Toxicology 225 (2006) 183–194

Maternal ethanol consumption during pregnancy enhancesbile acid-induced oxidative stress and apoptosis

in fetal rat liver�

Maria J. Perez a, Elena Velasco b, Maria J. Monte b,Jose M. Gonzalez-Buitrago a, Jose J.G. Marin b,∗

a Research Unit, University Hospital, University of Salamanca, 37007 Salamanca, Spainb Department of Physiology and Pharmacology, Laboratory of Experimental Hepatology and Drug Targeting,

University of Salamanca, Campus Miguel de Unamuno E.I.D. S-09, 37007 Salamanca, Spain

Received 24 March 2006; received in revised form 24 May 2006; accepted 25 May 2006Available online 2 June 2006

Abstract

Ethanol is able to cross the placenta, which may cause teratogenicity. Here we investigated whether ethanol consumption duringpregnancy (ECDP), even at doses unable to cause malformation, might increase the susceptibility of fetal rat liver to oxidativeinsults. Since cholestasis is a common condition in alcoholic liver disease and pregnancy, exposure to glycochenodeoxycholic acid(GCDCA) has been used here as the oxidative insult. The mothers received drinking water without or with ethanol from 4 weeksbefore mating until term, when placenta, maternal liver, and fetal liver were used. Ethanol induced a decreased GSH/GSSG ratio inthese organs, together with enhanced �-glutamylcysteine synthetase and glutathione reductase activities in both placenta and fetalliver. Lipid peroxidation in placenta and fetal liver was enhanced by ethanol, although it had no effect on caspase-3 activity. Althoughthe basal production of reactive oxygen species (ROS) was higher by fetal (FHs) than by maternal (AHs) hepatocytes in short-termcultures, the production of ROS in response to the presence of varying GCDCA concentrations was higher in AHs and was furtherincreased by ECDP, which was associated to a more marked impairment in mitochondrial function. Moreover, GCDCA-inducedapoptosis was increased by ECDP, as revealed by enhanced Bax-�/Bcl-2 ratio (both in AHs and FHs) and the activity of caspase-8(only in AHs) and caspase-3. In sum, our results indicate that although AHs are more prone than FHs to producing ROS, at doses

unable to cause maternal liver damage ethanol consumption causes oxidative stress and apoptosis in fetal liver.© 2006 Elsevier Ireland Ltd. All rights reserved.

Keywords: Bax; Bile acid; Caspase; Ethanol; Glutathione; Gestation

Abbreviations: AHs, adult hepatocytes; ECDP, ethanol consumption during pregnancy; FHs, fetal hepatocytes; �-GCS, �-glutamylcysteinesynthetase; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; GCDCA, glycochenodeoxycholic acid; MDA, malondialdehyde; Mrp, multidrugresistance-associated proteins; Oatp, organic anion-transporting polypeptide; GSSG, oxidized glutathione; PCR, polymerase chain reaction; rPLII,rat placental lactogen II; ROS, reactive oxygen species; GSH, reduced glutathione; RT, reverse transcription

� We have not received any financial support that may affect in any way the conclusions of our article. The authors have no direct or indirectcommercial interest in any company that might be financially affected by the conclusions of the present study.

∗ Corresponding author. Tel.: +34 923 294674; fax: +34 923 294669.E-mail address: jjgmarin@usal.es (J.J.G. Marin).

0300-483X/$ – see front matter © 2006 Elsevier Ireland Ltd. All rights reserved.doi:10.1016/j.tox.2006.05.015

icology

184 M.J. Perez et al. / Tox

1. Introduction

Ethanol is a teratogenic agent that freely crosses theplacenta and reaches fetal tissues. Despite the many stud-ies carried out over the last 10 decades to investigatethe toxic effect of maternal alcohol assumption duringpregnancy, the mechanisms underlying this devastat-ing phenomenon remain unclear. Maximum expressionof this teratogenic response is the fetal alcohol syn-drome (FAS), which is considered to be the leadingnon-genetic cause of birth defects in developed coun-tries (Krulewitch, 2005). The main clinical features ofFAS include intrauterine growth retardation, central ner-vous system malformation and dysfunction, heart andskeletal abnormalities, etc. In rodents, ethanol ingestionhas also been reported to affect a variety of fetal organs,including lung, pituitary, nervous system, brain, liver andplacenta (Driscoll et al., 1990).

The wide variety of the cellular/biochemical effects ofethanol on fetal tissues suggests that ethanol toxicity maybe multifactorial. Many of these responses can be relatedto effects on membrane structure and function. Thus,ethanol affects membrane transport systems, membranefluidity, Na+–K+ pump expression, and EGF receptorexpression (Henderson et al., 1999). Compelling evi-dence suggests that oxidative stress may be one mecha-nism by which ethanol produces these membrane-relatedevents. Ethanol induces oxidative stress in rat fetal brainand liver following in utero administration (Reyes et al.,1993). The means by which ethanol is able to induceoxidative stress in fetal cells is currently under inves-tigation. However, low levels of antioxidants in fetaltissues and exaggerated response of fetal mitochondriato prooxidant stimulation in vitro may lead to enhancedsusceptibility of fetal cells to oxidative stress. Addition-ally, fetal tissues are prone to the formation and subse-quent accumulation of toxic lipid peroxidation products(Henderson et al., 1999).

During intrauterine life, when the liver is not yetmature, an efficient transfer of fetal bile acids acrossthe placenta, together with normal maternal hepatobil-iary function, maintains fetal bile acid levels within thephysiological range (Marin et al., 2005). In humans,even normal pregnancy is frequently associated withmild sub-clinical cholestasis, known as asymptomatichypercholanemia of pregnancy due to a characteristicaccumulation of bile acids (Fulton et al., 1983; Pascual etal., 2002). In overt intrahepatic cholestasis of pregnancy,

it has been shown that fetal complications arise when bileacid levels in maternal serum are above 40 �M; this isaccompanied by serious repercussions for the concep-tus, including increased fetal distress, premature deliv-

225 (2006) 183–194

ery, and perinatal mortality and morbidity (Glantz et al.,2004). Using an experimental model of hypercholanemiaduring pregnancy in rats, impairment in placental trans-port functions has been described (Serrano et al., 2003).This elicits a moderate accumulation of bile acids in thefetal compartment that is sufficient to induce markedoxidative stress and apoptosis in fetal liver (Perez et al.,2005) as well as in the placenta (Perez et al., 2006).

