BENOXACOR IS ENANTIOSELECTIVELY METABOLIZED BY MICROSOMES AND CYTOSOL FROM THE HUMAN LIVER

Abstract

Benoxacor is a safener added to current-use herbicide formulations to protect the target crop from herbicidal toxicity. It is an emerging environmental contaminant that has been detected in surface waters, raising the possibility of human exposure via drinking water. Because it is not subject to the same regulations as active pesticide ingredients, its metabolism and toxicity in humans have not been studied. Here we investigate the enantioselective metabolism of benoxacor in human subcellular fractions. Pooled human liver microsomes (pHLM) and cytosol (pHLC) were incubated with racemic benoxacor for up to 30-minutes. Gas chromatographic analyses were used to measure the enantioselective depletion of benoxacor. pHLMs with and without nicotinamide adenine dinucleotide phosphate (NADPH, co-factor for cytochrome P450 enzymes [CYPs]) and pHLC with glutathione (GSH, co-factor for glutathione S-transferases [GSTs]) metabolized benoxacor. These results demonstrate that microsomal CYPs, microsomal carboxylesterase (CESs), and cytosolic GSTs metabolize benoxacor. Females were predicted to have a higher clearance of benoxacor by GSTs than males. Male and female pHLM incubations with NADPH showed enrichment of the first eluting benoxacor enantiomer (E₁-benoxacor). pHLM incubations without NADPH and pHLC incubations with GSH showed an enrichment of the second eluting enantiomer of benoxacor (E₂-benoxacor). Our results indicate that human hepatic microsomal and cytosolic enzymes enantioselectively metabolize benoxacor, a fact that needs to be considered when investigating human exposures and toxicities of benoxacor.

1. Introduction

Benoxacor is a dichloroacetamide safener co-formulated with the chloroacetanilide herbicide S-metolachlor to prevent herbicidal damage in crops, especially in areas with cool, wet soil that reduces the ability of the target crops (e.g., corn) to metabolize metolachlor.^{1, 2} The application of benoxacor alongside the herbicide promotes rapid in vivo transformation of the herbicide to less toxic metabolites within cereal crops, but not weed species, thus protecting the target crop from herbicidal damage.³ As a result, benoxacor is widely used in the Midwest of the United States (US). Benoxacor, as an emerging environmental contaminant, has been detected in surface water in Iowa and Illinois in the US.² Therefore, it is likely that humans are exposed to benoxacor via their drinking water and their diet. However, little is known about the metabolism and toxicity of benoxacor and other dichloroacetamide safeners in humans, and possible biomarkers of exposure. This knowledge gap is, in part, due to their classification as “inert” under the U.S. Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA).

The statutory nomenclature within the FIFRA framework creates confusion, with many people believing that “inert” ingredients are harmless.⁴ Currently, only seven short-term acute toxicity tests are required to register a pesticide formulation in the U.S.; however, up to 50% of inert ingredients may pose a moderate health risk.⁴ The inert classification also prevents safeners from being held to the same extensive scrutiny as their co-formulated herbicides even though safeners also exhibit biological activity.^{2, 3} This is a major issue since inert ingredients may increase the exposure to pesticides, increase the toxicity of pesticides, or they may have adverse effects on their own.⁴ For example, high-throughput screening data suggest that benoxacor is a hepatotoxicant. Interestingly, benoxacor displayed little toxicity following sub-acute oral exposure to male mice, likely due to its rapid metabolism.⁵ Moreover, the assessment of exposures to benoxacor is difficult because manufacturers typically do not disclose the composition of pesticide formulations.

Limited information is available about the biotransformation of safeners, such as benoxacor, in mammals. The metabolism of the chloroacetanilide herbicides typically co-formulated with benoxacor provides insights into possible metabolism pathways for benoxacor because of structural similarities. These herbicides are metabolized by cytochrome P450 enzymes (CYPs), glutathione S-transferases (GSTs), and carboxylesterases (CESs) in plants and mammals.⁶ As with the herbicides, GSTs are responsible for the metabolism of the structurally related dichloroacetamide safeners, like benoxacor, in plants.^{3, 7} GSTs, as well as CYPs and CESs, also play a role in the metabolism of benoxacor by liver microsomes and cytosol from rats⁸ and mice.⁵ Our earlier studies also demonstrate the benoxacor, which contains a chiral center at the 3-position of the benzoxazine group and is produced and formulated as a racemate,⁹ undergoes enantioselective metabolism. However, it is unknown if the metabolism of benoxacor is also enantioselective in humans.

Herein, we elucidate the enantioselective metabolism of benoxacor using pHLM and pHLC in vitro. Our results demonstrate that benoxacor is enantioselectively metabolized by CYPs, GSTs, and CESs, and the metabolism in humans is different from the metabolism in other species, such as the laboratory rat and mouse.

