Effects of combined prenatal exposure to air pollution and maternal stress on social behavior and oxytocin and vasopressin systems in male and female mice
Maura C. Stoehr, Elise M. Martin, Joy T. Babalola, Jason Xue, Matthew J. Kern, Niki Y. Li, Madeline F. Winters, Sarvin Bhagwagar, Caroline J. Smith
TL;DR
Exposure to air pollution and maternal stress during pregnancy affects social behavior in mice, with changes in brain receptors linked to social behavior.
Contribution
This study reveals sex-specific effects of combined prenatal exposure on social behavior and V1aR expression in mice.
Findings
DEP/MS exposure reduced social interaction time in male mice and social novelty preference in females.
Avpr1a mRNA was significantly increased in the nucleus accumbens in both sexes after DEP/MS exposure.
No changes in OXT or AVP cell number or fiber density were observed in hypothalamic regions.
Abstract
Prenatal exposures to air pollution and maternal psychosocial stress are each associated with increased risk of neurodevelopmental disorders, including autism spectrum disorder (ASD), and epidemiological work suggests that concurrent exposure to these risk factors may be particularly harmful. This is important given that the same populations often bear the brunt of both toxicant and psychosocial stress burdens. Social impairments are a defining symptom in ASD. Previous work modeling combined prenatal exposure to diesel exhaust particles (DEPs) and maternal stress (MS) in rodents has found male‐biased social deficits in offspring, as well as changes to neuroimmune processes and the gut microbiome. However, the precise neural circuits on which these exposures converge to impact social behavior are unclear. Oxytocin (OXT) and vasopressin (AVP) are neuropeptides critical to the regulation…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
FIGURE 1
FIGURE 2
FIGURE 3
FIGURE 4
FIGURE 5| Outcome | Treatment | Embryonic day | Interaction |
|---|---|---|---|
| Maternal body weight (g) |
|
|
|
| Behavior | Treatment | Stimulus/chamber | Interaction |
|---|---|---|---|
|
| |||
| Males | |||
| Chamber time (s) |
|
|
|
| Post hoc tests: light versus dark | |||
| Females | |||
| Chamber time (s) |
|
|
|
| Post hoc tests: VEH/CON versus DEP/MS: light side | |||
| Post hoc tests: VEH/CON: light versus dark | |||
| Post hoc tests: DEP/MS: light versus dark | |||
|
| |||
| Males | |||
| Stimulus interaction (s) |
|
|
|
| Stimulus nose‐poking (s) |
|
|
|
| Stimulus climbing (s) |
|
|
|
| Chamber time (s) |
|
|
|
| Post hoc tests: (chamber time): animal versus middle | |||
| Females | |||
| Stimulus interaction (s) |
|
|
|
| Stimulus nose‐poking (s) |
|
|
|
| Stimulus climbing (s) |
|
|
|
| Chamber time (s) |
|
|
|
| Post hoc tests: (chamber time): animal versus middle | |||
|
| |||
| Males | |||
| Stimulus interaction (s) |
|
|
|
| Stimulus nose‐poking (s) |
|
|
|
| Stimulus climbing (s) |
|
|
|
| Chamber time (s) |
|
|
|
| Post hoc tests: (chamber time): novel versus middle | |||
| Females | |||
| Stimulus interaction (s) |
|
|
|
| Stimulus nose‐poking (s) |
|
|
|
| Stimulus climbing (s) |
|
|
|
| Chamber time (s) |
|
|
|
| Post hoc tests (stimulus interaction): VEH/CON versus DEP/MS: novel | |||
| Post hoc test results (stimulus nose‐poking): VEH/CON versus DEP/MS: novel | |||
| Post hoc tests (chamber time): VEH/CON versus DEP/MS: novel C. | |||
| Post hoc tests (chamber time): VEH/CON: novel versus middle | |||
| Post hoc tests (chamber time): DEP/MS: novel versus middle | |||
| Cell type | Treatment | Sex | Interaction |
|---|---|---|---|
| Oxytocin cell bodies | |||
| Level 1 PVN |
|
|
|
| Level 2 PVN |
|
|
|
| Vasopressin cell bodies | |||
| Level 1 PVN |
|
|
|
| Level 2 PVN |
|
|
|
| Fiber type | Treatment | Sex | Interaction |
|---|---|---|---|
| Oxytocin fibers | |||
| LH fiber density |
|
|
|
| LH fiber % area covered |
|
|
|
| AH fiber density |
|
|
|
| AH fiber % area covered |
|
|
|
| MPOA fiber density |
|
|
|
| MPOA fiber % area covered |
|
|
|
| Vasopressin fibers | |||
| LH fiber density |
|
|
|
| LH fiber % area covered |
|
|
|
| AH fiber density |
|
|
|
| AH % area covered |
|
|
|
| Brain region | Treatment | Sex | Interaction |
|---|---|---|---|
|
| |||
| NAc |
|
|
|
| LS |
|
|
|
| AMY |
|
|
|
| dHIPP |
|
|
|
| vHIPP |
|
|
|
|
| |||
| NAc |
|
|
|
| LS |
|
|
|
| AMY |
|
|
|
| dHIPP |
|
|
|
| vHIPP |
|
|
|
- —National Institute of Environmental Health Sciences10.13039/100000066
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsHealth, Environment, Cognitive Aging · Anesthesia and Neurotoxicity Research · Air Quality and Health Impacts
INTRODUCTION
1
Air pollution is a massive public health burden with over 7.3 billion people exposed to unsafe levels of air pollution globally.1, 2, 3 Furthermore, air pollution disproportionately impacts low‐income communities that face higher levels of psychosocial stress and lower resource availability as compared to more affluent communities.4, 5 Both air pollution and maternal psychosocial stress during pregnancy are risk factors for neurodevelopmental disorders, such as autism spectrum disorder (ASD).6, 7, 8, 9, 10, 11, 12 While 50–80% of autism liability can be attributed to genetic sources, this still leaves a substantial amount of risk that is driven by environmental exposures.13, 14 Cumulative developmental exposures to air pollution and psychosocial stressors may have compounding effects.15, 16 For example, McGuinn et al.17 found that PM_2.5_ (particulate matter less than 2.5 μm diameter) exposure during the first year of life was associated with increased ASD risk, and that association was moderated by neighborhood deprivation score.17 Similarly, maternal perceived stress is associated with decreased hippocampal volume and visuospatial reasoning in 5‐ to 7‐year‐old children, and this relationship is moderated by prenatal polycyclic aromatic hydrocarbon (a component of air pollution) exposure.17, 18 Taken together, the existing literature suggests that prenatal exposure to air pollution and maternal psychosocial stress may interact to alter neurodevelopmental trajectories in ways that may increase risk for neurodevelopmental and/or psychiatric disorders.18, 19, 20, 21
Animal models lend further insight into the neurobiological mechanisms underlying disrupted neurodevelopment following prenatal toxicant and/or maternal stress (MS) exposures. Several studies using rodent models of air pollution and/or MS during early postnatal development have observed changes in social behavior in offspring.22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34 In humans, disrupted social behavior is a defining symptom of ASD.35 Exposure to ultrafine particulate matter during the first week of postnatal life decreased time spent exploring a novel social stimulus by adult male but not female mice.31 Male rats exposed to PM_2.5_ from postnatal days (P) 8 to 22 display decreased ultrasonic vocalizations, social interaction, and social discrimination, and increased anxiety‐like behavior as compared to controls.29 Similarly, developmental exposure to particulate matter until the end of the lactation period decreased social play and allogrooming in adolescent male rats.30 Importantly, these studies all involved air pollution exposure directly to individuals during the postnatal period. This postnatal exposure is fundamentally different from maternal gestational exposure which indirectly impacts offspring in utero via consequences of maternal immune activation (MIA36), rather than the immune sequelae that would result from direct air pollution exposure to offspring. In a line of work using a mouse model of combined exposure to diesel exhaust particles (DEP) and MS during pregnancy,36, 37, 38 offspring are exposed to DEP throughout pregnancy and maternal resource limitation stress from embryonic day (E)13 to the end of pregnancy. In these studies, P7–P9 male and female offspring display increased ultrasonic vocalization (USV) frequency but reduced USV complexity following DEP/MS as compared to control mice.36 During adolescence, DEP/MS males, but not females, exhibit deficits in sociability and social novelty preference behavior as compared to controls.36, 38 In adulthood, DEP/MS offspring also display impaired contextual fear recall, appetitive social behavior, courtship behavior, and functional activation of social behavior circuits in males only.36, 37 Importantly, deficits in sociability and social novelty preference are not observed following either prenatal air pollution or MS exposure alone, suggesting that the addition of a stressor to mothers unmasks vulnerability to environmental toxicants,38 in line with findings from the human literature. However, which social behavior circuits are disrupted following DEP/MS behavior remains unknown.
