A distinct microarray gene expression profile in primary rat hepatocytes incubated with ursodeoxycholic acid

A distinct microarray gene expression profile in primary rat hepatocytes incubated with ursodeoxycholic acid

Journal of Hepatology 42 (2005) 897–906 www.elsevier.com/locate/jhep A distinct microarray gene expression profile in primary rat hepatocytes incubat...

171KB Sizes 0 Downloads 7 Views

Journal of Hepatology 42 (2005) 897–906 www.elsevier.com/locate/jhep

A distinct microarray gene expression profile in primary rat hepatocytes incubated with ursodeoxycholic acid* Rui E. Castro1,2, Susana Sola´1, Xiaoming Ma2, Rita M. Ramalho1, Betsy T. Kren2, Clifford J. Steer2,3, Cecı´lia M.P. Rodrigues1,* 1

Centro de Patoge´nese Molecular, Faculty of Pharmacy, University of Lisbon, Lisbon 1600-083, Portugal 2 Department of Medicine, University of Minnesota Medical School, Minneapolis, MN 55455, USA 3 Department of Genetics, Cell Biology, and Development, University of Minnesota Medical School, Minneapolis, MN 55455, USA

Background/Aims: Ursodeoxycholic acid (UDCA) and its taurine-conjugated derivative, TUDCA, modulate cell death and cell cycle regulators, such as E2F-1 and p53. However, precise pathways underlying UDCA’s effects are not fully understood. The aim of this study was to identify specific cellular targets of UDCA. Methods: The expression profile of primary rat hepatocytes incubated with UDCA was determined using Affymetrix GeneChip Rat 230A arrays. Hybridization data were processed to identify genes with significant expression changes. RT-PCR and immunoblot analyses of a selected target confirmed microarray data. Results: The results showed that O440 genes were modulated with UDCA by O1.5-fold; w25% were significantly different from controls. Genes affected by UDCA included new regulatory molecules, such as Apaf-1. RT-PCR and immunoblotting confirmed a decrease in Apaf-1. Other altered genes were directly involved in cell cycle (cyclin D1, cadherin 1, HMG-box containing protein 1) and apoptosis (prothymosin-a) events. The E2F-1/p53/Apaf-1 pathway appears to be targeted by UDCA. Finally, transcripts for proteins with kinase activity and transcription factors were specifically modulated by TUDCA. Conclusions: This study expands our knowledge of the biological effects of UDCA in hepatocytes. Most of the identified genes represent novel potential targets of UDCA, which may ultimately explain its therapeutic properties. q 2005 Published by Elsevier B.V. on behalf of the European Association for the Study of the Liver. Keywords: Apaf-1; Bile acids; Cyclin D1; DNA microarrays; Liver; p53 1. Introduction In contrast to the toxic effects of several hydrophobic bile acids, ursodeoxycholic acid (UDCA) Received 26 September 2004; received in revised form 9 January 2005; accepted 15 January 2005; available online 7 April 2005 * Presented, in part, at the annual meeting of the American Association for the Study of Liver Diseases, Boston, 2003. * Corresponding author. Tel.: C351 21 794 6400; fax: C351 21 794 6491. E-mail address: [email protected] (C.M.P. Rodrigues). Abbreviations: Apaf-1, apoptotic protease activating factor 1; Cdh1, cadherin 1; EST, expressed sequence tag; FBS, fetal bovine serum; Hbp1, HMG-box containing protein 1; ProT, prothymosin-a; RT-PCR, reverse transcriptase-polymerase chain reaction; TGF-b1, transforming growth factor b1; TUDCA, tauroursodeoxycholic acid; UDCA, ursodeoxycholic acid.

has been used over the last several decades for the treatment of certain liver diseases [1]. More recently, we have shown that UDCA and its conjugated derivative, tauroursodeoxycholic acid (TUDCA), play a unique role in modulating the apoptotic threshold in several different cell types. In fact, we initially demonstrated that UDCA inhibited apoptosis in vivo, in part, by preventing translocation of the pro-apoptotic protein Bax to mitochondria [2]. These studies were subsequently extended and showed that UDCA modulates apoptosis in both hepatic and non-hepatic cells, in response to a variety of agents, including ethanol, transforming growth factor b1 (TGF-b1), anti-Fas antibody, and okadaic acid suggesting a basic mechanism common to each of the different apoptotic pathways [3]. UDCA, as well as its taurine and glycine conjugates, prevent the opening of

0168-8278/$30.00 q 2005 Published by Elsevier B.V. on behalf of the European Association for the Study of the Liver. doi:10.1016/j.jhep.2005.01.026


R.E. Castro et al. / Journal of Hepatology 42 (2005) 897–906

the permeability transition pore, collapse of the transmembrane potential, and release of cytochrome c [4,5]. Therefore, the mitochondrial pathway of apoptosis appears to be a major regulatory target for this bile acid. Nevertheless, it appears that UDCA can also partially modulate death receptor-mediated cell death [6] and prevent endoplasmic reticulum-stress mediated apoptosis [7]. Finally, we have recently shown that UDCA can specifically modulate the E2F-1/p53/Bax cell death pathway, abrogating E2F-1-induced p53 and p53-associated Bax expression independently of its effect on mitochondria and/or caspases [8]. Certain bile acids may also prevent and counteract their inherent cytotoxicity by activating cell survival pathways [9–11]. These pathways include the nuclear factor kB, protein kinase C, phosphatidylinositol 3-kinase and mitogen-activated protein kinase pathways, calciumdependent signal transduction cascades, or even cyclic adenosine monophosphate. Curiously, UDCA has also been suggested to protect against the development of human colon cancer in several pre-malignant conditions [12,13], and to have a chemopreventive role in a rodent model of colonic carcinogenesis [14]. Other authors have suggested that UDCA may inhibit the growth of tumors when associated with photodynamic therapy in a murine tumor model [15]. Thus, it is possible that UDCA may act differentially on death and survival pathways, depending on the physiological conditions of the cell and/or stimulus. In this study, we performed a comprehensive analysis of gene expression modulated by UDCA incubation of primary rat hepatocytes. Using microarray technology, we analyzed the level of expression from essentially every gene in the rat genome. In fact, UDCA has been shown to up-regulate selective targets at the transcription level contributing to its anti-apoptotic effects [8]. In contrast, UDCA also appears to be a strong modulator of survival transduction pathways, by posttranscriptional mechanisms which are not identified in this approach. Nevertheless, genomic analysis by this technology allows one to detect differential expression of mRNAs and permits rapid screening of hundreds to thousands of genes. The expression patterns generated from parallel analysis of the whole genome can then provide clues to the functions of previously uncharacterized genes, as well as provide information about how drugs achieve their therapeutic effect. In the present study, highly parallel screening of gene expression demonstrated novel features of UDCAinduced transcriptional profiles in primary rat hepatocytes, while confirming others. The identification of specific cellular targets regulated by UDCA may prove useful to our understanding of UDCA mechanistic actions and to the recognition of prospective therapeutic targets.

