Effects of Environmentally Relevant Mixtures of Persistent Organic Pollutants on the Developmental Neurobiology in Rats

Abstract

We report the developmental neuropathology for rat pups at postnatal day (PND) 37 and PND 77 and the molecular biomarkers for PND 35, 75, and 350 after perinatal exposure to a reconstituted mixture of persistent organochlorine pollutants (POPs) based on the blood profiles of people living in the Great Lake Basin. The developmental neuropathology included routine histopathology evaluation, quantification of cell proliferation and death in the subventricular zone, linear morphometric measurements, and transcriptional analysis. No histopathological, structural, or stereological changes were observed in animals treated with the POPs or Aroclor 1254, on PND 37 or PND 77. While no transcriptional changes were found in Arcolor-treated animals, significant transcriptional changes were observed on PND 350 in female offspring perinatally exposed to 0.13 mg/kg of the POP mixture. Markers of the cholinergic system including acetylcholinesterase and the muscarinic receptors (subtypes M1–M5) were downregulated 2- to 6-fold. In addition, structural genes including neurofilaments (NFLs) and microtubule-associated protein (MAP-2) were downregulated at least 2-fold or greater. Our results support that in utero and lactational exposure to the chemical mixture of POPs lead to developmental changes in adult rat brains.

Keywords

biomarkers neuropathology toxicologic pathology environmental toxicology chemical mixtures.

Introduction

The human health risk posed by exposure to persistent organochlorine pollutants (POPs), including polychlorinated dioxins, polychlorinated dibenzofurans, and polychlorinated biphenyls (PCBs), present in food and the environment is one of widespread concerns throughout the industrialized world (Wigle et al. 2008 ; Qing Qing et al. 2005; Kodavanti 2005; Hites et al. 2004; Stellman et al. 1998). These organic compounds are major concerns for human and ecosystem health due to their high degree of bioaccumulation and persistence and long life. Although the use of most POPs is banned in most countries, some organochlorine pesticides are still being used in certain parts of the world, which could lead to the environmental pollution of the neighboring regions or even the entire world. The presence of these toxic substances in the Great Lakes continues to be a significant concern because POPs and other pollutants were routinely disposed of into lakes by heavy industries and agricultural activities until the 1970s. Despite decades of cleanup efforts, the concentration of POPs in some local fish populations remains elevated and may pose a continuing threat to people who heavily rely on fish in their diet. In many parts of the Great Lakes basin, there is little information available on exposure levels, body burdens, and health effects of POPs (Kodavanti 2005; Abelsohn et al. 2002; Rice 1995). These considerations are especially relevant to those populations whose traditional diet of fish results in higher toxicant body burden. Epidemiological studies have shown that young children of mothers consuming large amounts of contaminated fish from the Great Lakes exhibit decreased growth rate and neurological disturbances such as deficient intellectual capacity, altered attention, and memory deficits (Debes et al. 2006; IIwaniuk et al. 2006; Jacobson and Jacobson 1997; Evangelista de Duffard and Duffard 1996). Animal studies have also shown that these contaminants affect brain development, neurobehavior, and cognitive functions; however, many of these studies have employed high doses of chemicals that greatly exceeded the levels of human exposure (Kodavanti 2005; Newland and Rassmussen 2000; Schantz and Widholm 2001; Gilbert and Grant-Webster 1995). In addition, most of the animal studies conducted used single chemicals, simple mixtures, or commercially available mixtures (e.g., Aroclors), and it is not clear whether these produce toxicological effects that adequately estimate the toxic effects among human populations exposed to complex mixtures of chemicals (Kodavanti 2005; Moser et al. 2005; Carpenter et al. 1998). In addition, the potential additive, synergistic, or antagonistic effects of contaminant mixtures have not been adequately investigated (Cory-Slechta 2005; Carpenter et al. 2002). It is unclear whether the prenatal exposure to these mixtures has been linked to reduced fetal and postnatal growth, neurological deficits, delayed development of motor functions, and impaired short-term memory (Kodavanti 2005; Seegal 2000; Seegal, Pappas, and Park 1998). The developmental processes, including neurogenesis, migration, synaptogenesis, gliogenesis, and myelination, are extremely susceptible to damage from exposure to environmental insult (Garman et al. 2001; Tilson 2000; Rodier, Reynolds, and Roberts 1979; Rodier 1976). These sequences of events are genetically programmed and interference at any stage may alter subsequent development and result in permanent impairment. In adults, they have generally reached a steady state and are not potential targets for neurotoxic agents (Garman et al. 2001; Tilson 2000; Campbell et al. 1997; Rodier, Reynolds, and Roberts 1979; Rodier 1976). Extensive studies using lead, mercury, methylazoxymethanol, alcohol, methylmercury, nicotine, and chlorpyrifos have shown that these chemicals affect multiple pathways during normal development (Erickson and Talts 2000; Campbell et al. 1997; Chanda and Pope 1996).

Previously reported results from this study have shown that developmental exposure to the mixture resulted in maternal weight loss, produced significant and persistent reductions in offspring weight, caused high mortality rates at the highest mixture dose, and altered early neurobehavioral function (Bowers et al. 2004). In addition, systemic toxicity results showed that exposure to the POP mixture up to 13 mg/kg/day perinatally caused elevated serum cholesterol, increased liver microsomal enzyme activities, and immunopathological changes in the thymus and spleen, and lethality. Most of the effects were seen at dose levels much higher than the expected human exposure. The reconstituted POPs mixture produced greater toxicity than a comparable dose of the commercial PCB mixture Aroclor 1254.

In this article, we report the neurotoxicological neuropathology end points including quantification of cell proliferation and death in the subventricular zone, linear morphometric measurements, and transcriptional analysis of brain tissue. The brains were collected for neuropathologic assessment on postnatal day (PND) 37 and PND 77. For molecular analysis, brains were collected on PND 35, PND 75, and PND 350. These time periods correspond to juveniles, young, and mature adults, respectively. Necropsies were performed at the target age (±2 days) and were counterbalanced between treatment groups and gender. Because of the multidisciplinary nature of this study, tissues were collected for a number of analyses or end points, and no animals were available for neuropathologic assessment on PND 350.

