Multiple protective effects of melatonin against maternalcholestasis-induced oxidative stress and apoptosis in the rat fetalliver–placenta–maternal liver trio

Introduction

In cholestatic liver diseases, the intrahepatic accumulationof bile acids induces cytotoxicity, which results in hepato-

cellular injury. Some bile acid species, such as the morehydrophobic bile acids, can induce apoptosis directly [1] orthrough the generation of oxidant stress that causes

mitochondrial dysfunction [2]. In humans, even normalpregnancy is frequently associated with mild subclinicalcholestasis, known as asymptomatic hypercholanemia ofpregnancy, whose only known alteration involves bio-

chemical parameters, and is characterized by the accumu-lation of bile acids in maternal serum [3]. The etiology ofthis condition is not well understood but alterations in the

metabolism of progesterone [3] and trans-inhibition byprogesterone metabolites from the canalicular lumen of thebile salt export pump have been suggested to be involved

[4]. In contrast, although overt intrahepatic cholestasis ofpregnancy is usually a benign condition for the mother, it isaccompanied by serious repercussions for the conceptus,

including increased fetal distress, premature delivery, and

enhanced perinatal morbidity and mortality [5]. In addi-tion, the intrahepatic cholestasis of pregnancy impairsplacental functions, reducing the ability of the fetus to

eliminate bile acids toward the maternal blood [6], whichmay further aggravate the situation.In experimental maternal hypercholanemia, induced in

rats by complete obstructive cholestasis during the last third

of pregnancy (OCP), an impairment in placental transportfunctions has been described [7]. This elicits a moderateaccumulation of bile acids in the fetal blood, much lower

than that occurring in the maternal blood but sufficient toinduce marked oxidative damage and apoptosis in fetalliver [8] as well as in the placenta [9].

Obstructive cholestasis has been found to induce areduction in the activities of several enzymes involved inresistance to oxidative stress, such as glutathione peroxi-dase, glutathione-S-transferase, and catalase in the rat fetal

liver–placenta–maternal liver trio [8, 9]. However, biliver-din-IXa reductase (BVRa), the enzyme that transforms themain product of heme oxygenase activity biliverdin-IXainto bilirubin, was found to be moderately upregulated by

Abstract: Maternal cholestasis is usually a benign condition for the mother

but induces profound placental damage and may be lethal for the fetus. The

aim of this study was to investigate the protective effects in rat maternal and

fetal livers as also the placenta of melatonin or silymarin against the

oxidative stress and apoptosis induced by maternal obstructive cholestasis

during the last third of pregnancy (OCP). Melatonin or silymarin

administration (i.e. 5 mg/100 g bw/day after ligation of the maternal

common bile duct on day 14 of pregnancy) reduced OCP-induced lipid

peroxidation, and prevented decreases in total glutathione levels. However,

the protective effect on OCP-induced impairment in the GSH/GSSG ratio

was mild in the placenta and fetal liver, while absent in maternal liver.

Melatonin or silymarin also reduced OCP-induced signs of apoptosis

(increased caspase-3 activity and Bax-a upregulation) in all the organs

assayed. Moreover, melatonin (but not silymarin) upregulated several

proteins involved in the cellular protection against the oxidative stress in rats

with OCP. These included, biliverdin-IXa reductase and the sodium-

dependent vitamin C transport proteins SVCT1 and SVCT2, whose

expression levels were enhanced in maternal and fetal liver by melatonin

treatment. In contrast, in placenta only biliverdin-IXa reductase and SVCT2

were upregulated. These results indicate that whereas the treatment of

cholestatic pregnant rats with melatonin or silymarin affords a direct

protective antioxidant activity, only melatonin has dual beneficial effects

against OCP-induced oxidative challenge in that it stimulates the expression

of some components of the endogenous cellular antioxidant defense.

Maria J. Perez1, Beatriz Castano2,Jose M. Gonzalez-Buitrago1 andJose J. G. Marin2

1Research Unit, University Hospital,

Salamanca; 2Laboratory of Experimental

Hepatology and Drug Targeting, CIBERehd,

University of Salamanca, Salamanca, Spain

Key words: antioxidant, Bax, bile acid,

caspase, glutathione, pregnancy

Address reprint requests to Jose J. G. Marin,

Department of Physiology and Pharmacology,

Campus Miguel de Unamuno E.I.D. S-09,

37007 Salamanca, Spain.

E-mail: jjgmarin@usal.es

Received February 05, 2007;

accepted March 26, 2007.

J. Pineal Res. 2007; 43:130–139Doi:10.1111/j.1600-079X.2007.00453.x

� 2007 The AuthorsJournal compilation � 2007 Blackwell Munksgaard

Journal of Pineal Research

130

OCP [10]. Bilirubin has been reported to be a majorphysiological antioxidant cytoprotectant that acts througha BVRa antioxidant cycle; thus, in the presence of reactive

oxygen species bilirubin can be oxidized to biliverdin andthen recycled by BVRa back to bilirubin [11].

Another important component of the cellular antioxidantdefense system is ascorbic acid or vitamin C. Its physio-

logical role as an antioxidant probably accounts for theinverse relationship between plasma ascorbate concentra-tions and the degree of lipid peroxidation in healthy people

[12]. The requirements of several vitamins are increased inliver disease to allow tissue repair and to compensate forthe diminished storage capacity [13]. Regarding ascorbate,

cells may accumulate this vitamin through sodium-depend-ent vitamin C transport proteins (SVCT) [14]. Twoisoforms of this transporter (SVCT1 and SVCT2) havebeen described in humans [15] and rats [16].

