Metabolism, Toxicokinetics and Hemoglobin Adduct Formation in Rats Following Subacute and Subchronic Acrylamide Dosing

NeuroToxicology 22 (2001) 341±353

Metabolism, Toxicokinetics and Hemoglobin Adduct Formation in Rats Following Subacute and Subchronic Acrylamide Dosing D.S. Barber1,y, J.R. Hunt1, M.F. Ehrich1, E.J. Lehning2, R.M. LoPachin2,* 1

Virginia Polytechnic Institute, Virginia-Maryland Regional College of Veterinary Medicine, Duckpond Dr., Blacksburg, VA 24061, USA 2 Department of Anesthesiology, Albert Einstein College of Medicine, Monte®ore Medical Center, 111 E.210th Street, Bronx, NY 10467, USA Received 9 November 2000; accepted 14 February 2001

Abstract Long-term, low-dose (subchronic) oral acrylamide (ACR) exposure produces peripheral nerve axon degeneration, whereas irreversible axon injury is not a component of short-term, higher dose (subacute) i.p. intoxication [Toxicol Appl Pharmacol 1998;151:211]. It is possible that this differential axonopathic expression is a product of exposure-dependent differences in ACR biotransformation and/or tissue distribution. Therefore, we determined the toxicokinetics and metabolism of ACR following subchronic oral (2.8 mM in drinking water for 34 days) or subacute i.p. (50 mg/kg per day for 11 days) administration to rats. Both dosing regimens produced moderate levels of behavioral neurotoxicity and, for each, ACR was rapidly absorbed from the site of administration and evenly distributed to tissues. Peak ACR plasma concentrations and tissue levels were directly related to corresponding daily dosing rates (20 or 50 mg/kg per day). During subchronic oral dosing a larger proportion (30%) of plasma ACR was converted to the epoxide metabolite glycidamide (GLY) than was observed following subacute i.p. intoxication (8%). This subchronic effect was not speci®cally related to changes in enzyme activities involved in GLY formation (cytochrome P450 2E1) or metabolism (epoxide hydrolases). Both ACR and GLY formed hemoglobin adducts during subacute and subchronic dosing, the absolute quantity of which did not change as a function of neurotoxicant exposure. Compared to subacute i.p. exposure, the subchronic schedule produced approximately 30% less ACR adducts but two-fold more GLY adducts. GLY has been considered to be an active ACR metabolite and might mediate axon degeneration during subchronic ACR administration. However, corresponding peak GLY plasma concentrations were relatively low and previous studies have shown that GLY is only a weak neurotoxicant. Our study did not reveal other toxicokinetic idiosyncrasies that might be a basis for subchronic induction of irreversible axon damage. Consequently the mechanism of axon degeneration does not appear to involve route- or rate-dependent differences in metabolism or disposition. # 2001 Elsevier Science Inc. All rights reserved.

Keywords: Toxic axonopathy; Neurotoxicant metabolism; Glycidamide; Adduct formation

INTRODUCTION Acrylamide (ACR, 2-propenamide) is a water soluble vinyl monomer which has broad industrial and *

Corresponding author. Tel.: ‡1-718-920-5054; fax: ‡1-718-515-4903. E-mail address: [email protected] (R.M. LoPachin). y Present address: Center for Environmental & Human Toxicology, University of Florida, Building 471, Mowry Road, Gainesville, FL 32611-0885, USA.

scienti®c applications, e.g. water puri®cation, sewage treatment, ore processing, gel electrophoresis (Berger and Schaumburg, 1995; LeQuesne, 1980; Spencer and Schaumburg, 1974a). Early human epidemiological and laboratory animal studies indicated that ACR exposure produced skeletal muscle weakness and ataxia (Spencer and Schaumburg, 1974b; LoPachin and Lehning, 1994). Continued research suggested that this behavioral neurotoxicity was a product of nerve damage classi®ed as a central±peripheral distal axonopathy (Spencer and Schaumburg, 1977). Axon

0161-813X/01/$ ± see front matter # 2001 Elsevier Science Inc. All rights reserved. PII: S 0 1 6 1 - 8 1 3 X ( 0 1 ) 0 0 0 2 4 - 9

342

D.S. Barber et al. / NeuroToxicology 22 (2001) 341±353

degeneration, which begins as multifocal paranodal swellings of preterminal distal ®bers, has been considered to be the morphologic hallmark of this axonopathy (Fullerton and Barnes, 1996; Prineas, 1969; Suzuki and Pfaff, 1973; Spencer and Schaumburg, 1974b). We (see reviews by LoPachin and Lehning, 1994, 1997; LoPachin et al., 2000) have attempted to delineate the mechanism of ACR-induced distal axon degeneration and have proposed that irreversible axon injury is caused by failure of axonal Na‡/K‡-ATPase. ‡ The resulting loss of Na‡ i extrusion and Ko import ‡ promotes intra-axonal Na accumulation, membrane depolarization and reverse operation of axolemmal Na‡±Ca2‡ exchange. Exchanger reversal mediates toxic Ca2‡ entry that initiates an autodestructive sequence culminating in axon degeneration. Recent quantitative morphologic studies (Lehning et al., 1998, 2001), however, have demonstrated that axon degeneration in the PNS and CNS was a conditional effect, i.e. degeneration occurred in association with long-term, low-dose (subchronic) exposure (2.8 mM ACR in drinking water), and did not develop during short-term, higher dose (subacute) induction (50 mg/kg per day). This differential expression of degeneration was not due to a relative difference in neurotoxic state, since both dosing paradigms produced equivalent quanti®ed de®cits in neurobehavior (Lehning et al., 1998, 2001). Our ®nding that similar behavioral neurotoxicity occurred in the absence of axon degeneration suggested that axonopathy was an epiphenomenon related to subchronic dosing schedules (for details see LoPachin et al., 2000). It is possible that dose rate- or route-dependent differences in ACR metabolism or tissue distribution were responsible for the differential axonopathy. For example, glycidamide (GLY) is an epoxide metabolite of ACR (Calleman et al., 1990) and some evidence suggests it might produce axon degeneration and cerebellar Purkinje cell death in intoxicated rats (Abou-Donia et al., 1993). Previous studies have shown that the conversion of ACR to GLY is highest for subchronic low-dose exposure rates (Bergmark et al., 1991). Thus, in our studies, Na‡/K‡-ATPase inhibition and axon degeneration induced by subchronic oral ACR might be mediated by the relatively elevated production of GLYassociated with this dosing schedule (Lehning et al., 1998). We therefore conducted a detailed study of ACR metabolism and toxicokinetics following subchronic oral (2.8 mM in drinking water) and subacute i.p. (50 mg/ kg per day) induction of moderate behavioral neurotoxicity. Relatively little information exists (see Crofton et al., 1996) concerning the in¯uence of route or

