Interaction

.1ab 356±39.3a .03 NS NS

.7a 31.0±10.8a NS .001 NS.9a 441±55.5a NS b.001 NS

.4b 226±50.5a .02 NS .03

.7a 254±18.3a NS b.001 NS

.5b 116±7.29b NS NS .04

8ab 15.1±0.28a NS NS NS6a 17.3±1.71a NS b.001 NS2 15.7±1.92 NS NS NS

5b 42.2±7.68a .07 .003 NS

erscripts within a diet group were different at Pb.05. NS, not significant.

Fig. 2. Individual bile acid concentrations (μM) in serum (A), liver tissue (B) and gallbladder bile (C) from STAT5 fl/fl and STAT5 LKOmice fed either a control diet (CON) or a lithogenicdiet (Lith) for 12 weeks. Values are expressed as mean±S.E.M. (n=5–6). a–c Means with different superscript were different at Pb.05.

63M. Baik et al. / Journal of Nutritional Biochemistry 39 (2017) 59–67

(ApoB; P=.05), and those in the lithogenic diet group showed adecreasing trend in mRNA levels.

Next, we examined the mRNA levels of bile acid uptake, synthesis,detoxification and canalicular transport genes in the liver (Fig. 4). First,we examined mRNA levels of bile acid uptake genes from entero-hepatic circulation into hepatocytes, including the following: solutecarrier family 10 (sodium/bile acid co-transporter family), member 1(Slco10a1; Ntcp); solute carrier organic anion transporter family,member 1a1 (Slco1a1; Oatp1); solute carrier organic anion trans-porter family, member 1a4 (Slco1a4; Oatp2); and solute carrier

organic anion transporter family, member 1b2 (Slco1b2; Oatp4).STAT5 LKO mice in the control diet group had decreased (Pb.05)hepatic mRNA levels of Slco1a1 and Slco1b2, but this was not the casefor those in the lithogenic diet group. STAT5 LKO mice in the controldiet group had decreased trend of Slcola4 mRNA levels, but this wasnot the case for those in the lithogenic diet group. However, themRNAlevels of Slco10a1 were the same in both genotypes, regardless of diettype. The lithogenic diet caused down-regulation (Pb.05) of Slco10a1,Slco1a1 and Slco1b2 mRNA levels, regardless of genotype, but it hadno effect on Slco1a4 mRNA levels. Overall, both STAT5 gene deletion

Fig. 3. H&E, MT and ORO staining of liver sections and serum ALP and ALP levels from STAT5 fl/fl and STAT5 LKO mice fed either a control diet (CON) or a lithogenic diet (Lith) for 12weeks. (A) Ballooning and inflammation were determined by H&E staining. Fibrosis was determined by MT staining. (B) Quantification of ballooning, inflammation and fibrosis wasbased on a Kleiner scoring system, as described inMaterials andmethods. The degree of ORO stainingwas determined by reference ORO staining, as described inMaterials andmethods.a,b Means with different superscripts were different at Pb.05. A,B Means with different superscript were different at Pb.10.

64 M. Baik et al. / Journal of Nutritional Biochemistry 39 (2017) 59–67

and a lithogenic diet generally caused down-regulation of the genesinvolved in hepatic bile acid uptake from entero-hepatic circulation.

We measured the mRNA levels of the nuclear farnesoid X receptor(FXR) gene, which encodes a suppressor of bile acid synthesis. InSTAT5 LKO mice in the control diet group, increased FXRmRNA levelswere observed, but this was not the case in lithogenic diet group. Thelithogenic diet led to increased FXR gene expression, regardless ofgenotype.