Materials and Methods

Chemical Mixture

The POP mixture used in this study is shown in Table 1. This mixture was derived from measured contaminant blood levels in human fish eaters living in the Canadian Great Lakes basin (Health Canada 1998). The specific relative concentration for each chemical was generated from a weighted average of the lipid-adjusted PCBs and organochlorine concentrations found in tissues as reported in these studies. Compounds measured in human tissues that contributed 1.5% or more to the total mass of the measured chemicals were selected for inclusion in the final mixture. The contribution of each chemical to the total mass of all chemicals included was recalculated and this percentage contribution (by mass) defined the chemical mixture. Two PCB congeners (PCB 126 and PCB 169) did not satisfy the 1.5% inclusion criteria; however, because of their known high toxicity, they were added to the mixture at one-half of the detection limits reported in the human studies. Aldrin, p,p′-dichlorodiphenyldichloroethane (DDE), p,p′-dichlorodiphenyttrichloroethane (DDT), dieldrin, heptachlor epoxide (isomer B), hexachlorobenzene (HCB), β-hexachlorocyclohexane (HCH), mirex, cis-nonachlor, and trans-nonachlor were purchased from Sigma-Aldrich Canada Ltd. (Oakville, Ontario). The PCB congeners 74, 94, 99, 126, 156, 169, 170, 187, 194, and 201 (BZ 200/IUPAC 201) were purchased from AccuStandard Inc. (New Haven, CT). The PCB congeners 118, 138, 153, and 180 were purchased from Radian International (Austin, TX). Oxychlordane was generously donated by Julie Fillion of the Pest Management Regulatory Agency (Ottawa, Ontario). All of the chemicals were neat and the purities were at least 99%.

Table 1.

Persistent organochlorine pollutants in the Great Lakes mixture.

Component	% Total mass in mixture	Component	% Total mass in mixture
p, p’-DDE	50.43	PCBs (IUPAC #)
Hexachlorobenzene	1.89	PCB 74	2.08
cis-nonachlor	0.61	PCB 94	1.21
trans-nonachlor	2.18	PCB 99	2.03
p, p’-DDT	1.59	PCB 118	2.74
Mirex	1.00	PCB 126	0.01
Dieldrin	1.20	PCB 138	1.46
Aldrin	0.18	PCB 153	10.27
β-Hexachlorohexane	2.27	PCB 156	1.46
Heptachlor epoxide	0.83	PCB 169	0.002
Oxychlordane	1.07	PCB 170	2.67
		PCB 180	6.09
		PCB 187	3.23
		PCB 194	1.13
		PCB 201	2.38
Total		100%

Note. The relative proportions of the components in the Great Lakes Mixture were based upon the measured levels in fish consumers in the Great Lakes/St. Lawrence basin. The proportion of PCB 126 in the mixture is slightly higher than intended, possibly as a result of trace quantities of PCB 126 as an impurity in other more concentrated PCBs, or possibly as a measurement error since PCB 126 was added at very low concentrations. DDE = dichlorodiphenyldichloroethane; DDT = dichlorodiphenyttrichloroethane; PCBs = polychlorinated biphenyls.

The mixture was prepared by weighing the individual chemicals in clean vials and transferring them into small volumes of Mazola corn oil using preservative-free diethyl ether (Sigma-Aldrich Chemicals). Diethyl ether was then removed from corn oil solutions using a Savant Automatic Environmental Speedvac (Model AES 2000, Abbott Laboratories, Abbott Park, Illinois, USA) and confirmed by weighing the corn oil samples before adding ether and after processing in the Savant. The mixture was prepared at its final concentration (by mass) using clean corn oil. Because the mixture precipitated out of the liquid upon standing, 5% (wt/wt) diethyl ether was reintroduced into the PCBs/OC mixture as a cosolvent to ensure that all chemicals remained in the solution. The nominal concentration of this stock solution was 15 mg/ml and dosing solutions of 1.5, 0.15, and 0.015 mg/ml were prepared by serial dilution of the stock solution with corn oil containing 5% (wt/wt) of diethyl ether. The mixture composition was verified by an independent laboratory (Wellington Laboratories; Guelph, Ontario) using high-resolution gas chromatograpy–mass spectroscopy (GC-MS) analysis.

A separate group of animals was exposed to Aroclor 1254, a known neurotoxic agent, as a positive control treatment and the toxicological impact of this commercial and frequently used PCB mixture was compared against the complex mixture based on human blood contaminant profiles. Aroclor 1254 (AccuStandard Inc., lot number 124191) was prepared at a concentration of 15 mg/ml in corn oil with 5% diethyl ether. The control dosing solution consisted of corn oil containing 5% (wt/wt) of diethyl ether.

Animal Treatment

All animals, treatments, and dosing methods have been reported in detail previously (Bowers et al. 2004). Briefly, all treatment procedures and housing conditions conformed to the Canadian Council on Animal Care guidelines and had been approved by Health Canada’s Institutional Animal Care Committee. Seventy-eight nulliparous female Sprague-Dawley rats (200–230 g) and 38 male Sprague-Dawley rats (315–350 g) from Charles River, St Constant, Quebec, served as subjects. Females were housed two/cage and males were housed individually in polycarbonate hanging cages (35 cm × 30 cm × 16.5 cm) containing shaved wood bedding and maintained in rooms at 22°C ± 2°C and 50% ± 10% humidity. Animals were placed on a reverse dark–light cycle starting 24 hr after arrival in the facility by advancing the dark cycle by 1 hr/day for 12 days for a final light cycle of lights on: 20:00, lights off: 08:00. Breeding and dosing were conducted in two cohorts of litters of roughly equivalent size and representing all dose groups.

Food and water were available ad libitum. Permanent identification chips (Model IMI-1000, Bio Medic Data Systems, Seaford, DE) were implanted under light isoflurane anesthesia in females 7 days after arrival. Breeding females were weighed daily starting 10 days before breeding until they were sacrificed. Following 3 weeks of habituation breeding was conducted by placing the females into the cage of a breeding male. Females were monitored daily (7:00 – 8:00 hr) for the presence of sperm or a vaginal plug. Once detected (denoted as gestation day, GD-0), the female was housed individually in a cage with shaved wood bedding.

