|Home | About | Journals | Submit | Contact Us | Français|
A transgenic mouse model of the human hPON1Q192R polymorphism was used to address the role of paraoxonase (PON1) in modulating toxicity associated with exposure to mixtures of organophosphorus (OP) compounds. Chlorpyrifos oxon (CPO), diazoxon (DZO), and paraoxon (PO) are potent inhibitors of carboxylesterases (CaE). We hypothesized that a prior exposure to these OPs would increase sensitivity to malaoxon (MO), a CaE substrate, and the degree of the effect would vary among PON1 genotypes if the OP was a physiologically significant PON1 substrate in vivo. CPO and DZO are detoxified by PON1. For CPO hydrolysis, hPON1R192 has a higher catalytic efficiency than hPON1Q192. For DZO hydrolysis, the two alloforms have nearly equal catalytic efficiencies. For PO hydrolysis, the catalytic efficiency of PON1 is too low to be physiologically relevant. When wild-type mice were exposed dermally to CPO, DZO, or PO followed 4-h later by increasing doses of MO, toxicity was increased compared to mice receiving MO alone, presumably due to CaE inhibition. Potentiation of MO toxicity by CPO and DZO was greater in PON1−/− mice, which have greatly reduced capacity to detoxify CPO or DZO. Potentiation by CPO was more pronounced in hPON1Q192 mice than in hPON1R192 mice due to the decreased efficiency of hPON1Q192 for detoxifying CPO. Potentiation by DZO was similar in hPON1Q192 and hPON1R192 mice, which are equally efficient at hydrolyzing DZO. Potentiation by PO was equivalent among all four genotypes. These results indicate that PON1 status can have a major influence on CaE-mediated detoxication of OP compounds.
Examining exposures to mixtures of insecticides is important for understanding the underlying mechanisms of organophosphorus (OP) compound toxicity and assessing aggregate risk. Numerous studies have examined additive and synergistic effects of combinations of OPs in vitro and in vivo. Early studies on OP mixtures in rodents found that exposure to malathion or malaoxon (MO), in combination with compounds that inhibit carboxylesterase (CaE), led to a potentiation of MO toxicity (Aldridge et al., 1954; Cook et al., 1957; Dubois, 1958; Murphy et al., 1959; Seume et al., 1960; Casida et al., 1961; Casida et al., 1963; Cohen and Murphy, 1971a; 1971b), even at doses below those required for biologically significant inhibition of cholinesterase activity (Dubois, 1969; Su et al., 1971).
The CaEs are members of a multigene family of enzymes that are widely distributed in the body with the highest activity levels being expressed in the liver, gastrointestinal tract, and the brain (Satoh and Hosokawa, 2006). CaE activity is also present in the plasma of rodents, but not humans (Williams et al., 1989; Li et al., 2005). Significant individual variability in levels of CaE in human liver microsomes has been reported, with activities varying from 5.3- to 44.7-fold, depending on the substrate used for measuring CaE activity (Hosokawa et al., 1995). Inter-individual variability of OP detoxication and bioactivation has also been identified in cytochrome P450s (Tang et al., 2001; reviewed in Furlong et al., 2007). CaEs can catalytically hydrolyze the carboxylic esters of malathion and MO (March et al., 1956; O’Brien, 1957; Cook and Yip, 1958; Chen et al., 1969). The bioactivation of malathion to its oxon analog forms a potent direct inhibitor of AChE (DuBois et al., 1953; March et al., 1956; Murphy and DuBois, 1957; O’Brien, 1957). When MO and CaE are combined in vitro, MO can act as both a substrate for hydrolysis and as an irreversible inhibitor of CaE and the rate of each reaction is affected by the other (Main and Dauterman, 1967). In addition to catalytic hydrolysis of OPs by A-esterases, noncatalytic hydrolysis occurs when these compounds stoichiometrically phosphorylate serine esterases (B-esterases) that are inhibited by OPs but do not hydrolyze them catalytically (e.g., CaEs and butyrylcholinesterase). The irreversible binding of CaEs to some OPs [chlorpyrifos oxon (CPO), diazoxon (DZO), and paraoxon (PO)] allows the CaEs to act as scavengers, leaving less OP available to inhibit AChE at the target site (Chambers et al., 1990).
The importance of the HDL-associated enzyme paraoxonase 1 (PON1) in OP detoxication has been known for some time (reviewed in Costa, 2006; Furlong, 2008). PON1 exhibits broad substrate specificity with different rates of hydrolysis and substrate affinity for specific OP compounds (Furlong et al., 1989; Davies et al., 1996; Pond et al., 1998; Li et al., 2000). PON1 has two common amino acid polymorphisms (Q192R and L55M) (Hassett et al., 1991). The Gln (Q)/Arg (R) substitution at position 192 affects catalytic efficiency towards some OP substrates (Adkins 1993; Humbert et al., 1993; Davies et al., 1996; Li et al., 2000). Activity levels of plasma PON1 can vary by as much as 15-fold among individuals with the same PON1 (192) genotype (Q/Q; Q/R; R/R), determined in part by polymorphisms in the promoter region of PON1 (Furlong, 2007). PON1 “status” is a term used to encompass both the Q192R polymorphism and the level of PON1 activity (Li et al., 1993; Richter and Furlong, 1999; Li et al., 2000). An individual’s PON1 status can be determined using a two-substrate assay that provides both PON1Q192R functional phenotype and PON1 levels by plotting the plasma rates of DZO hydrolysis (at high salt) vs. PO hydrolysis (Li et al., 1993; Richter and Furlong, 1999; Costa et al., 1999; Jarvik et al., 2003). A recently-developed protocol for determining PON1 status makes use of the non-toxic substrates, phenyl acetate and 4-(chloromethyl)phenyl acetate (Richter et al., 2008; 2009).
