| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Published online before print June 28, 2007
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
**
From Mount Sinai Hospital,* Toronto, Ontario, Canada; The Hospital for Sick Children, Toronto, Ontario, Canada; the Departments of Physiology, Obstetrics and Gynecology, and Pediatrics,** Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada; the Department of Obstetrics and Gynecology,¶ Unit of Maternal-Fetal Medicine, University of Turin, Turin, Italy; and New Jersey Medical School,|| Newark, New Jersey
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
Preeclampsia, a unique human disease of placental origin, affects 5 to 7% of all pregnancies and is clinically manifested by the sudden onset of maternal hypertension and proteinurea.4 Preeclamptic placentae are characterized by altered trophoblast differentiation and increased trophoblast cell death and turnover.5-11 Failure in trophoblast-mediated maternal vessel remodeling may predispose the preeclamptic placenta to hypoxia/oxidative stress and as such may impact trophoblast cell death and differentiation, possibly via dysregulated expression or function of Bcl-2 family members.12
We recently reported that increased expression of proapoptotic Mtd-L and Mtd-P, a newly identified isoform of the Bcl-2 family member Mtd/Bok (Matador/Bcl-2-related ovarian killer), results in elevated trophoblast cell death in placentae of patients diagnosed with severe early-onset preeclampsia.11 In addition, we demonstrated that both Mtd-L and Mtd-P isoforms induce mitochondrial apoptotic events.11 The death-inducing Mtd-L molecule preferentially interacts with Mcl-1L protein (myeloid cell leukemia factor 1-L: long isoform), which neutralizes its proapoptotic activity.13,14 The Mcl-1 gene was first described as an early induction gene during myeloblastic leukemia cell differentiation15 and subsequently characterized as an important anti-apoptotic molecule.16,17 Mcl-1 is essential for maintenance of hematopoietic stem cells in bone marrow, for early lymphoid development and later in maintenance and survival of mature lymphocytes via selective inhibition of the proapoptotic protein BIM.18
In addition to the prosurvival Mcl-1L molecule, the Mcl-1 gene also gives rise to a death-inducing splicing isoform known as Mcl-1S. This variant is generated via exon II skipping,19,20 resulting in loss of the BH1, BH2, and transmembrane domains, hence generating a truncated proapoptotic BH3-only isoform. Mcl-1L is presently the sole known molecule capable of binding and antagonizing the killing function Mcl-1S.19 Recent studies have demonstrated that activated caspase-3 leads to proteolytic cleavage of conserved aspartate residues (D127 and D157) within the PEST domain of both Mcl-1 variants, hence generating truncated molecules21-25 that exhibit proapoptotic functions.21,25,26
Oxygen plays a key role in the transcriptional regulation of both Mcl-1 and Mtd.11,27,28 We recently reported that Mtd is overexpressed in preeclamptic placentae,11 but the expression status of Mcl-1 in this pathology is currently unknown. Herein, we tested the hypothesis that reduced pO2 as seen in preeclampsia alters the Mtd/Mcl-1 apoptotic rheostat regulating trophoblast turnover. We investigated Mcl-1 expression in human pregnancies complicated by preeclampsia and examined how conditions of in vitro and in vivo placental hypoxia, including altitude-induced chronic hypoxia, affect the Mcl-1/Mtd apoptotic rheostat. Our findings indicate that pathological conditions of oxidative stress compared with physiological low pO2 induces a detrimental switch in the mitochondrial apoptotic balance via caspase-mediated cleavage of prosurvival Mcl-1 into proapoptotic isoforms.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Tissue was collected in accordance with participating institutions’ ethics guidelines. Preeclampsia was defined and diagnosed based on the American College of Obstetrics and Gynecology criteria (blood pressure of 140 mmHg systolic or higher and 90 mmHg diastolic or higher occurring after 20 weeks of gestation in a woman with previously normal blood pressure accompanied by urinary excretion of 0.3 g of protein or higher).29
