Pericellular oxygen concentration of cultured primary human trophoblasts

Abstract

Introduction

Oxygen is pivotal in placental development and function. In vitro culture of human trophoblasts provides a useful model to study this phenomenon, but a hotly debated issue is whether or not the oxygen tension of the culture conditions mimics in vivo conditions. We tested the hypothesis that ambient oxygen tensions in culture reflect the pericellular oxygen levels.

Methods

We used a microelectrode oxygen sensor to measure the concentration of dissolved oxygen in the culture medium equilibrated with 21%, 8% or <0.5% oxygen.

Results

The concentration of oxygen in medium without cells resembled that in the ambient atmosphere. The oxygen concentration present in medium bathing trophoblasts was remarkably dependent on the depth within the medium where sampling occurred, and the oxygen concentration within the overlying atmosphere was not reflected in medium immediately adjacent to the cells. Indeed, the pericellular oxygen concentration was in a range that most would consider severe hypoxia, at ≤ 0.6% oxygen or about 4.6 mm Hg, when the overlying atmosphere was 21% oxygen.

Conclusions

We conclude that culture conditions of 21% oxygen are unable to replicate the pO₂ of 40–60 mm Hg commonly attributed to the maternal blood in the intervillous space in the second and third trimesters of pregnancy. We further surmise that oxygen atmospheres in culture conditions between 0.5% and 21% provide different oxygen fluxes in the immediate pericellular environment yet can still yield insights into the responses of human trophoblast to different oxygen conditions.

INTRODUCTION

Oxygen is the Janus gas[1], and the two-faced effect of this necessary but potentially damaging gas is especially important in human placental development. Low oxygen tension at < 20 mm Hg is a prerequisite for extra villous trophoblasts to optimally invade decidua and establish the early placenta [2]. Oxygen tension rises at 10–12 weeks’ gestation to offer an estimated 40–60 mm Hg pO₂ in the intervillous space where maternal blood bathes villous trophoblasts [3–5]. These critical variations in oxygen tensions in vivo have created a hotly debated issue in placental biology: what oxygen conditions should be used for in vitro experiments to best mimic conditions in vivo? Reviewers of papers and attendees at placental-focused conferences have strong opinions about this question, yet little data exist to support or refute the different perspectives. We do not resolve this controversy here. Instead we test the hypothesis that ambient atmospheric oxygen tensions reflect pericellular oxygen levels experienced by cultured trophoblasts. We used a microelectrode to directly measure pericellular oxygen concentrations in the culture medium of primary human trophoblasts. We discovered that this hypothesis is simplistic and wrong.

METHODS

Cell culture

This study was approved by the Institutional Review Board of Washington University in St Louis. Primary human cytotrophoblasts were isolated from term placentas, from uncomplicated pregnancies delivered by scheduled cesarean sections, by the trypsin-deoxyribonuclease-Dispase/Percoll method as described previously [6]. Three million cells were plated at a density of 3.4 × 10⁵ per cm² with 2 ml Dulbecco’s modified Eagle’s medium (DMEM, Sigma, St. Louis, MO) containing 10% fetal bovine serum (Sigma), 20 mM HEPES (pH 7.4, Sigma), 100 U/ml penicillin, 100 μg/ml streptomycin, and 0.25 mg/ml Fungizone (all from the Washington University Tissue Culture Support Center), at 37 °C in 5% CO₂-air atmosphere with 21% O₂ (standard conditions), in a 35 mm culture dish (total culture area: 8.8 cm²; total volume of dish: 9.3 ml; total height of 2 ml medium: 1.8~1.9 mm. Nunc, Rochester, NY). After allowing cells to attach for 4 h, the dishes were washed thoroughly three times with PBS to eliminate nonattached cells and villous fragments and then replenished with 2 ml of 37°C pre-warmed, atmospheric-air-equilibrated culture medium and used for dissolved oxygen concentration measurements.