In attempts to prevent cholestasis-induced oxidativedamage pharmacologically, several drugs have been tested.Among them, ursodeoxycholic acid was previously found

to be beneficial in experimental OCP in rats [8, 9]. Itsprotective effect on rat fetal liver and rat placenta isprobably partly due to the fact that it prevents decreases in

glutathione levels and hinders the activation of apoptosis.Melatonin has also been shown to exert a protective effectin experimental models of cholestatic liver disease induced

by a-naphthylisothiocyanate [17], CCl4 [18], or bile ductligation [19, 20]. Moreover, melatonin [21, 22] as well as itsmetabolites [23, 24] are potent antioxidants that scavengefree radicals directly and stimulate the activity of the

enzymes involved in antioxidant defense [25, 26]. On theother hand, silymarin is a compound with promisingtherapeutic effects, as has been suggested in intrahepatic

cholestasis related to pregnancy in humans and in animalstreated with paracetamol, CCl4 or ethynylestradiol [27] orwith biliary obstruction [28]. The therapeutic effect of

silymarin has also been related to its actions as anantioxidant, membrane-stabilizing and cell regeneration-promoting agent [29–31].

The present study was undertaken to evaluate theprotective effects of melatonin or silymarin administrationon pregnant rats against oxidative stress and apoptosisinduced by OCP in the fetal liver–placenta–maternal liver

trio and to determine whether these compounds would beable to stimulate the expression of components involved inthe antioxidant defense system, such as BVRa and the

sodium-dependent vitamin C transporters SVCT1 andSVCT2.

Materials and methods

Animals and experimental design

Pregnant Wistar CF rats (University of Salamanca, Spain)were used according to the experimental protocolsapproved by the Local Ethical Committee for the Use of

Laboratory Animals. On day 14 of pregnancy, the rats wereanesthetized with ether and a sham operation (controlgroup) or complete biliary obstruction (OCP group) wasperformed, as previously described [32]. In brief, using a

nonabsorbable suture, a double ligation separated by 2 mmwas carried out. The common bile duct was dividedbetween the ligations. During the following week, some of

these animals (OCP + Mel and OCP + Sil groups)received daily intragastric administration of melatonin orsilymarin (Sigma-Aldrich, Madrid, Spain), respectively

(5 mg/100 g bw), dissolved in propylene glycol/saline solu-tion (75/25; v/v). The control and OCP groups receivedvehicle alone daily. On day 21 of pregnancy, the animals

were killed under sodium pentobarbital anesthesia. Liverand serum samples from the mothers and fetuses werecollected, frozen in liquid nitrogen, and stored at )80�C forfurther use. Some morphological and biochemical charac-

teristics of these experimental groups are shown in Tables 1and 2.

Determination of glutathione contents

The total glutathione (GSH + GSSG) contents were

determined by an enzymatic method [33]. Then proteinsof liver and placenta homogenates were precipitated by theaddition of a 10% trichloroacetic acid solution. The proteinprecipitate was separated from the remaining solution by

centrifugation at 4500 g at 4�C for 5 min. The supernatantswere combined with 0.6 mm 5,5¢-dithiobis(2-nitrobenzoicacid) prepared in Na2HPO4 buffer. The absorbance of the

Table 1. Body, liver weight of mothers and fetuses and number of fetuses per pregnancy and placenta weight following OCP with orwithout treatment with melatonin (Mel) or silymarin (Sil)

Mothers Fetuses

Control OCP OCP + Mel OCP + Sil Control OCP OCP + Mel OCP + Sil

Body weight (g) 349 ± 12a 300 ± 9 321 ± 10 335 ± 16 4.58 ± 0.08a 4.24 ± 0.05 4.27 ± 0.05 4.10 ± 0.10Liver weight (g) 11.5 ± 0.9 13.0 ± 1.5 10.4 ± 0.7 10.5 ± 0.6 0.28 ± 0.01 0.29 ± 0.01 0.23 ± 0.01 0.21 ± 0.01Liver/body weightratio (%)

3.39 ± 0.20 4.42 ± 0.54 3.24 ± 0.21 3.16 ± 0.23 6.10 ± 0.29a 6.84 ± 0.29 5.49 ± 0.23a 5.10 ± 0.19a

Fetuses per pregnancy 12.5 ± 1.5a 8.8 ± 1.5 12.0 ± 0.5a 12.0 ± 0.7a

Placenta weight (g) 0.48 ± 0.02 0.47 ± 0.01 0.45 ± 0.01 0.45 ± 0.02

These parameters were determined in mothers and their fetuses on day 21 of pregnancy. On day 14, pregnant rats underwent sham operation(control), obstructive cholestasis (OCP), or OCP followed by, i.e. treatment with melatonin or silymarin (OCP + Mel, OCP + Sil: 5 mg/100 g bw/day). In all groups n ¼ 6 mothers and ‡25 fetuses.Values are expressed as mean ± S.E.M.aP < 0.05 on comparing with OCP by the Bonferroni method of multiple range testing.

Melatonin-induced protection in cholestasis

131

samples was monitored at 412 nm over 2 min. Glutathionereductase and NADPH were also added to enzymaticallyreduce GSSG to GSH. The GSH content was determinedusing a calibration curve prepared with an authentic

sample. The GSH/GSSG ratio was calculated after selectivemeasurement of the GSSG levels using 10 mm 2-vinylpy-ridine to derivatize GSH [34].

Lipid peroxidation assays

Lipid peroxidation was estimated by measuring thiobarbit-uric acid-reactive substances (TBARS) [35]. Maternal liver,fetal liver, and the placenta were homogenized in phos-

phate-buffered saline (pH 7.4) containing 1 mm ethylenedi-aminetetraacetic acid (EDTA) and 0.02% butylatedhydroxytoluene. A mixture of trichloroacetic acid andhydrochloric acid containing 0.5% thiobarbituric acid was

added to the tissue homogenates. The reaction mixtureswere placed in a water bath at 95�C for 15 min. Aftercentrifugation at 10,000 g for 5 min, the absorbance of the

supernatants was determined at 535 nm. TBARS concen-trations were calculated using a standard curve preparedwith malonaldehyde (MDA) bis (diethyl acetal) (Sigma-

Aldrich). The results are expressed as pmol MDA/mgprotein. Protein concentrations were determined [36], usingbovine serum albumin as standard.