The social decision‐making network (SDMN) is a complex set of interconnected brain regions crucial to the generation of social behaviors that includes brain regions such as the nucleus accumbens (NAc), lateral septum (LS), hippocampus, and amygdala (AMY39, 40). Oxytocin (OXT) and vasopressin (AVP) are critical neuropeptide modulators of social behavior across species. OXT signaling has been implicated in numerous behaviors including maternal behavior, reproductive behavior, social investigation and recognition, social novelty‐seeking, social avoidance, cooperative behaviors, and social play, often with sex‐specific effects (for recent reviews, see References 41, 42, 43). Meanwhile, AVP signaling regulates aggression, social recognition, cooperative behaviors, partner preference, social communication, social play, social investigation, and others (for recent reviews, see References 43, 44, 45). Both OXT and AVP are synthesized in hypothalamic nuclei, most largely the paraventricular nucleus of the hypothalamus (PVN) and the supraoptic nucleus of the hypothalamus (SON46, 47, 48) and in the extended AMY.49, 50, 51, 52, 53 Importantly, there is abundant expression of OXT and AVP fibers, as well as their respective receptors: the oxytocin receptor (Oxtr) and the vasopressin 1a receptor (V1aR) throughout the regions of the SDMN.48, 49, 52, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70
There is a wealth of literature examining the effects of perinatal environmental toxicant exposures on OXT and AVP systems, particularly plastics, flame retardants, and pesticides (for reviews, see References 71, 72). However, there is very little work on the effects of perinatal air pollutants on OXT, AVP, and their respective receptors in rodent models. DEP/MS induces MIA,36 and several studies using MIA models have found changes in OXT, Oxtr, and AVP.73, 74, 75, 76 Similarly, studies of MS during pregnancy have found decreased Oxtr mRNA expression in the prefrontal cortex of male offspring.27 However, it is unknown how combined exposures to air pollution and MS impact OXT and AVP cell and receptor expression in the brain.
For the present study, we aimed to assess the impact of prenatal DEP/MS on social behavior, OXT and AVP cells and their projections in the hypothalamus, and Oxtr and Avpr1a mRNA expression within regions of the SDMN as compared to VEH/CON in both males and female offspring. Specifically, we measured OXT‐immunoreactive (‐ir) and AVP‐ir cell number and fiber density within the PVN, lateral hypothalamus (LH), anterior hypothalamus (AH), and medial preoptic area (MPOA). We assessed Oxtr and Avpr1a mRNA expression in the NAc, LS, AMY, dorsal hippocampus (dHipp), and ventral hippocampus (vHipp) following DEP/MS in male and female offspring. Because our previous work only found social deficits following combined DEP/MS exposure but neither DEP nor MS alone, we chose to only include VEH/CON and DEP/MS groups in this study. Based on the previous work in this model that reported male‐specific decreases in social behavior,36, 38 we hypothesized that DEP/MS would reduce the expression of OXT and/or AVP system parameters in many of these regions in male offspring, corresponding to social deficits.
METHODS
2
Animals
2.1
Wild‐type C57Bl/6J mice were purchased from Jackson Laboratories and bred in the laboratory for two generations before use in experiments. Animals were group‐housed with same‐sex littermates under standard laboratory conditions (12‐h light/dark cycle [6 a.m. to 6 p.m.], 68°F, 50% humidity, ad libitum access to food and water) prior to experiments. Experiments were conducted following the NIH Guide to the Care and Use of Laboratory Animals and approved by the Institutional Animal Care and Use Committee (IACUC) at Boston College.
DEP/MS exposures
2.2
DEP instillations
2.2.1
DEP was obtained from Dr. Staci Bilbo at Duke University. These particles originate from Dr. Ian Gilmour at the Environmental Protection Agency and are consistent in composition to those used in previous studies.36, 37, 38, 77 Instillation procedures were conducted in accordance with previously published work using this model (Figure 1a).36, 37, 38 Nulliparous C57Bl/6J females were time mated and the presence of a vaginal plug was used as an indication of pregnancy and set as E0. Females were then pair‐housed based on embryonic day until E13, at which point they were singly housed in cages containing a thin layer of AlphaDri bedding (AlphaDri; Shepherd Specialty Papers). Beginning on E2, dams received instillations every 3 days throughout pregnancy, totaling 6 instillations on E2, E5, E8, E11, E14, and E17. For each instillation, females were weighed and anesthetized with 4% isoflurane. Dams were then suspended by their incisors from a plastic wire and received either 50 μg/50 μL of DEP in vehicle (VEH; 0.05% Tween20 in phosphate‐buffered saline [PBS]) or 50 μL of VEH alone via oropharyngeal instillation over the course of 60s. Females were monitored in the home cage until they regained consciousness and resumed typical behavior (<5 min).
*Perinatal following prenatal DEP/MS exposure. (a) Schematic of experimental timeline. (b) No effect of treatment was observed in maternal weight gain during gestation. (c–e) No effects of DEP/MS treatment on overall litter size or number of male or female pups per litter as compared to VEH/CON. (f) At P16, DEP/MS males weighed less than VEH/CON males (p = 0.017) while DEP/MS females did not differ from VEH/CON females (p = 0.17). (g) At P40, there was no difference in body weight between DEP/MS and VEH/CON males (p = 0.75), but DEP/MS females weighed more than VEH/CON females (p = 0.04); (b) two‐way ANOVA (treatment × embryonic day), (c–e) unpaired t‐tests. (f, g) Mann–Whitney non‐parametric tests. Data = mean ± SEM, p < 0.05. ANOVA, analysis of variance; CON, control bedding condition; DEP, diesel exhaust particles; E, embryonic day; MS, maternal stress condition; P, postnatal day; VEH, vehicle.
Maternal stress
2.2.2
To induce a maternal stressor (again in line with previous work using this model),36, 37, 38 during the last 5 days of pregnancy (beginning on E13) females in the DEP condition were singly housed in a cage containing a thin layer of AlphaDri bedding covered by an elevated aluminum mesh platform (0.4 cm × 0.9 cm mesh; McNichols Co., Tampa, FL) and given 2/3 (~1.9 g) of a cotton nestlet (MS condition). At the same timepoint, females receiving VEH instillations were placed into a cage with AlphaDri bedding and an entire nestlet (CON condition). On E17.5 (evening before delivery), all dams, regardless of treatment, were transferred to a clean cage of AlphaDri bedding with a full nestlet. Litters were left undisturbed during the first week of life. Litter size and sex ratios were recorded between P7 and P10, and pup body weights were measured on P16 and P40. Litters were not culled at any point. Offspring were weaned into cages with same‐sex littermates at P24.
Behavioral testing
2.3
During the adolescent period (P25–P37), offspring underwent a series of behavioral assays each conducted during the second half of the light phase (1–5 pm). Behavior was assessed sequentially on different days in the light–dark box test, the three‐chambered sociability assay, and the three‐chambered social novelty preference test. For all assays, males and females were tested using separate apparatuses on different days, and each apparatus was disinfected between each test. Animals were handled and habituated to the testing room and the testing apparatus the day before testing. On testing days, animals were moved to the testing room 1 h prior to the beginning of the test to acclimate to the room. All videos were scored using Solomon Coder (Solomon.andraspeter.com) by blinded observers. Subjects in each treatment group were drawn from four to five different litters (see Figure S1 for complete litter information).
