Weanling rats developmentally exposed to polychlorinated biphenyls (PCBs) exhibit aberrant dendritic growth and neurobehavioral deficits. PCBs promote dendritic arborization in cultured neurons via sensitization of the ryanodine receptor (RyR). However, whether RyR function influences PCB developmental neurotoxicity in vivo has yet to be tested. The objective of this study was to determine whether a gain-of-function human RYR mutation affects neurodevelopment and/or alters PCB developmental neurotoxicity in vivo. To address this, mice heterozygous for the R163C-RYR1 mutation (HET) and congenic wild-type (WT) littermates were exposed to vehicle or PCB 95 at 0.1, 1.0, or 6.0 mg/kg/day via the maternal diet throughout gestation and lactation.  PCB 95 at 1.0 mg/kg/day caused performance deficits in the Morris water maze in WT but not HET weanlings. Gas chromatography confirmed dose-dependent increases in PCB tissue burdens. Serum levels of progesterone, estradiol, and thyroid hormone (TH), and brain transcript levels of TH-responsive genes were unaffected by genotype, dose, or training. However, serum cortisol was elevated in all females, independent of genotype or dose. Sholl analyses of Golgi-stained CA1 pyramidal neurons indicated that dendritic arborization was significantly increased in WT weanlings by either expression of the R163C-RYR1 mutation or developmental exposure to PCB 95 at 1 mg/kg/day. PCB 95 significantly reduced activity-induced dendritic arborization in WT weanlings, and this effect was significantly attenuated in HET weanlings. These findings support a role for RyR-dependent mechanisms in normal dendritic growth and in PCB developmental neurotoxicity in vivo and identify a specific gene-environment interaction that modifies risk for adverse neurodevelopmental outcomes.

Animal care and husbandry: Animals were treated humanely and with regard to alleviation of suffering according to protocols approved by the Institutional Animal Care and Use Committee of the University of California, Davis.  Weanling animals used for these studies were part of a larger study examining the neurodevelopmental effects of PCBs. Male C57BL/6 mice heterozygous for the R163C-RYR1 mutation (HET) were obtained from an in-house breeding colony, and the heterozygous genotype was confirmed as previously described (Yang et al., 2006a). Congenic WT C57BL/6 dams (8 weeks old) were purchased from Charles Rivers Laboratories (Wilmington, MA). Husbandry practices were performed as previously described (Kania-Korwel et al., 2012). All animal experiments complied with ARRIVE guidelines.

PCB dosing: To examine the neurological effect of developmental exposure to PCB 95 in vivo, female mice were dosed via oral ingestion of vehicle, 0.1, 1.0, or 6.0 mg/kg/day PCB 95 (2,2΄,3,5΄,6-pentachlorobiphenyl; 99.7% purity) (AccuStandard; New Haven, CT). PCB 95 was dissolved in peanut oil (Spectrum Organic Products, LLC, Melville, NY), which was then homogeneously mixed into peanut butter (Trader Joe’s, Monrovia, CA). After one week of acclimatization to peanut butter, dams were randomly assigned to a dose group, and dosing was initiated. Each dam was weighed, and the amount of PCB mixture was adjusted for body weight to ensure the appropriate dose was consumed. Two weeks after dosing was initiated, dams were trio-mated with R163C-RYR1 heterozygous males, and dosing continued daily throughout gestation and lactation up to postnatal day (P) 21 when pups were weaned.  At P2, litters were randomly culled or cross-fostered within treatment groups to 7 to 8 pups per litter. When possible, the numbers of males and females were equalized within a litter. Individual pups (a maximum of 2 pups per genotype per sex from at least 8 litters) served as the experimental unit for these experiments.

Morris water maze: Weanling mice (1 male and 1 female, 1 WT and 1 HET weanling per litter) were trained in the MWM beginning on P23-25 as previously described (Barnhart et al., 2015). Untrained controls included age-, sex-, and genotype-matched littermates that were not trained in the MWM but were otherwise similarly handled and housed. Briefly, MWM-trained weanlings were trained for 6 days in blocks of 2 trials/day with an inter-trial interval of approximately 30 minutes. A probe test, in which the platform was removed from the pool, was administered 24 h after day 6 of training. A visual cue test, in which the platform was made obviously visible above the surface of the water was conducted approximately 30 min after the probe test. All swim paths were recorded and analyzed using SMART video tracking and analysis software (version 2.5, Panlab, Barcelona, Spain). Spatial learning was assessed as changes from the beginning to the end of the training period in: (i) the time to locate the platform [delta(escape latency)], and (ii) the percentage of the swim path spent in the target quadrant that contained the platform [delta(% path length)].  Spatial memory was assessed in the probe test by determining the percentage of swim time (% time) spent in the target quadrant and by the number of times the weanlings crossed the former location of the platform when the platform was removed (platform crossings). Motivation and sensorimotor function were assessed by determining escape latency and mean swim velocity in the visual cue test (Vorhees and Williams, 2006; D'Hooge and De Deyn, 2001).