Assays for caspase-3 activity

Caspase-3 activity was determined using Ac-DEVD-AMC

(Alexis Corp., San Diego, CA, USA) as specific substrate[37]. Liver and placenta homogenates were centrifuged at10,000 g for 5 min, and the supernatants were mixed with a

solution containing 100 mm Tris-HCl, 2 mm EDTA, and20 mm EGTA. Dithiothreitol and Ac-DEVD-AMC wereadded to the resulting mixture, up to a final concentration

of 1 mm and 50 lm, respectively. The reaction wasperformed at 37�C, pH 7.5. At 0, 10, and 20 min, analiquot was taken from the reaction mixture, and 10 lL ofHClO4 was added to terminate the reaction. After centrif-

ugation at 10,000 g for 5 min, the fluorescence of thesupernatant containing released AMC was determinedusing a fluorescence spectrophotometer with excitation at

380 nm and emission at 460 nm. All assays were essentiallylinear during this time interval. The results are expressed asFAU/min/mg protein.

Measurement of gene expression levels

To determine mRNA levels by RT followed by real-time

quantitative PCR (RT-QPCR), tissue samples were imme-diately immersed in the RNAlater RNA stabilizationreagent (Qiagen, Izasa, Barcelona, Spain) and stored at

)80�C until total RNA was isolated using RNAeasy spincolumns from Qiagen (Izasa, Barcelona, Spain). Aftertreatment with RNase-free DNase I (Roche, Barcelona,

Spain), RNA was quantified fluorimetrically with theRiboGreen RNA-Quantitation kit (Molecular Probes,Leiden, the Netherlands). cDNA was synthesized from2 lg of total RNA using random hexamers and the avian

myeloblastosis virus RT (Cloned AMV First-strand cDNASynthesis kit, Invitrogen, Groningen, the Netherlands),according to the instructions supplied by the manufacturer.

Real-time quantitative PCR was then performed usingAmpliTaq Gold polymerase (Applied Biosystems, Madrid,Spain) in an ABI Prism 5700 Sequence Detection System

(Applied Biosystems). The thermal cycling conditions wereas follows: a single cycle at 95�C for 10 min followed by 45cycles at 95�C for 15 s and at 60�C for 60 s. Detection ofamplification products was carried out using SYBR Green

I. Nonspecific products of PCR, as detected by 2.5%agarose gel electrophoresis or melting temperature curves,were not found in any case. The total liver RNA from a

healthy male adult rat was used in all determinations as anexternal calibrator. The results of mRNA abundance foreach target gene in each sample were normalized on the

basis of its 18S rRNA content, which was measured withthe TaqMan Ribosomal RNA Control Reagents kit(Applied Biosystems).

The primer oligonucleotide sequences and conditions formeasuring the relative abundance of BVRa were describedpreviously [10]. The primer oligonucleotide sequences forSVCT1 were: forward primer (position 343–361, GenBank

accession number NM_017315) 5¢-TGT TCC AGG CCAGTG CCT T-3¢ and reverse primer (position 523–503) 5¢-ACC ACG CTG GAC ACC ATG ATT-3¢. For SVCT2,

Table 2. Biochemical parameters of pregnant rats following OCP with or without treatment with melatonin (Mel) or silymarin (Sil)

Mothers

Control OCP OCP + Mel OCP + Sil

Total bilirubin (mg/dL) 0.22 ± 0.02a 6.75 ± 0.38 5.28 ± 1.07 5.56 ± 0.77Alkaline phosphatase (IU/L) 122.00 ± 21.36a 262.25 ± 29.06 193.20 ± 28.77a 149.80 ± 23.54a

GGT (IU/L) 10.20 ± 0.20a 53.00 ± 14.73 55.67 ± 11.45 47.50 ± 14.05GPT (IU/L) 25.50 ± 3.10a 34.40 ± 1.81 43.17 ± 10.65 30.50 ± 7.29GOT (IU/L) 166.20 ± 28.50a 345.00 ± 38.58 406.20 ± 70.73 362.20 ± 88.30Total proteins (g/dL) 5.38 ± 0.45 4.95 ± 0.16 4.87 ± 0.15 4.68 ± 0.18Albumin (g/dL) 3.26 ± 0.36a 2.60 ± 0.11 2.37 ± 0.16 2.48 ± 0.21

Biochemical parameters were determined in blood samples on day 21 of pregnancy. On day 14, pregnant rats underwent sham operation(control), obstructive cholestasis (OCP), or OCP followed by, i.e. treatment with melatonin or silymarin (OCP + Mel, OCP + Sil: 5 mg/100 g bw/day). In all groups n ¼ 6 mothers.Values are expressed as mean ± S.E.M.aP < 0.05 on comparing with OCP by the Bonferroni method of multiple range testing.

Perez et al.

132

they were: forward primer (position 918–942, GenBankaccession number NM_017316) 5¢-GAG GTA TAT TGGACC CTT GAC CAT C-3¢ and reverse primer (position

1024–1007) 5¢-GCA TGG CAA TGC CCC AGT-3¢. Allprimers were designed with the assistance of primer

express software (Applied Biosystems), and their specificitywere checked, using BLAST. They were obtained from

Sigma-Genosys (Cambridge, UK).

