Role of hepatic phospholipids in development of liver injury in Mdr2 (Abcb4) knockout mice: Phospholipids and liver injury in Mdr2 knockout mice

BAS IC STUDIES

Roleofhepaticphospholipids indevelopmentof liver injury inMdr2(Abcb4)knockoutmiceAnna Baghdasaryan1, Peter Fickert1, Andrea Fuchsbichler2, Dagmar Silbert1, Judith Gumhold1, GerdHorl3, Cord Langner2, Tarek Moustafa1, Emina Halilbasic1, Thierry Claudel1 and Michael Trauner1

1 Laboratory of Experimental and Molecular Hepatology, Department of Internal Medicine, Division of Gastroenterology and Hepatology, Medical

University, Graz, Austria

2 Institute of Pathology, Medical University, Graz, Austria

3 Department of Molecular Biology and Biochemistry, Center for Molecular Medicine, Medical University, Graz, Austria

Keywords

Mdr2 – phosphatidylcholine – portal fibrosis –

sclerosing cholangitis

Abbreviations

ALT, alanine aminotransferase; AP, alkaline

phosphatase; BA, bile acids; BW, body

weight; CA, cholic acid; CCT, choline-

phosphate cytidyltransferase; CDD, choline-

deficient diet; CKa, choline kinase a; CKb,

choline kinase b; Co, control diet; Cyp7a1,

cholesterol 7a-hydroxylase; H&E staining,

haematoxylin and eosin staining; HP,

hydroxyproline; Lyso-PC,

lysophosphatidylcholine; Lyso-PI,

lysophosphatidylinositol; LW, liver weight;

PC, phosphatidylcholine; PCD,

phosphatidylcholine-enriched diet; Pemt,

phosphatidylethanolamine

methyltransferase; PL, phospholipid; SM,

sphingomyelin; TG, triglycerides; TLC, thin-

layer chromatography; VCAM-1, vascular cell

adhesion molecule-1.

Correspondence

Michael Trauner, Laboratory of Experimental

and Molecular Hepatology, Department of

Internal Medicine, Division of Gastroenterology

and Hepatology, Medical University of Graz,

Auenbruggerplatz 15, A-8036 Graz, Austria

Tel: 11143 316 385 4388

Fax: 11143 316 385 7560

e-mail: [email protected]

Received 4 December 2007

Accepted 4 March 2008

DOI:10.1111/j.1478-3231.2008.01758.x

AbstractBackground/Aims: Multidrug resistance protein 2 (Abcb4) gene knockout mice

(Mdr2�/�) lack phosphatidylcholine (PC) excretion into bile and spontaneously

develop sclerosing cholangitis, biliary fibrosis and hepatocellular carcinomas. We

therefore aimed to test whether formation and hepatic retention of abnormal PC

metabolites contribute to the pathogenesis of liver injury in Mdr2�/� mice.

Methods: Mdr2�/�mice were either fed a diet supplemented with soybean lecithin

2.5% w/w [phosphatidylcholine-enriched diet (PCD), to increase hepatic PC

content] or a choline-deficient diet (CDD, to reduce hepatic PC content) for 4

weeks; controls received chow with energy and nutrient content equivalent to PCD

and CDD. Serum liver tests, liver histology, markers of fibrosis, cholangiocyte

activation, cell proliferation and thin-layer chromatography for phospholipid (PL)

composition were carried out. Results: PCD decreased serum alkaline phosphatase

and total bilirubin levels compared with controls, while liver histology as well as

hepatic hydroxyproline content as markers of liver fibrosis did not differ among

groups. Both PCD and CDD decreased hepatocellular proliferation compared with

controls. Hepatic, serum and biliary PLs remained unchanged despite dietary

manipulations and no potentially toxic PL metabolites were detected. Con-

clusions: Mdr2�/�mice maintain stable hepatic, serum and biliary PL metabolism

in response to dietary PC manipulations. Our findings therefore suggest that liver

injury in Mdr2�/�mice is not due to formation of toxic PL metabolites.

Biliary phosphatidylcholine (PC) excretion is essentialfor preventing bile toxicity by formation of mixedmicelles with bile acids (BA) and cholesterol (1, 2). PCexcretion into bile is mediated by a PC-specific flop-pase, the multiple drug-resistance protein 2 (Mdr2/

Abcb4 in rodents, MDR3/ABCB4 in humans) locatedon the hepatocyte canalicular membrane (3, 4). Thephysiological function of this transport system wasdetermined by the generation of mice homozygous fora disruption of the Mdr2/Abcb4 gene (5). Mdr2�/�

Liver International (2008)948 c� 2008 The Authors. Journal compilation c� 2008 Blackwell Munksgaard

Liver International ISSN 1478-3223

mice represent a murine model of chronic cholestaticliver disease, characterized by increased proliferationof bile ducts, sclerosing cholangitis and biliary type ofliver fibrosis (6). Among human hereditary cholestaticdisorders, a subtype of progressive familial intra-hepatic cholestasis 3, manifesting with high serumg-glutamyl transpeptidase levels, shares biochemicaland histological features with Mdr2�/�mice (6, 7). Inaddition, genetic MDR3 variants may also play asignificant role in a range of acquired cholestatic liverdiseases such as intrahepatic cholestasis of pregnancyand drug-induced cholestasis (4). Taken together,these findings highlight the importance of biliary PCexcretion in the development of cholestatic liverdiseases.

