Acetylcholinesterase Inhibition in Rats and Humans Following Acute Fenitrothion Exposure Predicted by Physiologically Based Kinetic Modeling-Facilitated Quantitative In Vitro to In Vivo Extrapolation

Worldwide use of organophosphate pesticides as agricultural chemicals aims to maintain a stable food supply, while their toxicity remains a major public health concern. A common mechanism of acute neurotoxicity following organophosphate pesticide exposure is the inhibition of acetylcholinesterase (AChE). To support Next Generation Risk Assessment for public health upon acute neurotoxicity induced by organophosphate pesticides, physiologically based kinetic (PBK) modeling-facilitated quantitative in vitro to in vivo extrapolation (QIVIVE) approach was employed in this study, with fenitrothion (FNT) as an exemplary organophosphate pesticide. Rat and human PBK models were parametrized with data derived from in silico predictions and in vitro incubations. Then, PBK model-based QIVIVE was performed to convert species-specific concentration-dependent AChE inhibition obtained from in vitro blood assays to corresponding in vivo dose–response curves, from which points of departure (PODs) were derived. The obtained values for rats and humans were comparable with reported no-observed-adverse-effect levels (NOAELs). Humans were found to be more susceptible than rats toward erythrocyte AChE inhibition induced by acute FNT exposure due to interspecies differences in toxicokinetics and toxicodynamics. The described approach adequately predicts toxicokinetics and acute toxicity of FNT, providing a proof-of-principle for applying this approach in a 3R-based chemical risk assessment paradigm.

Human liver microsomes (pooled from 150 donors, mixed gender) and pooled male rat liver microsomes (Sprague-Dawley) were purchased from Corning (Amsterdam, The Netherlands).Pooled female rat liver microsomes (Sprague-Dawley) were ordered from Sigma-Aldrich (St. Louis, MO, USA).Gendermixed rat liver microsomes used in in vitro incubations were pooled from the male and female ones mentioned above.Human plasma (pooled from 25 donors, mixed gender) and rat plasma (Sprague-Dawley, mixed gender) were ordered from BioIVT (West Sussex, UK), and corresponding total protein concentrations were determined using BCA protein assay (Supporting Information).Human whole peripheral blood and rat whole blood (Sprague-Dawley) for in vitro blood AChE inhibition assay were ordered from CTIBiotech (Lyon, France) and Innovative Research Inc. (Novi, MI, USA), respectively, both with K 2 EDTA as an anticoagulant and were gender unspecified.Rat whole blood (Sprague-Dawley rat, mixed gender, with LiHep as an anticoagulant) used for preparing erythrocyte AChE was ordered from BioIVT (West Sussex, UK).Recombinant human acetylcholinesterase (rhAChE) and bovine serum albumin (BSA) were ordered from Sigma-Aldrich (St. Louis, MO, USA).The Pierce™ BCA protein assay kit was ordered from Thermo Fisher (Landsmeer, The Netherlands).

In vitro incubations for metabolic conversions of FNT
In vitro incubations were performed to obtain kinetic parameters for the CYP450-catalyzed conversion of FNT to FNO and MNP using rat and human liver microsomes as described by Zhao et al. (2021) with modifications.Preliminary experiments were performed for optimization to ascertain that reactions were linear with respect to time and microsomal protein concentration (data not shown).To the microsomal incubations with FNT, EDTA and iso-OMPA were added as inhibitors for PON1 and B-esterases, respectively, to prevent untargeted metabolism of the formed FNO in the reaction system.Briefly, the incubations with a total volume of 200 µL contained 50 mM Tris-HCl (pH 7.4), 5 mM MgCl 2 , 1 mM EDTA, 50 µM iso-OMPA, 1 mM NADPH, and FNT at final concentrations ranging from 10 to 250 μM (added from 100 times concentrated stock solutions in acetonitrile).Controls were carried out by replacing NADPH with Tris-HCl.After 1-min pre-incubation in a shaking water bath at 37°C, 1 µL rat or 5 µL human liver microsomes (final concentration 0.1 and 0.5 mg microsomal protein/mL for rat and human samples, respectively) was added to initiate the reaction.After a 2-min incubation, the reaction was terminated by the addition of 20 µL ice-cold HClO 4 (10%, v/v), and samples were kept on ice until further extraction.All incubations were performed in triplicate.