2. Materials and methods 2.1. Hepatocyte isolation Rat primary hepatocytes were isolated from male Sprague-Dawley rats (100–150 g) by collagenase perfusion as described previously [16]. In brief, hepatocyte suspensions were obtained by passing collagenase digested livers though 125 mm gauze and washing cells in William’s E medium (Invitrogen Corp., Grand Island, NY) supplemented with 26 mM sodium bicarbonate, 23 mM HEPES, 0.01 units/ml insulin, 2 mM L-glutamine, 10 nM dexamethasone, 5.5 mM glucose, 100 units/ml penicillin, 100 units/ ml streptomycin (Life Technologies, Gaithersburg, MD), and 20% heatinactivated fetal bovine serum (FBS; Atlanta Biologicals Inc., Norcross, GA). Cell viability was determined by trypan blue exclusion and was typically 85–90%.

2.2. Cell cultures and treatment After isolation, hepatocytes were resuspended in complete William’s E medium and plated on Primariae tissue culture dishes (BD Biosciences, San Jose, CA) at a density of 5!103 cells/cm2. The cells were maintained at 37 8C in a humidified atmosphere of 5% CO2 for 3 h, to allow attachment. Plates were then incubated in William’s E medium containing 10% heatinactivated FBS, and treated with either 100 mM UDCA, 100 mM TUDCA (Sigma-Aldrich, St. Louis, MO), or no addition (control), for 12, 24, 36 and 48 h before harvesting for RNA and protein isolation. Preliminary studies have shown that labeled UDCA rapidly crosses the plasma membrane of rat hepatocytes in cell culture (unpublished data). In addition, lower concentrations of this bile acid are thought to be equally effective at modulating the apoptotic threshold [3]. In parallel studies, human hepatoma HuH-7 cells were maintained in Dulbecco’s modified Eagle’s Medium (DMEM; Invitrogen Corp.) supplemented with 100 U/ml penicillin, 100 U/ml streptomycin, and 10% FBS. Cells were plated at a density of 5!103 cells/cm2 and maintained at 37 8C in a humidified atmosphere of 5% CO2 for 12 h. UDCA or TUDCA were then added to cells for 12, 24, 36 and 48 h before harvesting for protein isolation. We have previously shown that UDCA is effective at inhibiting apoptosis in HuH-7 cells [3].

2.3. Isolation of mRNA Primary rat hepatocytes incubated with UDCA for 24 h or controls were used in the microarray analysis. This time point was chosen based on previous studies in our laboratory that showed robust cell death protection in cultures after 24 h UDCA treatment [8]. In fact, primary hepatocytes treated with UDCA for 6 h revealed fewer transcript changes (data not shown). Parallel experiments were performed after incubation of primary rat hepatocytes with TUDCA for 24 h. For isolation of mRNA from hepatocytes, the mTRAP Midi mRNA isolation kit (Active Motif, Carlsbad, CA) was used according to manufacturer instructions. All RNA samples showed A260/280 ratios between 1.9 and 2.1. The integrity of mRNA was further confirmed in each case using the Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA).

2.4. Preparation of labeled cRNA and array hybridization All procedures were performed according to the instructions described in the Affymetrix GeneChip Expression Analysis Manual (Affymetrix, Santa Clara, CA). Briefly, 0.5 mg of mRNA was converted to double-stranded cDNA using the Superscript Reverse Transcriptase (Invitrogen Corp.) and an oligo(dT) primer linked to the T7 RNA polymerase-binding site sequence (Integrated DNA Technologies, Coralville, IA). Purified cDNA was then converted to labeled cRNA using T7 RNA polymerase in the presence of biotinylated UTP and CTP (Enzo Diagnostics, Farmingdale, NY). The labeled cRNA was purified with RNeasy columns (Qiagen, Valencia, CA) and fragmented in buffer (40 mM Tris/acetate, pH 8.1, 100 mM KOAc, and 30 mM MgOAc) for 35 min at 94 8C. An aliquot of each sample was first hybridized to an Affymetrix Test 2 Array to determine sample quality according to manufacture’s criteria.

R.E. Castro et al. / Journal of Hepatology 42 (2005) 897–906 All preparations met Affymetrix’s recommended criteria for use on expression arrays. Fragmented cRNAs (25 mg/probe array) were hybridized to a set of nine Affymetrix GeneChip Rat 230A arrays, each containing probes for about 16,000 genes using the protocol described in the Affymetrix Expression Analysis Technical Manual. Briefly, hybridization was performed at 45 8C overnight, followed by washing, staining, signal amplification with biotinylated antistrepavidin antibody, and final staining. Data were acquired with the Affymetrix Microarray Suite Software (MAS) version 5.0.

2.5. Data filtering and analysis For the microarray studies, processing of the raw data involved normalization, filtering, and analysis to identify reproducible and reliable patterns of data. The 9 .CEL files (three replicate hybridizations from three different animals were performed to establish the reproducibility of the results) generated by the Affymetrix MAS version 5.0 were converted into .DCP files using the dChip 1.3 Software (http://www. dChip.org) [17,18]. The 9 .DCP files were normalized, and raw gene expression data were generated using the dChip system of model-based analysis. The purpose of normalization is to identify and remove systematic sources of variation in intensity values to allow betweenarray comparisons. For global gene expression profiles, untreated cells were designated ‘baseline’ (B) and UDCA- or TUDCA-treated cells were designated ‘experiment’ (E) in the dChip comparison software. Genes differentially expressed between the controls and UDCA- or TUDCA-treated samples were identified by a series of filtering steps. Filtering reduced data collection by removing uninformative genes whose expression levels did not change or were below a user-defined threshold. In this case, only genes that were over- and under-expressed O and !1.5-fold, respectively, in UDCA- or TUDCA-treated cells versus controls were identified. Furthermore, each candidate gene required a probability of differential expression of P!0.05 from a ttest and a 90% lower confidence bound for the fold change expression O1.5.