In the liver, PC represents the main phospholipid(PL) and its homoeostasis is tightly regulated bycoordinated biosynthesis, catabolism and excretioninto bile and plasma (8). Mammalian cells synthesizePC from choline via the cytidine diphosphate(CDP)–choline pathway (also known as the Kennedypathway), which requires an exogenous source ofcholine (9–11). In the liver, PC can also be generatedvia methylation of phosphatidylethanolamine (PE),the second abundant PL in cell membranes by theenzyme PE-N-methyltransferase (Pemt) (12). So far,hepatobiliary injury observed in Mdr2�/� mice hasbeen attributed to the absence of biliary PC resultingin toxic bile formation and injury of bile ducts (4–6).While impaired biliary PC excretion in Mdr2�/�micehas received considerable attention in the past withrespect to the pathogenesis of sclerosing cholangitisand biliary fibrosis (4–6, 13), hepatocellular PC meta-bolism has not been studied in great detail so far.Astonishingly, hepatic lipid content and particularly

hepatic PL content remains unchanged in Mdr2�/�

despite their virtual absence of biliary PC excretion(14–16). This argues for an altered PC synthesis ordegradation. Because enhanced PC degradation maygive rise to biologically highly active and frequentlytoxic compounds (e.g. Lyso-PC) (17), we hypothe-sized that degradation products of PC could contri-bute to liver injury in Mdr2�/� mice, while reductionof hepatic PC content (under choline-deficient condi-tions) would be expected to ameliorate liver damage inMdr2�/� mice via reduction of potentially toxic PCmetabolites. Therefore, the aim of this study was toexplore the putative pathogenetic role of an alteredhepatic PC homoeostasis in Mdr2�/�mice. To addressthis issue, we compared liver injury in Mdr2�/� micefed either a control diet, a choline-deficient diet(CDD) or a PC-enriched diet (PCD).

Materials and methods

Animal experiments

Mdr2�/� mice of the FVB/N background were ob-tained from Jackson Laboratory (Bar Harbor, ME,USA) and housed under a 12:12-h light/dark cyclewith water and a control mouse diet (SSNIFF, Soest,Germany) ad libitum. The experimental protocolswere approved by the local Animal Care and UseCommittee according to the criteria outlined in theGuide for the Care and Use of Laboratory Animalsprepared by the US National Academy of Sciences(National Institutes of Health publication 86-23, re-vised 1985).

Feeding protocol

Feeding of PCD and CDD in Mdr2�/� was initiated atthe age of 4 weeks, a time point when the principalliver pathology including sclerosing cholangitis andbiliary fibrosis are already present, but not yet fullyestablished in this model (6). Male mice were eitherfed a PCD, supplemented with soybean lecithin (2.5%w/w) or CDD, for 4 weeks. The duration of 4 weeks’feeding has been chosen based on the previous workshowing efficacy in this model after 4 weeks of BAfeeding (13, 18, 19). Controls received a chow equiva-lent in nutrient and energy content to CDD and PCD.Both CDD (E15327-24) and control diets (E15000-04)were obtained from SSNIFF. Fat-free soybean lecithingranules, containing 20–24% PC, were kindly pro-vided in the form of Lipoid S21 as a gift from LipoidGMBH (Ludwigshafen, Germany). Detailed diet com-position is presented in Table 1.

Table 1. Ingredients of experimental diets

Ingredients Co CDD PCD

Gross energy (MJ/kg) 18.0 18.0 18.0Crude protein (%) 20.8 20.8 20.8Crude fat (%) 4.2 4.2 4.2Starch (%) 46.8 47.0 46.8Sugar (%) 10.8 10.8 10.8Choline (mg/kg) 1040 o 50 1040Lecithin granules (g) – – 25

Vitamin, mineral and aminoacid content was equivalent in control, CDD

and PCD. Lecithin granules (Lipoid GMBH) from soybeans contained

97% total phospholipids, of which 20–24% was presented as phospha-

tidylcholine, 18–22% as phosphatidylethanolamine, 12–15% as phos-

phatidylinositol and o 3% as Lyso-phosphatidylcholine.

CDD, choline-deficient diet; Co, control diet; PCD, phosphatidylcholine-

supplemented diet.

Liver International (2008)c� 2008 The Authors. Journal compilation c� 2008 Blackwell Munksgaard 949

Baghdasaryan et al. Phospholipids and liver injury in Mdr2 knockout mice

Food intake, body weight/liver weight measurement

Food intake was controlled by monitoring daily foodconsumption/mouse at days 1, 3, 7, 14 and 28. Bodyweight (BW) gain was monitored during the entireexperiment. The liver weight (LW) was correlated withBW at the end of the experiment.

Routine serum biochemistry

Blood was collected at harvesting. Samples were leftfor 10 min at room temperature and centrifuged for15 min at 1200 g Serum samples were stored at� 80 1C until analysis. Assays for serum total bilirubin,cholesterol, triglycerides (TGs), alanine aminotrans-ferase (ALT) and alkaline phosphatase (AP) wereperformed by a Hitachi 917 analyzer (BoehringerMannheim, Mannheim, Germany)

Liver histology

For conventional light microscopy (haematoxylin andeosin and Sirius red stain), livers were fixed in 4%neutral-buffered formaldehyde solution for 24 h andembedded in paraffin. The preparation of paraffinsections was performed as described previously by us(20). The sections were coded and examined by apathologist (C. L.) in a blinded manner and scoredsemiquantitatively for five histopathological para-meters in four grades (0–3) of severity: portal inflam-mation, ductular proliferation, fibrosis, mitoticactivity and Councilman bodies in all animals asdescribed previously with a slight modification (21).The degree of inflammation in the portal triads wasgraded as follows: 0, no inflammatory cells present;1, only scattered inflammatory cells present in themesenchyma of the portal triads; 2, increasing inflam-matory infiltrate associated with cholangiolitis, poly-morph nuclear leucocytes invading bile ducts; and 3,destructive cholangiolitis with the presence of fibrosis.Ductular proliferation at a magnification of 250-foldwas graded as follows: 0, no proliferation present; 1, nomore than one portal tract involved; 2, one to fiveportal tracts involved; and 3, all portal tracts involved.The grade of fibrosis was scored as follows: 0, nofibrosis present; 1, expansion of the portal triad byincrease of fibrotic tissue; 2, sporadic presence ofporto-portal septa in one or two lobules; and 3, septasurrounding the majority of lobules. Mitotic activity(mitotic figures per 10 high-power fields at a magnifi-cation of 400-fold) was assessed as: 0, no mitoticfigures present; 1, less than three; 2, three to five; or 3,more than five. The number of acidophilic (‘Council-man’) bodies in each lobule at a magnification of

400-fold was graded as follows: 0, no acidophilicbodies present; 1, less than three; 2, three to 10; and 3,more than 10.

