| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,1
* Liggins Institute, University of Auckland, Auckland, New Zealand;
Department of Pharmacology and Clinical Pharmacology, Faculty of Medical and Health Sciences, University of Auckland, Auckland, New Zealand;
Genesis Research and Development Corporation, Auckland, New Zealand; and
Pregnancy Research Centre, Department of Perinatal Medicine, The Royal Women’s Hospital and Department of Obstetrics and Gynaecology, The Royal Women’s Hospital and University of Melbourne, Carlton, Victoria, Australia
1Correspondence: Liggins Institute and Department of Pharmacology and Clinical Pharmacology, University of Auckland, Private Bag 92019, Auckland, New Zealand. E-mail: j.keelan{at}auckland.ac.nz
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
/interferon-
)-induced apoptosis and phosphatidylserine externalization in primary trophoblast and trophoblast-like BeWo cells. Silencing of BCRP expression in BeWo cells significantly increased their sensitivity to apoptotic injury in response to cytokines and exogenous C6 and C8 ceramides. BCRP silencing also increased intracellular ceramide levels after cytokine exposure but did not affect cellular protoporphyrin IX concentrations or sensitivity to activators of the intrinsic apoptotic pathway. BCRP expression in placentas from pregnancies complicated by idiopathic fetal growth restriction was decreased compared with controls, suggesting reduced transport of its substrates from the placenta. We conclude that BCRP may play a hitherto unrecognized survival role in the placenta, protecting the trophoblast against cytokine-induced apoptosis and possibly other extrinsic activators via modulation of ceramide signaling. Decreased placental BCRP expression may result in reduced viability and hence functional deficit, contributing to the fetal growth restriction phenotype.—Evseenko, D. A., Murthi, P., Paxton, J. W., Reid, G., Emerald, B. S., Mohankumar, K. M., Lobie, P. E., Brennecke, S. P., Kalionis, B. Keelan, J. A. The ABC transporter BCRP/ABCG2 is a placental survival factor, and its expression is reduced in idiopathic human fetal growth restriction.
Key Words: breast cancer resistance protein • cytokines • ceramide
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2
3
4)
. BCRP transports a broad range of xenobiotics and some endogenous lipid substrates such as steroids and their conjugates; it also possesses lipid floppase activities (5
, 6)
. Apart from their drug transport potential, BCRP and some other ABC proteins provide protection to both nontransformed tissues and cancer cells from endogenous and exogenous stress factors, including proinflammatory mediators and hypoxia (7
, 8)
. They are also postulated to regulate intracellular levels of toxic metabolites (e.g., sphingolipids, porphyrins, and prostaglandins; refs 7
, 9
, 10
).
The physiological functions of BCRP have not yet been fully elucidated. In pregnancy, placental BCRP has been shown to reduce fetal exposure to toxic xenobiotics (11)
. Immunohistochemical studies have localized BCRP to the maternal facing placental syncytium, with some staining of fetal capillaries reported (12)
, implying that BCRP may pump its substrates both to and from the fetus, consistent with a role in placental, in addition to fetal, protection. Expression increases dramatically during trophoblast fusion and differentiation in vitro and is regulated negatively by cytokines and positively by estrogen, epidermal growth factor (EGF), and insulin-like growth factor II (IGF II; refs 13
, 14
). Paradoxically, in the breast BCRP concentrates drugs and environmental toxins into breast milk (15)
. Additional functions for this transporter include cellular protection from hypoxia, toxins, and maintenance of stem cell regeneration potential (7
, 16)
.
Fetal growth restriction [FGR; also known as intrauterine growth restriction (IUGR)] is one of the most common complications of pregnancy (17)
. Clinically, FGR is defined as a fetal birth weight below the 10th percentile for gestational age with a pathological restriction of fetal growth (18
, 19)
. Fetuses that are "small for gestational age" (SGA), but are biophysically well, are intentionally excluded from this definition. Whereas the majority of FGR cases can be accounted for by obvious maternal, fetal, and placental causes (20)
, the remainder of them were classified as idiopathic. Idiopathic FGR pregnancies are distinguished by abnormal umbilical artery diastolic velocities, asymmetric growth of the fetus, and reduced liquor volume (21)
; idiopathic FGR is also frequently associated with placental insufficiency (22
, 23)
. The effect of placental dysfunction in FGR results in a decreased supply of oxygen, reduced transfer of nutrients and growth factors to the fetus, and subsequent growth delay (24
, 25)
.
Defects in trophoblast differentiation and reduced survival as a result of exposure to elevated concentrations of proinflammatory cytokines such as tumor necrosis factor- (TNF-), hypoxia, or other stressors are considered to play an important role in pathogenesis of placental insufficiency and FGR (26
, 27)
. Trophoblasts express TNF- receptors and exposure to this cytokine leads to trophoblast death (28
, 29)
via death receptor-initiated activation of extrinsic apoptotic pathways, including rapid alteration of plasma membrane lipid composition and the generation of ceramides, potent intracellular messengers that directly activate apoptotic cascades (28
, 30)
.
