Is 8% O2 more normoxic than 21% O2 for long-term in vitro cultures of human primary term cytotrophoblasts?

STUDY QUESTION: Is 8% O2 a better percentage of atmospheric oxygen for long-term cultures of human primary term cytotrophoblasts than the conventional 21% O2 traditionally used in cell culture? SUMMARY ANSWER: Human primary term cytotrophoblasts are able to differentiate into syncytiotrophoblasts under both atmospheric oxygen levels. WHAT IS KNOWN ALREADY: Cell culture is traditionally done under 21% O2, which is equal to a pO2 of ~160 mm Hg. Based on the pO2 measured after instauration of the blood circulation within the placenta, it has been proposed that cytotrophoblasts culture should be under 8% O2, which is equivalent to 60 mm Hg, and that this percentage should be considered as the physiological normoxia for cytotrophoblasts. STUDY, DESIGN, SIZE, DURATION: Cytotrophoblasts were isolated and purified from human term placentas (n > 4). Cells were cultured under 21% O2 and 8% O2 for 12 days. Several cellular parameters were assessed on Days 2, 4, 8 and 12. PARTICIPANTS/MATERIALS, SETTING, METHODS: Placentas were obtained after vaginal or elective cesarean delivery from uncomplicated pregnancies at term (n ≥ 4). Cell viability was measured by a luminescent assay based on quantitation of the ATP content of living cells. Cell fusion was assessed by quantification of syncytin and e-cadherin mRNA expression by real-time PCR and determination of the fusion index by immunofluorescent microscopy. Trophoblast differentiation was assessed by measuring the expression levels of hCGß, inhibin α subunit (InhA) and placental growth factor (PlGF) by real-time PCR and ELISA. Finally, the effect of the two oxygen levels on apoptosis and cellular oxidative stress was also investigated by quantifying caspase 3/7 activation, superoxide dismutase 1 (SOD-1) mRNA expression and H2O2 generation. MAIN RESULTS AND THE ROLE OF CHANCE: There was no difference between 21% O2 and 8% O2 on cell viability. Cell fusion seemed to be enhanced during the first 4 days when the cells were cultured under 21% O2 compared to 8% O2. The expression level of hCGß was equivalent in both oxygen conditions, indicating that there was no difference in trophoblast differentiation. Interestingly, InhA expression was higher under 8% O2, while PlGF expression was inhibited compared to 21% O2. This latter result indicates that 8% O2 may be more hypoxic than normoxic for in vitro culture of primary term cytotrophoblast. This is further corroborated by the fact that 21% O2 did not significantly increase caspase 3/7 activities and the oxidative stress (SOD-1 mRNA expression and H2O2 generation) in our cell cultures. LARGE SCALE DATA: Not applicable. LIMITATIONS, REASONS FOR CAUTION: The in vitro culture of cytotrophoblasts is artificial and does not reflect the in vivo situation. The cell population is nearly 100% pure, cultured as a monolayer, and the cells bath in a chemically defined culture medium deprived of any oxygen carrier. The oxygen molecules available to the cells are passively dissolved in the medium. The gas dissolution properties of the medium and the cellular consumption rate of oxygen may allow the cells to sustain a wide range of oxygen percentages from 8% to 21%. WIDER IMPLICATIONS OF THE FINDINGS: It is possible to culture human primary term cytotrophoblasts for at least 12 days. The O2 percentage of the air does not negatively affect in vitro cytotrophoblast differentiation. For in vitro culture of cytotrophoblasts, it is not necessary to lower the percentage of atmospheric oxygen to 8%. STUDY FUNDING/COMPETING INTEREST(S): This work was fully supported by 'Fetus for Life' charity. The authors state that there is no conflict of interest to declare regarding the publication of this paper.

HIF1A and EPAS1 potentiate hypoxia-induced upregulation of inhibin alpha chain expression in human term cytotrophoblasts in vitro.

