الرئيسية NeuroToxicology Cholinesterase inhibition and depression of the photic after discharge of flash evoked potentials...
أبلغ عن مشكلةThis book has a different problem? Report it to us
إختار نعم إذا كان إختار نعم إذا كان إختار نعم إذا كان إختار نعم إذا كان
نجحت في فتح الملف
يحتوي الملف على كتاب (يُسمح أيضًا بالرسوم الهزلية)
محتوى الكتاب مقبول
يتوافق الإسم و المؤلف و لغة الملف مع وصف الكتاب. تجاهل الخانات الأخرى لأنها ثانوية!
إختار لا إذا كان إختار لا إذا كان إختار لا إذا كان إختار لا إذا كان
- الملف تالف
- الملف محمي بواسطة الوسائل التقنية لحماية حق المؤلف DRM.
- الملف ليس كتاب (علي سبيل المثال، xls, html, xml)
- الملف عبارة عن مقال
- الملف مقتطف من كتاب
- الملف عبارة عن مجلة
- الملف نموذج إختبار
- الملف سبام
هل تعتقد أن محتوى الكتاب غير مناسب ويجب حظره
لا يتطابق الإسم أو المؤلف أو لغة الملف مع وصف الكتاب. تجاهل الخانات الأخرى.
Change your answer
أكثر المصطلحات والعبارات المستخدمة
il y avait190
NeuroToxicology 33 (2012) 332–346 Contents lists available at SciVerse ScienceDirect NeuroToxicology Cholinesterase inhibition and depression of the photic after discharge of ﬂash evoked potentials following acute or repeated exposures to a mixture of carbaryl and propoxur§,§§ Jean-Claude Mwanza a,1, Danielle F. Lyke b, Richard C. Hertzberg c, Lynne Haber d, Melissa Kohrman-Vincent d, Ruosha Li e, Yi Pan e, Robert H. Lyles e, Jane Ellen Simmons f, Denise K. MacMillan g, R. Dan Zehr g, Adam E. Swank g, David W. Herr b,* a National Research Council, Washington, DC 20001, United States Neurotoxicology Branch, Toxicity Assessment Division, National Health and Environmental Effects Research Laboratory, Ofﬁce of Research and Development, U.S. EPA, Research Triangle Park, NC 27711, United States c Biomathematics Consulting, Atlanta, GA 30307, United States d Toxicology Excellence for Risk Assessment, Cincinnati, OH 45211, United States e Deparment of Biostatistics and Bioinformatics, Emory University, Atlanta, GA 30322, United States f Pharmacokinetics Branch, Integrated Systems Biology Division, National Health and Environmental Effects Research Laboratory, Ofﬁce of Research and Development, U.S. EPA, Research Triangle Park, NC 27711, United States g Analytical Chemistry Research Core, Immediate Ofﬁce, National Health and Environmental Effects Research Laboratory, Ofﬁce of Research and Development, U.S. EPA, Research Triangle Park, NC 27711, United States b A R T I C L E I N F O A B S T R A C T Article history: Received 9 September 2011 Accepted 6 February 2012 Available online 14 February 2012 Previously, we reported that acute treatment with propoxur or carbaryl decreased the duration of the Photic After Discharge (PhAD) of Flash Evoked Potentials (FEPs). In the current studies, we compared the effects of acute or repeated exposure to a mixture of carbaryl and propoxur (1:1.45 ratio; propoxur:carbaryl) on the duration of the PhAD and brain ChE activity in Long Evans rats. Animals were exposed (po) eithe; r to a single dose (0, 3, 10, 45 or 75 mg/kg), or 14 daily dosages (0, 3, 10, 30, 45 mg/ kg), of the mixture. Acute and repeated treatment with 3 mg/kg (or greater) of the mixture produced dose-related inhibition of brain ChE activity. Compared to controls, the PhAD duration decreased after acute administration of 75 mg/kg or repeated treatment with 30 mg/kg of the mixture. The linear relationship between the percent of control brain ChE activity and the PhAD duration was similar for both exposure paradigms. Dose–response models for the acute and repeated exposure data did not differ for brain ChE activity or the duration of the PhAD. Repeated treatment with the mixture resulted in slightly less (13–22%) erythrocyte ChE inhibition than acute exposure. Both acute and repeated treatment resulted in dose-additive results for the PhAD duration and less than dose-additive responses (6–16%) for brain ChE activity for the middle range of dosages. Acute treatment resulted in greater than dose-additive erythrocyte ChE inhibition (15–18%) at the highest dosages. In contrast, repeated treatment resulted in less than dose-additive erythrocyte ChE inhibition (16–22%) at the middle dosages. Brain and plasma levels of propoxur and carbaryl did not differ between the acute and repeated dosing paradigms. In summary, a physiological measure of central nervous system function and brain ChE activity had similar responses after acute or repeated treatment with the carbamate mixture, and brain ChE showed only small deviations from dose-additivity. Erythrocyte ChE activity had larger differences between the acute and repeated treatment paradigms, and showed slightly greater deviations from dose-additivity. Because Keywords: Neurophysiology Flash evoked potential Photic after discharge Carbamate Mixture § The information in this document has been funded wholly (or in part) by the U.S. Environmental Protection Agency. It has been subjected to review by the National Health and Environmental Effects Research Laboratory and approved for publication. Approval does not signify that the contents reﬂect the views of the Agency, nor does mention of trade names or commercial products constitute endorsement or recommendation for use. §§ Portions of this manuscript were presented as a poster at the 47th Annual Meeting of the Society of Toxicology, Seattle, WA, March 16–20, 2008. * Corresponding author at: 109 T.W. Alexander Drive, MD B105-05, NHEERL/TAD/NB, U.S. EPA, Research Triangle Park, NC 27711, United States. Tel.: +1 919 541 0380; fax: +1 919 541 4849. E-mail addresses: Jean_Claude_Mwanza@med.unc.edu (J.-C. Mwanza), Lyke.Danielle@epamail.epa.gov (D.F. Lyke), email@example.com (R.C. Hertzberg), Haber@tera.org (L. Haber), Kohrman@tera.org (M. Kohrman-Vincent), firstname.lastname@example.org (R. Li), email@example.com (Y. Pan), firstname.lastname@example.org (R.H. Lyles), Simmons.email@example.com (J.E. Simmons), MacMillan.firstname.lastname@example.org (D.K. MacMillan), Zehr.email@example.com (R.D. Zehr), Swank.firstname.lastname@example.org (A.E. Swank), Herr.email@example.com (D.W. Herr). 1 Current address: Department of Ophthamology, University of North Carolina, Chapel Hill, NC 27599, United States. 0161-813X/$ – see front matter . Published by Elsevier Inc. doi:10.1016/j.neuro.2012.02.006 J.-C. Mwanza et al. / NeuroToxicology 33 (2012) 332–346 333 these treatments utilized larger dosages than anticipated environmental exposures, concern for nonadditive effects in humans is minimized. The small magnitude of the deviations from dose-additivity also suggest that in the absence of repeated exposure data, results from an acute study of readily reversible carbamate toxicity can be used to estimate the response to repeated daily exposures. Published by Elsevier Inc. 1. Introduction There is a lack of research on physiological changes after exposure to mixtures of N-methyl carbamate pesticides, although exposure to mixtures of these chemicals can occur through food, drinking water as well as other environmental media (US EPA, 2007). Until recently, risk assessment for pesticides was often conducted to evaluate the potential risks from exposure to a single chemical, often using only a single route of exposure. Although it is important to evaluate individual toxic agents, exposures are often simultaneous or sequential, involving more than one toxic chemical (Pohl et al., 1997). Such exposures can result in unexpected cumulative effects, and the related combined risk may be greater or less than what would typically be predicted from data on individual chemicals. This led to the promulgation of the Food Quality Protection Act (Senate and House of Representatives of the United States of America, 1996), which requires the U.S. Environmental Protection Agency (EPA) to set regulatory measures based on potential human risks associated with cumulative effects from mixtures of pesticides that have a common mode of action. Although the EPA risk assessment for N-methyl carbamate pesticides is based on brain inhibition of cholinesterase (ChE) activity as the critical endpoint (US EPA, 2007), additional data based on other measures (i.e. behavioral and neurophysiological) will increase the scientiﬁc conﬁdence in the risk assessment. The visual system is known to be vulnerable to a large number of chemicals including heavy metals, solvents, and pesticides (Crofton and Sheets, 1989). These substances may damage the nerve cells or disrupt the mechanisms involved in cortical processing of visual information (Dyer, 1985). Neurophysiological methods such as visual evoked potentials, which include pattern evoked potentials and ﬂash evoked potentials (FEPs), have proven to be useful in screening for toxicity to the visual system and in determining the targeted neural systems within the visual system (Dyer, 1985; Mattsson and Albee, 1988). The FEP is a neurophysiological response to stimulation of the visual system using light ﬂashes. It has been considered to be composed of at least three distinct portions generated by different neural structures (Creel et al., 1974; Mattsson and Albee, 1988). The early portion, which includes peak N36, is produced by depolarization of cells in lamina IV of the visual cortex, and reﬂects the primary cortical response. It depends upon direct input from the retino-geniculate afferent volley (Kraut et al., 1985; Mitzdorf, 1987). The generator of the middle portion (P59–P102) remains speculative. However, studies have shown that stellate cell input to the granulate elements in lamina IV (Kraut et al., 1985; Schroeder et al., 1991) and input of stellate cells to neuronal elements in supragranular extrastriate cortex play a key role in triggering its elicitation (Givre et al., 1994). The third portion includes peak N166 and is followed immediately by a rhythmic activity known as the photic after discharge (PhAD). The PhAD is believed to arise from a reverberating thalamo-cortical circuit, which is activated by the afferent retinogeniculate volley (Bigler, 1977; Sumitomo and Klingberg, 1972). It is related to the cortical processing of visual information, and it is modulated by the cholinergic system (Bigler and Fleming, 1976; Fleming et al., 1972; Mwanza et al., 2008). Therefore, we chose to examine the early and later portions of the FEP as they may indicate functional alterations affecting the initial afferent retino-geniculate volley and disruption of cortical processing of visual information, respectively (Herr and Boyes, 1995). We recently reported that acute treatment with carbaryl or propoxur decreases the duration of the PhAD of FEPs, and that this decrease correlated with the inhibition of brain ChE activity (Mwanza et al., 2008). Because the PhAD is modulated by the cholinergic system, we hypothesized that a mixture of N-methyl carbamate pesticides would result in dose-additive effects on biochemical and neurophysiological endpoints. This hypothesis was tested using brain ChE activity and the PhAD of FEPs following acute or repeated administration of a mixture of carbaryl and propoxur. Measures of peripheral ChE activity (plasma and erythrocytes) were quantiﬁed for comparison. Predicted doseadditive effects models were constructed for the PhAD duration and brain and erythrocyte ChE activity. These models examined if the observed mixture responses could be predicted from doseaddition of the responses from the single chemicals, and if they differed between acute and repeated treatment with the mixture of carbamates. 