daily dosing rate on ACR biotransformation and disposition. Also, changes in these processes have not been fully characterized as a function of developing ACR-induced neurotoxicity. MATERIALS AND METHODS Materials Acrylamide (ACR, 99% pure), S-carboxyethyl-Lcysteine (SCEC), heptane sulfonic acid, chlorzoxazone, acrylonitrile, styrene oxide and styrene glycol were purchased from Sigma/Aldrich Chemical Co. (St. Louis, MO). Dowex 3 and 8N methanolic HCl were purchased from Supelco (Bellefonte, PA). 2,3,3-D3acrylamide and 3,3-D2-cysteine were purchased from Cambridge Isotope Laboratories, Inc. (Andover, MA). 2,3-14 C-acrylamide (5.0 mCi/mmol and 1 mCi/ml in ethanol) was purchased from American Radiolabeled Chemicals, Inc. (St. Louis, MO). 6-Hydroxychlorzoxazone was purchased from Gentest (Woburn, MA). Glycidamide (GLY) was synthesized from acrylonitrile by the method of Sugiyama et al. (1989) and its identity was con®rmed by NMR and mass spectrometry. Purity of GLY was determined to be >95%. S-(carboxyhydroxyethyl)-L-cysteine (SCHEC), deuterated SCEC and deuterated SCHEC were synthesized from glycidamide and cysteine, D3-ACR and cysteine, and glycidamide and D2-cysteine, respectively, as described by Calleman et al. (1990). Treatment of Animals and Quantitation of Neurotoxicity All aspects of this study were in accordance with the NIH Guide for Care and Use of Laboratory Animals and were approved by the local animal care committee. Adult male Sprague±Dawley rats (250±275 g, Dublin, VA) were maintained on an inverted 12 h light/dark cycle and provided food and drinking water ad libitum. All measurements were conducted during the dark period of the rats light/dark cycle (see ahead). Groups of rats (n ˆ 5 7 per end point) were exposed to ACR according to one of two dosing paradigms: daily i.p. injections at 50 mg ACR/kg per day up to 11 days (subacute induction of behavioral neurotoxicity) or oral exposure at 2.8 mM ACR in drinking water for up to 47 days (subchronic induction of behavioral neurotoxicity). Both dosing schedules have been used in previous morphological and biochemical studies of ACR axonopathy and both produce stereotypic

D.S. Barber et al. / NeuroToxicology 22 (2001) 341±353

behavioral de®cits (e.g. Burek et al., 1980; LoPachin et al., 1992; O'Shaughnessy and Losos, 1986). Body weights of all rats and water consumption of subchronic orally exposed rats were monitored daily. To assess development of ACR-induced neurological de®cits, animals in each treatment group were scored with respect to appearance of gait abnormalities (Moser et al., 1992). Gait scores were based on evaluation of open ®eld behaviors, which included level of ataxia, hopping, splay, hind limb weakness and foot placement. The scale for rating gait ranged from 1 to 4 and corresponded to normal gait and slight, moderate, or severe gait impairment, respectively (see Moser et al., 1992; Lehning et al., 1998). Experimental time points for each dosing schedule were determined by temporal characteristics of ACR-induced neurotoxicity as indicated by graded expression of neurobehavioral abnormalities. For both exposure protocols, control rats were age-matched to toxicant-treated counterparts. ACR Metabolism and Toxicokinetics To assess ACR metabolism and toxicokinetics following oral ingestion or i.p. injection, rats were administered 14 C-ACR at selected end points and respective metabolite plasma concentrations and tissue distributions of radiolabel were determined. For these studies, rats were allowed free access to drinking water but were fasted for 12 h prior to 14 C-ACR administration. Thus, on day 11 of the i.p. dosing schedule (50 mg/kg per day), animals exhibited moderate neurotoxicity and received a single i.p. injection (1 ml/kg) of a solution containing 50 mg ACR/ml with a ®nal speci®c activity of 4 mCi 14 C-ACR/mg of unlabeled ACR. For oral intoxication, preliminary studies showed that kinetic parameters and tissue distribution could not be reliably determined during exposure of rats to 14 C-ACR in drinking water. This was due to the short half-life of ACR (<2 h; see ahead) and the low-volume, intermittent drinking behavior of rats. As an alternative approach, rats were exposed to ACR through drinking water (2.8 mM) for 34 days to produce a moderate level of neurotoxicity. Twenty-four hours later, a solution containing 20 mg/kg ACR (average daily ACR intake for orally exposed rats, see ahead) with a ®nal speci®c activity of 6 mCi 14 C-ACR/mg was administered to rats by single oral gavage. To demonstrate that this approach represented a reasonable estimate of drinking-based intoxication, nocturnal drinking behavior was quanti®ed (at 1 h intervals) and 14 C-ACR was administered in repeated small gavage doses that approximated imbibment patterns. Results showed that

343

ACR half-life and respective AUCs for plasma concentration±time curves were similar when the single and multiple oral gavage paradigms were compared (data not shown). At a moderate level of behavioral neurotoxicity (i.p. 11 days; oral 34 days), blood (200 ml) was obtained from a tail vein incision at 0, 5 15, 30, 60, 120, 240, 360, 500 and 580 min following i.p. 14 C-ACR administration. For the oral experiments, blood was sampled at 0, 30, 60, 90, 120, 240, 360, 500 and 580 min after 14 C-ACR oral gavage. The earlier sampling times for i.p. determinations re¯ect the relatively more rapid uptake associated with this route. Initial studies showed that at later collection times (>12 h) radiolabeled ACR and GLY levels were not detectable. Blood was collected in heparinized syringes and centrifuged (10,000  g for 30 s) to separate plasma and erythrocytes. Plasma levels of 14 C-ACR and 14 C-GLY (mg/ml plasma) were quantitated by reversed-phase HPLC with radioactive detection (Flo-One, Packard Instruments, Meriden, CT) using the method of Barber et al. (2001). To examine tissue distribution of radiolabel, rats were killed by CO2 inhalation at the last time point (580 min) and lumbar spinal cord, sciatic nerve and portions of biceps femoris muscle were excised. Tissues were solubilized (NCA, tissue solubilizer, Amersham Pharmacia Biotech, Piscataway, NJ) and total tissue radioactivity was determined by liquid scintillation counting. To determine whether repeated exposure to ACR in¯uenced toxicokinetics, naive rats (i.e. not previously exposed to ACR) were administered 14 C-ACR by a single i.p. injection (50 mg/kg; 4 mCi 14 C-ACR/mg) or by single oral gavage (20 mg/kg; 6 mCi 14 C-ACR/mg). Blood was collected at sequential time points (see above) after 14 C-ACR administration and plasma levels of radiolabeled ACR and GLY were quantitated as indicated above. Results from naive animals were compared to those from rats exposed to ACR (i.p. or oral) on a daily basis. Liver Microsome and Cytosol Preparation In a separate experiment, liver microsomal and cytosolic fractions were prepared (Hammock and Ota, 1983) from rats exposed to ACR by either the subchronic or subacute exposure paradigm (see above). Animals were killed by CO2 inhalation and livers were perfused with 1.15% KCl to remove blood. Tissue was homogenized with a Dounce homogenizer in 76 mM phosphate buffer, pH 7.4. Homogenates were centrifuged at 10,000  g for 20 min and resulting supernatant centrifuged at 100,000  g for 75 min. The