In vitro incubations for metabolic conversions of FNO
Incubations with liver microsomes and plasma were performed to obtain kinetic parameters for PON1mediated detoxification of FNO to MNP based on a modified method (Zhao et al. 2021).Incubation conditions that were suitable for linearity with respect to time and microsomal protein or plasma protein concentration were defined (data not shown).The final incubation mixtures with a total volume of 200 µL contained 50 mM Tris-HCl (pH 7.4), 2 mM CaCl 2 as PON1 activity simulator, and FNO at final concentrations ranging from 25 to 5000 µM (added from 100 times concentrated stock solutions in acetonitrile).Controls differed from test incubations by the absence of liver microsomes or plasma, which was replaced with Tris-HCl.After 1-min pre-incubation in a 37°C water bath, 2 µL rat or 10 µL human liver microsomes (final concentration 0.2 and 1 mg microsomal protein/mL for rat and human samples, respectively) was added to initiate the reaction.For the plasma incubations, 4 µL of rat plasma or 10 µL human plasma (final concentration 1.2 and 3.3 mg plasma protein/mL for rat and human samples, respectively) was used.Reactions were terminated after incubating 10 min for rat samples or 20 min for human samples by adding 20 µL ice-cold HClO 4 (10%, v/v), and samples were kept on ice until further extraction.All incubations were performed in triplicate.

Sample extraction
DIPE extraction of FNT and its metabolites (Zhao et al. 2019;Wang et al. 2022) was performed before UPLC analysis.Briefly, 1 mL DIPE was added to the ice-cold samples and mixed well, then the upper DIPE layer containing the target compounds was transferred into glass tubes.After another two repeated extractions, all collected DIPE layers were combined and evaporated to dryness under an N 2 stream.
Finally, the obtained residue was redissolved in 100 μL methanol, and subsequently analyzed with UPLC-PDA.

UPLC-PDA analysis
The quantification of FNT, FNO and MNP was conducted using a Shimadzu Nexera X2 liquid chromatography system (LC-30AD, Kyoto, Japan), connected with a photodiode array detector (SPD-M30A, Shimadzu, Kyoto, Japan).Chromatographic separation was performed with a Waters Acquity UPLC BEH C18 column (50 mm × 2.1 mm, 1.7 μm) coupled with a Waters Xbridge UPLC BEH C18 pre-column (5 mm × 2.1 mm, 2.5 μm).Column temperature was maintained at 40°C and auto-sampler temperature at 10°C during analysis.Ultrapure water (containing 0.1% TFA, v/v) and acetonitrile were used as mobile phases.A 22-min linear gradient with a flow rate of 0.6 mL/min ran from 10% to 100% acetonitrile over 17 min and held at 100% acetonitrile for 1 min, then returned to 0% over 0.3 min and held at 0% acetonitrile for 1 min, finally returning to the initial conditions and remaining for 2 min before the next injection.Under these conditions the retention times of FNT, FNO and MNP were 7.86 min, 4.65 min and 3.55 min, respectively.Quantification of FNT and FNO was performed by integrating peak areas at 269 nm and for MNP at 315 nm using calibration curves (r 2 > 0.99) prepared with commercial standards.The limit of quantification was 0.1 µM for FNT, FNO and MNP when using an injection volume of 20 µL.During the whole work, the performance of the UPLC-PDA was stable and consistent, and no influence of the presence or absence of the biological matrix (liver microsomes or plasma) on the quantification was observed.

Calculation of kinetic parameters
The kinetic parameters for the conversions of FNT to FNO and MNP in incubations with rat and human liver microsomes, and of FNO to MNP in incubations with rat and human liver microsomes or plasma were determined by fitting the data to a standard Michaelis-Menten equation (Eq S1) where v represents metabolite formation rate in nmol/min/mg microsomal protein or nmol/min/mg plasma protein, [S] the substrate concentration in µM, K m the apparent Michaelis-Menten constant in µM, and V max the apparent maximum rate in nmol/min/mg microsomal protein or nmol/min/mg plasma protein.Data were analyzed in GraphPad Prism (version 5.04, San Diego, CA, USA) and each data point was presented as the mean value ± SEM.