Western blot analyses

Immunoblotting analyses of liver and placenta homogen-ates were carried out in 10% sodium dodecyl sulfate

polyacrylamide gels, using 100 lg of protein loading perlane. Rabbit polyclonal antibody OSA-400 against BVRawas from Stressgen Bioreagents (bioNova Cientifica,Madrid, Spain). Rabbit polyclonal antibody against Bax-

a (P19) and mouse monoclonal antibody against glycer-aldehyde-3-phosphate dehydrogenase (GAPDH) (6C5),which was used to confirm equal protein loading, were

from Santa Cruz Biotechnology (Santa Cruz, CA, USA).Antirabbit or antimouse IgG horseradish peroxidase-linkedantibodies and enhanced chemiluminescence reagents were

from Amersham Pharmacia Biotech (Freiburg, Germany).

Statistical analysis

Values are expressed as mean ± S.E.M. Statistical analy-sis, which was performed using the spss 10.0.6 software(SPSS Inc., Chicago, IL, USA) for Windows (Microsoft

Co., Seattle, WA, USA), was carried out using one-wayANOVA, followed by the Bonferroni method of multiplerange testing to compare results obtained in the OCP group

with those obtained in control, OCP + Mel or OCP + Silgroups.

Results

In agreement with previous results [8], the body weights ofboth the mothers and the fetuses, together with the number

of fetuses per gestation, were significantly lower in the OCPgroup. Moreover, OCP increased the liver-to-body weightratio in both the mothers and the fetuses. The treatment of

pregnant rats with melatonin or silymarin had no beneficialeffect on the fetal weight, but it partly improved maternalbody weight gain, the number of fetuses per pregnancy andthe liver-to-body weight ratio (Table 1). The changes

observed in several serum biochemical parameters in themothers with OCP were consistent with typical signs ofliver cell injury associated with cholestasis, such as

increased bilirubin concentrations, increased activities ofalkaline phosphatase, GGT, and transaminase and de-creased albumin concentrations. Melatonin or silymarin

treatment partly prevented the OCP-induced elevation inalkaline phosphatase activity in the mothers (Table 2).Oxidative damage, as indicated by the magnitude of lipid

peroxidation (Fig. 1), was increased by OCP to a similar

extent in maternal livers (A), the placenta (B) and fetallivers (C), as previously described [8, 9]. Melatonin orsilymarin partly prevented these alterations in the three

organs from being assayed. OCP enhanced the steady-statelevels of total glutathione in the maternal liver butdecreased them in the placenta and the fetal liver

(Fig. 2A–C). The GSH/GSSG ratio was significantlyreduced by OCP in the maternal liver, the placenta, andthe fetal liver (Fig. 2D–F). Melatonin or silymarin preven-

ted changes in total glutathione, but had a mild (theplacenta and the fetal liver) or no (maternal liver) effect onthe GSH/GSSG ratio.To evaluate potential alterations in other components

involved in resistance to oxidative stress under theseconditions, the expression of BVRa in the organs wasdetermined by RT-QPCR (Fig. 3) and by the Western blot

analyses (Fig. 3, insets). Reductase migrated as two bandswith molecular masses of 33 and 30 kDa. Previous studiesshowed that post-translational modification of reductase is

responsible for the generation of the molecular weight andcharge variants [38]. OCP induced a marked increase inBVRa mRNA and the protein levels in the maternal

(Fig. 3A) and the fetal (Fig. 3C) liver and, although to a

Fig. 1. Degree of lipid peroxidation in maternal liver (A), placenta (B), and fetal liver (C) on day 21 of pregnancy. On day 14, pregnant ratsunderwent a sham operation (control), obstructive cholestasis (OCP), or OCP followed by treatment with melatonin (OCP + Mel) orsilymarin (OCP + Sil). In all groups n ¼ 6 mothers and ‡9 fetuses.*P < 0.05 on comparing with OCP.

Melatonin-induced protection in cholestasis

133

lesser extent, in the placenta (Fig. 3B). A further significantincrease in this parameter was found in the organs of the

OCP group treated with melatonin, while only a slighttrend in this direction was observed in the OCP animalstreated with silymarin.

The relative abundance of vitamin C transport proteins,such as SVCT1 (Fig. 4A–C) and SVCT2 (Fig. 4D–F) was

also determined, as part of the antioxidant defense systemin the maternal liver, the placenta, and the fetal liver, byRT-QPCR. Regarding the relative abundance of SVCTs

Fig. 2. Total glutathione contents and the GSH/GSSG ratio in maternal liver (A and D), placenta (B and E), and fetal liver (C and F) onday 21 of pregnancy. On day 14, pregnant rats underwent a sham operation (control), obstructive cholestasis (OCP), or OCP followed bytreatment with melatonin (OCP + Mel) or silymarin (OCP + Sil). In all groups n ¼ 6 mothers and ‡9 fetuses.*P < 0.05 on comparingwith OCP.

Fig. 3. Relative abundance of mRNA for BVRa in maternal liver (A), placenta (B), and fetal liver (C) on day 21 of pregnancy. On day 14,pregnant rats underwent a sham operation (control), obstructive cholestasis (OCP), or OCP followed by treatment with melatonin(OCP + Mel) or silymarin (OCP + Sil). Values are expressed as percentages of the external calibrator. In all groups n ¼ 6 mothers and‡12 fetuses.*P < 0.05 on comparing with OCP. The insets show representative Western blot analyses of BVRa in liver and placentahomogenates from the same experimental groups; GAPDH was used to confirm equal protein loading.

Perez et al.

134

mRNA in the control animals, for SVCT1 this was the fetalliver > the maternal liver >> the placenta (Fig. 4A–C),and for SVCT2 placenta >> fetal liver > maternal liver(Fig. 4D–F). The expression levels of SVCT1 and SVCT2

were moderately elevated in the OCP group. The treatmentof pregnant rats with melatonin induced a marked upreg-ulation of SVCT2 in all organs assayed and of SVCT1 in

both fetal and maternal livers (Fig. 4). The latter transpor-ter was poorly expressed in the placenta, and its mRNAlevels were not increased by melatonin. In contrast,

silymarin did not share the stimulating effect on vitaminC transporters.