2. Materials and methods

2.1. Chemicals

Benoxacor (Fig. 1; 98.4% purity; InChIKey: PFJJMJDEVDLPNE-UHFFFAOYSA-N) and trifluralin (99.3% purity; InChIKey: ZSDSQXJSNMTJDA-UHFFFAOYSA-N; recovery standard) and were obtained from Sigma-Aldrich (St Louis, MO, USA). Acenaphthene (99.3% purity; InChIKey: CWRYPZZKDGJXCA-UHFFFAOYSA-N; Accustandard, New Haven, CT, USA) was selected as internal standard (volume corrector) in a modification of a validated method for the analysis of safeners.² 1-Aminobenzotriazole (CYP enzyme inhibitor¹⁰) was purchased from TCI (≥98% purity; InChIKey: JCXKHYLLVKZPKE-UHFFFAOYSA-N; Portland, OR, USA). Benzil (CES inhibitor^{11, 12}) was obtained from ACROS Organics (≥99% purity; InChIKey: WURBFLDFSFBTLW-UHFFFAOYSA-N; New Jersey, NJ, USA). Ethacrynic acid (GST inhibitor^{13, 14}) was purchased from Enzo Life Sciences (≥98%; InChIKey: AVOLMBLBETYQHX-UHFFFAOYSA-N; Farmingdale, NY, USA). Pesticide-grade ethyl acetate was purchased from Fisher Chemical (Fair Lawn, NJ, USA). Chemicals for the preparation of buffers were purchased from commercial sources and were of ACS grade (≥98% purity). pHLM (lot numbers 1210079 and 1210270 for pooled liver microsomes from female and male humans, respectively) and pHLC preparations (lot numbers 0710245 and 0710493 for pooled liver cytosol from female and male human livers, respectively) were purchased from Sekisui XenoTech (Kansas City, KS, USA).

Figure 1.

Chemical structure of the benoxacor enantiomers.

2.2. Microsomal Incubations

pHLM prepared from female and male donors were incubated for up to 30-minutes in triplicate at 37 °C, as described previously.⁸ The incubations were composed of phosphate buffer (0.1 M, pH 7.4) with magnesium chloride (3 mM), 0.1 mg/mL microsomal protein, 0.5 mM NADPH, and 5 μM benoxacor for a total incubation volume of 5 mL (6.5 μg benoxacor total per incubation). At 5-minute intervals, 500 μL aliquots were taken from the incubation, and the reaction was quenched with 2 mL of ice-cold 1% formic acid.

2.3. Cytosolic Incubations

Cytosolic incubations using pHLC from female and male donors were conducted as described previously.^{8, 15} Briefly, cytosolic incubations included 0.1 mg/mL cytosolic protein, 5 μM benoxacor, 5 mM reduced GSH (Sigma Aldrich, St. Louis, MO, USA), phosphate buffer (0.1 M, pH 7.4), and magnesium chloride (3 mM) at 37 °C for a total incubation volume of 5 mL. Aliquots (500 μL) were removed at 5-minute intervals and quenched with 2 mL of ice-cold 1% formic acid.

2.4. Extraction of Benoxacor

Sample aliquots from microsomal or cytosolic incubations were placed in a 110 °C furnace for 10-minutes to facilitate the release of protein-bound benoxacor, as reported.⁸ Samples were allowed to cool to ambient temperature and spiked with 500 ng of trifluralin as the recovery standard. Samples were extracted with 2 mL ethyl acetate and centrifuged at 1811 g for 5-minutes to facilitate phase separation. The organic layer was collected, and the samples were re-extracted with 2 mL ethyl acetate. The organic extracts were combined and washed with 2 mL of 1% potassium chloride to remove the remaining proteins and residual formic acid. The organic layer was transferred to a new vial containing 0.25 mg of magnesium sulfate to remove residual water. The extracts were concentrated to approximately 500 μL under a gentle stream of nitrogen and spiked with 500 ng acenaphthene as the internal standard.

2.5. Gas chromatographic analyses

Analyses were performed on an Agilent 6890 N gas chromatograph coupled to an Agilent 5975 MS in SIM mode using acenaphthene and trifluralin as internal and recovery standards, respectively.⁸ The injector was set at 280 °C, and samples were injected using splitless injection. An Agilent SLB-5MS column (30 m length, 0.25 mm inner diameter, 0.25 μm film thickness) was used with a constant helium flow rate set at 1 mL/min. The oven program was as follows: Initial temperature, 50 °C hold for 1 min, 15 °C/min to 240 °C, hold for 13 min, 15 °C/min to 300 °C, hold for 10 min. The temperatures of the MS source, the MS quadrupole, and the transfer line were 230 °C, 150 °C, and 280 °C, respectively. The quantitation ions (quant) and qualitative (qual) ions were m/z 120 and 259 for benoxacor, m/z 264 and 306 for trifluralin, and m/z 153 and 154 for acenaphthene, as described.²

We used an Agilent 7890 A gas chromatograph coupled to a ⁶³Ni-μECD detector for the enantioselective analyses.⁸ The benoxacor enantiomers were baseline separated on an Agilent CP-Chrialsil Dex CB (CB) column (25 m length, 0.25 mm inner diameter, 0.25 μm film thickness) using the following temperature settings: 50 °C, hold for 1 min, 10 °C/min to 145 °C, hold for 50 min, 15 °C/min to 200 °C, hold for 12 min. A constant helium flow rate of 2 mL/min was utilized, and the injector temperature was maintained at 250 °C. The enantiomeric fraction was calculated with the equation EF = A₁/(A₁+A₂), where A₁ is the area of the first eluting enantiomer of benoxacor (E₁-benoxacor), and A₂ is the area of the second eluting enantiomer of benoxacor (E₂-benoxacor). The enantiomers of benoxacor were identified based on their elution order on the chiral columns because the optical rotation and absolute configuration of E₁- and E₂-benoxacor are unknown.

2.6. Quality assurance/quality control

The following controls were collected in parallel to each replicate: benoxacor spiked buffer, vehicle spiked (DMSO) buffer, benoxacor spiked microsomes or cytosol without respective co-factor, heat-inactivated microsomes or cytosol containing the respective co-factors and benoxacor, and benoxacor spiked buffer containing NADPH or GSH. No loss of benoxacor was seen in any of our controls.