Moreover, cholestasis is a common clinical featureaccompanying alcoholic liver disease, affecting approx-imately 20% of such patients (Nissenbaum et al., 1990).Histological evidence of cholestasis and clinical jaun-dice may be seen in all stages of alcoholic liver dis-ease. This condition is experimentally reproducible.Thus, acute ethanol administration in rodents results in amarked inhibition of bile secretion. The inhibitory effectof ethanol on bile secretion has been attributed to theimpairment in many different functions including bilesalt transport, vesicular exocytosis and the dynamics ofcytoskeleton elements (Alvaro et al., 1995b).

The present study was undertaken to investigatewhether maternal intake of ethanol sensitizes the fetal ratliver against oxidative insults. To this end, we selectedas the insult the exposure to glycochenodeoxycholicacid (GCDCA), a major primary bile acid known toinduce oxidative stress and apoptosis in rat hepatocytes(Faubion et al., 1999; Yerushalmi et al., 2001). Wedecided to address this issue because chronic ethanolintake and cholestasis are two conditions that can coex-ist during pregnancy and both induce oxidative stress,due in part to a depletion of the antioxidant defense sys-tem (Henderson et al., 1999; Perez et al., 2005, 2006).

2. Materials and methods

2.1. Animals and experimental design

Female Wistar CF rats (Animal House, University of Sala-manca, Spain) were used. All animals received humane careaccording to the criteria outlined in the “Guide for the Careand Use of Laboratory Animals” prepared by the NationalAcademy of Sciences and published by the National Institutesof Health (NIH publication vol. 2, 2nd Ed., 2002). The experi-mental protocols were approved by the Ethical Committee forthe Use of Laboratory Animals of the University of Salamanca.Weight-matched female rats were housed individually in plas-tic cages with controlled conditions of temperature (20 ◦C) andthe light/dark cycle (12 h:12 h) and were fed commercial pel-

leted rat food (Panlab, Madrid, Spain).

The model of ethanol consumption during pregnancy(ECDP) was based on previously well-characterized experi-mental models of ethanol consumption in pregnant rats withprogressive increase in the concentration of ethanol in drinking

M.J. Perez et al. / Toxicology

Table 1Body and placenta weight and biochemical parameters

Control Ethanol

Maternal body weight (g) 377 ± 7 297 ± 10*

Fetal body weight (g) 5.2 ± 0.1 3.7 ± 0.1*

Placental weight (g) 0.68 ± 0.06 0.51 ± 0.02*

Fetuses per pregnancy 11.0 ± 0.4 12.4 ± 1.1GPT (UI/L) 33.7 ± 2.7 37.2 ± 1.5GOT (UI/L) 95 ± 11 172 ± 13*

Total bilirubin (mg/dL) 0.40 ± 0.01 0.78 ± 0.11*

Total proteins (g/dL) 53.2 ± 2.0 49.5 ± 1.5Albumin (g/dL) 29.4 ± 1.4 26.0 ± 0.5*

Biochemical parameters were determined in maternal blood sampleson day 21 of pregnancy. Four weeks before mating and through-out pregnancy, rats in the ethanol group were administered ethanolin drinking water. Control rats were studied in parallel. Values aremeans ± S.E.M. from measurements carried out in duplicate on fivemp

waswefos3wsncgattwfstm

2

apif(oece

The generation of reactive oxygen species (ROS) was

others per group and ≥10 fetuses and placentas. *p < 0.05 on com-aring the ethanol group with the control group.

ater (Testar et al., 1986). However, as previous studies usingsimilar model to that used here reported that the liver ultra-

tructure and enzyme maturation of the offspring of dams fedith a maximum rate of 25% ethanol were preserved (Buts

t al., 1992), we decided to enhance the challenge for theetal liver by increasing the dose of ethanol up to 30%. Thus,ne group of animals was accustomed to alcohol intake bytep-wise increases in the ethanol concentrations (15, 20, 25,0%; v/v) in their drinking water weekly over a period of 4eeks, after which they were mated overnight with males of the

ame strain. Pregnancy was detected by the presence of vagi-al plugs. Pregnant female rats were placed back in individualages and received 30% ethanol in drinking water (ethanolroup). Control pair-fed group of pregnant rats, studied in par-llel, received water without ethanol. On day 21 of pregnancy,he animals were sacrificed under sodium pentobarbital anes-hesia (i.p., 50 mg/kg b. wt.). Liver, placenta and serum samplesere collected from the mothers as well as the livers of their

etuses and all were immediately frozen in liquid nitrogen, andtored at −80 ◦C for further use. Some biochemical parame-ers in serum, shown in Table 1, were measured by automated

ethods.

.2. In vivo studies

To evaluate the state of oxidative stress, the followingssays were carried out in liver and placenta homogenatesrepared in ice-cold phosphate-buffered saline. Lipid perox-dation was estimated by measuring malondialdehyde (MDA)ormation (Niehaus and Samuelsson, 1968). Total glutathioneGSH + GSSG) contents in trichloroacetic acid supernatants

f liver and placenta homogenates were determined by annzymatic method (Tietze, 1969). The GSH/GSSG ratio wasalculated after selective measurement of GSSG levels (Perezt al., 2005).

225 (2006) 183–194 185

The activity of glutathione reductase was measured follow-ing the oxidation of NADPH in reaction mixtures containingGSSG as substrate (Carlberg and Mannervick, 1985). Theactivity of �-glutamylcysteine synthetase (�-GCS) was deter-mined by following the formation of inorganic phosphate (Pi)spectrophotometrically in reaction mixtures containing l-�-aminobutyrate, l-glutamate, and Na2ATP as substrates (Sekuraand Meister, 1977), using the PiPer Phosphate Assay Kit(Molecular Probes, Leiden, The Netherlands).

Caspase-3 activity was determined in liver and pla-centa homogenates using Ac-DEVD-AMC (Alexis Corp.,San Diego, CA) as specific substrate and protein concentra-tions were determined using bovine serum albumin as stan-dard according to methods previously reported (Perez et al.,2005).

The abundance of mRNA of three isoforms of the organicanion-transporting polypeptides Oatp1/1a1, Oatp2/1a4, andOatp4/1b2, three isoforms of multidrug resistance-associatedproteins, Mrp1, Mrp2, and Mrp3, and the trophoblastic markerplacental lactogen II (rPLII) were determined in placenta byRT followed by real-time quantitative PCR, using appropriateprimers as previously described (Serrano et al., 2003). RNAfrom placenta samples collected in RNAlater (QIAGEN, Izasa,Barcelona, Spain) was isolated using RNeasy spin columns(QIAGEN) and was measured with the RiboGreen RNA-Quantitation kit (Molecular Probes, Leiden, The Netherlands).RT was carried out with total RNA, using random nanomersand the Enhanced Avian RT-PCR Kit (Sigma-Genosys, Cam-bridge, UK). PCR amplification products were detected usingSYBR Green I, once it had been ascertained in all cases thatno non-specific products were generated during PCR. TotalRNA from the liver of a healthy male adult rat (for Oatps andMrps) or from the term placenta of a healthy rat (for rPLII)was used in all determinations as an external calibrator. To nor-malize the results, the level of 18S rRNA in each sample wasdetermined with an appropriate Taqman® probe (Serrano et al.,2003).