2. Methods 2.1. Animals and surgery The animals used in these studies were housed in an AAALAC International accredited facility, and all investigations were approved by the National Health and Environmental Effects Research Laboratory Animal Care and Use Committee of the United States Environmental Protection Agency that requires compliance with National Institute of Health guidelines. Male Long Evans rats (60 days old) were purchased from Charles River Laboratories (Wilmington, MA), and housed with a 12:12 h light:dark cycle (lights on at 06:00 a.m.). Housing conditions were maintained at 22 2 8C with 40 20% humidity, and 10–15 air changes per hour of 100% ﬁltered fresh air. Animals were housed singly with heat-treated hardwood chip bedding in transparent polycarbonate cages and provided ad libitum with food pellets (Purina Lab Chow, St. Louis, MO) and ﬁltered tap water, and were allowed to acclimate to the facility for 5–6 days prior to surgery. The rats were anesthetized with sodium pentobarbital (50 mg/ kg, i.p.) and surgically implanted with epidural stainless steel screw electrodes (00–90 1/16 in.) with an approximate surface area of 0.8 mm2 (J.I. Morris, Co., Southbridge, MA), as previously described (Herr et al., 1991). The active electrode was placed over the visual cortex area (1 mm anterior to lambda and 4 mm left of midline), the reference electrode was placed over the frontal cortex (2 mm anterior to bregma and 2 mm right of midline), and the ground electrode was located 2 mm anterior to bregma and 2 mm left of midline. After surgery, the incision was painted with 10% povidone-iodine ointment (E. Fougera & Co., Melville, NY), and the rats were administered 67,000 IU penicillin G (i.m.; Sigma Chemical Co., St. Louis, MO), and 12.5 mg (s.c.) of carprofen (0.25 ml of a 50 mg/ml solution; Pﬁzer Animal Health, NY, NY). The following day, a 5 g feed pellet containing 2 mg of carprofen (BioServ, Frenchtown, NJ) was placed in the animal’s cage. The animals were randomly assigned to treatment groups, which were counterbalanced over test chambers. A one week recovery period was allowed before beginning testing in the acute exposure 334 J.-C. Mwanza et al. / NeuroToxicology 33 (2012) 332–346 experiment. In the repeated-exposure study, dosing began 5 days after surgery. 2.2. Chemicals and exposure Carbaryl (1-naphthalenol methylcarbamate, CAS: 63-25-2, purity: 99.5%) and propoxur ((2-(1-methylethoxy) phenol methylcarbamate), CAS: 114-26-1, purity: 99%) were purchased from Chem Service (West Chester, PA). These chemicals were chosen based on their inclusion in the N-methyl carbamate cumulative risk assessment (US EPA, 2007) and the single chemical data previously generated in our laboratory (Mwanza et al., 2008). Both chemicals were suspended in a corn oil vehicle either as pure compounds or in a mixture consisting of 1:1.45 parts (w/w) propoxur:carbaryl. This mixture ratio was based approximately on the chemicals’ approximate relative potencies for inhibition of brain ChE activity as reported by other investigators (Padilla et al., 2006; Moser, personal communication), and is within the range of relative potencies based on BMD10 for inhibition of brain ChE activity (0.58–2.61 mg/kg) reported in the N-methyl carbamate cumulative risk assessment (US EPA, 2007). In both the acute and repeated exposure mixture studies, the lower dosages of the mixtures were prepared as serial dilutions of the highest mixture dosage. The dosing solutions were prepared one day prior to the ﬁrst treatment day, and used throughout the study (1 or 14 days of dosing). Treatment solutions were refrigerated when not in use, allowed to warm to room temperature, and were stirred thoroughly prior to use. Animals in the repeated exposure study were weighed daily, and the dosage volume adjusted based on any changes in the animal’s weight. All solutions were administered by oral gavage (1 ml/kg) using an 18 ga feeding needle with a blunt tip. Both the acute and repeated exposure studies involved dose– responses functions using four dosages for the mixtures plus vehicle controls. Additionally, each study included one dosage of each of the single chemicals as positive controls, to examine shifts in the acute dose–response compared to the responses observed in earlier experiments (Mwanza et al., 2008). The acute exposure study used mixture dosages of 0, 3, 10, 45 and 75 mg/kg, and single chemical dosages of 30 mg/kg for propoxur and 75 mg/kg carbaryl. All treatments had 12 animals per group. The repeated exposure study used mixture dosages of 0 (n = 14), 3 (n = 13), 10 (n = 14), 30 (n = 14), and 45 (n = 15). Again, acute exposure (day 14 only) positive control treatments of 30 mg/kg for propoxur (n = 14) and 75 mg/kg carbaryl (n = 13) were included. The positive control dosages were chosen to produce about 60% decrease in brain ChE activity, and resulted in a signiﬁcant decrease in the duration of the PhAD in our previous study (Mwanza et al., 2008). For the analytical chemistry, carbaryl and 13C6-carbaryl were purchased from Chem Service, Inc. (West Chester, PA). Propoxur and 1naphthol were obtained from Sigma–Aldrich (St. Louis, MO). Propoxur-D3 and 1-naphthol-D7 were purchased from CDN Isotopes (Quebec, Canada). Solvents (HPLC-grade) were obtained from Fisher Scientiﬁc. 2.3. FEP recording procedure Custom built sound and light attenuating Faraday boxes were used for testing (see Hamm et al., 2000 for detailed description). The ﬂash stimulus (duration: 10 ms, intensity: 142 lx-s, rate: 0.32 Hz, setting: 16) was generated by a photic stimulator (Model PS22, Grass Instrument Division, Astro-Med, Inc., West Warwick, RI) placed approximately 37 cm above the animal’s head. Flash stimuli were presented with an ambient illumination of about 14 lx. Testing was performed in awake subjects that were restrained in a plastic decapicone (Braintree Scientiﬁc, Inc., Braintree, MA) with the front portion removed to uncover the nose, eyes, and ears. Ampliﬁers (Model 12 Neurodata Acquisition System, Grass Instrument Co., Quincy, MA) were calibrated with a custom built signal generator (Hamm et al., 2000). Flash intensity and ambient illumination were calibrated using a photometer (Model DR-2550, Gamma Scientiﬁc, Inc., San Diego, CA). The chamber background noise and masking auditory white noise (80 dB SPL) were measured at the rat’s ear level in the testing chamber using a 1.27 cm microphone (Model 4166, Brüel & Kjær, Denmark) and a measuring ampliﬁer (Model 2636, Brüel & Kjær). Flash evoked potentials were recorded for two consecutive days prior to the ﬁnal measurements, to allow an optimal development of late FEP components (such as the PhAD) (Dyer, 1989; Herr et al., 1994). On these two days, animals in the repeated exposure experiment were dosed immediately following testing to avoid suppression of the PhAD related to treatment. On the third day, rats’ pupils were dilated with 1–2 drops of a freshly prepared solution of 0.75% tropicamide and 2.5% phenylephrine (Sigma Chemical Co., St. Louis, MO). Ten minutes later, the animals were gavaged with corn oil, the mixture of carbaryl and propoxur, or the positive control dosages of carbaryl or propoxur singly. Flash evoked potentials were recorded 30 min following dosing with the pesticide(s) or vehicle. This time point approximates the time of maximal ChE inhibition following treatment with carbamate pesticides (Padilla et al., 2007). The rats were presented with 75 light ﬂashes, in the presence of 80 dB SPL acoustic masking white noise using a 10 cm speaker placed approximately 57 cm above the rat’s ear level. Concurrent presentation of white noise during photic stimulation masks the click associated with the strobe discharge, avoiding the confounding auditory evoked potentials (Herr et al., 1996; Shaw, 1992). The cortical responses were bandpass ﬁltered (0.1–1000 Hz), ampliﬁed (5000), and averaged. Colonic temperature was recorded during testing using a temperature probe (Models RET-1, Physitemp Instruments, Inc., Clifton, NJ) inserted approximately 8 cm rectally and connected to an external thermometer (Model TH-8, Physitemp Instrument). Testing lasted about 5 min. 2.4. Tissue collection and preparation, and ChE activity determination After physiological testing, the animals were moved to a separate room and rapidly sacriﬁced by decapitation (approximately 40 min after dosing). Trunk blood was collected in heparinized tubes (100 IU in 5 mL) and stored on ice. The brain was quickly removed, placed on ice, rinsed with chilled physiologic saline, and sectioned sagittally. The brain halves were weighed and immediately frozen on dry ice. The blood was centrifuged at 4 8C for 10 min at 950 g. The plasma was removed and immediately frozen on dry ice. Two hundred microliters of the pelleted erythrocytes were added to tubes containing 400 mL of 0.1 M sodium phosphate buffer (pH 8.0) with 1% Triton X-100 (ﬁnal dilution 1:2, v:v). All tissues were stored at 80 8C until ChE activity analysis. ChE activity was determined using a radioenzymatic method (Johnson and Russell, 1975), with slight modiﬁcations as previously described (Gordon et al., 2006). The radiometric assay was used because it minimizes the risk of ChE reactivation due to factors such as sample dilution, elevated temperatures, and long incubation times (Hunter and Padilla, 1999; Hunter et al., 1997). Although the assay measures the hydrolysis of acetylcholine, the contribution of butyrylesterases cannot be excluded without the inclusion of speciﬁc inhibitors of butyrylesterase. This is especially true of tissues such as rodent plasma, which has signiﬁcant butyrylesterase activity. Thus, the data are presented in terms of total ChE activity. We previously used the radiometric assay to evaluate the result of inhibition of brain ChE produced by acute treatment with carbaryl or propoxur J.-C. Mwanza et al. / NeuroToxicology 33 (2012) 332–346 on the PhAD of FEPs (Mwanza et al., 2008). Additionally, use of the radiometric method allows comparison of these results with those reported for the effects of seven carbamates (including carbaryl and propoxur) on brain ChE activity (McDaniel et al., 2007; Padilla et al., 2007). 2.5. Determination of carbaryl, 1-naphthol, and propoxur in plasma and brain tissue Plasma samples (500 mL) were extracted with 1.5 mL cold ethyl acetate. Samples were vortexed for 1 min, then centrifuged for 5 min at 1819 g and 4 8C. The ethyl acetate layer was removed and the extraction repeated twice. The combined extracts were evaporated to dryness at 35 8C and 10 psi by using a TurboVap LV (Caliper Life Science, Hopkinton, MA). The extract was resuspended in 50:50 (v/v) methanol:organic-free reagent water and ﬁltered prior to analysis by atmospheric pressure chemical ionization (APCI) liquid chromatography/mass spectrometry (LC/ MS) on an Accela MSQ Plus system (Thermo Electron Corp., San Jose, CA) in selected ion monitoring (SIM) mode and equipped with a Thermo Electron Corp. Hypersil Gold 1.9 mm column (50 mm 2.1 mm). The chromatography gradient was held at 70:30 (v/v) water:methanol for 1 min, before going to 5:95 water:methanol (v/v) over 9 min. The high organic composition was held for 2 min before returning to initial conditions. The ﬂow rate was 200 mL/min. Detection on the mass spectrometer occurred in positive-negative ion switching mode, with the corona discharge current set to 7 mA. The heater tube temperature was set to 450 8C. Analytes were quantitated by matrix-matched calibration. Method detection limits were 1 ng/mL, 3 ng/mL, and 5 ng/mL for carbaryl, propoxur, and 1-naphthol, respectively. The limits of quantitation for each analyte were 3 times higher. Half brain sections were homogenized 1:4 (w/v) with organicfree reagent water. A protein precipitation with cold acetonitrile (1 mL) was used to extract the analytes of interest from 500 mL aliquots of homogenate. The mixture was pulse-vortexed for 5 min, then centrifuged at 1810 g for 10 min at 4 8C. The supernatant was diluted 1:2 (v/v) with organic-free reagent water and ﬁltered prior to analysis. The extracts were analyzed on a Thermo TSQ Quantum Ultra AM liquid chromatography/tandem mass spectrometry (LC/MS/MS) system in atmospheric pressure 335 chemical ionization mode. A Waters (Milford, MA) Xbridge C18 3.5 mm, 2.1 mm 150 mm column was used for chromatographic separation. Mobile phase A was 50:50 (v/v) methanol:water; mobile phase B was 100% methanol. Analytes were eluted with a 200 mL/min ﬂow rate and a gradient that was held at 100% A for 1 min, then shifted to 10% A over 10 min, before returning to 100% A at 15 min. The mass spectrometer operated in positive-negative ion switching mode, with the corona discharge current set to 12 mA for negative ions and 7 mA for positive ions. The vaporizer temperature was 400 8C and the ion transfer tube was heated to 250 8C. Analytes were quantitated by selected reaction monitoring (SRM) of a matrix-matched curve with collision energies optimized for the monitored transitions. Method detection limits were 2 ng/g for both carbaryl and propoxur, and 20 ng/g for 1-napthol. The limits of quantitation were 4.8 ng/g for carbaryl and propoxur, and 48 ng/g for 1-naphthol. For brevity, the data for 1-naphthol are not presented in this manuscript, but is available from the corresponding author upon request. 