2.6. Reverse transcriptase-polymerase chain reaction (RT-PCR) The semi-quantitative RT-PCR method was used to confirm the differential gene expression of apaf-1, a transcript flagged by the Affymetrix analysis. Total RNA was extracted from rat primary hepatocytes using the TRIZOL reagent (Invitrogen Corp.). For RT-PCR, 5 mg of total RNA was reverse-transcribed using oligo(dT) (Integrated DNA Technologies Inc.) and ImProm II reverse transcriptase (Promega, Madison, WI). Specific oligonucleotide primer pairs were incubated with cDNA template for PCR amplification using the Expand High Fidelity PCR System (Roche Applied Science, Indianapolis, IN). The following sequences were used as primers: apaf-1 sense: 5 0 -CCTTCCTCT TGTGTCTTCTTCCAG-3 0 ; apaf-1 antisense 5 0 -TTCTGCTGAATCGC ACTGACC-3 0 ; b-actin sense 5 0 -TGCCCATCTATGAGGGTTACG-3 0 ; and b-actin antisense 5 0 -TAGAAGCATTTGCGGTGCACG-3 0 . The product of the b-actin RNA served as control.

2.7. Immunoblotting Steady-state levels of Apaf-1 were determined by Western blot analysis. Briefly, 100 mg of total protein were extracted from primary rat hepatocytes or HuH-7 cells, and the extracts separated on a 12% sodium dodecyl sulfate-polyacrylamide electrophoresis gel. Following electrophoretic transfer onto nitrocellulose membranes and blocking with 5% milk solution, the blots were incubated with a primary rabbit polyclonal antibody to Apaf-1 (Chemicon International Inc., Temecula, CA), and finally with a secondary antibody conjugated with horseradish peroxidase (Bio-Rad Laboratories, Hercules, CA). The membranes were processed for protein detection using Super Signale substrate (Pierce Biotechnology, Inc., Rockford, IL). b-Actin was used as a loading control.


3. Results 3.1. Modulation of global hepatocyte gene expression by UDCA We determined the global profile of genes regulated by UDCA by using the Affymetrix GeneChipw Rat Expression Array 230A, consisting of w16,000 transcripts and variants. cRNA prepared from control cells was used for comparative analysis. The relative levels of gene expression after 24 h incubation of hepatocytes with 100 mM UDCA were compared by plotting the average difference between treated and non-treated cells, and determining the fold change in gene expression. Approximately 441 genes (2.76%) exhibited alterations in expression levels following UDCA treatment, with a magnitude of change O1.5-fold. Among these, approximately 25% (96 genes) fulfilled the filtering criteria for detection, in at least one of the arrays. Of these 96 genes, 28 were found to be up-regulated and 68 were down-regulated (Table 1). Differentially expressed genes were categorized based on the best available information regarding their biological functions. Genes with multiple functions were assigned to a single category. In parallel experiments with TUDCA incubation, approximately 50% of the 96 genes listed in Table 1 were similarly regulated. These included many genes of apoptosis and cell cycle control families, as well as almost all genes coding for enzymes. Other transcripts, such as those for proteins with kinase activity and transcription factors were specifically modulated by TUDCA. Our evaluation also focused on the specificity and sensitivity of the microarray analysis. Hierarchical clustering was performed using the subset of 96 genes with the greatest variation between controls and the UDCA-treated samples. As expected, all three controls clustered with remarkable identity and separately from the three UDCAtreated cells on the dendogram. A graphical representation of the cluster analysis is shown in Fig. 1, where red represents increased relative expression and blue decreased relative expression. The clustering algorithm is as follows: distance between two genes is defined as 1Kr, where r is the correlation coefficient between the standardized expression values of two genes. Two closest genes are merged into a supergene and connected by branches, with length representing distance. The expression value of this supergene is the average of standardized expression values of the two genes (centroid linkage). The next pair of genes (supergene) with the smallest distance is merged, and the process is repeated to merge all genes [18]. 3.2. Confirmation of microarray data Validation of microarray data was provided by concordance of the expression profiles for several transcripts with previous reports in the literature. In contrast, although


R.E. Castro et al. / Journal of Hepatology 42 (2005) 897–906

Table 1 Genes modulated by UDCA and TUDCA in primary rat hepatocytes Affymetrix tag number

Gene name

Amino acid metabolism 1367838_at CTL target antigen 1398286_at Cysteine-sulfinate decarboxylase 1368092_at Fumarylacetoacetate hydrolase 1370200_at Glutamate dehydrogenase 1 1369671_at Ornithine carbamoyltransferase 1368720_at Tryptophan 2,3-dioxygenase Androgen and estrogen metabolism 1387759_s_at UDP glycosyltransferase 1 family, polypeptide A6 Apoptosis 1368330_at Apoptosis antagonizing transcription factor 1369197_at Apoptotic protease activating factor 1 1387343_at CCAAT/enhancer binding, protein (C/EBP)d 1368490_at CD14 antigen 1368174_at EGL nine homolog 3 (C. elegans) 1389177_at ESTs, Highly similar to p53 apoptosis effector related to Pmp22; p53 apoptosisassociated target [Mus musculus] [M. musculus] 1384427_at ESTs, Highly similar to A42772 mdm2 protein - rat (fragments) [R.norvegicus] 1371913_at ESTs, Moderately similar to transforming growth factor, b induced, 68 kDa [Mus musculus] [M.musculus] 1373051_at Liver regeneration p-53 related protein 1368514_at Monoamine oxidase B 1368016_at Perosisomal 2-enoyl-CoA reductase 1370156_at Prion protein, structural 1370243_a_at Prothymosin a 1369943_at Tissue-type transglutaminase Bile acid metabolism and transport 1398310_at Aldo-keto reductase family 1, member D1 (d 4-3-ketosteroid-5-b-reductase) 1368769_at ATP-binding cassette, sub-family B (MDR/TAP), member 11 Carbohydrate metabolism 1369502_a_at Amylase 1 1387203_at Glucokinase regulatory protein Cell cycle control 1386947_at Cadherin 1 1383075_at Cyclin D1 1369590_a_at DNA-damage inducible transcript 3 1368549_at HMG-box containing protein 1 1387402_at Myosin, heavy polypeptide 9, non-muscle Cell stress 1370892_at Complement component 4 1370026_at Crystallin, a B 1369202_at Myxovirus (influenza virus) resistance 2 1386900_at Ribosome associated membrane protein 4 Cell-cell/matrix interactions 1370310_at 3-Hydroxy-3-methylglutaryl-Coenzyme A synthase 2 1398294_at Actinin, a 1 1374976_a_at Acyl-coenzyme A:cholesterol acyltransferase 1387358_at ADP-ribosylation factor-like 1 1387811_at Angiotensinogen

Gene symbol

UDCA Average (fold change)

TUDCA P value

Average (fold change)