2.3. In vitro studies

Hepatocytes from the adult livers of the control and ethanolgroups were isolated on day 21 of pregnancy by a two-stepcollagenase perfusion, and cultured as previously described(Martinez-Diez et al., 2000). Hepatocytes from fetal livers wereisolated by collagenase disruption and cultured as described(De Juan et al., 1992). Cells were incubated in 5% CO2 at37 ◦C for 90 min, allowing cell attachment to plates. Then, themedium was changed and cells were exposed to 0–500 �Mof glycochenodeoxycholic acid (GCDCA) or 1 mM tert-butylhydroperoxide (tBOOH), used as a positive control, for differ-ent time-periods.

measured spectrofluorometrically using the ROS-detectingprobe, 2′,7′-dichlorofluorescein diacetate (DCFH-DA)(Rosenkranz et al., 1992). Briefly, hepatocytes were loadedwith DCFH-DA for 30 min at 37 ◦C. DCFH-DA is trapped

186 M.J. Perez et al. / Toxicology 225 (2006) 183–194

Fig. 1. Lipid peroxidation (A) and caspase-3 activity (B) in maternal liver, fetal liver and placenta on day 21 of pregnancy. Four weeks beforeup wereried out0.05 on

mating and throughout pregnancy, rats in the ethanol (closed bars) grostudied in parallel. Values are means ± S.E.M. from measurements car*p < 0.05 on comparing the ethanol group with the control group. †p <

within cells and deesterified, yielding non-fluorescent DCFH,which is then oxidized to the fluorescent DCF by severalROS. Hepatocytes were then exposed to GCDCA or tBOOHand DCF fluorescence was monitored hourly at 485 nmexcitation and 538 nm emission on a Microplate FluorescenceReader (Fluoroskan Ascent FL, Thermo, Electron Corpo-ration, Finland). Formazan formation was measured in rathepatocytes by using the water-soluble tetrazolium compound3-(4,5-dimethylthiazol -2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium and an electron-couplingreagent (phenazine methosulfate) as part of a commercialkit (CellTiter 96AQ, Promega, Madison, WI) based on anadaptation of the supplier’s instructions for use.

Caspase-3 activity was determined as mentioned above.Caspase-8 activity was determined in cell lysates using Ac-IETD-AFC (Alexis Corp.) as specific substrate (Perez etal., 2005). Hepatocyte necrosis in cell cultures was assessedby the amount of lactate dehydrogenase released from thecells to the culture medium (referred to as LDHout). Then,LDH activity in the cell lysate supernatant was measured(referred to as LDHin). The results are expressed as theratio between LDHout and total LDH activity (LDHout +LDHin).

Immunoblotting studies on cell lysates were carried outas previously described (Perez et al., 2005), using rab-bit polyclonal antibodies to Bax-� (P19) and Bcl-2 (N19)and mouse monoclonal antibody against glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (6C5), which was usedto confirm equal protein loading, from Santa Cruz Biotechnol-ogy (CA, USA). Anti-mouse or anti-rabbit IgG horseradishperoxidase-linked antibodies and enhanced chemilumines-

cence reagents were from Amersham Pharmacia Biotech(Freiburg, Germany). Lysate from human promyelocyticleukaemia HL-60 cells (Santa Cruz Biotechnology), whichhighly express several members of the Bcl-2 family of pro-teins, was used as a positive control.

administered ethanol in drinking water. Control (open bars) rats werein duplicate on five mothers per group and ≥10 fetuses and placentas.comparing with control maternal liver.

2.4. Statistical analysis

Results are expressed as mean ± S.E.M. To calculate thestatistical significance of the differences between groups, thepaired t-test or the Bonferroni method for multiple-range test-ing were used, as appropriate.

3. Results

3.1. Changes in maternal liver, fetal liver and placenta

Comparisons between control and ethanol groupsrevealed lower body weights in both the mothers andfetuses, as well as lower weights in the placentas of theethanol group (Table 1). However, the number of fetusesper gestation was similar in both groups. In maternalserum, ECDP induced an increase in transaminase activ-ity and total bilirubin levels and a decrease in albuminlevels, which indicated the existence of certain impair-ment in liver function.

ECDP caused marked oxidative damage to fetal liverand placenta, as suggested by the magnitude of lipid per-oxidation, which was lower in maternal liver and onlymoderately elevated by ECDP (Fig. 1A). However, thiswas not accompanied by signs of apoptosis activation, assuggested by the absence of any enhancement in caspase-3 activity, which was markedly lower in fetal liver andplacenta (Fig. 1B). The steady-state levels of total glu-tathione were decreased by ECDP in fetal liver and pla-

centa but not significantly so in maternal liver (Fig. 2A),whereas the GSH/GSSG ratio was significantly reducedby ECDP in fetal liver and placenta, and also in mater-nal liver (Fig. 2B). The activity of the enzymes �-GCS

M.J. Perez et al. / Toxicology 225 (2006) 183–194 187

Fig. 2. Total glutathione content (A), ratio between reduced (GSH) and oxidized (GSSG) glutathione (B), �-glutamylcysteine synthetase (�-GCS)activity (C) and glutathione reductase activity (D) in maternal liver, fetal liver and placenta on day 21 of pregnancy. Four weeks before mating andt inisterp duplicao ompari

(ilEiw

t(etot

hroughout pregnancy, rats in the ethanol (closed bars) group were admarallel. Values are means ± S.E.M. from measurements carried out inn comparing the ethanol group with the control group. †p < 0.05 on c

Fig. 2C) and glutathione reductase (Fig. 2D), involvedn the synthesis and recycling of GSH, respectively, wasower in fetal liver and placenta than in maternal liver.CDP induced an increase in both enzymatic activities

n fetal liver and placenta, whereas in maternal liver theyere not significantly modified.ECDP has been associated with reduced placental

ransport function in humans and experimental animalsGordon et al., 1985). As an indirect evidence of the

ffect of the ethanol consumption on the excretory func-ion of the placenta we measured the relative abundancesf the mRNA of the organic anion-transporting polypep-ides Oatp1/1a1, Oatp2/1a4 and Oatp4/1b2 and the mul-

ed ethanol in drinking water. Control (open bars) rats were studied inte on five mothers per group and ≥10 fetuses and placentas. *p < 0.05

ng with control maternal liver.

