Reduced bioenergetics and mitochondrial fragmentation in human primary cytotrophoblasts induced by an EGFR-targeting chemical mixture

Abstract

Exposures to complex environmental chemical mixtures during pregnancy reach and target the feto-placental unit. This study investigates the influence of environmental chemical mixtures on placental bioenergetics. Recognizing the essential role of the epidermal growth factor receptor (EGFR) in placental development and its role in stimulating glycolysis and mitochondrial respiration in trophoblast cells, we explored the effects of chemicals known to disrupt EGFR signaling on cellular energy production. Human primary cytotrophoblasts (hCTBs) and a first-trimester extravillous trophoblast cell line (HTR-8/SVneo) were exposed to a mixture of EGFR-interfering chemicals, including atrazine, bisphenol S, niclosamide, PCB-126, PCB-153, and trans-nonachlor. An RNA sequencing approach revealed that the mixture altered the transcriptional signature of genes involved in cellular energetics. Next, the impact of the mixture on cellular bioenergetics was evaluated using a combination of mitochondrial and glycolytic stress tests, ATP production, glucose consumption, lactate synthesis, and super-resolution imaging. The chemical mixture did not alter basal oxygen consumption but diminished the maximum respiratory capacity in a dose-dependent manner, indicating a disruption of mitochondrial function. The respiratory capacity and ATP production were increased by EGF, while the Chem-Mix reduced both EGF- and non-EGF-mediated oxygen consumption rate in hCTBs. A similar pattern was observed in the glycolytic medium acidification, with EGF increasing the acidification, and the Chem-Mix blocking EGF-induced glycolytic acidification. Furthermore, direct stochastic optical reconstruction microscopy (dSTORM) imaging demonstrated that the Chem-Mix led to a reduction of the mitochondrial network architecture, with findings supported by a decrease in the abundance of OPA1, a mitochondrial membrane GTPase involved in mitochondrial fusion. In conclusion, we demonstrated that a mixture of EGFR-disrupting chemicals alters mitochondrial remodeling, resulting in disturbed cellular bioenergetics, reducing the capacity of human cytotrophoblast cells to generate energy. Future studies should investigate the mechanism by which mitochondrial dynamics are disrupted and the pathological significance of these findings.

Graphical Abstract

Introduction

Humans are exposed daily to man-made chemicals and growing evidence supports the notion that many of these chemicals pose a risk to human health (Ghassabian et al., 2022). Particularly concerning are exposures that occur during pregnancy, a vulnerable period for maternal and fetal health (Gingrich et al., 2020). While effects of chemicals are primarily evaluated as a result of exposure to single compounds, humans are ubiquitously exposed to chemical mixtures (Sarigiannis and Hansen, 2012; Hernandez and Tsatsakis, 2017), including during pregnancy. Given that chemical mixtures can display different effects when compared to single chemical exposures (Amorim et al., 2012; Bunay et al., 2018; Kar et al., 2018), evaluating their chemical effects when in mixture is much warranted.

Serving as the interface between the fetus and the mother, the placenta plays a key role in fetal growth by regulating hormone synthesis and nutrient transport. As a dynamic and metabolically active organ with significant nutrient and energy requirements for its own growth, protein and steroid synthesis, and nutrient transfer to the fetus (Vaughan and Fowden, 2016), the placenta uses a range of substrates, including carbohydrates, amino acids, and lipids to meet its energy needs (Vaughan and Fowden, 2016). Notably, it has the highest rate of oxygen consumption than any other adult or fetal tissues (Hay, 1991). Information derived from sheep has demonstrated that of the O₂ consumed by the placenta, ~70% is used to generate ATP by mitochondrial oxidative phosphorylation (Carter, 2000). This indicates that placental mitochondrial function is the major source of energy towards the demand of placental and fetal development. In the human and the rodent placenta, both nutritional and hypoxic stresses can alter mitochondrial function. Specifically, changes in mitochondrial biogenesis, morphology, apoptosis, and abundance of electron transport complexes have been reported during common pregnancy stressors and disorders, including maternal diabetes, obesity, pre-eclampsia, caloric restriction, protein deprivation, and high-altitude hypoxia (Belkacemi et al., 2011; Colleoni et al., 2013; Mayeur et al., 2013; Rebelato et al., 2013; Mando et al., 2014; Chiaratti et al., 2015). While mitochondrial function can be affected by chemical exposures (Reddam et al., 2022), whether environmental chemicals alter mitochondrial bioenergetics in placenta trophoblast cells remains unexplored.

Growing evidence supports the impact of chemical exposures on growth factors with a key role in placental function. One of such factors is the epidermal growth factor (EGF) whose signaling through the EGF receptor (EGFR), is required for proper control of cytotrophoblast growth and differentiation during placental development in mice (Filla and Kaul, 1997; Dackor et al., 2007). Aside from tumorigenic tissues, the human placenta is among the tissues with the highest abundance of EGFR protein expression (Filla and Kaul, 1997; Proteinatlas, 2024). This translates to EGFR playing a key role in numerous placental cell processes, such as proliferation, fusion, and invasion (Fang et al., 2021; Ticiani et al., 2021; Ticiani et al., 2022). Notably, EGF can stimulate oxygen consumption in cytotrophoblasts (Kolahi et al., 2017) and mitochondrial biogenesis in intestinal stem cells (Zhang et al., 2022a).

Several high production volume chemicals - with very different chemical structures - can directly target EGFR by activating it (methoxychlor (Sauer et al., 2017)) or blocking it (atrazine, polychlorinated biphenyls [e.g.,: PCB-153], trans-nonachlor, niclosamide, and bisphenol S) (Hardesty et al., 2018; Ticiani et al., 2021; Ticiani et al., 2022). Of significance, organochlorine and organophosphate pesticides, such as trans-nonachlor, and the chlorinated herbicide atrazine, are some of the most common pesticides in the agricultural industry worldwide (2011; Vonberg et al., 2014), with farm workers at heightened exposure risk (Cragin et al., 2011; Lebov et al., 2016; Hoppin et al., 2017; Zuniga-Venegas et al., 2022). Despite being phased out, PCBs, used as plasticizers and in building materials, and trans-nonachlor are still detected in serum and the U.S. food supply (Eden et al., 2016; Mangum et al., 2016) and their combined exposure remains associated with adverse effect on overall mortality (Fry and Power, 2017). On the other hand, BPS is the leading bisphenol A replacement (ECHA and Agency, 2020) and niclosamide, an FDA-approved anti-helminthic drug listed as one of the 100 essential drugs in 2021 by the World Health Organization (WHO, 2021) is now being repurposed as a chemotherapeutic agent for several cancers (Burock et al., 2018; Schweizer et al., 2018; Luo et al., 2019; Parikh et al., 2021) and for COVID-19 management (Cairns et al., 2022). These compounds can accumulate in and transfer to the fetus through the placenta (e.g.: PCBs (Ando et al., 1985; Park et al., 2008; Boutot et al., 2021), atrazine (Lin et al., 2013), trans-nonachlor (Yin et al., 2019), and BPS (Grandin et al., 2018; Gingrich et al., 2019; Gingrich et al., 2021)). Of significance, gestational exposures to most of these chemicals have been linked to altered placental gene expression, enzyme activity, mitochondrial function, and oxidative stress and poor birth outcomes, including smaller birth weights (Benjaminov et al., 1992; Chevrier et al., 2011; Chiapella et al., 2013; Alonso et al., 2015; Rivero Osimani et al., 2016; Gingrich et al., 2018; Gingrich et al., 2020; Padmanabhan et al., 2021; Suwannakul et al., 2021). Such EGFR antagonizing chemicals can co-exist in the environment and thus, exposure to multiple of them can result in a “mixed body burden” (Ribeiro et al., 2017). Yet, their effects in mixture and whether their effects are mediated by disruptions in EGFR remain unexplored.