Determination of hepatic hydroxyproline content

To quantify liver fibrosis, hepatic hydroxyproline (HP)was measured as described previously (22).

Immunohistochemistry for Ki-67

Polyclonal rabbit antibody to Ki-67 (dilution 1:500;Novocastra, Newcastle upon Tyne, UK) was used todetect Ki-67 protein in paraffin-embedded liver sec-tions. b-amino-9-ethyl-carbazole (AEC; Dako, Glostr-up, Denmark) has been used to detect binding of theantibody in the ABC system (Dako). The number ofKi-67-positive nuclei in 30 high-power fields wascounted to assess the hepatocellular proliferation insections of five animals in control (Co) and six animalsin PCD and CDD groups.

Western blotting for cytokeratin 19 and vascular celladhesion molecule-1

Protein isolation and Western blotting for cytokeratin19 (CK19) and vascular cell adhesion molecule-1(VCAM-1) were performed as described previously(23, 24).

Measurement of bile flow and biliary phospholipidcomposition

Bile flow and biliary PL composition were determinedin the different groups as described previously (13).Briefly, CDD, PCD and control diet-fed Mdr2�/�micewere anaesthetized (10 mg of avertin intraperitoneally)and the gall bladder was cannulated after common bileduct ligation for bile collection. After a 10-min equili-bration period, bile was collected in preweighed testtubes for 30 min and bile flow was determined grav-imetrically and normalized to LW as described pre-viously (13). Biliary PL concentration was determinedusing a commercial kit (Phospholipid B; Wako, Neuss,Germany).

Lipid extraction from the liver, serum and bile.Measurement of phospholipid content and thin layerchromatographical analysis of phospholipid pool

Phospholipids were extracted from frozen liver tissue,serum and bile using the Bligh and Dyer protocol (25).The PC concentration was determined in the liver,serum and bile using a commercially available kit,based on phospholipase D activity (Phospholipids B;


Phospholipids and liver injury in Mdr2 knockout mice Baghdasaryan et al.

Wako). Thin layer chromatographical (TLC) analysisof PL pools was performed as described previously(26), using TLC aluminium sheets 20 cm� 20 cmSilica gel 60 (Merck, Darmstadt, Germany) andchloroform/propionic acid/n-propanol/water (2/3/2/1, v/v) as a solvent. PL spots were visualized afterspraying primuline (Sigma Aldrich, Vienna, Austria),identified on dried plates under a ultraviolet lamp(366 nm) and scanned in a fluorescence scanner (Ty-phoon 9400 variable mode imager; Amersham Bio-sciences, Freiburg, Germany; software IMAGEQUANT 5.2;Molecular Dynamics, Sunnyvale, CA, USA). Standardlipids: PE, PC, phosphatidylserine, phosphatidylinosi-tol, phosphatidic acid, sphingomyelin (SM) as well asLyso-PC were purchased from Sigma Aldrich andloaded on the plate.

Messenger RNA analysis and polymerase chainreaction

Total RNA isolation, complementary DNA synthesisand TaqMan real-time polymerase chain reaction forcholesterol 7a-hydroxylase (Cyp7a1) choline kinase a(CKa), choline kinase b (CKb), choline-phosphatecytidyltransferase (CCT), Pemt and 18s rRNAwere performed as described previously (27). TaqManoligonucleotides and probes were as follows: forCKa: forward primer 50-GCC TAC CTG TGG TGTAAG GAA TTC-30, reverse primer 50-GTT ACTGAG ACC ACC CCT GAT GA-30 (NM_013490.2/NM_001025566.1), for CKb: forward primer 50-ACTATA GAG TTC GGC TAC TTG GAG TAT G-30, reverseprimer 50-GGT TGG ATC CTC AGG ATG ATG-30

(NM_007692.4), for CCT: forward primer 50-GGAGTT GAG TTA AAA GAA GAT GGA TG-30, reverseprimer 50-GCT GCA CTT TGG AAG GAA TTC-30

(Z12302.1/NM_078622), for Pemt: forward primer 50-GGG ACC TTT CTA GGT GAC TAC TTT G-30,reverse primer 50-CCA GCC TAG GTA GTT GGCTGT AC-30 (NM_008819/NM_013003), for Cyp7a1:forward primer 50-CAG GGA GAT GCT CTG TGTTCA-30, reverse primer 50-AGG CAT ACA TCC CTT

CCG TGA-30, probe 50-TGC AAA ACC TCC AATCTG TCA TGA GAC CTC C-30 (NM_007824), for 18s:forward primer 50-GTA ACC CGT TGA ACC CCATT-30, reverse primer: 50-CAA TCC AAT CGG TAG TAGCG-30 (X 00686). Sybr green assays without the use ofspecific probes were performed for determination ofCKa, CKb, CCT, Pemt and 18s mRNA.

Statistical analysis

Data are reported as means� standard deviation.Data were tested for normality by the Kolmogorov–Smirnov and the Shapiro–Wilk test. Statistical analysiswas performed using a one-way analysis of variance(ANOVA) test. A P value of o 0.05 was considered to besignificant. Data were evaluated using SPSS (release14.0, 2005; SPSS Inc., Chicago, IL, USA).

