Toxicological assessments of rats exposed prenatally to inhaled vapors of gasoline and gasoline–ethanol blends

Neurotoxicology and Teratology 49 (2015) 19–30

Contents lists available at ScienceDirect

Neurotoxicology and Teratology journal homepage: www.elsevier.com/locate/neutera

Toxicological assessments of rats exposed prenatally to inhaled vapors of gasoline and gasoline–ethanol blends☆,☆☆ Philip J. Bushnell a,⁎, Tracey E. Beasley a, Paul A. Evansky b, Sheppard A. Martin a, Katherine L. McDaniel a, Virginia C. Moser a, Robert W. Luebke b, Joel Norwood Jr. a, Carey B. Copeland b, Tadeusz E. Kleindienst c, William A. Lonneman c, John M. Rogers a a

Toxicity Assessment Division, National Health and Environmental Effects Research Laboratory, United States Environmental Public Health Division, National Health and Environmental Effects Research Laboratory, United States c Human Exposure and Atmospheric Sciences Division, National Exposure Research Laboratory, Office of Research and Development, U.S. Environmental Protection Agency, Research Triangle Park, NC 27711, United States b

a r t i c l e

i n f o

Article history: Received 24 November 2014 Received in revised form 13 February 2015 Accepted 16 February 2015 Available online 24 February 2015 Keywords: Biofuels Development Behavior Immunotoxicity Blood pressure Glucose homeostasis

a b s t r a c t The primary alternative to petroleum-based fuels is ethanol, which may be blended with gasoline in the United States at concentrations up to 15% for most automobiles. Efforts to increase the amount of ethanol in gasoline have prompted concerns about the potential toxicity of inhaled ethanol vapors from these fuels. The well-known sensitivity of the developing nervous and immune systems to ingested ethanol and the lack of information about the neurodevelopmental toxicity of ethanol-blended fuels prompted the present work. Pregnant Long–Evans rats were exposed for 6.5 h/day on days 9–20 of gestation to clean air or vapors of gasoline containing no ethanol (E0) or gasoline blended with 15% ethanol (E15) or 85% ethanol (E85) at nominal concentrations of 3000, 6000, or 9000 ppm. Estimated maternal peak blood ethanol concentrations were less than 5 mg/dL for all exposures. No overt toxicity in the dams was observed, although pregnant dams exposed to 9000 ppm of E0 or E85 gained more weight per gram of food consumed during the 12 days of exposure than did controls. Fuel vapors did not affect litter size or weight, or postnatal weight gain in the offspring. Tests of motor activity and a functional observational battery (FOB) administered to the offspring between post-natal day (PND) 27–29 and PND 56–63 revealed an increase in vertical activity counts in the 3000- and 9000-ppm groups in the E85 experiment on PND 63 and a few small changes in sensorimotor responses in the FOB that were not monotonically related to exposure concentration in any experiment. Neither cell-mediated nor humoral immunity were affected in a concentration-related manner by exposure to any of the vapors in 6week-old male or female offspring. Systematic concentration-related differences in systolic blood pressure were not observed in rats tested at 3 and 6 months of age in any experiment. No systematic differences were observed in serum glucose or glycated hemoglobin A1c (a marker of long-term glucose homeostasis). These observations suggest a LOEL of 3000 ppm of E85 for vertical activity, LOELs of 9000 ppm of E0 and E85 for maternal food consumption, and NOELs of 9000 ppm for the other endpoints reported here. The ethanol content of the vapors did not consistently alter the pattern of behavioral, immunological, or physiological responses to the fuel vapors. The concentrations of the vapors used here exceed by 4–6 orders of magnitude typical exposure levels encountered by the public. Published by Elsevier Inc.

1. Introduction ☆ This manuscript has been reviewed by the National Health and Environmental Effects Research Laboratory, U.S. Environmental Protection Agency and approved for publication. Approval does not indicate 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 research were presented at the Society of Toxicology meetings in 2011, 2012 and 2013. ⁎ Corresponding author at: Toxicology Assessment Division, MD B105-04, National Health and Environmental Effects Research Laboratory, Office of Research and Development, U.S. Environmental Protection Agency, Research Triangle Park, NC 27711, United States. Tel.: +1 919 541 7747. E-mail address: [email protected] (P.J. Bushnell).

http://dx.doi.org/10.1016/j.ntt.2015.02.004 0892-0362/Published by Elsevier Inc.

Sources of renewable fuels and alternative energy have been topics of legislation during the past decade. The Energy Policy Act of 2005 (P.L. 109–58) established a Renewable Fuel Standard (RFS) program to incorporate renewable fuels in the American automotive fuel supply. Two years later, the Energy Independence and Security Act of 2007 (EISA, P.L. 110–140) was enacted to encourage energy independence and to limit climate change by reducing greenhouse gas emissions from automobiles through increased use of renewable fuels. EISA required the U.S. Environmental Protection Agency (EPA) to revise the

20

P.J. Bushnell et al. / Neurotoxicology and Teratology 49 (2015) 19–30

RFS to increase the volume of renewable components of transportation fuels from 9 billion gal in 2008 to 36 billion gal by 2022 (EPA-HQ OAR-2010-0133; FRL-9234-6, Federal Register, Vol: 75, No. 236, Dec. 2010). Increased use of alternative fuels and the large population likely to inhale their vapors argue for information about the public health impacts of evaporative emissions of these fuels. Ethanol is the only current commercially-viable alternative fuel for spark-ignition engines, and it constituted 95% of the biofuel produced in the U.S. in 2009 (EPA, 2011). Health consequences of ingested ethanol are well known. In addition to its acute effects in adults, consumption during pregnancy has been linked to a range of teratogenic effects known as fetal alcohol spectrum disorder (FASD), which includes craniofacial malformations, persistent neurodevelopmental and neurocognitive deficits and immune system dysfunction (Abel, 2006; Kane et al., 2012; Mukherjee et al., 2005). The magnitude and severity of the deficits have been linked to the peak blood ethanol concentration (BEC) in the mother, but are also influenced by the number of exposures and the developmental stage at which exposure occurs (Driscoll et al., 1990; Riley and McGee, 2005; Zajac and Abel, 1992). Significant changes in cognitive function have been reported in rats whose mothers were exposed to ethanol by the oral route during pregnancy at maternal BECs of about 80 mg/dL (Savage et al., 2002), indicating that serious effects of prenatal ethanol exposure can occur in the absence of morphological abnormalities. These well-documented effects of ethanol were observed after dosing by the oral route, rather than by inhalation, the relevant route for exposures to fuel vapors. For these reasons, previous studies from this laboratory assessed the neurodevelopmental, immunological and physiological impacts of maternal inhalation of ethanol in a rat model. These studies found no systematic, dose-related effects of prenatal ethanol in adult offspring on sensory systems (Boyes et al., 2014), unconditioned behaviors, immune function, blood pressure or clinical markers of liver and kidney functions (Beasley et al., 2014), even at maternal BECs approaching 200 mg/dL. Clear effects on cognitive function were limited to increased anticipatory responses in a choice reaction time test in rats whose mothers had been exposed during pregnancy to 21,000 ppm ethanol and an estimated BEC of 192 mg/dL although other changes that were not concentration-dependent were observed on cue learning and reference memory (Oshiro et al., 2014). Compared to ethanol, the effects of gasoline vapors are less thoroughly studied. Low reproductive and developmental toxicity of gasoline vapors has been reported, albeit at concentrations several orders of magnitude above those expected in garages or from fueling operations, which rarely exceed 100 ppb (Zielinska et al., 2007). For example, McKee et al. (2000) used a vapor recovery unit to collect a relatively volatile fraction of gasoline, similar to the vapors likely to be airborne during fueling operations. They reported that rats exposed to 5000, 10,000, or 20,000 μg/m3 (approximately 1900, 3700, or 7500 ppm) of this vapor during gestation showed no treatment-related effects other than hyaline droplet nephropathy in the kidneys of the male rats. Because this effect is specific to the male rat kidney (Hard et al., 1993), it did not raise concern for human health. Reproductive performance was not affected: maternal growth and litter sizes in the F1 generation were normal and the offspring survived and grew normally; no effects were seen in the F2 generation. Generally null findings consistent with this study have been reported from rats exposed to vapors of unleaded gasoline (Roberts et al., 2001) and gasoline containing 10% ethanol (Gray et al., 2014; Roberts et al., 2014) at concentrations up to 9000 ppm. Similarly low toxicity has been reported from rodent studies with high flash aromatic naphtha, a product of petroleum refining that contains 70–80% 9-carbon aromatics (McKee et al., 1990; Schreiner et al., 2000). A subchronic study of the effects of vapors from a blend of gasoline and 10% ethanol (E10) in adult rats identified increases in glial fibrillary acidic protein in the brains of males (Clark et al., 2014), suggesting the possibility of neurotoxicity from ethanol-blended gasolines.