All GC/MS analyses were performed using a Standard Operating Procedure to ensure the reproducibility of the analyses. The responses of benoxacor, trifluralin (recovery standard), and acenaphthene (volume corrector) were linear in the GC/MS analyses within the concentration range used in this study (3.9 to 1000 ng/mL, R² ≥ 0.9983). The method limit of detection (LOD) was 0.4 ng. This LOD was calculated using the IUPAC method¹⁶ with blank buffer samples (n=10) using the equation LOD = mean blanks + k* Standard deviation of the blanks where k represents the student’s t value at the 99% confidence interval with degrees of freedom 10–1 = 9. Benoxacor resolution (R) was determined on an Agilent Chiralsil-Dex CB column as the difference of peak enantiomer retention times divided by the average peak width $R=\frac{t_{R2}-t_{R1}}{\frac{1}{2}\left(W_1+W_2\right)}$. Analyses were only integrated if baseline resolution was (R ≥ 1). After correcting the recovery standard’s loss, there was no loss of benoxacor in any control incubations (corrected benoxacor recovery = 101 ± 2%, (n = 24)). Actual recovery of trifluralin across all samples, a measure of reproducibility and precision, averaged 88 ± 5% and ranged between 81–104% (n = 432). For stereoselective metabolism analysis, racemic benoxacor was reported as EF = 0.48 ± 0.01 (n = 16).

2.7. Calculation of intrinsic and scaled clearance

We calculated apparent intrinsic clearance (CL_int) and scaled intrinsic clearance (scaled CL_int) for both microsomal and cytosolic incubations using previously reported methods.^{8, 17} Briefly, the percentages of benoxacor were natural logarithm transformed and plotted against time to calculate the initial rates of benoxacor depletion (k_dep). The time when 50% of the benoxacor was depleted (in vitro t_1/2) was determined by linear regression analysis. CL_int was predicted using equation (1), where V represents the volume of the incubation:¹⁷

$${\text{CL}}_{\text{int}}={\sum }_{\text{i}}^{\text{n}}〈\frac{{\text{V}}_{\text{max},\text{i}}}{{\text{K}}_{\text{m},\text{i}}}〉=\frac{0.693\times \text{V}}{\text{in}\text{vitro}{\text{t}}_{\frac{1}{2}}}$$ (1)

CYP and GST k_dep values were corrected for CES activity. The CL_int was corrected for microsomal or cytosolic protein concentration. The CL_int values were then used to predict the scaled CL_int using literature values for microsomal protein per gram liver (40 mg microsomal protein/g liver) and cytosolic protein per gram liver (80.7 mg cytosolic protein/g liver).¹⁷

2.8. Statistical analyses

All statistical analyses were performed on Graphpad Prism 9.0.0. The zero-minute time point was compared to later time points by an unpaired t-test. Incubations with pHLM and pHLC prepared from female livers were compared to analogous incubations with subcellular fractions prepared from male livers with an unpaired t-test. Benoxacor levels and enantiomeric fractions determined at the 30-min time point were compared to controls using a one-way Analysis of Variance (ANOVA). Tukey’s test for multiple comparisons was used to confirm the one-way ANOVA results. Tests were considered significant when p < 0.05.

3. Results and discussion

3.1. Metabolism of benoxacor by human liver microsomes in the presence of NADPH

Xenobiotics, including pesticides, undergo oxidative metabolism by CYPs, a superfamily of hemeproteins.¹⁸ For example, rat and human CYPs metabolize the chloroacetanilide herbicides metolachlor and acetochlor.¹⁹ Both herbicides are structurally similar to benoxacor and are co-formulated with this safener. Benoxacor is metabolized by hepatic CYPs in the rat and mouse,^{5, 8} whereas CYPs do not metabolize the safener dichlormid.¹⁵ Herein, we studied the oxidative metabolism of benoxacor using pHLM prepared from male and female donors.

Benoxacor was depleted in incubations containing pHLM prepared from female and male donors in the presence of NADPH (Fig. 2, Table S1). The depletion of benoxacor was linear over the 30-minute incubation time (Fig. 2A). After 30 minutes, 59 ± 1% and 43 ± 1% of benoxacor were left in incubations with pHLM from female and male donors. Under comparable experimental conditions, rat and mouse liver microsomes also readily metabolized benoxacor. Across species, the percent of benoxacor remaining in the incubations followed the rank order of female human > female rat > male rat ~ male human > male mouse (Table 1).^{5, 8} Thus, studies in mice may show lower toxicity of benoxacor because of its more rapid oxidative metabolism in mice than humans, as we suggested previously.⁵

Figure 2.

CYPs and CESs deplete Benoxacor in incubations with pooled microsomes from human livers. (A) Metabolism of benoxacor with and without the addition of NADPH. Comparison of (B) the protein-adjusted rate of depletion of benoxacor at 30 minutes for liver microsomes prepared from female and male livers. The percentage of benoxacor in microsomal incubations from (C) female and (D) male livers and control incubations at 30 minutes; see text for more details. Data are reported as the mean ± standard deviation for three independent replicates; for some data points, the standard deviation is smaller than the size of the symbols. An unpaired t-test determined significance in panel B. One-way ANOVA was used to determine significance in panels C&D. A Tukey’s multiple comparisons test confirmed a significant decrease in active incubations (p < 0.0001) and active incubations without NADPH (p <0.0001) (panels C&D). **** p < 0.0001 (one-way ANOVA); ### p = 0.0001. The percentages of benoxacor depletion are summarized in Table S1 & Table S2. NADPH, nicotinamide adenine dinucleotide phosphate; IM, heat-inactivated microsomes.