Caspase-3 activity, an enzyme involved in several alter-

native pathways of apoptosis was measured as an index ofthe degree of apoptosis activation. The levels of thecaspase-3 activity in the control group were the maternalliver > the placenta > the fetal liver. OCP induced a

significant enhancement in caspase-3 activity in these threeorgans (Fig. 5), whereas melatonin partly prevented thesechanges more efficiently than silymarin. These results were

consistent with changes in the steady-state mRNA levels ofthe proapoptotic protein Bax-a, because these were alsomarkedly increased in OCP and this upregulation was

partly prevented by melatonin and silymarin treatment.However, in all cases, the order of the mRNA levels of Bax-a was: the placenta > the fetal liver > the maternal liver(Fig. 6). The changes in the abundance of these proteins

were also investigated by Western blotting, and the results

were consistent with the measurements of mRNA abun-dance, as shown in the insets of Fig. 6.

Discussion

The overproduction of free radicals plays a key role in thepathogenesis of liver damage, associated with experimental

cholestasis in rats. In previous studies, we consistentlyfound that OCP induces an impairment of the antioxidantstatus, accompanied by oxidative damage in the rat

maternal liver, the fetal liver, and the placenta [8, 9]. Inthe present study, we investigated the effects of melatoninor silymarin in this experimental model of cholestasis

during the pregnancy.The reason for the choice of melatonin was that this

compound, which can be rapidly transferred from thematernal to the fetal circulation in humans [39], is a potent

endogenous antioxidant agent able to directly scavengehydroxyl free radicals [21] as well as highly destructiveperoxynitrite anions [22]. Moreover, melatonin is also able

to inhibit the formation of free radicals [21, 22] and toenhance the activity of antioxidant enzymes [25, 26].Furthermore, melatonin metabolites [40, 41] may also

contribute to the melatonin-induced protection againstoxidative stress [42]. Thus, it has been observed thatmelatonin protects mitochondria against the oxidativedamage induced by ischemia and reperfusion in rat placenta

[41, 43]. Regarding cholestasis, previous studies by others

Fig. 4. Relative abundance of mRNA for SVCT1 and SVCT2 in maternal liver (A and D), placenta (B and E), and fetal liver (C and F) onday 21 of pregnancy. On day 14, pregnant rats underwent a sham operation (control), obstructive cholestasis (OCP), or OCP followed bytreatment with melatonin (OCP + Mel) or silymarin (OCP + Sil). Values are expressed as percentages of the external calibrator. In allgroups n ¼ 6 mothers and ‡12 fetuses.*P < 0.05 on comparing with OCP.

Melatonin-induced protection in cholestasis

135

suggested the beneficial effect of melatonin when thiscondition was imposed on adult rats [19]. Our resultsregarding the fetal liver and the placenta fully agree withthe concept that melatonin is an efficient agent against

oxidant challenge.The other compound tested in the present study was

silymarin. This has already been used to treat patients with

intrahepatic cholestasis of the pregnancy to alleviatepruritus. In these women, the drug was not able to correctthe characteristic alterations of intrahepatic cholestasis of

the pregnancy in serum biochemical markers [44]. Silymarinis also able to directly scavenge free radicals [29] and toprevent glutathione depletion [30]. It also binds to mitoch-

ondrial membranes, and prevents oxidation-induced mit-ochondrial swelling [31]. Thus, silymarin has been reportedto protect the fetal liver and the brain against ethanol-induced toxicity [45]. In some previous studies using this

compound on experimental cholestasis, the results were

controversial. Some authors [28] found a reduction inhepatic lipid peroxidation, whereas others [46] found noeffect of this drug on prolonged biliary obstruction in rats.In our hands, the efficiency of silymarin in protecting the

fetal liver and the placenta against OCP-induced oxidativestress, was similar to that of melatonin.Regarding one of the most important endogenous

antioxidant defense mechanisms, i.e. the glutathione redoxcycle, owing to the interruption of the biliary pathwaythrough which glutathione is eliminated from the liver, an

initial accumulation occurs in this organ over the first weekof OCP in rats. This is followed by a subsequent depletionof glutathione if the obstruction is maintained longer.

Treatment with melatonin and silymarin has been reportedto reduce the magnitude of long-term cholestasis-inducedglutathione depletion [19, 28]. In the present study, weobserved that silymarin and melatonin had no significant

effect on glutathione accumulation in the maternal liver at

(A) (B) (C)

Fig. 6. Relative abundance of mRNA for Bax-a in maternal liver (A), placenta (B), and fetal liver (C) on day 21 of pregnancy. On day 14,pregnant rats underwent a sham operation (control), obstructive cholestasis (OCP), or OCP followed by treatment with melatonin(OCP + Mel) or silymarin (OCP + Sil). Values are expressed as percentages of the external calibrator. In all groups n ¼ 6 mothers and‡12 fetuses.*P < 0.05 on comparing with OCP. The insets show representative Western blot analyses of Bax in liver and placentahomogenates from the same experimental groups; GAPDH was used to confirm equal protein loading.

Fig. 5. Activity of caspase-3 in maternal liver (A), placenta (B), and fetal liver (C) on day 21 of pregnancy. On day 14, pregnant ratsunderwent a sham operation (control), obstructive cholestasis (OCP), or OCP followed by treatment with melatonin (OCP + Mel) orsilymarin (OCP + Sil). In all groups n ¼ 6 mothers and ‡9 fetuses.*P < 0.05 on comparing with OCP.

Perez et al.