The fifth Korotkoff sound was used for measuring diastolic blood pressure. Preeclamptic placentae (PE; early-onset, n = 49; late-onset/term, n = 10) and preterm normotensive age-matched control placentae (AMC, n = 38) were collected from deliveries at Mount Sinai Hospital. Areas with calcified, necrotic, or visually ischemic tissue were omitted from sampling. Patients suffering from diabetes, kidney disease, or infections were excluded. Early-onset preeclampsia was defined when the patient delivered before 34 weeks of gestation because of preeclampsia. Pregnant patients with essential hypertension (n = 4, at term) and pregnancies affected by intrauterine growth restriction (IUGR) (n = 6, gestational age 35 to 38 weeks with fetal weight less than fifth percentile) without preeclampsia were included as controls. Preterm deliveries were attributable to multiple pregnancies (30%), preterm labor attributable to incompetent cervix (40%), premature preterm rupture of membrane (20%), and spontaneous rupture of membranes (10%). Preterm and term control groups had no clinical or pathological signs of preeclampsia, infections, or other maternal or placental disease. First-trimester human placental tissues (6 to 20 weeks of gestation, n = 16) were obtained from elective terminations of pregnancies by dilatation and curettage. High-altitude (HA; 3100 m, n = 16) and moderate altitude (MA; 1600 m, n = 13) and sea-level/term control (SL, n = 15) placental samples were collected from healthy normal patients with term deliveries. Because of organ heterogeneity and the fact that perfusion can differ depending on location within the placenta,30
multiple specimens (approximately five per region) were randomly sampled from central and peripheral regions of pathological, HA, and control placentae. The SL, MA, and HA groups did not show clinical or pathological signs of preeclampsia, infection, or other placental disease. For summary of clinical data, see Table 1
.
|
Explant cultures were performed as previously described.31 In brief, placental tissues (5 to 8 weeks of gestation) were placed in ice-cold phosphate-buffered saline (PBS) and processed within 2 hours of collection. Tissues were aseptically dissected to remove decidual tissue and fetal membranes. Small fragments of placental villi (25 to 45 mg wet weight) were teased apart, placed on Millicell-CM culture dish inserts (Millipore Corp., Bedford, MA) precoated with 0.15 ml of undiluted Matrigel (Collaborative Biomedical Products, Bedford, MA), and transferred to a 24-well culture dish. Explants were cultured in serum-free Dulbecco’s modified Eagle’s medium/F12 (Life Technologies, Inc., Grand Island, NY) supplemented with 100 µg/ml streptomycin, 100 U/ml penicillin, and incubated overnight at 37°C in 5% CO2 in air to allow attachment. Explants were maintained in standard condition (5% CO2 in 95% air) or in an atmosphere of 3% O2/92% N2/5% CO2 for 48 hours at 37°C. The morphological integrity and viability of villous explants under the various oxygenation conditions were monitored daily as previously reported.11,31 Explants from more than 18 different placentae in more than 15 separate experiments were used. A minimum of three explants per experimental condition was used at all times. Explants were exposed to hypoxia-reoxygenation (HR) as previously described11,32 in the presence of 100 µmol/L of the pan-caspase inhibitor z-VAD-fmk (BioMol, Plymouth Meeting, PA) or caspase-3-specific inhibitor z-DEVD-fmk (R&D Systems, Minneapolis, MN) both dissolved in dimethyl sulfoxide (equivalent volume of dimethyl sulfoxide in the absence of inhibitors was used in control conditions).