Light–dark box test
2.3.1
The light–dark box is a square apparatus (40 cm × 40 cm) divided in half by a barrier with a small opening for animals to pass through (Figure 3a). One half of the box is made of clear plexiglass (the “light” side) and the other half is opaque black and has a lid over the top (the “dark” side). Mice were placed into the dark side of the box and allowed to freely roam between both sides of the box for 5 min. Time spent on the light side (s), dark side (s), and head‐poking behavior (defined as the mouse's head being on one side of the chamber while the rest of its body was on the other side; s), were quantified. Fecal boli deposition (N) was also quantified.
Sociability assay
2.3.2
The three‐chambered social preference test was performed according to Smith et al.38 and consists of a three‐chambered arena (60 cm × 40 cm) with openings allowing passage between the chambers. Stimuli were confined within smaller containers (Plexiglass rod sides; diameter 10 cm) in each of the side chambers. Subjects were placed into the middle chamber and freely allowed to investigate each stimulus (either a novel age‐, sex‐, and treatment‐matched conspecific or a novel rubber duck) over the course of 5 min. Stimulus animals were habituated to the testing arena 1 day prior to testing. Behaviors quantified included: social, middle, and object chamber times (s), social and object stimulus interaction times (s; defined as time spent investigating/nose‐poking into the stimulus container and climbing on it); social and object investigation time (s; nose‐poking specifically), and social and object stimulus climbing (s; climbing specifically on the stimulus container rather than the chamber more generally). Interaction time included both nose‐poking and stimulus climbing, which were also scored separately. Middle chamber entries were recorded as a measure of locomotor activity. Animals were excluded from the test if they spent more than 25% of the test time climbing/sitting on top of the smaller cylinders containing the stimuli as this prevents them from engaging with the task (three mice).
Social novelty preference test
2.3.3
The social novelty preference test was conducted 2–3 days after the sociability assay for 5 min. In the social novelty preference test (adapted from Reference78), subject animals were placed in the middle of the same three‐chamber apparatus used for the sociability assay and allowed to freely investigate each stimulus (either a novel age‐, sex‐, and treatment‐matched conspecific or a familiar age‐, sex‐, and treatment‐matched cage mate) over the course of 5 min. Behaviors quantified included novel, middle, and familiar chamber times (s), novel and familiar interaction times (s; defined as time spent investigating/nose‐poking into the stimulus container and climbing on it), novel and familiar investigation time (s; nose‐poking specifically), and novel and familiar stimulus climbing (s; climbing specifically on the stimulus container rather than the chamber more generally). Interaction time included both nose‐poking and stimulus climbing, which were also scored separately.
Immunohistochemistry
2.4
Tissue collection
2.4.1
For all immunohistochemistry, animals were euthanized via CO_2_ inhalation between P42 and P44, and brains were perfused with ice‐cold saline. Brains were then removed and post‐fixed for 48 h in 4% paraformaldehyde, followed by 48 h in 30% sucrose. Brains were then flash frozen in 2‐methylbutane and stored at −20°C until sectioning at 30 μm on a cryostat (Leica Biosystems). Subjects were obtained from multiple litters (1–4 pups/sex/litter from eight litters total) in all groups.
OXT and AVP staining in the hypothalamus
2.4.2
Free‐floating sections containing hypothalamic nuclei (Allen Mouse Brain Atlas: https://atlas.brain-map.org; Bregma 0.02–1.06 mm) were rinsed 5× in PBS, incubated in 10 mM sodium citrate for 30 min at 60°C for antigen retrieval, and rinsed 3× in PBS again. They were then incubated in a 1 mg/1 mL solution of sodium tetraborate in PBS at room temperature for 30 min, followed by 3× PBS rinses. Next, sections were blocked for 1 h in PBS with 10% normal goat serum, 0.3% Triton‐X, and 1% H_2_O_2_. Sections were then incubated overnight with primary antibodies (guinea pig anti‐OXT, 1:2000, Synaptic Systems; rabbit anti‐AVP, 1:2000, Millipore Sigma) at reverse transcriptase (RT) on a shaker. After 3× PBS rinses, sections were then incubated in secondary antibodies (488 goat anti‐rabbit IgG, 1:500, Invitrogen; 568 goat anti‐guinea pig IgG, 1:500, Invitrogen) for 5 h on a shaker, followed by 3× PBS rinses. Sections were mounted on gelatinized slides and cover slipped with Vectashield Antifade Mounting Medium with (4', 6‐Diamidino‐2‐phenylindole, Vector Laboratories).
Imaging
2.4.3
A Zeiss AxioImager Z2 microscope was used to take 10× magnification, tile‐scanned images of the hypothalamus, including the PVN, LH, AH, and the MPOA. A 3 × 3 tile was used to capture the regions of interest. Seven‐step z‐stacks were taken with a step‐size of 1.53 μm measured from the center of the image to span 10 μm. Images were taken centering the third ventricle. The number of sections analyzed for each brain region is detailed in the sections below.
Quantification of OXT and AVP cell bodies in the PVN
Analysis of OXT‐ir and AVP‐ir cell bodies in the PVN was categorized based on OXT and AVP staining patterns. We observed that the density of OXT‐ir and AVP‐ir cells varied substantially across the rostrocaudal extent of the PVN.79, 80 Furthermore, previous work has found that early life events can have selective impacts on rostral versus caudal OXT‐ir cells in the PVN.80 Therefore, for OXT‐ir quantification, sections were categorized into Level 1 (Bregma −0.38 to −0.56 mm; the more anterior portion of the PVN with only sparse OXT‐ir cell bodies) and Level 2 (Bregma −0.56 to −0.88 mm: dense OXT‐ir cell bodies in heart‐shaped formation). For AVP quantification, sections were also categorized into two levels; Level 1 (Bregma −0.28 to −0.56 mm: moderately dense AVP‐ir cell bodies extending down along the third ventricle) and Level 2 (Bregma −0.56 to −0.66 mm: densest AVP‐ir cell bodies in heart‐shaped formation). The expression patterns of OXT and AVP differ across the PVN (OXT expression is shifted slightly earlier in the PVN); thus, Level 1 is shifted slightly later for AVP. Z‐stacks were converted into maximum intensity projections and cell bodies for OXT and AVP were hand counted in FIJI software (https://imagej.net/software/fiji) by blinded observers. For each stain (OXT vs. AVP), 1–3 sections were analyzed in each level (average of 2, median of 2 in each). For all sections, cell bodies were counted bilaterally in each coronal section and values for each level of the PVN for a given animal were averaged to provide a single measurement per animal.
Quantification of OXT and AVP fibers and/or cells in the LH, AH, and MPOA
Using the same images in which cell bodies were quantified in the PVN, OXT‐ir and AVP‐ir fibers were quantified in the LH (Bregma −0.48 to −1.06 mm), AH (Bregma −0.48 to −0.96 mm), and MPOA (Bregma 0.02 to −0.28 mm) using FIJI. Regions of interest for each hypothalamic nucleus were defined using the Allen Mouse Brain Atlas. In FIJI, a 325 μm^2^ × 325 μm^2^ was placed within the LH and a 67826.2 μm^2^ ellipse was separately placed within the AH and MPOA to create the regions of interest (ROI) for quantification. The sizes for these ROIs were determined using the scale of the Allen Mouse Brain Atlas, and each image had two ROIs per brain region (one ROI/hemisphere). Fiber number (and cell number in the AH) was hand counted in each brain region by a blinded observer. Fibers were considered individual fibers when they could be clearly distinguished as an unbroken unit. Disconnected puncta were not counted. Fiber density and fiber caliber differed between hypothalamic regions, with denser, smaller caliber fibers in the AH as compared to the MPOA and LH. These counts were conducted by blinded observers. However, because hand counting is inherently limited, we also included a more objective measure of fiber density to complement this. Using FIJI, analysis of % area covered in each ROI was conducted by thresholding maximum intensity projection images. For each brain region analyzed, 1–4 sections were quantified per animal (LH [average 1.7, median 2], AH [average 2, median 2], and MPOA [average 2.5, median 3]), and measurements were averaged to provide a single measurement/brain region/subject. OXT‐ir and AVP‐ir cell bodies found in the AH were quantified following the same procedure as in the PVN.