Incubators and hypoxia chambers

Two Forma Scientific anaerobic chambers at 37°C, one set at 5% CO₂/8% O₂/87% N₂ (oxygen sensor reading 8% O₂), and the other set at 5% CO₂/10% H₂/85% N₂ (oxygen sensor reading <0.5% O₂), were used to equilibrate culture medium to make “8%” and “<0.5%” media, respectively. A Forma Scientific water-jacketed incubator (Forma Scientific, USA) that provides standard culture conditions (21% O₂) was used to equilibrate standard culture medium and for normal culture of primary human trophoblasts.

Calibration of the oxygen sensor

A micro dissolved-O₂ electrode (DO-166MT-1, LAZAR research laboratories, Los Angeles, CA), which has a 3 mm wide tip, was used to measure the dissolved oxygen concentration in the culture medium. The concentration was recorded every second with the ArrowDO sampler program provided with the electrode. The microelectrode oxygen sensor was calibrated according to the manufacturer’s instructions to 21% O₂ in the incubator that provides standard culture conditions, and to an atmosphere of 0% O₂ in nitrogen gas.

Measurement of dissolved oxygen concentration in medium without cells

Culture dishes (35 mm, n=3) containing 2 ml of fresh medium without cells were transferred to 21%, 8%, or <0.5% oxygen incubators for 24 h until measured. The microelectrode was clamped vertically to a Narishige micromanipulator (Tokyo, Japan), which allowed precise adjustment of the electrode height (Figure 1 A). The microelectrode was introduced through a 5 mm hole in the center of the dish cover (Figure 1 B), and placed as close to the bottom of the culture dish as feasible without damaging the tip; this position was designated as zero mm. All experiments were conducted with the microelectrode and micromanipulator within the incubators. For the two anaerobic chambers (<0.5% and 8% oxygen), the height of the microelectrode in the medium was adjusted from outside the incubator through gloves attached to the front of the chambers. For the incubator that provided 21% oxygen, the height of the microelectrode was adjusted from inside the incubator after briefly opening the door. The dissolved O₂ concentration of medium was measured at heights noted in the text, with 0 mm being the pericellular location and 1.8 mm being at the top of the medium. A stable reading was obtained after continuous measurement for five minutes at each site.

Figure 1.

Apparatus for measurement of oxygen concentration. A. Oxygen levels were measured using a DO-166MT-1 micro dissolved oxygen electrode. The oxygen microelectrode was held by a micromanipulator and inserted into the culture medium through a hole in the cover of a 35 mm culture dish. Blue arrow: pH/millivolt meter. White arrow: Attenuator. Red arrow: Amplifier. B. The enlarged area outlined in panel A. Yellow arrow: the tip of the microelectrode.

Measurement of dissolved oxygen concentration in medium with cells

For measurement of pericellular oxygen concentrations, the microelectrode was placed at the bottom (cell surface) of the culture dish (n=4, from same placenta) as described above, and the oxygen concentration was recorded for ≤ 5 h beginning immediately after replenishing the culture with fresh medium equilibrated with 21% oxygen.

For measurement of the oxygen concentration at different heights above cultured trophoblasts (n=3, from the same placenta), primary trophoblasts were plated for 4 h, washed and the medium replenished. After an additional 5 h of culture, the microelectrode was then placed immediately above the cells on the bottom of the dish at 0 mm for the first measurement. The microelectrode was then raised in a stepwise manner, to the heights indicated in the Figure for each subsequent recording. At each height, the dissolved oxygen concentration was continuously measured for at least 15 min before adjusting the tip to another height. Two controls were performed to ensure that the observed oxygen gradient was not due to cellular debris blocking the electrode tip at the bottom of the plate and slowly detaching as the tip was raised. In the first control, after being raised, the microelectrode was then lowered in a stepwise manner and the oxygen concentrations again measured, showing a stepwise decrease (data not shown). In the second control, the cultured trophoblasts (n=3 from the same placenta) were plated, washed and replenished as described above. The cells were then treated with methanol at −20°C for 10 min, washed with PBS and replenished at 37°C with fresh medium equilibrated to the ambient air in the incubator that provides 21% oxygen. Five hours after the medium was replenished, the oxygen concentration at different heights of the medium was measured as described above.