None of these prior studies evaluated the potential developmental neurotoxicity of gasoline vapors, or the effects of vapors from fuels containing more than 10% ethanol. The present experiments were conducted to address these issues and the widely-recognized concerns regarding the effects of gestational exposure to ethanol on fetal development. In these experiments, we exposed pregnant rats to vapors of ethanol, gasoline, and two gasoline–ethanol mixtures. We targeted the developing fetus by exposing pregnant rats during the second and third weeks of gestation, a critical period of development during which orally-administered ethanol impairs development of many physiological systems including the CNS (Becker, 1996; Savage et al., 2002), the immune system (Jerrells and Weinberg, 1998), neuroendocrine functions (Zhang et al., 2005) including glucose homeostasis (Chen and Nyomba, 2003) and insulin regulation (Chen and Nyomba, 2004), and cardiovascular functions (Ren et al., 2002). Our studies were designed to compare a wide range of concentrations of ethanol in gasoline, including neat ethanol (E100), gasoline blended with either 85% or 15% ethanol (E85 or E15), or gasoline without ethanol (E0). This strategy enables quantitative comparisons of dose–effect relationships as a function of ethanol concentration, so that the toxicity of any blend could be estimated. Given that public exposure to these mixtures would occur primarily by breathing vapors during fueling and parking in enclosed garages, the exposures were conducted via the inhalation route. Because gasoline is a complex mixture of hydrocarbons with a wide range of volatilities, and because repeated daily exposures were planned during the animals' pregnancies, it was important to generate an airborne mixture that remained consistent across hours within each exposure day and across days of gestation. Because vapors generated by heating the entire mixture would change in composition as the more volatile components evaporated first and the less volatile components followed later, we adopted the strategy of exposing the rats to the vapors of condensed evaporative emissions of these fuels (Henley et al., 2014). These vapor condensates (VCs) were generated prior to the experiments by collecting and compressing the most volatile 10% of each fuel mixture; exposures were then conducted simply by warming and releasing the vapor into chambers in which the rats were housed. The studies of inhaled E100 (ethanol) have been reported previously (Beasley et al., 2014; Boyes et al., 2014; Oshiro et al., 2014). The results of tests of the toxicity of the fuel blends are being reported similarly in three parts. This report documents effects of exposure to the fuels in pregnant rats and their offspring on tests not involving cognitive or sensory functions, which will be reported separately. Tests in the offspring reported here include screening-level assessments of behavioral development using a well-validated battery of tests of unconditioned behaviors (functional observational battery, FOB) and motor activity. Because published evidence (see above) suggested that heart function and glucose homeostasis might be targets of inhaled ethanol and gasoline vapors, systolic blood pressure and serum concentrations of glucose and glycated hemoglobin A1c (GHbA1c, a marker of long-term glucose status) were measured to test these hypotheses. Tests of immune function include humoral and cell mediated immunity. 2. Materials and methods 2.1. Overall design Three experiments were conducted with the same basic design that was used for the ethanol study (Beasley et al., 2014). In each experiment, pregnant rats were exposed to air or one of three concentrations of vapor of a single fuel blend (E0, E15, or E85) daily for 6.5 h from gestational day (GD) 9 to 20. To evaluate the effects of housing in the exposure chamber during gestation, an additional control group was added to the E0 experiment. This ‘cage-control’ group was kept in the vivarium during pregnancy and allowed to deliver their offspring

P.J. Bushnell et al. / Neurotoxicology and Teratology 49 (2015) 19–30

under standard housing conditions. Offspring from the rats in all experiments were raised, weaned, and distributed to participating laboratories in the EPA for assessment of the functions identified above. 2.2. Subjects Pregnant Long–Evans rats (n = 84 for E0 and n = 72 for E15 and E85) weighing 157–243 g were received from Charles River Laboratories (Raleigh, NC) on GD 2. Upon arrival, the rats were individually housed in polycarbonate cages with hardwood chip bedding (Beta-Chip®, Northeastern Products, Warrensburg NY) and Enviro Dri® nesting material (Shepherd Specialty Papers, Watertown, TN) for enrichment. The rats were allowed to acclimate for 7 days in the vivarium, a facility that follows the NIH guidelines for animal care and is fully accredited by the Association for Assessment and Accreditation of Laboratory Animal Care International. The vivarium was maintained at a temperature of 21 ± 1 °C and relative humidity of 40 ± 10%, with a 12:12 h light:dark cycle (lights on at 6:00 am). All procedures were approved by the Institutional Animal Care and Use Committee of the EPA's National Health and Environmental Effects Research Laboratory, which ensures conformance with the 1996 NRC “Guide for the Care and Use of Laboratory Animals”, the Animal Welfare Act, and Public Health Service Policy on the Humane Care and Use of Laboratory Animals. Rats were weighed daily during the acclimation period and allowed free access to filtered tap water and rat chow (Purina 5008, PMI Nutrition International, Richmond, IN). 2.3. Exposure methods 2.3.1. Animal exposures For exposure to the vapor condensates, rats were transported from their home cages in the vivarium and placed in one of four 1-m3 Hazelton 1000 Inhalation chambers (Lab Products, Seaford, DE) on GD 9, where they remained until GD 20 (except that the cage control rats in E0 were not moved from their home cages in the vivarium). Each rat was placed in an individual stainless steel wire mesh cage in the inhalation chamber and each cage was equipped with a stainless steel “loft” that provided environmental enrichment as well as reduced the potential for pressure-induced peripheral neuropathy from the wire bottom cages (Beasley et al., 2014; Benevenga et al., 2005; Mizisin et al., 1998). Dams were weighed every other day at the beginning of the day during the exposure period. Rats were allowed free access to rat chow except during the exposure periods; water was supplied ad libitum at all times. The lighting schedule of the inhalation chambers followed that of the vivarium. Each pregnant female rat (except the cage controls in the E0 study, n = 12) was randomly assigned to one of the four exposure groups in each experiment (0, 3000, 6000, or 9000 ppm; n = 18/group). Vapors of the condensates were generated in each of the inhalation chambers and exposures occurred for 6.5 h per day for 12 consecutive days, beginning on GD 9. Daily exposures included 30 min of rise time and 6 h at the target concentration. The rats were removed from the inhalation chambers following exposure on GD 20 and returned to the vivarium where they were housed individually for the remainder of their pregnancies, with food and water available ad libitum. Day of birth was designated as Postnatal Day (PND) 0. 2.3.2. Generation and characterization of the vapor condensates The vapor condensates (VCs) of the fuels were generated for this project by Chevron USA, Inc. (Richmond, CA) under contract with the EPA. The gasoline starting material was a blend of conventional premium and regular unleaded summer-blend gasoline purchased by Chevron from their commercial fuel pipeline and blended to meet specifications set by the EPA (Supplemental File 1). Fuel-grade denatured ethanol with a corrosion inhibitor (Octel DCI-11) was also purchased from Chevron's commercial fuels. The feedstock for generating the E0

21

VC was gasoline with no added ethanol. Feedstocks for the E15 and E85 VCs were made by blending ethanol with the gasoline at 15% and 85% (V:V), respectively. Feedstocks were stirred to equilibration before distillation and characterized (ASTM D-86) to produce full distillation curves (Supplemental File 2). Next, 100 gal of each VC was generated by stirring and slowly heating 1000 gal of each feedstock, drawing off the vapor and condensing it between −60 °F and −40 °F, until a volume equal to 10% of the feedstock was obtained. The VCs were characterized to quantify both the ethanol concentration in the E15 and E85 vapors (ASTM D-4815 for oxygenates in gasoline) and by gas chromatographic methods (Chevron SE-30) to quantify the hydrocarbon constituents. The VCs were then shipped to the EPA in 5-gal pressurized cylinders from which we generated the atmospheres to which the rats were exposed. Upon arrival at the EPA, the concentrations of more than 100 hydrocarbon components of the mixtures in the pressurized VCs were measured using a Gas Chromatography/Flame Ionization Detection (GC/FID) method. Operational methods for the GC/FID (temperatures, runtime, etc.) have been reported previously (Blunden et al., 2005; Lough et al., 2005). Supplemental File 3 provides details of the sampling and analysis methods. Fig. 1 shows a profile of the concentrations of the 20 most abundant constituents of each mixture; these compounds account for about 90% of the total volumes of the mixtures. Atmospheres containing vapors of E0, E15 and E85 were generated using a dynamic countercurrent evaporator that was adapted from a design by Miller et al. (1980) and applied as described by Beasley et al. (2014). The temperature of the evaporator was set to 15 °C above the boiling point of the VC with the carrier gas flowing (51 °C for E0 and E15 and 93 °C for E85). The liquid VC was transferred from a pressurized cylinder to the evaporator with nitrogen gas, which displaced the VC through a dip tube to a liquid mass flow controller. Once the evaporator output stabilized, condensate vapor was directed to the inlet of the exposure chamber for mixing and dilution. 2.3.3. Monitoring the vapor condensate exposures The initial flow rate of the VC was calculated based on the chamber air flow, and adjusted as needed to achieve target vapor concentrations in the exposure chambers. Vapor concentrations (as total hydrocarbon) in each exposure chamber were monitored with three Miran 1a Infrared Gas Analyzers (Wilks Foxboro Analytical, South Norwalk, CT). Each analyzer sampled one chamber containing vapor and also periodically sampled the air chamber in a systematic rotation. The analyzers were tuned to the methane-stretch wavelength of hydrocarbons (3.5 μm, or 2857 cm−1 wavenumbers) and calibrated with either n-pentane (E0 and E15) or 100% ethanol (E85), for the appropriate chamber concentration ± 10%. The CH3 wavelength was chosen because it is present in all compounds present in the vapor condensates. The total hydrocarbon concentration (THC) in each chamber was monitored from before the start of each daily exposure to 30 min after the end of each exposure. In addition to the THC measurements, at least two grab samples of the vapors were obtained from each chamber each day during the 12 days of exposure, at least one in the morning and one in the afternoon. These samples were drawn into evacuated 2.7-L or 6-L stainless steel canisters and analyzed for concentrations of ethanol and more than 100 individual hydrocarbons. The methods for collecting and processing the chamber samples, and the adjustments to the GC system that enabled analysis of the high concentrations of the vapors are described in Supplemental File 3. The organic composition of the liquid samples from the pressurized cylinders and vapor samples from the exposure chambers were identical within the measurement precision of the GC/FID analysis system, which was about 2.5%. The temperature and relative humidity in each chamber were monitored during the exposure and non-exposure periods using a Rotronic Model I 200 Thermo-Hygrograph (Huntington, NY) interfaced to the data acquisition system (DaqLab 2001 with DasyLab software v8; Measurement Computing Corp, Norton, MA). During the exposure