P value

Cth Csad Fah Glud1 Otc Tdo2

– – – – K2.420 –

– – – – 0.037 –




Aatf Apaf-1 Cebpd Cd14 Egln3 EST

– K2.070 2.675 1.795 K1.950 K1.885

– 0.021 0.024 0.045 0.047 0.031

1.735 K1.613 2.088 – – –

0.034 0.044 0.041 – – –











LOC246046 Maob Pecr Prnp Ptma Tgm2

– K1.745 – K2.505 2.655 K2.340

– 0.045 – 0.020 0.026 0.020

1.670 K1.905 K1.815 – – K2.233

0.041 0.043 0.021 – – 0.017

K1.990 K2.905 K1.860 K2.015 K2.448 K3.620 –

0.017 0.033 0.025 0.022 0.033 0.010 –







Amy1 Gckr

– K2.340

– 0.036

K1.970 K2.378

0.014 0.026

Cdh1 Ccnd1 Ddit3 Hbp1 Myh9

K2.355 K3.400 1.745 1.790 K1.655

0.033 0.013 0.027 0.045 0.037

– K2.158 – 1.733 –

– 0.040 – 0.043 –

C4-2C4a Cryab Mx2 RAMP4

– K4.840 – 2.645

– 0.037 – 0.028

K1.791 – K4.305 2.197

0.033 – 0.022 0.028




Actn1 Soat1

K3.400 2.760

0.021 0.033

K2.421 –

0.018 –

2.115 –

0.034 –

2.103 K2.066

0.035 0.027

Arl1 Agt

(continued on next page)

R.E. Castro et al. / Journal of Hepatology 42 (2005) 897–906


Table 1 (continued) Affymetrix tag number


Gene name

Blocked early in transport 1 homolog (S. cerevisiae) 1369224_at Cadherin 17 1387259_at Cadherin 2 1368379_at CD36 antigen (collagen type I receptor, thrombospondin receptor)-like 2 1367631_at Connective tissue growth factor 1370234_at Fibronectin 1 1387877_at Formiminotransferase cyclodeaminase 1371575_at Moesin 1398866_at Scaffolding protein SLIPR 1367570_at Transgelin (Smooth muscle 22 protein) Cholesterol biosynthesis and metabolism 1370310_at 3-Hydroxy-3-methylglutaryl-Coenzyme A synthase 2 1374976_a_at Acyl-coenzyme A:cholesterol acyltransferase 1368335_at Apolipoprotein A-1 Electron transport 1368155_at Cytochrome P450 15-b gene 1387243_at Cytochrome P450, 1a2 1387583_at Cytochrome P450, 26, retinoic acid 1370241_at Cytochrome P450, 2c39 1387914_at Cytochrome P450, family 27, subfamily a, polypeptide 1 1369424_at Cytochrome P450, subfamily 2A, polypeptide 1 1368608_at Cytochrome P450, subfamily 2F, polypeptide 1 1368467_at Cytochrome P450, subfamily IVF, polypeptide 14 (leukotriene B4 q hydroxylase) 1370475_at Cytochrome P450IIB3 1372476_at Putative fatty acid desaturase Enzymes 1370708_a_at 3-a-Hydroxysteroid dehydrogenase 1368794_at 3-Hydroxyanthranilate 3,4-dioxygenase 1368091_at 5-Oxoprolinase (ATP-hydrolysing) 1369630_at Adenosin kinase 1368021_at Alcohol dehydrogenase 1 1367794_at a-2-macroglobulin 1387825_at Androsterone UDP-glucuronosyltransferase 1368266_at Arginase 1 1387860_at Calpain 2 1370151_at Carboamyl-phosphate synthetase 1 1370936_at Dimethylglycine dehydrogenase precursor 1368163_at Dipeptidyl peptidase 4 1368491_at DNaseII-like acid DNase 1387314_at Dopa/tyrosine sulfotransferase 1370047_at Ectonucleotide pyrophosphatase/phosphodiesterase 1 1368304_at Flavin-containing monooxygenase 3 1369921_at Glutathione S-transferase Yb4 gene 1388122_at Glutathione S-transferase, pi 2 1368253_at Guanidinoacetate methyltransferase 1369225_at K-kininogen, differential splicing leads to HMW Kngk 1387223_at Kynurenine aminotransferase 2 1387090_a_at LIM motif-containing protein kinase 2 1371098_a_at MASP-2 protein

Gene symbol

UDCA Average (fold change)


TUDCA P value

Average (fold change)



Cdh17 Cdh2 Scarb2

K1.930 – –

0.018 – –

K1.785 K1.733 1.762

0.030 0.019 0.029

Ctgf Fn1 Ftcd Msn Slipr Tagln

K12.390 – – K1.985 K2.155 K4.950

0.042 – – 0.034 0.021 0.020

K11.025 K1.751 K2.917 – – K3.223

0.012 0.045 0.020 – – 0.033









0.023 – 0.047 0.034 –

K1.985 K2.020 – – K1.675

0.032 0.018 – – 0.031


Cyp2c12 Cyp1a2 Cyp26 Cyp2c39 Cyp27a1

K2.955 – 2.830 K1.665 –

P value –










Cyp2b3 Fads3

K2.200 K1.995

0.031 0.030

K2.393 K2.058

0.022 0.019

LOC191574 Haao Oplah Adk Adh1 A2m Ugt2b Arg1 Capn2 Cps1 Dmgdh Dpp4 Dlad LOC64305 Enpp1

– K2.055 K1.965 – – – – K3.530 K1.755 – K1.780 K1.930 K2.090 K5.405 –

– 0.035 0.033 – – – – 0.024 0.044 – 0.032 0.049 0.043 0.036 –

K2.065 K2.413 – K1.940 K4.930 K15.810 K1.791 K5.471 – K4.582 – K1.984 – K6.925 K1.781

0.020 0.034 – 0.025 0.007 0.012 0.026 0.017 – 0.016 – 0.027 – 0.013 0.021

Fmo3 GstYb4 Gstp2 Gamt Kng1

– K3.525 – K1.710 –

– 0.022 – 0.040 –

K1.910 K2.663 2.300 K2.088 K1.990

0.030 0.030 0.018 0.031 0.014

Kat2 Limk2 Masp2

K1.810 – –

0.044 – –

K2.425 K1.827 K1.650 (continued on

0.028 0.036 0.028 next page)