344

D.S. Barber et al. / NeuroToxicology 22 (2001) 341±353

supernatant fraction (cytosol) was removed and the pellet (microsomes) was resuspended in homogenization buffer. Microsomal and cytosolic fractions were stored at 708C until analyzed. Total microsomal P450 content was determined spectrophotometrically by the method of Omura and Sato (1964). Determination of Microsomal Cytochrome P450 2E1 Activity Conversion of chlorzoxazone to 6-OH-chlorzoxazone (6-OH-CLX) was determined in liver microsomes (200±700 mg protein) fractions (100 ml) as a measure of cytochrome P450 2E1 activity (Bernauer et al., 1999). The assay medium contained 10 ml fresh chlorzoxazone (50 mM) in KOH (60 mM) added to 645 ml of 0.01 M Tris (pH 7.4), 50 ml of NADP‡ (10 mM), 95 ml glucose-6-phosphate (25 mM) and 100 ml glucose-6phosphate dehydrogenase (10 U/ml). This mixture was incubated at 378C for 5 min after which the chlorzoxazone reaction was started by addition of liver microsomal protein. The reaction proceeded for 20 min and was terminated by addition of 43% H3PO4. As an internal standard, 50 ml of phenacetin (10 mM) was then added to the reaction mixture. Samples were extracted with ethyl acetate (3  2 ml), organic layers were pooled and dried under a nitrogen stream at room temperature. Samples were reconstituted in 100 ml of mobile phase and analyzed by HPLC using the following conditions: column Luna C18(2) 250  3 mm (Phenomenex, Torrance, CA); mobile phase acetonitrile:0.5% phosphoric acid ratio 32:68; ¯ow rate 0.65 ml/min; injection volume 10 ml. Absorbances at 297 and 249 nm were monitored for 6-OH-CLX and phenacetin, respectively. Measurement of Epoxide Hydrolase Activity in Liver Microsomal and Cytosolic Fractions Epoxide hydrolase activity was determined by measuring the conversion of styrene oxide to styrene glycol according to a modi®cation of Stock et al. (1986). Rat liver cell microsomal (200±700 mg protein) or cytosolic (600±900 mg protein) fractions (100 ml each) were incubated with 2 mM styrene oxide (100 ml) in 0.1 M HEPES (pH 7.6) for 20 min at 378C. The reaction was terminated by addition of cold methanol (200 ml). Samples and styrene glycol standards were extracted with ethyl acetate (3  800 ml) and the organic layer was dried in N2 at room temperature. Samples and standards were reconstituted in 200 ml of distilled water and analyzed by HPLC with the following

conditions: column Luna C18(2) 250  3 mm (Phenomenex, Torrance, CA); mobile phase acetonitrile:water ratio 25:75; ¯ow rate 0.75 ml/min; injection volume 10 ml. Styrene glycol was monitored at an absorbance wavelength of 209 nm. ACR and GLY Hemoglobin Cysteine Adduct Determinations Groups of ACR-exposed (i.p. 5 and 11 days; oral 15, 21, 34 and 47 days) and age-matched control rats were killed by CO2 inhalation and blood was collected from the inferior vena cava using heparinized syringes. Globin isolation, hydrolysis and anion exchange chromatography of cysteine adducts were performed as described in Bergmark et al. (1991). D3-SCEC and D2-SCHEC were used as internal standards for ACR and GLY adducts, respectively (Bergmark et al., 1991). Following Dowex chromatography, dried samples were treated with 200 ml of 1.25 M methanolic HCl for 2 h at 808C to produce methyl esters of each analyte (see below). 2,2-Dimethoxypropane (10 ml) was included in this reaction as a water scavenger. After derivatization, samples were dried under nitrogen, dissolved in 1 ml of ethyl acetate and washed with 500 ml of 1 M sodium carbonate in deionized water. The ethyl acetate layer was dehydrated with sodium sulfate and dried under reduced pressure. Samples were reconstituted in 100 ml of ethyl acetate and 1 ml was analyzed by gas chromatography (HewlettPackard, Model 5890 series II). Chromatography conditions were: column 15 m HP-5; carrier gas He; injector temperature 2508C; oven temperature 1008C for 2 min followed by a temperature ramp of 158C/min up to 2808C; transfer line temperature 2808C; detector Hewlett-Packard, Model 5972 mass selective detector running in the positive ion electron ionization mode. One deuterium was lost from the D3-SCEC during analysis (Bergmark et al., 1991) and, therefore, the resulting D2-SCEC ion was used for quantitation of ACR adducts. Sample aliquots were run in selected ion monitoring (SIM) mode and m/z values of 132 (SCHEC), 134 (D2-SCHEC and SCEC) and 136 (D2-SCEC). Adduct concentrations were calculated using a regression of SCEC/D2-SCEC ratio for ACR and SCHEC/D2-SCHEC ratio for GLY. Statistical Analysis and Pharmacokinetic Modeling Pharmacokinetic modeling was performed with Win Nonlin (Standard Edition, Ver. 1.5, Pharsight Inc., Mt. View, CA) using a one-compartment model with

D.S. Barber et al. / NeuroToxicology 22 (2001) 341±353

®rst-order absorption and elimination. The correlation of predicted and observed values for all models was 0.9±0.94. Equality of model parameters (Cmax, tmax, AUC) was determined using the independent samples t-test. Mean adduct levels and mean enzyme activities were compared using a one-way ANOVA followed by Fisher's LSD post-hoc test. Values were considered signi®cantly different at P < 0:05. RESULTS Acrylamide Neurotoxicity As in previous studies (Lehning et al., 1998; LoPachin et al., 1992), rats exposed to oral or i.p. ACR intoxication developed changes in body weight and classic signs of ACR behavioral neurotoxicity. ACR exposure through drinking water caused a reduction in the rate of body weight gain relative to controls, i.e. ACR-exposed rats gained only 8  13% (mean  S:E:M:) of starting body weight over the 47-day experimental period, whereas age-matched controls rats gained 26  3% over the same time period. In contrast, animals intoxicated with ACR by i.p. injection lost 24  9% of original body weight over the 11-day exposure period, whereas age-matched controls gained 17  2% of original body weight during the same time period. In conjunction with weight changes, ACR caused progressive neurobehavioral de®cits. Orally intoxicated rats exhibited slight gait abnormalities (1:9  0:4 gait score) after 22 days of exposure. As exposure continued, ataxia, hopping and hind limb skeletal muscle weakness became more prevalent, i.e. at 34 days a moderate level of neurotoxicity was achieved (3:0  0:2 gait score) and by day 49 the gait score advanced to a mean of 3:4  0:1. Rats given daily i.p. injections of ACR (50 mg/kg per day) displayed a different temporal pattern, i.e. slight changes in gait (1:8  0:5 score) were observed on day 5 which advanced to moderate symptoms (3:4  0:3 gait score) after 11 injections. Thus, both routes of ACR exposure produced moderate neurotoxicity, although the length of time required to achieve this level of neurotoxicity differed between routes, i.e. oral 34 days versus i.p. 11 days. Based on daily water consumption, rats in the oral exposure group consumed an average of 8:8  0:9 mg ACR per rat per day or 20  4 mg/kg per day. The daily exposure rate for the i.p. group (50 mg/kg per day) is therefore 2.5 times higher than that of the oral group (20 mg/kg per day). Correspondingly, the cumulative