Quantitative PCR and analysis
2.5
Tissue collection and punching
2.5.1
At P50‐52, animals from a separate cohort of animals that were not exposed to behavioral testing were euthanized with CO_2_. Subjects were obtained from multiple litters (3–7 litters/group; for detailed litter identity see Figure S2) in all groups. Brains were removed and flash‐frozen in 2‐methylbutane, then stored at −80°C until sectioning and tissue punching. For tissue punching, brains were mounted in a sterilized cryostat and tissue punches were collected by inserting a sterilized 1.5 mm diameter core sampling tool to a depth of 1 mm (Electron Microscopy Sciences). Tissue punches were collected from the NAc, LS, AMY, dHipp, and vHipp. These regions were chosen based on their known expression of Oxtr and V1aR, importance to social behavior,60, 68 and shape limits of the tissue punching procedure. Tissue was collected bilaterally except for the LS. Punches were stored in 500 μL Trizol (Thermo Fisher Scientific) and frozen at −80°C until RNA extraction.
RNA extraction
2.5.2
RNA was extracted using phenol‐chloroform extraction. Samples were homogenized in 500 μL Trizol (Thermo Fisher Scientific), vortexed at 2000 rpm for 10 min, and rested for 15 min. 100 μL chloroform (Sigma Aldrich) was added and samples were vortexed at 2000 rpm for 2 min and rested for 3 min. Samples were centrifuged at 11,800 rpm for 15 min at 4°C. The aqueous phase was extracted from the phase gradient; 250 μL isopropanol (Sigma Aldrich) and 2 μL Glycogen Blue (Thermo Fisher Scientific) were added to the aqueous phase, and samples were centrifuged at 11,800 rpm for 10 min at 4°C. RNA pellets were rinsed with 500 μL ice‐cold 75% ethanol (Thermo Fisher Scientific) twice, centrifuging at 9000 rpm for 5 min at 4°C after ethanol was added each time. Pellets were then re‐suspended in nuclease‐free (NF) H_2_O (Fisher Bioreagents) and stored at −80°C until cDNA synthesis.
cDNA synthesis
2.5.3
cDNA synthesis was performed using the QuantiTect Reverse Transcription Kit according to its standard protocol (Qiagen). RNA quantity and quality were assessed using a NanoDrop One C (Thermo Fisher Scientific). Briefly, samples were gDNAse treated in a Thermofisher Applied Biosystems MiniAmp Plus Thermal Cycler for 2 min at 42°C. Next, the Master mix was prepared with 1 μL RT, 4 μL RT buffer, and 1 μL RT primer mix per sample. Master mix was added to suspended RNA. Samples were incubated for 15 min at 42°C and 3 min at 95°C. Samples were then diluted to 10 ng/μL and stored at −20°C unless proceeding immediately with quantitative PCR (qPCR). No‐template and no‐RT controls were included to verify primer specificity during qPCR.
Quantitative PCR
2.5.4
qPCR was performed using the QuantiNova SYBR Green PCR Kit (Qiagen). PCR primers were designed in‐house and purchased from Integrated DNA Technologies (18S: F: GAATAATGGAATAGGACCGC, R: CTTTCGCTCTGGTCCGTCTT; Oxtr: F: ACCTGGACTCCCACCTATTT, R: GCCGTCTTTCACAAGATACCA; Avpr1a: F: ATCCGCACAGTGAAGATGAC, R: GGAATCGGTCCAAACGAAATTG). To prepare the SYBR master mix, 6.5 μL QuantiNova SYBR, 1 μL forward primer, 1 μL reverse primer, and 3.5 μL NF H_2_O per reaction were combined. Next, 12 μL SYBR master mix per well was plated on MicroAmp Optical 96‐well Reaction Plates (Applied Biosystems), then 1 μL template cDNA was added to each well. Samples were run in duplicates. The no‐template and no‐RT controls generated during cDNA synthesis were run on each plate to verify primer specificity. qPCR was run on a QuantStudio 3 Real‐Time PCR machine (Thermo Fisher Scientific). Samples were held at 95°C for 2 min to activate SYBR. Forty cycles of PCR were performed: samples were held at 95°C for 5 min, ramped to 49.4°C or 49.9°C at a speed of 1.6°C/s (for Oxtr and Avpr1a, respectively), and held at 49.4°C or 49.9°C for 11 s; then the cycle was repeated. After PCR was complete, a melt curve was performed.
2‐∆∆CT analysis
2.5.5
Relative gene expression was calculated using the 2‐∆∆CT method, relative to 18S gene expression and the lowest sample on the plate.81, 82 Microsoft Excel was used for 2‐∆∆CT calculations. Samples were excluded before unblinding if they failed to amplify specifically (i.e., no amplification or multiple melt curve peaks) or if replicates differed by >1 fold change. On average across datasets out of 48 total samples, 0.8 samples/dataset were removed due to failure to amplify, and 5.4 samples were removed due to replicates differing by >1 fold change (see Table S1 for complete details).
Statistics
2.6
All statistical analyses were conducted using GraphPad Prism 10 software. Data were tested for normality using the Kolmogorov–Smirnov and Shapiro–Wilk tests and homoscedasticity using Spearman's test. In instances where these assumptions were violated, non‐parametric testing was used as detailed below. For maternal body weight, a two‐way analysis of variance (ANOVA) (embryonic day × treatment) was conducted. Litter ratios were analyzed using unpaired t‐tests (VEH/CON vs. DEP/MS). Pup body weights were analyzed using non‐parametric Mann–Whitney U tests. Behavioral outcomes (interaction, nose‐poking, and stimulus climbing) were assessed using mixed effect two‐way ANOVAs (stimulus [repeated measure] × treatment [between‐subjects measure]) followed by Sidak's multiple comparisons tests in the case of significant interaction effects. Chamber time was analyzed using a mixed effects model (REML) followed by Sidak's multiple comparisons tests. Because males and females were tested using separate apparatuses on different days, male and female behavior results were analyzed separately. Fecal bolus deposition was not normally distributed and, therefore, analyzed with non‐parametric Mann–Whitney U tests. OXT‐ir and AVP‐ir cells in the PVN and fibers in the LH and AH were analyzed using two‐way ANOVAs (sex × treatment). No significant main effects were observed; therefore, no post hoc comparisons were made. OXT‐ir and AVP‐ir cells in the AH and AVP fibers in the MPOA were not normally distributed; therefore, they were analyzed using non‐parametric Mann–Whitney U tests. Oxtr and Avpr1a mRNA were analyzed using two‐way ANOVAs (sex × treatment). No significant main effects were observed; therefore, no post hoc comparisons were made. Statistical outliers were identified using the ROUT outlier test (Q = 1%). Data are expressed as mean ± SEM and statistical significance was set at p < 0.05.
RESULTS
3
Maternal and litter outcomes
3.1
While both VEH/CON and DEP/MS dams gained significant weight during pregnancy, this did not differ between treatment groups (see Table 1 for complete F‐statistics; Figure 1b). DEP/MS exposure had no effect on litter size (Figure 1c, t (12) = 0.28, p = 0.79) or on the number of males (Figure 1d, t (12) = 0.56, p = 0.59) or females (Figure 1e, t (12) = 0.00, p > 0.99) in each litter.
Pup body weights were measured at timepoints pre‐ and post‐weaning to assess pup developmental growth trajectory following prenatal DEP/MS. Prior to weaning, at P16, DEP/MS males weighed less than VEH/CON males (Mann–Whitney U = 112, p = 0.017) while DEP/MS females did not differ from VEH/CON females (Mann–Whitney U = 267, p = 0.17; Figure 1f). At P40, there was no difference in body weight between DEP/MS and VEH/CON males (Mann–Whitney U = 130.5, p = 0.75), but DEP/MS females weighed more than VEH/CON females (Mann–Whitney U = 188.5, p = 0.04; Figure 1g).
