CONCENTRATION-DEPENDENT EFFECTS OF CHLORPYRIFOS OXON ON PEROXISOME PROLIFERATOR-ACTIVATED RECEPTOR SIGNALING IN MCF-7 CELLS

Abstract

Chlorpyrifos oxon (CPO) is the active metabolite of the organophosphorus pesticide, chlorpyrifos. CPO is a potent inhibitor of acetylcholinesterase (AChE) and other serine hydrolases including fatty acid amide hydrolase (FAAH). AChE is critical in regulating cholinergic signaling while FAAH catalyzes the inactivation of fatty acid signaling lipids including the endocannabinoid (eCB) N-arachidonylethanolamine (anandamide, AEA) and eCB-like metabolites (e.g., oleoylethanolamide, OEA). AEA and OEA are both peroxisome proliferator-activated receptor (PPAR) agonists that regulate numerous genes involved in lipid metabolism and energy homeostasis. We used the MCF-7 human breast cancer cell line, which expresses AChE, FAAH and PPAR alpha and gamma subtypes, to evaluate the potential effects of CPO on PPAR-related gene expression in an in vitro human cell system. CPO elicited relatively similar concentration-dependent inhibition of both AChE and FAAH. Marked concentration- and time-dependent changes in the expression of four selected PPAR-related genes, LXRα, ACOX1, ABCG2 and AGPAT2, were noted. These findings suggest chlorpyrifos may influence lipid metabolism through blocking the degradation of eCBs or eCB-like metabolites and in turn affecting PPAR receptor activation. The results highlight the potential for non-cholinesterase actions of this common insecticide metabolite through disruption of PPAR signaling including effects on lipid metabolism, immune function and inflammation.

1. Introduction

Chlorpyrifos (CPF; O,O’-diethyl-3,5,6-trichloropyridinyl-phosphorothioate) is one of the most extensively used insecticides worldwide, although it has been banned for residential use in the United States (Kondakala et al., 2017). In the US, approximately six million pounds are used each year (Atwood and Paisley-Jones, 2017). Although risks to non-target organisms, including humans, have been extensively studied for safety and regulatory reasons, potential adverse effects of environmental exposure to chlorpyrifos remain a concern for public health (Casida and Quistad, 2005; Parran et al., 2005; Zeng et al., 2017). For humans, routes of exposure to chlorpyrifos can include dermal, inhalation, and ingestion, with ingestion being the major route of entry (Huang et al., 2019). After absorption, chlorpyrifos is metabolically converted to a highly reactive metabolite, chlorpyrifos oxon (CPO), through oxidative desulfuration (Sultatos and Murphy, 1983)(Crane et al., 2012). The cytochrome P450s, Cyp2B6 and cyp3A4 play important roles in activation of CPS (Sams et al., 2004; Mutch and Williams, 2006; Croom et al., 2010; Crane et al., 2012). Oxidative desulfuration of CPF leads to a much more potent serine hydrolase inhibitor than the parent insecticide (Crane et al., 2012; Sultatos, 1994).

The acute toxicity of OPs is initiated by acetylcholinesterase (AChE) inhibition. Acetylcholinesterase is critical in the dynamic regulation of cholinergic signaling at synapses of muscles, neurons and parasympathetic end-organs. Extensive AChE inhibition inhibits hydrolysis of the neurotransmitter acetylcholine at cholinergic synapses, leading to prolonged activation of cholinergic receptors and resulting cholinergic toxicity (Sultatos, 1994; Pope 1999). Classic signs and symptoms of acute cholinergic toxicity include blurry vision, vomiting, diarrhea, convulsions, muscle twitching, and respiratory failure, the ultimate cause of death with lethal exposures. While the molecular initiating event in acute toxicity of CPF is AChE inhibition, non-cholinesterase macromolecular targets for OPs that may either modulate acute cholinergic toxicity or contribute to other toxicity pathways have been the subject of investigation for decades (Carr et al., 2014; Leung et al., 2019; Casida and Quistad, 2004; Pope, 1999; Russom et al., 2014; Yozzo et al., 2013).

A number of other serine hydrolases besides AChE are sensitive to inhibition by some OPs (Casida and Quistad, 2004; Nomura and Casida, 2011)(Casida and Quistad, 2005). Fatty acid amide hydrolase (FAAH) is involved in the regulation of a variety of processes including sleep induction, analgesia, and energy homeostasis (Cravatt and Lichtman, 2003; Ahn et al., 2009; (Ueda et al., 2000). CPO and other OPs inhibit FAAH via covalent binding to its active site serine residue within a catalytic triad, in a manner analogous to inhibition of AChE (Mangas et al., 2017). In some comparative studies, FAAH is reported to be more sensitive to inhibition by CPF than AChE (Carr et al., 2020, 2014; Huang et al., 2007). Of particular importance is that FAAH is the primary enzyme involved in the inactivation of the endocannabinoid anandamide (N-arachidonoylethanolamine, AEA) and the endocannabinoid-like metabolites, palmitoylethanolamide (PEA) and oleoylethanolamide (OEA) (McKinney and Cravatt, 2005).

