Trichloroethylene metabolite S-(1,2-dichlorovinyl)-L-cysteine induces lipid peroxidation-associated apoptosis via the intrinsic and extrinsic apoptosis pathways in a first-trimester placental cell line

Abstract

Trichloroethylene (TCE), a prevalent environmental contaminant, is a potent renal and hepatic toxicant through metabolites such as S-(1, 2-dichlorovinyl)-L-cysteine (DCVC). However, effects of TCE on other target organs such as the placenta have been minimally explored. Because elevated apoptosis and lipid peroxidation in placenta have been observed in pregnancy morbidities involving poor placentation, we evaluated the effects of DCVC exposure on apoptosis and lipid peroxidation in a human extravillous trophoblast cell line, HTR-8/SVneo. We exposed the cells in vitro to 10–100 μM DCVC for various time points up to 24 h. Following exposure, we measured apoptosis using flow cytometry, caspase activity using luminescence assays, gene expression using qRT-PCR, and lipid peroxidation using a malondialdehyde quantification assay. DCVC significantly increased apoptosis in time- and concentration-dependent manners (p<0.05). DCVC also significantly stimulated caspase 3, 7, 8 and 9 activities after 12 h (p<0.05), suggesting that DCVC stimulates the activation of both the intrinsic and extrinsic apoptotic signaling pathways simultaneously. Pre-treatment with the tBID inhibitor Bl-6C9 partially reduced DCVC-stimulated caspase 3 and 7 activity, signifying crosstalk between the two pathways. Additionally, DCVC treatment increased lipid peroxidation in a concentration-dependent manner. Co-treatment with the antioxidant peroxyl radical scavenger (±)-α-tocopherol attenuated caspase 3 and 7 activity, suggesting that lipid peroxidation mediates DCVC-induced apoptosis in extravillous trophoblasts. Our findings suggest that DCVC-induced apoptosis and lipid peroxidation in extravillous trophoblasts could contribute to poor placentation if similar effects occur in vivo in response to TCE exposure, indicating that further studies into this mechanism are warranted.

Graphical Abstract

Introduction

Trichloroethylene (TCE) is a chlorinated volatile organic solvent most commonly used in chemical production and as an industrial metal degreaser (Waters et al., 1977; Chiu et al., 2013; NTP, 2015). Ranked as number sixteen on the U.S. Agency for Toxic Substances and Disease Registry’s Priority List of Hazardous Substances, TCE is a common environmental contaminant found in approximately 800 Environmental Protection Agency-designated Superfund sites (Chiu et al., 2013; ATSDR, 2015). Despite being classified as a “known human carcinogen,” (Guha et al., 2012; NTP, 2015), approximately 1.9 million pounds of TCE were released into the environment in 2015 (EPA, 2017). Because of its continued use and widespread persistent environmental contamination, TCE exposure continues to pose a threat to human health through ingestion of contaminated drinking water and inhalation of the volatilized chemical.

Although TCE is most commonly recognized as a renal and liver toxicant (Chiu et al., 2013), its effects during pregnancy are not well understood. TCE-induced fetal cardiotoxicity is particularly controversial: for example, some laboratory animal studies reported TCE effects (Das and Scott, 1994; Johnson et al., 2003; Rufer et al., 2010) and other studies reported no cardiac defects (Fisher et al., 2001; Carney et al., 2006). Furthermore, two systematic literature reviews concluded there was insufficient evidence to support an association between TCE exposure and congenital heart abnormalities (Watson et al., 2006; Bukowski, 2014). With regard to other pregnancy outcomes, an earlier study found no association between maternal TCE exposure and low birth weight (Lagakos et al., 1986), but more recent epidemiology studies report positive associations of TCE exposure during pregnancy with decreased fetal weight and preterm birth (Forand et al., 2012) (Ruckart et al., 2014).

Placental toxicity could potentially mediate adverse birth outcomes such as preterm birth and decreased birth weight (Ilekis et al., 2016). In particular, emerging recent studies implicate placental insufficiency, defined as inadequate maternal-fetal nutrient and waste exchange resulting from placental abnormalities, as a potential cause of premature labor (Morgan, 2014; Morgan, 2016) Furthermore, a recent epidemiology study found a significant association between pre-eclampsia and preterm birth (Davies et al., 2016). Because it is highly perfused, the placenta is readily exposed to circulating TCE and its metabolites, and may be a target for TCE toxicity (Laham, 1970). Moreover, the placenta is capable of metabolizing compounds, which puts it at risk for tissue generation of toxic TCE metabolites (Burton and Fowden, 2015).

Previous studies have shown that TCE exerts its toxic effects primarily through its metabolites (Lash et al., 2014). For example, exposure to the glutathione conjugation pathway metabolite S-(1, 2-dichlorovinyl)-L-cysteine (DCVC) is toxic in vitro to renal proximal tubular cells, the main putative target in the kidney, of rats, mice and humans (Lash and Anders, 1986; Darnerud et al., 1989; Chen et al., 2001; Lash et al., 2001; Xu et al., 2008). Furthermore, numerous studies demonstrated that DCVC induces mitochondrial dysfunction, excessive reactive oxygen species generation and subsequent lipid peroxidation in kidney cells (Lash and Anders, 1986; Chen et al., 1990; van de Water et al., 1994; van de Water et al., 1995; Chen et al., 2001; Lash et al., 2003; Xu et al., 2008). Similarly, our lab recently showed that DCVC induces a loss of mitochondrial membrane potential and increases ROS generation in the first trimester extravillous trophoblast HTR-8/SVneo cell line (Hassan et al., 2016). Taken together, the evidence indicates that mitochondrial dysfunction and aberrant ROS-generating lipid peroxidation play central roles in DCVC-mediated TCE cytotoxicity (Lash and Anders, 1986).

Mitochondria and lipid peroxidation are involved in the regulation of apoptotic pathways (Simon et al., 2000; Ayala et al., 2014), and multiple studies have demonstrated that DCVC initiates apoptosis in proximal tubular cells of humans and rodents (Van de Water et al., 1996; Chen et al., 2001; Lash et al., 2001; Xu et al., 2008). Apoptosis is a highly organized form of cell death that is tightly regulated by two primary signaling pathways depending upon the stimulus: the mitochondrial-dependent or intrinsic pathway and the cell surface death receptor-mediated or extrinsic pathway. Apoptosis is especially critical during pregnancy (Straszewski-Chavez et al., 2005). During placental development, apoptosis plays an important role in removing damaged cells without injuring surrounding tissues (Smith et al., 1997b; Straszewski-Chavez et al., 2005; Sharp et al., 2010). Despite its role in normal placentation, evidence suggests that an abnormal increase in apoptosis of extravillous trophoblasts contributes to multiple placental pathologies including intrauterine growth restriction and pre-eclampsia (Smith et al., 1997a; DiFederico et al., 1999; Genbacev et al., 1999). The present study investigated the effects of DCVC on the two primary apoptosis signaling pathways and lipid peroxidation in the human extravillous trophoblast cell line HTR-8/SVneo.