345

dose needed to reach a given level of neurotoxicity differed between routes, e.g. to induce moderate neurotoxicity the oral group consumed an estimated total cumulative dose of 680 mg/kg during 34 days of exposure, whereas i.p. intoxicated rats received 550 mg/kg after 11 consecutive daily injections. ACR Toxicokinetics Previous toxicokinetic studies have shown that ACR is rapidly absorbed from most sites of administration and then evenly distributed among tissues (Edwards, 1975; Hashimoto and Aldridge, 1970; Kadry et al., 1999; Miller et al., 1982). Fig. 1 presents plasma ACR uptake and disappearance in moderately intoxicated rats following either repeated i.p. injections (i.e. 50 mg/kg per day for 11 days; 3:4  0:3 gait score; Fig. 1A) or subchronic oral exposure (i.e. 2.8 mM p.o. for 34 days; 3:0  0:2 gait score; Fig. 1B). Results show that irrespective of route, ACR appeared rapidly in plasma (Fig. 1A and B). For both i.p. and oral ACR, plasma concentrations peaked after approximately 60± 90 min (tmax) and then declined steadily over time (Fig. 1A and B and Table 1). Also for both routes, the mean ACR plasma half-life (t1/2) was approximately 2 h (Table 1). However, the mean maximum plasma concentration (Cmax) and AUC for repeated i.p. intoxication were approximately three times larger than corresponding oral data (Table 1), e.g. mean peak ACR plasma concentration for repeated i.p. injections was 30:1  2:7 versus 11:6  0:6 mg/ml for oral exposure. These differences in pharmacokinetic parameters probably re¯ect respective daily exposure rates, i.e. 50 mg/kg per day i.p. versus 20 mg/kg per day oral, and relative differences in route-dependent bioavailability, i.e. ACR injected i.p. has greater systemic access resulting in larger peak plasma concentrations. To determine whether repeated daily exposure by either i.p. or oral administration affected ACR toxicokinetics, we compared single versus multiple exposure paradigms (Fig. 1 and Table 1). Results show that repeated daily i.p. dosing does not in¯uence toxicokinetics, whereas subchronic oral treatment slightly affected kinetic parameters (Table 1), e.g. Cmax and AUC for repeated oral exposures tended to be larger than respective single data. Following uptake, ACR can be oxidized to the epoxide, GLY (Calleman et al., 1990; Bergmark et al., 1991; Sumner et al., 1992, 1999). Fig. 2 shows the appearance and elimination of GLY in plasma following either subacute i.p. (Fig. 2A) or subchronic oral (Fig. 2B) ACR intoxication. Although the daily

346

D.S. Barber et al. / NeuroToxicology 22 (2001) 341±353

Fig. 1. Plasma concentrations of acrylamide (ACR) in rats as a function of time after subacute i.p. injection (A) or subchronic oral administration (B). Rats were intoxicated with ACR by either route up to a moderate level of neurobehavioral deficits (repeated paradigm: i.p. 50 mg/kg per day for 11 days; oral 2.8 mM in drinking water or 20 mg/kg per day for 34 days). Twenty-four hours later, rats were administered 14 C-labeled ACR (4±6 mCi/mg final specific activity) by either i.p. injection (50 mg/kg) or oral gavage (20 mg/kg) and plasma was sampled at sequential time points. Plasma ACR levels were measured by the method of Barber et al. (2001). To determine whether repeated administration influenced plasma kinetics (single paradigm), naõÈve rats (not previously exposed) were given a single i.p. injection ((A) 50 mg/kg; 4 mCi 14 C-ACR/mg) or oral gavage ((B) 20 mg/kg per day; 6 mCi 14 C-ACR/mg) and plasma ACR concentrations were determined at sequential time points. Data are presented as mean mg ACR/ml plasma  S:E:M: (n ˆ 5 7 per experimental group).

Fig. 2. Plasma concentrations of glycidamide (GLY) in rats as a function of time after acrylamide (ACR) administration by either subacute i.p. injection (A) or subchronic oral (B) exposure. Rats were intoxicated with ACR up to a moderate level of neurobehavioral deficits (repeated paradigm: i.p. 50 mg/kg per day for 11 days; oral 2.8 mM in drinking water or 20 mg/kg per day for 34 days). Twenty-four hours later, rats were administered 14 C-labeled ACR (4±6 mCi/mg final specific activity) by either i.p. injection (50 mg/kg) or oral gavage (20 mg/kg) and plasma was sampled at sequential time points. Plasma GLY levels were measured by the method of Barber et al. (2001). To determine whether repeated administration influenced plasma kinetics (single paradigm), naõÈve rats (not previously exposed) were given a single i.p. injection ((A) 50 mg/kg; 4 mCi 14 C-ACR/mg) or oral gavage ((B) 20 mg/kg per day; 6 mCi 14 C-ACR/mg) and plasma GLY concentrations were determined at sequential time points. Data are presented as mean mg GLY/ml plasma  S:E:M: (n ˆ 5± 7 per experimental group).

Table 1 Plasma kinetic parametersa for acrylamide Treatment

ACR dosing structure

Cmax (mg/ml)

ACR i.p. model

Single Repeated

29.4  1.3b 30.1  2.7a

ACR oral model

Single Repeated

7.9  0.4c 11.6  0.6

a

tmax (min) 60  7b,c 99  17 104  9 87  7

t1/2 (min)

AUC

136 144

9849  506b 9643  475b

98 118

2236  203 2828  227

All parameters were calculated using a one compartment model with first-order absorption and elimination (n ˆ 5±7 rats per experimental group). Values are presented as mean  S.E.M. Cmax is the peak plasma concentration, tmax the time to peak plasma concentration, t1/2 the plasma half-life and AUC is the area under the curve (plasma concentration integrated from 0 to 580 min post-14 C ACR administration). b The i.p. model parameter is significantly different (P < 0.05) from corresponding oral parameter. c Result of ``single'' administration paradigm is statistically different (P < 0.05) from that of corresponding ``repeated'' paradigm.

D.S. Barber et al. / NeuroToxicology 22 (2001) 341±353

347

Table 2 Plasma kinetic parametersa for glycidamide Treatment

ACR dosing structure

GLY Cmax (mg/ml)

GLY tmax (min)

GLY t1/2(min)

GLY AUC

ACR i.p. model

Single Repeated

2.8  0.3b 2.0  0.2

119  16 152  36

118 105

950  73a,b 730  102

ACR oral model

Single Repeated

1.3  0.1 1.7  0.2

163  44 185  22

112 127

486  86c 822  99

a

All parameters were calculated using a one compartment model with first-order absorption and elimination (n ˆ 5±7 rats per experimental group). Values are presented as mean  S.E.M. Cmax is the peak plasma concentration, tmax the time to peak plasma concentration, t1/2 the plasma half-life and AUC is the area under the curve (plasma concentration integrated from 0 to 580 min post-14 C ACR administration). b The i.p. model parameter is significantly different (P < 0.05) from corresponding oral parameter. c Result of ``single'' administration paradigm is statistically different (P < 0.05) from that of corresponding ``repeated'' paradigm.