PON1 knockout (PON1−/−) mice have no detectable liver or plasma hydrolytic activity toward PO or DZO and only very limited activity toward CPO. Accordingly, these animals exhibit dramatically increased sensitivity to the toxicity of CPO (Shih et al., 1998) and DZO (Li et al., 2000). Injection of purified PON1 increases the resistance of rats and mice to OP toxicity (Main, 1956; Costa et al., 1990; Li et al., 1995; Li et al., 2000). When PON1−/− mice were injected with purified human PON1 alloforms (hPON1R192 or hPON1Q192) to restore plasma PON1, hPON1R192 provided better protection than hPON1Q192 against CPO exposure, while both alloforms were equally effective in protecting against the toxicity of DZO (Li et al., 2000), and neither alloform provided protection against PO toxicity (Li et al., 2000). The extent of the protection provided by the hPON1Q192R alloforms against OP exposure was dependent on the catalytic efficiency of hydrolysis of the specific OP compound (Li et al., 2000). A study utilizing “humanized” PON1 transgenic mice provided further evidence of the role of PON1 in modulating OP toxicity. The genes of PON1−/− mice were replaced with either human hPON1R192 or hPON1Q192, and founders of each transgenic genotype were chosen that expressed hPON1 at equivalent levels (Cole et al., 2003). When these mice were exposed to CPO or its parent compound chlorpyrifos (CPS), the mice expressing hPON1Q192 were found to be significantly more sensitive to CPO/CPS exposure than the mice expressing hPON1R192 (Cole et al., 2005).
The current study demonstrates that these differences in OP detoxication between the hPON1Q192 and hPON1R192 alloforms can have consequences beyond cholinesterase inhibition. In a combined or sequential exposure, the alloform differences can affect the subsequent toxicity (or efficacy) of compounds metabolized by CaE, even when the compound is not metabolized directly by PON1. We demonstrate that CPO, DZO, and PO inhibit CaE in vitro and in vivo and increase MO toxicity in vivo, and that PON1 status modulates the degree of MO potentiation by virtue of its impact on the metabolism of CPO and DZO.
Chlorpyrifos oxon (CAS 5598-15-2; 98% purity), diazoxon (diazinon-O-analog; CAS 962-58-3; 96% purity), paraoxon (O,O-diethyl-O-p-nitrophenylphosphate; CAS 311-45-5; 98.4% purity), malaoxon (CAS 1634-78-2; 88% purity) and tri-ortho-cresyl phosphate (CAS 1330-78-5; 98.5% purity) were purchased from Chem Service (West Chester, PA, USA). Acetylthiocholine, 5,5′-dithio-bis-nitrobenzoic acid (DTNB), p-nitrophenyl valerate, and phenyl acetate were from Sigma-Aldrich (St. Louis, MO). All other analytical grade chemicals were obtained from commercially available sources.
Mice were 8–12 wk-old males and females, of congenic B6.129 strain background, >96% C57Bl/6J, backcrossed five times from the original C57Bl/6J×129/svEv strain. PON1 knockout (PON1−/−) mice (Shih et al., 1998) and mice expressing either the human hPON1R192 or hPON1Q192 transgene in place of endogenous mouse PON1 (Cole et al., 2003; Cole et al., 2005) were kindly provided by Drs. Diana M. Shih, Aaron Tward and Aldons J. Lusis (UCLA, Los Angeles, CA). Mice with at least one copy of the transgene were crossed with same-genotype animals to produce both PON1 knockout (PON1−/−) mice and transgenic mice (hPON1R192 or hPON1Q192) in the same litter. Wild-type (PON1+/+) mice were bred from the same congenic B6.129 strain background. Equal numbers of males and females were used for all experiments. Presence of hPON1Q192 or hPON1R192 enzyme activity in heparinized saphenous-vein plasma was detected by measuring the rate of hydrolysis of either the alloform-neutral substrate, phenyl acetate (Furlong et al., 1989; Furlong et al., 1993; Furlong et al., 2006), or diazoxon, which is alloform-neutral at 0.5 M NaCl (Richter and Furlong, 1999) and has no background activity in PON1−/− mice (Cole et al., 2003). PCR-based genotyping was used to determine the presence of the transgenes. Arylesterase (AREase) and diazoxonase (DZOase) assays were carried out in a microtiter plate reader (SpectraMax Plus, Molecular Devices), using 0.5 μL (AREase) and 1.25 μL (DZOase) of plasma per well. Only the initial linear rates of hydrolysis were used for calculations.