Materials and Methods

Chemicals and reagents

The trichloroethylene metabolite S-(1, 2-dichlorovinyl)-L-cysteine (DCVC) was synthesized by the University of Michigan Medicinal Chemistry Core according to procedures described by McKinney et al. (McKinney et al., 1959). Purity (98.7%) was determined by HPLC analysis. A stock solution was prepared in PBS and identity was confirmed by proton nuclear magnetic resonance spectroscopy performed at the University of Michigan Biochemical Nuclear Magnetic Resonance Core. Phosphate buffered saline (PBS) and 0.25% trypsin were purchased from Invitrog en Life Technologies (Carlsbad, CA, USA). BI-6C9 tBID inhibitor, tert-butyl hydroperoxide (TBHP) and (±)-α-tocopherol were purchased from Sigma-Aldrich (St. Louis, MO). Camptothecin was purchased from Cayman Chemical (Ann Arbor, MI). RPMI 1640 culture medium with L-glutamine and without phenol red, 10,000 U/mL penicillin/10,000 μg/mL streptomycin (P/S) solution, and fetal bovine serum (FBS) were purchased from Thermo Fisher Scientific (Waltham, MA, USA).

Cell culture and treatment

The HTR-8/SVneo cell line, a gift from Dr. Charles H. Graham (Queen’s University, Kingston, Ontario, Canada), models first-trimester extravillous cytotrophoblasts in vitro (Graham et al., 1993). HTR-8/SVneo cells were originally derived from first trimester female human cytotrophoblast cells and immortalized with simian virus 40 large T antigen (Graham et al., 1993). HTR-8/SVneo cells were cultured as previously described (Tetz et al., 2013; Hassan et al., 2016). Briefly, cells were cultured between passages 71–87 in RPMI 1640 medium supplemented with 10% FBS and 1% P/S at 37°C in a 5% CO₂ humidified incubator. Cells were maintained in RPMI 1640 growth medium with 10% FBS and 1% P/S prior to and during experiments to ensure optimal cell growth (Graham et al., 1993). Cells were grown to 70–90% confluence at least 24 h after subculture before starting any experiment. A stock solution of 1 mM DCVC was prepared in PBS and stored in 1-ml aliquots at −20°C to minimize freeze/thaw cycles. Prior to each experiment, a DCVC stock solution aliquot was quickly thawed in a 37°C water bath and then diluted in RPMI 1640 medium with 10% FBS and 1% P/S to final exposure concentrations of 10–100 μM DCVC. The DCVC concentrations were selected for the current study to include the mean peak blood concentration of 13.4 μM S-(1,2-dichlorovinyl)glutathione, the stable precursor of DCVC, measured in female volunteers exposed to 100 ppm of TCE by inhalation for 4 h (Lash et al., 1999), with higher concentrations consistent with DCVC-induced cytotoxicity in human placental cells and human proximal tubular cells in vitro (Xu et al., 2008; Hassan et al., 2016).

Cell line validation

HTR-8/SVneo cells were seeded at a density of 400,000 cells per well in a 6-well cell culture plate and allowed to adhere for 24 h. Cells were treated with RPMI 1640 medium alone for 24 h. Following exposure, DNA was extracted using QIAamp® DNA Mini Kit (Qiagen; Hilden, Germany). RNA samples were frozen at −20°C and transported to the University of Michigan DNA Sequencing Core for completion of the cell line validation process using microsatellite genotyping. At the core, AmpFLSTR Identifiler Plus PCR Amplification Kit run on an 3730XL Genetic Analyzer purchased from Applied Biosystems (Waltham, MA) was utilized to identify human genomic DNA for 8 tetranucleotide repeat loci and the Amelogenin gender determination marker. The short tandem repeat profile generated for our cells was compared to the short tandem repeat profile for HTR-8/SVneo (ATCC® CRL-3271^™) published by American Type Culture Collection (Manassas, VA) (ATCC, 2015). The short tandem repeat profile was an exact match: Amelogenin gender determination marker: X, CSF1PO: 12, D13S317: 9,12, D16S539: 13D5S818: 12, D7S820: 12, TH01: 6,9.3, vWA: 13,18, TPOX: 8 (ATCC, 2015).

Apoptosis assessment with flow cytometry

HTR-8/SVneo cells were seeded at a density of 50,000 cells per well in a 24-well plate and allowed to adhere for 24 h prior to treatment. Cells were treated with medium alone (control), DCVC (10, 20, 50 and 100 μM), or 4 μM camptothecin (positive control) in triplicate. After 12 or 24 h of exposure, viable, early apoptotic, late-apoptotic, and necrotic cells were measured using the Annexin V FITC Assay Kit (Cayman Chemical; Ann Arbor, MI). This assay uses two staining solutions simultaneously: annexin V, which exclusively binds to phosphatidylserines externalized on the plasma membrane during apoptosis, and propidium iodide (PI), which enters the cell when the plasma membrane is compromised in late -apoptotic and necrotic cells. The assay was performed according to the manufacturer’s protocol with modifications. Following exposure, cells were treated with 0.25% trypsin-EDTA and incubated for 2 min to detach adherent cells. Following incubation, trypsin was deactivated with cell culture medium and triplicate wells were pooled into one 5-ml round-bottom fluorescence-activated cell sorting tube per treatment group. Cells were washed with 1 ml Annexin V Binding Buffer and resuspended with 500 μl Annexin V FITC/Propidium Iodide Staining Solution diluted in Binding Buffer, incubated for 10 min at room temperature and transported to the University of Michigan Flow Cytometry Core for analysis. Early apoptotic, late apoptotic and necrotic cells were quantified using a FACSAria II Flow Cytometer (BD Biosciences; San Jose, CA). Quantification and dye-based visualization were performed using BD FACSDiva Software (BD Sciences; San Jose, CA) for PC. Cell aggregates and cellular debris were excluded using the forward scatter and side scatter modes and the analysis was performed using 10,000 events per sample. Annexin V-FITC was detected using the 488 nm laser, while PI was detecting using the 561 nm laser.

Measurement of caspase 3, 7, 8 and 9 activity

HTR-8/SVneo cells were seeded at a density of 10,000 cells per well in a 96-well white, clear-bottom plate and allowed to adhere for 24 h prior to treatment. Cells were treated in quadruplicate with RPMI 1640 medium alone (control) or DCVC (20 or 100 μM). Following 3, 6, 12, or 24 h treatments, caspase activity was measured using luminescence-based Caspase-Glo 3/7, 8 and 9 Assays (Promega; Madison, WI) following the manufacturer’s recommended protocols. Caspase assays utilize luminogenically engineered caspase substrates that are specific to each type of caspase being measured based on specific amino acid sequences embedded within the substrates: DEVD (Asp -Glu-Val-Asp) for caspase 3 and 7, LETD (Leu-Glu-Thr-Asp) for caspase 8, and LEHD (Leu-Glu-His-Asp) for caspase 9. Following cleavage of the substrates, the product generated, aminoluciferin, reacts with luciferase to produce light proportional to the amount of caspase activity (Niles et al., 2008). Briefly, for Caspase-Glo 3/7, 8 and 9 assays, respectively, 100 μL Caspase-Glo substrate in Caspase-Glo buffer was added to cell cultures and incubated at room temperature for 1 h following DCVC treatment. In order to reduce non-specific background signal, MG-132, a proteasome inhibitor, was added to caspase 8 and 9 cell cultures only. Following incubation, luminescence signal was measured on Glomax Multi Plus Detection System (Promega; Madison, WI).