R.E. Castro et al. / Journal of Hepatology 42 (2005) 897–906

Table 1 (continued) Affymetrix tag number

1368244_at 1387034_at 1387294_at

Gene name

Methyltransferase Cyt19 Phenylalanine hydroxylase SH3-domain binding protein 5 (BTK-associated) 1370019_at Sulfotransferase family 1A, phenol-preferring, member 1 1387963_a_at Urate oxidase 1368245_at Ureidopropionase, b Glucocorticoid metabolism 1371089_at Glutathione-S-transferase, a type2 1386953_at Hydroxysteroid dehydrogenase, 11 b type 1 Glycolysis and gluconeogenesis 1368651_at Pyruvate kinase, liver and RBC Lipid, fatty acid metabolism 1367854_at ATP citrate lyase 1367659_s_at Dodecenoyl-coenzyme A d isomerase DCI 1370939_at Fatty acid Coenzyme A ligase, long chain 2 1367857_at Fatty acid desaturase 1 1367707_at Fatty acid synthase 1370237_at L-3-hydroxyacyl-Coenzyme A dehydrogenase, short chain 1370831_at Monoglyceride lipase 1387271_at Phytanoyl-CoA hydroxylase (Refsum disease) 1368150_at Solute carrier family 27 (fatty acid transporter), member 32 1370355_at Stearoyl-Coenzyme A desaturase 1 Other 1371373_at Fasting-inducible integral membrane protein TM6P1 1388155_at Keratin complex 1, acidic, gene 18 1369716_s_at Lectin, galactose binding, soluble 5 (Galectin-5) 1398260_a_at Leuserpin-2 1367847_at Nuclear protein 1 1370541_at Nuclear receptor subfamily 1, group D, member 2 1387994_at Oxidative 17 b hydroxysteroid dehydrogenase type 6 1370073_at Protein kinase inhibitor p58 1370988_at Rat VL30 element mRNA 1390706_at Rattus norvegicus b II spectrin-short isoform mRNA, partial cds 1379497_at Rattus norvegicus Tclone4 mRNA 1388202_at RT1 class Ib gene 1387969_at Small inducible cytokine B subfamily (Cys-X-Cys), member 10 1368840_at TORID 1386321_s_at Tribbles homolog 3 (Drosophila) 1368762_at Ubiquitin D 1371091_at Unknown protein Proliferation/cell growth 1368431_at Hepsin 1370333_a_at Insulin-like growth factor 1 1368160_at Insulin-like growth factor binding protein 1 1387816_at Insulin-like growth factor binding protein, acid labile subunit 1368706_at Transmembrane 4 superfamily member 4

Gene symbol

UDCA Average (fold change)

Cyt19 Pah Sh3bp5

– – –

TUDCA P value

Average (fold change)

P value

– – –

K1.899 K2.485 K1.812

0.026 0.025 0.020






Uox Upb1

K3.855 –

0.020 –

K5.888 K2.510

0.018 0.020

Gsta2 Hsd11b1

K1.970 –

0.041 –

K2.045 K1.712

0.039 0.028




Acly Dci Acsl1

– K1.820 –

– 0.044 –

K1.870 – K1.705

0.031 – 0.015

Fads1 Fasn Hadhsc

K1.900 K1.750 –

0.047 0.036 –

K3.095 – K1.765

0.020 – 0.037

Mgll Phyh

K2.150 –

0.048 –

– K1.823

– 0.032










K1.755 K1.635

0.043 0.034

– K2.260

– 0.018

– 1.870 –

– 0.040 –

K2.050 – 1.681

0.015 – 0.032







1.870 K2.125 K1.755

0.037 0.043 0.048

– K2.115 –

– 0.028 –

2.720 – 2.020

0.016 – 0.028

– 1.981 –

– 0.018 –

Torid Trib3 Ubd LOC207125

K1.630 1.800 K2.745 1.625

0.040 0.050 0.045 0.048

– – – –

– – – –

Hpn Igf1 Igfbp1

K2.080 – 2.820

0.019 – 0.030




Krt1-18 Lgals5 Serpind1 Nupr1 NR1D2


RT1-Aw2 Cxcl10


K2.038 K1.791 2.778 –

0.029 0.025 0.027 –

K1.990 0.018 (continued on next page)

R.E. Castro et al. / Journal of Hepatology 42 (2005) 897–906


Table 1 (continued) Affymetrix tag number

Protein adaptors 1368618_at 1387201_at Protooncogenes 1370902_at 1371103_at 1386857_at Structural proteins 1370902_at Transcription factors 1367602_at 1368249_at 1369679_a_at 1367946_at 1371400_at Transport 1368316_at 1370464_at

Gene name

Gene symbol

UDCA Average (fold change)

TUDCA P value

Average (fold change)

P value

Growth factor receptor bound protein 14 Trif gene

Grb14 Trif-pending

K1.765 2.035

0.032 0.037

– 1.940

– 0.037

Aldose reductase-like protein RAB6, member RAS oncogene family Stathmin 1

LOC286921 Rab6 Stmn1

2.805 1.605 1.945

0.025 0.046 0.022

3.323 – 2.218

0.025 – 0.035

Aldose reductase-like protein






Cbp/p300-interacting transactivator, with Glu/Asp-rich carboxy-terminal domain, 2 Kruppel-like factor 15 Nuclear factor I/A PDZ and LIM domain 1 Thyroid hormone responsive protein (spot14)






Klf15 Nfia Pdlim1 Thrsp

2.930 – – K2.025

0.036 – – 0.017

– K1.915 K1.772 K3.790

– 0.026 0.035 0.012

Aqp8 Abcb1a

K1.680 2.195

0.042 0.039

K1.753 –

0.036 –

S100a6 Slc37a4 Slc6a9 Kif1c Lgals9

K6.465 – – K1.645 –

0.028 – – 0.048 –

K5.815 K1.715 2.010 – K1.962

0.028 0.032 0.022 – 0.033

Lman1 Nucb2 Nup62 Slc2a2

1.835 1.730 – K1.810

0.029 0.040 – 0.031

– – 1.681 –

– – 0.033 –

Slc21a10 Slc22a8 Slc28a2 Slc38a3 Scp2 Slc26a1 Pr1

K1.645 K4.725 K1.895 – – K2.950 –

0.044 0.030 0.046 – – 0.024 –

– K3.393 – K1.980 K1.690 K3.165 K2.543

– 0.019 – 0.019 0.030 0.017 0.017

Gpnmb Hrg Lox Wif1

K2.215 K1.770 K3.895 1.980

0.045 0.031 0.027 0.045

– K1.788 K3.621 –

– 0.033 0.032 –

Aquaporin 8 ATP-binding cassette, sub-family B (MDR/TAP), member 1A 1367661_at Calcium binding protein A6 (calcyclin) 1386960_at Glucose-6-phosphatase, transport protein 1 1387693_a_at Glycine transporter 1 1370930_at Kinesin 1C 1387027_a_at Lectin, galactose binding, soluble 9 (Galectin-9) 1368848_at Lectin, mannose-binding, 1 1370000_at NEFA precursor 1370909_at Nuclear pore glycoprotein 62 1387228_at Solute carrier family 2 A2 (glucose transporter, type 2) 1369746_a_at Solute carrier family 21, member 10 1368461_at Solute carrier family 22 member 8 1368227_at Solute carrier family 28, member 2 1370824_at Solute carrier family 38, member 3 1387896_at Sterol carrier protein 2, liver 1368600_at Sulfate anion transporter 1388103_at Voltage-dependent calcium channel g subunit-like protein Tumor supressor activity 1368187_at Glycoprotein (transmembrane) nmb 1368583_a_at Histidine-rich glycoprotein 1368172_a_at Lysyl oxidase 1369203_at Wnt inhibitory factor 1