Statistical Analysis

Data are presented as mean ± SD and comparisons were by ANOVA with Bonferroni corrections. A p < 0.05 was significant.

RESULTS

The microelectrode system set-up is shown in Figure 1. We first determined the time required for the electrode to obtain a stable reading for the oxygen tension using culture medium equilibrated in ambient atmospheres of 21%, 8%, and < 0.5% O₂. There was no significant difference in the time for equilibration among the three oxygen tensions (data not shown), with a stable reading consistently obtained by five minutes.

Medium oxygen levels in the absence of cells

We next compared the ambient oxygen concentration in the atmosphere of the incubator to the concentration of dissolved oxygen in medium without cells. Medium equilibrated with ambient atmospheres of 21%, 8%, and < 0.5% O₂ for 24 h at 37 °C yielded dissolved O₂ concentrations of ~19.3%, ~7.9% and ~0.5%, respectively (Figure 2).

Figure 2.

Concentration of dissolved oxygen (% O₂) in culture medium without cells. Medium without cells equilibrated in 21%, 8%, and <0.5% O₂. The concentrations of dissolved oxygen among the three media were significantly different (p<0.0001).

Pericellular oxygen concentrations in trophoblast cultures

We next measured the dissolved oxygen concentration in medium in four separate, near-confluent cultures of cells that were exposed to an atmosphere of 21% O₂ at 37°C. Immediately after introduction of fresh medium equilibrated in 21% O₂, the dissolved pericellular oxygen concentration (near the bottom of the dish) was initially ~20%. Between 11 and 75 min after introduction of the fresh medium, the pericellular oxygen level plummeted, eventually reaching a steady-state oxygen concentration between 0.1% and 0.6% (Figure 3 A).

Figure 3.

A. Time course of concentration of dissolved oxygen in medium with cells at the height of 0 mm of the culture dish in 21% O₂. The microelectrode was placed at height 0 mm. Cell cultures received fresh medium equilibrated with ambient air (21% O₂) at 37°C in the incubator. Recording of the oxygen concentration was begun immediately after replenishment of the medium.

B. Concentration of dissolved oxygen at different heights (0 mm, 0.25mm, 0.45 mm, 0.9 mm, 1.8 mm) in the culture medium with no cells, non-respiring cells, or live cells at 21% O₂. Recording of the oxygen concentration was begun 5 h after replenishing the culture medium at the indicated heights above the cells, starting at 0 mm.

Medium oxygen concentration vs. height above zero mm in the absence or presence of cells

Finally, we assayed the dependence of the dissolved oxygen concentration in the medium on the height from the bottom of the dish. The electrode was placed at the bottom of the dish (defined as 0 mm) and raised to the medium surface. As expected, in both medium without cells and in medium with dead cells the oxygen concentration was ~20% and was independent of height (Figure 3 B). By contrast, the dissolved oxygen concentration in dishes with trophoblasts was strongly dependent on the height above the cells, yielding 0.47% ± 0.2% at zero mm, 0.5% ± 0.3% at 0.25 mm, 4.3% ± 0.2% at 0.45 mm, 8.8% ± 0.1% at 0.9 mm and 12.7% ± 0.1% at 1.8 mm.

DISCUSSION

The data show that oxygen concentrations in medium without cells is within 2% of the oxygen level in the ambient atmosphere and is independent of the depth where the medium is sampled. By contrast, the oxygen concentration present in medium bathing trophoblasts attached to standard culture dishes is dependent upon the depth within the medium where sampling occurs and does not mimic the oxygen concentration within the overlying atmosphere. Moreover, the pericellular oxygen concentration is in a range that most would consider severe hypoxia, at ≤ 0.6% oxygen or about 4.6 mm Hg, in the medium in 21% oxygen atmosphere. We conclude that culture conditions are unable to replicate the pO₂ of 40–60 mm Hg commonly attributed to the maternal blood in the intervillous space in the second and third trimesters of pregnancy [7].