The endocannabinoid (eCB) system is a global neuromodulatory network that regulates neurotransmitter release at neuronal presynaptic terminals (Castillo et al., 2012). The CB1 receptors, along with the eCBs and enzymes responsible for their biosynthesis and degradation make up this pathway. Endocannabinoid signaling is important in a number of neurological functions including appetite regulation, pain perception, cognitive development, emotional state, seizures and many others (Osei-Hyiaman et al., 2005; Tegeder, 2016). The eCB system also plays an important role in lipid homeostasis and energy balance. The two primary eCBs, AEA and 2-arachidonoylglycerol (2-AG), are synthesized “on demand” from membrane phospholipids following neuronal depolarization (Di Marzo et al., 1994; Lu and MacKie, 2016; Nomura et al., 2008).

Peroxisome proliferator-activated receptors (PPARs) are ligand-regulated transcription factors linked to the regulation of genes involved in multiple pathways of lipid metabolism including fatty acid oxidation, lipid transport, lipogenesis, and cholesterol metabolism, making PPARs essential regulators of energy homeostasis (Mandard et al., 2004; Semple et al., 2006). PPARs are involved in multiple other processes including carcinogenesis, inflammation and immune modulation (see review of Hernandez-Quiles et al., 2021). In this study, we focused on two subtypes, PPARα and PPARγ. AEA is an agonist of both of these subtypes while the eCB-like metabolites OEA and PEA are agonists of PPARα, all with the potential to influence expression of their target genes (Brunetti et al., 2019).

To evaluate CPO’s effect on FAAH and PPAR signaling, we first compared the in vitro concentration-dependent effects of CPO on FAAH and AChE activity in extracts of MCF-7 cells. This human cell line was chosen based on species relevance, the presence of both “target enzymes” and the presence of functional PPARα and PPARγ (Nwankwo and Robbins, 2001; Suchanek et al., 2002). Four genes that are regulated by PPARs were selected based on their role in the regulation of lipid metabolism, and for their known expression in the MCF-7 cell line. The objective of this study was to evaluate the potential for CPO to influence PPAR signaling by inhibiting FAAH and putatively modifying the levels of endogenous or exogenous AEA or OEA. Such a mechanism could contribute to non-cholinesterase mechanisms of toxicity acting through changes in lipid metabolism. Changes in PPAR signaling as a consequence of OP induced inhibition of FAAH could possibly form a bridge between exposure to organophosphorus pesticides and changes in lipid metabolism related to obesity.

2. Materials and Methods

2.1. Chemicals

Eagle’s Minimum Essential Medium and fetal bovine serum were purchased from Quality Biological Inc. Chlorpyrifos oxon (CPO, >97% purity) was purchased from Chem Service and kept desiccated under nitrogen at −70°C. The desiccator containing CPO was brought to room temperature under the fume hood before opening to minimize any CPO hydrolysis. Bradford reagent, 5-amino-2-methoxypyridine (AMP), ethylenediaminetetraacetic acid (EDTA), 5,5-dithio-bis-(2-nitrobenzoic acid) (DTNB), Trizma base, Trizma hydrochloride, Triton X-100, sodium phosphate, sodium chloride, and bovine serum albumin were all purchased from Sigma-Aldrich (St. Louis, MO). Octanoyl methoxypyridine and was a kind gift from Dr. Bruce Hammock, UC Davis. TaqMan gene expression assays, TaqMan Fast Advanced Master Mix No AMP Erase UNG, and the High Capacity cDNA Reverse Transcription Kit were all purchased from Applied Biosystems. Anandamide, oleoylethanolamide, and GW9662 were purchased from Cayman Chemical. DMSO was purchased from ATCC. Human recombinant insulin in zinc solution was obtained from Gibco. Phosphate buffered saline (PBS) was purchased from Corning. Trizol reagent was purchased from Ambion Life Sciences. 5-amino-2-methoxypyridine (AMP) as well as all chemicals required for the MTT assay were purchased from Sigma Aldrich.

2.2. Cell Culture

Human MCF-7 cells were obtained from American Type Culture Collection (ATCC, HTB-22). Cells were initially seeded in 25 cm² culture flasks (Corning), and sub-cultured at a ratio of 1:3 once a confluency of ~90% was reached. For incubation with CPO, cells were seeded at a density of 4.2 × 10⁶ cells per 60 mm plate in 4 mL complete culture medium (Eagle’s Minimum Essential Medium containing 10% FBS and 0.01 mg/ml human recombinant insulin) and cultured at 37°C in 95% air:5% CO2 until they formed an adherent layer (generally 24–36 hours). Plates were removed from the incubator and the medium was aspirated. CPO, AEA and OEA were dissolved first in 100% DMSO and then added to fresh culture medium prior to adding to the cells, with a final DMSO concentration of 1% in all cases. Preliminary studies indicated that 1% DMSO was necessary for complete solubility of all reagents used in these assays.