RNA Analysis
RNA extraction was performed using a RNeasy mini kit (Qiagen, Valencia, CA), reverse-transcribed using a random hexamer approach, and amplified by 40 cycles of quantitative polymerase chain reaction (PCR) (15 minutes at 95°C, cycle: 30 seconds at 95°C, 30 seconds at 60°C, and 30 seconds at 72°C). Southern blot analysis was performed as previously described,11 using a 32P-labeled full-length Mcl-1L probe. Primer sequences used in reverse transcriptase (RT)-PCR/Southern blot analysis were: Mcl-1 (NM_021960): forward 5'-ATGTTTGGCCTCAAAAGAAACGCG-3' and reverse 5'-GGCTATCTTATTAGATATGCCAA-3' (Mcl-1L: predicted size = 1054 bp and Mcl-1S: predicted size = 806 bp) and ß-actin (NM_001101): forward 5'-CTTCTACAATGAGCTGCGTG-3', reverse 5'-TCATGAGGTAGTCAGTCAGG-3' (predicted size = 304 bp). All amplified and cloned products were confirmed by sequencing. Quantitative PCR was performed using the SYBR Green I dye DyNamo HS kit (MJ Research, Waltham, MA) based on the manufacturer’s protocol using isoform-specific primers for Mtd-L and Mtd-P (Mtd-L: forward 5'-GCCTGGCTGAGGTGTGC-3', Mtd-P: forward 5'-GCGGGAGAGGCGATGA-3', reverse (both L and P) 5'-TGCAGAGAAGATGTGGCCA-3', Mcl-1L: forward 5'-ATGCTTCGGAAACTGGACAT-3', Mcl-1S: forward 5'-CCTTCCAAGGATGGGTTTG-3', Mcl-1: reverse (both L and S) 5'-CTAGGTTGCTAGGGTGCAA-3'). For syncytin and cytokeratin 7 analyses, qRT-PCR was performed using Assays-on-Demand TaqMan primers and probe (Applied Biosystems, Foster City, CA). Analysis was done using the DNA Engine Opticon2 System (MJ Research). Data for all qPCR analyses were normalized against expression of 18S ribosomal RNA as previously described.33
Western Blot Analysis
Western blotting was performed as previously described.34 Fifty µg of total protein from placental tissue was subjected to 12% (w/v) sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Membranes were probed at 4°C overnight with a 1:1000 dilution of primary antibodies: rabbit polyclonal Mtd antibody capable of recognizing all isoforms as previously reported11 ; Mcl-1-specific rabbit polyclonal antibody (SC-819 clone S-19 from Santa Cruz Biotechnology, Santa Cruz, CA); specific cleaved caspase-3 (Asp175) (5A1); cleaved caspase-8 (Asp384) rabbit polyclonal antibodies (Cell Signaling, Beverly, MA), and anti-PARP p85 (active PARP) fragment rabbit polyclonal antibodies (1:300 dilution; Promega, Madison, WI). Rabbit polyclonal antisera against N-terminal amino acids (28 to 41) of syncytin (NM_014590) was generated by injecting rabbits first with 400 µg of KHL-conjugated peptide in Freund’s complete adjuvant followed by three boosts of 100 µg in incomplete adjuvant. Titer was checked by enzyme-linked immunosorbent assay using bovine serum albumin-conjugated peptide as antigen. For anti-Mtd, Mcl-1, and syncytin antibodies, preimmune serum and competing peptides were used as controls. After overnight incubation, membranes were washed with TBS/T and incubated for 60 minutes at room temperature with 1:5000 diluted horseradish peroxidase-conjugated anti-rabbit (Santa Cruz Biotechnology). Blots were exposed to chemiluminescent ECL-plus reagent (Amersham, Piscataway, NJ). All blots were confirmed for equal protein loading and transfer using Ponceau staining.35
Immunohistochemistry
Immunohistochemical analyses were performed using an avidin-biotin-based immunoperoxidase approach, as previously described.32 Nonspecific binding sites were blocked using 5% (v/v) normal goat serum and 1% (w/v) bovine serum albumin in Tris-buffer. Slides were incubated overnight at 4°C with a 1:200 dilution of rabbit polyclonal anti-Mtd or anti-Mcl-1 antibodies. After washing, slides were probed with a 300-fold dilution of biotinylated goat anti-rabbit or goat anti-mouse IgG (Vector Laboratories, Burlingame, CA) for 1 hour at 4°C. Avidin-biotin complex was applied for 1 hour. Slides were developed in 0.075% (w/v) 3,3-diaminobenzidine in PBS (pH 7.6) containing 0.002% (v/v) H2O2, counterstained with hematoxylin, dehydrated in an ascending ethanol series, cleared in xylene, and mounted. In control experiments, primary antibodies were replaced with blocking solution [5% (v/v) normal goat serum and 1% (w/v) bovine serum albumin].