Results

Choline-deficient diet and phosphatidylcholine-enriched diet significantly reduce the liver to bodyweight ratio in Mdr2�/� mice

We first addressed whether the different diets had anyimpact on the general appearance, behaviour and BWof Mdr2�/� mice. Food consumption (average: 4.5 g/day) was similar and the survival rate (100%) did notdiffer between the groups. No apparent differenceswere observed in animal behaviour and/or develop-ment. As shown in previous studies, the LW/BW ratiois increased in 8-week-old Mdr2�/� mice comparedwith age-matched wild-type mice (5, 16) (mean valuesfrom our laboratory are as follows: 7.6� 0.5 vs.5.8� 0.5). In our experiments, the LW/BW ratio wassignificantly reduced in the CDD and PCD groupscompared with the control diet-fed animals (Table 2).

Phosphatidylcholine-enriched diet significantlyreduces serum liver enzymes but has no impact onbile duct disease in Mdr2�/� mice

The macroscopic appearance of livers from PCD andCDD was not changed compared with control diet-fed

Table 2. Characteristics of Mdr2�/�mice fed different experimental diets

Variable Co (n = 5) CDD (n = 6) PCD (n = 6)

Body weight (g) 24.9� 1.7 25.6� 1.4 26.2� 1.2Liver weight (g) 1.88� 0.11 1.70� 0.19 1.77� 0.16Liver weight to body weight ratio (%) 7.5� 0.2 6.6� 0.5# 6.9� 0.17�

Mdr2�/�male mice received either a control diet (Co), a choline-deficient diet (CDD) or a phosphatidylcholine-enriched diet (PCD) for 4 weeks. The

values are presented as means� SD.�Po 0.05 vs. Co.#Po 0.01,

SD, standard deviation.



Mdr2�/� mice. In line with these results, microscopicexamination showed that Mdr2�/� mice receivingeither CDD or PCD developed pronounced portalfibrosis with sclerosing cholangitis and proliferationof both small and large bile ducts to the same degree asobserved in the controls (Fig. 1A). Histological scoringby a pathologist (C. L.) in a blinded manner also didnot show any differences in the severity of the lesionsamong study groups (Table 3). As reported previously,serum parameters of liver damage and cholestasis(ALT, AP and bilirubin) are significantly elevated inMdr2�/� mice compared with age-matched Mdr21/1

mice (5, 6, 13, 18). In our experiments, total bilirubin,ALT and AP levels were not altered in CDD animalscompared with the corresponding controls (Fig. 1B).Interestingly, PCD feeding in Mdr2�/� mice signifi-cantly reduced AP and total bilirubin levels comparedwith controls as well as ALT levels compared withCDD animals (Fig. 1B), despite a lack of effects onliver histology. Collectively, these data show that diet-ary choline deficiency does not reduce overall liverinjury in Mdr2�/� mice. On the contrary, dietary PCsupplementation may be partially beneficial by im-proving liver biochemistry.

Dietary choline content does not influence bile flowand BA synthesis in Mdr2�/� mice

Because toxic bile is considered to trigger the cascadeof pathogenetic events involved in liver damage ofMdr2�/� mice (6), we next determined bile flow andexpression levels of the key enzyme involved in BAsynthesis Cyp7a1. Bile flow measurements did notreveal significant differences between the groups (Co:50.7� 6.0 ml/30 min/mg LW, CDD: 44.5� 9.4 ml/30 min/mg LW, PCD: 40.1� 13.6 ml/30 min/mg LW),showing that dietary choline content has no impacton bile flow in Mdr2�/�mice. Moreover, BA synthesisreflected by mRNA expression of the Cyp7a1 gene didnot differ between control, CDD and PCD groups (Co:100� 104%, CDD: 160� 79%, PCD: 120� 43%).Taken together, our data indicate that the dietarycholine content most likely does not influence BAhomoeostasis in Mdr2�/�mice.

Dietary choline content does not influencedevelopment of hepatic fibrosis in Mdr2�/� mice

As described previously, Mdr2�/�mice develop biliarytype of liver fibrosis and demonstrate two- to three-

Fig. 1. Serum biochemistry but not liver histology is affected by dietary choline content in Mdr2�/�mice. (A) Representative liverhistology (H&E) of 8-week-old Mdr2�/�mice receiving control (Co), choline-deficient diet (CDD) or phosphatidylcholine-enriched diet(PCD) for 4 weeks is shown (original magnification, 10-fold). Cholangiopathy is not influenced by dietary choline content as reflectedby the similar histological appearance of liver injury in all the experimental groups. (B) Serum parameters of liver damage andcholestasis. PCD but not CDD improved liver biochemistry. Data are represented as means� SD. �Po 0.05 vs. Co, #Po 0.05compared with CDD. ALT, alanine aminotransferase; AP, alkaline phosphatase; bd, bile duct; H&E, haematoxylin and eosin; pv, portalvein; SD, standard deviation.



fold increased levels of HP: a marker of hepatic fibrosiscompared with age-matched wild-type animals (13,28). We next assessed the influence of PCD and CDDon biliary fibrosis in Mdr2�/�mice. As shown by Siriusred staining, collagen deposition was similar in differ-ent experimental groups independently of the dietarycomposition (Fig. 2A). In line with these data, HPcontent did not differ between the control diet-, PCD-and CDD-fed mice (Fig. 2B). Collectively, these datatherefore indicate that dietary choline content doesnot play a pivotal role in the development of biliaryfibrosis in Mdr2�/�mice.