2.3. Cell sampling and processing

Sample collection and preparation were performed the same for all assays. Plates were removed from the incubator, medium was aspirated, and the cells were rinsed three times with 1 mL ice-cold PBS. Lysates were made by adding 0.5 mL ice-cold lysis buffer (prepared by adding 1 mL 50 mM EDTA, 25 mL 50 mM Tris HCl pH 7.4, 50 μL Triton X-100, and 7.5 mL 1 M NaCl, diluted to 50 ml with deionized water) to each plate, followed by ten minutes incubation on ice. Lysates were homogenized for 30 seconds using an Eberbach Con-Torque tissue homogenizer. Homogenates were then transferred to pre-chilled 1.5 mL Eppendorf tubes and centrifuged at 12,000 × g for 15 minutes at 4°C. The supernatants were collected, diluted with lysis buffer and used for enzyme assays.

2.4. Enzyme Assays

Preliminary assays determined dilutions and times of incubation that led to linear rates of substrate hydrolysis with both enzymes. Acetylcholinesterase activity was measured by a modified Ellman (Ellman et al., 1961) method as described previously (Pope et al., 2018). Plates (96-well, clear, flat-bottom plastic) were prepared on ice by first adding 25 μL of either lysis buffer (for blanks) or diluted supernatant to wells in quadruplicate. A cocktail (175 μl) containing the chromogen DTNB (1.086 mM), the substrate acetylthiocholine (ATC, 7.62 mM), and disodium EDTA (1 mM) in Tris buffer, pH 7.2 was added to all wells to give final concentrations of 0.1 mM DTNB and 1 mM ATC. Enzyme activity was measured kinetically at 412 nM (every 60 seconds for 10 minutes) in a SpectraMax 340PC plate reader.

FAAH activity was measured essentially by the method of Huang and coworkers (Huang et al., 2007). A 10 mM stock solution of the synthetic substrate octanoyl methoxypyridine (OMP) was prepared in a vehicle of 50% DMSO/50% ethanol. Plates (96-well, black, flat-bottom plastic) were prepared on ice by first adding 30 μL of either lysis buffer (blanks) or diluted supernatant to each well. Assay buffer (125 mM sodium phosphate, pH 8 containing 1% glycerol and 0.1% Triton X-100; 160 μl), was then added and 10 μl of the stock substrate solution was added to all wells immediately before loading into a CLARIOstar plate reader (BMG Labtech, Cary, NC) with fluorescence measured every 60 seconds for 10 minutes (excitation, 320 nM; emission, 396 nM). FAAH activity (hydrolysis of OMP) was calculated by relative fluorescence unit changes, with interpolation using a standard curve of the product, AMP.

Enzyme activities were normalized relative to protein content. Protein concentration was measured using the Bradford method (Bradford, 1976) with bovine serum albumin as a standard. Lysis buffer only was used in blank wells.

2.5. Treatments in Culture

Cells were cultured as in section 2.2. CPO, AEA and OEA were first dissolved in DMSO before being added to the culture medium. Concentration-dependent effects of CPO were studied by adding culture medium containing 0, 250, 500 or 1,000 nM CPO to the cells and culturing for 8 hours, 1 day, and 3 days. Cells were collected and processed as in section 2.3 to measure AChE and FAAH activities. For the portion of the study using the PPAR antagonist GW9662, the cells were seeded and cultured as described above with the antagonist diluted in DMSO and and then added to complete medium to produce final concentrations of 0.5, 2.0, and 10 μM (1% DMSO final concentration). Cells were incubated with GW9662 for 2 hours. The medium was then aspirated and replaced with fresh complete medium containing 1 μM CPO with 1% DMSO and incubated for 24 hours. Complete culture medium containing 1% DMSO was used for control.

2.6. MTT Assay

To determine possible effects of DMSO and GW9662 on cell viability, an MTT assay was performed. Cells were seeded in 24-well plates and cultured for 24 hours with 250, 500, or 1000 nM CPO, 10 μM GW9662, or vehicle only (1% DMSO). Blank wells contained complete culture medium only. Thiazolyl blue tetrazolium bromide (MTT) was dissolved as a 5 mg/mL solution in sterile PBS, then filter sterilized through a 0.2 μM filter into an autoclaved amber bottle. The solution was then stored (protected from light) at 4° C. After 24 hours in culture, the plates were removed from the incubator, placed under a laminar flow hood and 50 μL of MTT solution was added to each well. The plates were placed back into the incubator for 4 hours. Plates were then removed, culture medium was aspirated, and 500 μL DMSO was added to each well. The plates were swirled/mixed for a few seconds and then incubated for 2 hours at room temperature (protected from light). Absorbance at 570 nM was measured using a BMG Labtech POLARstar Optima plate reader.