We hypothesized that the remarkably abundant expression of BCRP in the placenta reflects a function of fundamental importance in trophoblast biology, in particular a role in syncytiotrophoblast survival through protection from endogenous (as opposed to xenobiotic) compounds. We predicted that reduced BCRP expression and activity in the placenta would be correlated with conditions of placental compromise such as FGR. To test this hypothesis, we investigated the ability of BCRP to abrogate the apoptotic effects of TNF- and ceramide in primary trophoblast cells and trophoblast-like BeWo cells, which highly express this transporter (13)
. We used both pharmacological inhibition of BCRP and sequence-specific silencing of BCRP expression. We also investigated expression of BCRP in placentas from pregnancies complicated with idiopathic FGR. In this study, we provide evidence that BCRP acts as a survival factor in placental trophoblasts, mitigating the effects of extrinsic apoptotic stimuli and enhancing trophoblast viability via diminished ceramide accumulation. Furthermore, we report that decreased expression of BCRP is associated with idiopathic FGR and negatively correlates with adverse fetal outcomes and evidence of impaired trophoblast function.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
was from PreProTech (Canton, MA, USA), and hygromycin from was BD Bioscience (Bedford, MA, USA). Ko 143 was a generous gift from J. Allen (Centenary Institute of Cancer Medicine and Cell Biology, NSW, Australia). Recombinant human interferon- (IFN-) and M30 antibodies were purchased from Roche Diagnostics (Mannheim, Germany), mitoxantrone [3H] was from Moravek Biochemicals Inc. (Brea, CA, USA), and apoptosis activators 1-(3,4–dichlorobenzyl)-1H-indole-2,3-dione and (–)-deguelin were purchased from Calbiochem, EMD Biosciences Inc. (Darmstadt, Germany). N-hexanoyl-D-erythro-sphingosine (C6 ceramide), N-octanoyl-D-erythro-sphingosine (C8 ceramide), N-palmitoyl-D-erythro-sphingosine (C16 ceramide), N-stearoyl-D-erythro-sphingosine (C18 ceramide), N-arachidoyl-D-erythro-sphingosine (C20 ceramide), N-lignoceroyl-D-erythro-sphingosine (C24 ceramide), and N-palmitoyl (D31)-D-erythro-sphingosine (C16-D31 ceramide) were all from Avanti Polar Lipids (Alabaster, AL, USA). General chemicals (analytical grade) were from Serva (Heidelberg, Germany), Scharlay Chemie (Barcelona, Spain), or AppliChem (Darmstadt, Germany).
Patient details and tissue sampling
Placentas from pregnancies complicated by idiopathic FGR (n=25) and gestation-matched control pregnancies (n=25) were collected with informed patient consent and approval from the Research and Ethics Committees of The Royal Women’s Hospital (Melbourne, Victoria, Australia). Ultrasound data were used to prospectively identify pregnancies complicated by FGR and growth restricted fetuses. Clinical features of the FGR-affected pregnancies as well as the gestation-matched controls used in this study are described in Table 1
. The inclusion criteria for FGR were birth weight less than the 10th centile for gestation age using Australian growth charts (31)
, plus any two of the following criteria diagnosed on antenatal ultrasound: abnormal umbilical artery Doppler flow velocimetry, oligohydramnios as determined by amniotic fluid index (AFI) <7 on antenatal ultrasound performed before delivery, or asymmetric growth of the fetus as quantified from the head circumference (HC) to abdominal circumference (AC) ratio (>1.2). Last menstrual period dates were used to calculate the gestation times for both FGR and control patients, and these were confirmed by early pregnancy ultrasound. Control patients were selected to match FGR cases according to gestational age. For both control and FGR-affected pregnancies, the exclusion criteria were multiple pregnancy, chemical dependency, maternal smoking, preeclampsia, prolonged rupture of the membranes, placental abruption, suspicion of intrauterine viral infection, and fetal congenital anomalies. Only normotensive patients with idiopathic FGR were included. Control patients were included if they required elective delivery by induction of labor/Caesarean section or presented with spontaneous labor. Preterm control patients presented with spontaneous idiopathic preterm labor or underwent elective delivery for conditions not associated with placental dysfunction (e.g., breast cancer). None of the control group patients had clinical evidence of preeclampsia, FGR, placental abruption, ascending infection, or prolonged rupture of the membranes. All control patients gave birth to normally formed babies with birth weights appropriate for gestational age, and the placentas from these patients were grossly normal with no obvious infarcts.
|
All samples were processed within 10 min of placental delivery. Placental tissue samples were excised from randomly selected areas of central placental cotyledons with any attached decidua carefully removed by dissection. Tissues were divided into small pieces and thoroughly washed in phosphate buffered 0.9% saline (PBS) to minimize blood contamination and then snap frozen and stored at –80°C for RNA analyses.