Doses of the chemical mixture were based on a pilot study in which pregnant Sprague-Dawley rats were dosed orally with 0.15, 1.5, and 15 mg/kg/day from GD 10 to 17. These doses produced no effects on maternal weight gain, reproductive rates, litter size, pup growth, development, or mortality. Mixture doses used in the principal study were 0.015, 0.15, 1.5, and 15 mg/kg/day. Pregnant females were dosed by placing 1 ul/g body weight of the assigned dose on a Teddy Graham® cookie (approximately 2.0 g; Nabisco Ltd., Toronto, Ontario), placing the uncovered cookies in a fume hood overnight to allow the ether to evaporate and providing the dosed cookies to pregnant rats the next day. All dams had been pre-exposed to cookies dosed with 250 ul of clean corn oil for 5 days prior to breeding, and all females readily consumed the dosed cookies. Dams were dosed daily from GD 1 to weaning on PND 23. Parturition was monitored starting on GD 18 at 07:00, 15:00, and 23:00 hr. Day of birth was denoted as PND 0. For each litter, pups were counted on PND 0 but not weighed or handled. Pups were sexed on PND 1 and confirmed on PND 2, 3, and 4. Total litter weight for each litter was obtained on PND 1, 2, 3, and 4. Mortality and morbidity were monitored daily and dead pups removed immediately. Litters were culled to 8 pups on PND 4 by randomly selecting 4 males and 4 females (where possible) from each litter to remain and the excess pups sacrificed. Pups were fostered between litters from the same dose group where fostering was needed, but fostered pups were excluded from all analyses. Pups were individually identified on PND 4 with footpad injections (Ketchum Animal Tattoo Ink, Ketcham Manufacturing, Ottawa Ontario) and weighed individually each day from PND 4 until PND 35. Any offspring remaining after PND 35 were weighed 3 times weekly. Permanent identification chips (Model IMI-1000, Bio Medic Data Systems) were implanted in pups on PND 27–28 under light isoflurane anesthesia.

The schedule and sample size for tissues collection are shown in Table 2. One male and one female per litter were sacrificed by decapitation on PNDs 35, 75, and 350 for collection of tissues for brain neurochemistry and biomarkers analysis. Brains were divided into two hemispheres with one hemisphere assigned for neurochemical analysis (reported separately) and the other hemisphere assigned for brain biomarker analysis. The tissue was immediately frozen on crushed dry ice and stored at −80°C until analysis. The assignment of brain hemisphere for neurochemistry or biomarker analysis was counterbalanced between dose groups and sexes to address potential lateralization effects.

Table 2.

Schedule and sample size for tissues collection.

Dosing	Tissue collections (PND)
GD1–PND 23	PND 35		PND 37	PND 75		PND 77	PND 350
Purpose	Brain biomarkers	Residue	Neuropathology	Brain Biomarkers	Neurochemistry	Neuropathology	Brain biomarkers	Neurochemistry
Sample/litter	1 M and 1 F	1 M and 1 F	1 M or 1 F	1 M and 1 F	1 M and 1 F	1 M or 1 F	1 M and 1 F	1 M and 1 F

Note. Text in boldface indicate samples relevant for the current report. Note that 1 male and 1 female offspring were collected for all post-mortem analyses except that for neuropathology analysis where tissues were collected from either 1 male or 1 female per litter. PND = postnatal day.

The remaining cohort of male and female offspring was assigned for neuropathology with half the animals per litter (1 male or 1 female per litter) sacrificed by decapitation under isoflurane anesthesia on PND 37 and the remainder sacrificed under isoflurane anesthesia on PND 77. Perfusion fixation was not used since multiple tissues from these animals were also collected for other analyses. For logistical reasons, animals assigned for neuropathology analysis on PNDs 37 and 77 were sacrificed 2 days after the animals were sacrificed for brain biomarker analysis. The skull was opened and the entire head submersed in 10% formalin until later processing in 4% buffered paraformaldehyde and postfixed in 10% formaldehyde. Control and treated animals were handled under the same conditions and tissue sampling was processed at the same time. The brains were removed from the skull and sliced into coronal sections using the appropriate rodent matrix suitable to the age of the rat (RBM Activational Systems Inc., Electron Microscopy Sciences, Hatfield, PA). The tissues were processed for paraffin embedding and sectioned on a microtome at 6 µm (Truelove et al. 1996).

For histology examination and morphometric measurements, brain coronal sections were prepared and referenced according to the rat brain atlas of Paxinos and Watson (1986). Several anatomical plates were used to ensure the inclusion of specific structures including subventricular zone (SVZ, plates 16–17); hippocampus/median eminence/arcuate nucleus (plates 30–32); hippocampus, substantia nigra, pineal recess (plates 38–40); facial nucleus/dorsal cochlear nucleus/locus coeruleus/cerebellum (plates 58–60); and area postrema (plates 72–74). The atlas of Amaral and Witter (1995) was used as a reference for the anatomy of the hippocampus.

The histological evaluation was conducted using Mayer’s hematoxylin and eosin Y (H&E; Luna 1968) and specific histological stains to evaluate morphological structures. Myelin and nerve cells were stained with Kluver–Barrera/luxol fast blue (LFB) with cresyl violet as the counterstain; Nissl substance were stained with Einarson gallocyanin; nucleic acids were stained with Feulgen; neuronal processes were stained with silver stain (Chen–Bodian); and microglia cells were stained with phosphotungstic acid hematoxylin ([PTAH] Luna 1968). The stained slides were evaluated using a light microscope.

Detailed evaluation of the brain morphology included routine histology (H&E), histomorphometric analysis including linear measurements, and quantification of pyknosis and mitosis in the SVZ of the lateral ventricles. For the current study, a detailed evaluation of these regions was performed using histomorphometric analysis, which included (1) linear measurements and (2) cell quantification of pyknosis and mitosis in the SVZ of the lateral ventricles.

Linear Measurements

For each brain, two highly homologous and symmetrically, vertically cut sections at the level of the anterior part of anterior commissure (plate 30–32; Paxinos and Watson 1986) were stained with H&E and subsequently used for quantitative linear measurements. All measurements were done using a Zeiss microscope (Ob-Objective x1, Oc-Occular x10). Calibration was performed using a calibrated ocular scale (1 mm in 100 parts, Zeiss, Zeiss Axioskop 2 with Axiocam Camera, Zeiss Canada Ltd, Toronto, CA) and the object micrometer slide (1 mm in 100 parts, Zeiss). All measurements (except measuring the Ammon’s horn height) were performed as shown in Figure 1 (Rodier and Gramann 1979; Duffell, Soames, and Gunby 2000). The following were measured: a = the vertical line through the middle of dorsoventral diencephalon (thalamus and hypothalamus) from the dorsal III ventricle to the median eminence; b = lateromedial width of the brain hemisphere, perpendicular to (a), just below the anterior hippocampus; c1 = cortical thickness in the hind limb area of the cortex; c2 = cortical thickness in the parietal 1 area of the cortex; d1 = length of the dorsal arm of the dentate gyrus; d2 = length of the ventral arm of the dentate gyrus; e1 = lateromedial length of the Ammon’s horn from the area CA 2 to the dorsal III ventricle; e2 = height of the Ammon’s horn, perpendicular to (e1) from alveus through areas CA1 and CA3 just laterally to the end of the ventral arm of the dentate gyrus. All measurements, except (a), were taken in both the right and left hemispheres.