Study Question: Are hypoxia-inducible factors (HIF) responsible for the potentiation of inhibin alpha subunit (INHA) gene expression in primary cultures of human term cytotrophoblasts under low-oxygen tension? Summary Answer: Both HIF1A and endothelial PAS domain protein 1 (EPAS1) are involved in the potentiation of INHA gene upregulation in cytotrophoblasts cultured under hypoxia. What Is Known Already: During the in vitro differentiation of cytotrophoblasts into syncytiotrophoblasts under 21% O2, INHA expression increases. This expression is further increased when cells are cultured under low-oxygen tension (e.g. 2.5% O2). Moreover, in pregnancy-related diseases, such as pre-eclampsia or intrauterine growth restriction (IUGR), in which hypoxia is suspected to be responsible for the abnormal placental development, maternal serum concentration of inhibin A is elevated. Study Design, Size, Duration: Cytotrophoblasts were isolated and purified from human term placentas (n = 6). Cells were cultured under 21% O2, and allowed to differentiate for 48 h. A first group of cells was treated for 16 h under 21% O2 with dimethyloxalylglycine (DMOG) or deferoxamine (DFX), molecules that mimic hypoxia by inhibiting HIF1 proteasomal degradation. Involvement of HIF1A and EPAS1 (also known as HIF2A), two HIF isoforms expressed in trophoblasts, was shown by treating another group of cells cultured under 2.5% O2 with specific inhibitors of HIF1A and EPAS1 for 16 h. INHA mRNA expression was assessed by real-time PCR and secreted inhibin A was quantified by ELISA. The role of HIF1A and EPAS1 in INHA transcriptional regulation was further confirmed by cotransfecting primary cytotrophoblasts with a luciferase reporter plasmid containing a 3.9 kb INHA promoter and plasmids allowing overexpression of HIF1A and EPAS1. Participants/Materials, Setting, Methods: Placentas were obtained after vaginal or elective cesarean delivery from uncomplicated pregnancies at term (n≥ 4). The methods used were hormone measurements in the cell supernatants by enzyme-linked immunosorbent assay, real-time quantitative PCR, western blotting, immunofluorescence microscopy and transient transfection. Main Results and the Role of Chance: HIF1 protein stabilization with DMOG and DFX increased 21% O2-induced INHA mRNA and protein upregulation (P < 0.05 versus control), while hypoxia-induced INHA upregulation was repressed by HIF1A and EPAS1 inhibitors (P < 0.05 versus control). In transfection experiments of primary term cytotrophoblasts, cloned INHA promoter transcriptional activity was increased by 2.5% O2 compared to 21% O2 (P < 0.05). Overexpression of both HIF1A and EPAS1 under 21% O2 increased cloned INHA transcriptional activity (P < 0.001 versus control). Large Scale Data: Not applicable. Limitations, Reasons for Caution: HIF1A and EPAS1 may regulate INHA expression by binding to an hypoxia-responsive element within the promoter, but we were unable to identify such an element. Inhibition of HIF1A and EPAS1 did not completely suppress upregulation of INHA expression suggesting that other transcription factors, not identified or studied here, are involved. Wider Implications of the Findings: Our data suggest that the effect of HIF1 proteins on INHA gene promoter activity may be indirect. By demonstrating the role of HIF1A and especially EPAS1 in INHA gene upregulation under hypoxia, the results suggest that HIF1 proteins may become new therapeutic targets in the treatment of pregnancy-related diseases such as pre-eclampsia or IUGR. Study Funding/Competing Interest(s): This work was fully supported by 'Fetus for Life' charity. C. Depoix was supported by a fellowship 'Fonds de Recherche Clinique' from 'Fondation Saint-Luc', Belgium. The authors declare that there is no conflict of interest regarding the publication of this paper.

Reversible effects of oxygen partial pressure on genes associated with placental angiogenesis and differentiation in primary-term cytotrophoblast cell culture.

Timely regulated changes in oxygen partial pressure are important for placental formation. Disturbances could be responsible for pregnancy-related diseases like preeclampsia and intrauterine growth restriction. We aimed to (i) determine the effect of oxygen partial pressure on cytotrophoblast differentiation; (ii) measure mRNA expression and protein secretion from genes associated with placental angiogenesis; and (iii) determine the reversibility of these effects at different oxygen partial pressures. Term cytotrophoblasts were incubated at 21% and 2.5% O2 for 96 hr, or were switched between the two oxygen concentrations after 48 hr. Real-time PCR and enzyme-linked immunosorbent assays (ELISAs) were used to evaluate cell fusion and differentiation, measuring transcript levels for those genes involved in cell fusion and placental angiogenesis, including VEGF, PlGF, VEGFR1, sVEGFR1, sENG, INHA, and GCM1. Cytotrophoblasts underwent fusion and differentiation in 2.5% O2 . PlGF expression was inhibited while sVEGFR1 expression increased. VEGF and sENG mRNA expressions increased in 2.5% compared to 21% O2 , but no protein was detected in the cell supernatants. Finally, GCM1 mRNA expression increased during trophoblast differentiation at 21% O2 , but was inhibited at 2.5% O2 . These mRNA expression effects were reversed by returning the cells to 21% O2 . Thus, low-oxygen partial pressure does not inhibit term-cytotrophoblast cell fusion and differentiation in vitro. Lowering the oxygen partial pressure from 21% to 2.5% caused normal-term trophoblasts to reversibly modify their expression of genes associated with placental angiogenesis. This suggests that modifications observed in pregnancy diseases such as preeclampsia or growth retardation are probably due to an extrinsic effect on trophoblasts.