22

P.J. Bushnell et al. / Neurotoxicology and Teratology 49 (2015) 19–30

A) Percentages in pressurized shipping cylinders. 60

Percent of Mixture

50

E0 E15 E85

40

30

20

10

0 l e e e e e e e e e e e e e e e e e e e ano tan tan tan tan tan xan uen tan tan tan xan non ten xan tan tan tan tan ten Eth sopen n-Pen n-Bu ylpen ylpen n-He Tol ylpen hylbu lopen clohe enta -2-Bu hylhe ylpen ylpen -Hep Isobu 2-Pen n I h eth t -P hyl th eth th met lcyc Cy t s e 3 e e e n i + et 2-M Dim Dim tra 3-M ,3-D ethy 2-M -TriM ne 2-M 2 M ,4 exa 2,3 2,4 2,2 ylh h t e 3-M

B) Percentages in the inhalation chamber air. 60

Percent of Mixture

50

E0 E15 E85

40

30

20

10

0 l e e e e e e e e e e e e e e e e e e e ano tan tan tan tan tan xan uen tan tan tan xan non ten xan tan tan tan tan ten Eth sopen n-Pen n-Bu ylpen ylpen n-He Tol ylpen hylbu lopen clohe enta -2-Bu hylhe ylpen ylpen -Hep Isobu 2-Pen n I t -P hyl th met lcyc Cy th eth th eth s e 3 e e e n t a i tr e + Me 2-M -Dim -Dim 3-M ,3-D ethy 2-M -TriM xan 22 M ,4 2,3 2,4 lhe 2,2 thy e 3-M

Fig. 1. Major constituents of the vapor condensates. The twenty compounds of the three mixtures that were present in highest concentration are plotted as a percent of the total volume of each mixture. These 20 compounds comprised about 90% of the total volumes of the mixtures. Traces of ethanol were present in the E0 condensate. Iso-pentane and n-butane were the constituents most strongly displaced by the ethanol in the E15 and E85 mixtures. A. Percentages in pressurized shipping cylinders. B. Percentages in the inhalation chamber air.

periods the temperature ranged from 21.8 °C to 24.3 °C and the relative humidity from 45% to 55%. 2.4. Assessments of the animals during exposure All dams, except for the cage-control animals in E0, were weighed daily during the acclimation period in the vivarium (GDs 2–8) and on alternate days during the exposure period. Cage-control dams were weighed on GDs 2, 3, 8, 11, 15 and 19. Dams in the inhalation chambers were also monitored for adverse effects during and after each exposure. Daily food intake was monitored in the pregnant rats during exposure to discriminate any potential effects of exposure from those related to maternal and fetal nutrition (Kalev-Zylinska and During, 2007;

Shankar et al., 2007). Total daily intake of food was measured every morning following the ~18 h overnight period in clean air. Food hoppers were removed before daily exposures to prevent ingestion of vapor residues in the food. The hoppers were filled with rat chow pellets, weighed, and replaced into each rat's compartment after each daily exposure. The following morning, hoppers were removed, placed in a plastic bag, and weighed. The daily change in weight was recorded as total daily intake of food (g/day). 2.5. Assessments of offspring After the last exposure on GD 20, dams were moved from the inhalation chambers to the vivarium where they were housed in standard

P.J. Bushnell et al. / Neurotoxicology and Teratology 49 (2015) 19–30

23

plastic home cages with bedding. Pups were born 1–2 days later; the day of birth was defined as PND 0.

because each performs a different and important role in resistance to different types of infectious agents.

2.5.1. Redistribution of offspring On PND 3 (E0 and E15) or PND 2 (E85), all pups from the dams of a given exposure concentration were removed from their mothers, weighed, pooled by sex, and reassigned at random to the dams in that exposure group such that each dam received 6 male pups and 4 female pups whenever possible. No attempt was made to keep former littermates together or apart or to return pups to their biological mothers. In cases where it was not possible to give a dam 6 males and 4 females, these litters were filled with as many males as possible then adjusted to 10 with female pups. The basis for this process was 1) to standardize the litters across dams to eliminate the effects of unequal litter size or sex ratio (Agnish and Keller, 1997), and 2) to include dams with litters of fewer than 10 of her own pups by the addition of pups from other litters. Excess pups were euthanized, and their brains were weighed. Henceforth, the term ‘litter’ will refer to the unit of pups that were grouped together with a given dam during the redistribution process.

2.5.4.1. Humoral immunity. Eight offspring of each sex and exposure group were immunized by intravenous injection of a novel antigen, sheep red blood cells (SRBC; Rockland Immunochemicals Inc., Gilbertsville, PA), to stimulate a primary (IgM) antibody response. Serum samples were obtained 5 days after immunization (when serum concentrations of anti-SRBC IgM are typically at their peak) and frozen for later analysis by ELISA.

2.5.2. Pup growth and weaning Dams and litters were weighed every other day throughout lactation, and litters were separated by sex for weighing. Pups were separated from their dams on PND 22 and then allocated among laboratories for toxicity assessments. No more than one male and one female pup from each litter were allocated to any single assessment. After weaning, offspring were provided rat chow (Purina 5001) and water ad libitum. 2.5.3. Neurobehavioral assessments Behavior was evaluated using motor activity and a functional observational battery (FOB) as described by Beasley et al. (2014), using one male and one female from each of 10 litters (n = 10/concentration/ sex). Motor activity and FOB tests were conducted in the same animals at each of several ages over the course of one to three days. 2.5.3.1. Motor activity. Motor activity was monitored in photobeambased figure-eight chambers (Reiter, 1983) that included a bank of photobeams 14 cm above the chamber floor to register vertical activity counts. Testing was conducted in pre-weaning pups on PNDs 13, 17, and 21 in 30-min sessions, and again in 60-min sessions immediately after FOB testing on PNDs 27–29 and 56–58 in E0, on PNDs 27–28 and 55–56 in E15, and on PNDs 27–28 and 63–64 in E85. Activity was defined as the total number of photobeam interruptions, and was collected in 5-min intervals during each session. Because post-weaning tests took place over several days, sex and treatment groups were counterbalanced across time of day and activity chamber. 2.5.3.2. Functional observational battery (FOB). A modified FOB protocol (McDaniel and Moser, 1993) was administered to the animals twice after weaning at the ages specified above, as described previously (Beasley et al., 2014). Functional assessments included general appearance (lacrimation, salivation, and muscle tone), handling reactivity, body posture, tremors, gait characteristics, activity, arousal, sensorimotor reactions to an audible click or tail pinch, urination, defecation, pinna reflexes, proprioceptive positioning reaction of a hindlimb, aerial righting reflex, forelimb and hindlimb grip strength, landing foot splay, and body weight. The observer was unaware of the rat's exposure condition during these tests. 2.5.4. Immune function Immune function was assessed in female and male offspring (n = 8/ concentration/sex) at 6 weeks of age, which corresponds to the onset of immune system maturity. General methods followed those described by Beasley et al. (2014) with assays conducted as detailed by DeWitt et al. (2006). Both humoral and cellular functions were assessed

2.5.4.2. Cellular immunity (Delayed Type Hypersensitivity (DTH)). Eight additional offspring of each sex and exposure group were anesthetized using isoflurane and sensitized to purified (fraction V) bovine serum albumin (BSA; Sigma-Aldrich, St. Louis, MO), emulsified in complete Freund's adjuvant, by subcutaneous injection at the base of the tail. Seven days later, sterile heat-aggregated BSA was injected into the right rear footpad and sterile normal saline was injected into the left rear footpad of anesthetized rats. Twenty-four hours later, the rats were anesthetized a third time and the thickness of both rear feet was measured to assess antigen-driven inflammation. Rats were euthanized after this final measurement. 2.6. Physiological assessments Physiological assessments included blood pressure (BP) measurements and analyses of blood serum for glucose homeostasis (glucose and glycated hemoglobin A1c concentrations). Blood samples and BP measurements were taken from each rat twice: at about 3 months (all experiments), and at about 6 months (E0 and E15) or 8.5 months of age (E85). Details of the methods are provided in Supplemental File 4. 2.6.1. Blood pressure Blood pressure (BP) was measured by tail-cuff plethysmography (IITC Life Science, Woodland Hills, CA) as described by Beasley et al. (2014). Briefly, the rat was restrained and warmed to dilate the tail vein, after which five consecutive readings of systolic BP were acquired. Sets of readings for a given animal with a range greater than 20 mmHg in systolic pressure were excluded from analysis. Stable sets of five readings of BP for each rat were averaged for analysis. 2.6.2. Glucose homeostasis Five days after measurement of BP, blood was drawn from each rat after an overnight fast for measuring serum markers of glucose homeostasis. Venous blood was drawn from the tail vein at 3 months of age; at 6 (E0 and E15) or 8.5 months of age (E85), trunk blood was collected after decapitation. Concentrations of glycated hemoglobin A1c (GHbA1c) were measured in whole blood to assess long-term glucose status. Blood serum was prepared and analyzed for analysis of glucose concentration in serum using a KONELAB-30 automated clinical chemistry analyzer (Thermo Clinical Labsystems, Espoo, Finland). 2.7. Statistical analyses Studies with developmental exposures using multiparous animals require careful statistical evaluation, due in part to demonstrably low within-litter variance relative to between-litter variance (Holson et al., 2008). In the present experiments, the birth litter unit was dissociated during redistribution, and new litters were assembled (without regard for the birth litter) to standardize the postnatal environment across litters and exposure groups. This process distributed any influence of the prenatal environment across all litters, thereby controlling for its potential effects on outcomes. Given the importance of the early postnatal period for functional development, however, we retained the litter as the statistical unit for analyzing the effects of exposure to the vapors, selecting either one male pup, or one pup per sex, from each litter for