Genes were identified by the dChip 1.3 software and assigned to a functional class based on their reported function in the literature. The Affymetrix tag number, name and symbol are shown for each gene. Fold-change data are presented for UDCA- and TUDCA-treated samples relative to unity. Criteria for selection of candidate transcripts are described in Section 2.

unexpected, hepatobiliary transport, metabolism and signaling genes were not significantly affected after exposure of hepatocytes to UDCA for 24 h. Apaf-1 was one of the newly described target genes of UDCA identified herein. Apaf-1 was recently described as a direct transcriptional target of p53 [23]. In addition, our previous reports suggest that UDCA modulates p53

expression in rat hepatocytes [8]. Thus, given Apaf’s pivotal role in the apoptotic process, it was chosen to further confirm the data obtained by the microarray analysis. The expression levels of Apaf-1 were characterized by RTPCR and Western blot. As expected, the expression changes identified through microarrays were well reproduced by both semi-quantitative RT-PCR and immunoblot analyses


R.E. Castro et al. / Journal of Hepatology 42 (2005) 897–906 A




β-actin -

Control 3

Control 2


Control 1




- 340 bp -147

apaf-1 Time (h)

Cd14 ProT Cebpd EST Mdm-2 Egln3 EST p53 apoptosis-associated target Maob EST TGF-β-induced protein Apaf-1 Prnp

Fold change



1.00 1.00

B Hepatocytes



1.00 1.00









0.91 0.79§ 0.63§ 0.78§ 0.96 0.84§ 1.08 1.05



TUDCA − 130 kDa

Apaf-1 -

− 43

β-Actin Time (h) Fold change



1.00 1.00

HuH-7 cells











1.00 1.00 0.76§ 0.91 0.80§ 0.72* 0.73§ 1.10 0.91 0.90



TUDCA −130 kDa

Apaf-1 -






Gene Expression (fold-change) Fig. 1. Cluster analysis of controls and UDCA-treated experimental samples using the dChip software. For both conditions, each column represents primary rat hepatocytes isolated from a different animal. A representative panel of the genes with a biological role in apoptosis, which are significantly modulated by UDCA is shown (for more information please refer to Table 1). Shades of red indicate increased relative expression, shades of blue indicate decreased relative expression, and white indicates no variation. Genes were assigned to a functional class based on their reported function in the literature.

(Fig. 2). TUDCA was less effective in down-regulating Apaf-1 expression.

4. Discussion The mechanisms by which UDCA modulates cell death in different cell types are not fully understood. We have previously shown that the cytoprotective effects of UDCA and TUDCA result, in part, from their ability to inhibit apoptosis. They regulate apoptosis by modulating mitochondrial membrane perturbation [3–5], as well as by acting on survival pathways [10,11]. However, the exact molecular targets and genetic cascades induced by these bile acids remain incomplete. Therefore, we conducted a microarray analysis in search of UDCA-responsive genes, specifically those important in the modulation of cell death. Of the 96 genes found to be significant and differentially expressed in this study, some have already been thoroughly classified as UDCA-responsive genes in the literature. However, the overwhelming majority represent newly described targets for UDCA. For obvious reasons, not all genes modulated by UDCA and listed herein can be extensively discussed in relation to their potential role on cytoprotective mechanisms. Furthermore, these genes fall into several broad categories, although the most prominent

− 43

β-Actin Time (h) Fold change











1.00 1.00 1.00 1.00 0.99 0.88 0.69§ 0.83 0.80 0.78



0.87 1.13

Fig. 2. RT-PCR and Western blot analysis of Apaf-1 levels. Cells were incubated with 100 mM UDCA, TUDCA or no addition (control) for 12, 24, 36 and 48 h. Total RNA was obtained for RT-PCR analysis, and total proteins were extracted for Western blot analysis as described in Section 2. (A) Representative RT-PCR of apaf-1 and mean densitometry values in primary rat hepatocytes exposed to UDCA or TUDCA. Consistent with microarray data, apaf-1 is down-regulated at 24 h after UDCA incubation, and remains below control levels throughout the time-course. TUDCA is less effective at down-regulating apaf-1 levels. (B) Representative immunoblots of Apaf-1 protein in primary rat hepatocytes and HuH-7 cells incubated with UDCA or TUDCA. As expected, Apaf-1 is significantly down-regulated at 36 and 48 h after UDCA incubation in both cell types. TUDCA showed similar results only at shorter time-points. The blots were normalized to either endogenous b-actin mRNA or b-Actin protein levels, and the results are expressed as mean fold change for at least three different experiments. *P!0.01 and §P!0.05 from the respective control.

are involved in cell cycle, proliferation and apoptosis. Interestingly, primary rat hepatocytes incubated with UDCA for shorter periods showed fewer transcript changes. Nevertheless, cell signaling genes, specifically those involved in amino acid dephosphorylation and phosphorylation, and genes coding for proteins with a role in cellular ion transport appeared to be the most affected. Similarly, TUDCA regulated many genes of apoptosis and cell cycle control events, as well as almost all genes encoding for enzymes. Other transcripts, such as those for proteins with kinase activity and transcription factors were specifically modulated by TUDCA. The array analysis indicated that Apaf-1 is significantly down-regulated in rat hepatocytes in response to incubation with UDCA. However, TUDCA was less effective in downregulating Apaf-1 expression, perhaps reflecting a more robust effect of the unconjugated bile acid in modulating expression of the apoptosis molecule. A number of different factors may be responsible for this observation, including