Transfection Experiments
Human choriocarcinoma JEG-3 cells (75% confluence) were transfected with 1.5 µg/35-mm dish of either empty pcDNA3.1 vector or pcDNA-3-Mcl-1L vector (kindly provided by Dr. Ruth Craig, Dartmouth Medical School, Hanover, NH) using Lipofectamine 2000 (Invitrogen, Carlsbad, CA). Twenty-four hours after transfection, cells were treated with 10 mmol/L of the NO donor sodium nitroprusside (SNP; Sigma, St. Louis, MO) or exposed to HR for 5 hours (3 hours at 3% O2 followed by 2 hours at 20% O2) as previously described.11 Cells were then harvested, and the percentage of dead cells was assessed by trypan blue staining.
Statistical Analysis
Data are represented as mean ± SEM of at least three separate experiments performed in triplicate. For comparison of data between multiple groups, we used Kruskal-Wallis one-way analysis of variance with posthoc Dunnett’s test. For comparison between two groups, we used a paired and unpaired Student’s t-test as appropriate. Significance defined as P < 0.05.
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Placental transcript and protein expression of Mcl-1 was first assessed in placentae of severe early-onset preeclamptic patients (PE) relative to AMCs. Quantitative real-time PCR (qRT-PCR) using isoform-specific primers for anti-apoptotic Mcl-1L and proapoptotic Mcl-1S demonstrated that Mcl-1L mRNA expression was unchanged between PE and AMC, whereas Mcl-1S expression was significantly increased in PE (fourfold, P = 0.001) relative to AMCs (Figure 1a)
. Sequence analysis of the PCR products confirmed identity of Mcl-1L and Mcl-1S (data not shown).
|
37 kd) was decreased in early-onset PE relative to AMC (Figure 1b)
despite no change in transcript levels (Figure 1a)
. In addition, two shorter Mcl-1-immunoreactive bands migrating at relative molecular weights of 29 and 26 kd, respectively, were observed, both of which seemed to be more abundant in PE tissues (Figure 1b)
. Preabsorption of anti-Mcl-1 antibody with corresponding peptide abolished the appearance of all Mcl-1 immunoreactive bands (data not shown). Densitometric analysis confirmed that in early PE Mcl-1L, protein expression was decreased by 25% (P = 0.01), whereas Mcl-1 p29 and Mcl-1 p26 expression increased by twofold (P = 0.003) and 3.6-fold (P = 0.01), respectively (Figure 1c)
. No notable differences were observed in Mcl-1L and Mcl-1 p29 expression between late preeclampsia relative to normal term deliveries and nonlabor C/S deliveries (Figure 1d)
. Furthermore, the controls age-matched to the severe early-onset preeclampsia (Figure 1b)
, despite clinical heterogeneity in the cause of preterm delivery, did not differ in Mcl-1 isoform profile expression relative to the term deliveries with and without labor as shown in Figure 1d
. When Mcl-1 protein expression was compared between early-onset PE placentae and other placental pathologies (IUGR pregnancies, late-onset preeclampsia associated with IUGR, essential hypertension) and normal term tissues, the switch in Mcl-1 protein profile was predominantly once again observed only in placentae of patients with the early severe form of PE (Figure 1e)
. Mcl-1 Cleavage in Human Villous Explants Is Regulated by Caspase Activation
The appearance of several specific Mcl-1-immunoreactive bands suggested the presence of various Mcl-1 isoforms. Taking into account our transcript data, the band migrating at 26 kd is consistent with the predicted molecular weight of Mcl-1S, originally cloned from a placental cDNA library.19,20