Viability of primary term cytotrophoblast cell culture in normoxia and hypoxia.

Cytotrophoblast (CT) cells isolated and purified from term placenta are able to differentiate into syncytiotrophoblast cells. Previous reports suggested that hypoxia is an inhibitor of this differentiation and also increases apoptosis. As visual observations of our CT cell cultures indicated a better development in hypoxia than in normoxia (defined as 2.5 and 21% O(2), respectively), we decided to assess the effect of low oxygen tension on in vitro CT cell differentiation by measuring cell viability, apoptosis and CT cell fusion and differentiation. We observed a 45% decrease in cell viability 24 h after plating both in normoxia and in hypoxia but no difference between the two oxygen conditions. Cell viability remained stable during the 4-day culture. Apoptosis also did not increase in hypoxia. Apoptotic index and caspase activation were even lower in hypoxia than in normoxia at Day 1 and Day 4 of the culture. Finally, we observed a 5-fold and 6-fold increase in Syncytin-1 mRNA expression in normoxia and in hypoxia, respectively, indicating that hypoxia did not inhibit CT cell fusion. CT cells differentiated as well in hypoxia as an increase in inhibin α subunit mRNA was evidenced during the 4-day culture. This increase was even higher in hypoxia than in normoxia. In conclusion, hypoxia defined as 2.5% O(2) based on first trimester placental pO(2) did not decrease term primary CT cell viability and did not increase apoptosis. Moreover, it did not inhibit either CT cell fusion or differentiation.

Control of retinoic acid receptor heterodimerization by ligand-induced structural transitions. A novel mechanism of action for retinoid antagonists.

Heterodimerization of retinoic acid receptors (RARs) with 9-cis-retinoic receptors (RXRs) is a prerequisite for binding of RXR.RAR dimers to DNA and for retinoic acid-induced gene regulation. Whether retinoids control RXR/RAR solution interaction remains a debated question, and we have used in vitro and in vivo protein interaction assays to investigate the role of ligand in modulating RXR/RAR interaction in the absence of DNA. Two-hybrid assay in mammalian cells demonstrated that only RAR agonists were able to increase significantly RAR interaction with RXR, whereas RAR antagonists inhibited RXR binding to RAR. Quantitative glutathione S-transferase pull-down assays established that there was a strict correlation between agonist binding affinity for the RAR monomer and the affinity of RXR for liganded RAR, but RAR antagonists were inactive in inducing RXR recruitment to RAR in vitro. Alteration of coactivator- or corepressor-binding interfaces of RXR or RAR did not alter ligand-enhanced dimerization. In contrast, preventing the formation of a stable holoreceptor structure upon agonist binding strongly altered RXR.RAR dimerization. Finally, we observed that RAR interaction with RXR silenced RXR ligand-dependent activation function. We propose that ligand-controlled dimerization of RAR with RXR is an important step in the RXR.RAR activation process. This interaction is dependent upon adequate remodeling of the AF-2 structure and amenable to pharmacological inhibition by structurally modified retinoids.

Alteration of the glucocorticoid receptor subcellular localization by non steroidal compounds.

The glucocorticoid receptor (GR) engages transient or stable interactions with chaperones (hsp90, hsp70), co-chaperones (p60/hop, hsp40) and several other polypeptides such as immunophilins (Cyp40, FKBP59) and p23 to achieve a high affinity ligand binding state. This complex dissociates in response to hormonal stimuli and holo-GR translocates into the nucleus, where it regulates the activity of glucocorticoid-sensitive genes. GR activity is controlled through its ligand binding domain by steroids displaying either agonistic or antagonistic activity. An alternative approach to modulate GR activity is to target receptor-associated proteins (RAPs), and several non steroidal compounds binding to RAPs affect GR transcriptional activity. We have studied the effect of such drugs on the intracellular localization of a EGFP-GR fusion protein, which has wild type GR pharmacological properties. Agonist and antagonist binding induced nuclear translocation of GR, whereas rifampicin was found to be inactive in our system. Immunosuppressants FK506 and cyclosporin A were able to induce partial nuclear translocation of GR, suggesting that potentiation of glucocorticoid action by these compounds may also proceed through enhanced GR nuclear transfer. Short treatment of cells with the hsp90 inhibitor geldanamycin (GA) did not prevent nuclear translocation of GR. However, longer treatments, in parrallel to the inhibition of GR transcriptional activity, strongly perturbed GR subcellular localization concomitantly to the disruption of the actin network, and caused GR aggregation and down-regulation. The GA-induced transcriptional shutdown was also observed for other nuclear receptors which do not interact stably with hsp90. Thus RAP-binding compounds may exert their effects at least in part through perturbation of the GR cytosol to nucleus partitioning, and identify these proteins as valuable therapeutic targets to control nuclear receptor activity.