tidrug resistance-associated proteins Mrp1, Mrp2 andMrp3 (Marin et al., 2005) (Table 2). Rat placental lacto-gen type II (rPLII) mRNA, a specific trophoblast markerduring the last stages of rat gestation was used as amarker of the amount of trophoblastic tissue in wholeplacenta, in order to extrapolate changes in the expres-sion of the transporters to the overall transport function ofthis organ. No difference between the placentas from thecontrol and ethanol groups in the mRNA levels of rPLII,

Oatp1, Oatp2, Oatp4, Mrp1, Mrp2 and Mrp3 was found(Table 2). However, although no effects were observedon expression of genes involved in placental transport,and the relative abundance of trophoblastic tissue was

188 M.J. Perez et al. / Toxicology 225 (2006) 183–194

Fig. 3. Production of reactive oxygen species (ROS) in adult and fetal hepatocytes isolated from rats on day 21 of pregnancy. Four weeks beforemating and throughout pregnancy, rats in the ethanol group were administered ethanol in drinking water. Control rats were studied in parallel. Adult

) groupsd out in

and fetal hepatocytes from control (open bars) or ethanol (closed barsfor 6 h. Values are means ± S.E.M. from three different cultures carriegroup. †p < 0.05 on comparing with and without GCDCA or tBOOH.

probably not affected, the placentas were smaller. Nev-ertheless the fetuses were also smaller (Table 1), andhence the placenta-to-fetus body weight ratio was simi-lar in control group (13%) and in ethanol group (14%).In sum, these results constitute indirect evidence sug-gesting that placental function relative to fetal size wasnot likely significantly reduced.

3.2. Studies in fetal and adult hepatocytes

Preliminary experiments revealed that ROS produc-tion increased progressively over 6 h of exposure to

Table 2Expression of genes involved in placental transport function

Control Ethanol

rPLII 145 ± 2.40 144 ± 3.29Oatp1/1a1 2.44 ± 0.50 2.54 ± 0.49Oatp2/1a4 2.05 ± 0.42 2.76 ± 0.56Oatp4/1b2 0.59 ± 0.14 0.60 ± 0.12Mrp1 504 ± 38 565 ± 33Mrp2 0.60 ± 0.13 0.61 ± 0.06Mrp3 181 ± 30 183 ± 35

Relative abundance of mRNA in placentas at term determined usingreal-time quantitative RT-PCR. Four weeks before mating and through-out pregnancy, rats in the ethanol group were administered ethanolin drinking water. Control rats were studied in parallel. Values areexpressed as percentages of external calibrators (RNA from the liverof a healthy male adult rat for Oatps and Mrps or from the placenta atterm of a healthy rat for rPLII). Values were normalized by the determi-nation of 18S rRNA in each sample. Values are means ± S.E.M. frommeasurements carried out in triplicate on eight placentas. No signifi-cant difference (p < 0.05) between the control and ethanol groups wasfound.

were obtained and incubated with 1 mM tBOOH or 25 �M GCDCAtriplicate. *p < 0.05 on comparing the ethanol group with the control

25 �M GCDCA (data not shown). Thus, at the end of thisperiod the differences between groups were seen moreclearly and this incubation time was therefore chosenfor comparative purposes in further experiments. ROSproduction was markedly lower in adult (AHs) than infetal (FHs) hepatocytes of the control group (Fig. 3).ECDP enhanced basal ROS production by AHs but notby FHs. The addition of 25 �M GCDCA to the incuba-tion medium stimulated the production of ROS, whichwas more marked in AHs. Moreover, this effect wasstronger in both the AHs and FHs isolated from theethanol group as compared to control group (Fig. 3).

Exposure to the pro-oxidant compound tert-butylhydroperoxide (tBOOH) was used as a positive control ofmaximal stimulation of ROS production by hepatocytes.When this agent was added to the incubation mediumfor 6 h, a similar high production of ROS in control andethanol groups, in both FHs and AHs, was found (Fig. 3).

When the stimulation of ROS production was plottedversus GCDCA concentrations, a significant correlationwas found in all experimental groups (Fig. 4). The dose-dependent GCDCA-induced increase in ROS generationwas similarly low in the AHs and FHs of the controlgroup. ECDP induced an enhanced dose-dependent ROSproduction in response to GCDCA (Fig. 4). These curvessuggested that regarding the magnitude in the responseto GCDCA the order would be control AHs < controlFHs < ethanol FHs ≈ ethanol AHs.

We next investigated whether there was a relation-ship between GCDCA-induced ROS generation and theimpairment in mitochondrial function (Fig. 5). The plotsof GCDCA-induced ROS production versus impair-

M.J. Perez et al. / Toxicology 225 (2006) 183–194 189

Fig. 4. Relationship between the generation of reactive oxygen species(ROS) and GCDCA concentrations in adult (circles) and fetal (squares)hepatocytes isolated from rats on day 21 of pregnancy. Four weeksbefore mating and throughout pregnancy, rats in the ethanol groupwere administered ethanol in drinking water. Control rats were studiedin parallel. Adult and fetal hepatocytes from the control (open symbols)or ethanol (closed symbols) groups were obtained and incubated with0–500 �M GCDCA for 6 h. Values are means from three differentcultures carried out in triplicate. The S.E.M. values range from 4 to 13%of the mean and are not shown for the sake of clarity. The equationsfor both regression curves (p < 0.005 for all of them) were: controlAy

mctcwF

pEiGc(

hdecsF8ab

L

Fig. 5. Relationship between the GCDCA-induced generation of reac-tive oxygen species (ROS) and impairment in formazan formation inadult (circles) and fetal (squares) hepatocytes isolated from rats on day21 of pregnancy. Four weeks before mating and throughout pregnancy,rats in the ethanol group were administered ethanol in drinking water.Control rats were studied in parallel. Adult and fetal hepatocytes fromthe control (open symbols) or ethanol (closed symbols) groups wereobtained and incubated with 0–500 �M GCDCA for 6 h. Values aremeans from three different cultures carried out in triplicate. The S.E.M.values range from 2 to 10% of the mean and are not shown for the sakeof clarity. The equations for both regression curves (p < 0.005 for all of

Hs, y = 0.9 ln(x) + 6.2; control FHs, y = 2.5 ln(x) + 35.5; ethanol AHs,

= 4.6 ln(x) + 65.1; ethanol FHs, y = 4.6 ln(x) + 41.4.

ent in formazan formation revealed a significant linearorrelation. These curves were interpreted as showinghat at similar GCDCA-induced impairment in mito-hondrial function, the magnitude of ROS productionas control AHs < control FHs < ethanol AHs < ethanolHs.

Caspase-3, which participates in several alternativeathways of apoptosis activation, was increased byCDP only in FHs, but not in AHs (Fig. 6A). This activ-

ty was significantly enhanced by incubation with 25 �MCDCA for 6 h in both AHs and FHs isolated from

ontrol rats. The effect was further enhanced by ECDPFig. 6A).