Mice were housed in SPF (specific pathogen-free) or modified SPF facilities with a 12-h dark–light cycle and unlimited access to food and water. The animal use protocols used were approved by the Institutional Animal Care and Use Committee at the University of Washington. All animal experiments were carried out in accordance with the National Research Council Guide for the Care and Use of Laboratory Animals, as adopted by the National Institutes of Health.
Chlorpyrifos oxon (CPO), diazoxon (DZO), paraoxon (PO), malaoxon (MO) and tri-ortho-cresyl phosphate (TOCP) were dissolved in acetone and applied (1 μl/g body weight) to a shaved area (4 cm2) on the upper back. Control mice received acetone only. For experiments involving mixtures, four hours following the first OP exposures (0.75 mg/kg CPO, 0.5 mg/kg DZO, or 0.35 mg/kg PO), MO was applied dermally (1 μl/g body weight) at doses of 0, 30, 60, or 100 mg/kg. Mice were euthanized by cervical dislocation four hours following the second exposure, at the time of maximal AChE inhibition. During the time between the second dosing and euthanasia, animals were monitored for cholinergic symptoms. At the time of sacrifice, each mouse was assessed for signs of reduced body temperature, hypoactivity, eye secretions, and tremors. Following sacrifice, trunk blood was collected directly into a centrifuge tube, and serum was frozen at −80°C until analysis. Brains, diaphragms, livers, and tails were dissected, placed immediately on dry ice, and stored at −80°C until analysis. Blood was collected from the saphenous vein at least 24-h prior to the experiment, for the purpose of measuring PON1 activity (DZOase or AREase) in the plasma. For time course experiments, plasma CaE activity was monitored in individual mice by repeated blood draws (~50 μl) from the saphenous vein at 4, 8, 12, and 24-hr following dermal exposure to the OP compounds.
Brain and diaphragm AChE activities were measured using a microtiter plate assay based on the method of Ellman et al. (1961), essentially as described (Cole et al., 2005) with minor modifications. Brain and diaphragm tissues were homogenized in 10 volumes (w/v) of ice-cold 0.1 M sodium phosphate buffer (PB), pH 8.0 using a Polytron homogenizer (Brinkmann Instruments, Westbury, NY) for 30 s (brain) or 45 s (diaphragm). The samples were then diluted in the same buffer to 4 mg/mL (brain) or 25 mg/ml (diaphragm). For triplicate assays, 35 μL of diluted tissue homogenate were added to 315 μL 0.1 M PB, pH 8.0, and 100 μL of the mixture were added to each of three wells of a 96-well plate. Following initiation of the kinetic assay by addition of 100 μL of freshly-prepared 2X substrate mix (2.0 mM acetylthiocholine, 0.64 mM 5,5′-dithio-bis-nitrobenzoic acid (DTNB), 0.1 M PB, pH 8.0), the formation of 5-thio-2-nitrobenzoate, the hydrolytic product of acetylthiocholine, was monitored continuously for 10 min at room temperature at 412 nm in a microtiter plate reader (SpectraMax Plus, Molecular Devices). A reagent blank and a previously-assayed standard sample were included in each plate to normalize the samples and ensure consistency among plate assays. The initial rates of 5-thio-2-nitrobenzoate formed during the assay (mOD/min) were converted to U/g of wet tissue using its extinction coefficient of 13.6 mM−1 cm−1. AChE activity is expressed as U/g of wet tissue (U=μmol of acetylthiocholine hydrolyzed per minute). Values are presented as percent of control (%C). Control values (± SEM) for brain AChE and diaphragm AChE were 9.7 ± 0.8 U/g (n = 48) and 2.7 ± 0.1 U/g (n = 48), respectively.
Liver, plasma, and serum carboxylesterase (CaE) activities were measured using a microtiter plate assay based on the method of Munger et al. (1991) with minor modifications. Liver tissues were homogenized in 10 volumes (w/v) of ice-cold 0.1 M Tris-HCl, pH 8.0, then diluted in the same buffer to a final tissue concentration of 1.0 mg/ml. The liver homogenate was centrifuged at 10,000 g for 10 minutes at 4°C, and the supernatant was used for the assays. For triplicate assays, 20 μL of diluted tissue homogenate or 1:50 diluted serum or plasma were added to each of three wells of a 96-well plate. Following initiation of the kinetic assay by the addition of 200 μL of freshly-prepared p-nitrophenyl valerate (PNV) substrate mix (1 mM PNV, 0.1 M Tris-HCl, pH 8.0), the formation of p-nitrophenol was monitored continuously for 10 min (23°C) at 405 nm. The initial linear rates of p-nitrophenol formed during the assay were calculated using an extinction coefficient of 18.5 mM−1 cm−1. CaE activity is expressed as U/g of wet tissue or U/mL plasma or serum (U=μmol of PNV hydrolyzed per minute). Values are presented as percent of control (%C). Control values (± SEM) for liver CaE and plasma CaE were 167.0 ± 12.1 U/g (n = 8) and 8.1 ± 0.2 U/ml (n = 48), respectively.
