Uric acid attenuates trophoblast invasion and integration into endothelial cell monolayers

Cell culture materials.

Culture media, media additives, and fetal bovine serum were purchased from Lonza (Walkersville, MD). Cell tracker fluorescent dyes (CMFDA, green; CMTPX, red) were purchased from Molecular Probes, Invitrogen (Eugene, OR). All other chemical reagents and materials, unless otherwise stated, were purchased from Sigma-Aldrich (St. Louis, MO).

Cell lines.

Human uterine myometrial microvascular endothelial cells (UtMVEC, passage 3) were purchased from Lonza and maintained in endothelial basal medium-2 (Lonza) supplemented with supplier-recommended concentrations of human recombinant epidermal growth factor, fibroblast growth factor, vascular endothelial growth factor, ascorbic acid (vitamin C), hydrocortisone, recombinant insulin-like growth factor, gentamicin, penicillin-streptomycin (100 IU/ml), and 5% fetal bovine serum (endothelial cell growth media, EGM). These cells were used between passages 4 and 8.

The human HTR-8/SVneo trophoblast cell line (HTR8) was a gift from Dr. Charles Graham (Queen's University, Kingston, ON, Canada). This cell line was established from explant cultures of first trimester (weeks 8–10) chorionic villi. The cells were immortalized through transfection with cDNA encoding the simian virus-40 large T antigen (14). HTR-8/SVneo cells were grown in RPMI 1640 medium (GIBCO-BRL, Grand Island, NY) supplemented with 5% fetal bovine serum and 100 IU/ml penicillin-streptomycin (trophoblast cell growth media, TGM). The cells were used at passages 70–85.

Effects of uric acid on trophoblast invasion.

We examined the effects of uric acid on trophoblast invasion in a matrigel-based invasion assay (36). Trophoblast cells were seeded (50,000 cells/well) in the top of transwell inserts (6.5-mm diameter polycarbonate membrane, 8-μm pore; Corning Costar, Lowell, MA) that had been precoated with 100 μl Matrigel (1 mg/ml; BD Biosciences, San Jose, CA) and housed in a 24-well plate. The cells were allowed to invade through the reconstituted extracellular matrix for 24 h in the presence or absence of increasing concentrations of uric acid (0, 3, 5, and 7 mg/dl) dissolved in TGM (n = 6). Trophoblast cells located on the undersurface of the transwell membrane were fixed with ice-cold methanol and stained with hematoxylin, and brightfield images were obtained with an upright Zeiss Axioskop 40 microscope. All invaded cells per membrane were counted using Image J freeware (National Institutes of Health). To correct for interassay variation, cell counts were expressed as fold change in invasion relative to untreated controls.

Trophoblast integration into endothelial cell monolayers.

To model the ability of trophoblast to integrate into and replace the uterine vascular endothelium, we employed an in vitro trophoblast-endothelial cell monolayer coculture system as previously described with slight modifications (8, 12). Endothelial cells were seeded into gelatin-coated wells of a 6-well plate (200,000 cells/well), grown to confluence in EGM, and labeled with green fluorescent cell tracker CMFDA (10 μM) for 30 min, as per manufacturer's instructions. The endothelial cell monolayers were pretreated with experimental agents described below for 2 h after which HTR8 trophoblast cells in suspension, initially labeled with red fluorescent cell tracker CMTPX (10 μM, 30 min), were seeded (400,000 cells/well) onto the endothelial cell monolayers. The combined cells were coincubated in a 1:1 mixture of EGM and TGM for 24 h (37°C, 21% O₂-5% CO₂) in the continued presence of the experimental agents described below. A mixed monolayer of trophoblast and endothelial cells was thus generated over the 24-h coincubation, with varying degrees of trophoblast integration. The resulting mixed monolayer was then washed twice with PBS and fixed with 4% paraformaldehyde for 1 h at room temperature, and mounting media was added to each well. We examined each well using a Zeiss Axiovert 40 CFL inverted fluorescent microscope, capturing 4 fields/well at 5× magnification using Zeiss Axiovision software. A grid system was used to ensure that the four captured fields were in similar locations for all wells imaged. Trophoblast integration into endothelial cell monolayers was quantified as a percentage of total field area occupied by trophoblast cell islands (red label) using locally developed software. The effect of treatment was expressed as a fold change of trophoblast integration area relative to untreated coculture controls from the same experiment.

Effects of uric acid on trophoblast integration into endothelial cell monolayers.