dosing rates differed for i.p. and oral ACR administration (i.e. 50 versus 20 mg/kg per day, respectively), GLY pharmacokinetic parameters were quantitatively similar. Thus, both routes (repeated exposure paradigm) were associated with statistically similar mean peak GLY plasma concentrations (2.0 mg/ml), onset times (165 min) and plasma half-lives (2.0 h; Table 2). The mean AUC for GLY associated with subacute i.p. ACR intoxication (730  102) was statistically similar to the GLY AUC for subchronic ACR exposure (822  99; Table 2). With the exception of tmax, mean GLY kinetic parameters for single i.p. ACR injections tended to be higher than those for repeated i.p. administration data, i.e. Cmax and AUC (Table 2). In contrast, mean data for single oral administration was signi®cantly less (P < 0:05) than those for multiple oral exposures, e.g. AUCsingle ˆ 486  86 versus AUCrepeated ˆ 822  99. Considered in concert, the ACR and GLY pharmacokinetic data (compare respective AUCs from Tables 1 and 2) suggest that repeated low-dose oral exposure favors metabolism of ACR to GLY, i.e. a larger portion of plasma ACR was converted to GLY during subchronic oral (30%) versus subacute i.p. dosing (8%). Furthermore, the plasma appearance of GLY was delayed relative to that of ACR (compare Figs. 1 and 2) and the time to peak GLY concentration was signi®cantly longer, i.e. tmax ACRoral ˆ 87 7 min versus tmax GLYoral ˆ 185  22 min (compare Tables 1 and 2). This temporal relationship is consistent with metabolic conversion of ACR to GLY. Cytochrome P450 and Epoxide Hydrolase Activities Evidence suggests that ACR metabolism to GLY is mediated by the cytochrome P450 system (Calleman et al., 1990; Kotlovsky et al., 1984; Ortiz et al., 1981); in particular cytochrome P450 2E1 (P450 2E1; Sumner et al., 1999). Our pharmacokinetic ®ndings suggest that

when compared to subacute i.p. intoxication, a larger percentage of plasma ACR is converted to GLY during subchronic oral dosing (Figs. 1 and 2 and Tables 1 and 2). It is possible that this difference in metabolic conversion is due to differential enzyme induction (el-Din et al., 1993). Therefore, to determine the effects

Fig. 3. Cytochrome P450 2E1 (CYP P450 2E1) isotype activity in liver microsomes prepared from rats intoxicated with acrylamide (ACR) at a subacute ((A) 50 mg/kg per day for 11 days, i.p.) or subchronic ((B) 2.8 mM in drinking water for 15, 34 and 47 days) dosing rate and their respective age-matched controls. Conversion of chlorzoxazone to 6-OH-chlorzoxazone (6-OH-CLX) was measured as an index of CYP P450 2E1 activity. Enzyme isotype activities were normalized to total cytochrome P450 content and data are expressed as mean nmol 6-OH-CLX/min/nmol P450  S:E:M: (n ˆ 4 6 rats/experimental group). Asterisk () denotes statistically significant difference (P < 0:05) from agematched control.

348

D.S. Barber et al. / NeuroToxicology 22 (2001) 341±353

of subacute or subchronic ACR dosing on enzyme function, we measured the activity of P450 2E1 in liver microsomes prepared from i.p. or oral intoxicated rats and age-matched controls (Fig. 3). Results showed that daily i.p. injection of ACR (50 mg/kg per day for 11 days) signi®cantly increased microsomal P450 2E1 activity by approximately 33% (Fig. 3A). Oral ACR exposure (2.8 mM p.o. for 15 days) also produced early, quantitatively similar increases in P450 2E1 activity, although as exposure continued (i.e. 34 and 47 days) enzyme activity returned to control levels (Fig. 3B). It has been suggested that GLY is metabolized to glyceramide via epoxide hydrolases (Calleman et al., 1990; Sumner et al., 1992, 1999). The elevated formation of GLY and its hemoglobin adducts during subchronic oral ACR intoxication might be related to a relative slowing of epoxide metabolism. To examine this possibility, we measured epoxide hydrolase activity in liver microsomal and cytosolic fractions

Fig. 4. Epoxide hydrolase activity in liver microsomal and cytosolic fractions prepared from rats intoxicated with acrylamide (ACR) at a subacute ((A) 50 mg/kg per day for 11 days, i.p.) or subchronic ((B) 2.8 mM in drinking water for 15, 34 and 47 days) dosing rate and their respective age-matched controls. Conversion of styrene oxide to styrene glycol was measured as an index of epoxide hydrolase activities in liver fractions. Data are presented as mean nmol styrene glycol/min/mg protein  S:E:M: (n ˆ 4 6 rats per experimental group). Asterisk () denotes statistically significant difference (P < 0:05) from age-matched control.

prepared from i.p. or oral intoxicated rats and their age-matched controls (Fig. 4). Subacute intoxication with i.p. ACR caused a signi®cant increase in activity of both microsomal and cytosolic hydrolases (Fig. 4A). Subchronic oral exposure produced an early (day 15) increase in hydrolase activity, although as treatment progressed, microsomal activity declined to below control (i.e. day 47) while cytosolic hydrolase activity remained signi®cantly elevated (Fig. 4B). Hemoglobin Cysteine Adduct Formation ACR and GLY are electrophilic and therefore both can form adducts with sulfhydryl groups on hemoglobin and other SH containing proteins (Hashimoto and

Fig. 5. Hemoglobin adducts in rats intoxicated with acrylamide (ACR) at a subacute ((A) 50 mg/kg per day for 11 days, i.p.) or subchronic ((B) 2.8 mM in drinking water for 15, 34 and 47 days) dosing rate and their respective age-matched controls. ACR derivatized hemoglobin cysteine and resulting adducts were determined as S-(2-carboxyethyl)cysteine (SCEC), whereas glycidamide (GLY) adducts of cysteine were determined as S-(2carboxy-2-hydroxyethyl)cysteine (SCHEC). Data are presented as mean mmol adduct/g globin  S:E:M: (n ˆ 5 9 rats per experimental group). Mean  S.E.M. baseline data derived from control animals for SCEC and SCHEC are 0:05  0:02 and 0:006  0:002 mmol adduct/g globin, respectively. (1) Mean ACR adduct concentration of i.p. dosing schedule (A) is statistically larger (P < 0:05) than that of the oral paradigm (B). (2) Mean GLY adduct concentration of i.p. dosing schedule (A) is statistically smaller (P < 0:05) than that of the oral paradigm (B).

D.S. Barber et al. / NeuroToxicology 22 (2001) 341±353

349

Fig. 6. Tissue distribution of 14 C-ACR derived radioactivity following subacute i.p. or subchronic oral dosing. Rats were intoxicated with ACR up to a moderate level of neurobehavioral deficits (repeated paradigm: i.p. 50 mg/kg per day for 11 days; oral 2.8 mM in drinking water or 21 mg/kg per day for 34 days). Twenty-four hours later, rats were administered 14 C-labeled ACR (4 mCi/mg final specific activity) by either i.p. injection (50 mg/kg) or oral gavage (21 mg/kg). At 12 h post-14 C administration, rats were killed and blood and tissues were sampled. Tissue radioactivity was determined by liquid scintillation counting. Data are expressed as mean mg ACR equivalents/g tissue  S:E:M: (n ˆ 3 6 rats per experimental group).

