For the HTR8 human plasma co-culture experiment, written informed consent was obtained from each participant, and the study was approved by the institutional review boards at Rhode Island Hospital and the Philippines Research Institute of Tropical Medicine.

We used the immortalized first trimester cell line HTR8/SVneo, originally obtained from a human pregnancy terminated in the first trimester, and displaying properties of invasive extravillous cytotrophoblast cells [18]. Cells were maintained at 37°C with 5% CO₂ in 1∶1 DMEM/F-12 media (Invitrogen, Grand Island, NY) supplemented with 1% L-glutamine (Invitrogen), 1% penicillin/streptomycin (Invitrogen) and 5% fetal bovine serum (Atlanta Biologicals, Lawrenceville, GA). All experiments were performed with the addition of SEA (25 µg/ml) in complete media for 24 h or media only control, unless otherwise noted. This dose was chosen based on dose response curves performed previously in our laboratory using purified primary trophoblast cells [19].

Schistosoma japonicum SEA was generously donated by Dr. Chuan-Xin Yu (Jiangsu Provincial Institute of Schistosomiasis Control, Wuxi, Jiangsu, China) after having been prepared according to standard procedure [20]. The SEA was prepared under endotoxin free conditions, with all reagents and equipment used to isolate the SEA from collected livers being LPS free prior to use. Preparations were evaluated for contaminating endotoxin using an FDA-cleared LAL-assay (Lonza Group, Basel, Switzerland). Endotoxin levels for all SEA preparations used were <6 EU/mg protein, which, in our culture conditions, is at least 1000-fold lower than levels that have been previously shown to influence human trophoblast cells [21].

Following the treatment period, media from HTR8/SVneo cells was collected and levels of multiple cytokines, chemokines, and fibrotic markers were assessed on a bead-based platform (BioPlex, Bio-Rad, Hercules, CA) using a sandwich antibody-based assay as previously described [22]. Cytokines evaluated included interleukin (IL)-1β, IL-6, interferon (IFN)-γ, tumor necrosis factor (TNF)-α, IL-4, IL-5, IL-10, IL-13, IL-12, IL-8 and IL-2. These specific cytokines were measured, as they were all components of a multiplex analysis developed and validated in our laboratory. The majority of these cytokines have been reported to be expressed by the placenta, although expression is highly variable depending on culture conditions, gestational age and disease status [23]–[26]. In addition, we measured levels of tissue inhibitor of metalloproteinases (TIMP)-2, TIMP-4, insulin-like growth factor binding protein (IGFBP)-5, matrix metalloproteinase (MMP)-9, tenascin C, syndecan 1, Fas ligand, osteopontin, TIMP-1, connective tissue growth factor (CTGF), macrophage inflammatory protein (MIP)-1α, MMP-1, IGF-1, MMP-8, monocyte chemotactic protein (MCP)-1, and TIMP-3. These analytes were developed into multiplexed assays due to their importance in schistosomiasis-associated and idiopathic pulmonary fibrosis [27]–[29]. However, they are also widely implicated in invasion and remodeling at the maternal-fetal interface, underscoring the similarities between these two processes.

For the human plasma assays, we utilized plasma collected from pregnant women at 32 weeks gestation residing in Leyte, the Philippines, an area endemic for S. japonicum. The study population and sample collection has been described elsewhere [8], with socioeconomic status (SES), gravida, parity, body mass index (BMI), smoking status and age determined via questionnaire [30]. Schistosomiasis and co-infections (Ascaris lumbricoides, Trichuris trichuria, and hookworm) were determined from stool samples using the Kato Katz method. Infection intensities for each were determined using the WHO guidelines [31]. From this larger cohort of 150 women, we selected 29 women infected with schistosomiasis (20 lightly infected [1–99 eggs per gram], 9 moderate infection [100–399 epg]), and 29 uninfected women matched to the infected women for SES, co-infections, gravida, parity, gestational age, BMI, smoking status and maternal age (Table 1). Schistosomiasis was evaluated as a nominal (yes/no) variable due to the low numbers of women with moderate-high infection intensities. Serum was collected at the only pre-natal study visit (32 wks gestation).

HTR8/SVneo cells were cultured to 80% confluency in complete media before being cultured for 48 hours in serum free media with the addition of 10% plasma from the aforementioned pregnancies. Following 48 h incubation, trophoblast culture supernatants were collected and analyzed for cytokine production as described above.

HTR8/SVneo cells were cultured to 100% confluence in complete media. Once completely confluent, a scratch was made across the well using a sterile pipette tip. The underside of each of the wells was cross-hatched for reference. Cells were briefly washed with PBS in order to remove all detached cells after scratching, and cultured in serum-free media with the addition of SEA (25 µg/ml) for 48 h. Phase contrast images of the denuded region were taken at 0 h, 24 h and 48 h after scratch formation using an Olympus IX70 inverted tissue culture microscope (Olympus Corp., Tokyo, Japan). The same region of each well was imaged at each time point, using the cross-hatching as reference. The area free of cells was quantified using ImageJ software (NIH, Betheseda, MD). The denuded area at 24 h and 48 h for a specific well was expressed as a percentage of the denuded area that had been present in that well at 0 h, thus controlling for well-to-well variation in original scratch sizes.