To investigate whether the p29 Mcl-1-immunoreactive band can be a result of caspase-mediated cleavage, we exposed first trimester placental explants to conditions of HR known to induce caspase activation and trophoblast apoptosis in vitro.11,36
Explants were exposed to HR in the presence or absence of z-VAD-fmk, a broad-based inhibitor of caspase activity and caspase-mediated Mcl-1L cleavage.21-23,26
Relative to untreated controls, tissues exposed to HR demonstrated a notable Mcl-1 isoform switch; ie, Mcl-1L protein decreased with a concomitant increase of p29 immunoreactive Mcl-1 protein (Figure 2a)
. Exposure of HR-treated explants to z-VAD-fmk abrogated the appearance of the immunoreactive p29 Mcl-1 band, restored Mcl-1L content, and increased Mcl-1S amount relative to HR and untreated tissues (Figure 2, a and c)
. To investigate whether Mcl-1 cleavage under HR was mediated by executioner caspases, explants under HR were incubated in the presence and absence of a caspase-3/7-specific inhibitor. Similar to z-VAD-fmk treatment, z-DEVD-fmk exposure under HR conditions prevented caspase-mediated Mcl-1 p29 formation and increased both Mcl-1L and Mcl-1S levels (Figure 2b)
. Thus, both Mcl-1L and Mcl-1S seem to be regulated by caspase cleavage under HR stress and p29 Mcl-1 is most likely the previously described proapoptotic cleavage product of Mcl-1L.21-25
|
Because preeclampsia is associated with placental hypoxia/oxidative stress, the effects of varying oxygenation on Mcl-1 expression were assessed in vitro using villous explants exposed to 20% oxygen, 3% oxygen, and HR. In 3% oxygen (normal oxygen tension for <8 weeks) relative to standard 20% O2 conditions, mRNA expression of Mcl-1L significantly increased (more than 2.5-fold, P = 0.0095), whereas that of Mcl-1S decreased (0.2-fold, P = 0.01) (Figure 3a)
. The opposite expression profile was observed under HR conditions in which Mcl-1L decreased (twofold, P = 0.01) and Mcl-1S increased (2.5-fold, P = 0.01) relative to 20% control (Figure 3a)
. The protein profile of Mcl-1 under 3% oxygen showed increased Mcl-1L expression relative to the 20% O2 controls (Figure 3b)
. Importantly, HR resulted in markedly decreased levels of anti-apoptotic Mcl-1L and a concomitant increased formation of proapoptotic Mcl-1c (p29 Mcl-1) relative to 20% O2 (Figure 3b)
. Although the expression of Mcl-1S at the transcript level was markedly increased in HR, this pattern was not evident at the protein level.
|
10 to 12 weeks of gestation when the intervillous space opens to maternal circulation. Rapid placental oxygenation, similar to a reperfusion injury, may hence result in a transient state of oxidative stress.37
Therefore, we hypothesized that Mcl-1 cleavage would occur at that period. Indeed, there was increased mRNA expression of both Mcl-1L and Mcl-1S (Figure 3c
, left) and formation of Mcl-1c (p29 Mcl-1) at 10 to 13 weeks (coinciding with the placental surge of oxidative stress), which declined at later gestational ages (Figure 3c
, right).
Next, we assessed the function of Mcl-1L under conditions of oxidative stress (HR) because this condition has been shown to be the strongest inducer of Mtd-L and -P expression.11
Because SNP, a nitric oxide donor, is a powerful inducer of cell death,38
SNP treatment was used as positive control. Both HR and SNP treatments of JEG-3 cells significantly increased cell death when compared with cells maintained at 20% O2 (Figure 3d)
. Mcl-1L overexpression in JEG-3 cells significantly decreased HR or SNP-induced cell death when compared with cells transfected with empty vector (Figure 3d)
.