Phosphatidylcholine-enriched diet and choline-deficient diet significantly reduce hepatocellular butnot ductular proliferation in Mdr2�/� mice

Bile duct disease in sclerosing cholangitis and inMdr2�/� mice is characterized by significantly in-creased number of proliferating hepatocytes and cho-langiocytes as well as acquisition of a ‘reactivecholangiocyte phenotype’ characterized by theover-expression of pro-inflammatory cytokines andchemokines such as VCAM-1 (6, 29). To furthercharacterize the influence of CDD and PCD on biliarylesions, we determined hepatic CK19 content as amarker of cholangiocyte cytoskeleton and VCAM-1content as a marker of the ‘reactive cholangiocytephenotype’ by Western blotting. Neither cholangiocytemass nor VCAM-1 expression was affected by PCD orCDD (Fig. 3), showing that dietary choline does notaffect bile duct disease in Mdr2�/� mice. To furtherdissect the differential effects of the studied diets onhepatocytes and cholangiocytes, we next stained liverswith the proliferation marker Ki-67. The amount ofKi-67-positive hepatocytes was significantly reducedin both CDD and PCD compared with controls(Fig. 4), while there were no significant effects on thebile duct level (data not shown). Taken together, thesedata indicate that PCD exhibits its beneficial effects, at

Table 3. Histological scoring of hepatic lesions in Mdr2�/�micefed diets with different choline content

VariableCo(n = 5)

CDD(n = 6)

PCD(n = 6)

Portal inflammation 1 1 1Ductular proliferation 3 3 3Biliary fibrosis 2 2 2Mitotic figures 1 1 1Councilman bodies 1 1 1Total histological score 8 8 8

Scoring was performed as described in ‘Materials and methods’.

CDD, choline-deficient diet; Co, control diet; PCD, phosphatidylcholine-

enriched diet.

Fig. 2. Biliary fibrosis in Mdr2�/�mice is not influenced by dietary choline content. (A) Sirius red staining shows a similar degree offibrosis with periductal collagen fibres (red) in the control diet (Co)-, 4 weeks of choline-deficient diet (CDD) and phosphatidylcholine-enriched diet (PCD)-fed animals (original magnification, 10-fold). (B) Hepatic hydroxyproline (HP) content was determined in liverhomogenates from Co, CDD and PCD groups to quantify fibrosis. As shown, dietary choline manipulations do not alter HP content inMdr2�/�mice. Values are represented as means� SD from five animals per group. bd, bile duct; pv, portal vein; SD, standard deviation.



least in part, at the hepatocellular but not at theductular level.

Effects of choline-deficient diet andphosphatidylcholine-enriched diet on lipidmetabolism in Mdr2�/� mice

Serum cholesterol and TG contents were not affectedby dietary modification of choline content. Moreover,no significant changes were identified in the overall PLcontent of liver, serum and bile (Table 4). PL composi-tion was also not affected as reflected by unchangedserum, hepatic and biliary PC, phosphatidic acid, PE,phosphatidylinositol, Lyso-PC and SM content in bothCDD and PCD groups compared with the controlgroup (Fig. 5A). To further exclude the influence ofdietary choline content on hepatic PC biosynthesis, weanalysed expression levels of key enzymes involved inboth classical and Pemt-mediated synthesis of PC. Inline with an unchanged PC content, hepatic expressionof CKa, CKb, CCT and Pemt did not differ betweenCo, PCD and CDD groups (Fig. 5B). These datatherefore show that modulation of the dietary PLcontent does not influence hepatic PC content as wellas the main PL composition in Mdr2�/�mice.

Discussion

This study aimed to determine the role of hepatic PChomoeostasis in the pathogenesis of liver injury inMdr2�/�mice.

Interestingly, Mdr2�/� mice were able to maintainnormal total hepatic PL levels despite virtually

absent biliary PL excretion (15, 16). Several hypothesescan be generated to explain this intriguing observa-tion: increased PC elimination into serum via lipopro-tein particles could maintain a stable hepatic PCcontent. Alternatively, either inhibition of PC synth-esis or enhanced degradation of PC may account forthe balanced PC content. Interestingly, the serumlipoprotein concentration in Mdr2�/� is even lowercompared with Mdr21/1 mice as a result of a decreasein the concentration of PC-rich high-density lipopro-teins (15, 30). Moreover, the activities of CDP–cholinetransferase (CT) and Pemt, two pathways of PCsynthesis in the liver (10–12), are not altered inMdr2�/� (15). Therefore, we hypothesized that in theabsence of biliary PC excretion without evidence forcompensatory secretion into blood or repression ofendogenous synthesis, degradation of PC could ex-plain normal hepatic PC levels in Mdr2�/� mice.However, as a potential flip side of the coin, formationand accumulation of potentially toxic intermediarymetabolites could contribute to liver injury in Mdr2�/�

mice. Indeed, recent findings from Mdr2�/� Pemt�/�

double knockout mice showed that mice deficient inboth PC export and synthesis via the Pemt-mediatedpathway demonstrate increased PC catabolism byphospholipase A2 (8, 31). Obviously, lack of PCelimination via bile stimulates PC catabolism (8).Therefore, it is tempting to speculate that in situationsof normal choline consumption and preserved synth-esis, but altered elimination, degradation of PC wouldbe increased even further. Choline deficiency, on theother hand, would be expected to be beneficial owingto lower production of potential toxic degradationproducts. To test our hypothesis, we therefore fedMdr2�/�mice control, CDD and PCD.

In line with a previous study (30), serum, hepaticand biliary PC and other PL contents did not differamong the groups, showing that dietary PC enrich-ment does not affect serum PL levels.