2.6. Statistical analysis Statistical analysis was performed using SAS (SAS Institute Inc., 1989, 1997, 2004). Amplitudes (mV) of peaks N36 and N166 were measured from baseline (deﬁned as average voltage over a 50 ms pre-stimulus period), and latencies (ms) from stimulus onset were calculated (see Fig. 1A and B for average FEPs and peak identiﬁcation). Durations (ms) and areas (mV-ms) of peak N166 and PhAD were measured as previously described (Mwanza et al., 2008). Brieﬂy, we deﬁned the duration of peak N166 in the current study as the time elapsing between peak P102 and the next positive ‘‘peak’’ below baseline, and the duration of the PhAD as the time between the end of peak N166 and the intersection of the last discharge activity and the baseline. The areas over the duration of peak N166 and PhAD were used to deﬁne respective peak areas. An analysis of variance (PROC GLM) with a Greenhouse–Geisser correction factor, where appropriate, was used to analyze the physiological and analytical chemistry data (Geisser and Greenhouse, 1958; Greenhouse and Geisser, 1959). Dunnett’s test (a = 0.05) was used to compare group means between treated animals and control responses (Dunnett, 1980). Only treatmentrelated changes in the FEPs are presented. Regression analysis was Fig. 1. Group average FEP waveforms for animals treated either (A) acutely or (B) for 14 days with a mixture of 1:1.45 propoxur:carbaryl. The representative various peaks, areas, and durations are indicated for the acute 10 mg/kg dosage waveform. Note the minimal changes in the peak amplitudes or latencies. There was a trend for a decrease in the duration of the PhAD as the dosage increased, which is visible in the group averages. J.-C. Mwanza et al. / NeuroToxicology 33 (2012) 332–346 336 used to examine the relationship between the neurophysiological measures and the percent control brain ChE activity. Because the data from the control subjects were used to calculate the percent control ChE activity, the results from the control group were removed from the regressions. Additionally, only data from animals with greater than 10% inhibition of brain ChE activity were included in these linear regression analysis. This cutoff was included because 10% inhibition of brain ChE activity has been selected as the point of departure in the N-methyl carbamate cumulative risk assessment (US EPA, 2007), thus including only animals which had been ‘‘adversely’’ affected by the treatments. The regressions used the brain ChE activity data, assuming that the FEP (which is generated by the brain), would be more sensitive to local alteration of ChE activity levels. Data are reported as mean SE. Assessment of dose-additivity of the mixtures was performed using SAS v9.1 (SAS Institute Inc., 2004) and JMP v7 (SAS Institute Inc., 2007). Dose-addition assumes two characteristics relative to the components and the mixture. First is that the components display toxicologic proportionality. This means that the dose– response curves are geometrically similar once each chemical’s dosage is scaled by its toxic potency, and that the response variance only depends on the response value. Second, the doseaddition model derived from the component dose–response information is assumed to adequately predict the mixture response. For example, with a binary mixture, an exponential dose–response model, yi = exp(a + b*dosei), for the ith component’s response gives rise to the dose-addition model, ymix = exp(a + b1*dose1 + b2*dose2), for predicting the mixture response from the two components’ doses. Data from the merged single chemical dose–response studies (Mwanza et al., 2008) were used to create the predicted dose-additive responses. Data from an additional 4 rats/dosage for 0, 0.3, and 3 mg/kg propoxur were included in the single chemical dose–response analysis for brain and erythrocyte ChE activity. These data were included to improve the modeling for the low dosage region of the propoxur dose– response function. No physiological measures were included from these additional animals. The signiﬁcance testing for ChE activity for the single chemical data has been previously reported (Mwanza et al., 2008), and the reader is referred to the previous publication for details. Data for percent control brain and erythrocyte ChE activity, and the duration of the PhAD, were individually ﬁt using the models listed in Table 1 (expanded from Lyles et al., 2008). Each of these models can smoothly decrease with increasing dose, some have plateaus, and the more complex logistic and Gompertz models can have a low dose shoulder to approximate a threshold region of no difference from controls. The models incorporated the individual control ChE data, so that the modeled response at zero dose was not ﬁxed at 100% and reﬂected the full variation in the control group. For each endpoint, the best ﬁtting model was chosen based on lack of ﬁt statistics and the model with the smallest corrected Akaike Information Criteria (Burnham and Anderson, 2004). For a given endpoint, the same model was chosen for all data sets. After selecting the best ﬁtting model, a dose-additive Combined Prediction Model (CPM) was constructed by ﬁtting the model to Table 1 Models used to ﬁt single chemical and mixture dose–response data. Model Formula Exponential Exponential with Plateau 3-Parameter Logistic 4-Parameter Logistic 4-Parameter Gompertz y = exp(a + b*dose) y = plateau + exp(a + b*dose) y = a0 + ((a1 a0)/(1 + exp(dose b0))) y = d + (c/(1 + exp(a + b*dose))) y = a + g*exp( exp( (b0 + b1*dose))) the merged single-chemical data sets. This ﬁtting is based on the dose-additive assumption of nearly congruent dose–response curves for the individual components of a mixture and the mixture itself (US EPA, 2000). The selected function was also ﬁt to the mixture data, resulting in the Mixture Model (MM). To assess doseadditivity (Hertzberg et al., 2011; Abstract), the following steps were performed. First, the ﬁt of the CPM to the combined singlechemical data sets was assessed. A good ﬁt supports the common shape and proportionality of the dose–response functions among the single chemicals, as interpreted by relative potency factors (Hertzberg et al., 1999). Second, the ﬁt of the MM to the mixture data is assessed. A good ﬁt supports consistency of the shape of the dose–response functions for the single chemicals and the mixture. Third, the agreement between the CPM and the mixture data was evaluated by comparing the mean mixture responses with the prediction intervals of the CPM at each dose level. Fourth, the CPM and MM parameters were compared for consistency using a Wald test (Wald, 1943). If differences were indicated, individual parameters were then compared, using a Holm’s correction for multiple comparisons (Holm, 1979). 3. Results 3.1. Acute exposure 3.1.1. Physiological measures Acute treatment with the carbamate mixture did not result in large changes in the FEP waveforms (Fig. 1A). The data showed that the latency of peak N36 was not altered, but there was an increase in the latency of peak N166 (Dose Effect: F[4,55] = 3.08, p = 0.0232) (Fig. 2A and B). Treatment with 75 mg/kg of the mixture increased the latency of peak N166 compared to controls (Fig. 2B). Acute treatment with the carbamate mixture did not alter the amplitudes of peaks N36 or N166 (p-values > 0.05). Neither the area nor the duration of peak N166 was altered by treatment (p-values > 0.05). Similar to peak N166, neither the area nor duration of the PhAD was signiﬁcantly changed (p-values > 0.05). However, the decrease in the duration of the PhAD (Fig. 3C) was nearly signiﬁcant (p = 0.0695). Although the decrease in the PhAD duration was not signiﬁcant (see above), there was a signiﬁcant linear relationship between the duration of the PhAD and the percent control brain ChE activity (greater than 10% inhibition) (Fig. 4A). The slope of the regression (0.0315 0.0112, r2 = 0.1577) was signiﬁcantly different from zero (p = 0.0076). These data show that there was a decrease in the PhAD duration as the inhibition of brain ChE activity increased after acute treatment with the mixture of carbaryl and propoxur. This result is similar to what we have previously reported for carbaryl and propoxur administered singly (Mwanza et al., 2008). Treatment with the carbamate mixture resulted in a decrease in the animal’s colonic temperature (Fig. 2C) (Dose Effect: F[4,55] = 3.34, p = 0.0160). Compared with controls, signiﬁcant decreases were observed after treatment with 10 or 45 mg/kg of the mixture. 3.1.2. Cholinesterase activity Acute treatment with the carbamate mixture decreased ChE activity in plasma, brain, and erythrocyte tissues [Dose Effect: Fvalues[4,55] 27.27, p-values < 0.0001] (Figs. 2D, 3A and B). Control ChE activity averaged 293.4 5.9 nmol/min/mL in plasma, 4751.7 88.3 nmol/min/g in brain, and 509.3 47.6 nmol/min/mL in erythrocytes. Signiﬁcant decreases in ChE activity were indicated for the 3, 10, 45, and 75 mg/kg dosages of the mixture in all of the tissues. When analyzed on a percent of control activity basis, there were no signiﬁcant differences in the dose–response functions between the three tissues (Dose Tissue Interaction: p = 0.1022). J.-C. Mwanza et al. / NeuroToxicology 33 (2012) 332–346 (A) + 34 33 32 31 * 170 165 160 155 150 145 30 0 3 10 (C) 30 45 (D) ChE Activity (%Control, Mean ± SE) Colonic Temperature 39.0 * 38.5 * 38.0 * 37.5 * 37.0 0 3 10 30 45 0 3 10 75 39.5 Temperature ( o C, Mean ± SE) Peak N166 Latency 175 Latency (ms, Mean ± SE) 35 Latency (ms, Mean ± SE) (B) Peak N36 Latency 337 75 30 45 75 Plasma 120 100 * 80 * 60 * * * * 40 * * 20 0 3 10 30 45 75 Acute Repeated Dosage (mg/kg) Fig. 2. Changes in the latencies of peaks (A) N36 and (B) N166, (C) colonic temperature, and (D) plasma ChE activity after acute and 14 day treatment with the carbamate mixture. No change from control responses was noted for peak N36 latency, but the latency of peak N166 was increased after acute treatment with 75 mg/kg of the mixture. There was reduction in colonic temperature after acute or repeated treatment with the carbamate mixture. A dose-related decrease in plasma ChE activity was observed after acute and repeated treatment scenarios. Minimal differences were observed in the responses between the acute and repeated exposure scenarios. In this and all subsequent ﬁgures, missing error bars were either removed (due to overlap) for clarity, or are contained within the symbols. See text for details. *Signiﬁcantly different from controls, + signiﬁcant difference between acute and repeated treatments. 