Table 1.

Comparison of the percent of benoxacor remaining in 30-minute incubations with human, rat, and mouse microsomes or cytosol. The data for experiments with rat and mouse subcellular fractions have been reported previously.^{5, 8}

Subcellular fraction	Female Human	Male Human	Female Rat	Male Rat	Male Mouse
	[%]	[%]	[%]	[%]	[%]
Microsomes with NADPH (microsomal CYP)	59 ± 1	43 ± 3*	48 ± 1	45 ± 1	36 ± 2
Microsomes without NADPH (microsomal CES)	89 ± 2	63 ± 1*	98 ± 2	94 ± 1*	77 ± 1
Cytosol with GSH (cytosolic GST)	73 ± 2	86 ± 1*	18 ± 1	9 ± 1*	43 ± 2
Cytosol without GSH (cytosolic CES)	100 ± 1	100 ± 1	100± 2	100 ± 1	100 ± 2

Data are reported of the average ± standard deviation of at least three determinations.

^*Significantly different from incubations with the corresponding subcellular fraction for female animals (unpaired t-test).

The rate of depletion was significantly slower in microsomal incubations with pHLM from females (18 ± 1 ng benoxacor/mg protein/min) than from males (25 ±1 ng benoxacor/mg protein/min) in the presence of NADPH (p = 0.0001) (Fig. 2B; Table 2). It is not surprising that the oxidation of benoxacor by pHLM is sex-dependent since there are well-documented sex-related differences in xenobiotic processing genes in humans, including the hepatic expression of CYPs.^20–22 In contrast, the depletion rate of benoxacor exhibited no significant sex differences in studies with rat hepatic microsomes.⁸ The benoxacor depletion rates were comparable for incubations with rat microsomes or pHLM, irrespective of sex (18 to 27 ng benoxacor/mg protein/min; Table 2). In contrast, in experiments with microsomes from male mice, the depletion rate was over 2-fold higher than in analogous studies using human or rat microsomes (61 ± 1 ng benoxacor/mg protein/min).

Table 2.

Comparison of the depletion rates of benoxacor (ng benoxacor/mg protein/min) in incubations with human, rat, and mouse microsomes or cytosol. The depletion rates of benoxacor in studies with rat and mouse subcellular fractions have been reported previously.^{5, 8}

Subcellular fraction	Female Human	Male Human	Female Rat	Male Rat	Male Mouse
Microsomes with NADPH (microsomal CYP)	18 ± 1	25 ± 1*	24 ± 1	27 ± 2	61 ± 3
Microsomes without NADPH (microsomal CES)	5 ± 1	13 ± 1*	N/A	3 ± 1*	12 ± 1
Cytosol with GSH (cytosolic GST)	12 ± 1	6 ± 1*	7 ± 1	9 ± 1*	25 ± 1
Cytosol without GSH (cytosolic CES)	N/A	N/A	N/A	N/A	N/A

Data are reported of the average ± standard deviation of at least three determinations.

^*Significantly different from incubations with the corresponding subcellular fraction for female animals (unpaired t-test).

Control incubations with heat-inactivated pHLM or without pHLM showed no disappearance of benoxacor (101 ± 1% and 100 ± 1% for microsomes from female and male human livers, respectively) (Fig. 2C & 2D). These results are consistent with the oxidative metabolism of benoxacor by human hepatic CYPs.

3.2. Metabolism of benoxacor by human liver microsomes in the absence of NADPH

CESs in microsomes can also metabolize a broad range of pesticides with amide, ester, and thioester bonds, including carbamate, organophosphate, and pyrethroid pesticides.^{23, 24} CESs are highly expressed in the liver endoplasmic reticulum. Interspecies differences, including different activities and the number of CES isoforms, have been noted between many model organisms, including humans.^{25, 26} For example, CES1 and CES2 are expressed in the human liver. In contrast, rats do not express CES2 within the liver.²⁶ CESs make only a minor contribution to the metabolism of benoxacor in incubations with hepatic microsomes from male rats.⁸ Based on these previous studies, the involvement of CESs in benoxacor metabolism was investigated in incubations with pHLM from female and male donors.

In incubations with pHLM, levels of benoxacor decreased linearly for up to 30-minutes in the absence of NADPH (Fig. 2, Table S2). Without the co-factor NADPH, 89 ± 2% and 70 ± 2% of benoxacor were left after a 30-minute incubation with pHLM from female and male donors, respectively. (Fig. 2A, Table S2). In the absence of NADPH, the depletion of benoxacor was significantly lower in hepatic microsomes from females than from male humans (p = 0.002). No metabolism was observed in control experiments without pHLM or with heat-inactivated pHLM. These findings provide compelling evidence that human microsomal CESs and microsomal CYPs metabolize benoxacor in a sex-dependent manner.

This study provides further evidence that the metabolism of benoxacor by microsomal CESs is species-specific. The depletion of benoxacor was higher in experiments with pHLM than in analogous studies with rat liver microsomes, where 98 ± 2% and 94 ± 1% of benoxacor remained in incubations with liver microsomes from female and male rats, respectively.⁸ The depletion of benoxacor (77 ± 1%) in the absence of oxidative metabolism in analogous studies with microsomes from male mice was within the range observed for pHLM.⁵ Overall, the percentage of benoxacor remaining in microsomal incubations without NADPH follows the approximate rank order female rat ~ male rat > female human > male mouse > male human (Table 1). Consistent with this rank order, the highest benoxacor depletion rates were observed in experiments with male human and male mouse microsomes (Table 2). Conversely, the lowest depletion rate was observed in studies using female rat microsomes, where essentially no CES-mediated metabolism occurred.