Brain and liver homogenates were prepared from three PON1+/+ mice by the methods described above. Liver and brain homogenates were pooled and diluted to final tissue concentrations of 1.0 mg/ml and 4.0 mg/ml, respectively. Commercial mouse plasma (Pel-Freez) was diluted 1:50. Mouse liver homogenate, brain homogenate, or plasma was added in triplicate to 96-well plates. The liver and plasma samples were diluted in 50 mM Tris-HCl, 1 mM EGTA buffer, pH 8.0 and the brain samples were diluted in 0.1 M PB, pH 8.0 as described above. The samples were then incubated with CPO (0–100 μM), DZO (0–100 μM), PO (0–100 μM), or MO (0–200 μM) at 23°C for 30 minutes. AChE and CaE assays were performed in triplicate as described above, in the presence or absence of inhibitor. Three separate inhibition curves were generated for each homogenate and with each compound. GraphPad Prism software was used for nonlinear regression curve fit analysis. Final IC50 values are reported as the mean ± SD of the IC50 values calculated from the three replicate experiments.
The data collected from the enzyme assays were compiled using Microsoft Excel and values beyond a 2X inter-quartile range (IQR) criteria for outliers were excluded. Only two data points were excluded based on this criterion. Statistical analysis of the data was performed using the STATA 8.0 software package (STATA Corporation, College Station, Texas). Multifactorial ANOVA was used to determine the statistical significance of potentiation for each genotype and to determine statistically significant effects of genotype on the relevant dependent variable (AChE or CaE activity). Values of p<0.05 were considered to be significant. Results are presented as means ± SEM.
The first objective of this study was to determine, through the use of PON1+/+ and PON1−/− mice, whether PON1 modulates susceptibility to exposures of OP insecticide mixtures. The second objective was to determine the impact of PON1 status (hPON1Q192R genotype and plasma PON1 level) on the toxicity of OP insecticide mixtures, through the use of humanized hPON1Q192 and hPON1R192 transgenic mice. The design involved measuring inhibition of AChE (brain and diaphragm) and CaE (liver and serum) in mice, resulting from acute exposures of specific combinations of OP insecticide metabolites (CPO, DZO, PO) with MO, and determining whether these effects were modulated by PON1 status. The study was designed to create a timed inhibition of CaE by exposure to one of the three tested OP compounds and then 4 hours later, at the time of maximal CaE inhibition, to challenge the mice with exposure to malaoxon (MO), a CaE substrate. This protocol allowed determination of the degree to which PON1, based on genotype, was able to modulate the mixed exposure. First, inhibition of CaE by CPO, DZO, and PO was measured in vitro, then the “humanized” PON1 transgenic mouse model was employed to evaluate the degree of potentiation in vivo and the effect of PON1 status on potentiation. We hypothesized, based on previous reports, that prior exposure to an OP that inactivates CaE would potentiate the effect of MO exposure, and that PON1 status would, in cases where the OP is a physiologically relevant PON1 substrate, be important in determining an individual’s sensitivity to the mixed exposure.
OP inhibition of plasma CaE, liver CaE, and brain AChE was assessed by measuring IC50 values, using different concentrations of CPO, DZO, PO, and MO incubated with tissues prepared from PON1+/+ (B6.129) mice (Table 1). CaE was inhibited in vitro by all four compounds (CPO, DZO, PO and MO), in both liver homogenates and plasma samples, following exposure to the compounds for 30 minutes at 23°C (Table 1, Fig. 1). Brain AChE was also inhibited in vitro by all four compounds (CPO, DZO, PO and MO) following exposure of homogenates for 30 minutes at 23°C (Table 1, Fig. 1). Liver CaE [mean IC50 values (nM): 4.3 ± 0.8 (CPO); 3.9 ± 0.6 (DZO); 1.1±0.3 (PO)] demonstrated a greater sensitivity to these three OP compounds than did plasma CaE [mean IC50 values (nM): 33.7 ± 8.3 (CPO); 16.6 ± 4.2 (DZO); 18.7 ± 2 (PO)]. CPO and PO were more potent inhibitors [mean IC50 values (nM):14 ± 1.9 and 9.5 ± 1.3 respectively] of brain AChE than DZO [mean IC50 value (nM): 97 ± 7.6]. MO was a rather weak inhibitor of plasma and liver CaE but a stronger inhibitor of brain AChE than DZO [mean IC50 value (nM): 75 ± 7.8] (Table 1).
Plasma PON1 levels were determined for all mice used in this study by measuring either arylesterase (AREase) or diazoxonase (DZOase) activity (Fig. 2). The hPON1Q192 mice had the highest PON1 levels, ~50% higher than the levels in hPON1R192 mice and ~20% higher than the levels in PON1+/+ mice (Fig. 2). As seen previously, PON1−/− mice had some background AREase activity that was not due to PON1 (Fig. 2A), most likely due to the presence of plasma CaE. In contrast, plasma DZOase activity was nearly zero in the PON1−/− mice (Fig. 2B), indicating that the DZOase activity was due entirely to PON1.