Mcl-1 Expression in in Vivo Chronic Placental Hypoxia
As previously demonstrated, preeclamptic placentae have a global gene expression similarity to placental tissues obtained from HA pregnancies because both conditions may be affected by aberrant placental oxygenation.39
As such, Mcl-1 expression was next examined under physiologically reduced placental oxygenation using HA placentae from normal pregnancies. Placental Mcl-1 isoform mRNA expression was different between HA and MA relative to SL control placentae (SL). Although Mcl-1L transcript was significantly increased in HA (2.3-fold, P = 0.005) and MA (2-fold, P = 0.022) relative to SL, Mcl-1S transcript levels were significantly decreased in HA and MA relative to SL samples (HA versus SL, 0.4-fold, P < 0.05; MA versus SL, 0.5-fold, P = 0.016) (Figure 4, a and b)
.
|
.
Similar to Mcl-1 mRNA expression, Mcl-1L protein expression was increased in HA and to a lesser extent in MA when compared with SL samples (Figure 4e)
. Both Mcl-1Lc and Mcl-1S molecules were hardly detectable at the protein level in SL, MA, and HA samples with no apparent change in expression (Figure 4e)
. Although expression of all known Mtd proteins (L: 28 kd; S: 18 kd; and P: 15 kd) could be observed, no apparent difference in Mtd expression was noted between the conditions tested (Figure 4f)
.
Mcl-1 immunoreactivity in placental sections from SL, MA, and HA samples was predominantly observed in trophoblast cell layers (Figure 4g
, top) with low to absent staining in stromal cells. Mtd immunoreactivity was generally low and also restricted to trophoblast cells. Positive immunoreactivity for Mtd was greater in SL tissue compared with HA and MA tissues (Figure 4g
, bottom). No positive staining was observed in control sections in which Mtd and McL-1 antibodies were replaced with nonimmune IgG (data not shown).
Caspase Activation in in Vivo Chronic and Pathological Placental Hypoxia
Excessive cell death via death receptor is a well-known phenomenon occurring frequently in trophoblast cells of preeclamptic placentae.3,40,41
Although this death pathway was described previously, downstream events related to caspase-8 activation have not been explored in PE tissues. Relative to AMC tissues, cleaved caspase-8 was increased in severe early-onset preeclampsia (Figure 5a
, left), consistent with previous reports showing increased activation of caspase-3.42
|
, right). Similar to cleaved caspase-3, expression of activated caspase-8 in HA was decreased by 40% relative to controls (MA and SL) (P = 0.0335) (Figure 5b
, left). We validated our apoptosis findings by measuring the p85 fragment of cleaved poly-ADP-ribose polymerase (PARP), another classical marker of apoptosis. As anticipated, p85 PARP was increased in preeclamptic tissue relative to age-matched normotensive controls (Figure 5a
, right), whereas a marked decrease in p85 PARP content was observed in HA samples relative to SL and MA tissues (Figure 5c)
. Trophoblast Cell Fusion in Pathological and in Vivo Chronic Placental Hypoxia
Because placentae from HA pregnancies exhibit thinning of trophoblast membranes possibly because of reduced trophoblast cell death/turnover,43
we next investigated the expression of syncytin,44,45
a typical marker for cytotrophoblast fusion into syncytium. HA samples had significantly reduced syncytin mRNA expression relative to MA or SL tissues (Figure 6a)
. The decreased syncytin expression at HA was not because of a shift in trophoblast cell population because expression of trophoblast-specific cytokeratin 7 was similar in HA and SL placentae (Figure 6b)
.
|
59 kd (the theoretical molecular weight of syncytin) was recognized only when the postimmune serum of rabbits was used (Figure 6c)
. This band was significantly reduced when competed with the peptide used for immunization (Figure 6c)
. Similar to transcript levels, syncytin protein expression was decreased in conditions of chronic placental hypoxia as determined by expression in HA placentae relative to control samples (Figure 6d)
. Exposing first trimester explants to 3% O2 also resulted in decreased syncytin expression (Figure 6e)
. We also tested syncytin expression in pathological conditions of placental hypoxia and found that syncytin mRNA and protein expression levels were reduced in preeclamptic placentae relative to controls (Figure 6, f–h)
, in agreement with previous reports.46-48
Last, we examined the spatial localization of syncytin (Figure 7)
. In AMC SL placentae, syncytin immunostaining localized to the trophoblast layers. Positive immunoreactivity for syncytin was markedly decreased in the trophoblast layers of both HA and PE placentae.