We examined the effects of uric acid on the degree of trophoblast cells integration into uterine microvascular endothelial cell monolayers. UtMVEC monolayers were pretreated with uric acid (0, 0.5, 3, 5, or 7 mg/dl) for 2 h and then coincubated with trophoblast cells for 24 h in the presence of the same concentrations of uric acid (n = 6). The uric acid concentration range was based upon concentrations in the circulation of women at 15 wk gestation who go on to have healthy pregnancies [2.7–3.1 mg/dl (33)] and preeclamptic pregnancies [3.5–4.4 mg/dl (33)] and concentrations typical of women with clinically evident preeclampsia [>5.5 mg/dl at term, with an upper limit ∼10 mg/dl (33)].

To test whether cellular uptake is necessary for the effects of uric acid, we also treated the cocultures with the highest concentration of uric acid (7 mg/dl, as above) in the presence or absence of probenecid (100 μM), an organic anion transporter inhibitor (n = 6). Matched coculture wells were treated with probenecid alone.

Effects of antioxidants on trophoblast integration into endothelial cell monolayers.

To examine the role of redox signaling in uric acid-induced alterations to trophoblast integration into endothelial cell monolayers, endothelial monolayers were pretreated for 2 h with or without dehydroascorbic acid (DHA, 75 μM) or a combination of catalase (17 mU/ml)-superoxide dismutase (SOD, 33 mU/ml), or the intracellular antioxidant N-acetyl cysteine (NAC, 75 μM, Sigma-Aldrich), or the NADPH oxidase inhibitor apocynin (100 μM, Sigma-Aldrich) in the presence (DHA + uric acid, n = 5; catalase/SOD + uric acid, n = 5; NAC + uric acid, n = 5; apocynin + uric acid, n = 5) and absence (DHA, n = 5; catalase-SOD, n = 5; NAC, n = 5; apocynin, n = 10) of uric acid (7 mg/dl). DHA, intracellularly reduced to ascorbic acid, was used to limit redox cycling of ascorbic acid with adventitious transition metals in culture media. Trophoblast cells (400,000/well) were then added and coincubation was maintained for 24 h in 1:1 EGM/TGM containing the same concentrations of test reagents.

Trophoblast integration into endothelial cell monolayers: role of apoptosis.

We examined the effect of uric acid on apoptosis in the coculture model. Total apoptotic cell death was assayed in the presence or absence of uric acid (7 mg/dl; n = 5) or the antioxidants DHA (n = 4) or catalase-SOD (n = 4). After 24 h, the mixed monolayer and overlying media were combined and spun down (200 g for 10 min at room temperature), and the resulting cell pellet was isolated and fragmented DNA-histone complexes measured by ELISA (Cell Death Detection ELISA, Roche Applied Science, Indianapolis, IN). This same protocol was performed on endothelial cells alone or trophoblast cells alone in monolayer in the absence and presence of uric acid (24 h incubation, 7 mg/dl).

To determine the proportion of each cell type undergoing apoptosis, we performed fluorescent transferase-mediated dUTP nick-end labeling (TUNEL) staining (In Situ Cell Death Detection Kit, Roche Applied Science) on mixed monolayers grown in the absence or presence of uric acid (24 h incubation, 7 mg/dl, n = 4). As the TUNEL probe fluoresces red, the coculture protocol was modified to label only the endothelial cells (CMFDA, green). We used filters to selectively view the rhodamine and fluorescein spectrums, capturing 10 fields/well at 20× magnification using Axiovision software. A grid system was employed to ensure that the 10 captured fields were in similar locations for all mixed monolayers imaged. Analysis of immunofluoresent TUNEL staining included cell counts of all apoptotic cells, proportion of endothelial cells versus trophoblast cells undergoing apoptosis and percentage of all endothelial or trophoblast cells undergoing apoptosis. This analysis was conducted using Image J freeware, with the operator blinded as to treatment.

Effects of uric acid on trophoblast and endothelial cell migration, cell number, and cell viability.

We tested the effects of uric acid on cell migration, cell number, and viability for each cell type individually. Cell migration was assessed using a “scratch wound healing” assay (26). Confluent monolayers of either cell type were scratched with a sterile P200 pipette tip. The ability of the cells to migrate into the wound in the presence or absence of uric acid (7 mg/dl; n = 6 for trophoblast cells, n = 6 for endothelial cells) in growth media was observed over a 6-h time period and analyzed using Image J software. For determination of changes in cell number each cell type was grown to 70% confluence in 6-well plates. After 12 h in serum-free media, cells were treated with growth media alone or containing uric acid (7 mg/dl) for 24 h. Cells were counted at 0 and 24 h (n = 5 for trophoblast cells, n = 5 for endothelial cells). Lactate dehydrogenase (In vitro toxicology LDH assay kit, Sigma-Aldrich) was measured in the media of cells treated for 24 h with either growth media alone or with uric acid (7 mg/dl; n = 6 for trophoblast cells, n = 6 for endothelial cells).