Once in vitro inhibition of CaE was measured for each of the compounds, experiments were conducted to determine the time course of inhibition in vivo. Dermal exposure of mice (PON1+/+, hPON1R192, hPON1Q192, PON1−/−) to 0.75 mg/kg CPO (n=6–10 per genotype; male and female) revealed that maximal inhibition of plasma CaE occurred at four hours in all four genotypes (Fig. 3A). The time course of inhibition of plasma CaE in PON1−/− mice exposed to 0.5 mg/kg DZO (n=4) or 0.35 mg/kg PO (n=4) was also maximal at 4 hours (Fig. 3B,C). Similar to what was predicted based on genotype, sensitivity to CPO 4-h after exposure was as follows (from least sensitive to most sensitive): PON1+/+, hPON1R192, hPON1Q192, PON1−/−. The hPON1Q192 mice were more sensitive than the hPON1R192 mice despite the presence of ~50% higher levels of PON1 in their plasma (Fig. 2). At 8-h, 12-h, and 24-h post-exposure, CaE inhibition was greater in PON1−/− mice than in PON1+/+ mice, and was similar in the hPON1R192 and hPON1Q192 mice. Sensitivity to DZO 4-h after exposure was greatest in PON1−/− mice, with equivalent sensitivities between the hPON1Q192 and hPON1R192 mice (Fig. 3B). The higher levels of plasma PON1 in the hPON1Q192 mice may have provided some additional protection against CaE inhibition (Fig. 3B). Sensitivity to PO 4-h after exposure was equivalent among all four genotypes (Fig. 3C). At the time of sacrifice, 24-h following exposure, plasma CaE in PON1+/+, hPON1R192, hPON1Q192, and PON1−/− mice had recovered to 60–80% of control values (Fig. 3) and there was no detectable inhibition of liver CaE compared to controls (not shown). In a separate experiment, inhibition of CaE was measured in both serum and liver in PON1+/+, hPON1Q192, hPON1R192 and PON1−/− mice 4-h after exposure to each of the OP compounds (0.75 mg/kg CPO, 0.5 mg/kg DZO, or 0.35 mg/kg PO). Again, no measureable inhibition of liver CaE was observed, with the exception of PON1−/− mice exposed to CPO, which had CaE activities that were 17% lower than controls (not shown).
The lack of inhibition of CaE in the liver was an unexpected result. Dose response curves for all three compounds were generated in the most sensitive mice (PON1−/−) to determine whether higher doses might inhibit liver CaE (Fig. 4). Each dose group consisted of four mice and all animals were sacrificed at 4 h. Compared to serum CaE (Fig. 4A), liver CaE was resistant to OP inhibition in the PON1−/− mice (Fig. 4B), perhaps due to interaction of the oxon forms of the OP compounds with plasma serine hydrolases before they reached the liver. Dermal exposure to 0.75 mg/kg CPO, 1.0 mg/kg DZO, or 0.35 mg/kg PO resulted in inhibition of liver CaE by 8%, 7%, and 4%, respectively (Fig. 4B). Exposure to higher doses resulted in greater inhibition of liver CaE at 4-h (Fig. 4B). OP concentrations sufficient to inhibit liver CaE by at least 20% produced substantially higher inhibition of brain AChE, with a range of inhibition by all three compounds between 80 and 95%. These higher concentrations of OP (e.g., exposure of PON1−/− mice to 1 mg/kg CPO) were lethal when combined with MO doses as low as 30 mg/kg (data not shown).
Dermal exposures of PON1+/+ and PON1−/− mice to CPO, DZO, and PO followed 4-h later by exposure to MO assessed the role of PON1 status in modulating these mixed exposures. Use of the transgenic mouse model, where mouse PON1 was replaced by human PON1, addressed the relative importance of the two transgenic hPON1Q192R alloforms in protecting against OP toxicity during sequential exposures. The dose of each compound was chosen based on its ability to maximally inhibit plasma CaE while minimally inhibiting brain AChE. Doses used for the initial exposure were 0.75 mg/kg CPO, 0.5 mg/kg DZO, and 0.35 mg/kg PO, while the doses of the CaE substrate, MO, administered four hours later were 0, 30, 60, 100 mg/kg. Animals were sacrificed eight hours following administration of the first OP.
Following dermal exposure of the mice, CaE inhibition by CPO, DZO, or PO was associated with increased MO-mediated inhibition of brain and diaphragm AChE (Figs. 5–7). The effect (p value) of potentiation for each genotype, as measured by the inhibition of brain and diaphragm acetylcholinesterase (AChE), is shown in Table 2. A significant increase (p<0.01) in the average inhibition of brain AChE across doses was observed in all genotypes and for all three compounds (Table 2; Figs. 5–7) with one exception: brain AChE inhibition was not significantly increased in the DZO + MO mixture in PON1+/+ mice (p<0.10; Fig. 6A vs. Fig. 6C). Diaphragm AChE inhibition was also significantly increased (p<0.01) with prior exposure to CPO, DZO, and PO (Figs. 5–7; Table 2) with a few exceptions: diaphragm AChE inhibition was not significantly increased (p<0.08) in the CPO + MO mixture in PON1+/+ or hPON1R192 mice (p<0.32; Fig. 5B vs. Fig. 5D,E; Table 2), nor was it increased (p<0.41) in the DZO + MO mixture in PON1+/+ mice (Fig. 6B vs. Fig. 6D; Table 2). Table 2 lists the p-values associated with potentiation in each of the four genotypes.