Chemical mixtures can result in exacerbated effects (Silva et al., 2002; Rajapakse et al., 2004; Kortenkamp, 2007; Orton et al., 2012; Kortenkamp, 2014; Martin et al., 2020). Yet, underlying molecular mechanisms driving these effects remain largely unexplored in mixture studies. Previous studies using chemical mixtures sharing a common pathway (androgen, estrogen, and PPARγ receptor) have made seminal contributions (Silva et al., 2002; Rajapakse et al., 2004; Orton et al., 2012; Watt et al., 2016; Schlezinger et al., 2020; Nielsen et al., 2022). Leading efforts towards this goal, we have recently demonstrated that the combination of chemicals known to block EGFR activation (PCB-126, PCB-153, atrazine, trans-nonachlor, niclosamide, and BPS), results in a mixture that competes with EGF for EGFR binding, interferes with EGFR activation, and reduces trophoblast cell proliferation in an EGF-dependent manner (Waye et al., 2024). In this study, we aimed to continue shedding light into how chemical mixtures sharing a common pathway (EGFR), can elicit stronger effects compared to single chemical exposures in placental cells in the context of mitochondrial bioenergetics.

Since evaluating mixtures of chemicals at environmentally relevant doses is needed to assess their cumulative, additive, or synergistic effects in the context on placental function, we hypothesized that a mixture of chemicals reported to interfere with EGFR phosphorylation (Hardesty et al., 2018; Waye et al., 2024) disrupts mitochondrial morphology and bioenergetics in trophoblast cells. To test this, we used a combination of approaches, including RNA sequencing, metabolic stress tests, gene and protein expression, metabolite concentration assays, and mitochondrial imaging in primary human cytotrophoblasts and/or an extravillous trophoblast cell line.

Materials and Methods

Human primary cytotrophoblast cell isolation, purification, and culture

Healthy placentas derived from elective C-section term pregnancies were collected after written informed consent and IRB approval (IRB # 2020-0903) for isolation of primary human cytotrophoblast (hCTB) cells as previously described (Petroff et al., 2006; Ticiani et al., 2021). Subjects were enrolled considering the following exclusion criteria: fetal developmental abnormalities, multiple pregnancies, maternal diagnosis of HIV, hepatitis B or C, polycystic ovary syndrome, and/or congenital adrenal hyperplasia. In brief, within an hour of delivery, placental tissue was processed for hCTB isolation. First, placental tissue was rinsed with 0.9 % NaCl and then transferred to a 150-mm Petri dish. After gently mincing with scissors, the villous tissue was dissociated with an enzyme digestion solution (Hank’s Balanced Salt Solution, 25 mM HEPES, 2.5% trypsin, and ~300 Kunitz U/ml DNase (Cat#: 15090-046, Gibco, Toronto, ON, Canada)). The resulting cell suspension was then collected, layered on fetal bovine serum (FBS, Cat#: 35-010-CV, Corning, Woodland, CA, USA), and centrifuged at 1,000 x g for 15 min. The cell pellet was resuspended in Iscove’s Modified Dulbecco’s Medium (Cat#: I3390, MilliporeSigma, Saint Louis, MO, USA). The cell suspension was filtered through a 100 μm nylon cell strainer, followed by a Percoll gradient centrifugation at 1,200 x g for 20 min in a swinging bucket rotor without brake at 4 °C. The visible cell band between 30 and 50% Percoll was collected and resuspended in Iscove’s modified Dulbecco’s (IMD) medium supplemented with 10% FBS. Cytotrophoblast cell purity was evaluated with cytokeratin-7 by flow cytometry as a previously described (Petroff et al., 2006; Ticiani et al., 2021) which yields ∼98% cytotrophoblast cell purity. Cells were then stored in liquid nitrogen until needed. hCTBs were cultured overnight to allow cell attachment and then washed the following day with IMD medium to remove unattached cells. hCTBs were supplemented with 10% of FBS, 2 mM L-glutamine, 10 mM HEPES, 100 IU/ml penicillin, and 100 μg/ml streptomycin. Cells were incubated at 37 °C and 5% CO₂ and cultured for up to 96 h.

Extravillous trophoblast cell culture

The first trimester human extravillous trophoblast cell line HTR-8/SVneo was cultured in basic cell culture medium consisting of Dulbecco’s modified Eagle’s medium/F12 medium (Cat#: 124000-024, Millipore Sigma, Saint Louis, MO, USA) supplemented with 10% of fetal bovine serum (FBS), 2 mM L-glutamine, 10 mM HEPES, 100 IU/ml penicillin, and 100 μg/ml streptomycin. Cells were incubated at 37 °C and 5 % CO₂.

Chemical exposure

Chemicals used in this study were: 2,2’,4,4’,5,5’-hexachlorobiphenyl (PCB-153, Cat#: C-153N, Accustandard, New Heaven, CT, USA), 3,3’,4,4’,5-pentachlorobiphenyl (PCB-126, Cat#: C-126N, Accustandard, New Heaven, CT, USA), 4,4’-sulfonyldiphenol (BPS, Cat#: 80-09-11, Acros Organics, Geel, Belgium), atrazine (ATZ, Cat#: P-005N, Accustandard, New Heaven, CT, USA ), niclosamide (NIC, Cat#: N3510, Milliporesigma, Saint Louis, MO, USA), and trans-nonachlor (TNC, Cat#: 36846, Milliporesigma, Saint Louis, MO, USA). Di-methyl sulfoxide (DMSO, Cat#: BP231-100, Thermo-Fisher, Rockford, IL, USA), human EGF (Cat#: E9644, Sigma Aldrich, St. Louis, MO, USA), and afatinib (Cat#: G-7208, Advanced ChemBlocks Inc, Hayward, CA, USA) were used as controls in the study. DMSO was used as the vehicle control and added to a final concentration of 0.1 % in all exposure groups.

HTR-8/SVneo and hCTBs cells were cultured in basic or IMD medium, respectively, and exposed to one of four different culture conditions: 1) control (C; 0.1 % DMSO), 2) Chem-Mix (ATZ, BPS, NIC, PCB-126, PCB-153, and TNC at doses of 1, 10, or 100 ng/ml each), 3) EGF (30 ng/ml), or 4) Chem-Mix + EGF. After 24 h, cells and medium were harvested for protein and metabolites (ATP, glucose, and lactate) quantification and RNA extraction. For conducting mitochondrial stress assays (see below), hCTBs cells were initially exposed to the control (C; 0.1 % DMSO) or the Chem-Mix at 1, 10, and 100 ng/ml of each of the chemicals. Based on the results from this dose-response experiment, the concentration leading to the maximum effect on maximal oxygen consumption rate was chosen to further probe the effects of the Chem-Mix. To evaluate individual chemical effect on mitochondrial stress, hCTB cells were exposed to each one of the chemicals (ATZ, BPS, NIC, PCB-126, PCB-153, or TNC) at 100 ng/ml in combination with EGF (30 ng/ml). Chemical doses used in the study are based on environmentally relevant doses (Dearth and Hites, 1991; Newsome and Ryan, 1999; Perry et al., 2000; Rylander et al., 2005; Costopoulou et al., 2006; Diamanti-Kandarakis et al., 2009; Zhu et al., 2009; Ritter et al., 2011; Mendas et al., 2012; Zoeller et al., 2012; 2014; National Academy of Medicine, 2014; Megersa, 2015; Faroon and Ruiz, 2016; Namulanda et al., 2017; Schweizer et al., 2018; Wang et al., 2018b), which have been previously summarized (Waye et al., 2024).