2.7. Gene Expression Analysis

Cells were cultured with either vehicle or 250, 500 or 1000 nM CPO as described above. In some cases, AEA (1 μM; a PPARα and PPARγ agonist) or OEA (1 μM; a PPARα agonist) was added to study the potential interaction between exogenous PPAR agonists and CPO. After 8, 24 or 72 hours, culture medium was aspirated before adding Trizol reagent to isolate RNA, according to the manufacturer’s instructions. Total RNA was quantified by absorbance (A₂₆₀/A₂₈₀ ratio) using a Beckman Coulter DU530 UV/Vis spectrophotometer. An Applied Biosystems High Capacity cDNA Reverse Transcription Kit was used for complete cDNA synthesis using 1 μg total RNA. Quantitative real-time polymerase chain reactions were carried out in an Applied Biosystems 7500-Fast Real Time PCR System. The reaction mixture consisted of TaqMan Fast Advanced Master Mix, TaqMan Gene Expression Assay Primers/Probes, and the synthesized cDNA, for a final reaction volume of 20 μL in 96-well fast optical plates (Applied Biosystems). Thermal protocol settings were UNG incubation 50° C for 2 minutes (1 cycle), enzyme activation 95° C for 20 seconds (1 cycle), denature 95° C 3 seconds (40 cycles), and anneal/extend 60° C for 30 seconds (40 cycles). Experiments were set up in the system software as comparative Ct, using the vehicle-treated samples as reference, and β-Actin as the endogenous control. Each target gene for each sample type was run in triplicate. No treatment-controls (NTCs) for blanks contained nuclease-free water in place of cDNA. The dyes in each reaction mix were FAM (reporter), NFQ-MGB (quencher), and ROX (passive reference).

2.8. Western Blots

Cells cultured with one of the three concentrations of CPO, as well as those treated with GW9662 (± CPO, 1,000 nM), were collected for Western blot analysis. Vehicle-only samples were used as controls. Samples were collected and prepared by aspirating culture medium and adding 200 μL ice cold RIPA buffer to each sample well. After 30 minutes at 4° C with mild shaking, cells were collected and centrifuged (20 min, 12,000 × g) in 1.5 mL Eppendorf tubes. Supernatants were stored at −20° C. On the day of blotting, samples were thawed on ice prior to being prepared for addition to the running gel. Samples were prepared by adding 75 μL thawed supernatant to 25 μL 4x loading buffer and boiled for 5 minutes. Polyacrylamide gels were prepared in advance (12% resolving gel/4% stacking gel). In each sample lane, 25 μL prepared sample in loading buffer was added and separated at 120 V for approximately 2 hours. Transfers to PVDF membrane were then carried out at 100 V for 90 minutes. After protein transfer, the PVDF membranes were removed from the apparatus and placed in blocking buffer (PBST/5% BSA). Membrane blocking was carried out at room temperature for 1 hour with gentle shaking. All antibody dilutions were prepared in PBST/5% BSA. Primary antibodies diluted 1:1000 were as follows: rabbit anti-ABCG2 (Abnova), mouse anti-AGPAT2 (Abnova), rabbit anti-B-actin (Novus), and mouse anti B-actin (EMD Millipore). Secondary antibodies were diluted 1:10,000 in PBST/5% BSA. Secondary antibodies used were rabbit anti-mouse IgG, HRP conjugated and goat anti-rabbit IgG, HRP conjugated (both from Millipore). After 1 hour incubation in blocking buffer, membranes were incubated overnight with a primary antibody at 4° C with gentle shaking. Membranes were then rinsed three times with PBST for five minutes each andthen incubated at room temperature with the appropriate secondary antibody for 1 hour with gentle shaking. Following incubation with the secondary antibody, membranes were again rinsed with PBST three times for five minutes each. Membranes were then coated with SuperSignal Western Pico Plus Chemiluminescent Substrate (ThermoFisher) for five minutes. Excess substrate was removed, membranes were placed in plastic film to prevent drying, and images were taken using a VersaDoc 5000 MP imaging system.

2.9. Data Analysis

All enzyme activities were calculated in terms of nmol/min/mg protein and plotted as percent inhibition relative to vehicle controls. Results from three independent replicate assays were evaluated using GraphPad Prism version 6 software. Enzyme data were evaluated by 2-way ANOVA with treatment and time as main effect variables. Tukey’s post hoc test was used to compare each treatment group within each time of incubation.

MTT assay results were calculated as percent viability compared to control values by subtracting the absorbance value of the blanks from the absorbance value of the sample, then dividing that by absorbance value of the control minus the absorbance value of the blank, and multiplying the result by 100. A one-way ANOVA was performed to evaluate differences among treatment groups.