Cell culture
Placentas were obtained with informed consent from women after delivery by Caesarean section at term. Cytotrophoblasts were isolated using trypsin digestion as described previously (14
, 32)
. In brief, villous tissue from term placenta was subjected to eight consecutive digestions in 0.25% trypsin, supernatants were collected, and cells were isolated by centrifugation at 300 g for 7 min. Erythrocytes were removed by incubation of the cell pellet in red cell lysis buffer (50 mM NH4Cl, 10 mM NaHCO3, and 0.1 mM EDTA), and cytotrophoblasts were purified by centrifugation at 1,200 g for 20 min on a discontinuous Percoll gradient (20–60%). Cells between the 40 and 50% Percoll layers were collected and plated in 96-well plates (5x104 cells/well) for apoptosis and viability assays. The cells were grown for 24 h in M199 media, supplemented with 10% FCS, EGF (10 ng/ml), insulin (5 ng/ml), transferrin (10 ng/ml), sodium selenite (0.2 nM), and penicillin/streptomycin (100 U/ml) in a 5% CO2 humidified atmosphere at 37°C. After 24 h in culture, cells were washed with PBS and further cultured in M199 media supplemented with 10% FCS. Trophoblasts differentiated over the proceeding 5 days in culture, expressing the syncytial marker protein ß-chorionic gonadotrophin (13
, 14)
. BeWo cells were cultured in 1:1 DMEM/F-12 with 10% FCS and 1x GlutaMAX.
Quantitative real-time PCR
Total RNA extraction, first-strand cDNA synthesis, and SYBR Green RT-PCR amplification and detection were performed using an ABI Prism 7900 HT (Applied Biosystems) as described previously (13)
. Primer sequences are shown in the Table 2
. The relative standard curve method was used to quantitate gene expression according to Applied Biosystem recommendations (7900 HT Real-time fast and SDS enterprise and database user guide). PCR efficiencies for all tested genes were >95%. Expression of target genes was normalized to the level of 18S ribosomal RNA (18S rRNA), TATA binding protein (TBP), and succinate dehydrogenase complex subunit A (SDHA) after calculation of a normalization factor and was expressed relative to control treatments according to previously published algorithm (33)
.
|
Immunoblotting
Total cell lysates were prepared as described previously (13)
. Protein (20–30 µg) was separated under reducing conditions on a 4–12% BisTris precast polyacrylamide gradient gel and transferred to a nitrocellulose membrane in an XCELL transfer module (Invitrogen). Membranes were blocked in 2% nonfat milk powder and incubated overnight with anti-BCRP monoclonal antibody. Membranes were then washed and incubated with horseradish peroxidase-conjugated goat anti-mouse antibody and visualized by enhanced chemiluminescence recorded on CP-BU New X-ray film (Agfa, Westerlo-Heultje, Belgium). Band intensity was quantitated by densitometry using Quantity One software (Bio-Rad Laboratories, Auckland, NZ). Equivalence of protein loading was confirmed by secondary immunoblotting with anti-ß-actin antibodies.
Immunocytochemical detection of cytokeratin 18 neoepitope with M30 antibody
Cytokeratin 18 neoepitope detection was used as a marker of execution stages of apoptosis (34)
. BeWo cells were seeded in 96-well plates and incubated with compounds as described, after which cells were fixed in ice-cold methanol and immunohistochemical detection of cytokeratin 18 neoepitope was carried out using M30 antibody. As a negative control, the primary antibody was omitted.
Analysis of chromatin condensation and nuclear fragmentation
Apoptotic cell death was also measured by fluorescent microscopic analysis of chromatin condensation with the karyophilic dye Hoechst 33258. Cells were seeded to in 24- or 96-well plates and incubated with the various compounds. After 48 h in culture, cells were fixed in 4% paraformaldehyde in PBS (pH 7.4) and stained with Hoechst 33258 (5 µg/ml) for 10 min at room temperature. After being washed with PBS, nuclear morphology was examined under an ultraviolet-visible inverted fluorescence microscope (Olympus IX 71, Tokyo, Japan). Apoptotic cells were distinguished from viable cells by their nuclear condensation and fragmentation and higher intensity of the blue nuclear fluorescence. For statistical analysis, 
1000 cells were counted in eight random microscopic fields at x400.
Detection of phosphatidylserine externalization
The appearance of phosphatidylserine on the cell surface was detected with anti-phosphatidylserine antibodies. Cells were cultured in 24-well plates, washed three times before experimentation, and cultured without fixation with 1:1000 primary antibody (clone 4B6) in PBS with 0.5% albumin, followed by incubation with Cy3-labeled detection antibody. After incubation, cells were washed thoroughly with PBS and fixed in 10% formaldehyde. Quantitative analysis was carried out using ImageJ 1.36b software from National Institutes of Health (Bethesda, MD, USA) and expressed in normalized fluorescent units (NFU), representing an area of positive phosphatidylserine staining normalized to the total area of nuclei in the same field.
Cell viability assay
The 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay was used to assess BeWo cell viability in 96-well plates after 24, 48, and 72 h of exposure to the various compounds.