Behavior
3.2
Light–dark box test
3.2.1
In the light–dark box test (Figure 2a), we found that DEP/MS did not alter chamber time or fecal boli deposition in males (see Table 2 for complete F‐statistics and post hoc results; fecal boli: Mann–Whitney U = 16, p = 0.19; Figure 2b,c). Males spent more time in the dark side of the box than in the light (post hoc light vs. dark: p < 0.01) regardless of treatment. In females, there was no main effect of treatment, but a significant treatment × chamber interaction effect (p = 0.037). Post hoc tests found that DEP/MS females tended to spend more time in the light side (post hoc VEH/CON vs. DEP/MS in light chamber: p = 0.06) and less time in the dark side (post hoc VEH/CON vs. DEP/MS in dark chamber: p = 0.07) as compared to VEH/CON females. VEH/CON females spent more time in the dark side of the box (post hoc light vs. dark in VEH/CON females: p < 0.001) while DEP/MS females did not (post hoc light vs. dark in DEP/MS females: p = 0.683). There was no difference in fecal boli deposited in DEP/MS versus VEH/CON females (fecal boli: Mann–Whitney U = 21.5, p = 0.16; Figure 2d,e).
Results of light/dark box and social behavior assays. (a) Light–dark box test. (b, c) In males, there is a significant main effect of chamber (with both groups spending more time in the dark chamber), but no effect of DEP/MS, and no effect of DEP/MS on fecal boli. (d, e) In females, there is a main effect of chamber and an interaction such that VEH/CON females prefer the dark side of the box while DEP/MS females do not. There is no effect of DEP/MS on fecal boli. (f) Sociability assay. (g–i) In males, DEP/MS decreases interaction time (g) and stimulus climbing (i) but not nose‐poking (h). (j) In males, there is a main effect of chamber with both groups spending more time in the animal chamber. (k–m) In females, DEP/MS has no effect on interaction (k), stimulus climbing (m) or nose‐poking, but there is a main effect of stimulus on nose‐poking (l). (n) In females, there is no effect of DEP/MS on chamber time, but a main effect of chamber (with both groups spending more time in the animal chamber). (o) Three‐chambered social novelty preference test. (p–r) In males, DEP/MS has no effect on interaction (p) or nose‐poking (q) but tends to reduce stimulus climbing (r) as compared to VEH/CON. (s) In males, there is a main effect of chamber, with males spending more time in both the novel and familiar chambers than the middle chamber. (t–v) In females, there are significant interaction effects on interaction (t) and nose‐poking (u), but not stimulus climbing (v) with DEP/MS females interacting less with the novel stimulus as compared to VEH/CON females. (w) There is an interaction between chamber and treatment on chamber times, with DEP/MS females spending more time in the familiar chamber and less time in the novel chamber compared to VEH/CON females. Chamber times for all tests (b, d, j, n, s, w) are mixed effects models with Sidak's post hoc comparisons for chamber and (in the case of interaction effects, treatment). Post hoc comparisons for chamber time are marked with * or & (denoting difference from both other groups). For significant interaction effects, the interaction effect is marked () and post hoc comparisons are also labeled (* and p‐values if nonsignificant). Fecal boli (c, e) are analyzed with non‐parametric Mann–Whitney U tests; (g–i), (k–m), (p–r), and (t–v) are mixed effects two‐way ANOVA (chamber × treatment) with Sidak's post hoc tests in the case of significant interaction effects. s = seconds. Data = mean ± SEM, p < 0.05; #p < 0.08, &p < 0.05 versus all other groups. ANOVA, analysis of variance; CON, control bedding condition; DEP, diesel exhaust particles; MS, maternal stress condition; VEH, vehicle.
Sociability assay
3.2.2
In the sociability assay (Figure 2f) in males, there was a main effect of treatment such that total interaction time was lower in DEP/MS as compared to VEH/CON males (main effect of treatment: p = 0.027; see Table 2 for complete statistics and post hoc results; Figure 2g). Interaction time is comprised of nose‐poking into (investigating) and stimulus climbing on the social or object container. Interestingly, when broken down into its component parts, we found that the decrease in interaction time was largely driven by decreased stimulus climbing following DEP/MS (main effect of treatment: p = 0.04) rather than nose‐poking (main effect of treatment: p = 0.23; Figure 2h,i). There were no main effects of treatment on chamber time in males. However, there was a main effect of chamber (p < 0.001) with more time spent in the social chamber than the middle or novel object chambers (Figure 2j). DEP/MS did not affect middle chamber entries in the sociability assay in males (middle chamber entries: t (15) = 0.239, p = 0.815).
In females, DEP/MS treatment did not affect stimulus interaction, nose‐poking, or stimulus climbing in the sociability assay (Table 2 and Figure 2k–m). There was a main effect of stimulus on nose‐poking time, with females investigating the social stimulus more than the object (main effect of stimulus: p < 0.001; Figure 2l), regardless of treatment. In line with this, there was also a main effect of chamber, with females spending a greater amount of time in the chamber containing the social stimulus than the middle or novel object chambers, regardless of treatment (Figure 2n). DEP/MS did not affect middle chamber entries in the sociability assay in females (middle chamber entries: t (17) = 1.191, p = 0.250).
Social novelty preference test
3.2.3
In the social novelty preference test (Figure 2o) in males, DEP/MS had no effect on interaction time (for complete F‐statistics and post hoc results see Table 2 and Figure 2p) but tended to decrease stimulus climbing time (main effect of treatment, p = 0.07; Figure 2r). There was no significant effect of DEP/MS on nose‐poking (Figure 2q). There was no significant effect of treatment on overall chamber time, but there was a main effect of chamber, with males spending more time in the novel and familiar stimulus chambers than the middle chamber regardless of treatment (Figure 2s). DEP/MS tended to reduce middle chamber entries in the social novelty preference test in males, though this did not reach statistical significance (middle chamber entries: t (17) = 1.903, p = 0.07).
In females, there was a significant interaction effect on interaction time (stimulus × treatment interaction: p = 0.015; Figure 2t). Post hoc comparisons revealed a significant decrease in interaction with the novel conspecific following DEP/MS as compared to VEH/CON (post hoc VEH/CON vs. DEP/MS novel interaction: p = 0.04). This was mirrored in a significant interaction effect on nose‐poking time (stimulus × treatment interaction: p = 0.015; Figure 2u) with DEP/MS females shifting toward decreased novel nose‐poking but more familiar nose‐poking as compared to VEH/CON females. This shift in preference towards the familiar cage mate was also reflected in chamber times. DEP/MS females spent less time in the novel chamber (post hoc VEH/CON vs. DEP/MS novel chamber: p = 0.012) and more time in the familiar chamber as compared to VEH/CON females (post hoc VEH/CON vs. DEP/MS novel chamber: p < 0.01; Figure 2w). DEP/MS did not affect middle chamber entries in the social novelty preference test in females (middle chamber entries: t 18 = 0.436, p = 0.668).
OXT and AVP in the hypothalamus
3.3
OXT and AVP cell bodies in the PVN and AH
3.3.1
We found that in the more rostral Level 1 PVN sub‐region analyzed, DEP/MS tended to increase OXT‐ir cell number, regardless of sex as compared to VEH/CON (main effect of treatment p = 0.07; see Table 3 for complete F‐statistics; Figure 3a,b). This pattern did not continue more caudally in the Level 2 PVN sub‐region, where we did not observe any treatment effects on OXT cell number (Figure 3c,d). We observed no significant effects of sex or treatment on AVP‐ir cell number in either Level 1 or 2 of the PVN (Figure 3e–h), nor did we find any effects of DEP/MS on either OXT‐ir or AVP‐ir cell bodies in the AH which were sparse but present (OXT‐ir cells: males [Mann–Whitney U = 21, p = 0.71], females [Mann–Whitney U = 39, p = 0.91], AVP‐ir cells: males [Mann–Whitney U = 26.5, p = 0.61], females [Mann–Whitney U = 33.5, p = 0.55]; Figure 3i–k).