To assess the effect of hPON1Q192R genotype on potentiation, it was necessary to first determine if the response to MO alone was affected by genotype. Because MO is detoxified by CaE and is not a PON1 substrate, it was hypothesized that exposure to MO alone would not result in differences in AChE inhibition across hPON1Q192R genotypes. Indeed, there were no significant differences in brain or diaphragm AChE inhibition among hPON1Q192R genotypes for mice exposed to MO alone (p=0.35; Fig. 5A,B; Fig. 6A,B; Fig. 7A,B), nor were there differences in inhibition of serum CaE by MO alone (p<0.99; not shown). Exposures to the OP mixtures (CPO/MO, DZO/MO, and PO/MO) resulted in differential inhibition of brain and diaphragm AChE among mice of different hPON1Q192R genotypes (Figs. 5–7; summarized in Table 3), as discussed below.
In the case of pre-exposure to CPO, genotype influenced the degree of potentiation of MO toxicity; brain and diaphragm AChE inhibition were significantly different between genotypes and across doses (p<0.002). Differences in the potentiation of MO toxicity followed our hypotheses based on the relative sensitivities of PON1−/− and PON1+/+ mice (Fig. 5C,D), or hPON1Q192 and hPON1R192 mice (Fig. 5E,F; Table 3). Inhibition of brain AChE in PON1−/− mice was significantly greater (p<0.01) than that seen in the PON1+/+ mice (Fig. 5C; Table 3), and hPON1Q192 mice had significantly greater (p<0.02) brain AChE inhibition than that seen in the hPON1R192 mice (Fig. 5E; Table 3). Diaphragm AChE inhibition in PON1−/− mice was also significantly greater than in PON1+/+ mice (p<0.001; Fig. 5D; Table 3); however, inhibition of diaphragm AChE in hPON1Q192 mice was not significantly different from hPON1R192 mice (p<0.19; Fig. 5F; Table 3), possibly because of the higher PON1 levels present in the hPON1Q192 mice.
Brain and diaphragm AChE inhibition, due to the combined sequential exposure to DZO and MO, were significantly different between genotypes and across doses (p<0.002). Inhibition of brain and diaphragm AChE in the PON1−/− mice was significantly greater than that seen in PON1+/+ mice [p<0.001 (brain); p<0.001 (diaphragm); Table 3], with PON1−/− mice exhibiting higher sensitivity to DZO (Fig. 6C,D; Table 3). However, neither brain nor diaphragm AChE inhibition were significantly different between the hPON1Q192 and hPON1R192 mice [p=0.13 (brain); p=0.62 (diaphragm); Fig. 6E,F; Table 3], as expected based on their equivalent catalytic efficiencies of DZO hydrolysis (Li et al., 2000).
PO, whose acute toxicity is not influenced by PON1 due to its low catalytic efficiency of hydrolysis (Chambers et al., 1994; Li et al., 2000), increased the toxicity of MO. It was hypothesized that exposure to PO followed by MO would produce levels of brain and diaphragm AChE inhibition that would be equivalent among all four hPON1Q192R genotypes. Indeed, brain and diaphragm AChE inhibition were not significantly different between PON1−/− mice and PON1+/+ mice (p<0.88; Fig. 7C,D; Table 3), nor between hPON1Q192 and hPON1R192 mice (p<0.07; Fig. 7E,F; Table 3).
Cholinergic symptoms were recorded for all mice at 8 h post-exposure (Table 4). Mice that were exposed to 0.75 mg/kg CPO, 0.5 mg/kg DZO, or 0.35 mg/kg PO alone did not generally exhibit cholinergic symptoms. Mice that received MO alone did not exhibit any cholinergic signs for doses of 0, 30, or 60 mg/kg MO, despite inhibition of brain AChE by 5–50%. At doses of 100 mg/kg MO, brain AChE was inhibited 35–60% and slight hypoactivity and thermoregulatory problems (i.e., coolness to the touch) were observed. Following sequential exposure to CPO, DZO, or PO followed by MO, cholinergic symptoms were observed in the PON1−/−, PON1+/+, hPON1Q192 and hPON1R192 mice. Symptoms were (in the order of occurrence with increasing dose): hypoactivity, difficulty with thermoregulation, muscle fasciculations, consistent fine tremor in forelimbs and/or head, mild lacrimation, and at higher doses of MO, tremor extending more caudally or whole body tremor that was usually associated with moderate to severe lacrimation (Table 4). At comparable doses of CPO and DZO, the PON1+/+ and hPON1R192 mice were generally more resistant to developing cholinergic symptoms than the PON1−/− and hPON1Q192 mice. Following PO exposure, the dose at which the mice developed cholinergic symptoms was equivalent among genotypes (Table 4).