Viability assay

Cytotoxicity of chemicals to hCTBs (n = 5 primary cell cultures) and HTR-8/SVneo cells (n = 6 replicates) was tested using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay as previously described (Pu et al., 2019). Cells were cultured in basic cell culture medium in 96-well plates and exposed to either 0.1 % DMSO (control), or to the mixture of chemical (Chem-Mix) at doses of 0.01, 0.1, 1, 10, 100, or 1,000 ng/ml of each chemical for 24 h. Thereafter, the exposure medium was replaced with 100 μl of phenol red-free MTT working solution (50 μg/ml) and incubated for 3 h. The MTT working solution was then discarded and replaced with 100 μl of DMSO per well. Plates were vortexed and absorbances of each treatment were quantified at 570 nm in a microplate reader (Cytation 1; BioTek Instruments, Winooski, VT, USA).

RNA sequencing and bioinformatics analysis.

An RNA sequencing was performed in hCTB cells (n = 6 primary cell cultures) exposed for 24 h to one of the two treatment groups (control and Chem-Mix) as detailed in the Chemical exposure section. Total RNA was extracted using a RNeasy Mini kit (Qiagen, Hilden, Germany) following the manufacturer’s protocol as previously reported (Pu et al., 2022). RNA quality and concentration were measured by Nanodrop (Thermo Fisher Scientific, Wilmington, NC, USA). Samples with a concentration of ≥ 500 ng/μl and RIN value ≥ 8.0 were used for the library preparation and sequencing. Messenger RNA (mRNA) was extracted using oligo(dT) magnetic beads and fragmented with fragmentation master mix. Subsequently, complementary DNA (cDNA) was synthesized, and mRNA templates were removed. Illumina adaptors and indices were added to the cDNA, followed by the preparation of cDNA libraries according to the Illumina Stranded mRNA Prep protocol. The libraries underwent amplification and analysis using the Agilent TapeStation 4200. After pooling, the libraries were sequenced with paired-end reads (2x150 bp) on a NovaSeq-X flow cell, generating 30 million reads per sample. Transcriptome sequencing and data processing were performed at the Genomics Facility of the University of Chicago.

Upon receipt of the sequencing file, demultiplexing was conducted to segregate the sequenced reads into individual files corresponding to each sample within the sequencing run. Subsequently, cutadapt was utilized to eliminate adapters and low-quality reads. Clean reads were then aligned to the human reference genome (GRCh38/hg38) using Star software (Dobin et al., 2013), and mapped reads enumerated using featureCounts. The resulting gene read counts underwent normalization to RPKM, a metric representing gene expression levels in terms of reads per kilobase of transcript per million mapped reads. This normalization resulted in a meaningful comparison of relative gene expression levels across samples. DESeq2 was employed for differential gene expression (DEGs) analysis between the control and the Chem-Mix, quantifying differences in gene expression using logarithmically transformed fold change values, denoted as either log base 2 (log2FC) or log base 10 (log10FC). Statistical significance was determined using a two-tailed Wald test, which evaluates the significance of coefficients in statistical models (Love et al., 2014; Li et al., 2020). To further analyze the data, Gene Set Enrichment Analysis (GSEA) was carried out using the “clusterProfiler” R package, ranking all genes based on their log2FC values, with a cut off set to 0.5, derived from the differential expression analysis. Gene Ontology (GO) classifications were analyzed using PANTHER GO-Slim tool (Mi et al., 2019). Visual representation of GSEA findings was provided using the two directional bar plot.

Mitochondrial stress assay

Oxygen consumption rate (OCR) was measured using the Seahorse XFe24 FluxPak mini (Cat# 102342-100, Agilent Technologies, Cedar Creek, TX, USA) with the Agilent Seahorse XFe24 Analyzer. hCTBs (500,000/well) and HTR-8/SVneo cells (40,000/well) were plated into Seahorse XF24 24-well microplates and exposed for 24 h to one of the four treatment groups (control, Chem-Mix, EGF, or Chem-Mix + EGF) as detailed in the Chemical exposure section. After 24 h of exposure, the growth medium was removed and replaced with the pre-warmed Seahorse XF assay medium (Cat#103575-100, Agilent, Santa Clara, CA, USA) supplemented with 1 mM of pyruvate, 2 mM of glutamine, and 25 μM of glucose. For the mitochondrial stress assay the following reagents were added at 25 min intervals: 1) ATP synthase inhibitor oligomycin (1 μM; Cat#: 103015-100, Agilent Technologies, Cedar Creek, TX, USA), 2) mitochondrial uncoupling agent carbonyl cyanide p-trifluoromethoxyphenyl-hydrazone (FCCP; 4 μM; Cat#: 103015-100, Agilent Technologies, Cedar Creek, TX, USA), and 3) mixture of the mitochondrial complex I inhibitor rotenone (0.5 μM; Cat#: 103015-100, Agilent Technologies, Cedar Creek, TX, USA) and the mitochondrial complex III inhibitor antimycin A (0.5 μM; Cat#: 103015-100, Agilent Technologies, Cedar Creek, TX, USA). OCR was recorded for three cycles following each timed injection. After the assay, the cells were fixed with 10% formalin and nuclei were stained with DAPI (1:1,000) for cell quantification which was used to normalize OCR results by cell number. The assay was performed at least in triplicate in HTR-8/SVneo or in at least three different hCTB primary cell cultures.

The glycolytic ATP (glycoATP) production rate was measured based on the following stoichiometric equation: glucose + 2 ADP + 2 Pi = 2 lactate + 2 ATP + 2 H₂O + 2 H⁺. The rate of proton production due to glycolysis (glycolysis-associated proton efflux rate: GlycoPER) was considered equal to the glycoATP production rate as for each molecule glucose converted to lactate, two molecules of ATP and two protons are released. GlycoATP was calculated from the extracellular acidification rate (ECAR) measured by the Seahorse XFe 24 flux analyzer.

Mitochondrial ATP (mitoATP) production rate was calculated from the OCR before and after the addition of oligomycin (inhibition of ATP synthase, oligo OCR) using the following formula: OCR ATP (pmol of O₂/min) = basal OCR - oligo OCR. Basal OCR and oligo OCR were calculated after deducting the OCR due to non-mitochondrial respiration (after the addition of antimycin A and rotenone). MitoATP production rate was calculated from the OCR rate using the following equations: mitoATP production rate (pmol of ATP/min) = OCR rate (pmol of O₂/min) × 2 (number of oxygen atoms) × 2.75 (ATP / O₂ ratio; P/O ratio). The P/O ratio indicates the generation of ATP molecules for a single O₂ atom consumed (e.g.: 2.75 calculated under ideal Seahorse assay conditions).

Glycolytic rate assay

hCTBs or HTR-8/SVneo cells were seeded 500,000 or 40,000 cells/well, respectively, into a Seahorse XF24 24-well microplate and exposed for 24 h to one of the four treatment groups (control, Chem-Mix, EGF, or Chem-Mix + EGF) as detailed in the Chemical exposure section. Basal glycolytic rate and compensatory glycolytic rates were obtained and measured as previously published (Romero et al., 2017) and as per manufacturer’s instructions. After 24 h of exposure, growth medium was removed and replaced with pre-warmed Seahorse XF assay medium (Cat#: 103575-100, Agilent, Santa Clara, CA, USA) supplemented with 1 mM of pyruvate, 2 mM of glutamine. For the glycolytic rate assay the following reagents were added every 25 min: 1) glucose (10 μM; Cat#: 103020-100, Agilent Technologies, Cedar Creek, TX, USA), 2) mixture of the mitochondrial complex I inhibitor rotenone (0.5 μM; Cat#: 103020-100, Agilent Technologies, Cedar Creek, TX, USA) and the mitochondrial complex III inhibitor antimycin A (0.5 μM; Cat#: 103020-100, Agilent Technologies, Cedar Creek, TX, USA), and 3) glycolysis inhibitor 2- deoxy-d-glucose (2-DG, 0.5 μM; Cat#: 103020-100, Agilent Technologies, Cedar Creek, TX, USA). Extracellular acidification rate (ECAR) as a product of glycolysis was recorded for three cycles following each timed injection. After the assay, the cells were fixed with 10% formalin and nuclei were stained with DAPI (1:1,000) for cell quantification which was used to normalize ECAR results by cell number. The assay was performed at least in triplicate in HTR-8/SVneo cells or in five different hCTB primary cell cultures.