Next, we examined hepatic mRNA levels of bile acid synthesisgenes, including the following: cytochrome P450, family 7, subfamilya, polypeptide 1(Cyp7a1); hydroxy-delta-5-steroid dehydrogenase, 3beta- and steroid delta-isomerase 5 (Hsd3b5); cytochrome P450,family 8, subfamily b, polypeptide 1 (Cyp8b1); and cytochrome P450,family 7, subfamily b, and polypeptide 1 (Cyp7b1). In the control dietgroups, STAT5 LKOmice had decreased (Pb.05) hepaticmRNA levels of

Hsd3b5, Cyp8b1 and Cyp7b1 and showed a nonsignificant decrease inCyp7a1 mRNA levels compared with STAT5 fl/fl mice. However, thelevels of these genes did not significantly differ between micegenotypes in the lithogenic diet group. The lithogenic diet markedlydecreased (Pb.05) the expression of all bile acid synthesis genes,including Cyp7a1, Hsd3b5, Cyp8b1 and Cyp7b1, regardless ofgenotype. Thus, no changes in the expression of bile acid synthesisgenes in STAT5 LKO mice in the lithogenic diet group may have beendue to themasking of the STAT5deletion effect by the profounddown-regulation of bile acid synthesis genes under lithogenic diet feeding.

We also examined the hepatic expression of bile acid detoxificationgenes. Cytochrome P450, family 2, subfamily b, and polypeptide 9(Cyp2b9) mRNA levels were up-regulated (Pb.05) in STAT5 LKOmice,regardless of diet type (Fig. 4). STAT5 LKO mice in the lithogenic dietgroup also had increased (Pb.05) mRNA expression of cytochrome

Fig. 4. LivermRNA levels of bile acidmetabolismgenes fromSTAT5fl/flandSTAT5LKOmice fedeither a controldiet (CON)ora lithogenicdiet (Lith) for12weeks.mRNA levelsweredeterminedbyreal-time PCR and normalized toβ-actin levels (n=5–6). Values for STAT5 fl/flmice fed the normal control dietwere normalized to 1.0. Values are expressed asmean±S.E.M. (n=5–6). a–cMeanswith different superscripts were different at Pb.05 among groups.

65M. Baik et al. / Journal of Nutritional Biochemistry 39 (2017) 59–67

66 M. Baik et al. / Journal of Nutritional Biochemistry 39 (2017) 59–67

P450, family 2, subfamily b, polypeptide 10 (Cyp2b10), cytochromeP450, family 2, subfamily b, polypeptide 13 (Cyp2b13), and sulfo-transferase family 2 A, and dehydroepiandrosterone (DHEA)-preferring, member 1 (Sult2a1), but this was not the case in controldiet group. The lithogenic diet caused up-regulation (Pb.05) ofCyp2b10 and Sult2a1 gene expression in the liver.

We then examined the expression of bile acid canaliculartransporter genes, including ATP-binding cassette, sub-family B(MDR/TAP), member 11 (Abcb11); ATP-binding cassette, sub-familyG (WHITE), member 5 (Abcg5); ATP-binding cassette, sub-family G(WHITE), member 8 (Abcg8); ATP-binding cassette, sub-family C(CFTR/MRP), member 2 (Abcc2); and ATP-binding cassette, sub-family B (MDR/TAP), and member 4 (Abcb4; mdr2). Compared withSTAT5 fl/fl mice, STAT5 LKO mice had a nonsignificant increase inAbcb11 gene expression. However, when lithogenic diet groups werecompared, the Abcb11 mRNA levels were significantly increased(Pb.05) in STAT5 LKOmice (Fig. 4). STAT5 gene deletion did not affectthe expression of other transporter genes, including Abcg5, Abcg8 andMdr2, regardless of diet type. The lithogenic diet significantlyincreased (Pb.05) Abcg5 and Abcg8 gene expression, regardless ofgenotype, but it had no effect on Abcc2 and Mdr2 gene expression ineither genotype.

Overall, our results demonstrate that STAT5 gene deletion down-regulates hepatic expression of many genes involved in bile acidentero-hepatic uptake and synthesis and up-regulates the expressionof bile acid detoxification genes. Furthermore, STAT5 gene deletionup-regulated the expression of Abcb11 gene, which encodes a majorcanalicular transporter.