Apoptosis PCR array

Because DCVC stimulated apoptosis as shown by annexin V-FITC/PI staining, we evaluated changes in the gene expression of 84 genes specifically involved in apoptosis using the commercially available Apoptosis PCR Array manufactured by SABiosciences (Valencia, CA). HTR-8/SVneo cells were seeded at a density of 400,000 cells per well in a 6-well cell culture plate and allowed to adhere for 24 h. Following the adjustment period, the cells were treated with RPMI 1640 medium alone (control) or DCVC (20, 50 and 100 μM) and exposed for 24 h. The concentrations of 20, 50 and 100 μM were chosen to span a range of concentrations from no effect to effective concentrations that induced apoptosis in our earlier experiment. Following exposure, cell lysates were homogenized using QIA shredder (Qiagen; Hilden, Germany) and RNA was extracted using the RNeasy Plus Mini Kit (Qiagen) following the manufacture’s recommended protocol. RNA samples were frozen at −20°C and transported to the University of Michigan DNA Sequencing Core for completion of the Apoptosis PCR Array. At the core, cDNA was synthesized using the RT2 First Strand Kit (SABiosciences; Valencia, CA) following the manufacturer’s recommended protocol. For the array, cDNA from the culture medium-only control, 20 μM DCVC, 50 μM DCVC and 100 μM DCVC treatment groups was analyzed using the Applied Biosystems 7900HT Sequence Detection System following the SABiosciences recommended protocol. Fold changes were calculated from ΔCT values (gene of interest CT value – average of all housekeeping gene CT values) using the ΔΔCT method. Mean ΔCT values were compared between groups using paired t-tests from the Limma package of Bioconductor (Smyth, 2004). The resulting P-values were adjusted for multiplicity using the Benjamini and Hochberg false discovery rate method (Benjamini and Hochberg, 1995).

Real-time quantitative PCR (qRT-PCR) validation

We validated the findings of the PCR-based array using qRT-PCR for genes with significant differential mRNA expression at least 2 fold higher or 0.5 fold lower than control following 100 μM DCVC treatment for 24 h. These genes included: Harakiri, BCl-2 interaction protein (HRK), BCL2-related protein A1 (BCL2A1), growth arrest and DNA-damage-inducible, alpha (GADD45α), DNA fragmentation factor alpha polypeptide (c), Receptor-interacting serine-threonine kinase 2 (RIPK2), nuclear factor kappa B subunit 1 (NFKB1), caspase 1 (CASP1) caspase 4 (CASP4), BCL2-antagonist/killer 1 (BAK1), and TNF receptor-associated factor 3 (TRAF3). We also tested differential expression of the tumor suppressor protein p53 (TP53) and TNF receptor superfamily, member 6 (FAS) genes because they showed a trend towards upregulation in the PCR-based array. Primary sequences are shown in Supplemental Table 1. Primary sequences for BCL2A1, DFFA, RIPK2, TRAF3, NFKB1, TP53, CASP1, BAK1, FAS AND CASP4 were obtained from the online primer sequence database PrimerBank (Spandidos et al., 2010). HRK primer sequence was obtained from Zaker et al. (Zaker et al., 2016). GADD45α primer sequence was obtained from Zhang et al. (Zhang et al., 2011). qRT-PCR was performed on the above genes using samples from cells treated with RPMI 1640 medium alone (control), or DCVC (20 and 100 μM) for 24 h. qRT-PCR reactions were prepared with SYBR Green Mastermix (SABiosciences) and custom synthesized primers (Integrated DNA Technologies; Coralville, IA), and run on a Bio-Rad (Hercules, CA) CFX96 Real Time C1000 thermal cycler following the manufacturer’s recommended protocols. mRNA levels of each gene of interest were normalized to β-2-microglobulin (B2M) mRNA levels.

Visualization of intrinsic and extrinsic apoptosis pathways

We used PathVisio software (version 3.2.4) to visualize DCVC-induced gene expression changes in the intrinsic and extrinsic apoptotic pathways. Pathways were constructed from WikiPathways (Apoptosis-Homo Sapiens) (Kelder et al., 2012; Zambon et al., 2017) with modifications from Kegg Pathway Database (Apoptosis [map04210] and NOD-like receptor [map04621]-Homo sapiens signaling pathways) (Kanehisa et al., 2017) and other sources from the literature (Inohara et al., 1997; McCarthy et al., 1998; Gao and Abu Kwaik, 2000; Haupt et al., 2003; Roth et al., 2003; Perfettini et al., 2004; Amaral et al., 2010; Graupner et al., 2011; Nishizaki et al., 2014). Genes found to be differentially expressed in either the 84-gene qRT-PCR array experiments or in qRT-PCR experiments were included in the visualization (except CD70 and IGR1F).

ROS lipid peroxidation assessment

Because our evidence indicated that DCVC exposure activated the mitochondria-dependent intrinsic apoptosis pathway, we investigated the effect of DCVC treatment on the degradation of membrane lipids by ROS by measuring a lipid peroxidation by-product, malondialdehyde (MDA). HTR-8/SVneo cells were seeded at a density of 150,000 cells per well in 12-well plates and allowed to adhere for 24 h prior to treatment. Cells were treated with medium alone (control), DCVC (10, 20, and 50 μM), or 20 μM tert-butyl hydroperoxide (TBHP) (positive control) for 12 h. MDA was measured using a Lipid Peroxidation (MDA) Assay Kit (Sigma-Aldrich; St Louis, MO) according to the manufacturer’s protocol. The assay uses the reaction of MDA with thiobarbituric acid (TBA) to form an MDA-TBA adduct product proportional to the amount of MDA present in the samples. Briefly, cells were lysed using MDA lysis buffer containing butylated hydroxytoluene. Insoluble material was removed by centrifugation and lysates were transferred to microcentrifuge tubes. TBA solution was added to the lysate and incubated at 95°C for 1 h. Lysate was cooled to room temperature with an ice bath for 10 min and plated on a 96-well plate for analysis. MDA concentration was measured colorimetrically (OD = 532 nm) using a SpectraMax M2e Multi-Mode Microplate Reader (Molecular Devices; Sunnyvale, CA).

Modulation of DCVC-stimulated caspase 3 and 7 activity

In order to investigate the role of aberrant ROS generation on DCVC-stimulated caspase 3 and 7 activity, the antioxidant peroxyl radical scavenger (±)-α-tocopherol was used to block lipid peroxidation. HTR-8/SVneo cells were seeded at a density of 10,000 cells per well in a 96-well white, clear-bottom plate and allowed to adhere for 24 h prior to treatment. Cells were treated in quadruplicate with medium alone (control) or DCVC (20 μM) plus (±)-α-tocopherol (50 μM). The cells were pre-treated for 15 min with (±)-α-tocopherol prior to co-treatment with DCVC for 12 h. We used a concentration of 50 μM (±)-α-tocopherol because it was previously shown by our lab to attenuate DCVC-stimulated IL-6 release in HTR-8/SVneo cells (Hassan et al., 2016). In addition, it showed a protective role in TCE-induced human epidermal keratinocytes at a concentration of one order of magnitude higher (50 mM) (Zhu et al., 2005). Caspase 3 and 7 activity was measured with Caspase-Glo 3/7 Assay as previously described.

In order to gain more insight into the roles that the intrinsic and extrinsic apoptosis pathways play in DCVC-stimulated caspase 3 and 7 activity, we explored the involvement of the truncated BID protein, tBID, as a source of cross talk between the two pathways. HTR-8/SVneo cells were seeded and cultured as previously described. Cells were treated in triplicate with medium alone (control), DCVC (20 μM) or DCVC plus tBID inhibitor BI-6C9 (10 μM). We used a concentration of 10 μM BI-6C9 because it was shown in a previous study to moderately inhibit BID-induced apoptosis without cytotoxicity (Becattini et al., 2004). The cells were pre-treated for 1 h prior with BI-6C9, followed by treatment with DCVC for 12 h. Caspase 3 and 7 activity was measured with Caspase-Glo 3/7 Assays as previously described.