Total ATP concentration

ATP production in hCTB and HTR-8/SVneo cells was measured using a CellTiter-Glo luminescent cell viability assay (Cat#: G8091, Promega, Madison, WI, USA) following the manufacturer’s instructions. Cells were seeded into clear 96-well plates at 2.5 × 10³ cells per well and exposed to one of the four treatment groups (control, Chem-Mix, EGF, and Chem-Mix + EGF) as detailed in the Chemical exposure section. After 24 h of exposure, cells were incubated with 100 μl of CellTiter-Glo reagent per well for 10 min and luminescence recorded microplate reader (Cytation 1; BioTek Instruments, Winooski, VT, USA). The assay was performed at least in triplicate in HTR-8/SVneo cells or in three different hCTB primary cell cultures.

Quantitative RT-PCR

hCTBs exposed to one of the four treatment groups (control, Chem-Mix, EGF, and Chem-Mix + EGF) were collected for total RNA extraction using an RNeasy Mini kit (Cat# 74104, Qiagen, Hilden, Germany) following the manufacturer’s protocol. A total of 1 μg RNA (A260/A280: 2.0 ± 0.05, RNA concentration: 150 ± 50 ng/μl) was reverse transcribed into complementary DNA using a High-Capacity cDNA Reverse Transcription Kit (Promega, Madison, WI, USA) in 20 μl reaction volumes as previously described (Pascuali et al., 2024). Quantitative real time PCR (CFX Opus 96 Real-Time PCR Detection System, Bio-Rad, Hercules, CA, USA) was performed to examine the mRNA expression of the following genes: (LDHA, HK2, PKM2, and PFK2). Primer list and sequences are provided in Supplemental Table S1. All experiments and qRT-PCR were run in triplicate. The amplification reaction included template denaturation and polymerase activation at 95 °C for 30 s, followed by 40 cycles of denaturation at 95 °C for 15 s, annealing and extension at 60 °C for 30 s. Melt curve analyses were performed for all genes, and the specificity as well as integrity of the PCR products was confirmed by the presence of a single peak. mRNA expression levels encoding the indicated genes were calculated using the ΔΔCT method, normalized against GAPDH, and presented as relative fold change to the control.

Western blotting

Protein extraction was performed in HTR-8/SVneo cells as described previously (Pu et al., 2019). In brief, RIPA lysis buffer (Cat#: N653, VWR Life Science, San Francisco, CA, USA) supplemented with 20% 1M NaF, 1mM Na₃VO₄ and 1% protease inhibitor cocktail was utilized for cell lysis. Protein concentration was determined using a Pierce BCA protein assay kit (Cat#: 23225, Thermo Fisher, Rockford, IL, USA). Twenty micrograms of protein per sample from the extracted cell lysates were subjected to electrophoresis on a 10% SDS-polyacrylamide gel (120 V for 60 min). Protein was then transferred from the gel to a nitrocellulose membrane (200 mA for 90 min) and subjected to western blotting. Membranes were blocked with 5% non-fat dry milk in tris-buffered saline (TBS) containing 0.03% tween-20 (block solution) and incubated with primary antibodies diluted in block overnight at 4 °C. Primary antibodies used were: anti-phospho- pyruvate kinase muscle isozyme M2 (PKM2; 1:1,000; Cat#: 11456, Sabbiotech, Greenbelt, Maryland, USA), anti-metalloendopeptidase OMA1 (OMA1; 1:1,000; Cat#: 95473, Cell Signaling, Danvers, MA, USA) and anti-optic atrophy 1 (OPA1; 1:1,000; Cat#: 67589, Cell Signaling, Danvers, MA, USA), and anti-β-actin (1:5,000; Cat#: A1978, MilliporeSigma, Saint Louis, MO, USA). After three washes with TBS containing 0.03% tween-20, membranes were incubated with goat anti-mouse HRP-conjugated (Cat#:115-005-003, Jackson Immunoresearch, West Grove, PA, USA) and goat anti-rabbit HRP-conjugated (Cat#: 111-005-003, respectively; Jackson Immunoresearch, West Grove, PA, USA) secondary antibodies, respectively. Secondary antibodies were diluted 1:5,000 in block solution for 1 h at room temperature in the dark. WesternBright ECL (Cat#: K12045, Advansta, Menlo Park, CA, USA) was used for enhanced chemiluminescence and visualized using BIO-RAD ChemiDoc (Bio-Rad, Hercules, CA, USA). Quantification of band intensities was performed using ImageJ software. Differences in protein loading were accounted for by normalizing the target protein band by the control β-actin band for each sample. The assay was performed in triplicate for each of the proteins of interest.

Glucose Concentration Analysis

To measure trophoblast cell glucose production the Glucose-Glo^™ Assay (Promega, Madison, WI, USA) was used. hCTBs and HTR-8/SVneo cells were seeded in 6-well plates (50,000 cells per well) for 24 h. Thereafter, the medium was replaced with fresh medium containing one of the four treatments (control, Chem-Mix, EGF, or Chem-Mix + EGF) as detailed in the Chemical exposure section. The cells were incubated for an additional 24 h. Afterwards, 2 μl of cell culture medium from each well was diluted in 298 μl of PBS. The glucose concentration in the media was determined based on a linear standard curve, which was prepared in parallel. After 60 min of incubating diluted culture medium (50 μl per well) with a glucose detection reagent (50 μl per well), luminometric analysis of the glucose concentration was performed on a BioTek Cytation 1 imaging reader (Bio-Tek Instruments, Winooski, USA). The assay was performed at least in triplicate in HTR-8/SVneo cells or in five different hCTB primary cell cultures.

Lactate Concentration Analysis

The analysis of the lactate concentration in the culture medium was performed using the Lactate-Glo^™ Assay (Promega, Madison, WI, USA). hCTBs and HTR-8/SVneo cells were seeded in 12-well plates (50,000 cells per well) for 24 h. Then, the cells were exposed for 24 h to one of the four treatments (control, Chem-Mix, EGF, or Chem-Mix + EGF) as detailed in the Chemical exposure section. Afterwards, 2 μl of medium from each well was diluted in 198 μl of DPBS. The lactate concentration in the media was determined based on a linear standard curve, which was prepared in parallel. After 60 min of incubating diluted culture medium (50 μl per well) with a glucose detection reagent (50 μl per well), luminometric analysis of the glucose concentration was performed on a microplate reader (Cytation 1; BioTek Instruments, Winooski, VT, USA). The assay was performed at least in triplicate in HTR-8/SVneo or in five different hCTB primary cell cultures.