Moreover, PL composition remained unaffected asreflected by unchanged serum, hepatic and biliary PC,phosphatidic acid, PE, phosphatidylinositol, Lyso-PCand SM contents in both CDD and PCD groupscompared with the control group. Of note, unchangedLyso-PC content in all study groups indicates unaf-fected phospholipase A2 activity, thus ruling outenhanced degradation of PC by phospholipase A2. Ithas been shown recently that adaptation to cholinedeprivation is not restricted to the liver, but alsoinvolves choline redistribution from non-hepatic tis-sues to the liver (32). Indeed, no changes in expressionlevels of key enzymes involved in both CDP- andPemt-mediated pathways of PC biosynthesis by either

Fig. 3. Cholangiocyte proliferation and phenotype are notaltered by dietary choline content in Mdr2�/�mice. Cytokeratin19 (K19) reflecting cholangiocyte mass and vascular celladhesion molecule-1 (VCAM-1) indicating reactive phenotypesof cholangiocytes were assessed in liver homogenates from allexperimental groups by Western blotting. No changes wereseen after 4 weeks of choline-deficient (CDD) orphosphatidylcholine-enriched (PCD) diets compared withcontrol diet-fed (Co) animals. Densitometry data are expressedas the n-fold change relative to the Co group. Values arepresented as the average obtained from four to five animals pergroup� SD. SD, standard deviation



PCD or CDD rule out compensatory biosyntheticchanges. Most likely, blocked PC excretion into bileand loss via faeces and/or redistribution from othertissues are sufficient to maintain PC levels in the liver.Taken together, these findings suggest that livers fromMdr2�/� mice receiving different dietary choline con-tents are able to maintain stable PC levels (33–35).However, cholesterol and TG also remained un-changed in our experimental groups, also indicatingthat the dietary choline alterations do not affectlipoprotein export.

According to our hypothesis, we anticipated dete-rioration of liver damage in animals receiving PCD.Surprisingly, PCD as well as CDD reduced the LW/BW

ratio. In line with this, PCD resulted in significantlyimproved serum AP and bilirubin levels comparedwith the control group in contrast to CDD, furthersuggesting the potential beneficial effects of PC onliver damage. Additionally, PCD reduced serum ALTsignificantly compared with CDD. Interestingly, portalfibrosis and ductular proliferation as well as bile flowand BA synthesis did not differ between groups, rulingout the potential role of dietary choline in BA-inducedliver damage (6). However, the absence of beneficialPCD effects on liver histology contrasts with recentlyreported beneficial effects of dietary lecithin on liverfibrogenesis in cholic acid (CA)-fed Mdr2�/� mice(36). In this study, improvement of hepatic fibrosis

Table 4. Serum, hepatic and biliary phospholipid composition in Mdr2�/�mice fed diets with different choline content

Variable Co (n = 5) CDD (n = 6) PCD (n = 6)

Serum cholesterol (mg/dl) 63.6� 8.3 65.5� 12 58.50� 6.5Serum triglycerides (mg/dl) 138.6� 12.4 143.5� 14.6 137.5� 14.3Serum PL (mg/dl) 167� 24.5 176.2� 42 205.4� 106.6Hepatic PL (mg/mg LW) 1.68� 0.8 1.87� 0.15 1.19� 0.4Biliary PL (mg/dl) ND ND ND

Data are represented as means� SD.

CDD, choline-deficient diet; Co, control diet; ND, not detected; PCD, phosphatidylcholine-enriched diet; PL, phospholipid; SD, standard deviation.

Fig. 4. Hepatocellular proliferation is reduced by modification of dietary choline content in Mdr2�/�mice. (A) Representativeimmunohistochemical staining for Ki-67 as a cellular proliferation marker (red) in hepatocytes in a control diet (Co)-, choline-deficientdiet (CDD)- and phosphatidylcholine-enriched diet (PCD)-fed Mdr2�/�mouse is shown. Numerous Ki-67-positive hepatocytes are seenin Mdr2�/� receiving the control (Co) diet (arrowheads), while PCD as well as CDD reduce the number of proliferating hepatocytes. (B)The number of Ki-67-positive hepatocytes was counted in 30 high-power fields per animal. Both CDD and PCD reduce the number ofproliferating hepatocytes significantly. Values are mean� SD from five animals in the control group and six animals in the CDD andPCD groups. Po 0.05: �CDD and PCD vs. Co.



was only observed in CA-fed Mdr2�/�mice receiving alecithin-enriched diet, but not in the animals receivinga lecithin diet without CA as an aggravating factor. Apossible explanation for these discrepant effects couldbe that addition of lecithin together with CA leads tobetter packing of toxic CA into micelles and thiscontributes to less toxicity of ‘micelled’ CA comparedwith animals fed CA without lecithin. Another expla-nation would be the use of soybean lecithin withdifferent contents (78% triacylglycerols and 22% PL,of which 11% are PC in the study by Lamireau et al.(36) vs. 97% PL in total and 24% PC). The beneficialeffects on hepatic fibrosis and ductular proliferation ofsoybean lecithin may therefore be mediated by othercompounds, but not PC directly. In line with thesefindings, both bile duct mass (as reflected by un-changed hepatic CK19 protein levels) as well as un-changed activated cholangiocyte phenotype (reflectedby VCAM-1 expression) argue for a lack of a dietarycholine effect on Mdr2�/� associated cholangiopathyin our experiments. In contrast, the number of Ki-67-positive hepatocytes was significantly less in PCD,which correlates with decreased serum AP, indicating

some beneficial effects on hepatocytes. The hepato-protective effects of PC may be mediated by anincreased anti-oxidative response. S-adenosylmethio-nine (SAMe) is utilized by a trans-sulphuration path-way to replenish the cellular glutathione (37, 38). Theformation of PC requires SAMe for methylation.Therefore, PC supplementation would decrease SAMeutilization for PC synthesis and restore cellular SAMe.Additionally, PC implementation into cellular mem-branes, as shown in experiments with choline-defi-cient mice, may lead to increased membrane stability(31). However, hepatic PC measurement and TLCanalysis of hepatic lipid extracts did not show differ-ences in PC content among our study groups, thusruling out the hepatoprotective effects of PC viaincreased PC implementation into the cellular mem-brane in our model.