Figure 1.

Linear measurements were done at the level of the anterior part of anterior commissure. Abbreviations for measurements are a = vertical line through the middle of dorsoventral diencephalon from the dorsal III ventricle to the median eminence; b = lateromedial width of the brain hemisphere; c1 = cortical thickness in the hind limb area of the cortex; c2 = cortical thickness in the parietal 1 area of the cortex; d1 = length of the dorsal arm of the dentate gyrus; d2 = length of the ventral arm of the dendate gyrus; e1 = lateromedial length of the Ammon’s horn from the area CA 2 to the dorsal III ventricle; and e2 = height of the Ammon’s horn.

Counting of the Number of Mitoses and Pyknoses in the SVZ of the Forebrain Lateral Ventricles

For each brain, 2 H&E stained sections at the level of the anterior part of anterior commissure (plates 16–17; Paxinos and Watson 1986) were selected for quantification of pyknotic and mitotic cells in the SVZ (Figure 2). Cells were counted under the Zeiss microscope (Model number), Ob ×40, Oc ×12.5. Counting started from the upper lateral corner of the lateral ventricle and was performed by shifting the field of vision around the ventricle in clockwise direction eventually encircling it. The thickness of the SVZ cell layer spread along the ventricle never surpassed the chosen diameter of the field of vision of the microscope. Densely packed, small, moderate to dark stained subventricular (subependymal) germinal cells located along the lateral wall of both lateral ventricles, with predominant concentration in the upper lateral corner of each lateral ventricle, are distinguishable from sparsely distributed glial cells in the corpus callosum and caudate putamen.

Figure 2.

Cell quantification of pyknosis and mitosis ratio of the subventricular zone (SVZ) were done in hemotoxylin-eosin stained sections at the level of the anterior part of anterior commissure (Paxinos and Watson 1986). Densely packed, small, moderate to dark stained subventricular (subependymal) germinal cells located along the lateral wall of both lateral ventricles, with predominant concentration in the upper lateral corner of each lateral ventricle, are distinguishable from sparsely distributed glial cells in the corpus callosum and caudate putamen. Mitotic and pyknotic cells were counted. Small, round, densely stained pyknotic nuclei were easily discernable from mitotic cells within subventricular cells. Cells were counted under a Zeiss microscope, Ob _40, Oc _12.5. (Dr. R. Mueller).

This SVZ was chosen since it exhibits the largest population of proliferating cells in the brain of rodents, monkeys, and humans. Mitoses (metaphases and anaphases) and pyknoses were counted per a section of the lateral ventricle in left and right hemispheres, respectively. Small, round, extremely densely stained pyknotic nuclei (or their fragments) are easily discernable from subventricular cells. Apoptotic cells were characterized by small, round, densely stained pyknotic nuclei (or their fragments) and were easily distinguishable from subventricular cells.

Transcriptional Analysis by Semiquantitative Polymerase Chain Reaction

The transcriptional analysis was conducted on brain tissue collected from animals on PND 35, PND 75, and PND 350. A systematic analysis of the genes related to specific cell types of the brain and their relation to a range of relevant pathways was performed. These include (1) neurotransmitter systems (cholinergic muscarinic and nicotinic receptors) and gamma-aminobutyric acid (GABA) receptors; (2) cell cycle checkpoint genes including p53, p16, p21, cyclin E, and D; (3) early injury markers (EIMs) including c-fos, c-myc, c-jun, and the heat shock protein (Hsp-72), and oxidative stress markers (cyclooxygenase 2 [COX-2], also known as prostaglandin endoperoxide synthase), superoxide dismutase ([SOD] Zn- and Mn-dependent SOD), and catalase. We also used markers of specific cell types and markers for structural genes. These included markers for (1) astrocytes (glial fibrillary acidic protein [GFAP]) and S100; (2) neurons (neuron-specific marker—enolase—NSE); (3) microglia (F4/80 and leukocyte common antigen [LCA]; (4) oligodendrocytes (myelin basic protein [MBP]), myelin-associated protein (MAG), myelin proteolipid protein (PLP), and 2′,3′-cyclic nucleotide 3′-phosphodiesterase (CNP); and (5) structural genes (microtubule-associated protein [MAP-2], tubulin-1α, and neurofilaments [NFLs]).

Total RNA was extracted from the frozen brain hemispheres using Trizol reagent (Gibco/BRL, Burlington, Ontario), according to the manufacturer's protocol. Total RNA (5 animals per dose/sex/brain hemisphere) were pooled and the pooled samples analyzed. For RT (reverse-transcriptase) polymerase chain reaction (PCR) analysis, 2 mg of total RNA used for first-strand complementary DNA (cDNA) synthesis using the Superscript II reverse transcriptase (Gibco/BRL, Burlington, Ontario) and with the oligo(dT) primer. The cDNAs were then amplified by PCR using gene-specific primers and Taq enzyme (Promega, Madison, WI). The reactions were performed in triplicates. The PCR conditions included denaturation at 94°C for 1 min, annealing at 55 to 60°C (depending on the primer pair used) for 1 min, and extension at 68°C for 1 min. The cDNA fragments were amplified for 25 to 35 cycles followed by an extension at 68°C for 10 min. All the cycles chosen were in the range of linear relationship between the PCR products and the number of cycles (Marone et al. 2001). The cDNA products were separated on a 2% agarose gel (Gill et al. 2003; Marone et al. 2001).

The 18S (3:7 alt RNA; Applied Biosystems, Life Technologies Inc., Burlington, ON, CA) was used as the internal housekeeping gene for semiquantitative PCR analysis because its expression did not change with different treatments. All PCRs were performed in triplicates (Marone et al. 2001).

The amplified PCR-generated products were analyzed by electrophoresis on a 2% agarose gel using ethidium bromide staining. To evaluate the changes in relative levels of different genes, the band intensities for the genes obtained from each aliquot of PCR products were normalized against the band intensities of the housekeeping gene 18S (3:7 alt rRNA) in the same experimental run (Alpha Innotech Incorporation, San Leandro, CA). Values were expressed as arbitrary units of relative abundance of the specific target genes. A 2-fold change of the integrated density value was taken as a significant change in gene expression (Gill et al. 2003; Marone et al. 2001).

Statistical analysis

Morphometric measurements were analyzed by one-way analysis of variance (ANOVA) followed by Turkey’s multiple comparison test. Linear trend analysis was also performed (GraphPad, Prism 4.0).