|
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Among Bcl-2 family members, Mcl-1 function is uniquely regulated via complex transcriptional, posttranscriptional, and post-translational processes. The PEST domain found in both Mcl-1 splicing isoforms is recognized as a substrate for caspase-3-mediated cleavage.24 This cleavage is a unique regulatory mechanism conferring differential Mcl-1 protein function.21-26 Mcl-1L is a potent anti-apoptotic molecule believed to sequester other members of the proapoptotic channel forming Bcl-2 subfamily such as Bak.49 Although it has been established that Mcl-1L neutralizes the killing ability of Mtd/Bok,13 the precise molecular events leading to suppression of cell death in this pathway remain unclear.13
Oxygen is a potent regulator of apoptotic cell death. Several Bcl-2 family members, including BH3-only ligands such as Nix and Nip as well as Mcl-1,28 have been shown to be directly regulated by oxygen via HIF-1.50,51 Our observation of increased Mcl-1L expression under in vitro or in vivo reduced oxygenation is consistent with HIF-1 mediated Mcl-1L regulation.28 Furthermore, we found increased processing of proapoptotic Mcl-1c and Mcl-1S in vivo at 10 to 13 weeks, when trophoblast cells experiences a rapid surge in oxygenation, and in vitro in villous explants undergoing HR injury. Thus, transcriptional regulation of Mcl-1 and differential expression of its isoforms are likely attributable to specific oxygen conditions experienced by the placental tissue. Although the expression of Mcl-1S at the transcript level was markedly increased in HR, this pattern was not evident at the protein level. This could be explained by the fact that Mcl-1S, being a low abundant isoform (relative to the L), is rapidly degraded/cleaved at the protein level because of caspase activation under HR conditions or could also reflect alteration of splicing without a concomitant elevation in translation or increased McL-1S stabilization. Thus, to our knowledge this is the first evidence of regulation of Mcl-1 expression and cleavage by oxygen in a human organ and importantly dysregulation of these events in a human disorder.
In severe early-onset preeclampsia, excess trophoblast cell death, likely induced by hypoxia (10 mmHg or <1 to 2% O2)52,53
or intermittent oxygenation36
increases trophoblast shedding, which is believed to generate a Th1-type maternal inflammatory response and generalized maternal endothelial cell injury.54,55
A tilt in Mcl-1 expression toward generation of death-inducing molecules in severe preeclamptic placentae combined with increased expression of killer Mtd-P isoform in this disease,11
initiates a detrimental pathological switch toward trophoblast demise accompanied by increased shedding. Our transfection studies showing that anti-apoptotic Mcl-1L overexpression can rescue trophoblast (JEG-3) cells from undergoing apoptosis under detrimental oxygen conditions supports the idea that pathological reduced oxygen tilts the Mcl-1/Mtd balance toward cell death. In contrast to preeclamptic placentae, placental tissues from HA experience chronically reduced oxygenation, estimated at 20% below the intervillous pO2 of 60 mmHg (8% O2) observed in near term SL pregnancy or the equivalent of 5 to 6% O2. Interestingly, normal placentae from HA conditions display decreased markers of apoptosis and increased Mcl-1L expression. In support of our molecular findings, previous studies have reported fewer areas of syncytial damage or denudations in HA placentae relative to lower altitudes,56
again suggesting slowed apoptosis-mediated trophoblast turnover. Hence, conditions of chronic hypoxia may actually stimulate an adaptive molecular response by minimizing the burden of apoptotic-mediated trophoblast shedding in these pregnancies. Such speculation may be controverted since an association between HA residence and an increased incidence of preeclampsia has been reported.57
Although the altitude pattern of increased placental expression of anti-apoptotic molecules in HA tissues may to a certain extent protect against trophoblast cell death, other factors may contribute to the increased incidence of preeclampsia under chronic hypoxia.