24

P.J. Bushnell et al. / Neurotoxicology and Teratology 49 (2015) 19–30

each assessment. In addition, we nested sex within litter for the analyses of variance for each assessment in which both sexes were tested, to account for the litter-induced dependence between these observations. All analyses were conducted with SAS software (SAS, Cary, NC; Software Release 9.1, 9.2 or 9.3). Unless otherwise noted, a level of p b 0.05 was considered significant for each analysis. The cage-control group in E0 was compared to the air-control group to determine whether gestation in the inhalation chambers altered physical and behavioral development. All effects of the vapors were tested with respect to the air-control group only. 2.7.1. Dam and pup weights and dam food consumption The gestational weight gain of the dams was defined as the difference in body weight between GD 8, the last day before exposure, and GD 20, the last day of exposure. Daily food consumption amounts were summed across all days of exposure yielding a value of total grams of food consumed. Weight gain, food consumption, and the ratio of weight gain per gram of food consumed were analyzed for each experiment using one-factor analyses of variance (PROC GLM) with concentration as a between-subject factor. Pup weights on PND 2 and PND 20 were analyzed for males and females separately using concentration as a between-subjects factor in separate ANOVAs. Significant differences were followed with Dunnett's tests to compare each treated group to the control group. Losses of pups from litters were tested by a chi-square test comparing the number of affected litters across the four treatment groups combined across the three experiments. 2.7.2. Behavioral assessments Session totals of horizontal and vertical activity counts were analyzed with ANOVA (PROC GLM) with concentration as a between-subject factor and sex nested within litter. Habituation of activity counts (horizontal plus vertical) was analyzed with a repeated measures ANOVA (PROC GLM) as above, but including time interval as a second repeated measure. FOB data were converted to severity scores for each endpoint and then summed to produce overall scores for each functional domain (McDaniel and Moser, 1993; Moser et al., 1995, 1997), which were then analyzed using traditional ANOVAs with PROC GLM. Individual measures within domains were subsequently analyzed only if the overall domain effect was significant. Significant interactions between concentration and sex were followed by analyses of each sex separately, whereas male and female data were combined for each litter when only the concentration effect was significant. Post-hoc tests for differences between exposure groups and control used Dunnett's procedure. Non-parametric data (e.g., ranked FOB scores) were analyzed using Kruskal–Wallis test for each sex separately. 2.7.3. Immunological assessments Foot-pad thickness (cellular immunity) and IgM titers (humoral immunity) were analyzed using separate two-way ANOVAs with concentration as a between-group factor, sex nested within litter, and a compound symmetry covariance structure. IgM titers were analyzed as log(2) to reflect the serial 1:2 dilutions used in the assay. Both analyses were run in PROC Mixed with post-hoc comparisons across concentrations using Dunnett's tests. 2.7.4. Physiological assessments Measures of BP, glucose, and glycated hemoglobin A1c were subject to independent ANOVAs (PROC Mixed) with concentration as a betweensubject variable, age as a repeated measure, and sex nested within litter. The compound symmetry covariance structure was selected either because it yielded the best fit (lowest value of the corrected Akaike Information Criterion for model fit) or because of a lack of convergence with other structures. In the case of significant interactions with concentration, separate analyses by sex were examined to identify the factors contributing to the concentration effect. Post-hoc analyses used the Dunnett–Hsu procedure for comparing treatment groups with control.

3. Results 3.1. Exposures 3.1.1. Characterization of the vapor condensates The twenty constituents of the three vapor condensates that were present in highest concentration are plotted as a percent of the total volume of each mixture as pressurized condensates (Fig. 1A) and as components of the atmosphere in the inhalation chambers (Fig. 1B). These 20 compounds comprised about 90% of the total volumes of the vapors. Traces of ethanol were present in the E0 condensate; ethanol comprised 12% and 35% of the E15 and E85 condensates, and 20% and 57% of the chamber atmospheres of the E15 and E85 mixtures, respectively. Iso-pentane and n-butane were the constituents most strongly displaced by the ethanol in the E15 and E85 mixtures under both conditions. 3.1.2. Exposure conditions The atmospheric conditions during exposure in each inhalation chamber are summarized in Table 1. The mean vapor concentrations, measured as total hydrocarbon concentrations (THCs), were close to the target values. Of the 280–296 individual measurements of THC obtained per chamber in each of the three experiments, up to 6% fell above, and up to 20% fell below, the target window (nominal concentration ± 10%). Equipment failure during the exposures to E15 caused a loss of vapor generation for 1.75 h on GD 10 and for 2 h on GD 20 in the 9000-ppm chamber. With the exception of these 2 shortened sessions, all experimental rats were exposed to a vapor for 6.5 h/day for 12 days during pregnancy. Temperature and humidity values were well controlled. Table 1 also shows the average concentrations of ethanol measured in each chamber atmosphere. Ethanol concentrations in the chamber air were about 25% of THC in the E15 experiment and 70% of THC in the E85 experiment. Traces of ethanol were detected in the chamber receiving E0 vapors, amounting to about 0.6% of THC. 3.2. Assessments during exposure and pregnancy outcomes No signs of intoxication or distress were observed in the pregnant rats during exposures. All groups of dams gained equivalent amounts of weight between GD 8 and GD 20 in all experiments (Table 2). Whereas food consumption itself was not significantly affected by exposure to E0 or E15, there was a concentration-related decrease in food consumption in the E85 experiment, which was statistically significant (F(3,68) = 3.81, p b 0.02) for the 9000-ppm group only (Dunnett's test). However, the ratio of body weight gain to food consumption by the dams showed an upward trend across exposure concentrations for all experiments except E15 (Fig. 2). These ratios were significantly increased in the 9000-ppm groups of the experiments with E0 (F(3,67) = 2.91, p b 0.05) and E85 (F(3,67) = 2.70, p b 0.06), but not E15 (F(3,66) b 1). None of the vapors affected any of the measured pregnancy outcomes including litter size, sex ratio, or the pre-weaning growth of the pups of either sex (Table 2). In some litters, pups were cannibalized by their mothers during the first week after birth: loss of litters was most apparent in the E15 experiment, in which litters from all four groups were lost. The number of affected litters was not significantly related to the concentration of vapor (χ2 (3) = 3.09, ns). 3.3. Postweaning growth of offspring Body weights after weaning were not systematically affected by exposure to any of the vapors. No significant changes in growth were observed in offspring of rats exposed to E0 or E15 (data not shown). In the E85 experiment, a significant concentration by day interaction was found in males only (F(24,288) = 1.77, p = 0.0161); however,

P.J. Bushnell et al. / Neurotoxicology and Teratology 49 (2015) 19–30

25

Table 1 Fuel vapor exposure data. E0, E15 and E85 refer to the fuel mixtures used in the three experiments at nominal concentrations of 0, 3000, 6000, and 9000 ppm of the vapor. All concentrations are in parts per million (ppm). Experiment

E0

Nominal concentration

0

3000

6000

9000

0

3000

6000

9000

0

3000

6000

9000

0 0 0 0 296 0 0 21.8 55

2993 554 20 8 297 0 52 23.0 50

6068 678 37 9 293 0 32 23.6 52

8918 692 55 14 289 0 23 23.0 52

0 0 0 0 291 0 0 22.2 50

2933 355 703 167 292 8 25 23.7 45

5950 701 1531 487 292 15 19 24.2 46

8585 1503 2264 343 280 3 39 24.0 46

0 0 0 0 286 0 0 22.1 52

2750 465 1767 282 286 1 58 23.6 47

5758 706 4081 275 286 9 34 24.3 46

9001 1003 6758 583 286 18 14 24.1 46

a

Actual THC mean SD b Ethanol (ppm) meanc SD Number of readings d Number highe Number lowf Mean temperature (°C) Mean relative humidity (%) a b c d e f

E15

E85

THC = total hydrocarbon concentration. SD = standard deviation. Concentration of ethanol measured in the vapor in the chamber atmosphere during exposures. Number of measurements of the THC during all exposures. Number of THC measurements that were N110% of the target concentration. Number of THC measurements that were b90% of the target concentration.

no significant group differences appeared on any single day and greater weight gain was apparent only in the 3000 ppm group (data not shown). In the E0 experiment, differences in body weight between the air-control and cage-control groups emerged near the end of FOB testing at PND 50 (day by housing interaction F(8,144) = 3.52, p = 0.0010); this difference persisted until the animals were terminated on PND 120 (effect of housing F(1,18) = 4.56, p = 0.0467; Supplemental File 5).

3.4. Behavioral assessments 3.4.1. Motor activity Compared with controls, none of the vapors significantly affected motor activity through PND 29 (data not shown). On PND 63 (Fig. 3), prenatal exposure to the E85 vapor significantly increased vertical activity counts (main effect of concentration F(3,36) = 3.37, p b 0.03). Post-hoc tests showed that activity of the sexes combined was significantly increased in the 3000- and 9000-ppm groups (121% and 125% of control, respectively), and although the group mean of the 6000 ppm group was similar (119%) it did not differ significantly from control by Dunnett's test. While horizontal activity showed a similar pattern, analysis of a significant sex by concentration interaction (F(3,36) = 3.41, p b 0.03) revealed no significant difference between

controls and any of the treated groups for either sex alone (data not shown). Neither measure of motor activity was significantly altered by gestational exposure to E0 or E15 at any age. The change in activity across time within test sessions (habituation) was not significantly altered by gestational exposure to any of the vapors at any age (data not shown). 3.4.2. Functional observational battery The autonomic, neuromuscular, excitability, activity, and general health domains of the FOB were not affected by exposure to any vapor (data not shown). The tail-pinch response, a component of the sensorimotor domain, was altered in the male offspring in the E0 experiment only, with equivalently lower responses than control in the 6000-ppm group at PND 27/28, and in the cage controls and all exposed groups on PND 56/57. Similarly low absolute scores were also recorded in the E85 treated groups on PND 63/64; however, in that experiment the controls were also low, so it is not clear whether the effects of E0 resulted from higher control responses or from suppressed responses in the treated and cage-control rats (Supplemental File 6). 3.4.3. Cage vs air controls (E0 only) The offspring from the cage and air control groups in the E0 experiment did not differ in motor activity at any time during development