Statistical Analysis

All experiments were performed independently in at least triplicate and repeated at least three times. The technical replicates were averaged within each experiment, and these values were analyzed using student’s t-test, one-way or two-way analysis of variance (ANOVA), followed by Tukey’s or Dunnett’s post-hoc test for comparison of means using GraphPad Prism software (GraphPad Software Inc., San Diego, CA, USA). Data are expressed as means ± SEM. N=number of independent experiments. P < 0.05 was considered statistically significant.

Results

Effect of DCVC treatment on apoptosis

Because our laboratory previously demonstrated that DCVC treatment impacts cell viability, we investigated the effects of DCVC exposure on a specific form of cell death, apoptosis. Early apoptosis, late apoptosis, and necrosis were quantified in HTR-8/SVneo cells with annexin V-FITC and propidium iodide fluorescence indicators using flow cytometry (Fig. 1). Statistically significant concentration-dependent increases of total apoptosis were observed after 24 h but not 12 h of DCVC exposure (ANOVA interaction effect, P=0.001, Fig. 1). No statistically significant treatment-related differences were observed with 12 h of exposure, but treatment with 50 and 100 μM DCVC increased total apoptosis in a concentration-dependent manner at 24 h. Following 24-h treatment, total apoptosis increased from 6.38% (control) to 14.73% and 23.48% with 50 and 100 μM DCVC treatment, respectively (P<0.001). The 100 μM DCVC treatment increased apoptosis 2.3 fold at 24 h compared with 12 h (P<0.0001).

Figure 1. Effects of DCVC on apoptosis in HTR-8/SVneo cells.

Cells were treated for 12 h or 24 h with 0 (no treatment control), 10, 20, 50, or 100 μM DCVC. A) Annexin V-FITC and propidium iodide fluorescence indicators were used to quantify apoptosis using flow cytometry. The dot plots are representative of each experiment at 24 h. B) Graphical representation of total apoptosis (early+late apoptosis). Bars represent means±SEM. Data were analyzed by two -way ANOVA (interaction between time and treatment, P=0.001) with posthoc Tukey multiple comparisons. ^#Significantly different compared to all 12 h time points (P<0.0022). ^&Significantly different compared to control and 10 μM DCVC at 24 h time point (P<0.0088). ⁺Significantly different compared to control, 10 and 20 μM DCVC at 24 h time point (P<0.0001). *Significantly different compared to 50 μM DCVC at 24 h time (P=0.005). N=4 independent experiments for 12 h and N=5 independent experiments for 24 h. All experiments were performed in triplicate. Camptothecin (4 μM) was included as a positive control and increased total apoptosis 32.89% +/−15.23% at 12 h and 39.11% +/− 6.36 at 24 h.

Effect of DCVC treatment on caspase activity

We measured the effects of DCVC on caspase activity to evaluate the involvement of the intrinsic and extrinsic apoptotic pathways in DCVC-induced apoptosis (Fig. 2). Caspases 3 and 7, which are downstream executor caspases activated by both the intrinsic and extrinsic pathways, showed time - and concentration-dependent increases in activity following DCVC exposure for 12 and 24 h (ANOVA time and treatment interaction, P<0.0001, Fig. 2A). Caspases 3 and 7 measured after 3 and 6 h of DCVC treatment showed no significant changes in activity. In contrast, treatment with 20 and 100 μM DCVC increased caspase 3 and 7 activity by 86.2% (P=0.0015) and 187.5% (P<0.001) at 12 h, and 88.6% (P<0.001) and 155.5% (P<0.001) at 24 h, respectively, compared with time -matched controls. Likewise, increases of caspase 3 and 7 activity were observed at 12 h and 24 h compared to earlier time points, ranging from 2.4 to 3.5 fold within each DCVC treatment (P<0.0001).

Figure 2. DCVC effects on caspase activity in HTR-8/SVneo cells.

Cells were treated for 3, 6, 12 or 24 h with 0 (control), 20 or 100 μM DCVC. Caspase activity was measured using the Caspase Glo® luminescence assays. A) Time course of caspase 3 and 7 activity. B) Caspase 9 activity. C) Caspase 8 activity. Symbols represent means±SEM. N=3 independent experiments. Data were analyzed by two-way ANOVA (interaction between time and treatment, P<0.0001) with posthoc Tukey multiple comparisons. Given the very large number of significant mean comparisons, only significant comparisons within a treatment across time points and within a time point across treatments are indicated. Asterisks indicate significant differences with all lower concentrations within a single time point across treatments: * P<0.05, ** P<0.005, *** P<0.0001. Pound signs indicate significantly different compared with all earlier time points for same treatment: ^# P<0.05, ^### P<0.0001. N=3 independent experiments. All experiments were performed in quadruplicate.

Caspase 9, an initiator caspase specific to the mitochondria-mediated intrinsic pathway, showed time- and concentration-dependent increases in activity following DCVC exposure (ANOVA time and treatment interaction, P<0.0001, Fig. 2B). Caspase 9 activity measured after 3 and 6 h of DCVC treatment showed no changes in activity level compared to time-matched controls. However, caspase 9 activity measured at 12 h showed a 56.3% increase in activity compared to time-matched control for 100 μM DCVC, whereas activity measured at 24 h showed 41.4% and 69.6% increases in activity compared to time-matched controls (P<0.0001). The 20 μM DCVC treatment increased caspase 9 activity 1.32 fold at 12 h compared to 6 h (P=0.0228) and 1.42 fold at 24 h compared to 12 h (P<0.0001). Similarly, 100 μM DCVC treatment increased caspase 9 activity 1.66 fold at 12 h compared to 6 h and 1.29 fold at 24 h compared to 12 h (P<0.0001).

Caspase 8, an initiator caspase specific to the cell surface receptor-mediated extrinsic pathway, showed time- and concentration-dependent increases in activity following DCVC exposure (ANOVA time and treatment interaction, P<0.0001, Fig. 2C). Caspase 8 activity measured after 3 and 6 h post-DCVC treatment showed no significant differences in activity compared to time-matched controls. However, caspase 8 activity measured at 12 h showed 36.5% (P=0.029) and 72.0% (P<0.0001) increases in activity compared to time-matched control for 20 and 100 μM DCVC, respectively, and activity measured at 24 h showed 30.6% (P=0.0002) and 70.2% (P<0.0001) increases in activity compared to time-matched control. The 20 μM DCVC treatment increased caspase 8 activity 1.86 fold at 12 h compared to 6 h and 1.39 fold at 24 h compared to 12 h (P<0.0001). The 100 μM DCVC treatment increased caspase 8 activity 2.39 fold at 12 h compared to 6 h and 1.43 fold at 24 h compared to 12 h (P<0.0001).

Inhibition of DCVC-stimulated caspase 3 and 7 by tBID protein inhibitor

In order to ascertain whether the DCVC mechanism of cytotoxicity involves interaction between the intrinsic and extrinsic apoptosis pathways, we used BI-6C9, which inhibits tBID-mediated cross talk between the two pathways. HTR-8/SVneo cells were pre-treated for 1 h with 10 μM BI-6C9 followed by treatment with 20 μM DCVC for 12 h. BI-6C9 pre-treatment attenuated DCVC-induced caspase 3 and 7 activity by 60.77% compared to cells treated with DCVC without the inhibitor (ANOVA interaction between DCVC and BI-6C9 treatments, P=0.028; Fig 3).