Effects of preeclampsia and normal pregnancy serum on trophoblast integration into endothelial cell monolayers.

Pregnant subjects were recruited at the time of delivery as part of an ongoing study of preeclampsia pathogenesis as approved by the University of Pittsburgh/Magee-Womens Hospital Institutional Review Board. All subjects provided written informed consent. Eight women had preeclampsia defined as hypertension and proteinuria arising de novo after the 20th week of gestation, and hyperuricemia, with reversal of hypertension and proteinuria after delivery (9). Gestational hypertension was defined as systolic blood pressure >140 mmHg or diastolic blood pressure >90 mmHg arising after 20 wk gestation in a previously normotensive woman. Proteinuria was defined as >300 mg of protein in a 24-h urine collection, or 1+ or greater protein on dipstick of a catheterized urine sample, or 2+ or greater urine protein on dipstick of a voided urine sample, or a random urine protein-to-creatinine ratio >0.3. Hyperuricemia was defined as >1 standard deviation above normal for the index gestational age [at term >5.5 mg/dl (3.3 mM)]. Controls (n = 8) had uncomplicated, normotensive pregnancies and were delivered at term with healthy babies of appropriate weight for gestational age. Patients with chronic hypertension, diabetes, renal disease, or other significant preexisting metabolic disorders, or with a recent history of cigarette smoking or illicit drug use, or with multi-fetal gestation were excluded. Clinical characteristics of cases and controls are listed in Table 1.

Aliquots of each patients' serum were stored at −70°C without thaw until used. Two pools of serum were created from equal volumes from either the eight preeclampsia or the eight uncomplicated pregnancy patients. Some of the aliquots of each pool were pretreated with uricase (17 mU/ml, 30 min at 37°C) to remove endogenous uric acid, the removal confirmed by colorimetric uric acid assay (Pointe Scientific; Canton, MI).

To compare the effects of normal pregnancy versus preeclampsia serum (native or uric acid deficient) on trophoblast cell integration into preformed endothelial monolayers, the endothelial cell monolayers were incubated in EGM with 5% vol/vol pooled serum for 2 h, and this exposure was continued for 24 h during the formation of trophoblast/endothelial mixed monolayers (5% vol/vol pooled serum in 1:1 EGM-TGM; n = 5).

Statistical analysis.

Patient demographic data are represented as means ± SD. We reviewed the experimental data by plotting histograms and found a skewed distribution. Appropriate nonparametric tests were used due to the nonnormality coupled with the small sample sizes. All experimental results are presented as median with interquartile range. Concentration-dependent effects of uric acid on trophoblast invasion and trophoblast integration were analyzed using the nonparametric Kruskal-Wallis one-way analysis of variance with Dunn's post hoc analysis and Cuzick's test of trend. The asymptotic relative efficiency of the Kruskal-Wallis Test to the usual parametric F test is never less than 0.864 if, as we found, the distribution functions have identical shapes and differ only in the location parameter (15). Effects of individual experimental treatments (uric acid, probenecid, antioxidants) were compared with untreated controls using a Wilcoxon signed-rank test. Effects of individual experimental treatments in combination with uric acid (probenecid + uric acid, antioxidants + uric acid) were compared with uric acid alone using the Mann-Whitney U test. Differences were considered statistically significant if P < 0.05. With the use of a 5% level of significance and the given sample sizes, the power of the test of hypotheses by post hoc analysis was greater than 0.92 in most instances. Exceptions included the comparisons of control versus catalase-SOD + uric acid (0.87) and control versus catalase-SOD (0.60) in testing the effects of antioxidants on trophoblast integration into endothelial monolayers.

Effects of uric acid on trophoblast invasion.

Uric acid attenuated trophoblast invasion through a reconstituted extracellular matrix in a concentration-dependent fashion with reductions of 51.1% (38.9–65.3%) and 72.9% (66.8–74.5%) when compared with untreated controls with 5 and 7 mg/dl uric acid, respectively (Fig. 1, P < 0.05).

Effects of uric acid on trophoblast integration into endothelial cell monolayers.

Uric acid reduced the ability of trophoblast cells to displace endothelial cells from monolayers in a concentration-dependent fashion. In untreated control wells trophoblast cell islands comprised more than 60% of the monolayer following 24 h of coincubation (Fig. 2, A an B). Pretreatment of the endothelial cell monolayer and concurrent treatment during the 24-h incubation with uric acid attenuated trophoblast integration into the monolayer (Fig. 2, C and D). This inhibitory effect of uric acid on trophoblast integration into microvascular endothelial cell monolayers increased with increasing uric acid concentrations (P < 0.05; Fig. 2E).