Table 2 Pregnancy outcomes in the E0, E15 and E85 experiments. Pregnant female

Pups per litter

Pup weights PND2 (g)

Pup weights PND20 (g)

Postnatal pup loss

Experiment

Nominal concentration

Number pregnant

Food consumed (g)

Body weight gain (g)

Male

Female

Male

Female

Male

Female

Litters

E0

Air 3000 6000 9000 Cage Air 3000 6000 9000 Air 3000 6000 9000

18 17 18 18 12 18 18 16 18 18 18 17 18

307.7 ± 8.1 311.9 ± 8.5 298.9 ± 9.5 302.8 ± 8.0 nd 314.4 ± 5.5 294.9 ± 6.1 305.8 ± 7.4 311.7 ± 10.4 324.2 ± 6.3 319.1 ± 7.3 302.4 ± 6.7 294.5 ± 8.1a

101.1 ± 4.8 101.7 ± 4.9 102.1 ± 4.8 108.9 ± 3.5 nd 99.5 ± 3.0 94.1 ± 3.0 93.7 ± 4.8 97.7 ± 4.7 99.3 ± 2.8 105.9 ± 3.0 101.6 ± 4.8 99.5 ± 5.5

5.8 ± 0.6 6.6 ± 0.5 5.4 ± 0.5 6.0 ± 0.3 5.7 ± 0.5 5.9 ± 0.5 6.1 ± 0.5 5.8 ± 0.5 5.8 ± 0.5 5.7 ± 0.5 5.7 ± 0.6 5.8 ± 0.7 6.2 ± 0.4

5.3 ± 0.4 4.9 ± 0.3 6.0 ± 0.5 6.3 ± 0.5 6.3 ± 0.7 5.6 ± 0.4 5.6 ± 0.3 6.7 ± 0.4 5.5 ± 0.5 5.6 ± 0.5 6.2 ± 0.5 5.6 ± 0.5 5.7 ± 0.5

7.9 ± 0.6 8.2 ± 0.4 8.0 ± 0.3 7.8 ± 0.3 8.0 ± 0.6 7.6 ± 0.1 7.5 ± 0.1 7.4 ± 0.1 7.6 ± 0.1 6.8 ± 0.1 6.8 ± 0.0 6.8 ± 0.1 6.9 ± 0.0

7.8 ± 0.4 7.8 ± 0.4 7.9 ± 0.3 7.4 ± 0.3 7.9 ± 0.3 7.1 ± 0.1 7.5 ± 0.1 7.4 ± 0.1 7.4 ± 0.1 6.7 ± 0.1 6.6 ± 0.0 6.7 ± 0.1 6.5 ± 0.1

45.1 ± 0.6 46.8 ± 0.7 46.3 ± 0.7 45.7 ± 0.7 46.0 ± 0.7 44.9 ± 0.9 44.0 ± 0.8 42.6 ± 1.1 44.2 ± 0.9 46.6 ± 0.8 46.3 ± 1.4 44.3 ± 1.2 49.4 ± 2.1

43.9 ± 0.6 45.4 ± 0.5 45.5 ± 0.7 44.3 ± 0.9 45.1 ± 0.8 45.8 ± 0.7 43.3 ± 0.7 42.5 ± 0.9 42.5 ± 1.1 46.1 ± 0.4 44.9 ± 1.5 45.8 ± 1.2 43.7 ± 1.3

0 0 0 0 1 2 3 4 4 0 0 2 1

E15

E85

nd = not determined. Postnatal pup loss was due to death and cannibalization of pups during the first week after birth. The losses were greatest in the E15 experiment, but were not significantly related to the vapor concentration. a Significantly different from control (p b 0.05).

26

P.J. Bushnell et al. / Neurotoxicology and Teratology 49 (2015) 19–30

0.40 E0 E15 E85 E100

0.38

Table 3A IgM titers (log2) reflecting humoral immune function of offspring immunized at 6 weeks of age.a

* *

0.36

Sex

Concentration (ppm)

E0

E15

E85

Male

0 3000 6000 9000 0 3000 6000 9000

8.0 ± 0.4 7.3 ± 0.4 7.4 ± 0.4 8.1 ± 0.4 7.8 ± 0.4 6.1 ± 0.4b 8.9 ± 0.4 7.4 ± 0.4

7.4 ± 0.6 5.9 ± 0.6 7.1 ± 0.7 9.0 ± 0.6 6.9 ± 0.7 7.3 ± 0.7 6.7 ± 0.7 7.2 ± 0.6

6.6 ± 0.5 6.6 ± 0.5 6.5 ± 0.5 7.1 ± 0.5 5.8 ± 0.5 6.8 ± 0.5 5.4 ± 0.5 6.5 ± 0.5

*

0.34

Female

0.32

a

0.30

b

Least Squares Means ± standard error, 8 rats/sex/concentration. Significantly different from control (p b 0.02).

0.28 0

5000

10000

15000

20000

Vapor Concentration (ppm)

3.6. Physiological assessments

Fig. 2. Body weight gain of pregnant rats during exposure to the fuel vapors as a function of food consumption. Weight gain is the difference, in grams, between body weights on GD 20, the last day of exposure, and on GD 8, the last day before exposure began. Daily food consumption was summed across the 12 days of exposure. The figure shows the ratio of weight gain per gram of food consumed as a function of exposure concentration for the three fuel-vapor experiments and for ethanol (E100) (Beasley et al., 2014). Asterisks mark significant increases in weight gain per gram of food consumed in the high-concentration group of each experiment except for E15, in which no increase was observed.

(data not shown). On the other hand, cage-control males showed a lower tail-pinch response (Supplemental File 6) and cage-control females showed higher click responses compared to the corresponding air-control groups on PND 56/57 (Supplemental File 7).

3.5. Immunological assessments For humoral immunity (Table 3A), all statistical comparisons were negative except for a main effect of vapor concentration (F(3,28) = 4.21, p b 0.02), and a concentration by sex interaction on IgM titers (F(3,28) = 3.46, p b 0.03) in the E0 experiment. Dunnett's tests revealed that these effects resulted from lower IgM titers in the 3000 ppm females compared to control females. No other treatment groups differed significantly from control. Maternal exposure to the vapors did not affect cell-mediated immunity in the offspring in any experiment (Table 3B).

Mean (±SE) Vertical Counts

120 100

E0

Cage Air 3000 ppm 6000 ppm 9000 ppm

3.6.1. Blood pressure Different patterns of effects of prenatal exposure to the vapors, age, and sex were observed on systolic blood pressure in the three experiments (Tables 4A, 4B, 4C). In the E0 experiment (Table 4A), blood pressure was higher in females than in males (main effect of sex F(1,87.2) = 23.23, p b 0.0001) and did not change with age. Analysis of a significant concentration by sex interaction (F(3,86.1) = 3.28, p b 0.03) revealed that it was due to the difference between the male and female controls, in contrast to attenuated sex difference in the exposed groups. In the E15 experiment (Table 4B), blood pressure was again generally higher in females than in males (F(1,81.3) = 7.05, p b 0.01) and decreased with age in both sexes (F(1,80) = 6.06, p b 0.02). Exposure to E15 increased blood pressure at 6 months of age (Concentration × Age interaction F(3,80) = 4.58, p b 0.006): analysis of this interaction revealed that its significance was due to lower blood pressure at 6 months relative to 3 months in the controls only. Gestational exposure to E85 (Table 4C) did not affect blood pressure significantly or interact with age or sex, although blood pressure again decreased with age in this experiment (F(1,28) = 11.96, p b 0.002), and did so more in males than in females (Sex by Age interaction F(1,28) = 8.19, p b 0.008).

3.6.2. Glucose homeostasis None of the vapors significantly affected the concentration of glucose in serum or the percentage of glycated hemoglobin A1c in whole blood in either sex at either age (Tables 4A, 4B, 4C).

120 100

Air 3000 ppm 6000 ppm 9000 ppm

E15

120 100

80

80

80

60

60

60

40

40

40

20

20

20

0

PND27

PND56

0

PND27

PND55

0

Air 3000 ppm 6000 ppm 9000 ppm

PND27

E85 *

*

PND63

Fig. 3. Motor activity in post-weaning offspring. Vertical activity counts, summed over the 60-min test session, increased with age in all experiments. All three treated groups in the E85 experiment were more active than their controls at PND 63, although only the 3000- and 9000-ppm group means were significantly different from control. No significant differences in activity across treatment groups occurred in the other experiments, including the cage-control group in the E0 experiment. No significant differences in horizontal motor activity were found in any experiment (data not shown).

P.J. Bushnell et al. / Neurotoxicology and Teratology 49 (2015) 19–30 Table 3B Footpad thickness (mm) reflecting cellular immune function of offspring immunized at 6 weeks of age.a Sex

Concentration (ppm)

E0

E15

E85

Male

0 3000 6000 9000 0 3000 6000 9000

0.6 ± 0.1 0.8 ± 0.1 0.8 ± 0.2 1.1 ± 0.1 0.9 ± 0.1 1.0 ± 0.2 1.0 ± 0.1 1.0 ± 0.1

0.6 ± 0.1 0.8 ± 0.1 0.5 ± 0.1 0.6 ± 0.1 0.9 ± 0.1 0.8 ± 0.1 0.8 ± 0.1 0.7 ± 0.1

0.5 ± 0.1 0.8 ± 0.1 0.5 ± 0.1 0.7 ± 0.1 0.6 ± 0.1 0.7 ± 0.1 0.6 ± 0.1 0.8 ± 0.1

27

constituents, which kept a fraction of the ethanol in liquid form in the pressurized cylinders, but allowed it to vaporize when heated and released into the inhalation chambers at atmospheric pressure. Ethanol preferentially replaced iso-pentane and butane in the mixtures under both sampling conditions. 4.2. Maternal effects during exposures

Pregnant rats were exposed to vapors of gasoline and two gasoline– ethanol blends at high total hydrocarbon concentrations that were limited only by safety considerations. These concentrations were selected to permit characterization of any observed effects, not to model typical public exposures. Despite these high concentrations, the fuel vapors exerted few measureable effects on the pregnant rats or their offspring. No signs of distress or narcosis were observed during the period of exposure. The litters from exposed rats were of normal composition and size; the exposed offspring grew at the same rate as controls; and behavioral assessments using motor activity and FOB measures revealed a few statistically-significant changes, none of which showed a monotonic relationship to concentration. Immune functions of the offspring at 6 weeks of age were largely unaffected by gestational exposure to the vapors. The vapors did not affect systolic blood pressure or serum glucose or glycated hemoglobin A1c concentrations, suggesting that they did not interfere with cardiovascular functions or glucose metabolism.