PCR results were analyzed using the ΔΔCt method. Ct values below 16 and above 35 were not included in further analyses. Gene expression changes were normalized against the reference gene β-actin and were presented as relative fold-change over vehicle control samples (cultured in complete culture medium with DMSO, 1% final). Relative fold-changes in expression were calculated using 2^−ΔΔCt for each target gene (Adnan et al., 2011; Gaonkar et al., 2018; Natarajan et al., 2020). Results for each gene were grouped into each of the three time points. Western blots were analyzed visually for qualitative assessment of changes in protein translation relative to observed gene expression changes.

3. Results

3.1. MTT Assay

The MTT assay was performed to determine any changes is cell viability, due to treatments, or because of the vehicle concentration in the culture medium. As shown in Figure 1, there were no significant differences between vehicle, the selected concentrations of CPO, or with the highest concentration of GW9662 used (10 μM).

Figure 1. MTT assay for cell viability.

Data are expressed as mean percent viability (± SEM) in treated vs vehicle controls with no DMSO added (n=3).

3.2. AChE and FAAH inhibition by CPO

We first evaluated the in vitro inhibitory potency of CPO (37°C for 20 minutes) against AChE and FAAH in supernatants of MCF-7 cells. Figure 2 shows that both enzymes were relatively similar in sensitivity to CPO in vitro (IC₅₀ 36–46 nM). Based on these in vitro results, the presence of inactivating proteins in cells and FBS in the culture medium (e.g., carboxylesterases, cholinesterases), along with preliminary results with in culture inhibition in MCF-7 cells (data not shown), we selected 250, 500 and 1,000 nM CPO to study concentration-related changes. As enzyme inhibition with the highly reactive CPO would likely occur relatively rapidly after exposure, and nuclear receptor-mediated changes in gene expression would presumably occur later, we selected a range of time-points to study, i.e., 8 hours, 1 day and 3 days.

Figure 2. In vitro inhibition of A) acetylcholinesterase and B) fatty acid amide hydrolase activity from MCF-7 cells.

MCF-7 cells were harvested at ~90 % confluency, lysed and centrifuged to produce cell supernatants for enzyme assays as described in Methods. The supernatants were incubated at 37° C with vehicle or one of a range of concentrations of chlorpyrifos oxon (CPO) for 20 minutes prior to assay of residual enzyme activity. Data are expressed as mean (± SEM) percent inhibition relative to vehicle controls (n = 4 independent assays). IC₅₀ values (± 95% confidence intervals) for AChE and FAAH were 36 (± 2.0) and 46 (± 4.7) nM, respectively.

As shown in Figure 3, both AChE and FAAH activities in MCF-7 cells cultured in the presence of CPO were inhibited in a concentration- and time-dependent manner. With AChE, there was a significant main effect of treatment (F_2,6 = 73.3, p <0.0001), a main effect of time (F_2,12 = 18.3, p=0.0002) and a significant interaction (F_4,12 = 3.52, p =0.04). Similarly, with FAAH there was a main effect of treatment (F_2,6 = 125.2, p<0.0001), time (F_2,12 = 55.6, p<0.0001) and a significant interaction (F_4,12 = 6.1, p=0.007). The most extensive inhibition of both enzymes was noted at the earlier time-points (8 hr and 1 d), and significant recovery was noted with both enzymes by 3 d compared to the earlier time-points.

Figure 3. Effects of CPO on A) AChE and B) FAAH activity in MCF-7 cells exposed in culture.

Data are expressed as mean percent inhibition (± SEM) based on vehicle controls (n = 3). Control values (nmol/min/mg protein mean ± SEM): AChE: 8 hours 4.5 ± 0.3, 1 day 4.7 ± 0.3, 3 days 4.7 ± 0.4; FAAH: 8 hours 23.3 ± 0.7, 1 day 24.8 ± 2.4, 3 days 26.5 ± 2.0.

3.3. Gene Expression Changes

FAAH hydrolyzes the PPARα and PPARγ agonist AEA, and the PPARα agonist OEA (Guzman et al., 2004; O’Sullivan, 2007; (Sun and Bennett, 2007). We hypothesized that CPO-mediated inhibition of FAAH would block the inactivation of these endogenous lipid metabolites and thereby increase the activation of PPAR-related gene transcription. As it is unclear whether endogenous lipid metabolites in MCF-7 cells could achieve levels sufficient for modulating PPAR gene expression profiles in the presence of FAAH inhibition, we also studied the interactive effects of CPO on gene expression changes when exogenously added agonist (AEA or OEA) was included. We postulated that the effects of exogenously added agonist would be amplified by CPO compared to changes observed without including an exogenous agonist.