MTT assays were done on HTR8/SVneo cells in parallel to the migration assays. MTT (Sigma Aldrich, St. Louis, MO) was added to each well and the cells were incubated for 4 h at 37°C in a humidified environment. Media and MTT were aspirated from each well, MTT solvent (4 mM HCl, 0.1% Nonidet P-40, in isopropanol) added, and the plate incubated at 25°C in the dark with rotation for 15 minutes. Absorbance for each well was read at 560 nm and 630 nm.

HTR8/SVneo cells at 80% confluency were treated with SEA (25 µg/ml) for 24 h in complete media before being gently trypsinized, washed with complete media and resuspended in serum-free media. 25,000 cells/well were plated on matrigel-coated transwell inserts with an 8 µm pore size (Corning, Tewksbury, MA). The bottom chamber contained complete HTR8 media. Following 48 h incubation, cells and matrigel were gently removed from the top of the transwell, and those cells that had invaded through the matrigel, traversed the pores, and reached the bottom of the transwell were stained with hematoxylin (Sigma Chemical). Stained cells were visualized and counted using an Olympus BH-2 microscope (Olympus Corp.).

Data analysis was performed using JMP v.10 (SAS Institute, Cary, NC). All data were evaluated for normality using the Shapiro-Wilk test. Those experiments for which all data were normally distributed were further evaluated with ANOVA and t-tests, with means ± SEM reported. For data that was not normally distributed, Wilcoxon Signed Rank analyses were performed, with data reported as median ± IQR. Specifically, cytokine production by HTR8 cells was compared between cells exposed to uninfected plasma and those exposed to infected plasma (Figure 1) as well as cells cultured with media alone and media with SEA (Figures 2 and 3). Similarly, HTR8 migration and invasion were compared between cells cultured with media alone and those with SEA addition to the media (Figures 4 and 5). Statistical significance was considered as P<0.05.

For these experiments, we selected a sub-set of plasma samples collected from women at 32 weeks gestation as part of a previous study [8], matching samples from 29 infected women with 29 samples from uninfected women on key potential confounding covariates (Table 1). HTR8/SVneo cells were cultured in serum-free media supplemented with 10% plasma collected at 32 weeks gestation, and allowed to remain in culture for 48 h. Media from HTR8/SVneo cells cultured with plasma from schistosome-infected women had significantly higher levels of the pro-inflammatory cytokines IL-6 and IL-8 (29%, P<0.02 and 42%, P<0.01, respectively), compared with cells cultured in the presence of plasma from uninfected pregnant women (Fig. 1). Levels of both IL-6 and IL-8 in the plasma alone were in most cases undetectable, with the highest level of either cytokine across all plasma samples being 32 pg/ml (data not shown). These data indicate that schistosomiasis results in the production of some factor(s), present in maternal circulation, that stimulate a pro-inflammatory cytokine response by first trimester trophoblasts.

Given that SEA are found in the circulation of infected individuals and are known to cross the placental barrier, we treated HTR8/SVneo cells with purified SEA (25 µg/ml) in culture. Within 24 h of culture, HTR8 cells treated with SEA secreted higher levels of IL-6 (3.2-fold, P = 0.01), IL-8 (1.5-fold, P = 0.02) and the pro-inflammatory chemokine, MCP-1 (also known as CCL2; 1.7-fold, P 0.04) into the culture media, as compared to HTR8/SVneo treated with media alone (Fig. 2). Not only do first trimester cell models respond to SEA with a pro-inflammatory signature, they also secrete higher levels of a chemo-attractant protein that may help recruit specific immune cells to the placenta, exacerbating the inflammatory reaction initiated by SEA at the maternal-fetal interface.

In addition to cytokine analysis, media from HTR8/SVneo cells treated with SEA for 24 h in culture were assessed for altered levels of a number of fibrotic markers. Of these, TIMP-1 production was increased in media from HTR8 treated with SEA, compared to media from cells with no SEA exposure (Fig. 3). In contrast, production of CTGF, CCL18, TIMP-3 and fibronectin all showed no difference in levels secreted from HTR8/SVneo cells exposed to SEA compared to those with media alone (data not shown). Given the important role of TIMP-1 in the inhibition of MMPs, SEA may influence the ability of these first trimester cells to remodel and migrate into the maternal uterine wall.

We performed in vitro wound assays to assess migration of the first trimester cell line, HTR8/SVneo, following exposure to SEA. There was little difference in cell migration at 24 h (Fig. 4a). By 48 h however, HTR8 cells in culture with SEA had filled only 57±8% of the denuded area compared to 92±6% for untreated cells (P<0.01, Fig. 4). Cell proliferation was measured using MTT assay in all wells, however no differences were observed between those wells with SEA added compared to those with media alone indicating the wound closure was not due to a proliferative effect of SEA treatment (data not shown). Together, these data suggest that migration of first trimester trophoblast cells is decreased in the presence of SEA.

We next assessed the ability of the HTR8/SVneo cell line to invade through matrigel, a model of extracellular matrix, and a transwell insert after being treated with SEA, using a standard invasion assay [32]. Cells were pretreated with SEA for 24 h to minimize any direct effect of SEA on the matrigel. Enumeration of the cells that had traversed the matrigel and the pores of the transwell after an additional 48 h incubation showed 2.7-fold lower absolute cell numbers in those wells that had been pre-treated with SEA compared to the HTR8 cells that received media alone for the initial 24 h (P = 0.04, Fig. 5). These data indicate that SEA inhibits the migratory and invasive properties of the first trimester cell line, HTR8/SVneo.