No signs of distress or discomfort were observed in the pregnant dams during or between exposures. All rats, including the controls, were quiet and apparently sleeping at the end of each exposure. This behavior was expected, given that exposures occurred during their normal diurnal period of inactivity. Moreover, the behavior of the exposed rats during maintenance operations after exposures was not obviously different from controls. The prominent observation from the previous study of inhaled ethanol (Beasley et al., 2014) was that the pregnant rats maintained a normal trajectory of weight gain during the exposure period while eating up to 10% less food than the controls. A similar pattern was observed in the E85 experiment, in which total food consumption was about 10% lower than control in the 9000-ppm group, despite equivalent weight gain. However, no differences in food consumption were observed in the E0 and E15 experiments (Table 2). The trends in weight gain per unit food consumption in Fig. 2 suggest that the rats exposed to E0, E85 and E100 obtained some caloric benefit from the inhaled hydrocarbons, as we previously documented for inhaled ethanol (Beasley et al., 2014). It is also possible that the vapors simply reduced the animals' metabolic rate, so that they required fewer calories to maintain body weight; however, the fact that exposures occurred during the animals' daily period of inactivity when metabolic rate is already low (Sugano, 1983) reduces the plausibility of this argument. Regardless of the mechanism involved, the fact that weight gain per unit food consumption increased in E0 and E85 but not in E15 argues that inhaled ethanol was not solely responsible for this effect.

4.1. Exposures

4.3. Assessments of offspring

The vapor condensates were typical of the light fraction of gasoline– ethanol blends, dominated by pentanes and butane (Fig. 1A). Ethanol concentrations in the inhalation chambers (Fig. 1B) were nearly double those measured in the pressurized cylinders, reflecting the relatively high boiling point of ethanol relative to the lightest hydrocarbon

4.3.1. Exposure metrics In the previous studies of inhaled ethanol (Beasley et al., 2014; Boyes et al., 2014; Oshiro et al., 2014), we estimated the peak blood ethanol concentrations of the pregnant rats during the gestational period of exposure to be 2.3, 6.7 and 192 mg/dL at airborne ethanol concentrations

Female

a

Least squares means ± standard error, 8 rats/sex/concentration.

4. Discussion

Table 4A Physiological assessments of offspring from the E0 experiment. Values are means ± SD for each endpoint. The Air control group was exposed to clean air in an inhalation chamber during gestation; the cage control group was housed in the vivarium during gestation. The other groups were exposed to 3000, 6000, or 9000 ppm of E0 vapor during gestation (GD 9–GD 20). Age

Body weight

Blood pressure

Glucose

Glycated HbA1c

Fuel

Sex

Months

Group

g

n

mm Hg

n

mg/dL

n

%

n

E0

M

3

E0

M

6

E0

F

3

E0

F

6

Air Cage 3000 6000 9000 Air Cage 3000 6000 9000 Air Cage 3000 6000 9000 Air Cage 3000 6000 9000

522.1 ± 21.93 528.8 ± 42.54 549.6 ± 41.11 523.5 ± 49.38 546.1 ± 41.43 697.0 ± 38.31 697.0 ± 42.61 707.0 ± 53.84 682.0 ± 75.92 735.0 ± 47.45 293.9 ± 20.66 279.3 ± 17.09 288.4 ± 30.46 283.3 ± 27.43 304.1 ± 21.16 332.0 ± 18.91 332.0 ± 14.46 431.0 ± 37.69 346.0 ± 53.23 372.0 ± 30.93

8 8 8 8 10 8 8 8 8 10 8 8 8 8 8 8 8 8 8 8

125 ± 23.79 143 ± 7.50 136 ± 11.87 142 ± 7.52 144 ± 20.50 131 ± 4.59 139 ± 12.84 139 ± 4.65 137 ± 7.70 141 ± 10.92 148 ± 9.26 152 ± 9.87 150 ± 6.51 142 ± 6.90 143 ± 7.26 142 ± 6.80 156 ± 7.03 151 ± 7.22 145 ± 11.63 144 ± 8.82

8 8 8 8 10 8 8 8 8 10 8 8 8 8 8 8 8 8 8 8

134.99 ± 7.86 116.73 ± 10.98 129.94 ± 6.50 130.84 ± 11.66 125.33 ± 2.85 132.06 ± 5.86 131.54 ± 10.13 134.30 ± 11.35 129.50 ± 9.66 133.49 ± 7.21 108.44 ± 1.66 105.24 ± 5.75 111.21 ± 9.46 113.03 ± 2.66 111.79 ± 7.78 127.69 ± 7.38 125.00 ± 6.63 129.44 ± 7.09 125.18 ± 23.82 140.01 ± 33.36

4 3 8 7 5 4 3 8 8 5 8 8 8 8 8 8 8 8 8 8

4.5 ± 0.22 4.1 ± 0.33 4.7 ± 0.26 4.8 ± 0.49 4.6 ± 0.41 5.5 ± 0.60 5.2 ± 0.37 5.3 ± 0.32 5.4 ± 0.41 5.4 ± 0.35 4.5 ± 0.32 4.4 ± 0.37 4.1 ± 0.40 4.4 ± 0.24 4.5 ± 0.25 4.8 ± 0.12 4.6 ± 0.13 4.9 ± 0.28 4.7 ± 0.25 4.9 ± 0.19

4 3 8 8 5 4 3 8 8 5 8 8 8 8 8 8 8 8 8 8

28

P.J. Bushnell et al. / Neurotoxicology and Teratology 49 (2015) 19–30

Table 4B Physiological assessments of offspring from the E15 experiment. Values are means ± SD for each endpoint. The Air control group was exposed to clean air in an inhalation chamber during gestation; the other groups were exposed to 3000, 6000, or 9000 ppm of E15 vapor during gestation (GD 9–GD 20). Age

Body weight

Blood pressure

Glucose

Glycated HbA1c

Fuel

Sex

Months

Group

g

n

mm Hg

n

mg/dL

n

%

n

E15

M

3

E15

M

6

E15

F

3

E15

F

6

Air 3000 6000 9000 Air 3000 6000 9000 Air 3000 6000 9000 Air 3000 6000 9000

535.0 ± 41.03 518.0 ± 31.48 552.6 ± 49.19 519.2 ± 40.12 681.1 ± 48.23 761.8 ± 45.53 715.0 ± 79.56 674.6 ± 57.93 308.8 ± 14.82 285.5 ± 20.97 288.9 ± 19.85 314.0 ± 29.80 365.1 ± 17.39 336.0 ± 21.49 351.8 ± 32.57 386.1 ± 46.61

8 8 7 8 7 8 7 8 8 8 8 8 8 8 8 8

147 ± 13.80 138 ± 8.06 138 ± 11.55 144 ± 15.08 118 ± 14.69 141 ± 15.81 142 ± 23.43 137 ± 16.42 145 ± 15.52 148 ± 10.89 153 ± 15.59 142 ± 9.97 132 ± 10.27 135 ± 14.02 158 ± 15.36 143 ± 11.04

8 8 7 8 7 8 7 8 8 8 8 8 8 8 8 8

146.28 ± 10.89 147.49 ± 24.82 148.04 ± 13.66 151.44 ± 11.43 137.84 ± 7.24 138.43 ± 10.79 133.97 ± 10.00 135.18 ± 12.84 133.46 ± 9.83 148.05 ± 9.43 137.29 ± 9.47 133.42 ± 14.04 133.81 ± 11.25 126.62 ± 9.23 141.64 ± 9.60 134.71 ± 12.97

7 8 7 8 7 8 7 8 8 8 8 8 8 8 8 8

4.5 ± 0.33 5.1 ± 0.21 5.0 ± 0.22 4.9 ± 0.15 4.6 ± 0.44 4.4 ± 0.25 4.6 ± 0.12 4.7 ± 0.42 4.5 ± 0.15 4.4 ± 0.26 4.5 ± 0.21 4.5 ± 0.33 4.2 ± 0.23 4.4 ± 0.34 4.2 ± 0.25 4.7 ± 0.28