Since bile acid-mediated apoptosis in rat hepatocytesas been shown to be partly due to activation of theeath receptor-dependent pathway of apoptosis (Faubiont al., 1999), the activity of a mediator of this pathway,aspase-8, was also measured (Fig. 6B). This was notignificantly increased by ECDP either in AHs or inHs. Treatment with 25 �M GCDCA increased caspase-activity in AHs and FHs from control rats and this

ctivation was further increased in AHs, but not in FHs,y ECDP.

To investigate the sensitivity to activation of necrosis,DH release was measured after treating the hepatocytes

them) were: control AHs, y = 0.1x + 6.0, control FHs, y = 0.2x + 35.0,ethanol AHs, y = 0.5x + 60.2, and ethanol FHs, y = 0.7x + 40.1.

with low (25 �M) and high (500 �M) concentrations ofGCDCA. No effect of the low concentration was found inany group. In contrast, the high dose of GCDCA induceda moderate degree of necrosis only in the FHs from thecontrol group but in both the AHs and FHs from theethanol group (Table 3).

Changes in the abundance of the pro-apoptotic andanti-apoptotic proteins, Bax-� and Bcl-2, respectively,were analyzed by Western blotting as an index of sus-ceptibiliy to apoptosis. GAPDH protein was determinedin each sample in order to correct differences in pro-tein loading (Fig. 7). In the AHs and FHs of theethanol group, the expression levels of Bax-� and Bcl-2 were markedly increased by 25 �M GCDCA. Theratio between the abundances of these proteins was cal-culated by densitometric analysis of the Western blotsand was normalized using GADPH (Fig. 8). In con-trol animals, the Bax-�/Bcl-2 ratio was lower in AHsthan in FHs. ECDP markedly enhanced the value ofthis ratio, but with a more pronounced effect on FHs,

resulting in greater differences between AHs and FHs inthe ethanol group. Moreover, treatment with GCDCAincreased this ratio, which was more marked in FHs(Fig. 8).

190 M.J. Perez et al. / Toxicology 225 (2006) 183–194

Fig. 6. Caspase-3 (A) and caspase-8 (B) activities in adult and fetalhepatocytes isolated from rats on day 21 of pregnancy. Four weeksbefore mating and throughout pregnancy, rats in the ethanol groupwere administered ethanol in drinking water. Control rats were studiedin parallel. Adult and fetal hepatocytes from the control (open bars)or ethanol (closed bars) groups were obtained and incubated with orwithout 25 �M GCDCA for 6 h. Values are means ± S.E.M. from three

Table 3Ethanol- and GCDCA-induced necrosis in adult and fetal hepatocytes

Control Ethanol

Adult hepatocytesNo bile acid 9.4 ± 1.3 11.5 ± 2.525 �M GCDCA 7.0 ± 1.3 9.8 ± 2.5500 �M GCDCA 8.6 ± 2.1 16.5 ± 1.5*

Fetal hepatocytesNo bile acid 9.7 ± 0.5 9.3 ± 1.525 �M GCDCA 11.6 ± 0.6 10.1 ± 1.8500 �M GCDCA 17.5 ± 1.1* 20.7 ± 2.3*

The degree of cytotoxicity leading to cell death by necrosis wasassessed by the measurement of the LDH activity released by adultand fetal hepatocytes isolated from rats on day 21 of pregnancy to theculture medium (referred to as LDHout) and the LDH activity in thecell lysate supernatant (referred to as LDHin). Results are expressed asthe ratio between LDHout and total LDH activity (LDHout + LDHin).Four weeks before mating and throughout pregnancy, rats in the ethanolgroup were administered ethanol in drinking water. Control rats werestudied in parallel. Adult and fetal hepatocytes from the control orethanol groups were obtained and incubated with or without 25 or500 �M GCDCA for 6 h. Values are means ± S.E.M. from three dif-

different cultures carried out in duplicate. *p < 0.05 on comparing theethanol group with the control group. †p < 0.05 on comparing with andwithout GCDCA.

4. Discussion

Previous studies in laboratory animals have shownthat ECDP can reproduce some of the alterationsobserved in neonates born from women with chronicalcoholism. These alterations include reduced fetal andneonatal body weight, impaired brain development andskeletal growth, decreased fetal viability, and increased

neonatal mortality (Driscoll et al., 1990). Regarding fetalliver, ECDP has been reported to impair fetal rat liverdevelopment (Meyers et al., 2002). The present studyprovides evidence suggesting that ECDP also affects the

ferent cultures carried out in triplicate. *p < 0.05 on comparing withthe control group in the absence of GCDCA (no bile acid).

fetal rat liver by markedly increasing its susceptibility tooxidative insult, such as bile acid accumulation, whichmight occur associated with cholestasis in the mother.However, it should be kept in mind that both chronic(Dare et al., 2002) and acute (Cicero et al., 1994; Abel,1995) paternal consumption of ethanol prior to breedingalso have a negative impact in pregnancy outcome andfetal development.

Oxidative stress plays a role in the pathogenesis andprogression of several conditions, including alcohol-induced liver disease and cholestasis (Jaeschke et al.,2002). In the present study, although maternal liver totalGSH levels were not altered by ECDP, alterations in theGSH/GSSG ratio in the conceptus were observed. In thisrespect, conflicting results about the hepatic contents ofGSH have been observed in experimental animals afterchronic ethanol administration (Fernandez-Checa et al.,1993; Addolorato et al., 1997). Since GSH probablyplays a pivotal role in protecting the fetal liver againstthe noxious effects of oxidant agents, ECDP-induceddecreased levels of GSH in fetal liver may be involvedin some of the changes observed and discussed below.

Decreased levels of GSH may be the result ofenhanced consumption and/or decreased synthesis. Here

we found that the activity of �-GCS and glutathionereductase, the enzymes involved in the rate-limiting stepof GSH synthesis and in the GSH recycling, respectively,were increased by ECDP in both fetal liver and placenta.