Preparation of BCRP short hairpin RNA (shRNA) constructs for stable transfection
A specific sequence targeting BCRP within exon 7 (Genbank accession number: NM_004827) was used for shRNA preparation (sense: 5'-AAGATGATTGTTCGTCCCTGCT-3'; antisense: 5'-AAGCAGGGACGAACAATCATCT-3'). This sequence has previously been shown to efficiently and specifically knock down BCRP gene expression in BeWo cells. Oligonucleotides, synthesized by Invitrogen NZ Ltd., were annealed and ligated into pRNAT-CMV3.1/Hygro vector using AflII and BamH1 enzymes according to the manufacturer’s protocol. The shRNA inserts of several plasmids were sequenced to confirm correct nucleotide sequence before experimentation.
Stable transfection of BeWo cells
BeWo cells were cultured in 6-well plates, and after reaching 50–70% confluence were transfected with the pRNAT-CMV3.1/Hygro-BCRP shRNA construct with lipofectamine reagent according to the manufacturer’s protocol. Vector containing a scrambled small interfering RNA (siRNA) sequence was used for a negative control. In brief, the following procedure was used: 1 µg plasmid and 10 µl of transfection reagent were added to 200 µl of DMEM without serum and incubated for 20 min at room temperature, and another 600 µl of serum-free DMEM were added to a total volume of 800 µl. Growth medium was then removed from cells, and the transfection mix was added and incubated overnight. The following day culture medium with 20% serum added to the cells without removing the transfection mixture, and 72 h later antibiotic selection was initiated. BeWo clones stably expressing integrated plasmids were established after selection in hygromycin (50 µg/ml) for at least 15 days.
Transient transfection using BCRP antisence oligonucleotides
Three different Stealth BCRP siRNA duplexes, designed and synthesized by Invitrogen, were used in combination with transient inhibition of BCRP gene expression in BeWo cells (Table 3
). BeWo cells (5–10,000/well in 96-well plates and 50,000/well in 24-well plates) were transfected using lipofectamine RNAi MAX transfection reagent according to the reverse transfection protocol provided by the manufacturer. All Stealth BCRP siRNA were used as a pool in equal proportions at a final concentration 20 nM (shown to be nontoxic in the pilot experiments). As a control, cells were transfected with an equivalent amount of Stealth scrambled siRNA duplexes with the same GC content.
|
Protoporphyrin IX (PPIX) detection
BeWo cells cultured in 24-well plates were harvested with trypsin and resuspended in fresh media, and PPIX concentration was measured in a LSRII flow cytometer (BD Biosciences). To induce PPIX fluorescence, the excitation wavelength was set at 405 nm, and the emission filter was set at 695 nm/40 nm (7
, 35)
.
Lipid analysis by mass spectrometry
Ceramide and cholesterol analysis was performed on a Thermo Finnigan TSQ Quantum Ultra AM mass spectrometer operating in a multiple reaction monitoring, positive ionization mode. Ceramides were extracted by a previously described protocol (36)
: Cells were washed twice in PBS, and then 1 ml of ice-cold methanol were added to each well. Cells were scraped, sonicated, and extracted with methanol/chloroform (75:25 v/v) overnight at 4–8°C. Chloroform (1 ml) and water (2 ml) was then added, and the sample was centrifuged at 3000 g for 15 min. The lower fraction (1 ml) was then collected and evaporated in glass tubes. Each sample was then reconstituted in 90 µl of methanol/chloroform (25:75), and 8 µl were injected into a liquid chromatography/mass spectrometry (LC/MS) system. Chromatography was performed on a Phenomenex Luna 3 µ 50 x 3 mm C18 column with a 100% methanol mobile phase system at a flow rate 800 µl/min. Peaks corresponding with the target analytes and internal standards were identified, collected, and processed with the Xcalibur software system. Quantitative analysis was based on the calibration curves generated by spiking an artificial matrix with the known amounts of synthetic standards and an equal amount of the internal standards. The target analyte/internal standard peak area ratios were plotted against analyte concentration to generate the calibration curve. The target analyte/internal standard peak area ratios from the samples were similarly normalized to their respective internal standards and compared with the calibration curves by a linear regression.
Immunofluorescence staining
Placentas (n=4) were washed in PBS with (0.1%) Tween 20, and small pieces were quickly frozen on dry ice in Optimum Cutting Temperature (OCT) compound (Lab-Tec Products, ProSci Tech, Thuringowa, QLD, Australia) and stored at –80°C. Sections (10 µm) were then cut with a Leica CM 3050S cryostat (Leica Microsystems GmbH, Wetzlar, Germany) at –20°C, transferred onto Superfrost Plus glass slides (ProSci Tech), and fixed in acetone at –20°C for 10 min. Sections were blocked in PBS with 5% normal horse serum (NHS) for 30 min and incubated with selective primary antibody against BCRP (clone BXP-21) or NHS only (for negative control) overnight at 4°C. After incubation, the sections were washed three times in PBS and incubated with secondary Cy3-labeled fluorescent antibody for 1 h at room temperature. Sections were then washed three times in PBS, counterstained with DAPI (Vector Laboratories, Burlingame, CA, USA), and mounted with Vectashield (Vector Laboratories) for examination with a Zeiss LSM 150 MetaLaser confocal microscope (Carl Zeiss, Oberkochen, Germany).