Immunofluorescence staining

Mitochondrial structure in HTR-8/SVneo cells was analyzed using immunofluorescence staining. HTR-8/SVneo (2,000 cells per chamber) cells were seeded in an 8-well-chambered coverglass system (Cat#:C8-1-N; Cellvis, Mountain View, CA, USA) to reach 30% confluency overnight, and subsequently exposed to two of the four treatments (control, Chem-Mix, EGF, or Chem-Mix + EGF) or (control, afatinib, EGF, or afatinib + EGF) for 24 hours. Cells were fixed with 4% formalin in 1x DPBS for 30 min, then permeabilized with Triton X-100 (0.2%) in 1x DPBS for 15 min and blocked using a blocking buffer (1x DPBS supplemented with 3% BSA and 0.05% Triton X-100 0.05%) for 90 min. Thereafter, the cells were incubated with the primary antibody TOMM20 (Cat#: MA5-32148, Thermo Fisher Scientific, Rockford, IL, USA) in blocking buffer at 5 μg/ml and incubated for 60 min in a humid chamber. Cells were then washed with 1x DPBS four times. After washing, cells were incubated with the secondary antibody Alexa Fluor 647 (Cat#: A27040, Thermo Fisher Scientific, Rockford, IL, USA) in blocking buffer at 2 μg/ml for 30 min in a dark, humid chamber. After incubation, the cells were washed with 1x DPBS four times. All samples were maintained in DPBS.

Before dSTORM imaging, samples were incubated with gold colloids for 10 min (1:5 dilution in Milli-Q water; Cat#: 15711-20, Ted Pella Inc, Redding, CA, USA) to correct for the lateral stage drift during image acquisition. After incubation, samples were washed with 1x DPBS two times. The solution was replaced with STORM buffer consisting of GLOX solution (7 μl), 2-mercaptoethanol (7 μl), and Buffer B (690 μl). The composition of GLOX solution, Buffer A and Buffer B is as follows. GLOX solution: glucose oxidase (14 mg), catalase (17 mg/ μl) in Buffer A (200 μl). Buffer A: Tris (10 mM, pH 8.0), sodium chloride (50 mM). Buffer B: Tris (50 mM, pH 8.0), sodium chloride (10 mM), glucose (10 %).

Microscopy and super-resolution image reconstruction

Imaging experiments were performed using an inverted microscope (Nikon Instruments, Eclipse Ti2E). A 100x/1.49 oil objective was used with a 1.5x external magnifier, using a Prime 95B sCMOS camera at 16-bit. Immunofluorescence images of treated and non-treated samples were acquired using 640 nm excitation laser at a power density of 18 W/cm2 (measured after the objective) and 50 ms exposure time. dSTORM movies with 40,000 - 60,000 frames were acquired using 640 nm excitation laser at a power density of 1.1 kW/cm2 and 30 ms exposure time. A 405 nm laser was gradually increased from 0–32 W/cm2 to maintain a convenient density of fluorescent molecules. dSTORM image reconstruction was performed using the open-source ImageJ plug-in ThunderSTORM (Ovesny et al., 2014). Raw images were analyzed in the ThunderSTORM plug-in with the following camera settings: pixel size 147 nm, photoelectrons per A/D count 0.98, and base level of 100. Drift correction was performed using fiducials (immobilized gold colloid) that were present during the entire acquisition. Events with a sigma value greater than 200 appeared to produce a diffusive appearance in the 2D histogram and were removed in the post-processing. dSTORM images were visualized at a pixel size of 20 nm in the reconstructed image. The visualization method of “Normalized Gaussian” at a forced lateral uncertainty of 20 nm and LUT of “Yellow Hot” were used for the visualization. Analysis of 2D and 3D skeletons images of the mitochondria were performed using the “AnalyzeSkeleton” plug-in of the Fiji/ImageJ software (Arganda-Carreras 2010) and quantified as cell structure number and length.

Statistical analysis

All data are presented as mean ± SEM comparing groups by using a generalized linear model (MIXED procedure) after evaluation for normality. The model used included treatment groups and run as fixed effects. Significance was set at P < 0.05. Data were tested for normality by the Anderson-Darling normality test. LSMEANS were used to adjust the means and to compare treatments. All data were analyzed using SAS version 9.4.

Results

Chem-Mix resulted in dysregulation of transcripts regulating cellular energetics

Using unbiased RNA-sequencing analysis, hCTB cells exposed to Chem-Mix revealed significant alterations in gene expression patterns in comparison to the control group. Comparison of gene expression levels in Chem-Mix-treated samples identified 3,692 significantly altered genes, comprising 2,424 downregulated and 1,288 upregulated genes (P < 0.05 with a fold change >1) (Figure 1A). Gene ontology (GO) analysis revealed a dysregulation in signaling pathways associated with metabolic processes (Figure 1B). GSEA-KEGG analysis unveiled dysregulation in several biochemical pathways involved in energy production within the mitochondria, including carbon metabolism, glycolysis/gluconeogenesis, and cellular senescence (Figure 1C). Specific transcripts associated with these pathways are detailed in Table S2. The functional implications of the discovered pathways relative to cellular bioenergetics were further explored by functional assays.

Figure 1.

Effect of Chem-Mix exposure on human cytotrophoblast cells (hCTB) on gene expression profile. (A) Volcano plots illustrating the differentially expressed genes (DEGs) between control (C; 0.1 % DMSO) and Chem-Mix-treated (Chem-Mix at 100 ng/ml) groups following 24 h of exposure. Statistical analysis was conducted using two-tailed Wald tests. The dashed line represents a false discovery rate cutoff of 0.05, with DEGs exhibiting fold change >1 highlighted. Representative examples of highly enriched transcripts within the Metabolic Process pathway (see B) are noted in boxes. (B) Gene ontology analysis using the PANTHER GO-Slim tool was performed on DEGs between control (C; 0.1 % DMSO) and Chem-Mix (Chem-Mix at 100 ng/mL) after 24 h of exposure. (C) GSEA-KEGG analysis demonstrating the enrichment of pathways related to mitochondrial energetics in response to the Chem-Mix treatment (P < 0.05, absolute log fold change > 0.5). N = 6 primary hCTB cell cultures per group.

Chem-Mix reduced mitochondrial OCR in hCTBs in a dose-response manner

To investigate whether the Chem-Mix could affect mitochondrial respiration, primary hCTBs cells were exposed to doses of 1, 10, and 100 ng/ml for 24 h. The Chem-Mix did not affect the basal OCR (Figure 2). However, in the Chem-Mix group, the addition of the uncoupler of mitochondrial oxidative phosphorylation FCCP reduced the maximum respiratory capacity in a dose dependent manner (P < 0.05). None of the doses tested induced cytotoxicity in any of the primary hCTB cell cultures as tested by an MTT assay (Figure S1). Based on the results from this dose-response experiment, the concentration leading to the maximum effect on maximal oxygen consumption rate was chosen for all the following experiments to explore mechanistic insights of the reduced OCR.

Figure 2.

Effect of Chem-Mix exposure on human cytotrophoblast cells (hCTB) mitochondrial respiratory profile. Oxygen consumption rate (OCR) measurements were acquired over time (min) using an extracellular flux analyzer (Seahorse XF Analyzer). A) The mitochondrial stress test was conducted to resolve bioenergetic parameters (mean ± SEM) by adding the ATP synthase inhibitor oligomycin (OM; 1 μM), the mitochondrial uncoupler carbonyl cyanide p-trifluoromethoxyphenyl-hydrazone (FCCP; 4 μM), for maximal OCR calculation, and a mixture of 0.5 μM rotenone and 0.5 μM antimycin A (Rot/AA). The mitochondrial test was performed after 24 h exposure to: 1) control (C; 0.1 % DMSO), 2) Chem-Mix 1 (M1; Chem-Mix at 1 ng/ml), 3) Chem-Mix 10 (M10; Chem-Mix at 10 ng/ml), and 4) Chem-Mix 100 (M100; Chem-Mix at 100 ng/ml). B) Quantification of basal mitochondrial respiration (left), maximal respiration capacity (middle), and non-mitochondrial respiration (right). N = 4 primary hCTB cell cultures per group. Least squares means were used to compare treatments. Different letters denote statistical differences among treatment groups at P < 0.05.