Previous work has also shown inter-species differences between rats and humans in the CES-mediated metabolism of clopidogrel.²⁷ Such species differences could play a role in future toxicity studies. For example, our results suggest that studies in rats may underestimate the toxic effects of CES-mediated metabolites. Also, there is likely to be organ-specific metabolism, depending on which class of carboxylesterase is involved in the metabolism of benoxacor, a possibility that warrants further attention.²⁸ Similarly, the anticancer drug CPT-11 is differentially metabolized in the liver, intestine, kidney, and lung due to the organ-specific expression of the two families of CESs.²⁹

3.3. Metabolism of benoxacor in human liver cytosol in the presence of GSH

Cytosolic phase II metabolic enzymes often metabolize pesticides and, thus, may also contribute to the metabolism of benoxacor.^{30, 31} Indeed, benoxacor not only induces GSTs in plants^{9, 32} but is also metabolized by plant GSTs.⁷ For example, GSTs in Zea mays metabolize benoxacor to glutathione conjugates and formylcarboxamide metabolites.^{7, 33} A couple of studies revealed that benoxacor and another dichloroacetamide safener, dichlormid, are metabolized by mammalian GSTs.^{8, 15} Moreover, metabolism of benoxacor by GSTs was expected because GSTs readily conjugate structurally related chloroacetanilide herbicides, such as metolachlor and acetochlor.³⁴

Consistent with these earlier studies, benoxacor was depleted linearly over a 30-minute incubation time in experiments with pHLC when GSH was added (Fig. 3, Table S3). By 30-minutes, 73 ± 1% and 86 ± 1% of benoxacor remained in incubations with pHLC from female and male donors, respectively (Fig. 3A). This difference in the percent of benoxacor remaining in these incubations was statistically significant (p = 0.0003). Benoxacor was also metabolized in the presence of GSH in metabolism studies with liver cytosol from male and female rats and mice.^{5, 8} When comparing benoxacor metabolism across species, benoxacor was more rapidly metabolized in the presence of GSH by cytosol prepared from rats, with a rank order of male and female rat > male mouse > male and female human liver (Table 1).

Figure 3.

GSTs metabolize Benoxacor in incubations with pooled liver cytosol prepared from human livers. (A) Depletion of benoxacor over a 30-minute time-course in cytosol prepared from human livers with and without the addition of GSH. Comparison of (B) the protein-adjusted rate of depletion of benoxacor at 30 minutes for cytosol from female and male livers. The percentage of benoxacor in cytosolic incubations with active cytosol prepared from (C) female and (D) male human liver and controls at 30 minutes; see text for more details. An unpaired t-test determined significance in panel B. Significance in panels C & D were determined by one-way ANOVA. A Tukey’s multiple comparisons test confirmed a significant decrease in benoxacor in active incubations with GSH (p < 0.0001) (panels C&D). The percentages of benoxacor depletion are summarized in Table S3 & Table S4. Data are reported as the mean ± standard deviation for three independent replicates; for some data points, the standard deviation is smaller than the size of the symbols. GSH, glutathione; IC, heat-inactivated cytosol; ** p = 0.0015; **** p < 0.0001 (one-way ANOVA); ns, not significant.

The benoxacor depletion rate was sex-dependent, with pHLC from females having twice the depletion rate as pHLC from males (12 ± 1 vs. 6 ± 1 ng benoxacor/mg protein/min, respectively) (Fig 3B; Table 2). In analogous studies with rat liver cytosol, the rate of benoxacor depletion was higher in cytosol from male than female rats, highlighting another species difference between rats and humans.⁸ When comparing depletion rates across species, the highest benoxacor depletion rate was observed with cytosol from the male mouse liver. pHLC and rat liver cytosol metabolized benoxacor with depletion rates over 2-fold lower than those found with the mouse cytosol (Table 2).

Control incubations containing heat-inactivated human cytosol preparations or no cytosol showed no significant disappearance of benoxacor (101 ±1% and 101 ±1%, respectively, after 30 minutes) (Figs. 3C & 3D). These control experiments demonstrate that the benoxacor depletion is due to metabolism by cytosolic GSTs.

3.4. Metabolism of benoxacor in human liver cytosol without GSH

Cytosolic CESs can contribute to the overall clearance of pesticides, for example, pyrethroid insecticides.³⁵ In the present study, benoxacor was not metabolized in experiments with pHLC without the addition of GSH (Fig. 3A, Table S4). This finding indicates that cytosolic CESs did not contribute to the hepatic metabolism of benoxacor in humans. Similarly, benoxacor was not metabolized in our earlier experiments with cytosol from rats or mice without the addition of GSH.^{5, 8}

3.5. Predicted Intrinsic and Scaled Intrinsic Clearance

The data from the human hepatic microsomal and cytosolic incubations were used to approximate the intrinsic and the scaled intrinsic clearance of CESs, CYPs, and GSTs in the human liver (Table 1). This approach provides an estimate for metabolic clearance of a xenobiotic when more sophisticated metabolic schemes are unavailable.³⁶ In the human liver, the estimated clearance of benoxacor followed the order GSTs > CYPs > microsomal CESs for subcellular fractions from both female and male donors. GSTs made the highest contribution to the clearance of benoxacor in humans. Still, all tested enzymes except for cytosolic CESs played a role in the metabolism of benoxacor. In metabolism studies with microsomes from male mice, GSTs also made major contributions to the clearance of benoxacor, followed by CYPs and microsomal CESs.⁵ CYPs followed by GSTs made major contributions to the hepatic clearance of benoxacor in rats, with a negligible role of microsomal CESs.⁸