Statistics
All studies were performed at least three times, descriptive statistics were performed for each data set, and the data were combined for collective analysis. Graphs were plotted, and data were transformed with Microsoft Excel 2003 (San Diego, CA, USA). Statistical analysis was performed with Sigmastat software from Systat Software Inc. (Richmond, CA, USA). Ceramide data were analyzed by one-way ANOVA with repeated measures. For apoptosis and viability experiments, one-way ANOVA was applied, followed by Student-Newman-Keul’s test. P < 0.05 was considered to be significant. Rank sum test was used to analyze the significance of differences between the clinical characteristics of the FGR-affected pregnancies and the control patients. Spearman rank correlation and multiple linear regression analysis were used to model the relationship between expression of transporter mRNA and mRNA of regulatory gene expression.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
/IFN- with/without Ko143, a selective pharmacological inhibitor of BCRP. Ko143 was tested in preliminary experiments using the MTT assay and found to have no detrimental effects on cell viability at concentrations up to 5 µM. Exposure of trophoblasts to TNF-/IFN- for 48 h in the presence of Ko143 (5 µM) resulted in a significant reduction in cell viability (P<0.05) compared with trophoblasts treated with TNF-/IFN- alone (Fig. 1
A). Exposure to TNF-/IFN- in presence of the BCRP blocker also led to significantly higher levels of apoptosis (P<0.05), detectable by analysis of chromatin condensation and nuclear fragmentation, compared with control cells exposed to cytokines alone (Fig. 1B
). Apoptotic effects of TNF-/IFN- were then tested on trophoblast-like BeWo cells ± BCRP inhibition by Ko 143, and the results were consistent with the effects observed in primary trophoblast (Fig. 1C, D
), suggesting similar mechanisms operating in both cell types.
|
Silencing of BCRP expression in BeWo cells
To identify whether BCRP plays a protective role in trophoblasts, we first generated an shRNA construct targeting BCRP and stably expressed this shRNA in BeWo cells. For control purposes, we utilized the same vector with a scrambled sequence. Stable expression of BCRP shRNA led to a 65% reduction of BCRP mRNA and protein expression (Fig. 2
). Secondly, we transiently silenced BCRP gene expression in BeWo cells with Stealth BCRP siRNA using a pool of three different BCRP sequences at a final concentration 1, 10, 20, or 50 nM. We observed 75% reduction of BCRP mRNA and protein expression in BeWo cells 48 h after transfection with 10 and 50 nM of siRNA (Fig. 3
A, B). Silencing of BCRP gene expression was maintained for at least for 5 days after transfection (data not shown). Inhibition of BCRP resulted in a dramatic, almost 5-fold increase in apoptosis after exposure to mitoxantrone, a cytotoxic high affinity BCRP substrate (Fig. 3C, D
). In control experiments, in which cells were transfected with the same concentrations of the Stealth scrambled siRNA with the same GC content, BCRP expression, and mitoxantrone sensitivity remained unchanged.
|
|
Subsequent experimental procedures were undertaken using BeWo cells with both stable and transient silencing of BCRP. Both models demonstrated similar responses to treatment, with a high degree of consistency between different transfection approaches (stable or transient) and BCRP siRNA sequences. To avoid unnecessary duplication, only the results of the transient siRNA transfection experiments are represented here.
Silencing of BCRP leads to increased BeWo cell sensitivity to proinflammatory cytokines and short-chain exogenous ceramides
Exposure of BeWo cells for 48 h to a combination of TNF- (50 ng/ml) and IFN- (100 U/ml) (TI) after RNAi-mediated BCRP silencing reduced cell viability, as assessed by MTT assay, by almost 40% (P<0.05). Control cells, in contrast, demonstrated minimal changes in MTT reduction after exposure to cytokines (Fig. 4
A). Silencing of BCRP led to an almost 3-fold increase in the number of apoptotic cells after exposure to cytokines compared with control cells and this difference increased with duration of exposure (Fig. 4B, C
). Short-chain (C6 and C8) ceramides exhibited higher levels of toxicity in cells with silenced BCRP compared with controls (Fig. 4A
). The number of apoptotic cells detected by M30 cytokeratin 18 neoepitope staining was also significantly higher in the cells with silenced BCRP after exposure to short-chain ceramides (Fig. 4B
). In contrast, BCRP silencing had no significant effects on apoptotic death in response to serum withdrawal (Fig. 4A-C
), suggesting that the protective effect of BCRP is restricted to certain components of the apoptotic cascade.