The doses of CPO, DZO, and PO used above were not sufficient to inhibit liver CaE without severely compromising brain AChE. For example, prior exposure of mice to 1 mg/kg CPO inhibited liver CaE, but was lethal when combined with MO doses as low as 30 mg/kg (data not shown). TOCP has been shown previously to inhibit liver and serum CaE, and can potentiate the toxicity of MO while minimally inhibiting brain AChE (Casida et al., 1961; Cohen and Murphy, 1971a; 1971b). To assess the effect of inhibiting both liver and serum CaE, PON1+/+ mice (n=4 for each dose group) were exposed to 10 mg/kg TOCP followed 24-h later by MO exposure (0, 30, 60, 100 mg/kg).
Four hours after MO exposure, AChE activity was measured in brain and diaphragm and CaE activity was measured in liver and serum. Pretreatment with TOCP significantly increased (p<0.001) the inhibition of both brain and diaphragm AChE by MO at all doses (30, 60, 100 mg/kg), with a markedly greater potentiation at the 60 mg/kg dose of MO (Fig. 8B,C). TOCP caused significant inhibition (p<0.001) of both liver and serum CaE, with liver CaE being the more sensitive of the two (Fig. 8A). Prior exposure to TOCP at 10 mg/kg produced serum CaE inhibition similar to that produced from prior exposure to 0.75 mg/kg CPO, yet exposure to TOCP resulted in a greater potentiation of MO toxicity.
This research focused on the relative importance of the two hPON1Q192R alloforms in protecting against OP toxicity during sequential mixed OP exposures, using transgenic mouse models that included strains with different PON1 status (PON1+/+, PON1−/−, hPON1R192 and hPON1Q192). OP exposures are often multi-route and multi-chemical in nature. Based on a determination that OPs form a common mechanism group based on their shared ability to inhibit AChE in both the central and peripheral nervous systems (US EPA, 1999), the EPA carried out a cumulative risk assessment for this class of pesticides (US EPA, 2002), and recently reviewed and updated its cumulative assessment, eliminating or modifying many uses of these compounds (US EPA, 2006). While the assumption in these assessments was dose additivity of the OP compounds, we demonstrate here that greater-than-additive effects are possible when exposure to certain OP compounds precedes exposure to another OP compound, e.g., malaoxon. The endpoints where potentiation was observed were clinically relevant; that is, AChE inhibition in the brain and diaphragm. As CaE inhibition is known to influence the toxicity of malathion following pre-exposure to other AChE inhibitors (Casida, 1961; Cohen and Murphy, 1971a; 1971b; Verschoyle et al., 1982) we utilized specific combinations of OPs to explore our first hypothesis that prior exposure to a CaE inhibitor (CPO, DZO, PO) would increase toxicity of CaE substrates such as MO. In this study, the combined effects of the selected OP insecticides (CPO + MO, DZO + MO, and PO + MO) increased MO toxicity significantly.
The mechanism for this increased toxicity is most likely related to inhibition of carboxylesterase (CaE) enzymes in the blood and/or liver. In contrast to rodents, there is no CaE activity in human plasma (Williams et al., 1989; Li et al., 2005). In the current study, OP compounds inhibited CaE activity significantly in mouse plasma or serum, but not liver. This did not appear to be due to differences in time for the OP compounds to reach the liver versus blood, because liver CaE was not inhibited at 4-h, 8-h, or 24-h after exposure at the doses used for the study of mixtures. A more likely explanation is that the oxon forms of the OP compounds used in this study encounter and bind to serum B-esterases before passing through the liver. Exposure of mice to higher doses of OP compounds resulted in inhibition of CaE activity in liver as well as serum; however, the resulting potentiation of MO toxicity was lethal.
Of particular importance is the demonstration that PON1 status (PON1 levels and hPON1Q192R genotype) modulates the interactive toxicity of OP compounds. The results presented here show that PON1 status alters MO potentiation in a substrate-dependent manner, and that the potentiation is determined in part by the catalytic efficiency of the PON1 alloform present in serum. First, by comparing PON1−/− and PON1+/+ mice, it was evident that the absence of PON1 significantly increased the potentiation of MO toxicity by both CPO and DZO, but not PO. This is consistent with the role of PON1 in the in vivo hydrolysis of CPO and DZO, but not PO. Our previous studies demonstrated that PON1 did not provide protection against PO exposure in vivo (Li et al., 2000). This is of particular importance given the large differences in PON1 levels among individuals and across development (Cole et al., 2003; Furlong et al., 2006).