136

1 wk after imposing OCP. In contrast, both compoundswere efficient in preventing the OCP-induced drop in theglutathione contents that had been previously reported to

occur in the placenta and the fetal liver in this experimentalmodel [8]. The enhanced amount of total glutathioneobserved in the present study presumably reflected theincrease in GSH synthesis in these tissues. This was

employed to neutralize the OCP-induced free radicalsgeneration as suggested by the fact that the GSH/GSSGratio was maintained at low levels in the fetal liver–

placenta–maternal liver trio of OCP rats treated withmelatonin and silymarin.

The present study did not address itself the mechanism

through which melatonin and silymarin enhanced the totalglutathione content in these organs, because previousinvestigations by others extensively studied the effects ofthese compounds on the enzymes involved. Thus, in

experimental cholestasis on adult rats, melatonin attenuatesthe decrease in hepatic superoxide dismutase and glutathi-one reductase activities without affecting the hepatic

glutathione peroxidase activity [19]. Silymarin is also ableto restore the decrease induced by common bile ductligation in serum glutathione reductase [47], as well as

hepatic glutathione peroxidase and glutathione transferaseactivities [28].

Whereas GSH plays a crucial role in neutralizing soluble

oxidants, the bilirubin/biliverdin redox cycle seems to beinvolved in the detoxification of more lipophilic reactiveoxygen species. This results in the oxidation of bilirubin tobiliverdin, which is then regenerated back to bilirubin by

BVRa [11]. The expression of this enzyme can be stimulatedby several oxidants [48]. Here, OCP consistently induced anincrease in the expression of BVRa. The magnitude of this

change in the placenta was only moderate, and, in previousstudies using less efficient reverse-transcriptase technology,this effect of OCP was clearly seen in the maternal and fetal

livers but not in placenta [10]. However, in the present study,this defensive response was markedly enhanced by treat-ment with melatonin and, to a lesser extent, with silymarin.

Both compounds stimulated the expression of BVRa andthis stimulation was particularly prominent in fetal liver.

Vitamin C is another element of the antioxidant defensesystem that was investigated here. Under conditions of an

appropriate vitamin C supply in the diet, its availability forcells depends on both the absorption of the compound bythe intestine and on the efficiency of its uptake from the

extracellular medium. Both processes are mediated bysodium-dependent carriers (SVCT) that were identified onlyrecently [16]. Despite their close sequence homology and

similar functions, the two isoforms of SVCT are differen-tially expressed in tissues. Thus, SVCT1 is highly expressedin liver, whereas SVCT2 is highly expressed in the placenta[16], where this transporter is required for the transport of

ascorbic acid across this organ [49]. Moreover, a deficiencyof this transporter in newborn mice has been reported to belethal [49], uncovering a hitherto unrecognized requirement

of ascorbic acid in the perinatal period.In experimental intrahepatic cholestasis induced by

a-naphthylisothiocyanate in rats an alteration in the hepatic

content of vitamin C was found [50]. The OCP-inducedincrease in the expression of SVCT1 and SVCT2 probably

forms a part of the defensive response against oxidativestress, and may favor tissue repair [13]. In a previous study,it was reported that the exposure of a human lens epithelial

cell line to the chemical oxidant t-butylhydroperoxidestimulated the expression of SVCT2, which suggested thatthis transport system may be regulated at the transcrip-tional level by oxidant stress [51]. Importantly, we found

here that in both fetal and maternal livers and the placenta,melatonin, but not silymarin, was able to enhance theupregulation of both SVCT1 and SVCT2, which presum-

ably contributed to enhance the protection of these organsfrom oxidative stress in animals treated with melatonin.As a consequence of OCP, apoptosis was activated in the

maternal liver, the fetal liver, and the placenta, as revealedby the presence of typical signs of apoptosis, such asenhanced activity of caspase-3 and upregulation of theproapoptotic protein Bax-a [8, 9]. These alterations were, in

part, prevented when the mothers were treated withmelatonin or silymarin during the last week of pregnancy.As oxidative stress probably precedes OCP-induced apop-

tosis in the liver and the placenta, melatonin and silymarin,as antioxidants, may be useful in preventing cell death infetal tissues during maternal cholestasis. Additional non-

antioxidant effects may also be involved in the protectiveability of melatonin. Thus, in in vivo [52] and in vitro [52,53] experimental models of free radical-mediated apoptosis,

melatonin was found to inhibit the mitochondrial per-meability transition pore, which contributes to the apop-totic process by releasing cytochrome c from mitochondria.However, regarding the antiapoptotic properties of silyma-

rin, controversial results have been reported. Thus,although apoptosis induced by tumor necrosis factor-a inseveral cell types, has been reported to be inhibited by

silymarin [54], a strong anticancer activity due to theinhibition of cell proliferation as well as enhanced apoptoticdeath have also been reported for this drug [55]. In the

experimental model of OCP used in the present study,silymarin induced a clear antiapoptotic effect.In conclusion, the treatment of cholestatic pregnant

rats with melatonin has beneficial effects on the rat fetalliver–placenta–maternal liver trio by direct antioxidantactivity and by additional enhancement in several ele-ments of the antioxidant defense, which results in

reducing oxidative stress and correcting proapoptoticchanges. Although less efficient than melatonin, silymarinshares most, but if not all, the beneficial properties

described above. This may have interesting clinicalimplications regarding the use of these compounds inthe treatment of pregnant women with cholestasis, which

deserve further investigations.

Acknowledgments

This study was supported in part by the Instituto de SaludCarlos III, FIS (Grants PI051547 and CP03/00093), Spain.Fundacion Investigacion Medica, Mutua Madrilena (Conv-

III, 2006). Junta de Castilla y Leon (Grant SA059A05),Spain. The group belongs to the CIBERehd (Centro deInvestigacion Biomedica en Red) for Hepatology and

Gastroenterology Research (Instituto de Salud Carlos III,Spain).

Melatonin-induced protection in cholestasis

137

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

also due to Nicholas Skinner for the revision of the Englishtext of the manuscript.