Effects of DEP/MS exposure on OXT‐ir and AVP‐ir cell number in the PVN. (a) Representative image of OXT‐ir cells in OXT Level 1 in the PVN. (b) DEP/MS tended to increase OXT‐ir cell number in both males and females in OXT Level 1 PVN (main effect of treatment, p = 0.07). (c) Representative image of OXT‐ir cells in OXT Level 2 PVN. (d) No effect of DEP/MS or sex on OXT‐ir cell number in OXT Level 2. (e) Representative image of AVP‐ir cells in AVP Level 1. (f) No effect of DEP/MS or sex on AVP‐ir cell number in AVP Level 1. (g) Representative image of AVP‐ir cells in AVP Level 2 PVN. (h) No effect of DEP/MS or sex on AVP‐ir cell number in AVP Level 2 PVN. (i) Representative images of OXT‐ir and AVP‐ir cells in the AH. (j) No effect of DEP/MS or sex on OXT‐ir cell number in the AH. (k) No effect of DEP/MS or sex on AVP‐ir cell number in the AH. (b, d, f, h) Two‐way ANOVAs (sex × treatment); (j, k) Mann–Whitney U tests. Data = mean ± SEM, # p < 0.07. AH, anterior hypothalamus; ANOVA, analysis of variance; AVP‐ir, vasopressin‐immunoreactive; DEP, diesel exhaust particles; MS, maternal stress condition; OXT‐ir, OXT‐immunoreactive; PVN, paraventricular nucleus of the hypothalamus.
OXT and AVP fibers in the LH, AH, and MPOA
3.3.2
We also assessed OXT‐ir and AVP‐ir fiber density in three hypothalamic nuclei that play varying roles in social behavior: the LH, AH, and MPOA. We used two measures to assess fiber density: fiber number and % area covered by fibers. We did not observe any significant effects of treatment or sex in any of the regions analyzed on either of these measures (for complete F‐statistics, see Table 4 and Figure 4a–o; MPOA AVP‐ir fiber counts: males [Mann–Whitney U = 26.5, p = 0.41], females [Mann–Whitney U = 22.5, p = 0.0.36], MPOA AVP‐ir fiber % area covered: males [Mann–Whitney U = 22, p = 0.21], females [Mann–Whitney U = 28.5, p = 0.77]).
Effects of DEP/MS exposure on OXT‐ir and AVP‐ir fibers in the hypothalamus. (a) Representative images of OXT‐ir and AVP‐ir fibers in the LH. (b–e) No effect of DEP/MS or sex on OXT‐ir or AVP‐ir fibers in the LH (c). (f) Representative images of OXT‐ir and AVP‐ir fibers in the AH. (g–j) No effect of DEP/MS or sex on OXT‐ir or AVP‐ir fibers in the AH. (k) Representative images of OXT‐ir and AVP‐ir fibers in the MPOA. (l–o) No effect of DEP/MS or sex on OXT‐ir or AVP‐ir fibers in the MPOA. (b–e, g–j) Two‐way ANOVAs (sex × treatment); (l–o) Mann–Whitney U tests. Data = mean ± SEM, #p < 0.1. AH, anterior hypothalamus; ANOVA, analysis of variance; AVP‐ir, vasopressin‐immunoreactive; DEP, diesel exhaust particles; MPOA, medial preoptic area; MS, maternal stress condition; OXT‐ir, OXT‐immunoreactive.
Oxtr and Avpr1a gene expression in regions of the SDMN
3.4
We observed numerous effects of sex and treatment on Oxtr and Avpr1a mRNA expression (for complete F‐statistics see Table 5). In the NAc, males exhibited greater Oxtr mRNA expression compared to females (main effect of sex: p < 0.01), but there was no effect of DEP/MS on Oxtr mRNA expression (Figure 5a). Interestingly, DEP/MS increased Avpr1a mRNA expression in both males and females in the NAc as compared to VEH/CON (main effect of treatment: p = 0.015, Figure 5a). In the LS, females expressed more Oxtr and Avpr1a mRNA than males (main effects of sex: Oxtr p = 0.048, Avpr1a p = 0.009), but there was no effect of DEP/MS (Figure 5b). In the AMY, there were no significant effects of sex or treatment on Oxtr mRNA expression (Figure 5c). Additionally, females exhibited greater Avpr1a mRNA expression than males (main effect of sex: p = 0.036, Figure 5c). In the dHipp, neither treatment nor sex altered Oxtr mRNA expression (Figure 5d). There was a trend towards a significant increase in Avpr1a mRNA expression following DEP/MS as compared to VEH/CON (main effect of treatment: p = 0.05), as well as a trend towards greater Avpr1a mRNA expression in females compared to males (main effect of sex p = 0.06, Figure 5d). Finally, in the vHipp, there were no effects of treatment or sex on Oxtr or Avpr1a mRNA expression (Figure 5e).
*Effects of DEP/MS exposure on Oxtr and Avpr1a mRNA in the social brain. (a) In the NAc, males express more Oxtr mRNA than females. DEP/MS increases Avpr1a mRNA expression in the NAc in both males and females. (b) No effect of DEP/MS on Oxtr or Avpr1a mRNA expression in the LS, though females express more Oxtr and Avpr1a mRNA than males. (c) In the AMY, there is no effect of DEP/MS or sex on Oxtr mRNA expression. Females express more Avpr1a mRNA than males in the AMY. (d) In the dHipp, there is no effect of DEP/MS or sex on Oxtr mRNA expression. DEP/MS tends to increase Avpr1a mRNA expression in both males and females and females tend to express more Avpr1a mRNA than males. (e) No effect of DEP/MS or sex on Oxtr or Avpr1a mRNA expression in the vHIPP. Two‐way ANOVAs (sex × treatment). Data = mean ± SEM, p = 0.05, #p < 0.07. AMY, amygdala; ANOVA, analysis of variance; DEP, diesel exhaust particles; dHipp, dorsal hippocampus; LS, lateral septum; MS, maternal stress condition; NAc, nucleus accumbens; vHIPP, ventral hippocampus.
DISCUSSION
4
Taken together, our results indicate several important findings. First, in accordance with previous work, DEP/MS exposure does not alter maternal weight gain during pregnancy or overall litter composition but does appear to subtly impact offspring weight gain during the early postnatal period. Second, in contrast to our hypothesis of male‐specific phenotypes following DEP/MS, we observed effects of DEP/MS exposure on behavior in both male and female offspring, with different outcomes impacted in each sex. Finally, while DEP/MS exposure does not appear to significantly impact OXT‐ir or AVP‐ir cell or fiber density in the hypothalamus, our results show that DEP/MS exposure may alter Avpr1a mRNA but not Oxtr mRNA expression: it increases Avpr1a mRNA expression in the NAc and tends to increase expression in the dHipp (p = 0.05).
Divergent effects of DEP/MS exposure on social behavior in males and females
4.1
We assessed the impact of DEP/MS exposure on several behavioral assays (light/dark box, sociability, and social novelty preference) during the adolescent period in both male and female offspring. Interestingly, we observed effects of DEP/MS exposure on behavior in both sexes, although the specific behavioral outcomes impacted differed between males and females. DEP/MS‐exposed males spent less time engaged in overall interaction and stimulus climbing as compared to their VEH/CON counterparts in the sociability assay and tended to spend less time in stimulus climbing in the social novelty preference test (p = 0.07). The decrease in interaction time in the sociability assay in males following DEP/MS was more pronounced for the social stimulus as compared to the object, but this did not reach statistical significance. DEP/MS‐exposed females, on the other hand, exhibited a shift in their social preference towards the familiar cage mate in the social novelty preference test, as evidenced by reduced interaction, nose‐poking, and chamber times associated with the novel stimulus as compared to VEH/CON females. Our findings in the sociability assay in males are only somewhat in line with previous work.36, 38 We did observe a decrease in interaction time in male offspring following DEP/MS, however this was not specific to social interaction (i.e., it was driven by both social and object interaction). Moreover, we found that this decrease in total interaction was largely driven by a decrease in climbing of the stimuli rather than nose‐poking into them. Together, these findings may reflect a more general decrease in exploratory behavior than previously reported. The decrease in climbing behavior observed following DEP/MS exposure could reflect changes in either exploratory motivation or complex motor abilities (also disrupted in neurodevelopmental disorders83, 84) and this should be the subject of future investigations.