Aldridge, 1970; Calleman et al., 1990; Bergmark et al., 1991). Fig. 5 shows the formation of ACR and GLY hemoglobin adducts during subacute i.p. (Fig. 5A) or subchronic oral (Fig. 5B) dosing conditions. Regardless of route, the concentrations of ACR or GLY adducts formed remained constant throughout the exposure period and did not vary as a function of continued ACR exposure or developing neurobehavioral de®cits. For i.p. intoxication, GLY adduct formation was low relative to production of ACR adducts, i.e. approximately 4.5 times more ACR adducts were formed at each experimental time point (Fig. 5A). This is consistent with the relative plasma concentrations of ACR and GLY associated with the i.p. dosing rate (compare Tables 1 and 2). Subchronic oral ACR dosing produced approximately 30% less ACR adducts when compared to production associated with i.p.-induced neurotoxicity (Fig. 5A and B). However, the oral ACR dosing rate (Fig. 5B) produced quantitatively more GLY adducts than subacute i.p. intoxication (Fig. 5A). This is likely due to our observation that a higher percentage of plasma ACR is converted to GLY during subchronic oral (30%) versus subacute i.p. intoxication (8%; see above). Tissue Distribution of Radiolabel Fig. 6 presents tissue distribution of radiolabel after either subacute i.p. or subchronic oral 14 C-ACR

administration. Results indicate that following i.p. intoxication, ACR and its derivatives were equally distributed in the erythrocyte, muscle and nervous tissue compartments. Oral induction of neurotoxicity produced a similar pattern of distribution with the exception that mean concentration of radiolabeled material in erythrocytes was signi®cantly higher than that of other compartments (Fig. 6). When ACR was administered by repeated i.p. injection, resulting tissue concentrations were approximately two to three times higher than those following subchronic oral exposure (Fig. 6). Finally, for both the oral and i.p. routes, respective tissue distributions and concentrations of 14 C-ACR equivalents were similar irrespective of dosing paradigm, i.e. single versus repeated exposure (data not shown). This suggests that repeated exposure to ACR does not affect pharmacokinetic distribution. DISCUSSION We (Lehning et al., 1998, 2001) and others (Crofton et al., 1996) have reported that subacute ACR intoxication does not cause primary axon degeneration in the CNS or PNS; whereas low-dose, subchronic exposure is associated with abundant ®ber loss. The dose rate-dependent differential expression of degeneration suggests that this hallmark morphologic effect (Spencer and Schaumburg, 1974b, 1976) is not a principal

350

D.S. Barber et al. / NeuroToxicology 22 (2001) 341±353

neurotoxic event and is instead an epiphenomenon related to the length of ACR exposure (see LoPachin et al., 2000 for detailed discussion of dose rate phenomena). This possibility contradicts current concepts of ACR distal axonopathy, i.e. appearance of multifocal distal preterminal axon swellings and subsequent degeneration, and implicates non-axonal sites of neurotoxicant action such as the cell body, nerve terminal or neuronal±glial interactions (Lehning et al., 1998; LoPachin et al., 2000). Before focusing on alternative toxicodynamic mechanisms and sites of action, we examined the possibility that route- or rate-dependent differences in toxicokinetics might be involved, e.g. GLY-mediated axon degeneration during subchronic ACR intoxication (see Section 1). Our results show that, regardless of route, ACR appeared rapidly in plasma and rose to peak concentrations within 60±90 min (Fig. 1). Respective plasma half-lives (t1/2) were approximately 2 h, which is similar to previous determinations of ACR t1/2 in plasma, blood (Edwards, 1975; Miller et al., 1982; Calleman et al., 1992; Raymer et al., 1993) and nervous tissue (Miller et al., 1982; Raymer et al., 1993). Peak plasma levels for each route were directly related to the magnitude of the respective daily dose, i.e. the i.p. dose and resulting Cmax were both 2.5 times larger than comparable oral parameters (see Table 1). Our ®ndings also suggested that repeated i.p. or oral exposure did not substantially alter ACR toxicokinetic parameters (see Fig. 1 and Table 1). For both i.p. and oral dosing rates, the appearance of GLY in plasma and mean peak concentration times were delayed relative to ACR kinetics. This temporal relationship provides direct in vivo evidence that GLY is a metabolite of ACR (Calleman et al., 1990; Bergmark et al., 1991; Sumner et al., 1999). The GLY plasma t1/2 for both routes was slightly less than 2 h, which compares well with previous t1/2 estimates (Bergmark et al., 1991). However, the i.p. and oral models (repeated exposure paradigms) were associated with comparable GLY concentrations and AUCs, i.e. approximately 1.85 and 770 mg/ml min, respectively (Table 2). When considered relative to the respective parent plasma concentrations, it is evident that a larger proportion of plasma ACR was converted to GLY during subchronic intoxication, e.g. when compared on the basis of AUC for repeated exposure, 8% of plasma ACR was converted to GLY as a result of high-dose i.p. intoxication versus a 30% conversion rate for low-dose oral exposure. In addition, results of the single administration paradigm indicate that repeated daily ACR exposure alters GLY kinetics, e.g. i.p. ACR injections for 11 days reduced the

maximum GLY plasma concentration and corresponding AUC, whereas multiple oral exposures (for 34 days) tended to increase these kinetic parameters (Table 2). Thus, our results and those of a previous investigation (Bergmark et al., 1991) suggest that intoxication at a lower daily dosing rate enhances metabolic conversion of ACR to GLY. The mechanism underlying the disproportional production of GLY during subchronic ACR dosing is unknown. Low-dose, subchronic ACR exposure might induce or otherwise increase the activity of enzyme systems mediating its biotransformation and thereby elevate conversion to GLY. The cytochrome P450 system appears to be responsible for metabolic transformation of ACR to GLY (Ortiz et al., 1981; Kotlovsky et al., 1984; Sumner et al., 1999) and there is some evidence that ACR can induce cytochrome enzymes (el-Din et al., 1993). However, although we found that oral dosing increased P450 2E1 isotype activity (Sumner et al., 1999), the effect was transient and did not correspond temporally to the persistent increase in GLY production (compare Figs. 3 and 5). Moreover, subacute i.p. intoxication also produced a comparable increase in P450 2E1 activity but converted only 8% of plasma ACR to GLY (Fig. 3). Alternatively, it is possible that hydrolysis of GLY to glyceramide via epoxide hydrolase (Calleman, 1996) is slowed as a function of subchronic ACR intoxication. This seems unlikely, however, since epoxide hydrolase activity increased during both i.p. and oral ACR exposure (Fig. 4). Finally, subchronic oral ACR dosing might retard glutathione conjugation of GLY and thereby promote plasma retention of this metabolite. Although we did not speci®cally examine this possibility, such a dose rate-dependent effect seems unlikely since previous studies have demonstrated that ACR administered at either a low (25 mg/kg per day) or high (50 mg/kg per day) daily dose produced comparable decreases (20±30%) in glutathioneS-transferase (GST), the enzyme that conjugates glutathione (GSH) to ACR or GLY (Dixit et al., 1980; Das et al., 1982). Like earlier research (Calleman et al., 1990; Bergmark et al., 1991), we found that both ACR and GLY formed adducts with hemoglobin cysteine groups (Fig. 5). When compared to subacute i.p. intoxication, signi®cantly more GLY adducts were formed during subchronic oral ACR exposure. This is presumably related to the correspondingly higher conversion of ACR to GLY associated with slower induction of neurotoxicity and is consistent with the earlier ®ndings of Bergmark et al. (1991) indicating that GLY hemo-