Protein concentration determination for rat and human plasma
Total protein concentration of both rat and human plasma was determined following the manufacturer's protocol (Thermo Fisher 2020).Briefly, 25 µL plasma sample or protein standard solution was incubated with 200 µL working reagents in a 96-well plate (Greiner Bio-One, The Netherlands) at 37ºC for 30 min, after cooling it to room temperature the absorbance at 562 nm was measured.Protein concentrations of rat and human plasma were quantified using a calibration curve prepared with the protein standard in the assay kit.

In vitro AChE inhibition assay with rat and human blood
Inhibition of rat and human erythrocyte AChE by FNT and FNO was determined using the protocol from Kasteel et al. (2021).The final concentrations of FNT and FNO in rat and human blood ranged from 0.05 to 500 µM and from 0.01 to 20 µM, respectively (added from 1000 times concentrated stock solutions in ethanol).Absorbance at 436 nm was measured continuously for 60 min and 10 min to detect the remaining AChE activity in rat and human blood samples, respectively.Assays were conducted in triplicate, and the data obtained were analyzed in GraphPad Prism (version 5.04, San Diego, CA, USA) to define the concentration resulting in 50% inhibition (IC 50 ).

AChE
Based on the method described in Zhao et al. (2021), rat erythrocyte AChE was prepared from rat whole blood and the AChE activity was quantified (0.16 U/mL).The inhibition ability of FNT and FNO towards rhAChE and rat erythrocyte AChE was evaluated using a modified protocol (Zhao et al. 2021), and assays were conducted in triplicate.Briefly, series of increasing concentrations of FNT or FNO in ethanol, 5000 µM chlorpyrifos-oxon in ethanol (CPO, positive control) and 100% ethanol (solvent control) were all diluted 50× in 100 mM sodium phosphate (pH 7.4, containing 0.1 mg/mL BSA).The incubation mixtures with a total volume of 50 µL were incubated in a 96 well-plate and consisted of 44 µL sodium phosphate (pH 7.4), and 5 µL FNO solution (final concentrations ranging from 0.005 to 5 µM), or 5 µL FNT solution (final concentrations ranging from 0.5 to 500 µM), or 5 µL positive control (CPO at a final concentration of 10 µM), or 5 µL solvent control (ethanol at a final concentration of 0.2%).To initiate the inhibition reaction, 1 µL rhAChE (0.16 U/mL) or self-prepared rat erythrocyte AChE (0.16 U/mL) was added into each well.After 15 min incubation at 37℃, 150 µL reaction reagent (mixture of ATC at a final concentration of 150 µM and DTNB at a final concentration of 75 µM) was added, and the absorbance at 412 nm was measured continuously for 10 min at 37℃ to test the remaining AChE activity.The AChE activity was expressed as the remaining AChE activity relative to solvent control (100% activity) and positive control (0% activity) based on the equation (Eq S2): (Eq S2) AChE activity% =  412 ( 10 - 0 ) Test Compound - 412 ( 10 - 0 ) Positive Control  412 ( 10 - 0 ) Solvent Control - 412 ( 10 - 0 ) Positive Control × 100% where A 412 (t 10 -t 0 ) Test Compound is the change of the absorbance at 412 nm between 0 min and 10 min for the test compound; similarly, A 412 (t 10 -t 0 ) Positive Control is the change of the absorbance for the CPO sample, and A 412 (t 10 -t 0 ) Solvent Control the change of the absorbance for the 0.2% ethanol sample.

Sensitivity analysis
A local sensitivity analysis was performed to identify the influential parameters on model outputs.In the current study, the maximum blood FNO concentration was used as the model output, considering that FNO is a more potent AChE inhibitor compared to its precursor FNT and its internal concentration is relevant for the toxicity prediction following acute FNT exposure.The normalized sensitivity coefficients (SCs) were calculated with the equation (Eq S3) where P represents the original parameter value in the PBK model and P' is the parameter value with a 5% increase, C is the model output with the initial parameter values and C' is the model output with a parameter value after a 5% increase.Parameters with an absolute SC greater than 0.1 were considered to have a substantial influence on the model output (WHO 2010).The sensitivity analysis was carried out using oral dose levels of 0.25 mg/kg BW and 0.33 mg/kg BW for rats and humans, respectively, representing the rat and human NOAELs derived by US EPA (2010) and APVMA (2023).