The estimated hepatic clearance of benoxacor by microsomal CESs and cytosolic GSTs varied by sex in experiments with pHLM and pHLC. Briefly, the clearance by microsomal CESs was significantly lower in pHLM from female than male donors (unpaired t-test, p = 0.001). The clearance of benoxacor by GSTs was significantly higher in pHLC from female than male donors (unpaired t-test, p < 0.0001). In analogous experiments using rat liver microsomes, significant sex differences in the hepatic clearance of benoxacor were observed for GSTs but not CYPs or CESs.⁸ The clearance of benoxacor by GSTs was approximately 1.6-time higher in the male than the female rat liver, which is opposite to the sex-dependent clearance of benoxacor by GSTs in humans.

These intrinsic clearance values provide insights into the clearance of benoxacor in the human liver and suggest that sex differences may differently influence the disposition and toxicity of benoxacor in humans vs. rats. However, it is important to point out the limitations of these estimations. First, the scaling factors are based on literature values and may not accurately reflect the protein yields of the pHLM and pHLC preparations used in this study. Moreover, the calculations do not consider the complex interactions of the drug-metabolizing enzymes in the liver. Despite these limitations, the estimates reveal differences in the metabolism of benoxacor between enzymes from humans and toxicologically relevant animal models, which should be considered when studying the toxicity of benoxacor.

3.6. Enantiomeric enrichment from oxidative metabolism in microsomal incubations

The enantioselective metabolism of pesticides and other environmental contaminants by CYPs has received some attention in the literature.^37–39 The enantioselective oxidation of inert agrochemicals, like benoxacor, by mammalian CYPs has received much less attention. This study aims to close this critical knowledge gap. We observed a linear increase in the enrichment of E₁-benoxacor over the 30-minute time course in incubations with pHLM, irrespective of donor sex (Fig. 4, Table S5). At 30-minutes, the EF of benoxacor was 0.61 ± 0.01 and EF = 0.67 ± 0.01 for incubations with pHLM from female and male donors, respectively (Fig 5A). The EF value for incubations with pHLM from female donors was significantly lower than the EF values observed in incubations with pHLM from male donors (p < 0.0001) (Fig. 5B). All controls showed no deviation from the EF value reported for racemic benoxacor (EF = 0.48) (Fig. 5C & 5D).

Figure 4.

Representative gas chromatograms show time-dependent changes in the enantiomeric enrichment of benoxacor in incubations with human liver microsomes from (A) female and (B) male livers. Incubations with microsomes from human livers showed an enrichment of the benoxacor enantiomer eluting first on the enantioselective column (E₁-benoxacor). Microsomal incubations contained 0.1 mg/mL protein, 0.5 mM NADPH, and 5 μM benoxacor in incubation buffer. All samples were incubated for 30-minutes, and aliquots were taken every 5-minutes for analysis. After extraction, an Agilent ChiralsilDex-CB column was used to separate the enantiomers of benoxacor on an Agilent 7890 A gas chromatograph coupled to a ⁶³Ni-μECD detector as described under Materials and Methods.

Figure 5.

Benoxacor is enantioselectively metabolized by pooled liver microsomes prepared from human livers with or without the addition of NADPH. (A) Change in the enantiomeric fraction of benoxacor over time in 30-minute incubations with microsomes prepared from female and male human livers. Comparison of the calculated enantiomeric fraction of benoxacor at 30 minutes between (B) incubations with microsomes prepared from female and male human livers. Experiments with active microsomes prepared from (C) female and (D) male human liver microsomes and controls; see text. Data are the mean ± standard deviation for three independent replicates; for some data points, the standard deviation is smaller than the size of the symbols. The EF values at all time points were significantly different from the 0-minute time-point (p < 0.0020). EF values were also significantly different between incubations with microsomes from female and male humans at all time points≥ 5 minutes with NADPH (p = 0.0128) and at all time points ≥ 5 minutes (p < 0.0090) without NADPH. Significance in panel C was determined using an unpaired t-test. Significance in panels C & D was determined with one-way ANOVA. Tukey’s multiple comparisons test confirmed a significant increase in the EF values in active preparations from female (p < 0.0001) and male donors (p < 0.0001). **** p < 0.0001 (one-way ANOVA); #### p < 0.0001 (unpaired t-test). EF values are reported in Table S5 & Table S6. The dotted lines represent EF = 0.5. NADPH, nicotinamide adenine dinucleotide phosphate; IM, heat-inactivated microsomes.

Like our studies with pHLM, E₁-benoxacor was enriched in studies with rat and mouse liver microsomes.^{5, 8} Unlike in the experiments with pHLM, E₁-benoxacor showed a more pronounced enantiomeric enrichment in analogous incubations with female than male rat liver microsomes, with EF values of 0.68 ± 0.01 and 0.60 ± 0.01, respectively. These observations indicate clear species differences in the enantioselective metabolism of benoxacor. At the 30-minute time point, the rank order of the average overall change in the EF for the microsomes from the three species was female rat (0.68 ± 0.01) ~ male human (0.67 ± 0.01) > > female human (0.61 ± 0.01) > male mouse (0.60 ± 0.01) ~ male rat (0.60 ± 0.01).