M.J. Perez et al. / Toxicology 225 (2006) 183–194 191

Fig. 7. Representative Western blots of the expresssion of Bax-� and Bcl-2 in fetal and adult hepatocytes isolated from rats on day 21 of pregnancy.F anol grs nol gro6 highlya

Sc4irw�1t

FwiLc

our weeks before mating and throughout pregnancy, rats in the ethtudied in parallel. Adult and fetal hepatocytes from the control or ethah. Lysate from human promyelocytic leukaemia HL-60 cells, whichpositive control.

imilar results have been described in adult rats, in whichhronic ethanol ingestion decreased liver GSH levels by0%, although �-GCS activity was doubled; the reasons that both subunits of �-GCS are transcriptionally up-egulated by oxidative stress (Lu et al., 1999). However,

hen alcohol is administered in utero, the activity of-GCS in fetal rat liver is not affected (Reyes et al.,993). ECDP-induced increases in glutathione reduc-ase activity in maternal and offspring livers have been

ig. 8. Relationship between the expression of Bax-� and Bcl-2 in fetal aneeks before mating and throughout pregnancy, rats in the ethanol group we

n parallel. Adult and fetal hepatocytes from the control or ethanol groups wysates from these cells (three samples per group) were used for Western blotorresponding to GADPH in each sample was used to normalize the results.

oup were administered ethanol in drinking water. Control rats wereups were obtained and incubated with or without 25 �M GCDCA forexpress several members of the Bcl-2 family of proteins, was used as

described previously (Cano et al., 2001). Moreover, sincethe antioxidant status of the cells is highly dependent ontheir energy status, it has been suggested that the reduc-tion in ATP levels observed in the livers of fetuses frommothers exposed to ethanol may account for the GSH

depletion (Addolorato et al., 1997).

Decreased GSH availability could be invokedto explain why lipid peroxidation was significantlyenhanced only in the conceptus. Previous studies have

d adult hepatocytes isolated from rats on day 21 of pregnancy. Fourre administered ethanol in drinking water. Control rats were studiedere obtained and incubated with or without 25 �M GCDCA for 6 h.ting and subsequent densitometric analysis. The intensity of the band

icology

192 M.J. Perez et al. / Tox

already shown that a single exposure to ethanol in uterois sufficient to increase lipid peroxidation in wholeembryos (Henderson et al., 1995). Furthermore, serialexposure to ethanol causes an accumulation of toxicaldehyde products derived from lipid peroxidation infetal liver mitochondria, which occurs to a greater degreethan in maternal liver mitochondria (Henderson et al.,1999). One of the factors possibly involved in this differ-ence is the exaggerated mitochondrial response to pro-oxidant stimuli during rat development, together withthe immaturity of antioxidant mechanisms (Hendersonet al., 1999). Overall, this situation may account for theparticularly high sensitivity of the fetal liver to ethanol,implying an increased danger of liver damage as a con-sequence of any additional oxidative insult. This occursfor instance during the fetal-to-neonatal transition, whenimportant circulatory and respiratory changes alreadylead to a transient oxidative stress.

Oxidative stress may lead to cell death by apoptosis.Although some studies in rats have demonstrated thatapoptosis occurs in the liver during chronic alcoholism(Baroni et al., 1994), at least under the experimentalcircumstances used in the present work, ECDP-inducedoxidative damage in fetal liver was insufficient to causeapoptosis “in vivo”, although this was seen when hepato-cytes were subjected to the additional stress of isolationand culture.

Since normal pregnancy is frequently associated withhigh levels of serum bile acids (Fulton et al., 1983;Pascual et al., 2002), and since these compounds areinvolved in the hepatic disorders induced by chronicethanol feeding in rats (Montet et al., 2002), it couldbe speculated that the bile acid accumulation associ-ated with ECDP would lead to enhanced liver fragility.Indeed, enhanced susceptibility to GCDCA-inducedoxidative stress and apoptosis in the FHs and AHsobtained from ethanol-treated rats has been observed.This is consistent with previous results indicating thatsupplementation to chronic ethanol feeding with chen-odeoxycholate aggravates ethanol-induced liver dam-age, characterized by steatosis, lipoperoxidation andcytolysis (Montet et al., 2002). Moreover, modula-tion of ethanol hepatotoxicity by bile acids in iso-lated perfused rat livers has been shown to depend onthe hydrophilic–hydrophobic properties of these com-pounds (Alvaro et al., 1995a), whereas in rat hepato-cytes in primary culture hydrophilic bile acids, such asursodeoxycholic and tauroursodeoxycholic acids, have

been found to intensify ethanol- and acetaldehyde-induced cell damage (Henzel et al., 2004).

Since fetal alterations due to ECDP may be partlydue to impaired placental function, and indeed, ECDP

225 (2006) 183–194

induces a decrease in placental weight and changesin placental histology (Turan-Akay and Arzu-Kockaya,2005), in the present study we have investigated whetherECDP also induced oxidative damage in this organ. Ourresults indicated that, in agreement with other reportson human placental villi (Kay et al., 2000), ethanolexposure induces oxidative stress in the placenta. How-ever, no signs of apoptosis were found in this organ.Moreover, indirect results obtained here suggest that therelative amount of trophoblastic tissue was not lower inthe ethanol than in the control group.

Hydrophobic bile acids directly stimulate the gen-eration of ROS by hepatocytes and liver mitochon-dria (Sokol et al., 1995), and inhibition by antioxidantsprotects hepatocytes from cell necrosis and apopto-sis (Yerushalmi et al., 2001). In the present study, theincreased response to GCDCA observed in the FHs andAHs from the ethanol-treated animals could be due tothe alteration of the GSH/GSSG redox couple found inthe liver of the fetuses and their mothers.

The mechanism by which hydrophobic bile acids maycause hepatocellular apoptosis and necrosis, dependingon the severity of the injury caused by them (Yerushalmiet al., 2001; Gumpricht et al., 2000), may involve mito-chondrial dysfunction (Yerushalmi et al., 2001). In thepresent study, ECDP enhanced the sensitivity to thenecrosis induced by high concentrations of GCDCA,which was more marked in FHs than in AHs, proba-bly due to the different sensitivity to bile acid-mediatedoxidative insult (Perez et al., 2005). However, whenlow concentrations of GCDCA, able to induce oxida-tive stress and apoptosis, were used, no sign of necrosisin FHs and AHs, even in the ethanol group, was found.

The relationship between ROS generation and thedeterioration in mitochondrial function suggests that, inaddition to other mechanisms also involved in ROS pro-duction, such as cytochrome P450, which can be affectedby ethanol (Wu and Cederbaum, 2003), this organelleplays an important role in the different response of AHsand FHs to ECDP and GCDCA. In turn, ROS mightcompromise dehydrogenase enzymatic activities in hep-atocytes, as has been reported for isolated rat brain mito-chondria, which is probably due to the oxidation of theseenzymes (Galindo et al., 2003). Whether ROS genera-tion or mitochondrial dysfunction might be the primaryevent in response to incubation of the hepatocytes withGCDCA is not known, although it is clear that ECDPmarkedly enhanced both changes.