Results

Because dosing cookies were prepared based on body weight 24 hr before delivery to the animals so that the required ether could evaporate, the actual doses delivered was calculated based on body weight on the day the dosing cookies were delivered. These calculated doses based on this were 0.013, 0.13, 1.3, and 13.9 mg/kg for the mixture-treated animals, while the dose for Aroclor-treated animals was 14.9 mg/kg. As previously reported (Bowers et al. 2004), the highest mixture dose had been tested in a pilot study using a shorter exposure period and revealed no signs on maternal and offspring toxicity. However, in the current study using a longer dosing period (GD1 to PND 21), there was clear maternal and offspring toxicity at the highest dose (13.9 mg/kg) of the mixture, including maternal weight loss and high offspring mortality (66% mortality after PND4). There were few offspring available from this dose group for testing, and the maternal toxicity in this dose group would preclude any clear attribution of effects to direct toxicity on the offspring. Accordingly, all data from the 13.9 mg/kg group was excluded from analysis.

Histology Staining

For other dose groups of POPs and Aroclor 1254, general histological screening (H&E) and specific stains did not reveal any observable effects in any of the brain regions examined. Linear measurements were made for both the left and the right hemispheres. Since there are no significant differences between the hemispheres, brain hemisphere was not used as a factor in further analysis. For the PND 37 pups, the mixture at 1.3 mg/kg/day and the Aroclor 1254 groups showed a significant decrease (p < .01) in the length of the dorsal and the ventral arms of the dentate gyrus compared to the control animals. In addition, Aroclor 1254 significantly decreased (p < .01) the lateromedial width of the brain hemisphere relative to the control animals. There were no changes in the lateromedial length or height of Ammon’s horn (Table 3). The mixture doses of 0.13 and 0.013 mg/kg had no impact on any measure. On PND 77, the Aroclor 1254-treated animals showed a significant decrease (p < .01) in the length of the dorsal and the ventral arms of the dentate gyrus compared to control animals (Table 3). Because we have previously reported that the body weights of these animals were affected by treatments (Bowers et al. 2004) as were organ weights, including brain weights (Chu et al. 2005), we also normalized the morphometric measures using the lateromedial width of the brain hemisphere (measure b in Figure 1) as a reference point and expressed all other measures relative to this measure. If the treatment-related effects on morphometric measures are not attributable to indirect effects on brain size, then treatment-related effects should be evident after this normalization. Measurements of a,, c1, c2, d1, d2, e1, and e2 were divided by measurement b, and these new values were analyzed by ANOVA. Analysis on these normalized morphometric measures indicated that there no significant changes in any dose group as compared to the control groups.

Table 3.

Linear measurements at the level of the anterior part of anterior commissure on PND 37 pups.

	a	b	c1	c2	d1	d2	e1	e2
Vehicle	81.2 ± 2.52	112.5 ± 2.66	24.20 ± 1.436	30.12 ± 1.576	32.00 ± 1.59	24.90 ± 2.27	65.85 ± 3.10	23.95 ± 1.0
Aroclor	79.83 ± 3.20	107.8 ± 4.07*	23.75 ± 1.52	29.45 ± 1.57	28.88 ± 3.222*	21.19 ± 2.37*	64.31 ± 5.17	23.00 ± 15
POPs (1.3mg/ml)	79.56 ± 3.4	111.3 ± 3.41	23.80 ± 0.90	30.77 ± 1.423	29.35 ± 3.297*	22.05 ± 4.14*	66.50 ± 4.12	23.85 ± 1.20

Note. a = vertical line through the middle of dorso-ventral diencephalon(thalamus and hypothalamus) from the dorsal III ventricle to the median eminence; b = lateromedial width of the brain hemisphere perpendicularly to (a) just below the anterior hippocampus; c1 = cortical thickness in the hind limb area of the cortex; c2 = cortical thickness in the parietal 1 area of the cortex; d1 = length of the dorsal arm of the dentate gyrus; d2 = length of the ventral arm of the dentate gyrus; e1 = lateromedial length of the Ammon’s horn from the area CA2 to the dorsal III ventricle; e2 = height of the Ammon’s horn perpendicular to e1 from alveus through areas CA1 and CA3 just laterally to the end of the ventral arm of the dentate gyrus; PND = postnatal day; ANOVA = analysis of variance.

*Data are significantly different from corresponding controls using one-way ANOVA. Bold represents p < 0.05.

In the PND 37 group, there were no POPs or Aroclor 1254 treatment-related effects in the number of mitotic cells. The number of pyknotic cells was significantly (p < .05) changed in the Aroclor 1254 group but not in the POPs-treated animals as compared to the control groups.

There were no significant changes in the mitotic or pyknotic cells on PND 77 in any treatment group as compared to the control groups. The number of mitotic cells in the SVZ of PND 37 rats significantly exceeded that in PND 77 rats. The number of pyknotic cells in the SVZ on PND 37 was higher than that in PND 77 rats.

Transcriptional Analysis by Semiquantitative PCR

A systematic examination of genes or biomarkers related to a range of relevant pathways for all the dose groups including vehicle, 0.013, 0.13, and 1.3 mg/kg/day was performed. While there were no significant changes in gene expression on PND 35 and PND 75 (data not shown), significant changes were observed on PND 350 (Table 4). These changes were observed only at the 0.13 mg/kg/day dose, and these changes were only observed in female offspring. Cell cycle checkpoint genes p53, p16, p21, and cyclin E or D were not affected. Although immediate EIMs, c-fos, and c-jun were not affected by any treatment Hsp-72 was downregulated more than 4-fold. Oligodentrocyte-specific markers MBP, MAG, and PLP were not affected but CNP was downregulated 3.7-fold. Markers for the structural genes, which included MAP-2 and NFLs, were downregulated 10-fold and 1.6-fold, respectively (Table 4 and Figure 3), while tubulin-1α was not affected.

Figure 3.

Semi quantitative RT-PCR analysis of the some of genes of interest for different doses of POPs (see also Table 2), using varied cycle numbers, the exponential phase for each primer pair was determined and used to allow semi quantitative analysis of the respective reactions. 18S was used as an internal control. NFLs - neurofilaments; MAP2 = microtubule-associated protein; CNP = 2,3-cyclic nucleotide- 3′- hosphohydrolase; MBP = myelin basic protein; 18S = internal control; PCR =polymerase chain reaction; POPs = persistent organochlorine pollutants. The numbers 1, 2, 3, 4 and 5 represent the dose groups of the animals. 1 = control group; 2 = 0.013 (mg/kg/day); 3 = 0.13 (mg/kg/day); 4 = 1.3 (mg/kg/day); and 5 = Aroclor.