3.1.3. Assessment of dose-additivity Both the carbaryl and propoxur single-chemical dose–response curves for brain ChE activity from (Mwanza et al., 2008) were best ﬁt by the exponential function with a plateau (Fig. 5A). The parameter estimates (and those from the CPM and MM) are shown in Table 2. The CPM model was compared to the single chemical curves using a likelihood ratio test (LRT = 12.9, p = 0.005), which indicated that the single chemicals have different dose–response curves from the predicted dose-additive CPM model. The difference is related to the different variances of the carbaryl (0.199) and propoxur (0.828) models, which involves the strict interpretation of toxicologic similarity, i.e. that the variance only depends on the response value, not the chemical or dosage. However, the CPM accurately predicted the mean responses for each of the individual components. The exponential function with a plateau also ﬁt the brain ChE mixture data (MM), with a lack of ﬁt p-value = 0.17 (Fig. 3A). The standard error of the slope parameter from the MM is slightly smaller than for the single chemicals in the CPM, indicating a slightly better ﬁt to the mixture data. Next the CPM model was compared to the actual mixture data. The early plateau in the CPM results in an overestimation of the predicted brain ChE inhibition at the middle doses (Fig. 3A). Finally, the parameters of the CPM and MM were compared as an index of dose-additivity, and the similarity of the curves was rejected by a Wald test (WT[3,285] = 5.37, p = 0.0013), indicating a lack of dose-additivity. The individual parameters of the CPM and MM were compared and the background and plateau parameters were not signiﬁcantly different (p-values = 0.7288). But as suggested in Fig. 3A, the slope parameters differed between the CPM and MM (p = 0.0057). When comparing predicted responses from the CPM and the acute MM at individual dosages, it appears that the lack of dose-additivity is evident at the 3 (p = 0.0001) and 10 mg/kg dosages (p = 0.0002). The actual brain ChE activity was greater than the dose-additive (CPM) predicted values by 11.8% and 10.7% at the 3 and 10 mg/kg dosages, respectively. Similar to the brain ChE data, the single chemical erythrocyte ChE data from Mwanza et al. (2008) were best ﬁt by the exponential function with a plateau (Fig. 5B), and the parameter estimates are shown in Table 2. The dose-additive CPM model was compared to the single chemical dose–response curves using a likelihood ratio test (LRT = 2.7, p = 0.44), which indicated that the CPM, carbaryl, and propoxur dose–response curves were congruent. The exponential function with a plateau also ﬁt the erythrocyte ChE mixture data (MM), with a lack of ﬁt p-value = 0.87 (Fig. 3B). J.-C. Mwanza et al. / NeuroToxicology 33 (2012) 332–346 338 (A) (B) Brain 120 120 Acute Repeated Combined Prediction Model 100 + 700 100 * 80 * * 80 * * 60 * * 60 * PhAD Duration 800 Duration (ms, Mean ± SE) ChE Activity (Percent Control, Mean ± SE) (C) Erythrocytes + * * 600 500 400 300 40 40 * * * * 20 20 03 10 30 45 03 10 75 200 * 30 * 100 45 75 03 10 30 45 75 Mixture Dosage (mg/kg) Fig. 3. Cholinesterase activity in (A) brain and (B) erythrocyte tissues, and (C) the duration of the PhAD, after acute and 14 day treatment with the propoxur:carbaryl mixture. Brain and erythrocyte tissues showed dose-related decreases in ChE activity after acute or repeated treatments. No differences between acute and repeated exposures were observed for brain ChE activity, with slightly reduced levels of inhibition of ChE in erythrocytes after repeated exposures to the mixture. There was a trend for a decrease in the duration of the PhAD after acute treatment, with a signiﬁcant reduction in duration after repeated dosing with the carbamate mixture. Green (acute) and red (repeated) symbols are the mean response at each mixture dosage with standard error bars. Missing error bars were removed to increase clarity. See text for details. *Signiﬁcantly different from controls, +signiﬁcant difference between acute and repeated treatments. The model ﬁt to the actual mixture data (MM) are shown as solid green (acute) or red (repeated) lines, and the corresponding 95% conﬁdence limits are shown with dotted lines. The predicted dose-additive model (CPM) is shown as the black line, with black symbols representing the predicted response and 95% conﬁdence limits at the individual dosages of the mixture. The standard errors for the slope of the CPM and the MM were similar. When the CPM model was compared to the mixture data, it was observed to underestimate the degree of ChE inhibition at 45 and 75 mg/kg of the mixture. The parameters for the CPM and MM were compared, and the conclusion of dose-additivity was rejected by a Wald test (WT[3,285] = 3.58, p = 0.0144). The individual parameters of the CPM and MM had similar background (p = 0.2107) and slope parameters (p = 0.0590), but different plateau values between the predicted and actual mixture data (A) (p = 0.0039) (Fig. 3B). When comparing the predicted MM and CPM values at individual dosages, the estimates for 45 and 75 mg/kg of the mixture were found to differ between the two models. The dose-additive (CPM) model underestimated the actual amount of erythrocyte ChE inhibition following acute treatment by 14.7% and 18.4% at the 45 and 75 mg/kg dosages, respectively. In contrast to the ChE data, the exponential function without a plateau was found to be the best ﬁt to the PhAD duration data (Fig. 5C) presented in Mwanza et al. (2008), and the resulting (B) Acute Repeated 140 y = 0.0315x + 39.4233 r2 = 0.1577, p = 0.0076 120 slope = 0.0315 ± 0.0112 100 3 80 10 3 3 3 33 3 10 10 10 10 10 10 75 10 4545 10 45 45 1010 45 10 75 75 75 45 45 45 75 45 7545 75 45 7575 45 75 75 75 60 40 20 0 200 400 600 800 1000 1200 PhAD Duration (ms) % Control Brain ChE Activity % Control Brain ChE Activity 140 y = 0.0194x + 46.5637 r2 = 0.1422, p = 0.0076 120 slope = 0.0194 ± 0.0069 100 3 3 80 1030 10 4530 30 30 45 30 30 45 30 45 30 60 40 3 3 33 10 10 10 45 10 30 45 10 10 10 45 10 45 30 45 10 45 103030 30 10 45 45 30 45 45 10 20 0 200 400 600 800 1000 1200 PhAD Duration (ms) Fig. 4. Linear regressions of PhAD duration and percent control brain ChE activity (greater than 10% inhibition) after (A) acute or (B) repeated treatment with the mixture of propoxur and carbaryl. The regression slopes differed from zero (p-values < 0.05) for both exposure scenarios, indicating the presence of a linear relationship between the duration of the PhAD and brain ChE activity. The duration of the PhAD decreased as the brain ChE activity decreased. The solid lines indicate the regression line and the dashed lines indicate the 95% conﬁdence intervals. The symbols represent the dosages (mg/kg or mg/kg/day). J.-C. Mwanza et al. / NeuroToxicology 33 (2012) 332–346 (A) Carbaryl Propoxur 100 Erythrocytes 120 PhAD Duration 800 700 100 600 80 80 60 60 Duration (ms, Mean ± SE) ChE Activity (Percent Control, Mean ± SE) (C) (B) Brain 120 339 500 400 300 40 40 200 20 20 0 13 10 30 50 75 100 013 10 30 50 75 013 10 30 50 75 Dosage (mg/kg) Fig. 5. Cholinesterase activity in (A) brain and (B) erythrocyte tissues, and (C) the duration of the PhAD, after acute treatment with the single chemicals propoxur (pink triangles) or carbaryl (blue squares). As described in the text, the mean levels (SE) of ChE activity and PhAD duration have been described in Mwanza et al. (2008). Brieﬂy, brain or erythrocyte ChE was inhibited by 3 mg/kg or greater dosages of carbaryl or propoxur. The duration of the PhAD was decreased by 75 mg/kg carbaryl or 30 mg/kg propoxur. The dose-response models ﬁt to the single chemical data (exponential with a plateau for ChE data, and exponential for the duration of the PhAD), are shown as solid pink (propoxur) or blue (carbaryl) with the corresponding 95% conﬁdence limits are shown with dotted lines. The parameters for the models are detailed in Table 2. These single-chemical dose response functions were used to construct the predicted dose-additive CPM function. parameter estimates are shown in Table 2. A likelihood ratio test (LRT = 1.1, p = 0.58) compared the CPM model to the propoxur and carbaryl single chemical dose–response curves, and indicated that the three functions were congruent. The PhAD duration mixture data were also ﬁt using the exponential function without a plateau as the MM, with a lack of ﬁt p-value = 0.52 (Fig. 3C). The background and slope parameters were comparable with those from the CPM. The CPM was found to ﬁt the actual mixture data, with the mean responses within the curve’s conﬁdence limits (Fig. 3C). The parameters for the CPM and MM were compared, and the conclusion of dose-additivity was not rejected by a Wald test (WT[2,273] = 1.28, p = 0.2807). The conclusion of dose-additivity is further supported by nonsigniﬁcant differences for the predicted values of the CPM and MM at any of the individual dosages. 3.1.4. Tissue levels of carbamates Approximately 40 min after acute treatment with the carbamate mixture, both carbaryl and propoxur were detected in brain and plasma tissues (Fig. 6A–D). Neither carbaryl nor propoxur was detected in the brain or plasma of control animals. At 3 mg/kg of the mixture, carbaryl was detected in only 10 of 12 (10/12), and propoxur in only 9/12, brain samples. Data from several plasma samples were lost due to errors in sample preparation. At 3 mg/kg, carbaryl was detected in 7/10, and propoxur in 5/10, plasma samples. At 10 and 45 mg/kg, carbaryl was detected in 10/10, and propoxur in 9/10, plasma samples. At 75 mg/kg, both carbaryl and propoxur were detected in 11/11 plasma samples. Both carbaryl (Dose Effect: F-values[3,33] 4.49, p-values 0.0095) and propoxur (F-values[3,29] 3.30, p-values 0.0340) levels increased in both brain and plasma tissues in a dose-related manner. There was not a difference in the brain vs. plasma levels for either carbaryl or propoxur, as indicated by the lack of signiﬁcant tissue effects or tissue by dose interactions (p-values 0.0827). Signiﬁcant (p < 0.05; compared to 0 ng/mg) levels of carbaryl and propoxur were detected in brain and plasma after treatment with 3, 10, 45, or 75 mg/kg of the mixture. 3.2. Repeated exposure 3.2.1. Physiological measures Similar to acute treatment, dosing for 14 days with the mixture of propoxur and carbaryl did not result in dramatic changes in the FEP waveforms (Fig. 1B). Testing on the last treatment day produced signiﬁcant changes in the latency of peak N36 (Dose Effect: F[4,65] = 2.71, p = 0.0377) (Fig. 2A), but not in the latency of peak N166 (p > 0.05) (Fig. 2B). Post hoc tests did not indicate signiﬁcant differences in peak N36 latency from control values, and suggest that the ANOVA results are related to the differences in latencies between the 10 and 30 mg/kg/day groups. Neither of the amplitudes of peaks N36 or N166 were affected by repeated treatment with the carbamate mixture (p-values > 0.05). Similarly, the area and duration of peak N166 were not altered by the 14 day treatment regime (p-values > 0.05). Although the area of the PhAD was not signiﬁcantly altered (p > 0.05), the duration of the PhAD (Fig. 3C) was signiﬁcantly changed after 14 days of dosing with the mixture of propoxur and carbaryl (Dose Effect: F[4,65] = 3.23, p = 0.0177). Speciﬁcally, repeated treatment with 30 mg/kg/day of the carbamate mixture decreased the duration of the PhAD compared to the control group. As observed for the single chemicals propoxur and carbaryl (Mwanza et al., 2008), and in the acute treatment with the mixture of propoxur and carbaryl, there was a signiﬁcant linear relationship between the duration of the PhAD and the percent control brain J.