|
To investigate whether silencing of BCRP affects plasma membrane architecture and leads to externalization of phosphatidylserine in the outer leaflet of the plasma membrane, we stained cells with anti-phosphatidylserine antibody with and without prior exposure (16 h) to TNF-/IFN- (Fig. 5
A, B). We found a significant increase in phosphatidylserine staining in cells treated with BCRP siRNA compared with scrambled control siRNA under both basal conditions (4.6±1.9 vs. 2.6±1.1 NFU; P<0.05) and after cytokine treatment (11.3±3.6 vs. 5.7±1.3; P<0.05).
|
Silencing of BCRP expression increases endogenous ceramide levels after exposure to proinflammatory cytokines
Since exogenous short-chain ceramides were found to be more apoptotic in cells with silenced BCRP expression, we next investigated levels of endogenous long-chain ceramides in cytokine-treated BeWo cells after BCRP silencing. Ceramides were extracted and analyzed by LC-MS, normalized to the total cholesterol content of the extracts. The results show that cells with silenced BCRP accumulate much higher amounts of endogenous ceramides in response to cytokine exposure. Levels of the four ceramide species examined, C16, C18, C20 and C24, have similar patterns of changes in response to treatments. For simplicity, therefore, these measurements were pooled and shown together as "endogenous ceramides levels" (Fig. 6
). Ceramide levels increased in all cells after exposure to TNF-/IFN- (Fig. 6)
. However, in BCRP-silenced cells intracellular ceramide accumulation was modestly but significantly increased (P<0.05) after cytokine treatment compared with control cells (Fig. 6A
). To assess differences in ceramide efflux, we also measured ceramide levels in conditioned medium. In complete medium (Fig. 6B
), background ceramide levels are high due to the contribution of plasma-derived lipids. Nevertheless, after the background levels were subtracted from the measurements, a significant increase was observed in ceramide concentrations in the media from BCRP-silenced cells compared with controls. To better quantitate ceramide efflux, these experiments were repeated in media omitting FCS and supplemented with fatty acid-free human albumin (0.5%) to act as a sphingolipid acceptor/carrier. Under these conditions, the differences in intracellular ceramide levels between control and BCRP-silenced cells were reduced but remained significant (P<0.05; Fig. 6C
); levels of ceramides in conditioned media were also significantly higher in cells with silenced BCRP (P<0.05; Fig. 6D
).
|
BCRP silencing did not affect levels of PPIX after exposure to TNF- and IFN-
Since BCRP has previously been shown to be implicated in regulation of cell survival via regulation of porphyrin levels in blood stem cells, we measured PPIX accumulation in BeWo cells using flow cytometry to investigate whether a similar mechanism operates in trophoblast cells. Our results showed that levels of PPIX were not changed after exposure to TNF- and IFN- at 12 and 24 h (data not shown), arguing against a role for PPIX is the mechanism of cytokine-induced apoptosis.
BCRP does not rescue BeWo cells from the intrinsic apoptotic pathway
To determine whether BCRP protects trophoblast from apoptosis via a nonspecific mechanism, BeWo cells were exposed to two intrinsic apoptotic activators, (–)-deguelin (iAA1) and 1-(3,4-dichlorobezyl)-1H-indole-2,3-dione (iAA2; Fig. 7
). These compounds act either via a cytochrome c-dependent mechanism (iAA1) or by promoting mitochondrial permeability transition (iAA2). Both agents induced cell death (P<0.05), and BCRP silencing had no effect on the extent of apoptosis, suggesting that BCRP exerts its cytoprotective effect upstream of a common terminal pathway of apoptotic death.
|
Levels of placental BCRP expression are markedly reduced in pregnancies with FGR syndrome
The clinical features of the FGR-affected pregnancies, and the gestation-matched controls are represented in Table 1
. Mean birth weight and placental weight were significantly lower in FGR-affected pregnancies than control (P=0.002 and P=0.03, respectively). In contrast, levels of TNF- mRNA expression were higher in the FGR group (P=0.02). IL-1ß, IL6, and HIF-1 expression was highly variable in FGR placentas and was not significantly different from controls (Fig. 8
). Analysis of BCRP mRNA expression in placentas from FGR pregnancies showed a significant reduction (median [range] 0.021, [0.0009–0.06]) compared to gestational age-matched control tissues (0.036 [0.007–0.07]; P=0.006). Levels of another apical transporter, ABCB subfamily member 1 (ABCB1)/multidrug resistance protein 1 (MDR1), were also significantly lower in FGR placentas (control: 0.285 [0.038–0.81]; FGR, 0.163 [0.014–0.538]; P=0.035). In contrast, expression of ABCC2/multidrug resistance-associated protein 2 (MRP2, apical) and ABCC1/MRP1 (basolateral) were not different between normal and FGR placentas. BCRP and MDR1 expression was strongly correlated (P<0.001) as revealed by the regression analysis. No significant correlations between BCRP or MDR1 expression and TNF-, IL-1ß and IL-6 mRNA levels were found.