Interestingly, CDD reduced the LW/BW ratio sig-nificantly compared with control diet-fed animals.However, serum parameters of liver injury and choles-tasis (ALT, AP and total bilirubin) were not improvedby CDD, indicating that dietary choline deprivation isnot beneficial in the treatment of liver disease in

Fig. 5. Dietary choline alterations do not influence phospholipid composition in the bile, liver and serum of Mdr2�/�mice. (A) Thin-layer chromatographical analysis of phospholipid species of pooled lipid extracts from the livers, serum and bile of Mdr2�/�mice wereanalysed. After 4 weeks of feeding, neither the choline-deficient diet (CDD) nor the phosphatidylcholine-enriched diet (PCD) alteredthe main hepatic, serum and biliary PL species compared with the control (Co) diet-fed animals. (B) The expression of key enzymesinvolved in PC biosynthesis in the liver is not affected by manipulation of dietary choline content in Mdr2�/�mice. Expression levels arenormalized to 18s gene expression, and levels of Co are expressed as 100%. Values are presented as means� SD from five animals inthe control group and six animals in the CDD and PCD groups. CCT, choline-phosphate cytidyltransferase; CKab, choline kinase a,choline kinase b; Lyso-PC, lysophosphatidylcholine; Lyso-PI, lysophosphatidylinositol; PA, phosphatidic acid; PC, phosphatidylcholine;PE, phosphatidylethanolamine; Pemt, phosphatidylethanolamine N-methyltransferase; PI, phosphatidylinositol; PS, phosphatidylserine;SM, sphingomyelin.



Mdr2�/� mice. Moreover, hepatocellular proliferation(reflected by Ki-67 staining) was reduced by CDD,reflecting most likely the influence of choline depriva-tion on proliferation, but not hepatocellular damage.Findings on hepatocellular proliferation influenced bycholine may be important in other clinical settings.Total parenteral nutrition (TPN) solutions, containingfat, are believed to be superior to fat-free TPN for liverregeneration after partial hepatectomy (39–41). How-ever, our results suggest that manipulation of cholinecontent in patients receiving TPN after partial hepa-tectomy may be critical in liver growth and regenera-tion; therefore, underlying mechanisms mediatingreduced proliferation in both PCD and CDD need tobe studied in detail. Furthermore, a better under-standing of hepatocellular proliferation in Mdr2�/�

mice, influenced by dietary choline, could help toprevent hepatocellular carcinomas development inthese animals. However, these intriguing conceptsneed to be explored in future studies, which arebeyond the scope of our current work.

In conclusion, we show that dietary manipulation ofPC synthesis did not affect the PL content andcomposition in the liver, bile and serum of Mdr2�/�

mice, indicating that abnormalities in hepatocellularPL metabolism may not be major pathogenic oraggravating factors. In addition, our findings furthersuggest the potential importance of choline metabo-lism for hepatocellular proliferation, which deservesfurther studies.

Acknowledgements

We gratefully acknowledge Dr W. Erwa (Graz) and collea-

gues for performing biochemical analyses of serum liver

tests and Ms Susanne Iten (Cham) for providing us with

soybean lecithin granules.

Grants: This work was supported by the Austrian

Science Foundation, grant number: P18613-B05 (to M. T.).

References

1. Trauner M, Boyer JL. Bile salt transporters: molecular

characterization, function, and regulation. Physiol Rev 2003;

83: 633–71.

2. Trauner M, Meier PJ, Boyer JL. Molecular pathogenesis of

cholestasis. N Engl J Med 1998; 339: 1217–27.

3. Ruetz S, Gros P. Phosphatidylcholine translocase: a physiolo-

gical role for the mdr2 gene. Cell 1994; 77: 1071–81.

4. Trauner M, Fickert P, Wagner M. MDR3 (ABCB4) defects: a

paradigm for the genetics of adult cholestatic syndromes.

Semin Liver Dis 2007; 27: 77–98.

5. Smit JJ, Schinkel AH, Oude Elferink RP, et al. Homozygous

disruption of the murine mdr2 P-glycoprotein gene leads to a

complete absence of phospholipid from bile and to liver

disease. Cell 1993; 75: 451–62.

6. Fickert P, Fuchsbichler A, Wagner M, et al. Regurgitation of

bile acids from leaky bile ducts causes sclerosing cholangitis

in Mdr2 (Abcb4) knockout mice. Gastroenterology 2004; 127:

261–74.

7. De Vree JM, Jacquemin E, Sturm E, et al. Mutations in the

MDR3 gene cause progressive familial intrahepatic cholesta-

sis. Proc Natl Acad Sci USA 1998; 95: 282–7.

8. Li Z, Agellon LB, Vance DE. Phosphatidylcholine home-

ostasis and liver failure. J Biol Chem 2005; 280: 37798–802.

9. Kennedy EP. The synthesis of cytidine diphosphate choline,

cytidine diphosphate ethanolamine, and related compounds.

J Biol Chem 1956; 222: 185–91.

10. Kennedy EP. Metabolism of lipides. Annu Rev Biochem 1957;

26: 119–48.

11. Walkey CJ, Yu L, Agellon LB, Vance DE. Biochemical and

evolutionary significance of phospholipid methylation. J Biol

Chem 1998; 273: 27043–46.

12. Vance DE, Ridgway ND. The methylation of phosphatidy-

lethanolamine. Prog Lipid Res 1988; 27: 61–79.

13. Fickert P, Wagner M, Marschall HU, et al. 24-norUrsodeoxy-

cholic acid is superior to ursodeoxycholic acid in the treat-

ment of sclerosing cholangitis in Mdr2 (Abcb4) knockout

mice. Gastroenterology 2006; 130: 465–81.

14. Elferink RP, Ottenhoff R, Van Marle J, Frijters CM, Smith AJ,

Groen AK. Class III P-glycoproteins mediate the formation of

lipoprotein X in the mouse. J Clin Invest 1998; 102: 1749–57.

15. Voshol PJ, Havinga R, Wolters H, et al. Reduced plasma

cholesterol and increased fecal sterol loss in multidrug

resistance gene 2 P-glycoprotein-deficient mice. Gastroenter-

ology 1998; 114: 1024–34.