4. Discussion

In the current study, hepatic down-regulation of the STAT5 geneand of the STAT5 target IGF-1 gene and a corresponding significantreduction in serum IGF-1 levels were confirmed in STAT5 LKO mice,which is consistent with previous reports [4,19].

In our study, STAT5 LKO mice did not have significantly alteredexpression of cholesterol uptake LDLR and cholesterol synthesisHmgcr genes, regardless of diet type, although they had numericalincrease in serum total and LDL cholesterol levels. Liver-specific STAT5gene deletion also did not alter the expression of canalicularcholesterol export genes, Abcg5 and Abcg8, in both diet groups. Inthe liver, cholesterol levels in both diet groups were not changed bySTAT5 gene deletion. Taken together, our results demonstrate thatSTAT5 gene deletion did not significantly affect hepatic cholesterolmetabolism. We found that STAT5 LKO mice had about 80% reductionin hepatic mRNA levels of STAT5 genes compared with STAT5 fl/flmice, revealing that the knock-out animals still have expressed about20% of the STAT5 mRNA. This may be one of reasons that can explainthat many of the mRNA levels observed were not dramaticallychanged by STAT5 LKO.

In our study with male mice, STAT5 gene deletion did not affecttranscription of slco10a1 (Ntcp) gene, although lithogenic diet down-regulated slco10a1 mRNA levels. However, prolactin activated STAT5in female postpartum suckling rat liver, resulting in transcriptionalactivation of Ntcp gene [20]. Species or gender may affect thisdifferential responses of Ntcp transcriptional regulation.

In our study, serum concentrations of all bile acids examined werenot changed by STAT LKO. However, we found that STAT5 genedeletion increased hepatic concentrations of CA and GCA in thelithogenic diet group.

The lithogenic diet group had significant increases in CA and GCAlevels in the liver. The lithogenic diet consisted of 0.5% CA, and thishigh quantity of CA may have increased the uptake of these bile acidsin themice on this diet. STAT5 LKOmice from the lithogenic diet grouphad increased concentrations of both DCA and TDCA in gallbladder

bile.We also found increasedAbcb11 gene expression. Abcb11 exportsbile acid from the liver to the gallbladder [21] andhas a high affinity forconjugated bile acids, including TCDCA, TCA, TDCA and GCA [21].Therefore, increased Abcb11 gene expression may have partiallycontributed to increased TDCA concentrations in the gallbladder bileof lithogenic diet-fed STAT5 LKOmice. In contrast, GCA concentrationswere lower in the gallbladder bile of lithogenic diet-fed STAT5 LKOmice. Perhaps, this may be related to decreased bile acid excretionfrom liverwith unclear reason, considering that liver contained higherGCA concentration in lithogenic diet-fed STAT5 LKO mice with nodifference in gene expression for de novo synthesis of bile acids.

In the control diet group, we found that liver sections from STAT5LKOmice had a greater tendency of ballooning, inflammation, fibrosisand fat accumulation compared with liver sections from STAT5 fl/flmice. Previously, it was found that disruption of GH–STAT5–IGF-1signaling aggravated liverfibrosis inmicewith a deletion of the Stat5a/blocus in both hepatocytes and cholangiocytes, although this changewasnot detected when the Stat5a/b locus was deleted only in hepatocytes[22]. Recently, a study showed that GH resistance (high circulating GHlevels and low IGF-1 levels) induced by loss of GH receptor signalingexacerbated liver fibrosis in amousemodel of inflammatory cholestasis[23]. It has been shown that TβMCA can increase biliary HCO3

−

concentration and a HCO3− umbrella as a protective mechanism against

bile acid-induced injury in human cholangiocytes [24,25]. In the controldiet group, we observed decreased hepatic concentration of TβMCA inSTAT5 LKOmice compared with STAT5 fl/fl. TβMCA has been shown tohave beneficial effects in cholestasis-induced liver damage in primaryrat hepatocytes [26]. Thus, a decrease in TβMCA concentration maypartially explain the increase in liver damage in STAT5 LKO micecompared with STAT5 fl/fl mice in the control diet groups.