Probenecid, an inhibitor of cellular uric acid uptake, abrogated the inhibitory effects of uric acid on the ability of trophoblast cells to integrate into the endothelial cell monolayers (Fig. 3). Probenecid alone had no effect on trophoblast integration.

Effects of antioxidants on trophoblast integration into endothelial cell monolayers.

The potential role of redox signaling in uric acid-induced effects was examined by testing the effect of antioxidants (DHA, catalase-SOD, NAC, apocynin) alone and in combination with uric acid (7 mg/dl). Each antioxidant attenuated trophoblast integration into uterine microvascular endothelial cell monolayers compared with untreated controls (DHA, catalase-SOD, NAC: P < 0.05; apocynin: P < 0.01) (Fig. 4A; supplemental data, Figs. S1 and S2), in a fashion similar to uric acid treatment alone (P = 0.49 for DHA vs. uric acid; P = 0.89 for catalase-SOD vs. uric acid; P = 0.92 for NAC vs. uric acid; P = 0.71 for apocynin vs. uric acid). Treatment with the combination of antioxidants and uric acid had no further attenuating effects (P = 0.55 for DHA + uric acid vs. uric acid alone; P = 0.42 for catalase-SOD + uric acid vs. uric acid alone; P = 0.35 for NAC + uric acid vs. uric acid alone; P = 0.18 for apocynin + uric acid vs. uric acid alone; Fig. 4B; supplemental data, Figs. S1 and S2).

Trophoblast integration into endothelial cell monolayers: role of apoptosis.

Total apoptotic cell death in the mixed trophoblast-endothelial monolayers was lower in the presence of uric acid (P < 0.05; Fig. 5A). The antioxidants DHA or catalase-SOD similarly reduced apoptosis (P < 0.05; Fig. 5A). Interestingly, uric acid did not reduce apoptotic cell death in monloayers of only trophoblast or only endothelial cells in culture (trophoblast cells, P = 0.31; Fig. 5B) (endothelial cells, P = 0.44; Fig. 5C).

Immunofluorescent TUNEL staining indicated that 6.6% (6.5–7.6%) (Fig. 6E) of all cells in untreated mixed monolayers were undergoing apoptosis. Endothelial cells made up 80.2% (79.6–85.7%) of the apoptotic cell population (Fig. 6E). The majority of endothelial cells undergoing apoptosis were along the leading edge of trophoblast cell islands (green fluorescent endothelial cells situated above unlabeled trophoblast cells in representative Fig. 6, A–D). Uric acid treatment reduced the overall number of TUNEL-positive cells to 3.6% (3.4–4.5%, P = 0.02) (Fig. 6E). The reduction of cells undergoing apoptosis was due almost entirely to decreases in endothelial cells undergoing apoptosis: 13.2% (12.1–15.6%) in control vs. 5.8% (4.8–9.3%) with uric acid treatment, P = 0.04 (Fig. 6F).

Effects of uric acid on trophoblast and endothelial cell number and migration.

To further address the mechanism by which uric acid might be altering integration, we tested the effects of uric acid on cellular behavior of either the trophoblast cells or endothelial cells alone in culture. The highest concentration of uric acid (7 mg/dl) caused a slight increase in trophoblast cell number after 24 h incubation (P < 0.05; Fig. 7A), with no measurable differences in cell viability as measured by LDH release (P = 0.44; Fig. 7C). The ability of trophoblast cells to migrate in the scratch wound assay system was unchanged by uric acid treatment (P = 0.70; Fig. 7E). Uric acid treatment had no significant effect on endothelial cell number following a 24-h incubation (P = 0.13; Fig. 7B), although there appeared to be a trend toward lower cell numbers. Twenty-four hour treatment with uric acid did not affect endothelial cell viability (P = 0.88; Fig. 7D) or migratory properties (P = 1.00; Fig. 7F).

Effects of preeclampsia and normal pregnancy serum on trophoblast integration into endothelial cell monolayers.

To attempt to test the relevance of the uric acid effect, we compared the ability of pooled serum from preeclamptic women with hyperuricemia and pooled serum from normal pregnant controls to reduce the integration of trophoblast cells into endothelial cell monolayers. Patient demographic data are presented in Table 1. The uric acid concentration was 9.3 mg/dl in the preeclampsia pool and 4.7 mg/dl in the control pool.

Trophoblast integration was attenuated by 5% vol/vol preeclampsia serum more than by the same dilution of serum from healthy pregnant controls: 44.3% (28.5–47.3%) vs. 62.2% (55.3–68.6%) trophoblast integration, respectively (P < 0.05; Fig. 8). Prior removal of endogenous uric acid with uricase did not change the effect of serum from controls but eliminated the effect of the preeclampsia serum pool such that integration (61.3%; 46.7–65.8%) was similar to healthy pregnant controls (P > 0.05; Fig. 8).