R.E. Castro et al. / Journal of Hepatology 42 (2005) 897–906

differences in uptake, half-life, binding affinities, as well as the primary and secondary effects associated with the target genes in Table 1. Apaf-1 plays a pivotal role in the mitochondrial pathway of apoptosis, as it activates caspases in a cytochrome c-dependent manner [19]. By recruiting procaspase-9 into a multimeric complex, Apaf-1 triggers its auto-activation [20], which in turn results in cleavage of procaspase-3 and subsequent activation of the proteolytic cascade [21]. Thus, UDCA-induced Apaf-1 down-regulation may ultimately decrease caspase activity, providing a possible mechanism for the reported anti-apoptotic actions of this bile acid. However, the complexity of the mitochondrial pathway of cell death is subjected to many factors involved in its regulation, including modulation by Bcl-2 family members, or inhibitor of apoptosis proteins. Recently, the oncoprotein prothymosin-a (ProT) was described as an inhibitor of the formation of the apoptosome [22], providing an additional step for this tight regulation. Notably, UDCA was also found to significantly induce ProT expression levels in the microarray analysis, thus supporting a role for this bile acid in the disruption of the apoptosome, and blockage of apoptosis. Ongoing experiments are further exploring the role of Apaf-1 in UDCA’s anti-apoptotic effect. These results are in agreement with our previous reports demonstrating that UDCA modulates the E2F-1/p53 apoptotic cascade [8]. Interestingly, Apaf-1 was recently identified as a transcriptional target of E2F-1 and p53 [23]. Although microarray data did not reveal a statistically significant and direct effect of UDCA on E2F-1 and p53 gene expression, we have previously shown that UDCA is able to up-regulate the retinoblastoma protein (pRb), while slightly altering E2F-1 and Mdm-2 protein levels [24,25]. In addition, the bile acid efficiently prevented the increase in p53 induced by either TGF-b1 incubation or E2F-1 overexpression in primary rat hepatocytes [8]. Nevertheless, our microarray analysis revealed that UDCA increased an expressed sequence tag (EST) highly similar to Mdm-2 protein while decreasing another EST highly similar to p53 apoptosis-associated target. All together, it appears that the p53/Apaf-1 pathway is strongly modulated by UDCA. These effects are most likely specific, as incubations with other bile acids such as deoxycholic or hyodeoxycholic acids failed to show similar protective effects [3,5]. E2F-1 can also be activated by cyclin D1, which may play a role in apoptosis [26]. In the present study, UDCA decreased cyclin D1 expression by almost 3.5-fold, while increasing one of its transcriptional repressors, the HMGbox containing protein 1 (Hbp1). Interestingly, Hbp1 is itself a target of pRb [27]. Still among the cell-cycle regulated genes, cadherin 1 (Cdh1), or E-cadherin, a negative regulator of cell proliferation [28] decreased its expression as a result of UDCA incubation. The modulation of cell survival by UDCA via different biological pathways is also apparent in its cyclin D1 and E2F-1/p53 targeting


actions. In fact, despite playing a crucial role in the activation of genes responsible for cell cycle progression and acting at times as an oncogene [29], excess E2F-1 can inhibit growth and promote apoptosis [30–32]. Thus, cyclin D1-induced pRb expression may under some circumstances enhance cell cycle arrest or apoptosis [26]. Our previous studies have suggested that a given stimulus may be required for UDCA and its conjugates to trigger certain cell survival pathways [24,33]. Further, it has been reported that UDCA was able to inhibit azoxymethaneinduced increase in cyclin D1 mRNA and protein levels, only in presence of the carcinogen [34]. UDCA has additionally been studied in considerably different biological systems, such as human colon cancer cells [35]. Thus, it is possible that the effects reported here may be further modulated in the presence of a stimulus. A number of other genes were either induced or repressed by UDCA and future investigation may clarify their role in UDCA-mediated cytoprotection. The main advantage of microarray technology is the ability to facilitate simultaneous assessment of genetic expression patterns that might otherwise not have been discovered through individual gene screening. In fact, our results identified a number of differences in gene expression between UDCA-treated and control cells, some of them revealing novel candidates in characterizing mechanistic targets for UDCA.

Acknowledgements The authors thank Dr. Jo¨rg Becker, Instituto Gulbenkian de Cieˆncia, Lisbon, Portugal, for excellent assistance with data analysis. Supported, in part, by grant POCTI/BCI/44929/2002 from Fundac¸a˜o para a Cieˆncia e a Tecnologia (FCT), Lisbon, Portugal (to C.M.P.R), and Ph.D. fellowships SFRH/BD/12655/2003, SFRH/BD/4823/2001 and SFRH/BD/12641/2003 (to R.E.C., S.S. and R.M.R., respectively) from FCT.

References [1] Paumgartner G, Beuers U. Ursodeoxycholic acid in cholestatic liver disease: mechanisms of action and therapeutic use revisited. Hepatology 2002;36:525–531. [2] Rodrigues CMP, Fan G, Wong PY, Kren BT, Steer CJ. Ursodeoxycholic acid may inhibit deoxycholic acid-induced apoptosis by modulating mitochondrial transmembrane potential and reactive oxygen species production. Mol Med 1998;4:165–178. [3] Rodrigues CMP, Fan G, Ma X, Kren BT, Steer CJ. A novel role for ursodeoxycholic acid in inhibiting apoptosis by modulating mitochondrial membrane perturbation. J Clin Invest 1998;101:2790–2799. [4] Botla R, Spivey JR, Aguilar H, Bronk SF, Gores GJ. Ursodeoxycholate (UDCA) inhibits the mitochondrial membrane permeability transition induced by glycochenodeoxycholate: a mechanism of UDCA cytoprotection. J Pharmacol Exp Ther 1995;272:930–938.