References

1. Faubion WA, Guicciardi ME, Miyoshi H et al. Toxic bile

salts induce rodent hepatocyte apoptosis via direct activation

of Fas. J Clin Invest 1999; 103:137–145.

2. Yerushalmi B, Dahl R, Devereaux MW et al. Bile acid-

induced rat hepatocyte apoptosis is inhibited by antioxidants

and blockers of the mitochondrial permeability transition.

Hepatology 2001; 33:616–626.

3. Pascual MJ, Serrano MA, El-Mir MY et al. Relationship

between asymptomatic hypercholanemia of pregnancy and

progesterone metabolism. Clin Sci 2002; 102:587–593.

4. Vallejo M, Briz O, Serrano MA et al. Potential role of

trans-inhibition of the bile salt export pump by progesterone

metabolites in the etiopathogenesis of intrahepatic cholestasis

of pregnancy. J Hepatol 2006; 44:1150–1157.

5. Lammert F, Marschall HU, Matern S. Intrahepatic chol-

estasis of pregnancy. Curr Treat Options Gastroenterol 2003;

6:123–132.

6. Serrano MA, Brites D, Larena MG et al. Beneficial effect

of ursodeoxycholic acid on alterations induced by cholestasis

of pregnancy in bile acid transport across the human placenta.

J Hepatol 1998; 28:829–839.

7. Serrano MA, Macias RI, Vallejo M et al. Effect of ur-

sodeoxycholic acid on the impairment induced by maternal

cholestasis in the rat placenta-maternal liver tandem excretory

pathway. J Pharmacol Exp Ther 2003; 305:515–524.

8. Perez MJ, Macias RI, Duran C et al. Oxidative stress and

apoptosis in fetal rat liver induced by maternal cholestasis.

Protective effect of ursodeoxycholic acid. J Hepatol 2005;

43:324–332.

9. Perez MJ, Macias RI, Marin JJ. Maternal cholestasis

induces placental oxidative stress and apoptosis. Protective

effect of ursodeoxycholic acid. Placenta 2006; 27:34–41.

10. Briz O, Macias RI, Perez MJ et al. Excretion of fetal bili-

verdin by the rat placenta-maternal liver tandem. Am J Physiol

Regul Integr Comp Physiol 2006; 290:R749–756.

11. Sedlak TW, Snyder SH. Bilirubin benefits: cellular protec-

tion by a biliverdin reductase antioxidant cycle. Pediatrics

2004; 113:1776–1782.

12. Block G, Dietrich M, Norkus EP et al. Factors associated

with oxidative stress in human populations. Am J Epidemiol

2002; 156:274–285.

13. Leevy CM, Thompson A, Baker H. Vitamins and liver injury.

Am J Clin Nutr 1970; 23:493–499.

14. Wilson JX. Vitamin C transport in animals and plants. In:

Vitamin C: Its Function and Biochemistry in Animals and

Plants. Asard H, May JM, Smirnoff N, eds. BIOS Scientific

Publishers, Oxford, UK, 1970; pp. 97–117.

15. Daruwala R, Song J, Koh WS et al. Cloning and functional

characterization of the human sodium-dependent vitamin C

transporters hSVCT1 and hSVCT2. FEBS Lett 1999; 460:480–

484.

16. Tsukaguchi H, Tokui T, Mackenzie B et al. A family of

mammalian Na+-dependent L-ascorbic acid transporters.

Nature 1999; 399:70–75.

17. Ohta Y, Kongo M, Sasaki E et al. Protective effect of

melatonin against alpha-naphthylisothiocyanate-induced liver

injury in rats. J Pineal Res 2000; 29:15–23.

18. Daniels WM, Reiter RJ, Melchiorri D et al. Melatonin

counteracts lipid peroxidation induced by carbon tetrachloride

but does not restore glucose-6 phosphatase activity. J Pineal

Res 1995; 19:1–6.

19. Lopez PM, Finana IT, De Agueda MC et al. Protective effect

of melatonin against oxidative stress induced by ligature of

extra-hepatic biliary duct in rats: comparison with the effect of

S-adenosyl-L-methionine. J Pineal Res 2000; 28:143–149.

20. Ohta Y, Imai Y, Matsura T et al. Successively postadmin-

istered melatonin prevents disruption of hepatic antioxidant

status in rats with bile duct ligation. J Pineal Res 2005; 39:367–

374.

21. Tan DX, Chen LD, Poeggeler B et al. Melatonin: a potent,

endogenous hydroxyl radical scavenger. Endocrine J 1993;

1:57–60.

22. Allegra M, Reiter RJ, Tan DX et al. The chemistry of

melatonin’s interaction with reactive species. J Pineal Res 2003;

34:1–10.

23. Rosen J, Than NN, Koch D et al. Interactions of melatonin

and its metabolites with the ABTS cation radical: extension of

the radical scavenger cascade and formation of a novel class of

oxidation products, C2-substituted 3-indolinones. J Pineal Res

2006; 41:374–381.

24. Tan DX, Manchester LC, Terron MP et al. One molecule,

many derivatives: a never-ending interaction of melatonin with

reactive oxygen and nitrogen species? J Pineal Res 2007; 42:28–

42.

25. Rodriguez C, May JC, Sainz RM et al. Regulation of anti-

oxidant enzymes: a significant role for melatonin. J Pineal Res

2004; 36:1–9.

26. Tomas-Zapico C, Coto-Montes A. A proposed mechanism

to explain the stimulatory effect of melatonin on antioxidative

enzymes. J Pineal Res 2005; 39:99–104.

27. Kuntz E. Silymarin in the Treatment of Liver Diseases. Falk

Foundation e. V, Freiburg, 1998; pp. 8–50.