-C. Mwanza et al. / NeuroToxicology 33 (2012) 332–346 340 Table 2 Parameter estimates for single chemical dose–response functions, Combined Prediction Models (CPMs), and Mixture Models (MMs) after acute or repeated treatment. Chemical/model Brain ChE as percent of control mean Carbaryl Propoxur CPM MM Parameter Estimate SE Background Slope Plateau Background Slope Plateau Background Carbaryl Slope Propoxur Slope Plateau Background 100.84 1.98 0.15 0.02 40.80 1.57 96.44 2.71 0.24 0.06 41.02 2.79 98.65 1.65 0.14 0.02 0.26 0.05 40.92 1.47 100.44 2.31a 101.86 2.10b 0.10 0.01 0.12 0.01 39.54 2.31 45.57 1.78 96.93 0.19 37.69 91.06 0.36 35.47 95.40 0.18 0.35 38.02 95.82a 97.68b 0.13 0.15 35.82 42.00 104.75 0.12 43.91 101.81 0.12 46.56 101.89 0.11 0.17 43.82 105.07a 106.05b 0.08 0.09 43.27 49.14 97.53 4.24 0.19 0.06 41.65 3.14 102.23 3.94 0.17 0.06 50.51 4.65 99.93 2.91 0.23 0.06 0.12 0.03 44.04 2.84 99.14 5.43a 103.69 5.95b 0.12 0.03 0.07 0.03 27.42 4.25 38.76 8.76 89.13 0.30 35.43 94.42 0.28 41.29 94.18 0.35 0.18 38.44 88.27a 91.83b 0.17 0.13 18.91 21.29 105.93 0.08 47.86 110.05 0.05 59.74 105.67 0.10 0.05 49.63 110.00a 115.55b 0.06 0.01 35.92 56.22 470.23 0.01 417.91 0.02 464.97 0.009 0.018 464.59a 471.78b 0.012 0.021 567.70 0.003 538.94 0.0004 540.72 0.002 0.004 638.01a 665.94b 0.001 0.003 Mixture Slope Plateau Erythrocyte ChE as percent of control mean Carbaryl Background Slope Plateau Propoxur Background Slope Plateau CPM Background Carbaryl Slope Propoxur Slope Plateau MM Background Mixture Slope Plateau PhAD duration Carbaryl Propoxur CPM MM Background Slope Background Slope Background Carbaryl Slope Propoxur Slope Background Slope a b 518.96 24.63 0.006 0.002 478.42 30.45 0.009 0.004 502.84 19.22 0.006 0.002 0.01 0.004 551.30 43.35a 568.86 48.67b 0.007 0.003 0.012 0.005 Lower 95% CI Upper 95% CI Estimates from acute exposure study. Estimates from repeated exposure study. ChE activity (greater than 10% inhibition) (Fig. 4B). The slope of the regression (0.0194 0.0069, r2 = 0.1422) was signiﬁcantly different from zero (p = 0.0076). Again, the data shows that there was a decrease in the PhAD duration as the inhibition of brain ChE activity increased after treatment with the mixture of carbaryl and propoxur. Similar to acute exposures, repeated treatment with the carbamate mixture affected the animal’s colonic temperature (Fig. 2C) (Dose Effect: F[4,65] = 15.55, p 0.0001). Treatment with 30 or 45 mg/kg/day of the carbamate mixture resulted in a decrease in colonic temperature compared to controls. Repeated treatment with the mixture of propoxur and carbaryl had minimal effects on the animals’ gain in body weight over the duration of the experiment (Fig. 7). There were no signiﬁcant differences in the body weight of the different treatment groups at the time of surgery (p > 0.05). The treatment groups had different rates of body weight gain over the 14 days of dosing (Dose Day Interaction: F[52,845] = 5.95, e = 0.1882, p < 0.0001). On individual days, differences in body weight were indicated on days 4–14 (Dose Effect: Fvalues[4,65] 2.66, p-values 0.0403). Dunnett’s post hoc tests indicated that the 30 mg/kg/day group had reduced body weight compared to controls on days 7, 11, 12, 13, and 14. 3.2.2. Cholinesterase activity As observed in the acute exposure, repeated treatment with the mixture of propoxur and carbaryl decreased ChE activity in plasma, brain, and erythrocyte plasma tissues [Dose Effect: F-values[4,65] 15.41, p-values < 0.0001] (Figs. 2D, 3A, and B). Control ChE activity averaged 292.8 11.2 nmol/min/mL in plasma, 4764.2 99.3 nmol/min/g in brain, and 481.8 26.8 nmol/min/mL in erythrocytes. Signiﬁcant decreases in ChE activity were indicated for the 3, 10, 30 and 45 mg/kg dosages of the mixture in brain and plasma, and the 10, 30, and 45 mg/kg/day dosages in erythrocytes. As in the acute exposures, there were no signiﬁcant differences in the dose– response functions between the three tissues (Dose Tissue Interaction: p = 0.2019) for percent of control ChE activity. J.-C. Mwanza et al. / NeuroToxicology 33 (2012) 332–346 (B) (A) Brain Carbaryl 4 3 * 2 1 # * * # # # 30 45 * 4 * 3 2 1 0 0 0 3 10 Plasma Carbaryl 5 * Carbaryl (ng/ µl, Mean ± SE) Carbaryl (ng/mg, Mean ± SE) 5 +* * # # # 0 3 10 75 30 45 75 (D) (C) 5 5 Brain Propoxur 4 3 2 * 1 0 # * * 0 3 10 # * # 30 45 Propoxur (ng/ µl, Mean ± SE) Propoxur (ng/mg, Mean ± SE) 341 Plasma Propoxur 4 3 2 1 + *# * # # 0 75 Dosage (mg/kg) 0 3 10 * * 30 45 75 Dosage (mg/kg) Acute Repeated Fig. 6. Levels of (A) brain carbaryl, (B) plasma carbaryl, (C) brain propoxur, and (D) plasma propoxur after acute and 14 day treatment with the propoxur:carbaryl mixture. Both tissues showed dose-related increases in pesticide levels with increasing dosages of the mixture. Low levels of carbaryl were detected in the brains of two animals in the repeated dosing study. The large variability in the plasma levels of propoxur and carbaryl for the 30 mg/kg/day group was due to a single animal. With the exception of the plasma for the 3 mg/kg groups, similar tissue levels were observed in brain and plasma with acute and 14 days of treatment. *Greater than zero tissue levels for acute study. # Greater than zero tissue levels for 14 day treatment study. +Signiﬁcant difference between acute and repeated treatments. 3.2.3. Comparison of acute vs. repeated treatment physiological responses and ChE activity The data were analyzed for changes in response between acute and repeated treatment with the carbamate mixture, with only minimal differences detected. Because some of the dosages differed between the two studies, only the common dosages were included in the analysis. The latency of FEP peak N36 for the 3 mg/ kg dosage was greater in the acute study than in the 14 day study (Study Effect: F[1,23] = 6.85, p = 0.0154) (Fig. 2A). None of the other physiological measures differed between the two exposure durations (p-values > 0.05). Minimal differences were noted between the studies for ChE inhibition and no differences were observed in the levels of ChE inhibition for brain or plasma tissues (p-values > 0.05). However, repeated treatment with the carbamate mixture resulted in less ChE inhibition in erythrocytes than acute exposures (Fig. 3B). These differences were signiﬁcant for the 3 (Study Effect: F[1,23] = 5.68, p = 0.0258) and 45 mg/kg dosages (Study Effect: F[1,25] = 7.74, p = 0.0101). When comparing the percent control brain ChE activity in either the acute or repeated treatment studies, or with previously reported acute single chemical dose–response experiments (Mwanza et al., 2008), there were no signiﬁcant differences after treatment with the positive control dosages of propoxur (30 mg/ kg; p-value = 0.3240) and carbaryl (75 mg/kg; p-value = 0.3359). For the 30 mg/kg propoxur dosage, the percent control brain ChE activities were: 37.4 4.3 (single chemical dose–response; Mwanza et al., 2008), 44.3 2.2 (acute study), and 41.2 2.8 (repeated treatment study). For the 75 mg/kg carbaryl dosage, the percent control brain ChE activities were: 42.7 1.5 (single chemical dose– response; Mwanza et al., 2008), 45.5 1.5 (acute study), and 42.9 1.2 (repeated treatment study). These results show that there was not a signiﬁcant shift in the dose–response functions for brain ChE inhibition by propoxur or carbaryl over the time required to complete these studies. 3.2.4. Assessment of dose-additivity Because single chemical dose–response functions were not generated for a 14 day dosing paradigm, the acute exposure dose– response functions (Fig. 5A–C) and CPM model (Fig. 3A–C) were used to evaluate deviations from dose-additivity for the repeated dosing study. This approach assumes that the acute data are an appropriate surrogate for the actual repeated exposure response. This assumption may be best applied to compounds with a similar mode of action and whose effects are reversible with a 24 h time frame, and is in agreement with the EPA’s N-methyl carbamate risk assessment (US EPA, 2007). J.-C. Mwanza et al. / NeuroToxicology 33 (2012) 332–346 342 Repeated Body Weight (g, Mean ± SE) 440 420 400 380 360 340 * * * 320 Surg 1 2 3 4 5 6 7 8 * * 9 10 11 12 13 14 Day 0 mg/kg 3 mg/kg 10 mg/kg 30 mg/kg 45 mg/kg Fig. 7. Changes in body weight over the 14 day dosing period with the carbamate mixture. No differences between groups were observed at surgery. The different treatment groups gained weight at different rates over the dosing period. The 30 mg/kg/day group had a reduced weight gain compared to controls. See text for details. *Signiﬁcantly different from controls. As observed in the acute exposure study, the exponential function with a plateau produced the best ﬁt for the brain ChE activity data (MM) after 14 days of treatment (Fig. 3A), and the parameter estimates are shown in Table 2. The MMs for acute and repeated exposures were then compared for similarity using a Wald test, which indicated that the acute and repeated MMs did not differ for brain ChE activity (WT[3,130] = 2.00, p = 0.1165). None of the individual model parameters was different between the acute and repeated MMs (p-values 0.0639). These comparisons indicate that the acute and 14 day exposures had similar dose–response functions for brain ChE activity. Next, the repeated exposure MM was compared to the acute CPM (Fig. 3A) for brain ChE activity. The standard error of the slope parameter in the repeated MM is slightly smaller than that for the single chemicals in the CPM, indicating a slightly better ﬁt to the mixture data. The acute CPM predicted lower ChE activity at the plateau, and had a steeper slope over the lower dosages, than observed in the repeated treatment data. The overall comparison of the acute CPM and repeated MM curves was rejected by a Wald test (WT[3,295] = 10.05, p 0.0001), indicating a lack of dose-additivity. The individual parameters of the acute CPM and repeated MM were then compared. The background and plateau parameters were not signiﬁcantly different (p-values = 0.6713, 0.0918), but as suggested in Fig. 3A, the slope parameters differed between the two models (p = 0.0387). When the acute CPM and the repeated MM were compared at individual dosages, differences were detected at the 3, 10, and 30 mg/kg dosages (p-values 0.0088). The actual brain ChE activity after repeated treatment was greater than the dose-additive predicted (CPM) values by 15.5%, 10.6%, and 5.8% at the 3, 10, and 30 mg/kg/day dosages, respectively. The erythrocyte ChE data (MM) for the repeated dosing study was also best ﬁt by the exponential function with a plateau, and the parameter estimates are shown in Table 2. The acute and repeated exposure MMs for erythrocyte ChE activity were statistically different (WT[3,130] = 3.95, p = 0.0099) (Fig. 3B). However, none of the individual model parameters differed between the acute and repeated dose–response models (p-values 0.2460). This is likely due to the size of the standard errors associated with the model parameters. When examined at the individual dosages, the predicted mean responses differed between the acute and repeated models for the 10 and 30 mg/kg estimates (pvalues 0.0177). When the parameters for the acute dosing CPM and repeated dosing MM were compared, the hypothesis of dose-additivity was rejected by a Wald test (WT[3,295] = 2.68, p = 0.0470). However, comparison of the individual parameters of the acute dosing CPM and repeated dosing MM did not indicate signiﬁcant differences (p-values 0.0957), indicating that the differences between the two models can not be explained by any single parameter. When comparing acute exposure dose-additive CPM predicted values to the repeated dosing MM predictions, or to the actual repeated dosing data, the CPM over predicted RBC ChE inhibition at the 3 and 10 mg/kg/day dosages (Fig. 3B). This over prediction was approximately 22.1% and 15.7% at the 3 and 10 mg/ kg/day dosages, respectively. As in the acute dosing study, the exponential function without a plateau was found to be the best MM ﬁt to the PhAD duration data after 14 days of repeated dosing (Fig. 3C). The parameter estimates are shown in Table 2. Comparison of the acute and repeated exposure dose–response MM curves indicated no statistical differences (WT[3,130] = 3.95, p = 0.5096), and none of the individual parameters associated with the two curves were signiﬁcantly different (p-values 0.2872). Similarly, none of the mean predicted values at any of the common doses were signiﬁcantly different for the two dose–response functions (pvalues 0.5744). These results are likely due to the size of the standard errors associated with the two functions. The acute exposure dose-additive CPM and repeated dosing observed MM were not signiﬁcantly different as indicated by a Wald test (WT[2,283] = 0.87, p = 0.4205), and there were no differences for any of the individual model parameters (p-values 0.1892). No signiﬁcant differences were indicated when the predicted values from the repeated dosing MM and acute exposure CPM were compared at individual dosages (p-values 0.2082). 