16. Oude Elferink RP, Ottenhoff R, Van Wijland M, Smit JJ, Schinkel

AH, Groen AK. Regulation of biliary lipid secretion by mdr2

P-glycoprotein in the mouse. J Clin Invest 1995; 95: 31–8.

17. Angelico M, Alvaro D, Cantafora A, et al. Impaired hepatic

handling and processing of lysophosphatidylcholine in rats

with liver cirrhosis. Gastroenterology 1991; 101: 228–37.

18. Fickert P, Zollner G, Fuchsbichler A, et al. Ursodeoxycholic

acid aggravates bile infarcts in bile duct-ligated and Mdr2

knockout mice via disruption of cholangioles. Gastroenterol-

ogy 2002; 123: 1238–51.

19. Van Nieuwkerk CM, Elferink RP, Groen AK, et al. Effects of

Ursodeoxycholate and cholate feeding on liver disease in FVB

mice with a disrupted mdr2 P-glycoprotein gene. Gastro-

enterology 1996; 111: 165–71.

20. Fickert P, Trauner M, Fuchsbichler A, Stumptner C, Zatlou-

kal K, Denk H. Cytokeratins as targets for bile acid-induced

toxicity. Am J Pathol 2002; 160: 491–9.

21. Van Nieuwerk CM, Groen AK, Ottenhoff R, et al. The role of

bile salt composition in liver pathology of mdr2 (� /� )

mice: differences between males and females. J Hepatol 1997;

26: 138–45.



22. Cho JJ, Hocher B, Herbst H, et al. An oral endothelin-A

receptor antagonist blocks collagen synthesis and deposition

in advanced rat liver fibrosis. Gastroenterology 2000; 118:

1169–78.

23. Bradford MM. A rapid and sensitive method for the quanti-

tation of microgram quantities of protein utilizing the

principle of protein-dye binding. Anal Biochem 1976; 72:

248–54.

24. Trauner M, Arrese M, Soroka CJ, et al. The rat canalicular

conjugate export pump (Mrp2) is down-regulated in intra-

hepatic and obstructive cholestasis. Gastroenterology 1997;

113: 255–64.

25. Bligh EG, Dyer WJ. A rapid method of total lipid extrac-

tion and purification. Can J Biochem Physiol 1959; 37:

911–7.

26. White T, Bursten S, Federighi D, Lewis RA, Nudelman E.

High-resolution separation and quantification of neutral

lipid and phospholipid species in mammalian cells and sera

by multi-one-dimensional thin-layer chromatography. Anal

Biochem 1998; 258: 109–17.

27. Wagner M, Fickert P, Zollner G, et al. Role of farnesoid X

receptor in determining hepatic ABC transporter expression

and liver injury in bile duct-ligated mice. Gastroenterology

2003; 125: 825–38.

28. Popov Y, Patsenker E, Fickert P, Trauner M, Schuppan D.

Mdr2 (Abcb4)� /� mice spontaneously develop severe

biliary fibrosis via massive dysregulation of pro- and anti-

fibrogenic genes. J Hepatol 2005; 43: 1045–54.

29. Lazaridis KN, Strazzabosco M, Larusso NF. The cholangio-

pathies: disorders of biliary epithelia. Gastroenterology 2004;

127: 1565–77.

30. Voshol PJ, Koopen NR, De Vree JM, et al. Dietary cholesterol

does not normalize low plasma cholesterol levels but induces

hyperbilirubinemia and hypercholanemia in Mdr2 P-glyco-

protein-deficient mice. J Hepatol 2001; 34: 202–9.

31. Li Z, Agellon LB, Allen TM, et al. The ratio of phosphatidyl-

choline to phosphatidylethanolamine influences membrane

integrity and steatohepatitis. Cell Metab 2006; 3: 321–31.

32. Li Z, Agellon LB, Vance DE. Choline redistribution during

adaptation to choline deprivation. J Biol Chem 2007; 282:

10283–9.

33. Agellon LB, Walkey CJ, Vance DE, Kuipers F, Verkade HJ. The

unique acyl chain specificity of biliary phosphatidylcholines

in mice is independent of their biosynthetic origin in the

liver. Hepatology 1999; 30: 725–9.

34. Kulinski A, Vance DE, Vance JE. A choline-deficient diet in

mice inhibits neither the CDP–choline pathway for phospha-

tidylcholine synthesis in hepatocytes nor apolipoprotein B

secretion. J Biol Chem 2004; 279: 23916–24.

35. Waite KA, Cabilio NR, Vance DE. Choline deficiency-in-

duced liver damage is reversible in Pemt(� /� ) mice. J Nutr

2002; 132: 68–71.

36. Lamireau T, Bouchard G, Yousef IM, et al. Dietary lecithin

protects against cholestatic liver disease in cholic acid-fed

Abcb4-deficient mice. Pediatr Res 2007; 61: 185–90.

37. Lieber CS. Liver diseases by alcohol and hepatitis C: early

detection and new insights in pathogenesis lead to improved

treatment. Am J Addict 2001; 10(Suppl.): 29–50.

38. Lu SC. Regulation of hepatic glutathione synthesis: current

concepts and controversies. FASEB J 1999; 13: 1169–83.

39. Delegge MH, Basel MD, Bannister C, Budak AR. Parenteral

nutrition (PN) use for adult hospitalized patients: a study of

usage in a tertiary medical center. Nutr Clin Pract 2007; 22:

246–9.

40. Grau T, Bonet A, Rubio M, et al. Liver dysfunction associated

with artificial nutrition in critically ill patients. Crit Care

2007; 11: R10.

41. Wessel JJ, Kocoshis SA. Nutritional management of infants

with short bowel syndrome. Semin Perinatol 2007; 31:

104–11.



Role of hepatic phospholipids in development of liver injury in Mdr2 (Abcb4) knockout mice: Phospholipids and liver injury in Mdr2 knockout mice

Documents