Contrary to our hypothesis that we would see male‐specific social deficits, here, we find decreases in social novelty preference in female offspring following DEP/MS exposure. Our iteration of this test includes the use of a cage mate as the familiar conspecific.78 Because this is a very familiar individual, and subjects are taken directly from co‐housing with this conspecific into the test, it is likely that this shift reflects a decreased motivation to interact with the novel conspecific rather than a deficit in social memory. However, we cannot conclusively rule this out based on our data. It is also interesting to consider given that we also observed a trend (p = 0.05) to an effect of DEP/MS on V1aR in the hippocampus—a brain region in which AVP signaling regulates social memory.85 Interestingly, effects on social behavior in females in this model have been observed previously, but at an earlier developmental timepoint. Block et al.36 measured ultrasonic vocalizations, a measure of social communication at P8 and found effects of DEP/MS exposure in both sexes on number of calls and total time spent calling, which supports the idea that female social development is also impacted by DEP/MS exposure. Decreases in social behavior in females have also been observed in other models of prenatal air pollution or stress exposure in rodents. For example, studies of prenatal through postnatal exposure to air pollutants have found decreases in sociability, social novelty preference, and altered ultrasonic vocalizations in both males and females.22, 34 Prenatal maternal restraint stress also impairs social behavior in both male and female offspring in adulthood.23, 27 These findings are important given that neurodevelopmental disorders are male‐biased but not male‐specific; females are also diagnosed. Further, current animal models of behavior and neurodevelopmental disorders were mostly developed using male subjects only and are likely not adequately capturing female‐specific mechanisms and behaviors.86, 87, 88
In our previous work, we found that social behavior deficits induced by DEP/MS exposure in males were accompanied by changes in the composition of the gut microbiome.38 Furthermore, these social behavior deficits were prevented by cross‐fostering at birth, a manipulation that restored the gut microbiome to the control phenotype. The rodent gut microbiome is known to shift depending on factors including housing conditions, institution, diet, and vendor.89, 90 While we designed our experiments so that bedding, diet, and other conditions were consistent with previous studies, these experiments were conducted at a new institution. It would be interesting to determine whether the changes we observe here in behavior in both sexes are accompanied by changes in the gut microbiome and whether this is a factor that might at least partially explain the differing behavioral results. The gut‐brain axis has been implicated in the regulation of social behavior in numerous studies91, 92, 93, 94, 95, 96, 97, 98 and there is emerging evidence that sex differences exist in the composition of gut microbiota, in both humans and mice.99 C57Bl/6J mice—the mouse strain used in the present study—exhibit some of the most robust sex differences in gut microbiota, including expression of the genus Bacteroides, with greater abundance in females than males.100 Bacteroides is differentially expressed in individuals with ASD101 and decreased in abundance in males exposed to DEP/MS.38 One possibility is that differences in the abundance of bacterial genera such as Bacteroides between sexes, exposures (such as DEP/MS), and studies may contribute to sex differences in social behavior outcomes.
In the light–dark box test, we observed no main effect of treatment in females, but a significant interaction effect between chamber and treatment. Post hoc tests revealed that DEP/MS exposed females tended to spend more time in the light side of the box compared to VEH/CON females (p = 0.06). There was no difference in the tendency to deposit more fecal boli during behavior, which is sometimes interpreted as a measure of anxiety.102, 103, 104 This is the first study to assess behavior in the light–dark box test following prenatal DEP/MS exposure. However, previous findings in the DEP/MS model have been mixed with respect to the assessment of anxiety‐like behavior using other assays. Smith et al.38 found no differences in behavior in the open field test following DEP/MS exposure during adolescence.38 In contrast, in adulthood, Bolton et al.37 found that DEP/MS increased anxiety‐like behavior in the elevated zero maze in both male and female offspring.37 More sensitive and/or naturalistic assays of anxiety‐like behavior, such as the looming‐shadow task,105 may be helpful in clarifying this behavioral dimension.
No significant effects of DEP/MS exposure on OXT‐ir or AVP‐ir cell/fiber density in the hypothalamus
4.2
Following behavioral testing, offspring brains were collected to determine whether there are long‐lasting developmental changes to OXT/AVP innervation following DEP/MS that corresponded to behavioral effects. We found that DEP/MS exposure had no significant effects on OXT‐ir or AVP‐ir cell number in the PVN or AH, nor on fiber density in the AH, LH, or MPOA. We observed a trend towards higher OXT‐ir cell number in the more anterior portion of the PVN assessed, but this did not reach statistical significance (p = 0.07). Still, this finding is interesting given that OXT‐ir cell number has been shown to increase in the PVN with manipulations to the gut microbiome such as treatment with L. reuteri.80 The present study is the first to quantify OXT and AVP innervation following prenatal air pollution or DEP/MS exposure, and these systems have been much more thoroughly characterized following perinatal exposure to other environmental toxicants, the effects of which on OXT and AVP vary considerably depending on dosage, timing of exposure, sex and other factors (for review, see Reference 71). For example, Reilly et al.106 also found no overall effects of prenatal exposure to the endocrine disruptor Aroclor 1221 (a polychlorinated biphenyl) on either OXT or AVP cell number in the PVN, but the suggestion of a rostro‐caudal shift.106 DEP/MS exposure induces an MIA cascade in dams36 and our findings are largely in line with those of two studies of OXT and AVP expression in the hypothalamus following MIA, both in rats.73, 76 Taylor et al.76 used in situ hybridization to assess AVP mRNA in the PVN, supraoptic nucleus (SON) and superchiasmatic nucleus of the hypothalamus and found no effects of maternal lipopolysaccharide (LPS)‐induced immune activation on AVP mRNA expression in any of these nuclei, although AH, LH, and MPOA were not assessed.76 Similarly, Breach et al.73 used an allergic asthma‐induced MIA model and found no effects of MIA on OXT‐ir or AVP‐ir cell or fiber density in the PVN, SON, but found a male‐specific decrease in AVP‐ir fiber density in the LH; MPOA and AH were not assessed.73 Interestingly, both studies did observe effects, particularly on AVP, in extrahypothalamic regions such as the medial amygdala (MeA) and bed nucleus of the stria terminals (BNST).73, 76 These effects will be discussed further in the context of our receptor expression findings. A limitation of this investigation is that while we have assessed Oxtr and Avpr1a mRNA in extrahypothalamic regions we did not include analysis of OXT‐ir or AVP‐ir cell or fiber densities in regions such as the LS, AMY, or BNST. Therefore, it is possible that DEP/MS impacts OXT and AVP expression in these regions and should be the subject of future investigations.
DEP/MS effects on Avpr1a, but not Oxtr, mRNA in regions of the SDMN
4.3
We observed treatment effects on Avpr1a mRNA expression and sex effects on both Avpr1a and Oxtr mRNA expression. Most notably, Avpr1a mRNA expression was significantly increased in the NAc in both males and females following DEP/MS. This is especially interesting given that the NAc is a brain region in which numerous changes have been observed following DEP/MS exposure,38 including male‐specific decreases in dopamine receptor expression and shifts in microglial morphology and gene expression. This finding is also in line with Breach et al.,73 who observed higher AVP‐ir fiber density in the NAc after MIA, albeit only in male offspring. There is limited work indicating the functional relevance of changes to V1aR expression in the NAc in mice. AVP acting at V1aRs in the NAc and ventral pallidum has been shown to mediate social behaviors such as partner preference formation in monogamous prairie voles.107 For example, the formation of a partner preference induces V1aR upregulation in the NAc of female prairie voles and partner preference formation is prevented by V1aR antagonist administration.108 This is relevant given that females exposed to DEP/MS display an increased preference for the familiar cage mate along with higher NAc‐Avpr1a expression. One possibility is that higher NAc‐Avpr1a following DEP/MS exposure serves to increase familiar rather than novel interactions. Oxtr mRNA was not impacted by DEP/MS exposure, but we did observe a sex difference in the NAc with higher Oxtr mRNA in males as compared to females. This sex difference is consistent with previous autoradiography work in rodents showing higher Oxtr binding in males than in females.60
Avpr1a mRNA also tended to increase in the dHipp following DEP/MS exposure as compared to VEH/CON (p = 0.05). Very little previous work has characterized the impacts of prenatal air pollution and/or stress on hippocampal V1aR expression in either sex. In typically developed rodents, AVP signaling at hippocampal V1b receptors is critical for social recognition and social aggression.109, 110 One study of maternal restraint stress during pregnancy found that Avpr1a mRNA in the hippocampus of female offspring was positively correlated with the amount of maternal care they received.111 Still, much remains unknown as to what the implications of higher Avpr1a mRNA in females following DEP/MS are for behavioral outcomes.