In this study, lithogenic diet feeding caused more severe liverdamage (ballooning, inflammation, fibrosis) in STAT5fl/flmice than inSTAT5 LKO mice. Our study revealed that minor liver damage wasinduced in STAT5 LKO mice fed a normal diet and that moderate liverdamagewas induced in STAT5fl/flmice fed a lithogenic diet. However,STAT5 gene deletion ameliorated lithogenic diet-induced liverdamage. Our results showed a profound up-regulation of bile aciddetoxification genes, including Cyp2b9, Cyp2b10, Cyp2b13 andSult2a1 in STAT5 LKO mice. Detoxification renders bile acids lesstoxic and more hydrophilic and amenable for urinary excretion[27,28]. Therefore, in the lithogenic diet group, an up-regulation in bileacid detoxification genes may have lessened liver damage in STAT5LKO mice compared with STAT5 fl/fl mice.

In lithogenic diet group, gallbladder bile volume was higher inSTAT5 LKO mice than in STAT5 fl/fl mice. This may have been due toincreased Abcb11 gene expression, resulting in increased bile acidtransport from the liver to the gallbladder in lithogenic diet-fed STAT5LKO mice.

We found an abundance of aggregated birefringent cholesterolcrystals and/or gallstones in the bile of both STAT5 fl/fl and STAT5 LKOmice fed a lithogenic diet. In this study, the gene expression of thecanalicular cholesterol exporters, Abcg5 and Abcg8, was profoundlyincreased in lithogenic diet groups. Increased Abcg5 and Abcg8 geneexpression in lithogenic diet-fed mice indicated that a significantamount of cholesterol was secreted from the liver to the canalicularlumen, promoting cholesterol crystal formation. However, theincidence and severity of cholesterol crystals and gallstones did notsignificantly differ between STAT5 fl/fl and STAT5 LKOmice. In clinicalstudies, the effects of GH on cholesterol gallstone development havebeen controversial. Dowling et al. [9] found an increased prevalence ofgallstones in patients with acromegaly. Erlinger et al. [29] also foundthat acromegalic patients had cholesterol-supersaturated bile. Incontrast, another study reported that GH therapy in healthy adultmales did not lead to an increased risk of developing cholesterolgallstones [5]. Furthermore, GH administration to GH-deficient

67M. Baik et al. / Journal of Nutritional Biochemistry 39 (2017) 59–67

children also did not lead to altered biliary lipid composition [10]. Ourstudy showed that diminished STAT5 signaling did not significantlyaffect cholesterol metabolism or influence cholesterol crystal andgallstone development, despite profound hepatic changes in thetranscription of genes involved in the uptake, synthesis, anddetoxification of bile acids.

In conclusion, we found that STAT5 gene deletion had a minoreffect on cholesterol metabolism, as evidenced by no changes inexpression of cholesterol transport and cholesterol synthesis genes,and numeric changes in serum and liver cholesterol levels. Theseminor changes in cholesterolmetabolismmaynot have been sufficientto influence on cholesterol crystal and gallstone formation inlithogenic diet-fed STAT5 LKO mice. Thus, STAT5 signaling may notsignificantly influence cholesterol metabolism.

In contrast, we found profound changes in the expression of bileacid metabolism genes, including decreased bile acid synthesis anduptake genes and increased detoxification genes, in STAT5 LKO mice.STAT5 gene deletion induced minor liver damage, as evidenced byslight increases in ballooning, inflammation and fibrosis in the liversections from control diet-fed STAT5 LKOmice. In STAT5 fl/flmice, thelithogenic diet induced further liver damage; however, the degree ofliver damage in lithogenic diet-fedmice was lower in STAT5 LKOmicethan in STAT5 fl/flmice. Therefore, lithogenic diet-fed STAT5 LKOmicemay have had a better defense system with which to counteract theoversupply of hepatic cholesterol and bile acids through activation ofbile acid detoxification genes than did STAT5 fl/fl mice.