Similar to increased Mtd-L and Mtd-P expression in early-onset severe preeclampsia,11 cleavage of Mcl-1L in this pathology may be another unique placental feature of the early severe form of this disease, which does not occur in control aged-matched patients, in late-onset preeclampsia even with IUGR or in other placental pathologies, including idiopathic IUGR and essential hypertension. This molecular evidence once again reiterates the need and importance for proper pathological classification of preeclampsia, conceivably because of different underlying etiologies involved in the pathogenesis of this multifaceted and often misconstrued disorder.
Normal physiological regulation of trophoblast cell fusion utilizes a molecular apoptotic cascade in which mononucleated cytotrophoblast cells fuse to maintain a multinucleated syncytiotrophoblast cell layer.58
Normal SL term placentae have a well-defined syncytiotrophoblast layer, which is in contrast to HA placentae, which show thinning of the syncytiotrophoblast layer and a hyperproliferative cytotrophoblast phenotype59
suggesting decreased syncytial renewal under conditions of reduced oxygenation.56,60,61
Decreased expression of active caspase-8, a promoter of trophoblast cell fusion,62
in conjunction with reduction of other cell death markers in HA placentae, as demonstrated by decreased cleaved p85 PARP formation, suggests a slowdown of syncytial formation and consequently trophoblast shedding during chronic placental hypoxia. Interestingly, in pathological placental hypoxia (preeclampsia), increase in caspase-336,42
and caspase-8 activation together with cleavage of PARP indicate increased apoptosis. Activation of caspase-8 in PE could also be secondary because of aberrant expression of its upstream activators such as activation of death receptors by elevated FasL and tumor necrosis factor-
as previously reported.6,7,9-11,40,41
Decreased syncytin expression, a key regulator of trophoblast cell fusion,44,45
may also negatively impact normal trophoblast differentiation resulting in thinning of the placental syncytium at HA. Studies have shown that in conditions of reduced oxygenation, syncytin as well as its binding receptor (ASCT2) are down-regulated relative to standard oxygenation conditions.63
Decreased syncytin expression in pregnancies complicated by preeclampsia has also been reported.46,48
These findings support our observations of reduced syncytin expression in conditions of altitude-induced placental hypoxia or pathological hypoxia (PE) relative to control. Impaired regulation of trophoblast fusion in HA placentae may limit de novo syncytial synthesis and maintenance at the expense of normal shedding/deportation of syncytial debris. Hence, decrease in the dynamic rate of syncytial renewal and shedding may provide a molecular explanation with respect to the thinned syncytial phenotype observed in HA placentae.
In conclusion, this study provides evidence that the Mcl-1/Mtd rheostat regulates trophoblast apoptosis in a different manner under physiological and pathological conditions of placental hypoxia. Although both preeclampsia and HA placentation are characterized by aberrant villous trophoblast differentiation, possibly because of down-regulation of syncytin expression; only preeclamptic placentae experience an apoptotic insult, believed to be responsible for accelerated trophoblast turnover. In HA pregnancies, decreased trophoblast cell death in conjunction with decreased trophoblast turnover/differentiation may present an important adaptive response to chronic placental hypoxia. The impact of aberrant oxygenation on Mcl-1 isoform expression and stability in preeclamptic placental tissues may influence events leading to the clinical manifestations of this disease.
|
Acknowledgements |
|---|
|
Footnotes |
|---|
Supported by the Canadian Institutes of Health Research (grant MOP-62845, a new investigator award to A.J., and a mid-career award from the Ontario Women’s Health Council to I.C.), the National Institutes of Health (grant HD042737), the Hospital for Sick Children (Restracomp training program grant to N.S.), and the University of Toronto (Ontario graduate scholarship program grant to N.S.).
M.P. is the holder of a Canadian research chair (tier 1) in respiration.
Accepted for publication April 13, 2007.
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
C(T) method. Methods 2001, 25:402-408[CrossRef][Medline]
B and placental apoptosis in pregnancies complicated with intrauterine growth restriction and preeclampsia: an immunohistochemical study. Tohoku J Exp Med 2004, 204:195-202[CrossRef][Medline]
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS |
