Main

In recent years, considerable interest has been generated in the role of cytokines in normal pregnancy(1). A range of cytokines can be detected in amniotic fluid during throughout gestation, including IL-1β, tumor necrosis factor-α, and IL-6(2). Furthermore, ample experimental evidence has demonstrated the expression of multiple cytokines by both placental(3) and decidual(4) tissues. It is believed that levels and timing of cytokine expression regulate placental maturation (and, therefore, the onset of labor), and there is clinical evidence to support the hypothesis that disruption in the normal course of cytokine expression at the maternal-fetal interface may provoke or be involved in the pathogenesis of premature labor(5, 6).

Special interest in this field has focused on placental trophoblasts(7). These cells, which form the barrier between mother and fetus, are clearly capable of expressing a broad range of cytokines and, indeed, in many respects resemble cells of the monocyte/macrophage lineage(8). Guilbert and Wegmann have argued that these cells in fact do function as monocytic cells, communicating with and modulating the maternal immune system during the course of pregnancy(8, 9, 10).

A common feature of cells of the monocyte/macrophage system is the ability to express cytokines such as IL-1β when stimulated via their FcγR(11). Whether trophoblasts or trophoblast-like cells share this characteristic with monocytes is unknown, but is of importance in understanding their biologic function. It is interesting to note, for example, that rises in amniotic fluid levels of both IL-1α and IL-1β, which occur during the late second and early third trimesters(12), parallel the transport of maternal IgG across the placenta. Another clue that stimulation of trophoblast Fc receptors may have important immunologic or biologic consequences is taken from SLE, a disease characterized by high levels of complement-activating circulating immune complexes(13). Women with this illness have a greater risk for poor fetal outcome with their pregnancies, which are characterized by a higher rate of miscarriage, stillbirth, and small for gestational age infants(14, 15). Furthermore, flares or worsening of disease activity are also known to complicate the course of pregnant SLE patients(16).

The experiments described here were undertaken to examine whether and how FcγR engagement might stimulate cytokine expression in a human choriocarcinoma cell line. The cytokine selected for study, IL-1β was chosen for its importance in the biology of the maternal-fetal interface and for its potential role in chronic inflammatory diseases such as SLE.

METHODS

Cells and cell cultures. Human choriocarcinoma JAR cells were a gift from A. Scott Goustin, Center for Molecular Medicine and Genetics, Wayne State University School of Medicine. Expression of FcγR has recently been demonstrated(17). Cells were maintained in RPMI with 10% heat-inactivated fetal bovine serum and 1.0% antibiotic/antimycotic solution (Sigma Chemical Co., St. Louis, MO).

Proteins and antibodies. Rabbit IgG fraction to BSA was obtained from Organon-Teknika (Durham, NC). A nonspecific rabbit IgG fraction and BSA (in fatty acid-free form) were obtained from Sigma Chemical Co. MAb to IL-1β and the 31-kD IL-1β precursor were obtained from Cistron (New Brunswick, NJ).

Immune complexes. Model BSA-anti-BSA immune complexes were formed at 2× antigen excess using rabbit IgG anti-BSA and BSA. Antibody and antigen were incubated at 37°C for 2 h, then subjected to centrifugation to remove insoluble material. Complexes were then further purified by affinity chromatography on protein A-agarose (Sigma Chemical Co.) and bound IgG eluted with 0.1 M glycine, pH 2.5. Complexes were then dialyzed against PBS with 0.01% sodium azide, and their concentration estimated by measuring OD at 280 nm.

Incubation of JAR cells with IgG and immune complexes. JAR cells were grown to 50% confluence in 25-cm2 flasks containing 10.0 mL of medium. All experiments were performed in the presence of 10.0 μg/mL polymyxin B, which inhibits lipopolysaccharide-induced IL-1β secretion. Cells were incubated with 25, 50, or 100 μg/mL of either BSA-anti-BSA immune complexes or monomeric rabbit IgG antibody. Concentrations of IgG/immune complexes were chosen based on published reports of immune complex levels of patients with rheumatic diseases such as SLE(18). To remove potentially confounding IgG aggregates, monomeric rabbit IgG (10 mg/mL) was diluted 1:10 in medium and subjected to centrifugation at room temperature at 14 000 × g for 30 min before addition to the JAR cells. Incubation times varied with the different experiments. To evaluate IL-1β protein expression, cells were incubated for 72 h, and medium (0.50 mL) was harvested at 24, 48, and 72 h. Harvested medium was aliquotted and stored at -70°C until used for cytokine ELISA assays. In Western blotting experiments and experiments to examine cytokine mRNA expression, cells were incubated over time frames ranging from 30 min to 18 h. These incubation times were derived empirically from our observations with immune complex-mediated cytokine expression in lymphocytes and monocytic cells lines (our unpublished data). JAR cells were solubilized in Trizol reagent (GIBCO, Gaithersburg, MD) and then either subjected immediately to total RNA extraction and mRNA analysis (see below) or stored at -70°C until used.

Requirements for new protein biosynthesis. To assess the requirements for new protein biosynthesis for cytokine expression, in certain experiments, JAR cells were incubated with maximum concentrations of immune complexes or monomeric IgG (100 μg/mL) in the presence of 2-10 μg/mL cycloheximide. Medium was collected after 24 h, and IL-1β concentrations were measured by ELISA.

ELISA for IL-1 β. Quantification of IL-1β under the different experimental conditions was undertaken using commercially available ELISA assays (T Cell Diagnostics, Cambridge, MA). The assays used were “sandwich” ELISAs using murine MAb to the cytokine of interest as the capture antibody and horse radish peroxidase-conjugated MAb as the detection antibody. Color development was achieved with 3,3′,5,5′-tetramethylbenzidine, and the colorimetric reaction was stopped with 0.18 M sulfuric acid. OD was read at 450 nm, and results were compared with samples of known concentration provided with the assay kit. All experimental samples were run in duplicate. Negative controls (blanks) consisted of duplicate wells incubated with medium alone and duplicate wells incubated with the sample diluent provided with the assay kits.

Each experiment was performed three times. Statistical analysis of the differences between cytokine levels produced by immune complexes versus monomeric IgG was performed by 2-tailed independent t tests using commercially available statistical analysis software(SPSS, Prentice Hall, Inc., Des Moines, IA).

Western blotting of cell-associated IL-1 β. Immune complexes or monoclonal rabbit IgG were incubated with JAR cells as described above. Cells were harvested at the time intervals of interest, washed three times with PBS, and lysed in 200 μL of solubilizing buffer consisting of 1.0% Nonidet P-40, 10 mM EDTA, 25 mM iodoacetamide, and 1.0 mM phenylmethylsulfonly fluoride. Cell lysates were subjected to centrifugation to remove insoluble material, and 20 μL of the cell lysate were subjected to SDS-PAGE in 10% minigels under reducing conditions. Proteins were then transferred electrophoretically to nitrocellulose membranes. Membranes were blocked with PBS, 2.0% powdered milk (“blocking buffer”) at room temperature for 1 h, then at 4°C overnight. Membranes were then washed with blocking buffer and incubated with primary antibody diluted 1:200 in blocking buffer. After 3 h of incubation, membranes were washed and incubated with horseradish peroxidase-conjugated goat anti-mouse IgG antibody diluted 1:200 in blocking buffer. Membranes were incubated with secondary antibody for 3 h at room temperature and washed three times. Visualization of the specific proteins was undertaken using the Renaissance chemiluminescent substrate(Du-Pont, Wilmington, DE). Membranes were exposed to x-ray film, and the resulting luminograms were developed after 20-60 s.

Reverse transcriptase-PCR IL-1 β mRNA. Cells exposed to immune complexes or monomeric IgG for 30, 60, or 120 min were dissolved in Trizol reagent, and total RNA extraction was performed exactly according to the manufacturer's recommendations. Total RNA was then immediately transcribed to cDNA using murine Maloney leukemia virus reverse transcriptase and oligo(dT) primers (Clontech, Palo Alto, CA) under the following conditions: 50 mM KCl, 10 mM Tris-HCl (pH 8.3), 5.0 mM MgCl2, 1.0 mM each dNTPs, 2.5 μM oligo(dT), 2.5 U/μl reverse transcriptase, and 1.0 U/μl RNase inhibitor (Promega, Madison, WI) with 2.0μL of target RNA in a total volume of 20 μL. Reactions were performed in a single three-step cycle: 42°C for 15 min, 99°C for 5 min, and 5°C for 5 min. DNA generated by this process was either stored at-20°C or used immediately for PCR.

Oligonucleotide primers for β-actin and IL-1β were obtained from Clontech. Sequences for the upstream and downstream primers are shown in Table 1. The IL-1β, but not β-actin, primers amplify cDNA segments that overlap introns, so that amplification of genomic DNA can be readily identified by the size of the fragment generated by the PCR reaction. Amplification reactions were performed using Thermus aquaticus DNA polymerase (2.0 U/reaction) in a volume of 50 μL under the following conditions: 50 mM KCl, 10 mM Tris-HCl (pH 8.3), 1.5 mM MgCl2, 0.2 mM dNTPs, 0.4 μM each 5′ and 3′ primers, with 2.0 μL of target cDNA. Cycling was performed at 94°C for 45 s for denaturation, 60°C for 45 s for annealing, and 72°C for 2 min for elongation for 35 cycles, followed by a final extension step (72°C for 7 min). Samples were solubilized in 5× sample buffer, consisting of 0.25% bromphenol blue, 0.25% xylene cyanol, and 30% glycerol, and subjected to electrophoresis in 1.5% agarose/ethidium bromide gels. Gels were examined for bands of the appropriate molecular weight and compared with positive controls generated from cDNA fragments provided with the primer sets encompassing the sequences amplified in the PCR reactions.

Table 1 Cytokine primer sequences
Full size table

RESULTS

Incubation of model immune complexes with JAR cells induced IL-1β at the protein level. The highest levels of IL-1β were achieved at the highest dose of immune complexes used, 100 μg/mL. Figure 1,A and B, shows the dose-response curve for IL-1β using both BSA-anti-BSA immune complexes and monomeric rabbit IgG. After 72 h, levels of IL-1β were produced by model immune complexes that were roughly 1.5-2 times higher than those produced by monomeric IgG. That is, at 25 μg/mL, IL-1β levels were (mean ± SD) 75 ± 13 pg/mL for monomeric IgG versus 140 ± 75 pg/mL for immune complexes. At 50μg/mL, these values were 147 ± 46 pg/mL versus 244± 57 pg/mL, and at 100 μg/mL these values were 292 ± 140 pg/mL versus 428 ± 275 pg/mL. The differences at 100 μg/mL were statistically significant (p = 0.007). Kinetic studies showed that approximately 60% of the IL-1β synthesized could be detected within 24 h in both IgG and immune complex-stimulated cells. Substantial increases in IL-1β production were not seen after 48 h (data not shown).

Figure 1
figure 1

Bar graphs show comparison of IL-1β production in JAR cells incubated for 72 h with monomeric IgG and model BSA-BSA-anti-BSA immune complexes (IC) at three different concentrations: 25 μg/mL(A), 50 μg/mL (B), and 100 μg/mL (C). Highest levels of IL-1β were seen with 100 μg/mL IgG or immune complexes. Differences between IL-1β levels achieved with 100 μg/mL immune complexes vs the same concentration of monomeric rabbit IgG were statistically significant (p = 0.007). Figures represent results from three different experiments.

Full size image

Cycloheximide did not substantially inhibit IL-1β secretion in these experiments, suggesting that most of the IL-1β secretion by JAR was not dependent upon new protein synthesis. Figure 2 shows the low levels of IL-1β mRNA that could be detected within 30 min of stimulation of JAR cells by immune complexes.