Figure 3. Effect of tBID inhibitor pre-treatment on DCVC-stimulated caspase 3/7 activity.

Cells were exposed to 20 μM DCVC for 12 h without pretreatment or after pretreatment with the tBID inhibitor BI-6C9 (10 μM for 1 h). Controls were not exposed to DCVC. Bars represent means±SEM. Data were analyzed by two-way ANOVA (interaction between DCVC and BI-6C9 treatments, P=0.028), followed by Tukey’s multiple comparison of means. ^#Significantly different compared to control with no inhibitor (P=0.001). *Significantly different compared to 20 μM DCVC with no inhibitor (P<0.010).

Apoptosis PCR array and qRT-PCR Validation

Because there are potentially hundreds of genes involved in the activation of apoptosis in cells, we used a commercially available PCR-based apoptosis array to screen 84 apoptosis-related genes for DCVC-induced changes in mRNA expression following treatment with 100 μM DCVC for 24 h (Supplement Table 2; P<0.05). Using this array, we identified ten genes that were significantly upregulated at least 2 fold or significantly downregulated at least 0.5 fold compared to control, and two other genes that showed a trend towards upregulation (Fig 4; P<0.05). We used qRT-PCR to follow up on these array results. In agreement with the PCR-based array results, 20 and 100 μM DCVC increased mRNA expression of the BCL-2 family genes HRK, BAK1 and BCL2A1 by 13.1 and 19.4 fold, 1.3 and 1.4 fold, and 7.6 and 8.0 fold, respectively, compared to control (P<0.05). Tumor suppressor gene TP53 showed 1.7 and 1.3 fold increases in mRNA expression following 24-h treatment with 20 and 100 μM DCVC, respectively (P<0.05). Cell cycle arrest gene GADD45α expression increased 9.8 and 9.7 fold, while FAS death receptor gene increased 1.8 fold with 20 and 100 μM DCVC treatments, respectively, compared to control (P<0.05). Inflammatory response genes NFKB1, RIPK2, and CASP4 were upregulated 2.0 and 1.8 fold, 2.0 and 1.6 fold, and 1.7 and 2.5 fold, respectively, while CASP1 was downregulated 0.5 and 0.2 fold, respectively, compared to control (P<0.05). Contrary to the array data, TRAF3 and DFFA did not yield a significant increase in expression, although DFFA showed a trend towards upregulation of 1.5 fold compared to control with 100 μM DCVC treatment (P=0.0582). Relevant genes significantly up or down regulated in the apoptosis array and/or qRT-PCR are shown in Figure 5.

Figure 4. DCVC effects on HTR-8/SVneo cell mRNA expression of genes implicated in apoptosis signaling.

Based on the results of an apoptosis PCR array, we identified ten genes that were significantly upregulated by at least 2 fold or downregulated at least 0.5 fold compared to control, following treatment with 100 μM DCVC for 24 h. The PCR array data results were confirmed using qRT-PCR. Two other genes were also included in the qRT-PCR analysis because they showed a trend of upregulation. Cells were exposed to 0 (control), 20 μM DCVC or 100 μM DCVC for 24 h followed by qRT-PCR analysis. A) The mRNA expression of TP53, FAS, DFFA and BAK1. B) The mRNA expression of NFKB1, RIPK2, TRAF3 and CASP1. B) The mRNA expression of HRK, GADD45α, BCL2A1 and CASP4. Bars represent means±SEM. * Significantly different compared to control (P<0.05). ** Significantly different compared to control (P<0.01). *** Significantly different compared to control (P<0.001). Data were analyzed by one -way ANOVA with posthoc Dunnett’s multiple comparison. N=3–6 independent experiments. All experiments were performed in triplicate.

Figure 5. Visualization of gene expression changes in intrinsic and extrinsic apoptosis pathways in HTR-8/SVneo cells placental after exposure to DCVC.

In order to elucidate the gene expression patterns that regulate the apoptosis observed in cells after treatment with DCVC, we mapped gene expression changes for extrinsic and intrinsic apoptosis pathways using Pathvisio software. Gene expression changes (log2 fold-change/control) are shown with red=upregulation, green=downregulation, grey=gene not evaluated. All gene expression data were obtained from an 84-gene qRT-PCR array except that genes indicated with an asterisk (*) had expression data obtained from separate qRT-PCR experiments. Two genes, IGF1R and CD70, were differentially expressed in the array but are not included in the diagram. This figure was constructed in part from Wikipathways: Apoptosis (Homo sapiens) (Zambon et al., 2017).

Effect of DCVC treatment on ROS lipid peroxidation

In order to further investigate the mechanism of DCVC-induced cytotoxicity, we evaluated the effects of DCVC on lipid peroxidation in HTR-8/SVneo cells by measuring cellular malondialdehyde concentrations, a byproduct and proxy measurement for lipid peroxidation. We detected a significant concentration-dependent increase in cellular malondialdehyde concentration following 24-h exposure to 10, 20 and 50 μM DCVC compared to control (Fig 6; P<0.05).

Figure 6. Effects of DCVC on lipid peroxidation in HTR-8/SVneo cells.

Cells were treated for 24 h with 0 (control), 10, 20, and 50 μM DCVC or 20 μM tert-butyl hydroperoxide (positive control). As a proxy measure for lipid peroxidation, malondialdehyde (MDA) concentrations were quantified using thiobarbituric acid (TBA), which generates a MDA-TBA adduct when in contact with MDA. The MDA concentrations were quantified colorimetrically (OD = 532). Bars represent means±SEM. *Significantly different compared to control (P<0.004). ^#Significantly different compared to 10 μM DCVC (P=0.0020). ^&Significantly different compared 20 μM DCVC (P= 0.0389). Data were analyzed by one -way ANOVA with posthoc Tukey multiple comparisons. N=4 independent experiments. All experiments were performed in triplicate. TBHP (20 μM) was used as a positive control (2.87 nmol +/− 0.167 nmol).

Attenuation of DCVC-stimulated caspase 3 and 7 activity by an antioxidant

To test the hypothesis that lipid peroxidation mediates DCVC-stimulated caspase 3 and 7 activity and apoptosis, we co-treated HTR-8/SVneo cells with either 20 μM DCVC only or 20 μM DCVC plus (±)-α-tocopherol (50 μM), an antioxidant peroxyl radical scavenger. Co-treatment with (±)-α-tocopherol significantly attenuated DCVC-stimulated caspase 3 and 7 activity by 80% compared to cells treated only with 20 μM DCVC for 12 h (ANOVA interaction between DCVC and α-tocopherol treatments, P=0.0103; Fig Discussion

Trichloroethylene continues to be a pervasive environmental contaminant in soil and groundwater despite reduction of improper disposal of the parent compound in recent decades. When TCE is ingested or inhaled, it is metabolized by two primary pathways into harmful metabolites, one of which is S-(1, 2-dichlorovinyl)-L-cysteine (DCVC), the focus of the current study. Although DCVC has been studied and established as a renal toxicant and carcinogen (Lash et al., 2014), epidemiological and animal studies linking TCE to pregnancy complications and adverse birth outcomes thus far remain inconclusive (Chiu et al., 2013), prompting our laboratory to examine DCVC as a placental toxicant. The general objectives of the current study were to further test the biological plausibility of DCVC as a placental toxicant and to examine its possible mechanism of action. In this study, we demonstrated that DCVC induced lipid peroxidation-associated apoptosis in placental cells through activation of both the intrinsic and extrinsic signaling pathways, providing new insights into the biological plausibility of DCVC-induced placental injury and the cytotoxic mechanism of action of DCVC.