Morphometric analysis of dendritic arborization: Brains were collected from pups 24 h after completion of MWM testing. MWM-trained pups and their untrained littermates were euthanized by cervical dislocation. Brains were carefully excised from the skull, bisected into hemispheres that were immediately immersed in Golgi solution according to the manufacturer’s instructions (FD NeuroTechnologies, Columbia, MD) (Milatovic et al., 2010). After immersion in Solution C, brains were transferred to 10% sucrose in phosphate-buffered saline (PBS) for 3-4 h at 4°C and then held in 30% sucrose in PBS at 4°C until sectioning. Prior to sectioning, brains were transferred to 70% ethanol in distilled water for 4 h at 4°C. Brains were cut into 100 µm coronal sections in chilled 70% ethanol using a vibratome (VT-1000, Leica, Solms, Germany). Sections were mounted on gelatin-subbed slides and stained according to the manufacturer’s instructions (FD NeuroTechnologies). Brightfield image stacks of CA1 hippocampal pyramidal neurons were captured using an IX-81 inverted microscope (Olympus, Shinjuku, Japan) and MetaMorph Image Analysis Software (version 7.1, Molecular Devices, San Jose, CA). Criteria for selection of CA1 neurons for Golgi analyses have been described previously (Lein et al., 2007). Basilar dendritic arbors of selected neurons were traced using NeuroLucida (version 11, MBF Bioscience, Williston, VT), and arbor complexity was quantified by Sholl analysis using NeuroLucida Explorer (version 11, MBF Bioscience), wherein 10 μm concentric circles (shells) were centered on the soma and dendritic branches that intersected with individual shells were counted (Lein et al., 2007). Six to seven neurons were quantified in each of 5 brains from each experimental group for a total of at least 30 neurons per experimental condition.

Quantification of PCB levels in tissues: PCB 95 and five of its hydroxylated metabolites (i.e., 3-103, 4-95, 4’-95, 5-95, and 4,5-95) were quantified in whole blood, brain, adipose tissue, and liver by gas chromatography with electron capture detection as previously described (Kania-Korwel et al., 2012) and were adjusted for tissue wet weights to facilitate the comparison with earlier studies. To obtain sufficient signal, samples of the same sex, genotype, training status, and dose group were pooled within cohorts. A description of the quality assurance/quality control, including limits of detection, is provided in Table S1.

Thyroid hormone (TH) and steroid hormone (SH) assays: Whole blood (50-200 µL) was collected 24 h after MWM testing, and blood was allowed to clot for a minimum of 30 min. After clotting, samples were spun down at 2500 x g for 10 min, and then the serum was collected and frozen at -20°C until extracted. Extraction was performed according to the manufacturer’s acetonitrile precipitation protocol with the modification that volumes were reduced by a factor of 5 to 15 and samples were reconstituted in 60 µL of assay buffer to accommodate the small amount of blood collected. Bead sets were analyzed using a flow-based Luminex® 100 suspension array system (Bio-Plex 200, Bio-Rad Laboratories Inc., Hercules, CA). Circulating levels of TH (triiodothyronine, or T3, and thyroxine, or T4) and SH (progesterone, estradiol, and cortisol) concentrations in samples were calculated using Bio-Plex Manager software (version 6.0, Bio-Rad Laboratories Inc.) based on a standard curve derived from known reference hormone concentrations supplied by the manufacturer. Final concentrations are expressed as (ng hormone/mL)/mL extracted serum.

Quantitative PCR (qPCR): Transcript levels of the TH-responsive genes myelin basic protein (MBP) and neurogranin (RC3) and the reference genes hypoxanthine-guanine phosphoribosyltransferase (HPRT), phosphoglycerate kinase (PGK1), and peptidylprolyl isomerase A (PPIA) were measured in hippocampi, cortices, and cerebella as previously described (Wu et al., 2014).

Preparation of brain cortical homogenate for equilibrium [^3^H]Ry binding assays: Cortices from weanlings of all experimental conditions were isolated and kept in ice-cold buffer containing 300 mM sucrose, 5 mM imidazole, 100 μM phenylmethanesulfonylfluoride, 10 μg/mL leupeptin, 0.5 mM sodium orthovanadate, 10 mM sodium fluoride, 2 mM β-glycerophosphate, and 5 mM sodium pyrophosphate. The tissue was homogenized with an electric drill in a Teflon glass homogenizer 3 times for 30 min each, followed by centrifugation at 3500 x g for 3 min. The supernatant was then centrifuged at 110,000 x g for 1 h at 4°C. The pellet was resuspended in 300 mM sucrose and 20 mM HEPES at pH 7.4, flash-frozen in liquid nitrogen, and stored at -80°C until use.

Equilibrium [^3^H]Ry binding assays: Specific binding of [^3^H]Ry to brain cortex homogenate was performed as previously described (Wong and Pessah, 1997) with some modifications. Briefly, protein (500 µg/ml) was incubated for 3 h at 37°C in binding buffer containing 250 mM KCl, 15 mM NaCl, 20 mM HEPES at pH 7.4, 5nM [^3^H]Ry (specific activity of 86.6 µCi/mmole; Perkins Elmer, MA), and 200 µM free calcium. Each experiment was performed with 5 data points. Non-specific binding of [^3^H]Ry was determined by samples with 1000-fold excessive unlabeled ryanodine.

Statistics: All data besides qPCR data were statistically analyzed in SPSS (version 24, IBM, Armonk, NY). Data were assessed for normality, homogeneity of variance, and sphericity with the Shapiro-Wilk test, Levene’s test, and Mauchly’s test. Data that did not have a normal distribution were log-transformed, and the F statistics from data that violated the assumption of sphericity were corrected using the Greenhouse-Geisser correction. Linear mixed effects models were fit to toxicity, behavior, gas chromatography, TH/SH, and [^3^H]Ry binding data with dose, sex, genotype, training day, or training status as the fixed factors and cohort and individual pup IDs as random factors. Pairwise comparisons were used to identify significant differences between experimental groups (effects of dose, sex, genotype, or training), and significant differences between specific experimental conditions (effects of genotype, sex, or training within or between dose groups), with differences considered statistically significant at p < 0.05. Golgi stain data were analyzed using the Kruskal-Wallis H test; Mann-Whitney U tests were used to identify significant differences between experimental groups. For qPCR data, C_q values for the target genes were normalized to the geometric mean of C_q values for the reference genes. Relative expression ratios between experimental groups were statistically analyzed using REST2009 software as previously described (Stamou et al., 2014).