Figure 4 shows concentration-dependent effects of CPO, alone or in the presence of AEA or OEA, on expression of the selected PPAR-regulated genes (LXRα, ACOX1, AGPAT2, ABCG2) after eight hours of exposure. With exposure to CPO alone (Figure 4A), there was a marked increase in AGPAT2 expression (16- to 23 -fold) that was relatively independent of CPO concentration. In contrast, there was a CPO concentration-related increase in ACOX1 expression (from 9- to 50-fold). ABCG2 and LXRAa were also increased but to a lesser degree (4- to 5-fold) compared to the other two genes. Figure 4B shows the effects of CPO in the presence of AEA. Interestingly, adding AEA led to decreased expression of LXRa, ACOX1 and ABCG2 expression (relative fold changes <1.0) and a mean AGPAT2 fold increase of only 1.2). Thus, in contrast to an expected enhancement of PPAR-related gene transcription with added AEA, there was a reversal or dampening of the relatively large increases noted with CPO only. Figure 4C shows changes with CPO in the presence of OEA. In this case, 250 nM and 500 nM CPO decreased expression of ACOX1 and ABCG2 but the highest concentration (1000 nM) led to 4- to 5-fold increased expression of ACOX1, ABCG2 and LXRa. AGPAT2 was upregulated with all three concentrations of CPO (16- to 45-fold). While OEA appeared to enhance the effects of CPO on AGPAT2 expression (45-fold vs 23-fold increase), addition of OEA reversed or reduced the increase in ACOX1 seen with CPO only (8- to 53-fold with CPO only vs 0.25- to 4-fold in the presence of added OEA).

Figure 4. Effects of chlorpyrifos oxon alone (A), in the presence of anandamide (B), or in the presence of oleoylethanolamide (C) on PPAR-related gene expression after 8 hours of exposure.

Gene expression was assayed as described in methods and data are expressed as relative fold-change.

Figure 5 shows changes in gene expression after one day of exposure. With CPO alone (Figure 5A), relatively minimal changes in expression of LXRa, ACOX1 and ABCG2 were noted (fold change 0.93 to 2.09) with 250 or 500 nM CPO. The highest CPO concentration however was associated with 16- to 18-fold increased expression of all four genes. Concentration-dependent increases were noted with all four genes when AEA was included with CPO (Figure 5B), in particular with AGPAT2 (44 to 64-fold with 500 and 1,000 nM CPO). Lesser increases in expression of the other three genes were noted. Figure 5C shows changes elicited by CPO in the presence of exogenous OEA. ABCG2 expression was decreased (0.5 to 0.7-fold) and LXRa, ACOX1 and AGPAT2 showed relatively minimal increases (1.1 to 3.7-fold) with 250 and 500 nM CPO. In contrast, with 1000 nM CPO all four genes showed marked increases (7- to 17-fold).

Figure 5. Effects of chlorpyrifos oxon alone (A), in the presence of anadamide (B), or in the presence of oleoylethanolamide (C) on PPAR-related gene expression after 1 day of exposure.

Gene expression was assayed and data expressed as in Figure 4.

Figure 6 shows the effects of CPO with and without added AEA or OEA at the 3 day time-point. Compared to earlier time-points (8 hr and 1 d), lesser increases in PPAR-related gene transcription were generally noted. With exposure to CPO alone (Figure 6A), LXRa, ACOX1 and ABCG2 expression was decreased (0.7- to 0.8-fold) with 250 nM CPO. Relatively minimal changes were noted with the 500 nM CPO exposure. With 1,000 nM CPO, 4- to 8-fold increases were observed. Figure 6B shows changes when AEA was included. The lowest concentration of CPO (250 nM) had minimal effects (0.77- to 1.9-fold) while 1000 nM CPO was associated with higher expression (3.8- to 13.5-fold) increases. When OEA was included (Figure 6C), compared to the magnitude of changes seen at earlier time-points there were again relatively minimal changes. With 250 or 500 nM CPO, there was a relatively minimal change with LXRa, ACOX1, and ABCG2. The increase in AGPAT2 expression at 3 days after dosing was much lower than noted with the 8-hour exposure (Figure 6C). Supplemental Table 1 lists all gene expression changes represented in Figures 4–6.

Figure 6. Effects of chlorpyrifos oxon alone (A), in the presence of anadamide (B), or in the presence of oleoylethanolamide (C) on PPAR-related gene expression after 3 days of exposure.

Gene expression was assayed and data expressed as in Figure 4.

Figure 7 shows the effects of GW9662 on AGPAT2 and ABCG2 expression, both by itself and when CPO (1,000 nM) was added after preincubation with the inhibitor. The most extensive increases in AGPAT2 (Figure 7A) and ABCG2 (Figure 7B) expression were observed in samples containing only 1,000 nM CPO (12- to 19-fold, data not shown). While GW9662 had relatively minimal effects on AGPAT2 and ABCG2 expression, when the cells were preincubated with GW9662 (0.5–10 μM), there was a concentration-related modulation of CPO-elicited gene expression, most evident with AGPAT2.

Figure 7. Effects of GW9662 on CPO-elicited increases in AGPAT2 and ABCG2 expression.

Gene expression data were measured and expressed as mean fold-change as described in Methods (n=3, ± SEM).

Figure 8 shows blots of protein products of ABCG2 and AGPAT, relative to β-actin.

Figure 8. Protein products of ABCG2 and AGPAT2 collected after one day culture with CPO (250, 500 or 1000 nM).