We found that the pericellular oxygen concentration at steady-state in our trophoblast cultures was low, ranging from 0.1% to 0.6%, and this finding is in agreement with studies of non-trophoblast, cell lines where dissolved oxygen levels were lowest adjacent to cells and higher near the medium surface [8–10]. However, we do not conclude that the cells cultured in 21% oxygen atmosphere are physiologically identical to cells cultured in hypoxia (despite<0.6% oxygen at the pericellular space) since we and others have detected clear differences between trophoblasts cultured under these two conditions [6, 11–14]. Instead, we surmise that measurements made in the immediate pericellular space reflect a balance of diffusion of oxygen through the culture medium from the overlying atmosphere and the rate of oxygen consumption by the cells, and, thus that the availability of more oxygen in the ambient atmosphere (e.g. in 21% vs <0.5% O₂) allows for different oxygen flux through the cells. This flux allows investigators to gain insights into the responses of human trophoblast to different oxygen tensions, even though the specific oxygen tensions studied may, or may not, exactly reflect in vivo conditions. Notably in vivo oxygen tensions are not commonly measured pericellularly, but instead are measured in variable locations within the lumens of vascular spaces. Such measurements likely reflect the overall level of dissolved oxygen, but may not reflect pericellular oxygen levels, suggesting ambient oxygen levels in the pericellular region of villous trophoblasts in vivo are likely to be lower than predicted from blood gas analyses.

Burton et al. [15] and Tuuli et al. [7] highlighted issues to consider about oxygen in culture conditions, including the absence of oxygen carriers, the data for in vivo oxygen concentrations, and the role of oxidative stress. Cells in vivo obtain oxygen from hemoglobin in red blood cells, and to compensate for the lack of red blood cells in culture medium, some investigators have used up to 95% oxygen in the ambient atmosphere. However, this approach predisposes to formation of excess damaging reactive oxygen species [7, 16–18]. In an attempt to increase oxygen availability in cell culture. investigators have developed oxygen carriers, such as hemoglobin, hydrocarbons, perfluorochemicals, and cyanobacterial gas vesicles. However, the instability, potential toxicity and side effects of these agents complicate their employment [19–23]. Despite this, it may be worthwhile to investigate their effects in trophoblast cell cultures.

We found that the oxygen concentration in medium without cells or with inviable cells at equilibrium approximates that in the overlying atmosphere. Glove-box incubators such as the ones used herein for non-air gas mixtures are generally expensive and thus not easily available to most investigators. Alternative models use closed culture systems, but after exposure to air the time for equilibration of medium to a newly introduced gas mixture can be hours, as shown by careful experiments from Lyall and colleagues [24]. Thus, each manipulation in air, however brief, requires re-equilibration.

Importantly, we observed that in the medium of near-confluent cultures of primary trophoblasts there is an oxygen concentration gradient. The lowest oxygen concentration is near the cells, with oxygen concentrations increasing to approximately that of the atmosphere at the air-medium interface. These data can be explained if trophoblasts use dissolved oxygen at a faster rate than oxygen diffuses into the medium from the ambient air, with the slope of the gradient depending on these two rates. One way to eliminate this oxygen gradient would be to mobilize the medium. However, this experimental design risks introduction of yet another independent variable, shear stress. Villous explants can be maintained at the gas-medium interface to minimize the diffusion distance for oxygen delivery in tissue culture [7]. However, this approach is not feasible for cell culture, as sufficient medium bathing the cells is required to provide nutrients, to buffer the pH of the medium, and to avoid artifacts created by inadvertent drying.

Our data suggest that we are unlikely to find any single in vitro culture condition that is an exact mimic for in vivo conditions, regardless of which ambient oxygen level is used. Thus, we cannot resolve the question of which oxygen paradigm is “best” in vitro. The varied oxygen conditions used among investigators [25–29] should thus be tolerated, but should be kept in mind when comparing and interpreting experimental results.