3.7. Enantiomeric enrichment from the non-oxidative metabolism in microsomal incubations

The hydrolysis of xenobiotics by microsomal CESs can be highly stereoselective, as shown for methylphenidate and naproxen methyl ester.^{40, 41} Therefore, we investigated the enantioselective metabolism of benoxacor by human microsomal CESs. In contrast to studies with NADPH, where E₁-benoxacor was enriched, E₂-benoxacor was enriched in incubations with pHLM from female and male donors without NADPH. Similar to analogous studies with rat liver microsomes,⁸ sex differences were observed in the EF values of benoxacor in studies with pHLM in the absence of NADPH (Fig. 5A, Table S6). For example, the average EF values of the remaining benoxacor at the 30-minute time point were 0.43 ± 0.01 and 0.40 ±0.01 for incubations with pHLM from female and male donors, respectively (Fig. 5B).

The enrichment of E₂-benoxacor in experiments with pHLM without NADPH was more pronounced than in analogous studies with male and female rat and male liver microsomes.^{5, 8} However, the were clear species differences in the extent of the enantiomeric enrichment of benoxacor. For example, the rank order of the overall change in the EF due to non-oxidative metabolism in the microsomes was female rat (0.48 ± 0.01) > male rat (0.44 ± 0.01) > female human (0.43 ± 0.01) > male human (0.40 ± 0.01) > male mouse (0.36 ± 0.01) at the 30-minute time point. No changes were seen in the EF value of benoxacor in heat-inactivated pHLM or any control not containing pHLM compared to the racemic benoxacor standard. This result indicates that, without the addition of NADPH, the enantiomeric enrichment in experiments with active microsomes is due to the metabolism of benoxacor by microsomal CESs (Fig. 5C & 5D).

3.8. Enantiomeric enrichment in cytosolic incubations with GSH

GST-mediated mediated metabolism of xenobiotics is often highly stereospecific.⁴² In our study, E₂-benoxacor was slightly enriched in experiments with pHLC in the presence of GSH (Fig. 6A and S1, Table S7). For example, the EF values of benoxacor in pHLC from female and male donors were 0.45 ± 0.01 and 0.45 ± 0.01 at the 30-minute incubation timepoint, respectively (Table S7). These EF values were significantly different from the racemic benoxacor (EF = 0.48 ± 0.01) used in this study. An enrichment of E₂-benoxacor was also observed in analogous studies with male mouse liver cytosol, with an EF of 0.41 ± 0.01 at the 30-minute time point.⁵ A more complex trend in the enantiomeric enrichment of benoxacor was observed in experiments with rat cytosol.⁸ Briefly, E₁-benoxacor was enriched at 30-minutes in experiments with cytosol from female rats (EF = 0.55), whereas E₂-benoxacor was enriched at 30-minutes in studies with cytosol from male rats (EF = 0.43). In contrast to rats, no sex differences were observed in the EF of benoxacor remaining in incubations with active pHLC and GSH (Fig. 6B). Controls showed no deviation from the reported EF value of racemic benoxacor (EF = 0.48). These findings confirm that the enrichment of E₂-benoxacor in metabolism studies with pHLC is due to its metabolism by GST (Fig. 6C & 6D).

Figure 6.

Benoxacor is enantioselectively metabolized by liver cytosol prepared from female and male human livers. (A) Change in the enantiomeric fraction over of benoxacor in experiments with female and male human liver cytosol over a 30-minute time-course. (B) Change in the enantiomeric fraction of benoxacor over time in 30-minute incubations with microsomes prepared from female and male human livers. Experiments with active cytosol prepared from (C) female and (D) male human liver microsomes and controls; see text for more details. An unpaired t-test was used to determine significance in (B). Significance in panels C & D were determined with one-way ANOVA. A Tukey’s multiple comparisons test confirmed a significant decrease in the EF values in active preparations from cytosol prepared from female (p < 0.0001) and male donors (p < 0.0001). EF values are reported in Table S7 and Table S8. Data are reported as the mean ± standard deviation for three independent replicates; for some data points, the standard deviation is smaller than the size of the symbols. The dotted lines represent EF = 0.5. GSH, glutathione; IC, heat-inactivated cytosol; ****, p < 0.0001 (one-way ANOVA); ns, not significant.

3.9. Enantiomeric enrichment in cytosolic incubations without GSH

Liver cytosol contains active CES enzymes that enantioselectively metabolize different pesticides, such as pyrethroids.⁴³ As previously stated, benoxacor was not depleted in incubations with pHLC without GSH (Fig. 6, Table S8), and there was no change in the EF over the 30-minute incubation time course (Fig. 6A). Similarly, no change in the EF values was seen in analogous incubations with rat and mouse cytosol.^{5, 8} These results demonstrate that cytosolic CESs do not contribute to the metabolism of benoxacor in several mammalian species.