Our study shows that local STAT5 signaling in the liver has animportant role in regulating the transcription of genes associatedwiththe synthesis, detoxification and transport of bile acids, but it has onlya minor role in bile acid metabolism.

Authors' disclosure

The authors have no conflicts of interest to disclose.

Appendix A. Supplementary data

Supplementary data to this article can be found online at doi:10.1016/j.jnutbio.2016.09.012.

References

[1] Ram PA, Park SH, Choi HK, Waxman DJ. Growth hormone activation of Stat 1, Stat3, and Stat 5 in rat liver. Differential kinetics of hormone desensitization andgrowth hormone stimulation of both tyrosine phosphorylation and serine/threonine phosphorylation. J Biol Chem 1996;271:5929–40.

[2] Smit LS, Meyer DJ, Billestrup N, Norstedt G, Schwartz J, Carter-Su C. The role of thegrowth hormone (GH) receptor and JAK1 and JAK2 kinases in the activation ofStats 1, 3, and 5 by GH. Mol Endocrinol 1996;10:519–33.

[3] Barclay JL, Nelson CN, Ishikawa M, Murray LA, Kerr LM, McPhee TR, et al. GH-dependent STAT5 signaling plays an important role in hepatic lipid metabolism.Endocrinology 2011;152:181–92.

[4] Baik M, Nam YS, Paio MY, Kang HJ, Park SJ, Lee J-H. Liver-specific deletion of thesignal transducer and activator of transcription 5 gene aggravates fatty liver inresponse to a high-fat diet in mice. J Nutr Biochem 2016;29:56–63.

[5] OlivecronaH, Ericsson S, Angelin B.Growth hormone treatment doesnot alter biliarylipid metabolism in healthy adult man. J Clin Endocrinol Metab 1995;80:1113–7.

[6] Matasconi M, Parini P, Angelin B, Rudling M. Pituitary control of cholesterolmetabolism in normal and LDL receptor knock-out mice: effects of hypophysec-tomy and growth hormone treatment. Biochim Biophys Acta 1736;2005:221–7.

[7] Baik M, Yu JH, Hennighausen L. Growth hormone-STAT5 regulation of growth,hepatocellular carcinoma, and liver metabolism. Ann N Y Acad Sci 2011;1229:29–37.

[8] Holloway MG, Cui Y, Laz EV, Hosui A, Hennighausen L, Waxman DJ. Loss ofsexually dimorphic liver gene expression upon hepatocyte-specific deletion ofSTAT5a–STAT5b locus. Endocrinology 2007;148:1977–86.

[9] Dowling RH, Hussaini SH, Murphy GM, Besser GM, Wass JA. Gallstones duringoctreotide therapy. Metabolism 1992;41:22–33.

[10] Heubi JE, Burstein S, Sperling MA, Gregg D, Subbiah MT, Matthews DE. The role ofhuman growth hormone in the regulation of cholesterol and bile acid metabolism.J Clin Endocrinol Metab 1983;57:885–91.

[11] Wang DQ, Paigen B, Carey MC. Phenotypic characterization of Lith genes thatdetermine susceptibility to cholesterol cholelithiasis in inbred mice: physical–chemistry of gallbladder bile. J Lipid Res 1997;38:1395–411.

[12] Xie Y, Fung HY, Newberry EP, Kennedy S, Luo J, Crooke RM, et al. Hepatic Mttpdeletion reverses gallstone susceptibility in L-Fabp knockout mice. J Lipid Res2014;55:540–8.

[13] Kleiner DE, Brunt EM, Van Natta M, Behling C, Contos MJ, Cummings OW, et al.Design and validation of a histological scoring system for nonalcoholic fatty liverdisease. Hepatology 2005;41:1313–21.

[14] Kwon DH, Kang W, Nam YS, Lee MS, Lee IY, Kim HJ, et al. Dietary proteinrestriction induces steatohepatitis and alters leptin/signal transducers andactivators of transcription 3 signaling in lactating rats. J Nutr Biochem 2012;23:791–9.