We found no effects of treatment on Oxtr or Avpr1a mRNA in the AMY, LS, or vHipp. Of interest, while air pollution/stress effects on AMY V1aR have not previously been characterized, perinatal exposure to the pesticide chlorpyrifos and the plasticizer bisphenol A have both been shown to alter AVP and V1aR in the MeA.112, 113, 114 Similarly, Breach et al.73 found higher AVP‐ir fiber density in the MeA in both males and females following MIA. In contrast, Taylor et al.76 found lower AVP mRNA expression in the AMY following a different, LPS‐induced MIA. V1aRs in the AMY regulate social behaviors including social recognition in the MeA115, 116 and maternal aggression in the central AMY.117, 118 One limitation of our approach is that our tissue punches of the AMY include all subregions (medial, central, and basolateral). Therefore, specific investigation of receptor mRNA and/or protein within the various subregions of the AMY and extended AMY is an important avenue for future research.
We observed sex differences in both Oxtr and Avpr1a mRNA in the LS with higher expression in females than in males for both. These sex differences reflect previously observed sex differences in these receptors in the LS. Specifically, Smith et al.68 found higher Oxtr binding density in the dorsal LS and higher V1aR binding density in the intermediate LS in female rats as compared to male rats using receptor autoradiography. This is in line with the previously discussed sex difference in Oxtr in the NAc, which is also consistent with previous work using receptor autoradiography.60 A limitation of the present study is the quantification of mRNA which does not necessarily translate to the protein level. However, the consistency of our findings with previous work using receptor autoradiography supports the idea that our observed effects on mRNA could be indicative of protein level changes. Future work will aim to characterize differences in V1aR in the NAc and AMY using methods that directly quantify protein expression.
CONCLUSIONS
5
In conclusion, our results have several important implications. First, our finding of behavioral changes following prenatal exposure to DEP/MS in both males and females underscores the need to strongly consider the effects of neurodevelopmental insults on females in addition to males. Our observation of different behavioral effects in each sex following DEP/MS highlights the need to expand our behavioral repertoire to fully capture male‐ and female‐specific responses to prenatal toxicant and stress exposures. We also find specific effects of prenatal DEP/MS exposure on Avpr1a mRNA expression in the NAc, in the absence of robust changes in Oxtr mRNA or changes in hypothalamic OXT‐ir and AVP‐ir cells or fibers. Overall, this work suggests that prenatal exposures to combined air pollutants and MS have long‐term impacts on behavior and V1aR expression in the social brain.
AUTHOR CONTRIBUTIONS
MCS and CJS designed the study and wrote the paper. MCS, EMM, and CJS analyzed the data. MCS performed the DEP/MS exposures and collected the maternal and litter outcome data. MCS, EMM, CJS, JX, and JTB performed the behavioral experiments and MCS, JTB, JX, and SB conducted behavioral scoring. MCS performed immunohistochemistry and MCS, JTB, and JX conducted OXT‐ir and AVP‐ir cell and fiber quantification. EMM, MJK, NYL, and MFW extracted RNA, made cDNA, and performed qPCR and analysis. CJS supervised all phases of the project.
FUNDING INFORMATION
This work was supported by the National Institute of Environmental Health Sciences (NIEHS) grant R00ES033278 to CJS.
CONFLICT OF INTEREST STATEMENT
The authors declare no conflicts of interest.
Supporting information
Figure S1. Effects of DEP/MS exposure on male sociability and female social novelty preference by litter. (A–D) Male behavior measures in the sociability assay plotted by litter. Each dot represents an individual animal, and dots are color‐coded by litter. In the sociability assay, there were four litters used for the VEH/CON male group and five litters used for the DEP/MS males group. (E–H) Female behavior measures in the social novelty preference test plotted by litter. Each dot represents an individual animal, and dots are color‐coded by litter. In the social novelty preference test, there were 4 litters used for the VEH/CON female group and 4 litters used for the DEP/MS female group.
Figure S2. Effects of DEP/MS exposure on Oxtr and Avpr1a mRNA by litter. (A–J) mRNA expression for each gene and brain region plotted by litter. Each dot represents an individual animal, and dots are color‐coded by litter. In the VEH/CON groups (both males and females), there were seven litters used. In the VEH/CON female group, one subject from one of the litters was housed with non‐littermates and was non‐distinguishable so that animal is represented separately in this plot. In the DEP/MS groups, five or six litters were used (males vs. females). One litter represents individuals from two litters that could not be distinguished because they were not ear tagged and were co‐housed. In the DEP/MS female group, one subject from one of the litters was housed with non‐littermates and was non‐distinguishable so that animal is represented separately in this plot.
Table S1. Social behavior outcomes by litter. (A–D) Seconds spent in each behavior in the sociability test by litter. (E–H) Seconds spend in each behavior in the social novelty preference test by litter. Each symbol and color combination represents a different litter within each treatment group.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Fuller R , Landrigan PJ , Balakrishnan K , et al. Pollution and health: a progress update. Lancet Planet Health. 2022;6(6):e 535‐e 547. doi:10.1016/S 2542-5196(22)00090-0 35594895 PMC 11995256 · doi ↗ · pubmed ↗
- 2Rentschler J , Leonova N . Global air pollution exposure and poverty. Nat Commun. 2023;14(1):4432. doi:10.1038/s 41467-023-39797-4 37481598 PMC 10363163 · doi ↗ · pubmed ↗
- 3Southerland VA , Brauer M , Mohegh A , et al. Global urban temporal trends in fine particulate matter (PM 2·5) and attributable health burdens: estimates from global datasets. Lancet Planet Health. 2022;6(2):e 139‐e 146. doi:10.1016/S 2542-5196(21)00350-8 34998505 PMC 8828497 · doi ↗ · pubmed ↗
- 4Earnshaw VA , Rosenthal L , Lewis JB , et al. Maternal experiences with everyday discrimination and infant birth weight: a test of mediators and moderators among young, urban women of color. Ann Behav Med. 2013;45(1):13‐23. doi:10.1007/s 12160-012-9404-3 22927016 PMC 3562380 · doi ↗ · pubmed ↗
- 5Jbaily A , Zhou X , Liu J , et al. Air pollution exposure disparities across US population and income groups. Nature. 2022;601(7892):228‐233. doi:10.1038/s 41586-021-04190-y 35022594 PMC 10516300 · doi ↗ · pubmed ↗
- 6Dutheil F , Comptour A , Morlon R , et al. Autism spectrum disorder and air pollution: a systematic review and meta‐analysis. Environ Pollut. 2021;278:116856. doi:10.1016/j.envpol.2021.116856 33714060 · doi ↗ · pubmed ↗
- 7Kinney DK , Miller AM , Crowley DJ , Huang E , Gerber E . Autism prevalence following prenatal exposure to hurricanes and tropical storms in Louisiana. J Autism Dev Disord. 2008 a;38(3):481‐488. doi:10.1007/s 10803-007-0414-0 17619130 · doi ↗ · pubmed ↗
- 8Kinney DK , Munir KM , Crowley DJ , Miller AM . Prenatal stress and risk for autism. Neurosci Biobehav Rev. 2008 b;32(8):1519‐1532. doi:10.1016/j.neubiorev.2008.06.004 18598714 PMC 2632594 · doi ↗ · pubmed ↗