Chem-Mix enhanced reduction on mitochondrial oxygen consumption in cytotrophoblast cells compared to single chemicals

To investigate if any of the chemicals was the driver of the lower OCR induced by the Chem-Mix, eight primary hCTBs cell cultures were exposed to each one of the chemicals at 100 ng/ml in combination with EGF at 30 ng/ml for 24 h. OCR was measured using a Seahorse XFe 24 flux analyzer by a mitochondria stress test protocol. None of the chemicals alone were able to significantly reduce basal or maximal respiratory capacity (Figure 3A and 3B) compared to the EGF group. Niclosamide tended to reduce OCR (P = 0.06). The OCR profiles for each one of the eight primary cell cultures are shown in Figure S2.

Figure 3.

Effect of Chem-Mix and single chemical exposure on human cytotrophoblast cells (hCTB) mitochondrial respiratory profile. Oxygen consumption rate (OCR) measurements were acquired over time (min) using an extracellular flux analyzer (Seahorse XF Analyzer). A) The mitochondrial stress test was conducted to resolve bioenergetic parameters (mean ± SEM) by adding the ATP synthase inhibitor oligomycin (OM; 1 μM), the mitochondrial uncoupler carbonyl cyanide p-trifluoromethoxyphenyl-hydrazone (FCCP; 4 μM), for maximal OCR calculation, and a mixture of 0.5 μM rotenone and 0.5 μM antimycin A (Rot/AA). The mitochondrial test was performed after 24 h exposure to: 1) control (C; 0.1 % DMSO), 2) Chem-Mix (Mix; Chem-Mix at 100 ng/ml), 3) epidermal growth factor (EGF; 30 ng/ml), 4) Chem-Mix +EGF, 5) PCB-126 (100 ng/ml) + EGF, 6) PCB-153 (100 ng/ml) + EGF, 7) atrazine (ATZ; 100 ng/ml) + EGF, 8) niclosamide (NIC; 100 ng/ml) + EGF, 9) trans-nonachlor (TNC; 100 ng/ml) + EGF, and 10) bisphenol S (BPS; 100 ng/ml) + EGF. B) Quantification of basal mitochondrial respiration (left), maximal respiration capacity (middle), and non-mitochondrial respiration (right). N = 8 primary hCTB cell cultures per group. Least squares means were used to compare treatments. Different letters denote statistical differences among treatment groups at P < 0.05. The OCR profiles for each of the 8 primary cell cultures tested per exposure group are shown in Figure S3.

Chem-Mix impact on OCR and ATP production in trophoblast cells

The effect of the Chem-Mix on OCR, ATP production, and whether the Chem-Mix would affect the EGFR induced mitochondrial respiration was evaluated. Both, primary hCTBs and HTR-8/SVneo cells were exposed to 100 ng/ml of the Chem-Mix with or without EGF for 24 h. The exposure to the Chem-Mix reduced OCR in HTR-8/SVneo, but not hCTBs. In hCTBs cells, EGF exposure led to an increase in the basal and maximal respiratory capacity and in non-mitochondrial respiration (Figure 4A and 4C; P < 0.05). HTR-8/SVneo cells, however, were not responsive to EGF regarding mitochondrial respiration (Figure 4B and 4D). In both cell types, the combination of Chem-Mix + EGF had lower OCR when compared to the EGF group. The ATP production was estimated using data from the oxygen consumption and medium acidification measured using a Seahorse XFe 24 flux analyzer (Figure 4E and 4F) and was also measured directly using a CellTiter-glo luminescent cell viability assay kit (Figure 4G and 4H). The mitochondrial ATP and total ATP production was reduced by the Chem-Mix alone or when co-exposed EGF (P < 0.05) in both cell types, but no changes were observed in glycolytic ATP induced by Chem-Mix.

Figure 4.

Effect of Chem-Mix exposure on human cytotrophoblast cells (hCTB) and extravillous trophoblast cells (HTR-8/SVneo) mitochondrial respiratory profile and ATP production. A-B) The mitochondrial stress test was conducted to resolve bioenergetic parameters (mean ± SEM) by adding the ATP synthase inhibitor oligomycin (OM; 1 μM), the mitochondrial uncoupler carbonyl cyanide p-trifluoromethoxyphenyl-hydrazone (FCCP; 4 μM), for maximal OCR calculation, and a mixture of 0.5 μM rotenone and 0.5 μM antimycin A (Rot/AA). The mitochondrial test was performed after 24 h exposure to: 1) control (C; 0.1 % DMSO), 2) chem-Mix (Mix; Chem-Mix at 100 ng/mL), 3) human epidermal growth factor (EGF; 30 ng/mL), and 4) Mix + EGF. C-D) Quantification of basal mitochondrial respiration (left), maximal respiration capacity (middle), and non-mitochondrial respiration (right). E-F) Calculated mitochondrial ATP (mitoATP; pmol/min) and glycolytic ATP (glycoATP; pmol/min) production rate derived from mitochondrial stress test (see A for assay details and text for calculation details). G-H) Total ATP concentrations measured using a CellTiter-Glo luminescent cell viability assay after exposure to the same treatment groups as described in A. Box plots represent median flanked by first and third quartiles and whiskers represent minimum and maximum. For all endpoints (A-H), N = 4 primary hCTB cell cultures/group or 3 replicates/group of the HTR-8/SVneo cells were used. Least squares means were used to compare treatments. Different letters denote statistical differences among treatment groups at P < 0.05.

Chem-Mix affected the abundance of mitochondrial fusogenic proteins

To test whether the Chem-Mix-induced effects on OCR and ATP were driven by changes in mitochondrial membrane proteins, their abundances was evaluated. HTR-8/SVneo cells were exposed to the Chem-Mix in the presence or absence of EGF. The Chem-Mix did not affect OMA1 (P = 0.83), but reduced OPA1 (P = 0.01) abundance compared to the control group. EGF had no effect in either protein compared to the control group (Figure 5A). However, in the presence of EGF, the Chem-Mix reduced OMA1 abundance by ~ 40% (P = 0.02) and OPA1 by ~ 60% (P < 0.01) compared to EGF group. To further confirm that changes in OCR and ATP production were due to changes in mitochondrial remodeling, HTR-8/SVneo cells were exposed to the same exposure treatments as above and immunostained with the outer mitochondrial membrane protein TOM20. Immunofluorescence images revealed that the Chem-Mix reduced the mitochondrial structure length (P < 0.05; Figure 5C) with changes in mitochondrial number (P < 0.05). Validation of changes in the mitochondrial network were further observed using super-resolution microscopy. Specifically, smaller, and rounder mitochondria were observed when hCTBs were exposed to the Chem-Mix (Figure 5). Exposure to the EGFR antagonist afatinib or afatinib in combination with EGF did not change mitochondrial structure in HTR-8/SVneo cells (Figure S4).

Figure 5.

Effect of Chem-Mix exposure on mitochondrial fusiogenic proteins and mitochondrial structure in extravillous trophoblasts cells (HTR-8/SVneo). A) Representative western blot images and quantification of OMA1 zinc metallopeptidase (OMA1), optic atrophy type 1 (OPA1), and the reference protein β-actin upon 24 h exposure to: 1) control (C; 0.1 % DMSO), 2) Chem-Mix (Mix; Chem-Mix at 100 ng/ml), 3) human epidermal growth factor (EGF; 30 ng/ml), and 4) Mix + EGF. N = 3 replicates/group. B) Representative dSTORM images and C) quantification (mean ± SEM) of mitochondrial structure length (top) and structure number (bottom) using immunofluorescence staining against the mitochondrial protein TOM20 after exposure to the same treatment groups as described in A. N = 12 replicates/group. Least squares means were used to compare treatments. Different letters denote statistical differences among treatment groups at P < 0.05.