7 8 7 8 7 8 7 8 8 8 8 8 8 8 8 8

of 5000, 10,000, and 21,000 ppm, respectively. Consistent with the lower airborne ethanol concentrations in the vapors in the present experiments (Table 1), estimated maternal and fetal BECs from exposure to these vapors (containing up to 7000 ppm ethanol in the 9000-ppm group of E85) did not exceed 5 mg/dL (Martin SOT poster 2014). Preliminary estimates of concentrations of other individual hydrocarbons in the tissues of the pregnant rats and their fetuses do not exceed 1 mg/dL (Martin et al., 2014), due to the very low concentrations of the vast majority of them in the vapor and the very low uptake and retention of prominent constituents like butane and pentane. 4.3.2. Unconditioned behavior Whereas the ontogeny of motor activity in the offspring of the vapor-exposed rats did not differ significantly from controls, vertical activity counts were slightly elevated on PND 63 in the animals with gestational exposure to E85. This effect did not increase with concentration and was not evident in the E0 and E15 experiments (Fig. 3). A similar pattern was seen in horizontal (but not vertical) activity counts at this age in the previous experiment with ethanol (Beasley et al., 2014). The biological significance of this apparent coincidence remains uncertain, as does the differential relative sensitivity of vertical vs horizontal activity counts in response to E85 and ethanol. Results of FOB tests were negative for the autonomic, neuromuscular, excitability, activity, and general health domains. Changes in the

sensorimotor domain were evident in male offspring after gestational exposure to E0, but were not incrementally altered by concentration. They did not occur after exposure to the other vapors, nor were they apparent after exposure to ethanol, which generated a few changes in the neuromuscular domain only (Beasley et al., 2014). It is possible that ethanol and the hydrocarbons in E0 affect different functional domains, but in both cases the magnitude of the effect was small and not concentration-related. In summary, the few effects seen with these behavioral assessments were relatively small, and the few effects even of high concentrations of inhaled fuel vapors, with or without ethanol, suggest a low biological significance of these observations. 4.3.3. Immune function IgM titers in the female offspring of rats exposed to 3000 ppm E0 were significantly lower than control values. However, this single change, in one sex of the lowest exposure group of one experiment, provides scant evidence for effects of these vapors on immune function, and is more likely to represent a statistical false positive event. Impaired immune function was one of the few positive findings among the extensive health assessments of oxygenated fuels sponsored by the American Petroleum Institute. In those studies, the antibodyforming cell response to sheep erythrocyte protein, a T-dependent antigen, was reduced in adult rats after 20 days of exposure to vapors of

Table 4C Physiological assessments of offspring from the E85 experiment. Values are means ± SD for each endpoint. The Air control group was exposed to clean air in an inhalation chamber during gestation; the other groups were exposed to 3000, 6000, or 9000 ppm of E85 vapor during gestation (GD 9–GD 20). Age

Body weight

Blood pressure

Glucose

Glycated HbA1c

Fuel

Sex

Months

Group

g

n

mm Hg

n

mg/dL

n

%

n

E85

M

3

E85

M

8.5

E85

F

3

E85

F

8.5

Air 3000 6000 9000 Air 3000 6000 9000 Air 3000 6000 9000 Air 3000 6000 9000

500.3 ± 31.23 501.7 ± 32.39 529.4 ± 38.15 501.7 ± 33.28 720.6 ± 56.24 734.4 ± 64.63 750.9 ± 79.69 701.1 ± 58.64 298.2 ± 12.37 305.2 ± 24.45 310.0 ± 25.88 301.9 ± 20.65 378.1 ± 33.38 392.0 ± 48.00 392.3 ± 45.13 378.5 ± 35.12

8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8

151 ± 17.44 143 ± 26.99 158 ± 13.90 147 ± 19.19 133 ± 9.01 131 ± 10.12 142 ± 12.09 131 ± 2.32 135 ± 9.21 142 ± 10.08 144 ± 10.44 142 ± 11.17 137 ± 12.35 143 ± 8.24 138 ± 13.47 138 ± 12.00

8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8

121.42 ± 9.62 135.52 ± 25.80 125.16 ± 14.08 138.16 ± 23.59 154.76 ± 8.31 165.80 ± 7.74 172.04 ± 25.67 153.10 ± 13.45 125.14 ± 12.03 114.98 ± 5.09 129.31 ± 7.82 123.92 ± 10.80 149.98 ± 12.83 159.93 ± 16.56 171.59 ± 32.24 163.15 ± 29.82

8 8 7 8 8 8 8 8 8 8 8 8 8 8 8 8

4.7 ± 0.17 4.6 ± 0.28 4.5 ± 0.27 4.7 ± 0.16 4.8 ± 0.19 5.0 ± 0.18 4.6 ± 0.34 4.9 ± 0.18 4.5 ± 0.18 4.5 ± 0.21 4.3 ± 0.15 4.4 ± 0.16 4.6 ± 0.22 4.5 ± 0.21 4.6 ± 0.24 4.6 ± 0.20

8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8

P.J. Bushnell et al. / Neurotoxicology and Teratology 49 (2015) 19–30

gasoline plus 10% ethanol or di-isopropyl ether (White et al., 2014). The lack of effect in this study suggests that changes in immune function require direct exposure to the animal, and that exposure during gestation is insufficient. 4.3.4. Blood pressure Systolic blood pressure was elevated at 3 months of age in male offspring of rats exposed to ethanol at 10,000 ppm but not at 5000 or 21,000 ppm (Beasley et al., 2014). In the present experiments, systolic blood pressure tended to be higher in females than in males and to decrease with age (Tables 4A, 4B, 4C). Statistical interactions involving vapor concentration were due to differences in blood pressure between the control groups, either across sex (E0) or age (E15). The reasons why sex and age differences might be attenuated by prenatal exposure to the fuel vapors are not clear. Thus no consistent pattern of effect of ethanol or the fuel vapors is evident in measures of systolic blood pressure. To our knowledge, blood pressure has not been evaluated previously in animal models of gasoline vapor exposure, although the association of hypertension with chronic ethanol consumption is well known (Marchi et al., 2014), and prenatal oral ethanol exposure yielding a BEC of 100 mg/dL in pregnant rats has been shown to elevate arterial blood pressure by about 10% in male and female offspring in association with reduced nephron number renal insufficiency (Gray et al., 2010). It may be that detection of changes in blood pressure of this magnitude requires measurement techniques more precise than the tail-cuff plethysmography used here. 4.3.5. Glucose homeostasis Neither glucose nor glycated hemoglobin, a marker of glucose concentration integrated over a time span corresponding to the half-life of the erythrocyte, was affected by prenatal exposure to any of the fuel vapors, nor was it affected by inhaled ethanol (Beasley et al., 2014). These results do not confirm evidence that developmental exposure to ethanol affected glucose metabolism (Chen and Nyomba, 2003) or insulin regulation (Chen and Nyomba, 2004). 4.4. Context The concentrations of fuel vapors used in these experiments were selected for purposes of determining concentration–effect relationships and not to reflect ambient concentrations or typical exposures in residences or workplaces. Most exposure studies report values of airborne VOCs in units of μg/m3; thus, assuming an average molecular weight of the vapor condensates to be about 60 g/mol, the present concentration values of 1000, 3000, and 9000 ppm correspond to values of about 2.7 × 106, 8.0 × 106, and 2.4 × 107 μg/m3. By comparison, mean total concentrations of 34 volatile organic compounds in enclosed garages were measured at 635.1 μg/m3, and outdoors at 5.18 μg/m3 (Batterman et al., 2006). Similarly, total hydrocarbon concentrations of ~ 30 μg/m3 were measured in the ambient air in Toronto, Canada (Su et al., 2010), and concentrations along major roads in Nanjing, China averaged 44.5 μg/m3 (Wang and Zhao, 2008). These more typical public exposure levels are 4 to 6 orders of magnitude below the experimental concentrations used here. In contrast, occupational exposure limits for hydrocarbon fuels are much higher (100–500 ppm for 8-h exposures), and up to 1800 ppm for transient (15-min) episodes, and some extreme jobs (e.g., cleaning fuel tank filters in aircraft) can result in exposures exceeding the concentrations used in this study (Ritchie et al., 2001). Clearly, public exposures to these concentrations will be extremely rare events. 5. Conclusions No systematically concentration-related changes in neurobehavioral development, immune function or glucose homeostasis were observed in the offspring of pregnant rats exposed to vapors of three gasoline–

29

ethanol fuel blends at concentrations up to 9000 ppm. Pregnant rats exposed to 9000 ppm of E0 and E85, but not E15, consumed less food during 12 days of exposure while maintaining a normal trajectory of weight gain. An increase in vertical motor activity and small changes in a few sensorimotor responses were observed, but were not monotonically related to exposure concentration. These results suggest a LOEL of 3000 ppm of E85 for the vertical activity counts, 9000 ppm of E0 and E85 for the effects on maternal food consumption, and a NOEL of 9000 ppm for the other endpoints assessed here. The ethanol content of these fuel vapors did not systematically alter the outcomes of the assessments. The concentrations of the vapors to which the rats were exposed exceed by 4–6 orders of magnitude typical exposure levels encountered by the public. Transparency document The Transparency document associated with this article can be found, in the online version. Acknowledgments We thank Drs. William Boyes, Linda Roberts and John Cowden for reviews of the manuscript, and Dr. Greg Travlos of NIEHS for advice regarding clinical chemistry assays. We also thank the animal care staff of Alpha BioServices, especially Kay Rigsbee, for her meticulous care of the mothers and offspring while in the animal facility, and Judy Richards for the clinical chemistry analyses. We also thank Pam Phillips and Danielle Lyke for assisting with the ethanol exposures, Wendy Oshiro for her assistance with the exposures, redistribution, and weaning of the pups, and Allen Ledbetter for assistance with blood pressure measurements. Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.ntt.2015.02.004. References Abel EL. Fetal alcohol syndrome: a cautionary note. Curr Pharm Des 2006;12(12):1521–9. Agnish ND, Keller KA. The rationale for culling of rodent litters. Fundam Appl Toxicol 1997;38(1):2–6. Batterman S, Hatzvasilis G, Jia CR. Concentrations and emissions of gasoline and other vapors from residential vehicle garages. Atmos Environ 2006;40(10):1828–44. Beasley TE, Evansky PA, Martin SA, McDaniel KL, Moser VC, Luebke RW, et al. Toxicological outcomes in rats exposed to inhaled ethanol during gestation. Neurotoxicol Teratol 2014;45C:59–69. Becker HC. Effects of ethanol on the central nervous system: fetal damage — neurobehavioral effects. In: Deitrich RA, Erwin VG, editors. Pharmacological effects of ethanol on the nervous system. Boca Raton, FL: CRC Press; 1996. p. 409–42. Benevenga NJ, Kaiser M, Clagett-Dame M. Development of the rat loft. Tech Talk 2005; 10(4):3. Blunden J, Aneja VP, Lonneman WA. Characterization of non-methane volatile organic compounds at swine facilities in eastern North Carolina. Atmos Environ 2005; 39(36):6707–18. Boyes WK, Degn LL, Martin SA, Lyke DF, Hamm CW, Herr DW. Neurophysiological assessment of auditory, peripheral nerve, somatosensory, and visual system functions after developmental exposure to ethanol vapors. Neurotoxicol Teratol 2014;43:1–10. Chen L, Nyomba BL. Effects of prenatal alcohol exposure on glucose tolerance in the rat offspring. Metabolism 2003;52(4):454–62. Chen L, Nyomba BL. Whole body insulin resistance in rat offspring of mothers consuming alcohol during pregnancy or lactation: comparing prenatal and postnatal exposure. J Appl Physiol (1985) 2004;96(1):167–72. Clark CR, Schreiner CA, Parker CM, Gray TM, Hoffman GM. Health assessment of gasoline and fuel oxygenate vapors: subchronic inhalation toxicity. Regul Toxicol Pharmacol 2014;70(2 Suppl):S18–28. http://dx.doi.org/10.1016/j.yrtph.2014.07.003. DeWitt JC, Copeland CB, Luebke RW. Developmental exposure to 1.0 or 2.5 mg/kg of dibutyltin dichloride does not impair immune function in Sprague–Dawley rats. J Immunotoxicol 2006;3(4):245–52. Driscoll CD, Streissguth AP, Riley EP. Prenatal alcohol exposure: comparability of effects in humans and animal models. Neurotoxicol Teratol 1990;12(3):231–7. EPA US. Biofuels and the environment: the First Triennial Report to Congress (2011 Final Report). Washington, DC; 2011.