Our laboratory previously reported that DCVC decreased cell viability and increased cytotoxicity in HTR-8/SVneo, an immortalized cell line that retains key molecular and functional characteristics of extravillous trophoblasts and which serves as an in vitro model for these cells (Kilburn et al., 2000; Hannan et al., 2010; Khan et al., 2011; Hassan et al., 2016). The present study extends those findings with evidences that DCVC directly induced aberrant apoptosis, a specific form of cell death, in the same HTR-8/SVneo cell type. These findings are consistent with previous reports that DCVC exposure induces apoptosis in kidney proximal tubular cells of humans and rodents (Van de Water et al., 1996; Chen et al., 2001; Lash et al., 2001; Xu et al., 2008). In addition, aberrant and widespread apoptosis of extravillous trophoblasts has been observed in several pregnancy-related morbidities specifically characterized by poor placentation such as pre-eclampsia and intrauterine growth restriction (DiFederico et al., 1999; Genbacev et al., 1999; Reister et al., 2001). Aberrant placental apoptosis is linked to major pathological features of these pregnancy morbidities including insufficient extravillous trophoblast invasion into the maternal decidua and inadequate remodeling of the spiral arteries needed to accommodate increased maternal blood flow to the placenta (Brosens et al., 1972; Meekins et al., 1994; Caniggia et al., 1999; Aardema et al., 2001; Kadyrov et al., 2006; Pennington et al., 2012). Because a growing body of evidence ties excessive apoptosis to impaired trophoblast invasion and inadequate remodeling of the spiral arties commonly observed with pregnancy-related morbidities, our results suggest that DCVC-induced apoptosis in extravillous trophoblasts could be a mechanism by which TCE contributes to poor placentation and early pregnancy morbidities if it occurs in vivo.

In order to gain further insight into the cytotoxic mechanism of action, we evaluated specific signaling pathways involved in DCVC-stimulated apoptosis. Depending upon the stimulus, there are two well-characterized pathways that are capable of inducing apoptosis: the extrinsic and intrinsic pathways, as illustrated in Figure 8 (Ashkenazi, 2008; Beesoo et al., 2014). The intrinsic pathway is activated by non-receptor stress stimuli such as hypoxia, damaged DNA and ROS (Elmore, 2007). These stimuli induce apoptosis through interactions with the BCL-2 family of mitochondrial proteins including BAX and BAK, which control mitochondrial membrane permeability. Upon activation, the pro-apoptotic BCL-2 family proteins dimerize to increase the permeability of the mitochondrial membrane, allowing cytochrome c to leak into the cytoplasm and prompting activation of the intrinsic pathway-specific initiator caspase 9 (Elmore, 2007). In contrast, the extrinsic pathway is stimulated when specific tumor necrosis factor (TNF) superfamily cytokines bind to TNF cell surface death receptors, prompting cytoplasmic adapter proteins to form complexes that activate extrinsic pathway-specific initiator caspase 8 (Straszewski-Chavez et al., 2005; Elmore, 2007). Following activation of caspase 8 and 9, both apoptosis pathways converge in a common pathway when caspases 8 and 9 activate executioner caspases 3 and 7, ultimately leading to the cleavage of cellular proteins and destruction of the cell (Levy and Nelson, 2000; Elmore, 2007). In the current study, we showed that DCVC exposure simultaneously activated both major apoptosis pathways, as indicated by caspase 8 and 9 activity results, in time- and concentration-dependent manners. We chose to measure caspase enzymatic activity as an endpoint that reflects caspase zymogen activation as well as enzyme abundance (Niles et al., 2008). Although western blot analysis would provide specific information on protein expression, a study that compared the sensitivity and specificity of Caspase Glo assays to western blot analysis demonstrated a strong correlation between activity levels measured by Caspase Glo 3/7, 8, and 9 assays and western blot detection of active caspase enzymes (Alvero et al., 2008).

Figure 8. Intrinsic and extrinsic apoptosis pathways.

The mitochondria-mediated intrinsic pathway is activated in response to stress stimuli. BCL-2 family proteins with the BH3-only domain such as HRK translocate into mitochondria and dimerize with anti-apoptotic BCL-2, releasing pro-apoptotic BAK and BAX, which increase mitochondrial membrane permeability. In response, mitochondrial components such as cytochrome c leak out and bind to caspase 9 and APAF-1 to form the apoptosome. This complex activates caspases 3 and 7, which execute apoptosis. The extrinsic pathway is activated when TNF family ligands bind to receptors and activate caspase 8, which then activates caspases 3 and 7. Cross talk occurs when extrinsic pathway protein caspase 8 cleaves BID into truncated tBID. tBID translocates into the mitochondria and dimerizes with BCL-2 family proteins, engaging the intrinsic pathway.

To our knowledge, our study is the first to demonstrate that DCVC is capable of activating both the intrinsic and extrinsic apoptosis pathways simultaneously, as indicated by the caspase results, as part of its cytotoxicity mechanism. Interestingly, these results only partially agree with a previous study that reported involvement of the mitochondrial-mediated intrinsic but not extrinsic pathway in DCVC-induced apoptosis (Xu et al., 2008). The differences between the study findings may be attributed to the different cell types and/or exposure concentrations examined. For example, contrary to our study, Xu et al. investigated the effects of DCVC in concentrations up to 300 μM in primary human proximal tubular cells.

Because our results clearly showed that DCVC exposure activated both intrinsic and extrinsic apoptosis signaling pathways in HTR-8/SVneo cells, we examined potential interactions between the two pathways by evaluating cross talk between the pathways. The best characterized source of cross talk between the two pathways involves the truncated protein tBID. The pathway interaction typically occurs when extrinsic pathway-specific caspase 8 cleaves the cytosolic BCL-2 protein family member BID into its truncated form, tBID. tBID then translocates across the mitochondrial membrane and engages the intrinsic pathway by dimerizing with BCL-2 family proteins and increasing mitochondrial membrane permeability (Li et al., 1998; Wei et al., 2000). Our results indicated that pre-treatment with the tBID inhibitor Bl-6C9 partially attenuated DCVC-induced downstream activation of caspases 3 and 7, demonstrating that cross talk between the two pathways contributes to the DCVC mechanism of stimulating apoptosis in placental cells.

In agreement with these results, we further observed that DCVC treatment stimulated an increase in caspase 3 and 7 activity that was more than double the increase in activity observed in either caspase 8 or 9 when compared to control. As reported in previous studies, the latter results strongly suggest that the cross talk occurring between the two pathways amplifies the effect on the common terminal pathway, which may also explain the magnitude of the DCVC-induced apoptotic response (Kuwana et al., 1998; Chou et al., 1999). Although multiple studies have established that first-trimester extravillous trophoblasts are capable of undergoing apoptosis via either the intrinsic or extrinsic pathways (Belkacemi et al., 2009; Belkacemi et al., 2011; Huang et al., 2014), to our knowledge, our study is the first to demonstrate a simultaneous pathway activation involving tBID-mediated cross talk in response to exposure to an exogenous chemical stimulant.