3.10. Environmental and potential health implications

This study provides compelling evidence that CYPs, CESs, and GSTs play a role in the hepatic metabolism of benoxacor in humans. Analogous studies with rat and mouse subcellular fractions demonstrate clear species differences in the enantioselective metabolism of benoxacor in the liver.^{5, 8} These species differences in the contribution of drug-metabolizing enzymes to the hepatic clearance of benoxacor may affect the balance of toxication vs. detoxication pathways. Although the benoxacor metabolites formed by mammalian CYPs, CESs, and GSTs have not been identified to date, the metabolism of the structurally related chloroacetanilide herbicides offers some insights. These herbicides are metabolized into redox-active quinoneimines through N-dealkylation by CYPs.⁴⁴ Briefly, the N-dealkylation of these chloroacetanilides forms anilines.⁴⁴ The anilines are further oxidized to hydroxyanilines and undergo rearrangement to form quinoneimines. The quinoneimines formed from chloroacetanilides are implicated in the formation of nasal tumors in rats; however, the same effects were not seen with fresh human nasal epithelium.⁴⁵

Similarly, we posit that benoxacor is metabolized to a reactive quinoneimine metabolite and, thus, could be carcinogenic. Briefly, the hydrolysis of the amide bond on benoxacor by CESs is expected to form a benzoxazine metabolite, 3,4-dihydro-3-methyl-2H-1,4-benzoxazine, and dichloroacetic acid. Alternatively, the hydrolysis of benoxacor to a benzoxazine metabolite may occur within environmental systems by bacterial esterase activity⁴⁶, CES activity within the gut microbiome,^47–49 or blood albumin.⁵⁰ The benzoxazine metabolite can be further oxidized to a quinoneimine. Quinoneimines are redox-active electrophiles that have been well studied in toxicology. They can saturate conjugation pathways, react with cellular proteins, and ultimately lead to liver necrosis in humans and model organisms.⁵¹ The generation of quinonimines in rats is likely to be much lower because of lower hepatic carboxylesterase activity.⁸

The formation of dichloroacetic acid could also lead to potential toxicity as this compound is a well-known toxicant often formed through drinking water disinfection.⁵² Neurotoxic effects, including altered gate and hind limb grip, have been noted in rats exposed to dichloroacetic acid.⁵² Dichloroacetic acid has also been implicated in hepatotoxicity and increases inflammatory cytokines in mice.⁵³ In pregnant rats, dichloroacetic acid can cause decreased weight gain, a significant increase in resorbed implants per litter, and soft-tissue malformations, primarily in the cardiovascular system.⁵⁴ Exposure to dichloroacetic acid can also lead to abnormal sperm formation and morphologic abnormalities, including decreased epididymis weight.⁵⁵ Although we did not measure the levels of dichloroacetic acids, a study of the toxicity of benoxacor in male mice resulted in gene expression changes in the liver similar to those observed with dichloroacetic acid.⁵

The current maximum contamination level (MCL) set under the EPA for dichloroacetic acid is 60 μg/mL.⁵⁶ In our experiments, roughly 25% and 50% of the total benoxacor metabolism in pHLM from females and males were due to CES hydrolysis forming dichloroacetic acid. This percentage corresponds to an exposure of approximately 240 μg/mL and 120 μg/mL of benoxacor to generate the current MCL, assuming no interactions between the enzymes. These doses are much higher than the highest reported level of benoxacor (190 ng/L) in surface water,² suggesting that benoxacor toxicity may not be a major human health concern. However, benoxacor exposure can still contribute to the overall exposure of dichloroacetic acid in the environment.

Table 3.

Comparison of estimated intrinsic and scaled intrinsic clearance values of benoxacor for GSTs, CYPs, and CESs, determined with human, rat, and mouse subcellular fractions. The intrinsic and scaled intrinsic clearance values for experiments with rat and mouse subcellular fractions have been reported previously.^{5, 8}

Clearance of benoxacor	Female Human	Male Human	Female Rat	Male Rat	Male Mouse
Microsomal CYP CLint (mL/min/mg protein)	0.7	0.7	1.3	1.1	1.1
Microsomal CYP Scaled CLint (mL/min/g liver)	27	26	58	50	40
Microsomal CES CLint (mL/min/mg protein)	0.3	0.7	N/A	0.2	0.5
Microsomal CES Scaled CLint (mL/min/g liver)	11	28	N/A	6.7	21
Cytosolic GST CLint (mL/min/mg protein)	0.7	0.3	0.3	0.4	1.5
Cytosolic GST Scaled CLint (mL/min/g liver)	53	28	26	40	134
Cytosolic CES CLint (mL/min/mg protein)	N/A	N/A	N/A	N/A	N/A
Cytosolic CES Scaled CLint (mL/min/g liver)	N/A	N/A	N/A	N/A	N/A

Table 4.

Comparison of the average enantiomeric fraction (EFs) of benoxacor in 30-minute incubations with human, rat, and mouse microsomes or cytosol. The EF values for experiments with rat and mouse subcellular fractions have been reported previously.^{5, 8}

Subcellular fraction	Female Human	Male Human	Female Rat	Male Rat	Male Mouse
Microsomes with NADPH (microsomal CYP)	0.61 ± 0.01#	0.67 ± 0.01*,#	0.68 ± 0.01#	0.60 ± 0.01*,#	0.60 ± 0.01#
Microsomes without NADPH (microsomal CES)	0.43 ± 0.01#	0.40 ± 0.01*,#	0.47 ± 0.01#	0.46 ± 0.01#	0.36 ± 0.01#
Cytosol with GSH (cytosolic GST)	0.45 ± 0.01#	0.45 ± 0.01#	0.55 ± 0.01#	0.43 ± 0.01*,#	0.41 ± 0.01#
Cytosol without GSH (cytosolic CES)	0.48 ± 0.01	0.48 ± 0.01	0.48 ± 0.01	0.48 ± 0.01	0.48 ± 0.01

Data are reported of the average ± standard deviation of at least three determinations.

^*Significantly different from incubations with the corresponding subcellular fraction for female animals (unpaired t-test).

^#Significantly different from control incubations with racemic benoxacor, EF = 0.48 ± 0.01 (unpaired t-test).