30

P.J. Bushnell et al. / Neurotoxicology and Teratology 49 (2015) 19–30

Gray SP, Denton KM, Cullen-McEwen L, Bertram JF, Moritz KM. Prenatal exposure to alcohol reduces nephron number and raises blood pressure in progeny. J Am Soc Nephrol 2010;21(11):1891–902. Gray TM, Steup D, Roberts LG, O'Callaghan JP, Hoffman G, Schreiner CA, et al. Health assessment of gasoline and fuel oxygenate vapors: reproductive toxicity assessment. Regul Toxicol Pharmacol 2014;70(2 Suppl):S48–57. http://dx.doi.org/10.1016/j. yrtph.2014.04.014. Hard GC, Rodgers IS, Baetcke KP, Richards WL, McGaughy RE, Valcovic LR. Hazard evaluation of chemicals that cause accumulation of alpha 2u-globulin, hyaline droplet nephropathy, and tubule neoplasia in the kidneys of male rats. Environ Health Perspect 1993;99:313–49. Henley M, Letinski DJ, Carr J, Caro ML, Daughtrey W, White R. Health assessment of gasoline and fuel oxygenate vapors: generation and characterization of test materials. Regul Toxicol Pharmacol 2014;70(2 Suppl):S13–7. http://dx.doi.org/10.1016/j.yrtph. 2014.05.012. Holson RR, Freshwater L, Maurissen JP, Moser VC, Phang W. Statistical issues and techniques appropriate for developmental neurotoxicity testing: a report from the ILSI Research Foundation/Risk Science Institute expert working group on neurodevelopmental endpoints. Neurotoxicol Teratol 2008;30(4):326–48. Jerrells TR, Weinberg J. Influence of ethanol consumption on immune competence of adult animals exposed to ethanol in utero. Alcohol Clin Exp Res 1998;22(2):391–400. Kalev-Zylinska ML, During MJ. Paradoxical facilitatory effect of low-dose alcohol consumption on memory mediated by NMDA receptors. J Neurosci 2007;27(39): 10456–67. Kane CJ, Phelan KD, Drew PD. Neuroimmune mechanisms in fetal alcohol spectrum disorder. Dev Neurobiol 2012;72(10):1302–16. Lough GC, Schauer JJ, Lonneman WA, Allen MK. Summer and winter nonmethane hydrocarbon emissions from on-road motor vehicles in the Midwestern United States. J Air Waste Manage Assoc 2005;55(5):629–46. Marchi KC, Muniz JJ, Tirapelli CR. Hypertension and chronic ethanol consumption: what do we know after a century of study? World J Cardio 2014;6(5):283–94. Martin SA, Oshiro WM, Evansky PA, Ford J, Degn LL, El-Masri H, et al. PBPK models for gasoline–ethanol biofuels in adult and pregnant rats. Phoenix, AZ: Society of Toxicology; 2014. McDaniel KL, Moser VC. Utility of a neurobehavioral screening battery for differentiating the effects of two pyrethroids, permethrin and cypermethrin. Neurotoxicol Teratol 1993;15(2):71–83. McKee RH, Wong ZA, Schmitt S, Beatty P, Swanson M, Schreiner CA, et al. The reproductive and developmental toxicity of high flash aromatic naphtha. Toxicol Ind Health 1990;6(3–4):441–60. McKee RH, Trimmer GW, Whitman FT, Nessel CS, Mackerer CR, Hagemann R, et al. Assessment in rats of the reproductive toxicity of gasoline from a gasoline vapor recovery unit. Reprod Toxicol 2000;14(4):337–53. Miller RR, Letts RL, Potts WJ, McKenna MJ. Improved methodology for generating controlled test atmospheres. Am Ind Hyg Assoc J 1980;41:844–6. Mizisin AP, Kalichman MW, Garrett RS, Dines KC. Tactile hyperesthesia, altered epidermal innervation and plantar nerve injury in the hindfeet of rats housed on wire grates. Brain Res 1998;788(1–2):13–9. Moser VC, Cheek BM, MacPhail RC. A multidisciplinary approach to toxicological screening: III. Neurobehavioral toxicity. J Toxicol Environ Health 1995;45(2):173–210. Moser VC, Tilson HA, MacPhail RC, Becking GC, Cuomo V, Frantik E, et al. The IPCS Collaborative Study on Neurobehavioral Screening Methods: II. Protocol design and testing procedures. Neurotoxicology 1997;18(4):929–38.

Mukherjee RAS, Hollins S, Abou-Saleh MT, Turk J. Low level alcohol consumption and the fetus. BMJ 2005;330(7488):375–6. Oshiro WM, Beasley TE, McDaniel KL, Taylor MM, Evansky P, Moser VC, et al. Selective cognitive deficits in adult rats after prenatal exposure to inhaled ethanol. Neurotoxicol Teratol 2014;45C:44–58. Reiter LW. Chemical exposures and animal activity: utility of the figure-eight maze. Dev Toxicol Environ Sci 1983;11:73–84. Ren J, Wold LE, Natavio M, Ren BH, Hannigan JH, Brown RA. Influence of prenatal alcohol exposure on myocardial contractile function in adult rat hearts: role of intracellular calcium and apoptosis. Alcohol Alcohol 2002;37(1):30–7. Riley EP, McGee CL. Fetal alcohol spectrum disorders: an overview with emphasis on changes in brain and behavior. Exp Biol Med (Maywood) 2005;230(6):357–65. Ritchie GD, Still KR, Alexander WK, Nordholm AF, Wilson CL, Rossi J, et al. A review of the neurotoxicity risk of selected hydrocarbon fuels. J Toxicol Environ Health B Crit Rev 2001;4(3):223–312. Roberts L, White R, Bui Q, Daughtrey W, Koschier F, Rodney S, et al. Developmental toxicity evaluation of unleaded gasoline vapor in the rat. Reprod Toxicol 2001; 15(5):487–94. Roberts LG, Gray TM, Marr MC, Tyl RW, Trimmer GW, Hoffman GM, et al. Health assessment of gasoline and fuel oxygenate vapors: developmental toxicity in mice. Regul Toxicol Pharmacol 2014;70(2 Suppl):S58–68. http://dx.doi.org/10.1016/j.yrtph. 2014.06.011. Savage DD, Becher M, de la Torre AJ, Sutherland RJ. Dose-dependent effects of prenatal ethanol exposure on synaptic plasticity and learning in mature offspring. Alcohol Clin Exp Res 2002;26(11):1752–8. Schreiner C, Bui Q, Breglia R, Burnett D, Koschier F, Podhasky P, et al. Toxicity evaluation of petroleum blending streams: reproductive and developmental effects of light catalytic reformed naphtha distillate in rats. J Toxicol Environ Health A 2000;60(3): 169–84. Shankar K, Ronis MJJ, Badger TM. Effects of pregnancy and nutritional status on alcohol metabolism. Alcohol Res Health 2007;30(1):55. Su JG, Jerrett M, Beckerman B, Verma D, Arain MA, Kanaroglou P, et al. A land use regression model for predicting ambient volatile organic compound concentrations in Toronto, Canada. Atmos Environ 2010;44(29):3529–37. Sugano Y. Heat balance of rats acclimated to diurnal 2-hour feeding. Physiol Behav 1983; 30(2):289–93. Wang P, Zhao W. Assessment of ambient volatile organic compounds (VOCs) near major roads in urban Nanjing, China. Atmos Res 2008;89(3):289–97. White Jr KL, Peachee VL, Armstrong SR, Twerdok LE, Clark CR, Schreiner CA. Health assessment of gasoline and fuel oxygenate vapors: immunotoxicity evaluation. Regul Toxicol Pharmacol 2014;70(2 Suppl):S43–7. http://dx.doi.org/10.1016/j.yrtph.2014. 04.010. Zajac CS, Abel EL. Animal models of prenatal alcohol exposure. Int J Epidemiol 1992; 21(Suppl. 1):S24–32. Zhang X, Sliwowska JH, Weinberg J. Prenatal alcohol exposure and fetal programming: effects on neuroendocrine and immune function. Exp Biol Med (Maywood) 2005; 230(6):376–88. Zielinska B, Fujita EM, Sagebiel JC, Campbell DE. Section 211(B) tier 2 high end exposure study of conventional and oxygenated gasoline. A. P. Institute; 2007. p. 140.