3.2.5. Tissue levels of carbamates Approximately 40 min after 14 days of repeated treatment with the carbamate mixture, both carbaryl and propoxur were detected in brain and plasma tissues (Fig. 6A–D). Low levels of carbaryl were detected in the brain of 2 control animals. At 3 mg/kg of the mixture, carbaryl was detected in 12 of 13 (12/13), and propoxur in only 2/13, brain samples. In plasma, carbaryl was detected in 11/ 13, and propoxur in 4/13, samples from the 3 mg/kg treatment group. At 10 mg/kg, carbaryl was detected in 13/13, and propoxur in 10/13, plasma samples. In the 45 mg/kg treatment group, propoxur was detected in 14/15, while carbaryl was detected in 15/ 15, samples. A signiﬁcant dose-related effect was found only for brain carbaryl concentration (F[3,48] = 7.90, p = 0.0002). This was probably related to the low levels of brain propoxur detected after treatment with 45 mg/kg/day of the mixture (Fig. 6C). The lack of dose-related effects on plasma is due to the large variability in levels of carbaryl and propoxur in the 30 mg/kg/day group. This was caused by one animal with high plasma levels of both carbaryl and propoxur. This animal had only 15% of control erythrocyte ChE activity, providing further evidence for high plasma levels of pesticides. There was not a difference in the brain vs. plasma levels for either carbaryl or propoxur, as indicated by the lack of signiﬁcant tissue effects or tissue by dose interactions (pvalues 0.1055). Signiﬁcant (p < 0.05; compared to 0 ng/mg) levels of brain carbaryl were detected at all dosages. Brain propoxur was signiﬁcantly elevated at 10, 30, and 45 mg/kg day of the mixture. Plasma carbaryl and propoxur levels were greater than 0 at 3, 10, and 45 mg/kg/day, with the large variability at 30 mg/kg/day preventing statistical signiﬁcance. J.-C. Mwanza et al. / NeuroToxicology 33 (2012) 332–346 The pesticide tissue levels in brain and plasma were then analyzed for differences between the acute and repeated treatment studies using only the common dosages (0, 3, 10, and 45 mg/kg). In the brain, no differences in carbaryl or propoxur were observed between the two treatment paradigms. In plasma, differences between the acute and repeated treatments were observed only for the 3 mg/kg group (Fig. 6B and D). Levels of carbaryl after acute treatment (0.1351 0.0475 ng/mL) were greater than those measured following 14 day repeated treatment (0.0241 0.0049 ng/mL). Similarly, plasma levels of propoxur were greater after acute treatment (0.0280 0.0038 ng/mL) than after 14 days of dosing (0.0074 0.0011 ng/mL). 4. Discussion Both carbaryl and propoxur are members of the N-methyl carbamate class of pesticides which are reversible inhibitors of ChE enzymes. This similarity in mode of action is one factor that has led to a common grouping for cumulative risk assessment (US EPA, 2007). Recovery of brain ChE activity is nearly complete by 24 h after dosing with carbaryl (30 mg/kg) or propoxur (20 mg/kg) as single compounds (Padilla et al., 2007). At lower dosages, recovery of brain ChE activity may be complete within 24 h, with half-life for recoveries of 1.83 h for carbaryl and 2.69 h for propoxur (US EPA, 2007). The studies in this manuscript have expanded upon results following acute treatment with the single chemicals of propoxur and carbaryl (Mwanza et al., 2008), and examined a binary mixture of these carbamates on peripheral and central ChE inhibition, and on changes in neurophysiological activity in the central nervous system. Acute exposure to a mixture of seven carbamates (equitoxic formulation) was found to be dose-additive for inhibition of brain ChE activity (US EPA, 2007). The present studies expanded current knowledge through inclusion of a neurophysiological measure of brain function and comparing acute responses with those following 14 days of treatment. To our knowledge, this is the ﬁrst report of both acute and repeated exposure to a carbamate mixture. There were two primary questions of interest in this series of experiments. The ﬁrst was whether biological responses (PhAD duration, ChE inhibition) would change after repeated treatment with a carbamate mixture when compared to the responses following acute exposure. The second question was if the acute and repeated data were dose-additive when compared to the biological responses following single chemical exposure to carbaryl or propoxur. We examined peaks N36 and N166 because they reﬂect different aspect of neurophysiological processing of the visual signal. Peak N36 reﬂects the primary cortical response to the ﬂash-evoked thalamocortical afferent volley, and is thought to be generated by depolarization of cells in cortical lamina IV (Bigler, 1977; Brankačk et al., 1990; Givre et al., 1994; Kraut et al., 1985; Mitzdorf, 1987). Neither acute or repeated treatment with the carbamate mixture signiﬁcantly altered the amplitude of peak N36, suggesting that the initial cortical response to the light ﬂash was not altered by inhibiting brain ChE activity. These ﬁndings are consistent with our previous ﬁndings after acute treatment with propoxur or carbaryl (Mwanza et al., 2008). The PhAD depression produced by the carbamate mixtures is similar to the effects reported following acute exposure to the single chemicals (Mwanza et al., 2008). It is consistent with publications showing that treatment with ChE inhibitors decreased the amplitude of late components of the cortical response to ﬂash stimuli (Bigler and Fleming, 1976; Bigler et al., 1978; Fleming et al., 1974; Hetzler and Smith, 1984). The depression of the PhAD duration after acute treatment with the carbamate mixture was related to percent control brain ChE activity with a 343 slope (0.0315 0.0112) similar to that observed after acute treatment with the single chemicals carbaryl (0.0234 0.0082) or propoxur (0.0319 0.0100) (Mwanza et al., 2008). This ﬁnding shows that acute treatment with the mixture of propoxur and carbaryl did not alter the relationship between the inhibition of brain ChE activity and depression of the PhAD duration observed after acute treatment with the single chemicals. The current data again suggest a role of acetylcholine in the modulation of the PhAD, and by extension the cortical processing of visual information. There did not appear to be large differences in response magnitude of the physiological or biochemical measures between the acute and repeated treatment paradigms (Figs. 2A–D, 3A–C). No differences were indicated for peak N166 latency, or colonic temperature. A reduced latency for peak N36 was indicated after repeated exposure compared with acute treatment. In addition to comparisons at common dosages, the analysis for ChE activity in brain and erythrocytes, and the duration of the PhAD, included comparison of the dose–response MM functions ﬁt to the actual data. No differences were found for plasma or brain ChE activity, or the duration of the PhAD, when the acute and repeated exposures were compared. Additionally, the relationship between the PhAD duration and brain ChE activity was ﬁt using a linear regression following either acute or repeated exposures (Fig. 4A and B). The similarity of the slope estimates suggests that the sensitivity of the physiological response to ChE inhibition did not change over the 14 day dosing period. Thus, the overall conclusion from these studies is that the animals’ responses to repeated treatment with the carbamate mixture did not differ from those following a single treatment. This suggests that there was no accumulation of, nor adaptation to, the biological effects of the mixture. Additionally, as predicted, the PhAD duration was related to the level of brain ChE activity, independent of which carbamate or mixture was responsible for the ChE inhibition. The only measure where a difference between the acute and repeated exposures was indicated was for erythrocyte ChE activity (Fig. 3B) with the 14 day treatment paradigm resulting in less inhibition of erythrocyte ChE activity than the acute treatment at the 3 (22%) and 45 (13%) mg/kg dosages. Additionally, when the data were ﬁt using the MM exponential function with a plateau, overall differences in the two curves were found, and less predicted inhibition of erythrocyte ChE activity after repeated treatment with 10 or 30 mg/kg of the carbamate mixture was indicated. The slight reduction in levels of RBC ChE inhibition after repeated exposure to the mixture may reﬂect small compensation in peripheral tissues following repeated exposures. This ﬁnding awaits replication for determining its biological implications. For several of the dependent measures in the repeated exposure study (colonic temperature, PhAD duration, body weight gain), it appeared that the 30 mg/kg dosage of the carbamate mixture resulted in larger changes from controls than the 45 mg/kg dosage (Figs. 2C, 3C, and 7). The response produced by the 30 mg/kg dosage also appeared to diverge from the acute exposure dose– response function for these same endpoints. It is important to note that the inhibition of brain and erythrocyte ChE activity did not appear to be greater than predicted at this dosage from either the acute or repeated exposure dose–response functions (Fig. 3A and B). It is interesting to note that the animals in the 30 mg/kg group had the lowest body weights on the ﬁrst treatment day, and reduced weight gain was observed throughout the study (Fig. 7). However, it is currently unknown why repeated treatment with the 30 mg/kg dosage may have resulted in greater changes in certain physiological measures than the 45 mg/kg dosage. Because neurophysiological measures are known to be altered by changes in body temperature (Boyes et al., 1993; Hetzler et al., 1988; Janssen et al., 1991), and inhibition of ChE is known to result in hypothermia in rodents (Mack and Gordon, 2007; Moser, 1995; 344 J.