|
Localization of BCRP expression in term placenta
The cellular localization of BCRP protein in term placenta was examined by laser confocal microscopy using an anti-BCRP/ABCG2-specific primary antibody (clone BXP-21). As shown in Fig. 9
A, robust immunofluorescence staining was detected on the apical (maternal-facing) side of the trophoblast layer (small arrow). An equally strong signal was also observed on the luminal (fetal-facing) surface of placental capillaries (large arrow) in the majority of terminal and intermediate villi examined (Fig. 9A
). Negative controls exhibited negligible staining (Fig. 9B
).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
/IFN- stimulation compared with control cells. Third, trophoblast cells with silenced BCRP have a much higher rate of phosphatidylserine externalization after exposure to inflammatory cytokines. Finally, we found a significant reduction of BCRP and MDR1 expression in placentas from FGR pregnancies, suggesting reduced transport of their substrates from the placenta and fetus to the maternal compartment. Collectively, these findings indicate that placental BCRP and most likely other ABC transporters play an important role in pregnancy, not only as drug transporters, limiting fetal and placental exposure to xenobiotics, but also in protecting trophoblast from extrinsic apoptotic stimuli, at least in part via prevention of accumulation of endogenous apoptotic mediators. Our data also suggest that a potential reduction in placental and fetal protection from drugs and xenobiotics may be associated with reduced BCRP expression in FGR pregnancies. These findings support the hypothesis that BCRP plays a fundamental protective role in placental cell survival. Our demonstration of dual syncytial and endothelial BCRP localization, consistent with bidirectional transfer of substrates across the placenta, is consistent with this view. It remains to be determined whether this protective mechanism has parallels in other cells and tissues in which ABC transporters have been found to play an antiapoptotic role.
We have recently demonstrated significant stimulatory effects of EGF, IGF-II, and estrogens on BCRP expression in primary trophoblast cells (14)
. These regulators, particularly EGF, are known to be potent survival factors for placental epithelial cells, protecting them from various apoptotic injuries including TNF--induced apoptosis (37
, 38)
. Importantly, EGF has been shown to lower ceramide levels and inhibit ceramide-induced apoptosis in primary trophoblast cells (28)
. It is possible, therefore, that this effect of EGF and other survival agents may be achieved at least in part by up-regulation of BCRP expression.
Our studies strongly argue for a role of ceramides in the mechanisms through which BCRP exerts its antiapoptotic effects. Ceramides belong to the sphingolipid family of 300–400 distinct molecular species involved in regulation of many important cellular functions, including proliferation and apoptosis, in multiple cell types (30
, 39)
. Levels of ceramides markedly increase when cells are exposed to stressors such as TNF-, hypoxia, environmental factors such as heat shock or ultraviolet light, or growth factor withdrawal (30)
. There are a large number of pathways involved in the regulation of ceramide levels in the cell (30)
. To mimic effects of endogenous, highly hydrophobic long-chain ceramides, less hydrophobic short-chain exogenous ceramides are commonly used (40)
, although some limitations of this approach have been reported (41)
. Although we observed a significant elevation of ceramide concentrations and rates of apoptosis in cells with depleted BCRP expression/activity, just how BCRP regulates ceramide levels in the trophoblast cells remains unclear. Several hydrophobic lipids, including cholesterol and phosphatidylcholine, as well as sphingomyelin (a ceramide precursor) are carried by low- density lipoproteins in plasma (42)
and are known to be translocated from cells to plasma via transporters of the ABCA and ABCG subfamilies (41
, 42)
. Ceramides are present in plasma at relatively high concentrations and, being extremely hydrophobic, are also transported by a specific carrier (43)
. A potential explanation for our observations is that BCRP limits the effects of ceramides by promoting their direct transport to the outer leaflet of the plasma membrane, where they are transferred to an acceptor/carrier in the plasma/medium and thus removed from the intracellular environment. We did not, however, observe the decrease in ceramide concentrations in the media from BCRP-silenced cells that would be necessary if this was the explanation. Although ceramide levels are clearly raised in cells with low BCRP expression, they are also increased in complete or albumin-supplemented media taken from these cells, compared with controls. The obvious interpretation of these data is that BCRP is not a major route of ceramide efflux in these cells and that increased intracellular generation is associated with similarly increased ceramide efflux. However, it cannot be completely excluded that the observed increase in apoptosis with BCRP-silencing resulted in greater number of dead or dying cells in the media, contributing to enhanced ceramide levels and thus possibly masking an effect via modulation of BCRP efflux.
Not only did BCRP silencing result in accumulation of ceramides, but BCRP siRNA treatment also resulted in enhanced sensitivity to exogenous toxic short-chain C6 and C8 ceramides, evidence that BCRP can limit trophoblast exposure to these toxic metabolites and protect them from ceramide-induced apoptosis. There is also some evidence that ceramides can regulate ABC activity (43)
, thereby indirectly influencing their cytoprotective function.