Chem-Mix reduces glycolysis dependent and independent of EGFR

Whether the Chem-Mix could affect the glycolytic capacity and the cellular acidification of trophoblast cells, both hCTB and HTR-8/SVneo cells were investigated using a glycolytic stress test. In hCTB cells, the Chem-Mix reduced glycolytic acidification with or without EGF (P = 0.04 and P = 0.01, respectively; Figure 6A and 6C). In contrast, in HTR-8/SVneo cells, the glycolytic acidification was affected by the Chem-Mix only when co-exposed to EGF (P = 0.05; Figure 6B and 6D). In both cell types, EGF increased glycolytic and non-glycolytic acidification compared to the control group (P < 0.05). When cellular acidification was partitioned into mitochondrial and glycolytic acidification, the Chem-Mix in the presence and absence of EGF decreased mitochondrial acidification only in hCTBs (P < 0.05; Figure 6E). Changes in glycolytic acidification were only observed when HTR-8/SVneo and hCTB cells were exposed to EGF (Figure 6E and 6F).

Figure 6.

Effect of Chem-Mix exposure on human cytotrophoblast cells (hCTB) and extravillous trophoblast cells (HTR-8/SVneo) glycolytic profile. A-B) Glycolytic stress test was conducted to resolve bioenergetic parameters (mean ± SEM) by adding glucose (Glu; 10 μM), the ATP synthase inhibitor oligomycin (OM; 1 μM), and glycolysis inhibitor 2- deoxy-d-glucose (2-DG, 0.5 μM). The glycolytic test was performed after 24 h exposure to: 1) control (C; 0.1 % DMSO), 2) Chem-Mix (Mix; Chem-Mix at 100 ng/ml), 3) human epidermal growth factor (EGF; 30 ng/ml), and 4) Mix + EGF. C-D) Quantification of glycolysis (left), glycolytic capacity (middle), and non-glycolytic acidification (right). E-F) Calculated mitochondrial acidification (pmol/min) and glycolysis acidification (pmol/min) rate derived from glycolytic stress test (see A for assay details and text for calculation details). For all endpoints (A-F), N = 5 primary hCTB cell cultures/group or 5 replicates/group of the HTR-8/SVneo cells were used. Least squares means were used to compare treatments. Different letters denote statistical differences among treatment groups at P < 0.05.

Chem-Mix dysregulates glycolytic enzymes and metabolites

To investigate whether the changes on the glycolytic capacity were due to changes in glycolytic enzymes, their expression and/or abundance was evaluated in HTR-8/SVneo cells. The Chem-Mix tended to upregulate HK2 (P = 0.05) in comparison to the control group but did not affect the expression of the other glycolytic enzymes. EGF increased HK2 (P < 0.01) and LDHA (P = 0.04), reduced PFK2 (P < 0.01) but did not change PKM2 expression compared to the control group (Figure S3). In the presence of EGF, the Chem-Mix tended to upregulate PFK2 (P = 0.05) compared to EGF (Figure S3). Protein abundance of PKM2 and pPKM2 was increased by EGF (P < 0.01) and Chem-Mix (P < 0.01), with a further increase when in combination (P < 0.05; Figure 7). No changes were observed in the pPKM2/PKM2 ratio. To further investigate functional changes in the glycolytic pathway, glucose (Figure 7A and 7B) and lactate (Figure 7C and 7D) media concentrations were measured in cell media. Interestingly, the Chem-Mix alone or in combination with EGF had no effect on glucose concentration, but increased lactate in the cell medium in both, hCTBs and HTR-8/SVneo cells (P < 0.01). In both cell types, EGF reduced glucose concentration (P < 0.01), but increased lactate (P < 0.05) compared to the control group.

Figure 7.

Effect of Chem-Mix exposure on human cytotrophoblast cells (hCTB) and extravillous trophoblast cells (HTR-8/SVneo) metabolic profile and pyruvate kinase muscle 2 (PKM2) abundance. A-B) Glucose and C-D) lactate concentration in the media upon 24 h exposure to exposure to: 1) control (C; 0.1 % DMSO), 2) Chem-Mix (Mix; Chem-Mix at 100 ng/ml), 3) human epidermal growth factor (EGF; 30 ng/ml), and 4) Mix + EGF. N = 3 replicates/ group. E) Representative images and F) quantification of PKM2 and phosphorylated PKM2 (pPKM2) western blots after exposure to the same treatment groups as described in A-D. N = 3 replicates/ group. Least squares means were used to compare treatments. Different letters denote statistical differences among treatment groups at P < 0.05.

Discussion

In this study, we have demonstrated that a mixture of chemicals known to interfere with EGFR activation (Waye et al., 2024) can also interfere with human trophoblast cells’ bioenergetics. In support of this, our data demonstrated that the Chem-Mix reduces the mitochondrial network size, reducing mitochondrial oxygen consumption and ATP production, in an EGFR dependent and independent manner. The impact on cellular bioenergetics was not limited to the mitochondria, with the Chem-Mix reducing glycolysis, further contributing to a reduction in energy production. Overall, these findings indicate that the exposure to a mixture of EGFR-disrupting chemicals has the potential to affect trophoblast cell function, by reducing the ability of placental cells to generate energy. Ultimately, this may limit the biosynthetic capacity of the placenta to support its own rapid growth and fetal growth during mid- to late pregnancy (Aye et al., 2022).

It is well established that chemicals in a mixture can contribute to toxicity even if they are present below their own effect threshold (Kortenkamp, 2014). Chemicals with the same modes of toxic action tend to follow the mixture concept of “concentration addition,” whereas those with different modes of action act according to “independent action” (Kortenkamp, 2014). Our results support the concept of concentration addition, where there was not a significant change in oxygen consumption when cytotrophoblasts where exposed to the single chemical, but when in a mixture with each chemical at the same concentration, the mixture reduced EGF-induced oxygen consumption. At lower doses (1 and 10 ng/ml), the Chem-Mix also reduced oxygen consumption in primary hCTBs. This is relevant because most of those chemicals (BPS: 0.37 μg/l in adults and 0.29 μg/l in children urine (Lehmler et al., 2018); PCBs: 0.04 to 5.37 ng/ml maternal serum and 0.004 to 1.21 ng/ml cord serum (Park et al., 2008); TNC: 48 ng/g of lipid (Trabert et al., 2012); ATZ: residue limits for atrazine in foods between 0.25 and 15 mg/kg (WHO, 1990)) can be found in human body fluids in this concentration range (Dearth and Hites, 1991; Newsome and Ryan, 1999; Perry et al., 2000; Rylander et al., 2005; Costopoulou et al., 2006; Diamanti-Kandarakis et al., 2009; Zhu et al., 2009; Ritter et al., 2011; Mendas et al., 2012; Zoeller et al., 2012; 2014; National Academy of Medicine, 2014; Megersa, 2015; Faroon and Ruiz, 2016; Namulanda et al., 2017; Schweizer et al., 2018; Wang et al., 2018b).

EGF stimulated the cellular metabolic rate in primary hCTBs, which is in agreement with prior work where EGF enhanced OCR in hCTBs (Kolahi et al., 2017). In this study, in hCTBs, the Chem-Mix reduced OCR in the presence of EGF in support of an EGFR-dependent effect interfering with maximal respiration. However, in HTR-8/SVneo cells, the Chem-Mix reduced OCR independently of EGF. This discrepancy is likely related to the fact that HTR-8/SVneo cells were relatively unresponsive to the EGF stimuli in both, basal and maximal respiration. Almost all bioenergetics studies take advantage of the principle of steady-state mass balance to determine metabolic fluxes. Measuring mitochondrial metabolic fluxes via OCR is essential to understand mitochondrial physiology (Jones et al., 2021). Changes in oxygen consumption and media acidification allow the integration of energy flux through mitochondrial (OXPHOS) and non-mitochondrial (glycolysis) metabolic pathways from OCR and ECAR data and how much each pathway contributes to total ATP synthesis. Given that EGF did not affect ATP derived from the mitochondria (mitoATP) and that the Chem-Mix lowered ATP independent of EGF, we could not determine if the Chem-Mix effect on ATP production was EGFR dependent.