ABCG2 MW ~72 kDa, AGPAT2 MW~33 kDa, β-Actin~42 kDa. β-Actin was used as the loading control for verification of equivalent protein quantities in each well.

The two genes chosen for protein analysis are both regulated primarily by PPARγ. Figure 8 shows the changes in protein product levels after 24-hour culture of cells in the presence of CPO. While there was little apparent change in protein levels between vehicle and 250 nM CPO, a marked increase in both proteins was observed in cells exposed to 500 and 1000 nM CPO.

4. Discussion

Roughly 20 million pounds of OPs are still used each year in the US alone, with chlorpyrifos use accounting for about 20–40% of the total amount (Atwood and Paisley-Jones, 2017). The molecular initiating event for the acute toxicity of organophosphorus (OP) insecticides is inhibition of acetylcholinesterase (AChE). Extensive risk assessments, typically based on AChE inhibition, along with periodic regulatory re-registrations, have generally supported their overall continued use when applied as recommended (Naughton and Terry, 2018). A multitude of studies have reported neurodevelopmental and other effects of OPs, however that may involve mechanisms distinct from AChE inhibition, adding to the uncertainty and widespread debate on their continued use (Aldridge et al., 2005; Carr et al., 2020; Nomura and Casida, 2011; Berg et al., 2020; Chiu et al., 2021; Guo et al., 2019.

One area that has received more attention recently is the possible role of disruption of eCB and eCB-like metabolite signaling in OP toxicity. The eCB AEA is an agonist at the −α and −γ subtypes of PPARs while the eCB-like metabolites PEA and OEA appear to be selective agonists at the α-subtype (Sun et al., 2007; Sun and Bennett, 2007). Of importance, while the eCBs and eCB-like metabolites activate different signaling pathways, AEA, PEA and OEA all interact metabolically, i.e., they are inactivated by the same hydrolytic enzyme, FAAH. Inhibition of FAAH could modulate a number of physiological processes including signaling through PPARs. Furthermore, while as noted above very little is known about AOPs in the neurodevelopmental effects of chlorpyrifos, a very recent study in children (Chiu et al., 2021) reported that prenatal chlorpyrifos exposures may affect DNA methylation of PPARγ to disrupt neurodevelopment.

A number of studies reported that chlorpyrifos and some other OP anticholinesterases are potent FAAH inhibitors (Casida and Quistad, 2004). FAAH inhibition, along with AChE inhibition, has been reported after chlorpyrifos exposures in adult and immature animals (Carr et al., 2020; Howell et al., 2018; Liu et al., 2015, 2013; Quistad et al., 2001). In fact, some studies suggest FAAH may be more sensitive than AChE to inhibition by chlorpyrifos (via chlorpyrifos oxon) (Carr et al., 2020; Nomura and Casida, 2011). Thus, FAAH inhibition could disrupt the inactivation of eCBs and eCB-like metabolites and contribute to possible non-cholinesterase mechanisms of toxicity elicited by chlorpyrifos, as well as with other OPs.

The human MCF-7 breast cancer cell line has been previously studied to evaluate possible effects of eCBs and eCB-like metabolites on PPAR signaling. The MCF-7 cells express essentially all of the necessary cellular machinery including FAAH Li et al., 2013), CB1 and CB2 cannabinoid receptors (De Petrocellis et al., 1998), PPARα (Zhou et al., 2012), PPARγ (Bonofiglio et al., 2009), OEA and PEA (Li et al., 2013), as well as AChE (Ye et al., 2015). In our studies, CPO inhibited both AChE and FAAH in MCF-7 cells in a concentration- and time- (8 hours, 1 day and 3 days) dependent manner (Figures 2 and 3). While some studies suggest FAAH may be more sensitive in vivo to low dose CPF exposures, AChE and FAAH in MCF-7 cells were relatively similar in sensitivity to CPO under the culture conditions used herein.