Table 4.

Transcriptional analysis by semiquantitative PCR of female rats (PND 350) treated with POPs.

Genes	0.013 (mg/kg/day)	0.13 (mg/kg/day)	1.3 (mg/kg/day)	Aroclor
Oligodendrocytes
CNP	2.3	–3.7	No change	2.0
Neuron cell
NSE	No change	–2.0	No change	No change
Structural genes
NFL	No change	–1.6	No change	No change
MAP 2	No change	–10.0	No change	No change
Astrocyte marker
GFAP	No change	–2.4	No change	No change
Microglia cells
F4/80	No change	–2	No change	No change
LCA	No change	–2.2	No change	No change
Cholinergic system markers
M1	No change	–4	No change	No change
M4	No change	–3.5	No change	No change
M5	No change	–4	No change	No change
Ache	No change	–6.0	No change	No change
COX-2	No change	–2.1	No change	No change
Early injury markers
HSP 72	No change	–4	No change	No change
c-Fos	No change	–3.5	No change	No change

Note. There were no lateral difference between the left and right hemispheres. Transcriptional analysis was done for the cell cycle checkpoints (p53, p16, p21, and cyclin E and D); early injury markers (c-fos, c-myc, c-jun; heat shock protein [HSP-72]); oxidative stress markers (cyclooxygenase [COX-2], also known as prostaglandin endoperoxide synthase, superoxide dismutase [Zn- and Mn dependent-SOD] and catalase); structural genes (microtubule-associated protein [MAP-2], tubulin-1, and neurofilaments [NFL]). Markers for specific cell types included (1) astrocytes: (glial fibrillary acidic protein [GFAP] and S100); (2) neurons: (neuron specific marker—enolase—NSE) (3) microglia: (F4/80 and leukocyte common antigen [LCA]); (4) oligodendrocytes: (myelin basic protein [MBP], myelin-associated protein [MAG], and myelin proteolipid protein [PLP]) and 2′,3′-cyclic nucleotide 3′-phosphodiesterase (CNP). Markers for the cholinergic system included M1 (muscarinic acetylcholine receptor subunit 1), M4 (muscarinic acetylcholine receptor subunit 4), M5 (muscarinic acetylcholine receptor subunit 5), and Ache (acetylcholine esterase); PCR = polymerase chain reaction; POPs = persistent organochlorine pollutants.

Markers for the GABA and cholinergic neurotransmitters systems were also examined. For the 14 subunits of the GABA_A receptors, there were no changes in gene expression at any age, gender, or dose. However, the cholinergic markers of muscarinic receptor subtypes (M1, M4, and M5) were all downregulated approximately 4-fold and acetycholin esterase (Ache) was downregulated 6-fold.

The oxidation marker COX-2 (PTGS) showed a 2-fold decrease, while SOD (Zn and Mn dependent) and catalase were not affected (Table 4).

Discussion and Conclusions

The current study examined the neurodevelopmental impact of perinatal exposure to a mixture of PCBs and organochlorine pesticides based on human blood levels on brain morphology and transcriptional changes in young, juvenile, and mature animals. Potential damage to the central nervous system (CNS) is difficult to assess because of wide variations in its normal function as well as its plasticity, residual capacity, and compensatory mechanisms. Furthermore, different brain regions have their own specific time window for the proliferation, migration, differentiation, and scheduled apoptosis of cells. During the specific time window, some regions may be more adversely affected than others, often resulting in changes in cell numbers or cell structures. Hence neuropathologic assessment of chemically induced developmental alterations in the nervous system for regulatory purposes is a multifactorial, complex process. In this study, we performed qualitative and quantitative morphologic analysis of numerous brain regions including general histological evaluation (H&Es), a battery of histological special stains, linear and stereologic measurements to evaluate morphological structures. In our study, H&Es and specific stains for structural damage did not reveal any injury either with the POPs or Aroclor groups. Although absolute morphometric measurements were affected by treatments, analysis of normalized results to adjust for treatment effects on body weight indicated that changes in morphometric measures observed on PND 37 could be attributed to treatment-induced changes in brain weights that are associated with altered body weights. Based on these results, there was no indication that any treatment directly altered brain morphometry.

It has been shown that linear measurements of adult mice exposed in the fetal period to the effect of antimitotic agents showed reduction in specific brain structures (Rodier et al. 1979). These tests were performed in the hippocampus and SVZ since they are the sites of neurogenesis where developmental insults of POPs may alter the growth of specific regions of the brain in the absence of overt lesions It is known that the progenitor cells within the SVZs are vulnerable to chemical, radiation, and ischemia-induced damage. With moderate insults, the SVZ can recover, but it cannot recover after more severe injury. Thus, the vulnerability or insult to these cells has important ramifications when considering therapeutic interventions for the treatment of brain tumors and for the prospect of recovery after ischemia. The cells of the perinatal and adult SVZ not only have the capacity to replenish their own numbers, but they also have the capacity to replace neurons and glia after ischemic and traumatic brain injuries. However, certain insults such as perinatal hypoxia/ischemia can deplete the progenitor cells in the SVZ (Romanko et al. 2007), leading to permanent damage.

There was a decrease in the number of mitotic and pyknotic cells in the SVZ on PND 77 as compared to PND 37. This decrease was expected between young and more mature animals. There was no indication that any mixture dose affected the mitotic index on PND 37 or 77 compared to control animals. This absence of difference between the experimental and control group could partly be due to the time between exposure and the age when the measurements were performed, which would suggest a transient effect. Interestingly, Aroclor-treated animals did show an increase in the number of pyknotic cells on PND 37, suggesting that the lack of effects of the mixture are not likely to be related to transient effects of exposure. These results also add to the evidence that the effects of Aroclor may be substantially different than the effects of mixtures that more completely mimic human exposure profiles on PND 37.

Age-related molecular and cellular alterations in the CNS are known to show selectivity for certain cell types and brain regions (Rutten et al. 2007). We chose general markers for transcriptional analysis, since the effects of the POPs were not known and the proper neurological function requires a coordinated and functional relationship between all the different cell types. For normal development, the acetylcholine receptors (muscarinic) are important for learning and cognition. Further, the cholinergic neurons make contact with noncholinergic structures, suggesting that they are needed for modulating the activity, acting on other types of neurons, and modulating the release of norepinephrine, dopamine, GABA, serotonin, and acetylcholine (Stein et al. 2002). To assess this, we conducted transcriptional analysis for a variety of cell types at each time period. Although there were no transcriptional changes on PND 35 and PND 75, significant transcriptional changes were observed in females exposed to 0.013 mg/kg of the mixture on PND 350. While Aroclor 1254 had no effects at any age, gender-specific transcriptional changes in exposure to POPs have also been reported by others (Powers et al. 2005; Adeeko et al. 2004); and these could be explained by differences in gender-specific metabolism of POPs. Most of the biomarkers including the EIMs, cell-specific markers, structural genes, oxidation markers, and markers of the cholinergic system were affected only at the 0.13 mg/kg/day group.