BMD analysis
BMD modeling was used to derive POD values from the predicted in vivo dose-response curves for rats and humans.The EFSA web-tool (https://efsa.openanalytics.eu/)integrated with the R package PROAST (version 70.0) developed by the Dutch National Institute for Public Health and the Environment (RIVM) was used for the BMD analysis.A BMD value resulting in a 10% benchmark response (BMR) change with lower 95% confidence limit was defined as BMDL 10 .Briefly, the continuous data were fitted to a set of models (Exponential, Hill, Inverse Exponential, and Log-Normal family models), and all fitted models excluding the FULL and NULL models were used for model averaging via a weighted average model.More weight was given to the models with lower Akaike's Information Criterion (AIC), and an averaged confidence interval was estimated using the recommended defaults.*: in the reference, it mentioned in the text that "whole-blood concentrations of fenitrothion" were shown in Table 1 and 2, while in the captions of Table 1 and 2, it was written as "fenitrothion plasma concentrations".We used the data in Table 1 and 2 as blood concentrations of fenitrothion (FNT) for the model evaluation, considering that the extraction and quantification were based on the whole blood samples as described in the "Analysis" section (Materials and Methods, Meaklim et al. 2003).

Results of BMD analysis for rats
Results from a BMD analysis for the predicted in vivo dose-response curve for rat erythrocyte AChE inhibition upon acute oral FNT exposure (Figure 7a).The table and figures present the characteristics of fitted models, the weights for model averaging and the final benchmark dose for 10% effect with the 95% lower-upper confidence limit values of the benchmark dose (BMDL-BMDU).

Figure S1
Figure S1Results of a local sensitivity analysis for the predicted maximum blood FNO concentration at dose levels of 0.25 mg/kg BW (rats) and 0.33 mg/kg BW (humans).Model parameters with a normalized SC with an absolute value higher than 0.1 are shown.VFc, fraction of fat tissue; VLc, fraction of liver tissue; VBc, fraction of blood; VSc, fraction of slowly perfused tissue; QFc, fraction of blood flow to fat tissue; QSc, fraction of blood flow to slowly perfused tissue; PFFNO, fat:blood partition coefficient of FNO; PSFNO, slowly perfused tissue:blood partition coefficient of FNO; Fa, fraction of dose absorbed; kaI, absorption rate constant from intestine to liver; ksI, transfer rate constant from stomach to intestine; MPL, liver microsomal protein yield scaling factor; Vmax1c, maximum rate for conversion of FNT to FNO; Km1, Michaelis-Menten constant for conversion of FNT to FNO; Vmax2c, maximum rate for conversion of FNT to MNP; Km2, Michaelis-Menten constant for conversion of FNT to MNP; Vmax3c, maximum rate for conversion of FNO to MNP in liver; Km3, Michaelis-Menten constant for conversion of FNO to MNP in liver; MPB, plasma protein concentration; Vmax4c, maximum rate for conversion of FNO to MNP in plasma; Km4, Michaelis-Menten constant for conversion of FNO to MNP in plasma; GFR, glomerular filtration rate; FupFNO, unbound fraction of FNO in plasma; FupliverFNO, adjusted unbound fraction of FNO in plasma for hepatic biotransformation.

Figure S2
Figure S2 PBK modeling-based predictions of maximum blood FNT and FNO concentrations in rats and humans under increasing FNT dose levels (0.01 -100 mg/kg BW).

Figure S3
Figure S3 Activity of (a) self-prepared rat erythrocyte AChE and (b) recombinant human AChE upon incubation with increasing FNT and FNO concentrations.Results are presented as means ± SEM from three independent experiments.The IC 50 values of FNT are 8.24 and 30.93 µM for rats and humans, and the IC 50 values of FNO are 0.18 and 0.25 µM for rats and humans, respectively.

Table S1
Summary of physiological and physicochemical parameters used for the rat and human PBK models.

Table S2
Summary of LogP and pKa values of FNT, FNO and MNP.Values used in the current study are presented in bold.

Table S3
Summary of available in vivo kinetic studies in rats and humans following single FNT exposure.

Table S4
Summary of in vitro kinetic parameters for the biotransformation of FNT and FNO.
*: calculated as V max /K m .