Figure 2
figure 2

Ethidium bromide-stained 1.5% agarose gel showing reverse transcriptase-PCR analysis for IL-1β mRNA in JAR cells stimulated with 100 μg/mL immune complexes. Time points (in minutes) are shown above each lane. Cells, JAR cells cultured in the absence of immune complexes or IgG.

Full size image

Western blotting further supported the data obtained by cycloheximide experiments and reverse transcriptase-PCR. Western blots using both pro-IL-1β and mature IL-1β antibodies revealed a 31-kD protein that was constitutively expressed by JAR cells. Initial decreases in this protein were seen within 2 h of stimulation of the cells with either immune complexes or monomeric IgG as shown in Figure 3. The Mr 17 000 mature IL-1β protein was not detected by Western blot.

Figure 3
figure 3

Western blot using pro-IL-1β MAb. Time frames (in hours) are shown above the blot. Cells stimulated with monomeric IgG or model BSA-anti-BSA immune complexes (IC) are indicated below the blot. A protein of approximately 31 kD is visualized in JAR cell lysates. Levels of this protein decline after 2 h and return to or near baseline within 18 h. Similar results were obtained using antibody to the mature form of Il-1β.

Full size image

DISCUSSION

The studies reported here provide further evidence to support the hypothesis that syncytiotrophoblasts (or, in this case, syncytiotrophoblast-like choriocarcinoma cells) share common features with cells of the monocyte/macrophage lineage. We have shown that these cells, like monocytes/macrophages(11), secrete the proinflammatory cytokine IL-1β when stimulated through their FcγRs. There are several interesting implications to these findings, both with respect to our understanding of normal pregnancy and with respect to certain diseases in pregnant women.

That cytokines are involved in numerous aspects of pregnancy, from implantation(19, 20) to the onset of labor(1), is well established. Experimental evidence supports the hypothesis that both maternal and fetal tissues(21) secrete a broad range of cytokines and other immunoregulatory substances(22), and are themselves targets of cytokine regulation. The production of IL-1β, which was associated with the binding of monmeric IgG to JAR cells, may therefore have physiologic relevance to normal pregnancy and the onset of labor. Concentrations of IL-1β are known to increase in amniotic fluid during the course of the third trimester, paralleling the increases in IgG transport across the placenta to the fetus. Our data suggest that the process of binding of IgG to placental FcγRs may result in a state of “activation” by the trophoblasts, resulting in the secretion of IL-1β. Both the levels and relative concentration of this cytokine, in turn, appear to play critical roles in placental function and maturation, partly through their interactions with local endocrine mediators(23, 24). Thus, in addition to providing the fetus with passive immunity, the transport of IgG across the placenta may also initiate signals, which are crucial to the process of placental growth and maturation and the onset of labor.

One interesting aspect of our data concerns the mechanism of IL-1β secretion in JAR cells in response to IgG or immune complexes. In monocytes, IL-1β secretion involves a complex series of events(25), all of which may be independently regulated(26). In many experimental systems, IL-1β secretion requires neither mRNA transcription nor new protein biosynthesis, as this cytokine is present in monocytes and monocyte-like cells as a 31-kD precursor(27), which is cleaved by IL-1β converting enzyme(28). Our data are consistent with the hypothesis that JAR cells, like other monocyte-like cells, contain pro-ILβ, which can be cleaved and released with the appropriate stimulus. We did not, however, see release of the mature, active IL-1β molecule. Whether normal trophoblasts exhibit this defect has not, to our knowledge, been studied.

In conclusion, we have shown that human choriocarcinoma JAR cells are capable of secreting IL-1β when stimulated through their FcγRs. These data therefore support the hypothesis that syncytiotrophoblasts share many functions and properties with cells of the monocyte/macrophage lineage and are active, rather than passive, participants in the complex immunologic regulation at the maternal-fetal interface.