D.S. Barber et al. / NeuroToxicology 22 (2001) 341±353

globin adduct formation was comparatively higher in response to subchronic i.p. ACR dosing regimens. However, the present results show that the formation of both ACR and GLY hemoglobin adducts does not accurately re¯ect toxicant exposure or development of neurotoxicity, i.e. blood adduct concentrations peaked early and then remained constant during continued i.p. or oral ACR intoxication (Fig. 5). This is in contrast to the data of Crofton et al. (1996), which showed that hemoglobin adduct levels accumulated as a dosedependent function of ACR exposure. The reason for this discrepancy is not known. Crofton et al. (1996) showed that a subacute ACR treatment paradigm of 30 mg/kg per day for 10 days produced nearly linear accumulation of hemoglobin cysteine adducts, whereas longer exposure times with lower daily dose rates (15 mg/kg per day for 30 days or 10 mg/kg per day for 90 days) produced curvilinear data that appear to approach a maximum of 10±12 mmol adduct/g hemoglobin. This suggests that cysteine adduct levels increase initially and then plateau. Accordingly, our ®nding of constant adduct levels during subacute i.p. or subchronic oral intoxication suggests that the selected experimental time points coincide with steady-state accumulation. As indicated above, considerable evidence now suggests that, regardless of exposure route, low-dose subchronic intoxication favors oxidative metabolism of ACR to GLY. GLY readily forms adducts with sulfhydryl groups (this study; Calleman et al., 1990; Bergmark et al., 1991) and, therefore, we hypothesized (Lehning et al., 1998) that the relative excess of GLY formed during subchronic exposures adduct target proteins that mediate axon degeneration. Indeed, some evidence suggested that GLY was an active metabolite that produced ®ber degeneration and behavioral neurotoxicity similar to that of parent compound (Abou-Donia et al., 1993). However, other studies have characterized the behavioral, functional and morphologic effects of GLY and have concluded that this epoxide metabolite does not produce peripheral axon degeneration and is a relatively weak general neurotoxicant (Costa et al., 1992, 1995; Brat and Brimijion, 1993). In our study, the oral dose rate produced low plasma GLY concentrations that were quantitatively similar to those formed in response to higher i.p. dosing and only a two-fold relative increase in corresponding hemoglobin adduct levels. This level of metabolite production and adduct formation is unlikely to offset the low neurotoxic potency of GLY. Together these ®ndings suggest that GLY does not mediate axon degeneration during subchronic ACR intoxication.

351

We also determined the effect of route and dosing rate on ACR tissue distribution. Results indicate an equal distribution of radiolabel among muscle, spinal cord and sciatic nerve following subchronic oral or subacute i.p. ACR dosing and that repeated daily administration at either dosing rate did not affect this disposition (Fig. 6). Moreover, radiolabel tissue levels were related to plasma ACR concentrations (Table 1), i.e. tissue ACR equivalents associated with subacute i.p. exposure were three to ®ve times higher than those of subchronic oral administration. These ®ndings are similar to earlier determinations of ACR pharmacokinetic parameters (Hashimoto and Aldridge, 1970; Miller et al., 1982; Edwards, 1975; Crofton et al., 1996; Kadry et al., 1999). Although it is recognized that the total radiolabel tissue levels measured in the present study represent parent compound, metabolite and adduct content, it is important to note that the highest concentration of ACR equivalents in nervous tissue are associated with a dosing model (subacute i.p.) that does not produce axon degeneration. This lends further support to the suggestion that axon degeneration is a secondary effect and is not due to the direct actions of ACR at axonal sites (Lehning et al., 1998). SUMMARY AND CONCLUSIONS In this study we examined ACR biotransformation and toxicokinetics as a function of subacute i.p. and subchronic oral intoxication. This research was based on the possibility that differential axon degeneration produced by subchronic ACR intoxication (see above) was related to route- or rate-dependent differences in toxicokinetics or metabolism. Our results have con®rmed previous ®ndings (Calleman et al., 1990; Bergmark et al., 1991) that the rate of ACR conversion to its epoxide metabolite GLY is higher during subchronic dosing conditions. Although GLY has been considered to be an active ACR metabolite (AbouDonia et al., 1993; Harris et al., 1994), corresponding neurotoxic potency (Costa et al., 1992, 1995; Calleman, 1996) and peak plasma concentrations (this study) were relatively low. Hence, it is unlikely that this metabolite mediates axon degeneration during low-dose ACR exposure. Our study did not reveal other idiosyncrasies that might be a basis for subchronic induction of irreversible axon damage, i.e. plasma kinetics and tissue distribution were directly related to respective daily dose rates of the i.p. (50 mg/kg per day) and oral (20 mg/kg per day) routes. Consequently the mechanism of axon degeneration during

352

D.S. Barber et al. / NeuroToxicology 22 (2001) 341±353

subchronic ACR intoxication remains to be determined but does not appear to involve route- or rate-dependent differences in disposition. Current research (Crofton et al., 1996; Lehning et al., 1998; LoPachin et al., 2000) indicates that peripheral axon degeneration is a product of subchronic ACR intoxication and occurs independent of behavioral and functional neurotoxicity. This suggests that degeneration is not due to the direct axonal actions of ACR but is instead secondary. We are currently exploring alternative sites and mechanisms of ACR action that might involve a secondary ``dying-back'' axonopathy (Cavanagh, 1964). ACKNOWLEDGEMENTS Research presented in this manuscript was supported by a grant from NIEHS (ES03830-15) to R.M.L. and by the Virginia/Maryland Regional College of Veterinary Medicine. The authors would like to thank Dr. David Dorman, CIIT, Research Triangle Park, NC, for his helpful comments and criticisms. REFERENCES Abou-Donia MB, Ibrahim SM, Corcoran JJ, Lack L, Friedman M, Lapadula DM. Neurotoxicity of glycidamide, an acrylamide metabolite, following intraperitoneal injections in rats. J Toxicol Environ Health 1993;39:447±64. Barber DS, Hunt J, LoPachin RM, Ehrich M. Determination of acrylamide and glycidamide in rat plasma by revered-phase high performance liquid chromatography. J chromatograph 2001 in press. Berger AR, Schaumburg HH. Human peripheral nerve disease (peripheral neuropathies). In: Waxman SG, Kocsis JD, Stys PK, editors. The axon: structure, function and pathophysiology. New York: Oxford University Press, 1995. p. 648±60. Bergmark E, Calleman CJ, Costa LG. Formation of hemoglobin adducts of acrylamide and its epoxide metabolite glycidamide in the rat. Toxicol Appl Pharmacol 1991;111:352±63. Bernauer U, Vieth B, Ellrich R, Heinrich-Hirsch B, Janig GR, Gundert-Remy U. CYP2E1-dependent benzene toxicity: the role of extrahepatic benzene metabolism. Arch Toxicol 1999;73:189±96. Brat DJ, Brimijion S. Acrylamide and glycidamide impair neurite outgrowth in differentiation N1E.115 neuroblastoma without disturbing rapid bi-directional transport of organelles observed by video microscopy. J Neurochem 1993;60:2145±52. Burek JD, Albee RR, Beyer JE, Bell TJ, Carreon RM, Morden DC, Wade CE, Hermann EA, Gorzinski SJ. Subchronic toxicity of acrylamide administered to rats in drinking water followed by up to 144 days of recovery. J Environ Pathol Toxicol 1980;4:157±82. Calleman CJ. The metabolism and pharmacokinetics of acrylamide: implications for mechanisms of toxicity and human risk estimation. Drug Met Rev 1996;28:527±90.

Calleman CJ, Bergmark E, Costa LG. Acrylamide is metabolized to glycidamide in the rat: evidence from hemoglobin adduct formation. Chem Res Toxicol 1990;3:406±12. Calleman CJ, Stern LG, Bergmark E, Costa LG. Linear versus nonlinear models for hemoglobin adduct formation by acrylamide and its metabolite glycidamide: implications for risk estimation. Cancer Epidemiol Biomarkers Prev 1992;1:361±8. Cavanagh JB. The significance of the dying-back process in experimental and human neurological disease. Int Rev Exp Pathol 1964;3:219±67. Costa LG, Deng H, Gregotti C, Manzo L, Faustman EM, Bergmark E, Calleman CJ. Comparative studies on the neuroand reproductive toxicity of acrylamide and its epoxide metabolite glycidamide in the rat. NeuroToxicology 1992; 13:219±24. Costa LG, Deng H, Calleman CJ, Bergmark E. Evaluation of the neurotoxicity of glycidamide, an epoxide metabolite of acrylamide: behavioral, neurochemical and morphological studies. Toxicology 1995;98:151±61. Crofton KM, Padilla S, Tilson HA, Anthony DC, Raymer JH, MacPhail RC. The impact of dose rate on the neurotoxicity of acrylamide: the interaction of administered dose, target tissue concentrations, tissue damage, and functional effects. Toxicol Appl Pharmacol 1996;139:163±76. Das M, Mukhtar H, Seth PK. Effects of acrylamide on brain and hepatic mixed-function oxidases and glutathione-S-transferase in rats. Toxicol Appl Pharmacol 1982;66:420±6. Dixit R, Mukhtar H, Seth PK, Murti CRK. Binding of acrylamide with glutathione-S-transferase. Chem-Biol Interact 1980;32: 353±9. Edwards PM. The distribution and metabolism of acrylamide and its neurotoxic analogues in rats. Biochem Pharmacol 1975;24:1277±82. el-Din AY, al-Maskati HA, Mohamed DY, Dairi MG. Acrylamide as an inducer of metabolic activation system (S9) in rats. Mutat Res 1993;300:91±7. Fullerton PM, Barnes JM. Peripheral neuropathy in rats produced by acrylamide. Br J Indust Med 1996;23:210±21. Hammock BD, Ota K. Differential induction of cytosolic epoxide hydrolase, microsomal epoxide hydrolase and glutathione-Stransferase activities. Toxicol Appl Pharmacol 1983;139:163±76. Harris CH, Gulati AK, Friedman MA, Sickles DW. Toxic neurofilamentous axonopathies and fast axonal transport. V. Reduced bidirectional vesicle transport in cultured neurons by acrylamide and glycidamide. J Environ Health Toxicol 1994;42:343±56. Hashimoto K, Aldridge WN. Biochemical studies on acrylamide, a neurotoxic agent. Biochem Pharmacol 1970;19:2591±604. Kadry AM, Friedman MA, Abdel-Rahman MS. Pharmacokinetics of acrylamide after oral administration in male rats. Environ Toxicol Pharmacol 1999;7:127±33. Kotlovsky Y, Grishanova A, Ivanov VV. Effects of methylacrylate and acrylamide on the system of microsomal oxidation in rat liver tissue. Vopr Med Khim 1984;30:44±7. Lehning EJ, Persaud A, Dyer KR, Jortner BS, LoPachin RM. Biochemical and morphologic characterization of acrylamide peripheral neuropathy. Toxicol Appl Pharmacol 1998;151: 211±21. Lehning EJ, Ross JF, Fix AS, LoPachin RM. Subacute ACR exposure causes widespread terminal and cerebellar Purkinje cell degeneration in rat brain. The Toxicologist 2001;60:366.

D.S. Barber et al. / NeuroToxicology 22 (2001) 341±353 LeQuesne PM., 1980. Acrylamide. In: Spencer PS, Schaumburg HH, editors. Experimental and clinical neurotoxicology. Baltimore (MD): Williams and Wilkins, 1980. p. 309±25. LoPachin RM, Lehning EJ. Acrylamide-induced distal axon degeneration: a proposed mechanism of action. NeuroToxicology 1994;15:247±60. LoPachin RM, Lehning EJ. Mechanism of calcium entry during axon injury and degeneration. Toxicol Appl Pharmacol 1997; 143:233±44. LoPachin RM, Castiglia CM, Saubermann AJ. Acrylamide disrupts elemental composition and water content of rat tibial nerve. I. Myelinated axons. Toxicol Appl Pharmacol 1992; 115:21±34. LoPachin RM, Lehning EJ, Opanashuk LA, Jortner BS. Rate of neurotoxicant exposure determines morphologic manifestations of distal axonopathy. Toxicol Appl Pharmacol 2000; 167:75±86. Miller MJ, Carter DE, Sipes IG. Pharmacokinetics of acrylamide in Fisher-334 rats. Toxicol Appl Pharmacol 1982;63:36±44. Moser VC, Anthony DC, Sette WF, MacPhail RC. Comparison of subchronic neurotoxicity of 2-hydroxyethyl acrylate and acrylamide in rats. Fundam Appl Toxicol 1992;18:343±52. Omura T, Sato R. The carbon monoxide-binding pigment of liver microsomes. I. Evidence for its hemoprotein nature. J Biol Chem 1964;239:2370±8. Ortiz E, Patel JM, Leibman KC. Specific inactivation of aniline hydroxylase by a reactive intermediate formed during acrylamide biotransformation by rat liver microsomes. Adv Exp Med Biol 1981;136:1221±7. O'Shaughnessy DJ, Losos GJ. Comparison of central and peripheral nervous system lesions caused by high-dose shortterm and low-dose subchronic acrylamide treatment in rats. Toxicol Pathol 1986;14:389±94. Prineas J. The pathogenesis of dying-back polyneuropathies. Part II. An ultrastructural study of experimental acrylamide intoxication in the cat. J Neuropathol Exp Neurol 1969; 28:598±621.

353

Raymer JH, Sparacino CM, Velez GR, Padilla S, MacPhail RC, Crofton KM. Analysis of acrylamide in rat serum and sciatic nerve by gas chromatography/electron capture detection of the derivative, 2-bromopropenamide. J Chromatogr Biomed Appl 1993;619:223±34. Spencer PS, Schaumburg HH. A review of acrylamide neurotoxicity. Part I. Properties, uses and human exposure. Can J Neurol Sci 1974;1:151±69. Spencer PS, Schaumburg HH. A review of acrylamide neurotoxicity. Part II. Experimental animal neurotoxicity and pathologic mechanisms. Can J Neurol Sci 1974;1:170±92. Spencer PS, Schaumburg HH. Central±peripheral distal axonopathy Ð the pathology of dying-back polyneuropathies. In: Zimmerman, editor. Progress in neuropathology, vol. 3. New York: Grune & Stratton, 1976. p. 253±76. Spencer PS, Schaumburg HH. Ultrastructural studies of the dyingback process. III. The evolution of experimental peripheral giant axonal degeneration. J Neuropathol Exp Neurol 1977; 36:276±99. Stock BH, Bend JR, Eling TE. The formation of styrene glutathione adducts catalyzed by prostaglandin H synthase. J Biol Chem 1986;261:5959±64. Sugiyama S, Ohigashi S, Sawa R, Hayashi H. Selective preparation of 2,3-epoxypropanamide and its facile conversion to 2,3-dihydroxypropanamide with acidic resins. Bull Chem Soc Jpn 1989;62:3202±6. Sumner S, MacNeela JP, Fennell T. Characterization and quantitation of urinary metabolites of [1,2,3-13 C]acrylamide in rats and mice using 13 C nuclear magnetic resonance spectroscopy. Chem Res Toxicol 1992;5:81±9. Sumner S, Fennell T, Moore T, Chanas B, Gonzalez F, Ghanayem B. Role of cytochrome P450 2E1 in the metabolism of acrylamide and acrylonitrile. Chem Res Toxicol 1999; 12:1110±6. Suzuki K, Pfaff L. Acrylamide neuropathy in rats. An electron microscopic study of degeneration and regeneration. Acta Neuropathol 1973;24:197±203.