We next investigated the repercussions of ECDP- andGCDCA-induced oxidative insult on cell viability. It hasbeen suggested that bile acids promote hepatocyte apop-tosis through activation of both the Fas receptor/caspase-

icology

8(Gsabr(wptp

mtsi

A

hMaEiCLyD0MNttTcM2

R

A

A

A

M.J. Perez et al. / Tox

(Faubion et al., 1999) and mitochondria-mediatedYerushalmi et al., 2001) pathways. In FHs and AHs,CDCA induced not only Bax-�, but also Bcl-2 expres-

ion. A similar result has been reported in both fetal anddult livers from cholestatic rats and such an increase haseen suggested to play a role as an adaptive response toesist, up to a certain extent, bile acid-induced injuryPerez et al., 2005). In the present study, this mechanismas insufficient to prevent the activation of caspases,robably because the Bax-�/Bcl-2 ratio was modifiedowards an enhanced predominance of the proapoptoticrotein Bax-�.

In sum, our results indicate that although AHs areore prone than FHs to producing ROS, at doses unable

o cause maternal liver damage, ECDP causes oxidativetress and apoptosis in the fetal liver, probably due tommaturity of the antioxidative enzyme systems.

cknowledgements

The authors thank Mrs. M.I. Hernandez Rodriguez forer secretarial help, and Mr. L. Munoz de la Pascua andr. J.F. Martin Martin for caring for the animals. Thanks

re also due to Nicholas Skinner for revision of thenglish text of the manuscript. This study was supported

n part by the Instituto de Salud Carlos III, FIS (GrantsP03/00093 and PI051547), Spain. Junta de Castilla yeon (Grant SA059A05), Spain. Ministerio de CienciaTecnologıa, Plan Nacional de Investigacion Cientıfica,esarrollo e Innovacion Tecnologica (Grant BFI2003-3208), Spain. Fundacion Investigacion Medica, Mutuaadrilena (Conv-III, 2006). The group is member of theetwork for Cooperative Research on Hepatitis, Insti-

uto de Salud Carlos III, FIS (Grant G03/015), andhe Network for Cooperative Research on Membraneransport Proteins (REIT, Red Espanola de Investiga-ion sobre Proteınas Transportadoras de Membrana),inisterio de Ciencia y Tecnologia (Grant BFU2005-

4983-E/BFI), Spain.

eferences

bel, E.L., 1995. A surprising effect of paternal alcohol treatment onrat fetuses. Alcohol 12, 1–6.

ddolorato, G., Gasbarrini, A., Marcoccia, S., Simoncini, M., Bac-carini, P., Vagni, G., Grieco, A., Sbriccoli, A., Granato, A., Ste-fanini, G.F., Gasbarrini, G., 1997. Prenatal exposure to ethanol inrats: effects on liver energy level and antioxidant status in mothers,fetuses, and newborns. Alcohol 14, 569–573.

lvaro, D., Gigliozzi, A., Bini, A., Della Guardia, P., La Rosa,T., Peri, C., del Vecchio, C., Furfaro, S., Capocaccia, L.,1995a. Acute ethanol hepatotoxicity is modulated by bile salthydrophilic–hydrophobic properties. Ital. J. Gastroenterol. 27,335–339.

225 (2006) 183–194 193

Alvaro, D., Gigliozzi, A., Piat, C., Carli, L., Bini, A., La Rosa, T., Fur-faro, S., Capocaccia, L., 1995b. Effect of s-adenosyl-l-methionineon ethanol cholestasis and hepatotoxicity in isolated perfused ratliver. Dig. Dis. Sci. 40, 1592–1600.

Baroni, G.S., Marucci, L., Benedetti, A., Mancini, R., Jezequel, A.M.,Orlandi, F., 1994. Chronic ethanol feeding increases apoptosis andcell proliferation in rat liver. J. Hepatol. 20, 508–513.

Buts, J.P., Sokal, E.M., Van Hoof, F., 1992. Prenatal exposure to ethanolin rats: effects on postnatal maturation of the small intestine andliver. Pediatr. Res. 32, 574–579.

Cano, M.J., Ayala, A., Murillo, M.L., Carreras, O., 2001. Protectiveeffect of folic acid against oxidative stress produced in 21-daypostpartum rats by maternal-ethanol chronic consumption duringpregnancy and lactation period. Free Radic. Res. 34, 1–8.

Carlberg, I., Mannervick, B., 1985. Glutathione reductase. Meth. Enzy-mol. 113, 484–490.

Cicero, T.J., Nock, B., O’Connor, L., Adams, M.L., Sewing, B.N.,Meyer, E.R., 1994. Acute alcohol exposure markedly influencesmale fertility and fetal outcome in the male rat. Life Sci. 55,901–910.

Dare, W.N., Noronha, C.C., Kusemiju, O.T., Okanlawon, O.A., 2002.The effect of ethanol on spermatogenesis and fertility in maleSprague–Dawley rats pretreated with acetylsalicylic acid. Niger.Postgrad. Med. J. 9, 194–198.

De Juan, C., Benito, M., Fabregat, I., 1992. Regulation of albuminexpression in fetal rat hepatocytes cultured under proliferative con-ditions: role of epidermal growth factor and hormones. J. Cell.Physiol. 152, 95–101.

Driscoll, C.D., Streissguth, A.P., Riley, E.P., 1990. Prenatal alcoholexposure: comparability of effects in humans and animal models.Neurotoxicol. Teratol. 12, 231–237.

Faubion, W.A., Guicciardi, M.E., Miyoshi, H., Bronk, S.F., Roberts,P.J., Svingen, P.A., Kaufmann, S.H., Gores, G.J., 1999. Toxic bilesalts induce rodent hepatocyte apoptosis via direct activation ofFas. J. Clin. Invest. 103, 137–145.

Fernandez-Checa, J.C., Hirano, T., Tsukamoto, H., Kaplowitz, N.,1993. Mitochondrial glutathione depletion in alcoholic liver dis-ease. Alcohol 10, 469–475.

Fulton, I.C., Douglas, J.G., Hutchon, D.J., Beckett, G.J., 1983. Is nor-mal pregnancy cholestatic? Clin. Chim. Acta 130, 171–176.

Galindo, M.F., Jordan, J., Gonzalez-Garcia, C., Cena, V., 2003. Reac-tive oxygen species induce swelling and cytochrome c release butnot transmembrane depolarization in isolated rat brain mitochon-dria. Br. J. Pharmacol. 139, 797–804.

Glantz, A., Marschall, H.U., Mattsson, L.A., 2004. Intrahepaticcholestasis of pregnancy: relationships between bile acid levelsand fetal complication rates. Hepatology 40, 467–474.

Gordon, B.H.J., Streeter, M.L., Rosso, P., Winick, M., 1985. Prena-tal alcohol exposure: abnormalities in placental growth and fetalamino acid uptake in the rat. Biol. Neonate 47, 113–119.