Our lab has studied the effects of different OPs on cholinesterase activities and functional/neurobehavioral signs of toxicity for decades (Pope et al., 1990; Chakraborti and Pope, 1992; Zheng et al., 2000; Liu et al., 2021). While it can be difficult to correlate changes in cholinesterase activities obtained in vitro in cultured cells with in vivo exposures eliciting similar enzyme changes, the degree of cholinesterase inhibition noted herein (Figure 3) could be assumed by some to necessarily lead to severe signs and even lethality if observed in an animal model(s). A number of our studies have reported extensive (>90%) inhibition of brain cholinesterase activity however with sublethal exposures to the parent insecticide chlorpyrifos as well as with other OPs. With chlorpyrifos administered by the subcutaneous route in rats, a number of laboratories have observed very extensive brain cholinesterase inhibition with relatively minimal signs of toxicity (e.g., Pope et al., 1992 showed >94% brain cholinesterase inhibition but relatively minor signs of cholinergic toxicity). A more recent publication (Liu et al., 2021) reported that the prototype OP cholinesterase inhibitor diisopropylfluorophosphate (2.25 mg/kg, sc) elicited >90% inhibition of brain cholinesterase activity and severe signs of acute cholinergic toxicity activity, but no lethality. Thus the degree of extensive cholinesterase inhibition noted in MCF-7 cells in the current study does not rule out the possibility that concurrent inhibition of FAAH and associated of PPAR signaling with similar levels of cholinesterase inhibition could realistically occur in vivo. We hypothesized that CPO-induced FAAH inhibition could indirectly increase activation of PPARs by elevating intracellular levels of eCB or eCB-like lipid metabolites, leading to increased transcriptional activity of selected genes. The PPAR-related genes evaluated here included AGPAT2, ACOX1, ABCG2 and LXRa (N1RH3). AGPAT2 encodes a member of the 1-acylglycerol-3-phosphate O-acyltransferase family of enzymes and is regulated by PPARγ (Yao-Borengasser et al., 2008). This enzyme converts lysophosphatidic acid into phosphatidic acid in the de novo pathway of phospholipid synthesis. LXRa (NR1H3) encodes a nuclear receptor that, like PPAR, forms a heterodimer with Retinoid X Receptor (RXR). It is involved in lipid homeostasis and inflammation and is both regulated by and has diverse molecular and metabolic interactions with PPARγ (Chawla et al., 2001). ACOX1 encodes the enzyme acyl coA oxidase 1, catalyzing the initial step in the β fatty acid oxidation pathway, and is regulated by PPARα (Rosen et al., 2008). ABCG2 is a member of the superfamily of ATP-binding cassette (ABC) transporters involved in xenobiotic efflux. It is also known as breast cancer resistance protein (BCRP) and is regulated by PPARα (Hoque et al., 2012; More et al., 2017). Relatively little is known about the potential effects of chlorpyrifos on the expression of these genes. Chlorpyrifos increased the expression of ABCG2 in placental JEG-3 cells and cells isolated from normal human term placenta (Ridano et al., 2017, 2012). Thus, evaluation of changes in expression of these four genes provides a snapshot of PPAR-related signaling changes sensitive to CPF (or more accurately its active form, CPO).

CPO caused marked concentration-dependent increases in the expression of all four PPAR-related genes in a complex, time-dependent manner. ACOX1 was more highly upregulated compared to the other genes at 8 hours after dosing, in particular with the highest concentration of CPO. This could suggest an early increase in endogenous levels of an eCB-like metabolite (e.g., OEA) and activation of PPARα.

Clearly, CPO had a marked effect on PPAR-related gene expression in MCF-7 cells in a concentration- and time-dependent manner. Less clear is the particular subtype(s) of PPAR involved, or the cellular mechanism(s). The PPAR antagonist GW9664 is thought to selectively block PPARγ signaling, and it indeed reversed some of the effects of CPO on increased gene expression in a concentration-dependent manner. CPO exposure was associated with a substantial dose-related reduction in FAAH activity (Figure 3), and the more substantial gene expression changes were noted in cells cultured with the highest concentration. Marked gene expression changes were also observed with the lower CPO exposure conditions (Figure 4), but the magnitude of changes in FAAH was relatively small across the range of CPO doses studied. Further studies should evaluate a wider range of exposure levels for comparison with gene expression.

While CPO may activate PPARs via FAAH inhibition as proposed, it is possible that CPO affects the expression of selected genes through other mechanisms. There is cross-talk between nuclear receptors (Motojima and Hirai, 2006; Pascussi et al., 2008) and some of the the same genes studied here are under the control of other nuclear receptors, e.g. AhR. Interestingly, a previous study from our lab studying gene expression in frontal cortex in neonatal rats one day following an acute chlorpyrifos exposure (2 mg/kg, sc, Ray et al., 2010) indicated AhR signaling was among the pathways most extensively affected (with both RT-PCR and microarray analyses suggesting greater than 50-fold increases in AhR). CPO itself may also be a nuclear receptor activator. It must be stressed that there is no information currently available however on the possible direct interaction of CPO with any PPARs. On the other hand, the reported effects of CPO on FAAH activity suggest a possible PPAR-mediated, non-cholinesterase pathway in the disruption of gene expression related to lipid homeostasis, inflammation and other processes. As multiple studies suggest that chlorpyrifos can influence lipid metabolism, disruption of PPAR signaling may play a role in these receptor-mediated metabolic changes (Olsvik et al., 2019; Wang et al., 2019).

Highlights.

• AChE and FAAH activity in MCF-7 cells were inhibited by CPO in a concentration- and time-dependent manner.

• The expression of selected target genes showed concentration- and time-dependent changes with CPO exposure.

• The PPAR antagonist GW9662 markedly reversed AGPAT2 and ABCG2 expression increases elicited by CPO.

• The data suggest a potential link between chlorpyrifos exposure and altered lipid metabolism, inflammation and immune function mediated by PPARs.