Another possible explanation for the increased ceramide levels and apoptosis rates in response to cytokines in BCRP-silenced cells is the reduced ability of these cells to maintain normal plasma membrane architecture, resulting in increased cleavage of sphingomyelin and enhanced accumulation of ceramide in the membranes. The distribution of lipids across biological membranes is highly asymmetric, with phosphatidylserine being located preferentially on the cytoplasmic inner leaflet and sphingomyelin/phosphatidylcholine located predominantly in the outer leaflet (31
, 44
, 45)
. Activation of death receptor signaling (extrinsic pathway) by TNF- may lead to the alteration of this membrane asymmetry and redistribution of phosphatidylserine/choline in the lipid bilayer (39
, 44)
. Some ABC transporters have previously been shown to be involved in maintenance of lipid asymmetry, moving different lipid molecules within the plasma membrane (31
, 41
, 45)
. We have found that when BeWo cells with silenced BCRP undergo fusion, a process known to be associated with transient loss of membrane lipid architecture, levels of apoptosis are markedly higher compared with control cells (46)
. In this present study, cells with silenced BCRP exhibited a significant increase in apoptosis and phosphatidylserine externalisation over control cells after exposure to cytokines. Taken together with data suggesting that fusion is associated with extreme fragility, plus our ceramide observations, these findings point to a role for BCRP in the regulation of plasma membrane architecture, specifically in preventing ceramide generation or/and accumulation in stress conditions. However, more studies are needed to confirm this hypothesis and clarify the mechanisms involved.
Idiopathic FGR is frequently associated with placental insufficiency, caused by alteration of essential metabolic, endocrine, and transport function of the placenta (26
, 47)
. Although hypoxia is one of the key factors affecting placental development during FGR, the effects of low oxygen tension, mediated via HIF-1, mainly targets placental angiogenesis and villogenesis, causing hypercapillarization of the villous vasculature (47
, 48)
. In contrast to highly sensitive capillaries, trophoblast is highly resistant to low oxygen tension; moreover, villous trophoblast proliferation is increased when oxygen tension is low, whereas it is reduced in high oxygen concentrations (48)
. In previous studies, BCRP expression has been shown to be strongly up-regulated by low pO2 in blood cells and trophoblast-like Jar cells, while a role for BCRP in protection from hypoxic injury has been postulated via regulation of toxic protoporphyrin levels in blood cells (7)
. In our study, we failed to obtain evidence of an association between FGR, BCRP expression, and hypoxia; levels of HIF-1 expression in the FGR group were extremely variable (up to 100-fold) and did not correlate with mRNA expression levels of BCRP or the other transporters.
In contrast to hypoxia, proinflammatory cytokines such as those observed in elevated concentrations in various pregnancy complications have direct apoptotic effects on trophoblast (27
, 28
, 49)
. The present study found elevated levels of TNF- expression and increased levels of TNF- induced apoptosis in the FGR group, consistent with previous studies (50
, 51)
. The data also clearly showed decreased expression of both BCRP and MDR1 transporters in the idiopathic FGR group; this may indicate reduced capacity for placental-maternal transport substrates such as drugs and endogenous toxins. Although our previous experiments showed inhibitory effects of TNF- on expression of BCRP and MDR1 mRNA and protein in cultured primary trophoblast (14)
, in this study we did not observe a correlation between the BCRP or MDR1 expression and any of three cytokines examined, despite the finding of elevated TNF- and reduced BCRP and MDR1 expression in idiopathic FGR group. The lack of correlation might reflect the heterogeneity of the group.
In conclusion, our results suggest that BCRP contributes to resistance of trophoblast cells to cytokine-induced apoptosis. This altered resistance is associated with increased intracellular accumulation of ceramides and reduced ability to maintain phosphatidylserine in the inner leaflet of the plasma membrane. The effect of BCRP silencing on apoptosis is minimal when apoptosis is activated via the intrinsic mitochondrial pathway, suggesting a particular specificity to extrinsic apoptotic stimulation. Finally, we report that decreased expression of BCRP is associated with idiopathic FGR, a condition associated with impaired trophoblast function. We conclude, therefore, that BCRP may play a hitherto unrecognized role in the placenta, protecting trophoblast against extrinsic apoptosis via modulation of the ceramide signaling pathway, contributing to the maintenance of placental function under adverse conditions.
|
ACKNOWLEDGMENTS |
|---|
Received for publication March 29, 2007. Accepted for publication May 17, 2007.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  O. Pfister, A. Oikonomopoulos, K.-I. Sereti, R. L. Sohn, D. Cullen, G. C. Fine, F. Mouquet, K. Westerman, and R. Liao Role of the ATP-Binding Cassette Transporter Abcg2 in the Phenotype and Function of Cardiac Side Population Cells Circ. Res., October 10, 2008; 103(8): 825 - 835. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Vore and M. Leggas Progesterone Acts via Progesterone Receptors A and B to Regulate Breast Cancer Resistance Protein Expression Mol. Pharmacol., March 1, 2008; 73(3): 613 - 615. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