While primary cells often lead to higher variability, the robust phenotype observed in primary hCTBs supports previous work where EGF has been reported to stimulate mitochondrial ATP production (Kolahi et al. 2017). Notably, the fact that EGF increased mitochondrial ATP in hCTBs, but not HTR-8/SVneo cells, may relate to the abundance of EGFR in each cell type. While we did not test the EGFR expression in each of the primary hCTB cell cultures used (it is likely to be very variable across primary cell cultures), at term, syncytiotrophoblasts and cytotrophoblasts have the highest expression of EGFR compared to extravillous trophoblasts (HTR-8/SVneo cell line) with five times lower expression (Proteinatlas, 2024).

EGF has been reported to enhance glucose consumption and lactate production in cancer cells (Kaplan et al., 1990; Baulida et al., 1992) and stimulate glycolysis of cytotrophoblasts (Kolahi et al., 2017). These findings were confirmed in our study through the glycolytic rate assay, that measures changes in the medium acidification, where glycolysis and the glycolytic capacity increased under EGF exposure in both hCTBs and HTR-8/SVneo cells. It is generally assumed that all lactate produced via glycolysis is the main driver of extracellular acidification as lactate is effluxed out of the cell (Schmidt et al., 2021). This is further reinforced by our results where EGF resulted in an increase in lactate detected in the cell media for both cell types.

Results obtained in the glycolytic assay and the medium acidification assays differed. While the Chem-Mix reduced medium acidification as measured by the glycolytic rate assay, the lactate concentration measured in the medium was higher compared to the control. We hypothesize that this paradoxical phenomenon may be due the fact that the Chem-Mix can regulate PKM2, an enzyme that catalyzes the final and a rate-limiting reaction in the glycolytic pathway (Zhang et al., 2019). In HTR-8/SVneo cells, the Chem-Mix increased PKM2 phosphorylation, with a further increase when in combination with EGF. PKM2 phosphorylation at the Y105 residue results in PKM2 dimerization (less active form) reducing glycolysis and promoting lactate production, while tetramer PKM2 (active form) directs glucose towards oxidative metabolism (Hitosugi et al., 2009). We hypothesize that the Chem-Mix shifts glucose metabolism from mitochondrial OXPHOS to aerobic glycolysis through phosphorylation and dimerization of PKM2, leading to lower glycolytic capacity, yet higher lactate production.

A change in the metabolic flux of glucose from mitochondrial OXPHOS to aerobic glycolysis is regarded as a hallmark of the Warburg effect. Previous work has demonstrated that PKM2 can interact with the mitochondrial protein mitofusin 2 (MFN2) and that the resulting PKM2:MFN2 complex can reduce glycolysis and promote OXPHOS and mitochondrial fusion (Li et al., 2019). Thus, given the effect of the Chem-Mix on oxygen consumption, ATP abundance, and phosphorylation of PKM2, we hypothesize that the Chem-Mix will also disrupt mitochondrial dynamics. The abundance of OPA1, a mitochondria membrane GTPase involved in mitochondrial fusion, was reduced by the Chem-Mix in HTR-8/SVneo cells. Super-resolution imaging confirmed that the Chem-Mix resulted in reduced mitochondrial network size. This finding is not solely EGFR dependent as we have also shown that afatinib, a potent EGFR inhibitor (Wang et al., 2018a), does not affect mitochondrial network size or structure number (Figure S4).

Other pathways such as TOR, Ras, Raf, cAMP/PKA, ERK1/2, p38 MAPK, STAT3, and Wnt/ β-catenin, have also been implicated in mitochondrial dynamics (Morita et al., 2013; Valsecchi et al., 2013; Cook et al., 2017; Nagdas and Kashatus, 2017; Bhargava et al., 2020; Lahiri et al., 2021) and are the target of some of the chemicals in the mixture (Llansola et al., 2009; Ren et al., 2010; Lu et al., 2011; Fa et al., 2013; Zhang et al., 2024). This suggests that although the mixture interferes with certain EGFR-mediated functions within trophoblast cells, it can also interfere with additional pathways related to mitochondrial dynamics that need to be further explored. Altogether, our data suggest that the Chem-Mix disrupts mitochondrial dynamics by promoting mitochondrial fission and inhibiting the mitochondrial fusion machinery, resulting in smaller individual mitochondria size and the overall mitochondrial network. Yet, our results do not support that this effect occurs via EGFR.

To note, changes reported here are restricted to a single 24-h exposure. While chronic exposure effect of the Chem-Mix remains to be evaluated, our previous work evaluating proliferation up to 36 h (Waye et al., 2024) and another study interrogating single chemical exposures (PCB-153, atrazine, and BPS) reporting organelle stress responses, including unfolded protein response activation, mitochondrial dysfunction, and ROS generation (Kim et al., 2022; Zhang et al., 2022b; Abarikwu et al., 2023) suggests that the Chem-Mix can have a prolonged impact in vitro.

As single exposures, several of the chemicals included in the Chem-Mix (atrazine, niclosamide, and PCBs) have been previously reported to disrupt cellular bioenergetics, but in doses much higher than those tested in the present study. For instance, a mixture of PCBs (Aroclor-1254) induced a reduction of OXPHOS and ATP production in the neuroblastoma cell line SH-SY5Y, at doses 50 times higher (5 μg/ml; (Cocco et al., 2015)) to those used in the present study. Also, a previous study has demonstrated that ATZ can impact mitochondrial morphology and reduce OCR in murine skeletal muscle and liver cells, but with doses a 1,000 times higher (100 μg/ml) compared to the highest dose used in this study (Lim et al., 2009). Niclosamide has also been reported to induce mitochondrial fragmentation in HeLa cells, with significant results starting at 1μM (327 ng/ml), a dose three times higher to that used here (Park et al., 2011). To our knowledge, neither BPS nor TNC have been reported to alter mitochondrial dynamics.

Previous transcriptomics analyses identified a metabolic shift consistent with a Warburg effect in metabolically stressed placentas which more often presented clinically as preeclampsia and/or fetal growth restriction (Ackerman et al., 2023). As our current study demonstrated a similar effect could be mimicked in vitro through exposure of human trophoblasts to environmental chemical mixtures it is critical for future studies to investigate the extent to which interactions between different environmental chemicals could be responsible for the increased prevalence of pregnancy complications in certain populations such as historically disadvantaged groups whose chemical exposure burden is higher compared to other groups (Varshavsky et al., 2020; Chan et al., 2021; Goin et al., 2022).

In conclusion, our findings demonstrate that a mixture of chemicals known to target EGFR can disrupt mitochondrial remodeling by reducing mitochondrial network size, altering cellular bioenergetics, and reducing the capacity of human cytotrophoblast cells to generate energy via OXPHOS and the glycolytic pathway. Considering that changes in mitochondrial biogenesis and morphology have been reported during common pregnancy disorders, including gestational diabetes (Sobrevia et al., 2020) and pre-eclampsia (Hu and Zhang, 2022), our findings suggest that the exposure to this mixture of chemicals during pregnancy can be deleterious to placental function and/or development. Future studies should investigate the mechanism by which EGFR-disrupting chemicals alter mitochondrial dynamics and its pathological significance on the placenta in an in vivo model.

Highlights.

An EGFR-targeting chemical mixture reduced mitochondrial OXPHOS in placental cells.

An EGFR-targeting chemical mixture reduced EGF-induced glycolytic acidification.

An EGFR-targeting chemical mixture causes mitochondrial remodeling towards fission.