Gumpricht, E., Devereaux, M.W., Dahl, R.H., Sokol, R.J., 2000. Glu-tathione status of isolated rat hepatocytes affects bile acid-inducedcellular necrosis but not apoptosis. Toxicol. Appl. Pharmacol. 164,102–111.

Henderson, G.I., Devi, B.G., Perez, A., Schenker, S., 1995. In uteroethanol exposure elicits oxidative stress in the rat fetus. Alcohol.Clin. Exp. Res. 19, 714–720.

Henderson, G.I., Chen, J.J., Schenker, S., 1999. Ethanol, oxida-tive stress, reactive aldehydes, and the fetus. Front. Biosci. 4,D541–D550.

Henzel, K., Thorborg, C., Hofmann, M., Zimmer, G., Leuschner, U.,2004. Toxicity of ethanol and acetaldehyde in hepatocytes treated

icology

194 M.J. Perez et al. / Tox

with ursodeoxycholic or tauroursodeoxycholic acid. Biochim. Bio-phys. Acta 1644, 37–45.

Jaeschke, H., Gores, G.J., Cederbaum, A.I., Hinson, J.A., Pessayre,D., Lemasters, J.J., 2002. Mechanisms of hepatotoxicity. Toxicol.Sci. 65, 166–176.

Kay, H.H., Grindle, K.M., Magness, R.R., 2000. Ethanol exposureinduces oxidative stress and impairs nitric oxide availability in thehuman placental villi: a possible mechanism of toxicity. Am. J.Obstet. Gynecol. 182, 682–688.

Krulewitch, C.J., 2005. Alcohol consumption during pregnancy. Annu.Rev. Nurs. Res. 23, 101–134.

Lu, S.C., Huang, Z.Z., Yang, J.M., Tsukamoto, H., 1999. Effect ofethanol and high-fat feeding on hepatic gamma-glutamylcysteinesynthetase subunit expression in the rat. Hepatology 30, 209–214.

Marin, J.J.G., Macias, R.I.R., Briz, O., Perez, M.J., Serrano, M.A.,2005. Molecular bases of the excretion of fetal bile acids and pig-ments through the fetal liver-placenta-maternal liver pathway. Ann.Hepatol. 4, 70–76.

Martinez-Diez, M.C., Serrano, M.A., Monte, M.J., Marin, J.J.G., 2000.Comparison of the effects of bile acids on cell viability and DNAsynthesis by rat hepatocytes in primary culture. Biochim. Biophys.Acta 1500, 153–160.

Meyers, A.F., Gong, Y., Zhang, M., Casiro, O.G., Battistuzzi, S., Pet-tigrew, N., Minuk, G.Y., 2002. Liver development in a rat modelof fetal alcohol syndrome. Dig. Dis. Sci. 47, 767–772.

Montet, A.M., Oliva, L., Beauge, F., Montet, J.C., 2002. Bile salts mod-ulate chronic ethanol-induced hepatotoxicity. Alcohol Alcohol. 37,25–29.

Niehaus, W.G., Samuelsson, B., 1968. Formation of malonaldehydefrom phospholipid arachidonate during microsomal lipid peroxi-dation. Eur. J. Biochem. 6, 126–130.

Nissenbaum, M., Chedid, A., Mendenhall, C., Gartside, P., 1990. Prog-nostic significance of cholestatic alcoholic hepatitis. VA Coopera-tive Study Group #119. Dig. Dis. Sci. 35, 891–896.

Pascual, M.J., Serrano, M.A., El-Mir, M.Y., Macias, R.I.R., Jimenez,

F., Marin, J.J.G., 2002. Relationship between asymptomatic hyper-cholanaemia of pregnancy and progesterone metabolism. Clin. Sci.(Lond.) 102, 587–593.

Perez, M.J., Macias, R.I.R., Duran, C., Monte, M.J., Gonzalez-Buitrago, J.M., Marin, J.J.G., 2005. Oxidative stress and apoptosis

225 (2006) 183–194

in fetal rat liver induced by maternal cholestasis. Protective effectof ursodeoxycholic acid. J. Hepatol. 43, 324–332.

Perez, M.J., Macias, R.I.R., Marin, J.J.G., 2006. Maternal cholestasisinduces placental oxidative stress and apoptosis. Protective effectof ursodeoxycholic acid. Placenta 27, 34–41.

Reyes, E., Ott, S., Robinson, B., 1993. Effects of in utero administrationof alcohol on glutathione levels in brain and liver. Alcohol. Clin.Exp. Res. 17, 877–881.

Rosenkranz, A.R., Schmaldienst, S., Stuhlmeier, K.M., Chen, W.,Knapp, W., Zlabinger, G.J., 1992. A microplate assay for the detec-tion of oxidative products using 2′,7′-dichlorofluorescin-diacetate.J. Immunol. Meth. 156, 39–45.

Sekura, R., Meister, A., 1977. �-Glutamylcysteine synthetase. Furtherpurification, “half of the sites” reactivity, subunits, and specificity.J. Biol. Chem. 252, 2599–2605.

Serrano, M.A., Macias, R.I.R., Vallejo, M., Briz, O., Bravo, A.,Pascual, M.J., St-Pierre, M.V., Stieger, B., Meier, P.J., Marin,J.J.G., 2003. Effect of ursodeoxycholic acid on the impairmentinduced by maternal cholestasis in the rat placenta-maternal livertandem excretory pathway. J. Pharmacol. Exp. Ther. 305, 515–524.

Sokol, R.J., Winklhofer-Roob, B.M., Devereaux, M.W., McKim Jr.,J.M., 1995. Generation of hydroperoxides in isolated rat hepato-cytes and hepatic mitochondria exposed to hydrophobic bile acids.Gastroenterology 109, 1249–1256.

Testar, X., Lopez, D., Llobera, M., Herrera, E., 1986. Ethanol adminis-tration in the drinking fluid to pregnant rats as a model for the fetalalcohol syndrome. Pharmacol. Biochem. Behav. 24, 625–630.

Tietze, F., 1969. Enzymatic method for quantitative determinationof nanogram amounts of total and oxidized glutathione: applica-tions to mammalian blood and other tissues. Anal. Biochem. 27,502–522.

Turan-Akay, M., Arzu-Kockaya, E., 2005. The effects of alcohol onrat placenta. Cell Biochem. Funct. 23, 435–445.

Wu, D., Cederbaum, A.I., 2003. Alcohol, oxidative stress, and free

radical damage. Alcohol Res. Health 27, 277–284.

Yerushalmi, B., Dahl, R., Devereaux, M.W., Gumpricht, E., Sokol,R.J., 2001. Bile acid-induced rat hepatocyte apoptosis is inhibitedby antioxidants and blockers of the mitochondrial permeabilitytransition. Hepatology 33, 616–626.