In addition to being gender specific, there was also a nonlinear dose-response for the brain biomarkers. Although there was little effect at doses of 0.013 and 1.3 mg/kg of the mixture, 0.13 mg/kg produced a dramatic effect. Though the reason for this U-shaped dose-response is not obviously clear, it is likely that multiple mechanisms of action, the presence of homeostatic mechanisms, and/or activation of compensatory or protective mechanisms among the different cell types are responsible (Davis and Svendsgaard 1989) as this has been observed in other systems (Marotta et al. 2002, 2003).

The transcriptional changes suggest that exposure to the mixtures during the rapid development of the neonatal rat brain (the brain growth spurt [BGS]) leads to irreversible changes in the function of the adult brain. Since there was no systemic or other effects on the dams at the 0.13 mg/kg dose (Chu et al. 2005), the transcriptional changes observed can be attributed to the perinatal exposure to the mixture. On PND 35 and PND 75, the insults that occurred during the BSG of the lactational period may be masked due to the reserve capacity and plasticity of the brain (Forcelli, Gale, and Kondratyev 2011; Slikker and Boyer 2005; Erickson and Talts 2000) only to be depleted over time by a number of factors, such as aging. Others have also shown that early exposures to environmental chemicals are reliably associated with late neurotoxicities in children (Bondy and Campbell 2005; Erickson and Talts 2000; Weiss 2000). A long latent or silent period is assumed to precede the expression of clinical signs, and this observation has been documented in both humans and laboratory animals (Bondy and Campbell 2005;Weiss 2000). The mechanisms of compensation may also make their own contribution to the process by generating neurotoxic by-products. In addition, concomitant aging of neuroendocrine and immune system function may amplify these CNS consequences. Many brain areas show a decline in nerve cell populations associated with aging, and additional exposure to environmental or food contaminants will exacerbate these effects. At that point, the aftermath of early damage may emerge in the form of impaired function detectable by transcriptional changes or even as overt neurodegenerative disease.

In summary, in utero and lactational exposure to the chemical mixture of POPs suggest developmental changes in female adult rats as observed by transcriptional changes.

Footnotes

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by a grant from the Toxic Substances Research Initiative from Health Canada and Environment Canada (TSRI # 209).

Abbreviations

Acknowledgments

The authors wish to thank Yangxun Hou and Wendy Prichett for their technical assistance and Bruce Martin for his excellent technical support.

References

Abelsohn

Brian

Gibson

B. L.

Margaret

Sanborn

M. D.

Weir

(2002). Identifying and managing adverse environmental health effects: 5. Persistent organic pollutants. CMAJ 166, 1549–54.

Amaral

D. G.

Winter

M. P.

(1995). Hippocampal formation. In The Rat Nervous System ( Paxinos

, ed.), pp. 443–93. Academic Publishers, San Diego, CA.

Bondy

S. C.

Campbell

(2005). Developmental neurotoxicology. J. Neurosci Res 81, 605–612.

Bowers

W. J.

Nakai

J. S.

Chu

Wade

M. G.

Moir

Yagminas

Gill

Pulido

Mueller

(2004). Early developmental neurotoxicity of a PCB/organochlorine mixture in rodents after gestational and lactational exposure. Toxicol Sci 771, 51–62.

Campbell

Seidler

Slotkin

(1997). Chlorpyrifos interferes with cell development in rat brain regions. Brain Research Bulletin 43, 179–89.

Carpenter

D. O.

Arcaro

K. F.

Bush

Niemi

W. D.

Pang

Vakharia

D. D.

(1998). Human health and chemical mixtures: An overview. Environ Health Perspect 106, 1263–70.

Carpenter

D. O.

Arcaro

Spink

D. C.

(2002). Understanding the human health effects of chemical mixtures. Environ Health Perspect 110, 25–42.

Chanda

Pope

(1996). Neurochemical and neurobehavioral effects of repeated gestational exposure to chlorpyrifos in maternal and developing rats. Pharmacol Biochem Behav 53, 771–76.

Chu

Bowers

Caldwell

Nakai

Pulido

Yagminas

Wade

Moir

Gill

Mueller

(2005). Toxicological effects of gestational and lactational exposure to a mixture of persistent organochlorine in rats: Systemic effects. Toxicol Sci 88, 645–55.

10.

Cory-Slechta

D. A.

(2005). Studying toxicants as single chemicals: Does this strategy adequately identify neurotoxic risk? Neurotoxicology 26, 491–510.

11.

Davis

J. M.

Svendsgaard

D. J.

(1989). U-shaped dose-response curves: Their occurrence and implication for risk assessment. J Toxicol Environ Health 30, 71–83.

12.

Debes

Budtz-Jorgensen

Weihe

White

R. F.

Grandjean

(2006). Impact of prenatal methylmercury exposure on neurobehavioral function at age 14 years. Neurotoxicol Teratol 28, 536–47.

13.

Duffell

S. J.

Soames

A. R.

Gunby

(2000). Morphometric analysis of the developing rat brain. Toxicol Pathol 28, 157–63.

14.

Erickson

Talts

(2000). Neonatal exposure to neurotoxic pesticides increase adult susceptibility. NeuroToxicology 21, 37–48.

15.

Evangelista de Duffard

A. M.

Duffard

(1996). Behavioral toxicology, risk assessment, and chlorinated hydrocarbons. Environ Health Perspect 104, 353–60.

16.

Forcelli

P. A.

Gale

Kondratyev

(2011). Early postnatal exposure of rats to lamotrigine, but not phenytoin, reduces seizure threshold in adulthood. Epilepsia 52, 20–22.

17.

Garman

R. H.

Fix

A. S.

Jortner

B. S.

Jensen

K. F.

Hardisty

J. F.

Claudio

Ferenc

(2001). Methods to identify and characterize developmental neurotoxicity for human health risk assessment. II: Neuropathology. Environ Health Perspect 109, 93–100.

18.

Gilbert

S. G.

Grant-Webster

K. S.

(1995). Neurobehavioral effects of developmental methylmercury exposure. Environ Health Perspect 103, 135–42.

19.