R.E. Castro et al. / Journal of Hepatology 42 (2005) 897–906

[5] Rodrigues CMP, Ma X, Linehan-Stieers C, Fan G, Kren BT, Steer CJ. Ursodeoxycholic acid prevents cytochrome c release in apoptosis by inhibiting mitochondrial membrane depolarization and channel formation. Cell Death Differ 1999;6:842–854. [6] Azzaroli F, Mehal W, Soroka CJ, Wang L, Lee J, Crispe N, et al. Ursodeoxycholic acid diminishes Fas-ligand-induced apoptosis in mouse hepatocytes. Hepatology 2002;36:49–54. [7] Xie Q, Khaoustov VI, Chung CC, Sohn J, Krishnan B, Lewis DE, et al. Effect of tauroursodeoxycholic acid on endoplasmic reticulum stress-induced caspase-12 activation. Hepatology 2002;36:592–601. [8] Sola´ S, Ma X, Castro RE, Kren BT, Steer CJ, Rodrigues CMP. Ursodeoxycholic acid modulates E2F-1 and p53 expression through a caspase-independent mechanism in transforming growth factor b1induced apoptosis of rat hepatocytes. J Biol Chem 2003;278: 48831–48838. [9] Rust C, Karnitz LM, Paya CV, Moscat J, Simari RD, Gores GJ. The bile acid taurochenodeoxycholate activates a phosphatidylinositol 3kinase-dependent survival signaling cascade. J Biol Chem 2000;275: 20210–20216. [10] Qiao L, Yacoub A, Studer E, Gupta S, Pei XY, Grant S, et al. Inhibition of the MAPK and PI3K pathways enhances UDCA-induced apoptosis in primary rodent hepatocytes. Hepatology 2002;35: 779–789. [11] Schoemaker MH, Conde de la Rosa L, Buist-Homan M, Vrenken TE, Havinga R, Poelstra K, et al. Tauroursodeoxycholic acid protects rat hepatocytes from bile acid-induced apoptosis via activation of survival pathways. Hepatology 2004;39:1563–1573. [12] Tung BY, Emond MJ, Haggitt RC, Bronner MP, Kimmey MB, Kowdley KV, et al. Ursodiol use is associated with lower prevalence of colonic neoplasia in patients with ulcerative colitis and primary sclerosing cholangitis. Ann Intern Med 2001;134:89–95. [13] Serfaty L, De Leusse A, Rosmorduc O, Desaint B, Flejou JF, Chazouilleres O, et al. Ursodeoxycholic acid therapy and the risk of colorectal adenoma in patients with primary biliary cirrhosis: an observational study. Hepatology 2003;38:203–209. [14] Earnest DL, Holubec H, Wali RK, Jolley CS, Bissonette M, Bhattacharyya AK, et al. Chemoprevention of azoxymethane-induced colonic carcinogenesis by supplemental dietary ursodeoxycholic acid. Cancer Res 1994;54:5071–5074. [15] Garbo GM, Vicente MG, Fingar V, Kessel D. Effects of ursodeoxycholic acid on photodynamic therapy in a murine tumor model. Photochem Photobiol 2003;78:407–410. [16] Mariash CN, Seelig S, Schwartz HL, Oppenheimer JH. Rapid synergistic interaction between thyroid hormone and carbohydrate on mRNAS14 induction. J Biol Chem 1986;261:9583–9586. [17] Li C, Hung Wong W. Model-based analysis of oligonucleotide arrays: model validation, design issues and standard error application. Genome Biol 2001;2: RESEARCH0032. [18] Li C, Wong WH. Model-based analysis of oligonucleotide arrays: expression index computation and outlier detection. Proc Natl Acad Sci USA 2001;98:31–36. [19] Zou H, Henzel WJ, Liu X, Lutschg A, Wang X. Apaf-1, a human protein homologous to C. elegans CED-4, participates in cytochrome c-dependent activation of caspase-3. Cell 1997;90:405–413.

[20] Srinivasula SM, Ahmad M, Fernandes-Alnemri T, Alnemri ES. Autoactivation of procaspase-9 by Apaf-1-mediated oligomerization. Mol Cell 1998;1:949–957. [21] Li P, Nijhawan D, Budihardjo I, Srinivasula SM, Ahmad M, Alnemri ES, et al. Cytochrome c and dATP-dependent formation of Apaf-1/caspase-9 complex initiates an apoptotic protease cascade. Cell 1997;91:479–489. [22] Jiang X, Kim HE, Shu H, Zhao Y, Zhang H, Kofron J, et al. Distinctive roles of PHAP proteins and prothymosin-a in a death regulatory pathway. Science 2003;299:223–226. [23] Moroni MC, Hickman ES, Denchi EL, Caprara G, Colli E, Cecconi F, et al. Apaf-1 is a transcriptional target for E2F and p53. Nat Cell Biol 2001;3:552–558. [24] Sola´ S, Castro RE, Laires PA, Steer CJ, Rodrigues CMP. Tauroursodeoxycholic acid prevents amyloid-b peptide-induced neuronal death via a phosphatidylinositol 3-kinase-dependent signaling pathway. Mol Med 2003;9:226–234. [25] Ramalho RM, Ribeiro PS, Sola´ S, Castro RE, Steer CJ, Rodrigues CMP. Inhibition of the E2F-1/p53/Bax pathway by tauroursodeoxycholic acid in amyloid b-peptide-induced apoptosis of PC12 cells. J Neurochem 2004;90:567–575. [26] Han EK, Ng SC, Arber N, Begemann M, Weinstein IB. Roles of cyclin D1 and related genes in growth inhibition, senescence and apoptosis. Apoptosis 1999;4:213–219. [27] Tevosian SG, Shih HH, Mendelson KG, Sheppard KA, Paulson KE, Yee AS. HBP1: a HMG box transcriptional repressor that is targeted by the retinoblastoma family. Genes Dev 1997;11:383–396. [28] Conacci-Sorrell M, Zhurinsky J, Ben-Ze’ev A. The cadherin-catenin adhesion system in signaling and cancer. J Clin Invest 2002;109: 987–991. [29] Helin K. Regulation of cell proliferation by the E2F transcription factors. Curr Opin Genet Dev 1998;8:28–35. [30] Field SJ, Tsai FY, Kuo F, Zubiaga AM, Kaelin Jr WG, Livingston DM, et al. E2F-1 functions in mice to promote apoptosis and suppress proliferation. Cell 1996;85:549–561. [31] Hunt KK, Deng J, Liu TJ, Wilson-Heiner M, Swisher SG, Clayman G, et al. Adenovirus-mediated overexpression of the transcription factor E2F-1 induces apoptosis in human breast and ovarian carcinoma cell lines and does not require p53. Cancer Res 1997;57:4722–4726. [32] Ganter B, Fu S, Lipsick JS. D-type cyclins repress transcriptional activation by the v-Myb but not the c-Myb DNA-binding domain. EMBO J 1998;17:255–268. [33] Castro RE, Sola´ S, Ramalho RM, Steer CJ, Rodrigues CMP. The bile acid tauroursodeoxycholic acid modulates phosphorylation and translocation of Bad via phosphatidylinositol 3-kinase in glutamateinduced apoptosis of rat cortical neurons. J Pharmacol Exp Ther 2004;311:845–852. [34] Wali RK, Khare S, Tretiakova M, Cohen G, Nguyen L, Hart J, et al. Ursodeoxycholic acid and F(6)-D(3) inhibit aberrant crypt proliferation in the rat azoxymethane model of colon cancer: roles of cyclin D1 and E-cadherin. Cancer Epidemiol Biomarkers Prev 2002;11: 1653–1662. [35] Im E, Martinez JD. Ursodeoxycholic acid (UDCA) can inhibit deoxycholic acid (DCA)-induced apoptosis via modulation of EGFR/Raf-1/ERK signaling in human colon cancer cells. J Nutr 2004;134:483–486.