28. Gonzalez-Correa JA, de la Cruz JP, Gordillo J et al.

Effects of silymarin MZ-80 on hepatic oxidative stress in rats

with biliary obstruction. Pharmacology 2002; 64:18–27.

29. Comoglio A, Tomasi A, Malandrino S et al. Scavenging

effect of silipide, a new silybinphospholipid complex, on eth-

anol-derived free radicals. Biochem Pharmacol 1995; 50:1313–

1316.

30. Campos R, Garrido A, Guerra R et al. Silybin dihemisuc-

cinate protects against glutathione depletion and lipid peroxi-

dation induced by acetaminophen on rat liver. Planta Med

1989; 55:417–419.

31. Bindoli A, Cavallini L, Silliprandi N. Inhibitory action of

silymarin of lipid peroxide formation in rat liver mitochondria

and microsomes. Biochem Pharmacol 1977; 26:2405–2409.

32. Monte MJ, Morales AI, Arevalo M et al. Reversible

impairment of neonatal hepatobiliary function by maternal

cholestasis. Hepatology 1996; 23:1208–1217.

33. Tietze F. Enzymatic method for quantitative determination of

nanogram amounts of total and oxidized glutathione: appli-

cations to mammalian blood and other tissues. Anal Biochem

1969; 27:502–522.

34. Griffith OW. Determination of glutathione and glutathione

disulfide using glutathione reductase and 2-vinylpyridine. Anal

Biochem 1980; 106:207–212.

Perez et al.

138

35. Niehaus WG, Samuelsson B. Formation of malonaldehyde

from phospholipid arachidonate during microsomal lipid per-

oxidation. Eur J Biochem 1968; 6:126–130.

36. Markwell MA, Haas SM, Bieber LL et al. A modification

of the Lowry procedure to simplify protein determination in

membrane and lipoprotein samples. Anal Biochem 1978;

87:206–210.

37. Hasegawa JI, Kamada S, Kamiike W et al. Involvement of

CPP32/Yama (-like) proteases in Fas-mediated apoptosis.

Cancer Res 1996; 56:1713–1718.

38. Huang TJ, Trakshel GM, Maines MD. Detection of ten

variants of biliverdin reductase in rat liver by two-dimensional

gel electrophoresis. J Biol Chem 1989; 264:7844–7849.

39. Okatani Y, Okamoto K, Hayashi K et al. Maternal-fetal

transfer of melatonin in pregnant women near term. J Pineal

Res 1998; 25:129–134.

40. Reiter RJ. Melatonin: lowering the high price of free radicals.

News Physiol Sci 2000; 15:246–250.

41. Leon J, Acuna-Castroviejo D, Escames G et al. Melatonin

mitigates mitochondrial malfunction. J Pineal Res 2005; 38:1–

9.

42. Ressmeyer AR, Mayo JC, Zelosko V et al. Antioxidant

properties of the melatonin metabolite N-acetyl-5-meth-

oxykynuramine (AMF): scavenging of free radicals and pre-

vention of protein destruction. Redox Rep 2003; 8:205–213.

43. Okatani Y, Wakatsuki A, Shinohara K et al. Melatonin

protects against oxidative mitochondrial damage induced in

rat placenta by ischemia reperfussion. J Pineal Res 2001;

31:173–178.

44. Reyes H. The spectrum of liver and gastrointestinal disease

seen in cholestasis of pregnancy. Gastroenterol Clin North Am

1992; 21:905–921.

45. Edwards J, La Grange L, Wang M et al. Fetoprotectivity of

the flavanolignan compound siliphos against ethanol-induced

toxicity. Phytother Res 2000; 14:517–521.

46. Muriel P, Moreno MG. Effects of silymarin and vitamins E

and C on liver damage induced by prolonged biliary obstruc-

tion in the rat. Basic Clin Pharmacol Toxicol 2004; 94:99–104.

47. Hagymasi K, Kocsis I, Lugasi A et al. Extrahepatic biliary

obstruction: can silymarin protect liver function? Phytother

Res 2002; 16:S78–S80.

48. Maines MD, Ewing JF, Huang TJ et al. Nuclear localization

of biliverdin reductase in the rat kidney: response to nephro-

toxins that induce heme oxygenase-1. J Pharmacol Exp Ther

2001; 296:1091–1097.

49. Sotiriou S, Gispert S, Cheng J et al. Ascorbic-acid trans-

porter Slc23a1 is essential for vitamin C transport into the

brain and for perinatal survival. Nat Med 2002; 8:514–517.

50. Ohta Y, Kongo M, Sasaki E et al. Change in hepatic anti-

oxidant defense system with liver injury development in rats

with a single a-naphthylisothiocyanate intoxication. Toxicol-

ogy 1999; 139:265–275.

51. Kannan R, Stolz A, Ji Q, Prasad PD, Ganapathy V.

Vitamin C transport in human lens epithelial cells: evidence for

the presence of SVCT2. Exp Eye Res 2001; 73:159–165.

52. Andrabi SA, Sayeed I, Siemen D et al. Direct inhibition of

the mitochondrial permeability transition pore: a possible

mechanism responsible for antiapoptotic effects of melatonin.

FASEB J 2004; 18:869–871.

53. Jou MJ, Peng TI, Reiter RJ et al. Visualization of the anti-

oxidative effects of melatonin at the mitochondrial level during

oxidative stress-induced apoptosis of rat brain astrocytes. J

Pineal Res 2004; 37:55–70.

54. Manna SK, Mukhopadhyay A, Van NT et al. Silymarin

suppresses TNF-induced activation of NF-kB, c-Jun N-ter-

minal kinase, and apoptosis. J Immunol 1999; 163:6800–6809.

55. Varghese L, Agarwal C, Tyagi A et al. Silibinin efficacy

against human hepatocellular carcinoma. Clin Cancer Res

2005; 11:8441–8448.

Melatonin-induced protection in cholestasis

139