-C. Mwanza et al. / NeuroToxicology 33 (2012) 332–346 Moser et al., 1988), we included this endpoint in our physiological assessments. When recording FEPs, peak latencies are increased by hypothermia to a greater extent than changes in peak amplitude (Boyes et al., 1993; Hetzler and Krekow, 1999; Hetzler et al., 1988). The increased peak latencies may be a result of the known decrease in nerve conduction velocity with decreases in temperature (Miyoshi and Goto, 1973; de Jesus et al., 1973; Petajan, 1968; Halar et al., 1980; Rutten et al., 1998). As shown in Fig. 2A, the latency of peak N36 did not differ from controls in the acute or repeated exposure studies. Acute treatment with 75 mg/kg of the mixture resulted in slight increases (about 13.6 ms) in the latency of peak N166 compared to controls with a decrease in body temperature of about 0.6 8C. This hypothermic response is less than that produced by 30 mg/kg of the mixture in the repeated dosing study (about 1.3 8C), when peak N166 latency was not altered. Because hypothermia would be expected to produce a generalized increase in peak latencies (rather than only affecting speciﬁc portions of the evoked response), it is unlikely that the observed results were simply due to hypothermia produced by inhibition of ChE activity. Prior to undertaking an analysis of dose-additivity, it was important to determine if there were any shifts in the dose– response curves for brain ChE inhibition between the single chemical studies (Mwanza et al., 2008), and the mixture experiments reported in this manuscript. We included positive control dosages of 30 mg/kg propoxur and 75 mg/kg carbaryl in both the acute and repeated exposure mixture studies. The lack of differences in the biological responses between the singlechemical and mixture experiments indicates that there were no large changes in the dose–response functions over time between the studies. In turn, this permits assessment of dose-additivity without concurrently determining full single-chemical dose– response functions. We used a predicted model (CPM) constructed from the single chemical data (Mwanza et al., 2008) (plus the additional propoxur ChE data) to assess deviations from dose-additivity. This approach has been used in many previous studies of mixtures of chemicals (Altenburger et al., 2005; Gennings et al., 2004a,b, 2005; Gordon et al., 2006; Moser et al., 2006). Because the effects of carbamate pesticides are largely dissipated 24 h after treatment (Moser et al., 2010; Padilla et al., 2007; US EPA, 2007), we used the data from the acute studies to predict the effects after repeated exposures. Deviations from the predicted response might be explained by either pharmacokinetic or pharmacodynamic changes after repeated exposures. This four step approach to assessing doseadditivity and using acute dose-additive predictions to assess additivity after repeated exposures are novel aspects of this report. We assessed dose-additivity for measures of central and peripheral ChE activity, and a measure of central nervous system physiological function. While non-additive effects were indicated for speciﬁc endpoints or dosing scenarios, the data overall support the conclusion of dose-additivity after acute or repeated exposure to the carbamate mixture. This conclusion is consistent with EPA’s regulatory ﬁnding of dose-additivity for inhibition of brain ChE activity by carbamates (US EPA, 2007). One should note that the EPA model requires the control mean to be 100% at a dosage equal to zero, and thus the model does not reﬂect the control group variability. Additionally, the EPA model also used robust iteratively re-weighted regression. These procedural differences may be involved in the small deviations from dose-additivity that we report, compared to EPA’s regulatory ﬁnding. Brain ChE activity showed some differences from the predicted dose-additive response. The actual mixture dose–response models (MM) were not different between the acute and repeated exposure scenarios for brain ChE activity. However, the acute exposure brain ChE results suggested a less than dose-additive response at 3 and 10 mg/kg (Fig. 3A) between the CPM and MM predicted responses. Similar to the acute data, repeated treatment with the mixture also indicated less than dose-additive responses at the 3, 10, and 30 mg/ kg/day dosages (Fig. 3A). In both the acute and repeated treatment scenarios, the deviations from dose-additivity were reﬂected in the signiﬁcant difference in the slope parameter between the predicted dose-additive (CPM) vs. acute or repeated exposure (MM) models. Although the CPM ﬁt the mean brain ChE activity levels in the single chemical dose–response functions, there was a difference in the variances between the CPM and the single chemical dose–response models (Figs. 3A and 5A). The strict interpretation of toxic similarity for a chemical class is that the response variance only depends on the response value, not factors such as the chemical or the dosage. In this case, that aspect of toxic similarity does not hold. However, the dose-additive predictive model (CPM) is highly accurate in estimating the mixture brain ChE response. The single chemical propoxur dose–response data (Fig. 5A) (Mwanza et al., 2008, with additional ChE data) have a non-monotonic shape, with slightly less ChE inhibition at 40 than at 30 mg/kg. The impact of that non-monotonicity on the CPM slope remains to be investigated. Additionally, the biological signiﬁcance of less than additive results at low-middle dosages remains to be replicated. The erythrocyte ChE data suggested non-additive effects after both acute and repeated treatments. In contrast to the brain ChE results, the ChE activity in erythrocytes suggested a greater than additive response at the high doses (45 and 75 mg/kg) after acute treatment (Fig. 3B). This effect was reﬂected in the signiﬁcantly different plateau estimates between the predicted dose-additive (CPM) and observed acute (MM) responses. Unlike the brain ChE models, the erythrocyte ChE models for the observed data (MMs) differed between the acute and repeated exposure scenarios. The modeled ChE activity following repeated treatment was greater (less inhibition) than the modeled ChE activity after acute exposure (MM) at the 10 and 30 mg/kg/day dosages. Also, the observed erythrocyte ChE activity after the repeated treatment was greater than predicted by the dose-additive CPM model at these same dosages. In both comparisons, no single parameters were responsible for the differences. The dose-additive prediction model (CPM) described the single-chemical dose–response functions, and had similar variance estimates, supportive of the concept of toxicologic proportionality between the two carbamates (Figs. 3B and 5B). These results would indicate greater than additive responses at high dosages after acute treatment, and less than additive effects at low-middle dosages after repeated exposures for erythrocyte ChE activity. However, readers should note the wider 95% conﬁdence limits in the erythrocyte data compared to the brain ChE data. This increased variability suggests that the high and middle dosage effects should be replicated before being used in a risk assessment process. The data from the functional measure of brain activity (PhAD duration), is consistent with dose-additivity after acute or repeated treatment with the carbamate mixture. The mixture models (MMs) for the data from the acute and repeated exposures did not differ in either parameter estimates or predicted response values. The predicted acute dose-additive model (CPM) was consistent with both the single-chemical dose–response functions, the observed acute and repeated responses (their respective MMs), and the actual mixture data from the two experiments (Figs. 3C and 5C). This conclusion must be tempered by the larger conﬁdence limits for this measure, requiring greater deviations from dose-additivity for detection, and the shallow dose–response functions (Fig. 5C) for the single chemicals (Mwanza et al., 2008) that are used in constructing the CPM. The PhAD is a higher level neurological function, involving thalamo-cortical circuits and multiple neurotransmitter systems (Bigler, 1975, 1977; Bigler et al., 1974). Thus, J.-C. Mwanza et al. / NeuroToxicology 33 (2012) 332–346 while depressed by ChE inhibitors, it is less directly linked in a causal manner to the inhibition of brain ChE activity. The PhAD data show that higher level endpoints that may not be directly caused by a single biochemical perturbation can be altered in a dose-additive manner after toxic insult. In general, the lack of differences in tissue levels of propoxur and carbaryl between the acute and repeated dosing paradigms supports the conclusion that there were no large changes in pharmacokinetics with repeated exposures. At the doses and time measured, neither pesticide showed evidence of bioaccumulation (increased tissue levels), nor large induction of metabolism (decreased tissue levels), with 14 days of repeated exposures (Fig. 6A–D). Therefore, alterations in tissue levels of the pesticides do not support any conclusions about differences in doseadditivity between the two dosing paradigms. In summary, we report that acute and 14 days of treatment with a mixture of propoxur and carbaryl resulted in similarly reduced levels of brain ChE activity and similar depression of the duration of the PhAD. The levels of ChE activity in erythrocytes was slightly less inhibited after repeated, compared to acute, treatment with the carbamate mixture. A dose-additive model constructed from data after acute treatment was used to predict the response following repeated exposures due to the reversible nature of the ChE inhibition produced by carbamates. Minor deviations from the predicted dose-additive model were observed in the low-middle treatment dosages after acute or repeated treatment for brain ChE, and after repeated treatment for erythrocyte ChE, suggesting a less than additive response. In contrast, acute treatment with the mixture resulted levels of erythrocyte ChE activity that suggested a greater than additive response at higher dosages. The results at the low-middle dosage levels are difﬁcult to interpret, and the magnitude of the differences was approximately in the 10–20% range. These differences are less than the order of magnitude changes of concern for greater than additive responses in human risk assessment. Additionally, the data do not suggest that repeated treatment with the mixture resulted in a greater response than that observed after acute treatment. Because the dosages used in these studies are greater than anticipated human exposures, concerns for non-dose-additive interactions following environmental exposures are minimal. Brain and plasma levels of propoxur and carbaryl were similar with the two dosing paradigms, indicating that large metabolic changes were not produced with repeated treatment. The results from these studies suggest that in the absence of repeated exposure data for the single chemicals, a dose-additive model constructed from acute data may serve as a good approximation for the response after repeated exposures for chemicals that have reversible effects over the dosing interval and are rapidly cleared from the body. Funding United States Environmental Protection Agency. Toxicology Excellence for Risk Assessment was funded by EPA Contract #GS10F0369N. Conﬂict of interest statement The authors declare that they do not have any actual or potential conﬂict of interest including any ﬁnancial, personal or other relationships with other people or organizations that could inappropriately inﬂuence, or be perceived to inﬂuence, this work. Acknowledgements We thank Drs. Anna Lowit and Bruce Hetzler for