Second, differences in potentiation were observed between mice expressing hPON1Q192 and hPON1R192 for CPO, but not DZO or PO, indicating that the hPON1Q192R genotype is important for modulating the interactive toxicity of OP compounds in a substrate-specific manner. The differences in genotype-modulation of potentiation among OP compounds are entirely consistent with the known catalytic efficiency differences in the hPON1Q192 and hPON1R192 alloforms. Compared to the hPON1Q192 alloform, the hPON1R192 alloform has a higher catalytic efficiency of CPO hydrolysis [150 vs. 250 (Li et al., 2000)], and in the current study, the hPON1R192 alloform was more protective than the hPON1Q192 alloform against CPO-potentiation of MO toxicity. This was the case despite the 50% higher PON1 levels present in the plasma of hPON1Q192 mice compared to the hPON1R192 mice. The hPON1Q192 and hPON1R192 alloforms have similar catalytic efficiencies of DZO hydrolysis (Li et al., 2000) and in the current study provided similar protection against DZO-potentiation of MO toxicity. For PO, the catalytic efficiency of both alloforms is too low to be protective in vivo (Li et al., 2000), and in the current study neither hPON1Q192, hPON1R192, nor mouse PON1 provided any protection against the effects of PO on MO toxicity. In addition to brain and diaphragm AChE inhibition, cholinergic symptoms were differentially affected among mice of different genotypes. Following sequential exposure to CPO, DZO, or PO, followed by MO, cholinergic symptoms were observed in the PON1−/− and hPON1Q192 mice. PON1+/+ and hPON1R192 mice were more resistant to developing cholinergic symptoms.
Third, the hPON1Q192 and hPON1R192 mice tended to have effects on potentiation that were intermediate between those observed in the PON1−/− and PON1+/+ mice. This is somewhat paradoxical, since the hPON1Q192 mice had 20% higher levels of PON1 compared to PON1+/+ mice. The most likely explanation would be a higher catalytic efficiency of mouse PON1 compared to human PON1, a possibility that is addressable experimentally. An additional possibility is that the human PON1 protein is not as fully functional as the endogenous mouse PON1 protein in the context of mouse physiology.
In this study, the degree to which CPO, DZO, or PO inhibited CaE predicted the degree of potentiation of MO toxicity. Other studies have examined the potentiation of MO toxicity by OPs that inhibit CaE both in vitro and in vivo. In one study, CPO showed a high inhibitory potency toward CaE-mediated detoxication of malathion in human liver microsomes (Buratti and Testai, 2005). Concurrent exposure to a mixture of five OPs, including chlorpyrifos and diazinon, was found to potentiate malathion toxicity (AChE inhibition and behavioral changes) in adult and pre-weanling rats, resulting in greater-than-additive effects for all endpoints measured (Moser et al., 2005; 2006). Previous work has reported both a lack of potentiation of the effects of malathion by pretreatment with parathion (Cohen and Murphy, 1971b) and a potentiation of effects (Su et al., 1971; Ramakrishna and Ramachandran, 1978). In addition, a study by Chambers and colleagues (1990) demonstrated that rat liver homogenates detoxified appreciable amounts of several organophosphates (including PO) stoichiometrically and this correlated well with aliesterase activity. A second study by this same group found that TOCP pretreatment potentiated PO toxicity, supporting the role of CaE in the detoxication of PO (Tang and Chambers, 1999). Timchalk and colleagues (2005) used a binary mixture of diazinon and chlorpyrifos to evaluate pharmacokinetic and pharmacodynamic interaction in the rat. At low doses (15 mg/kg), they found no apparent pharmacokinetic interactions, suggesting additive effects on AChE inhibition, whereas interactive effects between the two compounds were observed at a higher dose (60 mg/kg). The overall relative potency for AChE inhibition was chlorpyrifos + diazinon > chlorpyrifos > diazinon (Timchalk et al., 2005).
The data reported here indicate that the common hPON1Q192R polymorphism, coupled with large differences in PON1 levels among individuals, can have impacts beyond those of simple OP-mediated cholinesterase inhibition. By modulating OP-mediated inhibition of carboxylesterase, differences in PON1 status can have effects on the toxicity of other compounds that are detoxified by carboxylesterases. Presumably this would include not only malaoxon or malathion, but also pyrethroid compounds (Abernathy and Casida, 1973; Gaughan et al., 1980; Godin et al., 2001; Choi et al., 2004; Wheelock et al., 2005) that have widespread residential uses. Of additional interest in this regard is the potential in aircraft for simultaneous exposure to tricresyl phosphate as a cabin air contaminant and pyrethroids that are sprayed in the cabin for insect control (Winder, 2002). Variability of PON1 status would also impact the metabolism of many drugs and pro-drugs, and insecticides detoxified by carboxylesterases. The data presented here indicate that differences in PON1 activity levels, but not hPON1Q192R genotype, would be important for modulating mixed exposures involving diazinon. In contrast, individuals homozygous for the hPON1Q192 alloform would be the most susceptible for the interactive toxicity involving exposures to chlorpyrifos, particularly those with low plasma activity levels. These findings have particular relevance for newborns and young children, who have very low levels of PON1 (Cole et al., 2003).
The authors express their appreciation to Drs. Diana Shih, Aldons J. Lusis, and Aaron Tward for kindly providing the PON1−/− mice and the hPON1Q192 and hPON1R192 transgenic mice used in this study. This work was supported by the National Institute of Environmental Health Sciences (NIEHS) [grant numbers P42ES04696, R01ES09883; P30ES07033, and P01ES09601], and by grant number RD83170901 from the Environmental Protection Agency (EPA).
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.