Gill

Mueller

Murphy

Clausen

Richards

Quilliam

MacKinnon

LaBlanc

Pulido

(2003). Neural injury biomarkers of novel shellfish toxins, spirolides: A pilot study using immunochemical and transcriptional analysis. Neurotoxicology 24, 593–604.

20.

Health Canada. (1998). Persistent environmental contaminants and the Great Lakes Basin population: An exposure assessment. (Minister of Public Works and Government Services Canada, Cat No H46–2/98–218E).

21.

Hites

R. A.

Foran

J. A.

Carpenter

D. O.

Hamilton

M. C.

Knuth

B. A.

Schwager

S. J.

(2004). Global assessment of organic contaminants in farmed salmon. Science 303, 226–9.

22.

IIwaniuk

A. N.

Koperski

D. T.

Cheng

K. M.

Elliott

J. E.

Smith

L. K.

Wilson

L. K.

Wylie

D. R.

(2006). The effects of environmental exposure to DDT on the brain of a songbird: Changes in structures associated with mating and song. Behav Brain Res 173, 1–10.

23.

Jacobson

J. L.

Jacobson

S. W.

(1997). Evidence for PCBs as neurodevelopmental toxicants in humans. Neurotoxicology 18, 415–24.

24.

Kodavanti

P. R.

(2005). Neurotoxicity of persistent organic pollutants: Possible modes of action and further considerations. Dose-Response 3, 273–305.

25.

Luna

L. G.

(1968). Manual of Histological Staining Methods of the Armed Forces Institute of Pathology , 3rd ed. McGraw-Hill, New York, NY.

26.

Marone

Mozzetti

De Ritis

Pierlli

Scambia

(2001). Semi-quantitative RT-PCR analysis to assess the expression levels of multiple transcripts from the same samples. Biol Proced Online 3, 19–25.

27.

Marotta

Marini

Banaudha

Maharaj

S. V.

Ives

Morrissette

C. R.

Jonas

W. B.

(2002). Non-linear effects of cycloheximide in glutamate treated cultured rat cerebellar neurons. Neurotoxicology 23, 307–12.

28.

Marotta

Marini

Banaudha

Maharaj

S. V.

Jonas

W. B.

(2003). Nonlinear effects of glutamate and KCl on glutamate toxicity in cultured rat cerebellar neurons. Int J Neurosci.113, 491–502.

29.

Moser

V. C.

Casey

Hamm

Carter

W. H.

Simmons

J. E.

Gennings

(2005). Neurotoxicological and statistical analyses of a mixture of five organophosphorus pesticides using a ray design. Toxicol Sci 86, 101–15.

30.

Newland

Rassmussen

(2002). Aging unmasks adverse effects of gestational exposure to methylmercury in rats. Neurotoxicol Teratol 22, 819–28.

31.

Paxinos

Watson

(1986). The Rat Brain in Stereotaxic Coordinates , 2nd ed. Academic Press, Inc., San Diego, CA.

32.

Powers

B. E.

Lin

T. M.

Vanka

Peterson

R. E.

Juraska

J. M.

Schantz

S. L.

(2005). Tetrachlorodibenzo-p-dioxin exposure alters radial arm maze performance and hippocampal morphology in female AhR mice. Genes Brain Behav 4, 51–59.

33.

Qing

Q. L.

Loganath

Chong

Y. S.

Tan

Obbard

J. P.

(2005). Persistent organic pollutants and adverse health effect in humans. J Toxicol Environ Health 69, 1987–2005.

34.

Rice

D. C.

(1995). Neurotoxicity of lead, methylmercury, and PCBs in relation to the Great Lakes. Environ Health Perspect 103, 71–87.

35.

Rodier

P. M.

(1976). Critical periods for behavioral anomalies in mice. Environ Health Perspect 18, 79–83.

36.

Rodier

P. M.

Gramann

W. J.

(1979). Morphologic effects of interference with cell proliferation in the early fetal period. Neurobehavioral Toxicology 1, 129–35.

37.

Rodier

P. M.

Reynolds

S. S.

Roberts

W. N.

(1979). Behavioral consequences of interference with CNS development in the early fetal period. Teratology 19, 327–65.

38.

Romanko

M. J.

Zhu

Bahr

B. A.

Blomgren

Levison

S. W.

(2007). Death effector activation in the subventricular zone subsequent to perinatal hypoxia/ischemia. J Neurochem 103, 1121–31.

39.

Rutten

Schmitz

Gerlach

Oyen

de Mesquita

Steinbusch

Korr

(2007). The aging brain: Accumulation of DNA damage or neuron loss? Neurobiol Aging 28, 91–98.

40.

Schantz

S. L.

Widholm

J. J.

(2001). Cognitive effects of endocrine-disrupting chemicals in animals. Environ Health Perspect 109, 1197–206.

41.

Seegal

R. F.

(2000). The neurotoxicological consequences of developmental exposure to PCBs. Toxicol Sci 57, 1–3.

42.

Seegal

R. F.

Pappas

B. A.

Park

G. A.

(1998). Neurochemical effects of consumption of Great Lakes salmon by rats. Regul Toxicol Pharmacol 27, S68–75.

43.

Slikker

Bowyer

J. F.

(2005). Biomarkers of adult and developmental neurotoxicology. Toxicol Appl Pharmacol 206, 255–260.

44.

Stein

Schettler

Wakkinga

Valenti

(2002). In harms way: Toxic threats to child development. Dev Behav Pediatr 22, S13–22.

45.

Stellman

S. D.

Djordjevic

M. V.

Britton

J. A.

Muscat

J. E.

Citron

M. L.

Kemeny

Busch

Gong

(1998). Breast cancer risk in relation to adipose concentrations of organochlorine pesticides and polychlorinated biphenyls in Long Island, New York. Cancer Epidemiol Biomarkers Prev 9, 1241–9.

46.

Tilson

H. A.

(2000). Neurotoxicology risk assessment guidelines: Developmental neurotoxicology. Neurotoxicology 21, 189–94.

47.

Truelove

Mueller

Pulido

Iverson

(1996). Subchronic toxicity study of domoic acid in the rat. Food Chem Toxicol 34, 525–9.

48.

Weiss

(2000). Vulnerability to pesticide neurotoxicity is a lifetime issue. Neurotoxicology 21, 67–73.

49.

Wigle

Arbuckle

T. E.

Turner

M. C.

Bérubé

Yang

Liu

Krewski

(2008). Epidemiological evidence of relationhsip between reproductive and child health outcomes and environmental chemical contaminants. J Toxicol Environ Health, Part B 11, 373–517.