To elucidate the specific signaling mechanism of DCVC-induced apoptosis in placental cells, we evaluated expression of apoptosis-related genes using a targeted PCR-based array and qRT-PCR. The DCVC-induced gene expression changes for the extrinsic and intrinsic apoptotic pathways are summarized in Figure 6. We detected a notable DCVC-stimulated increase in TP53 (p53) tumor suppressor gene expression consistent with previous studies (Chen et al., 2002; Rehman et al., 2013). The p53 protein activates apoptosis through multiple transcription-dependent and independent mechanisms involving both apoptosis pathways (Haupt et al., 2003; Amaral et al., 2010). In fact, the effect of DCVC on p53 may explain, at least in part, why both pathways are activated in DCVC-induced apoptosis. For example, we observed a significant upregulation of the p53 transcriptional target FAS, a gene encoding a TNF family death receptor, indicating involvement of p53 in extrinsic pathway regulation. On the other hand, we also observed a significant gene upregulation of BAK1, another p53 transcriptional target which codes for a BCL-2 family protein, indicating simultaneous involvement of p53 in intrinsic pathway regulation (Perfettini et al., 2004; Graupner et al., 2011). Additionally, p53 regulates the intrinsic pathway in a non-transcriptional manner by directly translocating into the mitochondria following stress-induced upregulation and cytosolic accumulation of the protein (Haupt et al., 2003; Amaral et al., 2010). Once inside the mitochondria, p53 forms an antagonist complex with anti-apoptotic Bcl-XL, releasing pro-apoptotic BAX and BAK1 to dimerize and increase mitochondrial membrane permeability (Perfettini et al., 2004; Amaral et al., 2010).

In addition to p53-associated apoptosis, we observed that DCVC treatment substantially increased the expression of another p53 transcriptionally regulated gene, GADD45α, which induces cell cycle arrest in response to stress stimuli (Papathanasiou et al., 1991; Hollander et al., 1993). Although previous studies failed to reach a consensus on the role of GADD45α in apoptosis signaling (Sheikh et al., 2000), GADD45α-mediated gene expression has been implicated in physiological irregularities and insufficient trophoblast invasion in pre-eclamptic placentas (Xiong et al., 2009; Xiong et al., 2013; Liu et al., 2014; Liu et al., 2016b). Because DCVC exposure increased GADD45α gene expression by a magnitude of nearly ten-fold, our findings suggest that this gene may play a pivotal role in DCVC-induced placental cell cytotoxicity: however, further investigation is needed to clarify this role.

Several other salient apoptosis-related genes demonstrated differential DCVC-stimulated expression in the current study. For example, DNA fragmentation factor gene, DFFA, a common pathway caspase 3 cleavage target that promotes DNA fragmentation, showed a trend in upregulation following DCVC exposure. In addition, CASP4, the gene coding for caspase 4, was significantly upregulated, while CASP1 was significantly downregulated. Caspases 1 and 4 are part of a family of caspases that are strongly implicated in inflammatory processes and responses to pathogens (McIlwain et al., 2015). Although their role in the induction of apoptosis is not fully understood (Denes et al., 2012), several studies have suggested that caspase 4 participates in endoplasm-reticulum stress-induced apoptosis in neuronal cells (Hitomi et al., 2004; Yamamuro et al., 2011; Li et al., 2013). More research in needed to clarify the function of caspases 1 and 4 expression in placental cells. Lastly, HRK, which showed a large 19-fold increase in expression, belongs to a BCL-2 subfamily of genes called BH3-only because they only contain one domain in common with other BCL-2 family genes. This subfamily of genes plays a pivotal role in the induction of intrinsic apoptosis because they travel freely in the cytoplasm until they are activated by receptor-free stimuli. Upon activation, they translocate into the mitochondria and interact with other mitochondrial-sequestered BCL-2 proteins like BAK1, directly stimulating the mitochondrial components of the intrinsic pathway (Inohara et al., 1997). Interestingly, a previous study also reported an increase HRK mRNA expression in abnormally fragmented pre-implantation embryos (Jurisicova et al., 2003). These results may indicate that HRK plays an important role in regulating both normal and pathological cell death during early pregnancy.

Because our prior study showed that DCVC increases ROS generation in HTR-8/SVneo cells, we tested the hypothesis that aberrant ROS generation contributes to DCVC-induced apoptosis. We found that DCVC treatment increased malondialdehydes in HTR-8/SVneo cells, a byproduct and proxy measure for lipid peroxidation. These results are consistent with previous studies that indicate DCVC cause s lipid peroxidation in kidney tubular cells of humans and rodents (Beuter et al., 1989; Chen et al., 1990; Groves et al., 1991). Lipid peroxidation occurs when reactive oxygen species, including free radicals, attack carbon double bonds in lipids resulting in lipid peroxyl radicals and hydroperoxides (Ayala et al., 2014). This process damages lipid membranes within the cell and affects their ability to function properly, especially the plasma and mitochondrial membranes. Membrane damage caused by lipid peroxidation is capable of inducing a loss of mitochondrial membrane potential because it destroys the selective barrier and ion transport properties of the membrane (Stark, 2005). Our finding that DCVC induced lipid peroxidation offers a possible mechanistic explanation for a DCVC-induced loss of mitochondrial membrane potential in HTR-8/SVneo cells, previously reported by our lab (Hassan et al., 2016).

Consistent with our finding of DCVC-induced lipid peroxidation, we observed a significant increase in NFKB1 gene expression in placental cells, as well as its upstream regulator RIPK2 (McCarthy et al., 1998). NFKB1 (nuclear factor kappa-light-chain-enhancer of activated B cells) is a transcription factor primarily involved in inflammatory responses to a variety of stimuli, most notably, reactive oxygen species (ROS) (Schreck et al., 1991). NFKB1 directly regulates the transcription levels of many cytokines such as the pro-inflammatory cytokine interleuken-6 (IL-6) (Libermann and Baltimore, 1990). Our result indicating a DCVC-induced increase in NFKB1 gene expression is consistent with our previous study that showed DCVC caused an ROS-mediated increase in IL-6 production (Hassan et al., 2016). Our NFKB1 gene expression results are particularly significant because increased NFKB1 activity and concomitant aberrant ROS generation have been observed both in early pre-eclamptic pregnancies in vivo and in trophoblast-like cells in vitro (Luppi et al., 2006; Vaughan and Walsh, 2012). Additionally, we observed the upregulation of another NFKB1 transcriptional target BCL2A1, an anti-apoptotic BCL-2 family gene. The exact nature of the upregulation of this gene remains unclear but it is likely related to a stress response -induced biological feedback system.

Consistent with lipid peroxidation, we demonstrated that the antioxidant peroxyl radical scavenger (±)-α-tocopherol attenuated DCVC-induced increase caspase 3 and 7 activity. These results suggest that DCVC-stimulated apoptotic activity is dependent on increased ROS generation and lipid peroxidation. These data agree with prior findings that increased ROS generation is capable of inducing apoptosis (Smith et al., 1999; Simon et al., 2000; Myatt and Cui, 2004). Lipid peroxidation activity has been observed in pregnancy complications that involve poor placentation. For example, multiple studies reported that pre-eclamptic pregnancies have increased placental lipid peroxidation and serum circulating lipid peroxidation byproducts compared to normal pregnancies (Madazli et al., 2002; Atamer et al., 2005; Gupta et al., 2005). Studies have also implicated lipid peroxidation as a pathological feature of intrauterine growth restriction (Biri et al., 2007; Karowicz-Bilinska et al., 2007). The evidence presented in the current study suggests that increased ROS generation and lipid peroxidation are involved in the mechanism by which DCVC expo sure placental toxicity.