[15] Ballatori N, Fang F, ChristianWV, Li N, Hammond CL. Ostalpha–Ostbeta is requiredfor bile acid and conjugated steroid disposition in the intestine, kidney, and liver.Am J Physiol Gastrointest Liver Physiol 2008;295:G179–86.

[16] Alnouti Y, Csanaky IL, Klaassen CD. Quantitative-profiling of bile acids and theirconjugates in mouse liver, bile, plasma, and urine using LC–MS/MS. J ChromatogrB Analyt Technol Biomed Life Sci 2008;873:209–17.

[17] You J, Shi Y, Ming Y, Yu Z, Yi Y, Liu J. Development of a sensitive reagent, 1,2-benzo-3,4-dihydrocarbazole-9-ethyl-p-toluenesulfonate, for determination ofbile acids in serum by HPLC with fluorescence detection, and identification bymass spectrometry with an APCI source. Chromatographia 2004;60:527–35.

[18] Inoue Y, Yu AM, Inoue J, Gonzalez FJ. Hepatocyte nuclear factor 4alpha is a centralregulator of bile acid conjugation. J Biol Chem 2004;279:2480–9.

[19] Cui Y, Hosui A, Sun R, Shen K, Gavrilova O, Chen W, et al. Loss of signal transducerand activator of transcription 5 leads to hepatosteatosis and impaired liverregeneration. Hepatology 2007;46:504–13.

[20] Ganguly TC, O'Brien ML, Karpen SJ, Hyde JF, Suchy FJ, Vore M. Regulation of the ratliver sodium-dependent bile acid cotransporter gene by prolactin. Mediation oftranscriptional activation by Stat5. J Clin Invest 1997;99:2906–14.

[21] Kosters A, Karpen SJ. Bile acid transporters in health and disease. Xenobiotica2008;38:1043–71.

[22] Blaas L, Kornfeld JW, Schramek D, Musteanu M, Zollner G, Gumhold J, et al.Disruption of the growth hormone-signal transducer and activator of transcrip-tion 5-insulinlike growth factor 1 axis severely aggravates liver fibrosis in amousemodel of cholestasis. Hepatology 2010;51:1319–26.

[23] Stiedl P, McMahon R, Blaas L, Stanek V, Svinka J, Grabner B, et al. Growth hormoneresistance exacerbates cholestasis-induced murine liver fibrosis. Hepatology2015;61:613–26.

[24] Katagiri K, Nakai T, Hoshino M, Hayakawa T, Ohnishi H, Okayama Y, et al. Tauro-beta-muricholate preserves choleresis and prevents taurocholate-induced chole-stasis in colchicine-treated rat liver. Gastroenterology 1992;102:1660–7.

[25] Hohenester S, Wenniger LM, Paulusma CC, van Vliet SJ, Jefferson DM, Elferink RP,et al. A biliary HCO3

− umbrella constitutes a protective mechanism against bileacid-induced injury in human cholangiocytes. Hepatology 2012;55:173–83.

[26] Denk GU, Kleiss CP, Wimmer R, Vennegeerts T, Reiter FP, Schulz S, et al. Tauro-β-muricholic acid restricts bile acid-induced hepatocellular apoptosis by preservingthe mitochondrial membrane potential. Biochem Biophys Res Commun 2012;424:758–64.

[27] Claudel T, Gernot Z, Martin W, Trauner M. Role of nuclear receptors for bile acidmetabolism, bile secretion, cholestasis, and gallstone disease. Biochim BiophysActa 2011;1812:867–78.

[28] Paolini M, Pozzetti L, Piazza F, Cantelli-Forti G, Roda A. Bile acid structure andselective modulation of murine hepatic cytochrome P450-linked enzymes.Hepatology 1999;30:730–9.

[29] Erlinger S, Chanson P, Dumont M, Ponsot P, Warnet A, Harris AG. Effects ofoctreotide on biliary lipid composition and occurrence of cholesterol crystals inpatients with acromegaly. A prospective study. Dig Dis Sci 1994;39:2384–8.