helpful comments on an earlier version of this manuscript. Appreciation is 345 also expressed to Mr. Charles Hamm for computer assistance, to Mr. Jackie Farmer for helping with dosing, and to Mr. Mark Bercegay, Ms. Jaimie Graff, and Dr. Laura Degn with technical assistance with experimental procedures. References Altenburger R, Schmitt H, Schüürmann G. Algal toxicitiy of nitrobenzenes: combined effect analysis as a pharmacological probe for similar modes of interaction. Environ Toxicol Chem 2005;24(2):324–33. Bigler ED. Lateral geniculate multiple-unit activity related to metrazol potentiated after-discharges. Electroencephalogr Clin Neurophysiol 1975;39:491–7. Bigler ED. Neurophysiology, neuropharmacology and behavioral relationships of visual system evoked after-discharges: a review. Biobehav Rev 1977;1(2):95–112. Bigler ED, Fleming DE. Pharmacological suppression of photically evoked afterdischarges in rats: incremental dose, hippocampal EEG and behavioral correlates. Psychopharmacologia 1976;46(1):73–82. Bigler ED, Fleming DE, Shearer DE. Pharmacological modulation of photically evoked after discharge patterns in hooded Long–Evans rats. Bull Psychon Soc 1974; 4(3):179–81. Bigler ED, Shearer DE, Dustman RE, Fleming DE. Differential effects of convulsants on visually evoked responses in the albino rat. Pharmacol Biochem Behav 1978;8(6):727–33. Boyes WK, Hetzler BE, Dyer RS. Acute effects of ethanol on pattern reversal and ﬂashevoked potentials in rats and the relationship to body temperature. Int J Psychophysiol 1993;14:27–39. Brankačk J, Schober W, Klingberg F. Different laminar distribution of ﬂash evoked potentials in cortical areas 17 and 18b of freely moving rats. J Hirnforsch 1990; 31(4):525–33. Burnham KP, Anderson DR. Multimodel inference: understanding AIC and BIC in model selection. Sociolog Meth Res 2004;33:261–304. Creel D, Dustman RE, Beck EC. Intensity of ﬂash illumination and the visually evoked potential of rats, guinea pigs and cats. Vision Res 1974;14:725–9. Crofton KM, Sheets LP. Evaluation of sensory system function using reﬂex modiﬁcation of the startle response. J Am Colleg Toxicol 1989;8(1):199–211. de Jesus PV Jr, Hausmanowa-Petrusewicz I, Barchi RL. The effect of cold on nerve conduction of human slow and fast nerve ﬁbers. Neurology 1973;23:1182–7. Dunnett CW. Pairwise multiple comparisons in the homogenous variance, unequal sample size case. J Am Stat Assoc 1980;75(372):789–95. Dyer RS. The use of sensory evoked potentials in toxicology. Fundam Appl Toxicol 1985;5:24–40. Dyer RS. Peak N160 of rat ﬂash evoked potential: does it reﬂect habituation or sensitization? Physiol Behav 1989;45(2):355–62. Fleming DE, Rhodes LE, Wilson CE, Shearer DE. Time-drug modulations of photically evoked after-discharge patterns. Physiol Behav 1972;8(6):1045–9. Fleming DE, Shearer DE, Creel DJ. Effect of pharmacologically-induced arousal on the evoked potential in the unanesthetized rat. Pharmacol Biochem Behav 1974;2(2):187–92. Geisser S, Greenhouse SW. An extension of Box’s results on the use of the F distribution in multivariate analysis. Annal Mathemat Stat 1958;29:885–91. Gennings C, Carter WH Jr, Carchman RA, Teuschler LK, Simmons JE, Carney EW. A unifying concept for assessing toxicological interactions: changes in slope. Toxicol Sci 2005;88(2):287–97. Gennings C, Carter WH Jr, Carney EW, Charles GD, Gollapudi BB, Carchman RA. A novel ﬂexible approach for evaluating ﬁxed ratio mixtures of full and partial agonists. Toxicol Sci 2004a;80(1):134–50. Gennings C, Carter WH Jr, Casey M, Moser V, Carchman R, Simmons JE. Analysis of functional effects of a mixture of ﬁve pesticides using a ray design. Environ Toxicol Pharmacol 2004b;18:115–25. Givre SJ, Schroeder CE, Arezzo JC. Contribution of extrastriate area V4 to the surfacerecorded ﬂash VEP in the awake macaque. Vision Res 1994;34(4):415–38. Gordon CJ, Herr DW, Gennings C, Graff JE, McMurray M, Stork L, et al. Thermoregulatory response to an organophosphate and carbamate insecticide mixture: testing the assumption of dose-additivity. Toxicology 2006;217(1):1–13. Greenhouse SW, Geisser S. On methods in the analysis of proﬁle data. Psychometrika 1959;24(2):95–112. Halar EM, DeLisa JA, Brozovich FV. Nerve conduction velocity: relationship of skin, subcutaneous and intramuscular temperatures. Arch Phys Med Rehabil 1980;61:199–203. Hamm CW, Ali JS, Herr DW. A system for simultaneous multiple subject, multiple stimulus modality, and multiple channel collection and analysis of sensory evoked potentials. J Neurosci Meth 2000;102(2):95–108. Herr DW, Boyes WK. Electrophysiological analysis of complex brain systems: sensoryevoked potentials and their generators. In: Chang LW, Slikker Jr W, editors. Neurotoxicology. Approaches and methods. Academic Press: New York; 1995 , pp. 205–21. Herr DW, Boyes WK, Dyer RS. Rat ﬂash-evoked potential peak N160 amplitude: modulation by relative ﬂash intensity. Physiol Behav 1991;44(2):355–65. Herr DW, King D, Grifﬁn VT, Watkinson WP, Boyes WK, Ali JS, et al. Within-session changes in peak N160 amplitude of ﬂash evoked potentials in rats. Physiol Behav 1994;55(1):83–99. Herr DW, Vo KT, King D, Boyes WK. Possible confounding effects of strobe clicks on ﬂash evoked potentials in rats. Physiol Behav 1996;59(2):325–40. 346 J.-C. Mwanza et al. / NeuroToxicology 33 (2012) 332–346 Hertzberg RC, Rice G, Teuschler LK. Methods for health risk assessment of combustion mixtures. In: Roberts SM, Teaf CM, Bean JA, editors. Hazardous Waste Incineration. Evaluating the Human Health and Environmental Risks. Boca Raton, FL: CRC Press; 1999, pp. 105–48. Hertzberg R, Pan Y, Li R, Haber LT, Lyles R, Moser VC, et al. A four-step approach for evaluation of dose-additivity. Toxicologist CD – An Off J Soc Toxicol 2011;120(S2):496. Hetzler BE, Boyes WK, Creason JP, Dyer RS. Temperature-dependent changes in visual evoked potentials of rats. Electroencephalogr Clin Neurophysiol 1988;70:137–54. Hetzler BE, Krekow LK. Temperature dependence and independence of effects of pentobarbital on visual evoked potentials of rats. Neurotoxicol Teratol 1999;21(2):181–91. Hetzler BE, Smith TM. Cholinergic role in alcohol’s effects on evoked potentials in visual cortex of the albino rat. Alcohol 1984;1:305–13. Holm S. A simple sequentially rejective multiple test procedure. Scand J Stat 1979;6:65–70. Hunter DL, Marshall RS, Padilla S. Automated instrument analysis of cholinesterase activity in tissues from carbamate-treated animals: a cautionary note. Toxicol Meth 1997;7(1):43–53. Hunter DL, Padilla S. Inﬂuence of storage conditions on the stability of cholinesterase activity in plasma and brain tissue taken from carbamate or organophosphorus pesticide-treated rats. Toxicol Meth 1999;9(3):189–99. Janssen R, Hetzler BE, Creason JP, Dyer RS. Differential impact of hypothermia and pentobarbital on brain-stem auditory evoked responses. Electroencephalogr Clin Neurophysiol 1991;80:412–21. Johnson CD, Russell RL. A rapid, simple radiometric assay for cholinesterase, suitable for multiple determinations. Anal Biochem 1975;64:229–38. Kraut MA, Arezzo JC, Vaughan H.G.Jr. Intracortical generators of the ﬂash VEP in monkeys. Electroencephalogr Clin Neurophysiol 1985;62:300–12. Lyles RH, Poindexter C, Evans A, Brown M, Cooper CR. Nonlinear model-based estimates of IC50 for studies involving continuous therapeutic dose-response data. Contemp Clin Trials 2008;29:878–86. Mack CM, Gordon CJ. Differential sensitivity to anticholinesterase insecticides in the juvenile rat: Effects on thermoregulation. J Toxicol Environ Health 2007;70(Part A):439–44. Mattsson JL, Albee RR. Sensory evoked potentials in neurotoxicology. Neurotoxicol Teratol 1988;10(5):435–43. McDaniel KL, Padilla S, Marshall RS, Phillips PM, Podhorniak L, Qian Y, et al. Comparison of acute neurobehavioral and cholinesterase inhibitory effects of N-methylcarbamates in rat. Toxicol Sci 2007;98(2):552–60. Mitzdorf U. Properties of the evoked potential generators: current source-density analysis of visually evoked potentials in the cat cortex. Int J Neurosci 1987;33(1–2):33–59. Miyoshi T, Goto I. Serial in vivo determinations of nerve conduction velocity in rat tails. Physiological and pathological changes. Electroencephalogr Clin Neurophysiol 1973;35:125–31. Moser VC. Comparisons of the acute effects of cholinesterase inhibitors using a neurobehavioral screening battery in rats. Neurotoxicol Teratol 1995;17(6): 617–25. Moser VC, McCormick JP, Creason JP, MacPhail RC. Comparison of chlordimeform and carbaryl using a functional observational battery. Fundam Appl Toxicol 1988;11:189–206. Moser VC, McDaniel KL, Phillips PM, Lowit AB. Time-course, dose-response, and age comparative sensitivity of N-methyl carbamates in rats. Toxicol Sci 2010; 114(1):113–23. Moser VC, Simmons JE, Gennings C. Neurotoxicological interactions of a ﬁve-pesticide mixture in preweanling rats. Toxicol Sci 2006;92(1):235–45. Mwanza J-C, Finley D, Spivey CL, Graff JE, Herr DW. Depression of the photic after discharge of ﬂash evoked potentials by physostigmine, carbaryl and propoxur, and the relationship to inhibition of brain cholinesterase. Neurotoxicology 2008; 29(1):87–100. Padilla S, Marshall RS, Hunter DL, Lowit A. Time course of cholinesterase inhibition in adult rats treated acutely with carbaryl, carbofuran, formetanate, nethomyl, methiocarb, oxamyl or propoxur. Toxicol Appl Pharmacol 2007;219(2–3):202–9. Padilla S, Setzer W, Marshall RS, Hunter DL, Phillips P, McDaniel K, et al. A doseresponse study of the toxicity of a mixture of 7 N-methyl carbamate pseticides in adult, male rats. Toxicol CD – An Off J Soc Toxicol 2006;90:9. Petajan JH. Changes in rat ventral caudal nerve conduction velocity during cold exposure. Am J Physiol 1968;214(1):130–2. Pohl HR, Hansen H, Chou C-HSJ. Public health guidance values for chemical mixtures: current practice and future directions. Reg Toxicol Pharmacol 1997;26(3):322–9. Rutten GJM, Gaasbeek RDA, Franssen H. Decrease in nerve temperature: a model for increased temporal dispersion. Electroencephalogr Clin Neurophysiol 1998;109(1):15–23. SAS Institute Inc.. SAS/STAT User’s Guide Version 6 Fourth Edition Cary NC 27513, vol. 2. SAS Institute Inc.; 1989 pp 1–846.. SAS Institute Inc. SAS/STAT Software: Changes and Enhancements through Release 6.12. Cary, NC 27513, SAS Institute Inc., 1997. pp 1–1167. SAS Institute Inc. SAS/STAT1 9.1. 1-7, Cary, NC 27513, SAS Institute, Inc., 2004. pp. 1–5121. SAS Institute Inc., JMP Release 7 Statistics and Graphics. Cary, NC 27513, SAS Institute, Inc., 2007. pp. 1–1076. Schroeder CE, Tenke CE, Givre SJ, Arezzo JC, Vaughan HG Jr. Striate cortical contribution to the surface-recorded pattern-reversal VEP in the alert monkey. Vision Res 1991;31(7/8):1143–57. Senate and House of Representatives of the United States of America Food Quality Protection Act of 1996. Public Law 104-170, 110 Stat. 1489, pp. 1489–1538. Shaw NA. Auditory potentials elicited by the Grass photic stimulator in the rat. Physiol Behav 1992;52(2):401–3. Sumitomo I, Klingberg F. The role of lateral geniculate body in the generation of photically and electrically evoked afterdischarges in freely moving rats. Acta Biol Med Germ 1972;29:43–54. United States Environmental Protection Agency. Supplementary Guidance for Conducting Health Risk Assessment of Chemical Mixtures. 2000; EPA/630/R-00/002, 1-C-10. United States Environmental Protection Agency. Revised N-methyl carbamate cumulative risk assessment. 2007; http://www.epa.gov/oppsrrd1/REDs/nmc_revised_cra.pdf (accessed 05.11.2011), pp. 1–277. Wald A. Tests of statistical hypotheses concerning general parameters when the number of observations is large. Trans Am Mathemat Soc 1943;54(3):426–82.