Multiple studies have reported that HTR-8/SVneo cells retain key functional characteristics of extravillous trophoblasts such as invasive and proliferative capacities which change in response to stress and oxygen availability, expression of adhesion molecules and a mesenchymal proteomic phenotype (Kilburn et al., 2000; Hannan et al., 2010; Liu et al., 2016a; Szklanna et al., 2017). In addition, several studies have also shown that HTR-8/SVneo cells express specific markers that uniquely identify extravillous trophoblasts including cytokeratin 7 (CK7) and α5β1 integrin, and when grown on matrigel, histocompatibility antigen, class I, G (HLA-G) (Irving et al., 1995; Kilburn et al., 2000; Khan et al., 2011; Takao et al., 2011). Overall, HTR-8/SVneo cells have proven useful in modeling first-trimester extravillous trophoblasts because primary first-trimester placental tissue tends to have limited availability for research purposes. Regardless, in vitro cultured cells lack cell signaling and tissue interactions that would otherwise be present in in vivo models. Thus, in vivo studies are needed to further validate our results. In addition, HTR-8/SVneo is an immortalized cell line. Even though HTR-8/SVneo were originally derived from first-trimester non-cancerous placental cells, the process of immortalization changes cells to allow them to be cultured for extended periods of time (Graham et al., 1993). Moreover, cell culture conditions may also be responsible for reported changes in the genetic and epigenetic profiles of HTR-8/SVneo cells (Bilban et al., 2010; Novakovic et al., 2011). For these reasons, the mechanisms of cytotoxicity of DCVC should be further investigated in primary first-trimester extravillous trophoblasts.

The current study includes DCVC concentrations selected for relevance to human exposure. Although the EPA’s maximum contaminant level (MCL) for drinking water is 5 μg/L or parts-per-billion of TCE (NTP, 2015), some people may regularly be exposed to TCE levels in drinking water that exceed the EPA MCL (ATSDR, 2016). Moreover, the Occupational Safety and Health Administration (OSHA) Permissible Exposure Limit (PEL) is 100 ppm averaged over an 8-hour work day (ATSDR, 2016). We suggest that the 10 μM and 20 μM DCVC concentrations used in our study are relevant to occupational exposures because the average peak concentration in blood of female volunteers exposed to 100 ppm TCE by inhalation for 4 h was 13.4 μM for the metabolic precursor to DCVC, S-(1,2-dichlorovinyl)glutathione (Lash et al., 1999). Moreover, the most recent human physiologically based pharmacokinetic models are in good agreement with human data of the latter study, though the need for further study was noted (EPA, 2011). Adding support that the concentrations used in our study are plausible for some occupational exposures, a recent study detected concentrations up to 229 ppm using personal aerosolized TCE exposure of 80 workers (29% women) (Walker et al., 2016). We measured a number of significant effects on HTR-8/SVneo cells at the 20 μM DCVC treatment level, including increased caspase 3, 7, 8 and 9 activation, increased malondialdehyde formation, and increased expression of multiple apoptosis and inflammatory-related genes, suggesting that DCVC stimulated apoptosis, lipid peroxidation, and pro -inflammatory responses at a concentration relevant for human exposure. Additionally, we included higher DCVC concentrations in our study that allowed us to characterize DCVC-induced cell death more completely. The concentrations of 50 μM and 100 μM DCVC are still within an order of magnitude of occupationally relevant concentrations, and similar to (Hassan et al., 2016) or less than those used in previously published in vitro studies, with multiple studies using concentrations up to 500 μM and 1 mM DCVC (van de Water et al., 1995; Chen et al., 2001; Lash et al., 2001; Lash et al., 2003; Xu et al., 2008).

In summary, we detected significant DCVC-induced cytotoxic effects in placental cells utilizing concentrations that encompass those detected of the DCVC precursor in human serum samples (Lash et al., 1999). We demonstrated that the TCE metabolite DCVC stimulated activation of both the intrinsic and extrinsic apoptotic signaling pathways, culminating in aberrant apoptosis in a first-trimester extravillous trophoblast cell line. In addition, we demonstrated that DCVC-induced lipid peroxidation-associated apoptosis. To our knowledge, our study is the first to report that DCVC induces apoptosis via two signaling pathways simultaneously and that cross talk occurs between the two pathways. Our findings contribute to the biological plausibility of DCVC-induced placental toxicity, indicating that further studies into matter are warranted.

Figure 7. Effect of antioxidant co-treatment on DCVC-stimulated caspase 3/7 activity.

Cells were untreated (media only control) for 12 h or exposed to 20 μM DCVC without or with 50 μM (±)α-tocopherol pre-treatment for 15 min followed by co -treatment for 12 h. Bars represent means±SEM. Data were analyzed by two-way ANOVA (interaction between DCVC and α-tocopherol treatments, P=0.0103), followed by Tukey’s multiple comparison of means. ^#Significantly different compared to control with no (±)α-tocopherol treatment (P=0.0016). *Significantly different compared to 20 μM DCVC with no (±)α-tocopherol treatment (P<0.0075). N=3 independent experiments. All experiments were performed in quadruplicate.

Highlights.

• DCVC induced apoptosis in a placental cell line in a concentration-dependent manner.

• DCVC activated both intrinsic and extrinsic apoptotic pathways via tBID cross talk.

• DCVC stimulated differential intrinsic and extrinsic pathway gene expression.

• Cellular lipid peroxidation byproduct malondyaldehyde levels were increased by DCVC.

• Antioxidant (±)-α-tocopherol attenuated DCVC-induced caspase activity.

Abbreviations

BAK1

BCL2-antagonist/killer 1

BAX

BCL2-associated X protein

BCL-2

B-cell CLL/lymphoma 2

BCL2A1

BCL2-related protein A1

BID

BH3 interacting domain death agonist

CASP1

Caspase 1, apoptosis-related cysteine peptidase (interleukin 1, beta, convertase)

CASP3

Caspase 3, apoptosis-related cysteine peptidase

CASP4

Caspase 4, apoptosis-related cysteine peptidase

CASP7

Caspase 7, apoptosis-related cysteine peptidase

CASP8

Caspase 8, apoptosis-related cysteine peptidase

CASP9

Caspase 9, apoptosis-related cysteine peptidase

DCVC

S-(1, 2-dichlorovinyl)-L-cysteine

DEVD

Asp-Glu-Val-Asp

DFFA

DNA fragmentation factor

45kDa

alpha polypeptide

FAS

Fas (TNF receptor superfamily, member 6)

FBS

fetal bovine serum

FITC

fluorescein isothiocyanate

GADD45A

Growth arrest and DNA-damage-inducible, alpha

HRK

Harakiri

BCL2

interacting protein (contains only BH3 domain)

LEHD

Leu-Glu-His-Asp

LETD

Leu-Glu-Thr-Asp

IGF1R

Insulin-like growth factor 1 receptor

MCL

maximum contaminant level

MDA

malondialdehyde

NFKB1

Nuclear factor of kappa light polypeptide gene enhancer in B-cells 1

RIPK2

Receptor-interacting serine-threonine kinase 2

ROS

reactive oxygen species

TBA

thiobarbituric

TBHP

tert-butyl hydroperoxide

tBID

truncated BH3 interacting domain death agonist

TCE

trichloroethylene

TP53

tumor protein p53

TRAF3

TNF receptor-associated factor 3

Appendix A. Supplementary data

Supplementary data to this article can be found online at: