Microglial responses to inflammatory challenge in adult rats altered by developmental exposure to polychlorinated biphenyls in a sex-specific manner

Abstract

Polychlorinated biphenyls are ubiquitous environmental contaminants linked with peripheral immune and neural dysfunction. Neuroimmune signaling is critical to brain development and later health; however, effects of PCBs on neuroimmune processes are largely undescribed. This study extends our previous work in neonatal or adolescent rats by investigating longer-term effects of perinatal PCB exposure on later neuroimmune responses to an inflammatory challenge in adulthood. Male and female Sprague-Dawley rats were exposed to a low-dose, environmentally relevant, mixture of PCBs (Aroclors 1242, 1248, and 1254, 1:1:1, 20 μg / kg dam BW per gestational day) or oil control during gestation and via lactation. Upon reaching adulthood, rats were given a mild inflammatory challenge with lipopolysaccharide (LPS, 50 μg / kg BW, ip) or saline control and then euthanized 3 hours later for gene expression analysis or 24 hours later for immunohistochemical labeling of Iba1+ microglia. PCB exposure did not alter gene expression or microglial morphology independently, but instead interacted with the LPS challenge in brain region- and sex–specific ways. In the female hypothalamus, PCB exposure blunted LPS responses of neuroimmune and neuromodulatory genes without changing microglial morphology. In the female prefrontal cortex, PCBs shifted Iba1+ cells from reactive to hyperramified morphology in response to LPS. Conversely, in the male hypothalamus, PCBs shifted cell phenotypes from hyperramified to reactive morphologies in response to LPS. The results highlight the potential for long-lasting effects of environmental contaminants that are differentially revealed over a lifetime, sometimes only after a secondary challenge. These neuroimmune endpoints are possible mechanisms for PCB effects on a range of neural dysfunction in adulthood, including mental health and neurodegenerative disorders. The findings suggest possible interactions with other environmental challenges that also influence neuroimmune systems.

1. Introduction

The concept of Developmental Origins of Health and Disease (DOHAD) as applied to the brain recognizes that early life organization of neural systems can occur during critical or sensitive periods of development to set the trajectory for brain function and dysfunction later in life. The brain can regulate developmental processes via neuroimmune mechanisms (Uweru and Eyo 2019). In the developing brain, microglia sculpt synapses in an activity-dependent manner (Tremblay et al. 2010, Paolicelli et al. 2011, Rogers et al. 2011, Miyamoto et al. 2016, Weinhard et al. 2018), regulate cell numbers (Cunningham et al. 2013, Ueno et al. 2013) and guide neurite outgrowth (Pont-Lezica et al. 2014, Squarzoni et al. 2014). Microglia are also essential to adult brain function, as they constantly survey local environments to clear dying cells and defend against injury and pathogens while continuing to support and regulate neural activity (Nimmerjahn et al. 2005, Sierra et al. 2010, Bachstetter et al. 2011, Parkhurst et al. 2013, Badimon et al. 2020). Disruption to typical microglial functions can lead to diseased states in adulthood, as altered or enhanced inflammation is implicated in the progression of neurodegenerative diseases (Block et al. 2004, Hirsch and Hunot 2009, Liu et al. 2022, Moca et al. 2022) and neuropsychiatric disorders (van Berckel et al. 2008, Jones and Thomsen 2013, Whale et al. 2019, Carrier et al. 2022). Given that microglia are sensitive to the local milieu of pathogen and damage associated molecular patterns, hormones, and secreted factors (Bollinger et al. 2019, Perez-Pouchoulen et al. 2019, Villa et al. 2019, VanRyzin et al. 2020), they are vulnerable to perturbations by exposure to exogenous agents that may mimic or alter effects of immune signals, hormones, and neurotransmitters, including environmental contaminants.

Polychlorinated biphenyls (PCBs) are a class of persistent organic pollutants detectable throughout ecosystems that could alter microglial activity in animals, including humans (Thompson and Boekelheide 2013, Pavuk et al. 2014, Grimm et al. 2015, Ngoubeyou et al. 2022). Although intentional production was banned in the U.S. in 1977, PCBs continue to be present in the food chain (Kostyniak et al. 2005, Ampleman et al. 2015), are released from aging building materials over time (Rudel et al. 2008, Thomas et al. 2012), and are newly created as unintentional biproducts in pigment production (Hu and Hornbuckle 2010). The lipophilic compounds cross the placental barrier, concentrate in breastmilk, and are present in dust ingested via hand-to-mouth behavior, causing individuals to be heavily exposed in infancy (Kostyniak et al. 1999, Axelrad et al. 2009, Koh et al. 2015, Mitro et al. 2015, Panesar et al. 2020, Witczak et al. 2022). Critically, PCB exposure is linked to neurodevelopmental, mental health, neurodegenerative, immunological, and inflammation-linked hypertensive dysfunction in humans (Weisglas-Kuperus et al. 2000, Heilmann et al. 2010, Stølevik et al. 2013, Lyall et al. 2017, Kim et al. 2018, Pavuk et al. 2019, Pessah et al. 2019, Raffetti et al. 2020, Carlson et al. 2022, Sprowles et al. 2022, Wu et al. 2022).

Microglia may play a critical role in mediating interacting effects of PCBs on neural and immune systems. Animal and in vitro models have demonstrated effects of PCBs on neurons, including but not limited to initial effects on dopamine signaling, steroid and thyroid hormone receptor activity, intracellular calcium homeostasis, and oxidative stress (Fonnum and Mariussen 2009, Hamers et al. 2011, Sethi et al. 2019, Klocke et al. 2020), resulting in altered neuronal signaling, plasticity, and viability (Seegal et al. 1991, Morse et al. 1996, Mariussen and Fonnum 2001, Yang et al. 2009, Lee et al. 2012, Wayman et al. 2012). In addition, PCBs impact inflammatory processes throughout the body. While links between circulating PCBs and inflammatory markers in epidemiological studies are still under investigation (Turyk et al. 2015, Zota et al. 2018, Pavuk et al. 2019), they are associated with increased pro-inflammatory cytokine expression in stimulated adult human peripheral blood monocytes (Ghosh et al. 2011, Wens et al. 2011, Kuwatsuka et al. 2014, Leijs et al. 2019). Animal models and cell lines demonstrate that PCB exposure generally increase expression of proinflammatory factors, though effects depend on congener type and cell type (Kwon et al. 2002, Miller et al. 2010, Kim et al. 2012, Petriello et al. 2018, Rude et al. 2019, Wang et al. 2019). However, non-coplanar PCBs can also reduce phagocytic activity of human monocytes and reduce macrophage activation in murine cells (Levin et al. 2005, Santoro et al. 2015). Given the role of microglia in sculpting neural function and development, more research is needed to understand effects of PCBs on neuroimmune signaling.

Understanding effects of PCBs on microglia across the lifespan is critical to our understanding of disease etiology. Microglia can be primed in a two-hit manner, with exposure to pro-inflammatory signals such as maternal immune activation or stress early in life (a ‘first hit’) being associated with an exaggerated response to a secondary challenge (‘second hit’), even later in life (Bland et al. 2010, Schwarz and Bilbo 2012, Niraula et al. 2017, Makinson et al. 2019) which can then precipitate disease onset (Perry et al. 2007, Calcia et al. 2016, Frank et al. 2016). Studies in our lab have demonstrated that PCBs alter expression of genes associated with neuroimmune systems in a sex-dependent manner in the neonatal and adolescent brain (Bell et al. 2018, Liberman et al. 2020). However, our understanding of the effects of early life PCB exposure on neuroimmune endpoints in adulthood in both male and female adults is still incomplete; see (Sipka et al. 2008, Miller et al. 2010, Hayley et al. 2011, Rude et al. 2019, Matelski et al. 2020, McCann et al. 2021). In this study, we tested two complementary hypotheses: that perinatal exposure to PCBs a) alters basal levels of neuroimmune activity, and b) alters the neuroimmune response to a later ‘second hit’ inflammatory challenge in a sexdependent manner in adult rats. We also examined changes in corticotropin-releasing hormone, dopaminergic, opioid, and serotonergic signaling because of their bidirectional interactions with neuroimmune processes and relevance to brain health (Block et al. 2004, Yan et al. 2015, Varodayan et al. 2018, Hodo et al. 2020, Cuitavi et al. 2023). Comparing the results from the current study with those in neonatal and adolescent animals’ brains (Bell et al. 2018, Liberman et al. 2020) may reveal ages at which the brain is uniquely vulnerable to additional neuroimmune challenge and shed light on potential mechanisms of contaminant-associated developmental origins of health and disease.

2. Materials and Method

2.1. Experimental design

Here, we conducted two experiments on rats exposed perinatally to PCBs to investigate neuroimmune responses to LPS in adulthood (Figure 1). Prior work on their neonatal and adolescent littermates has been published (Bell et al. 2018, Liberman et al. 2020). Pregnant Sprague Dawley rats were exposed to a mixture of PCBs or oil-control via ingestion, throughout pregnancy until parturition. Offspring were challenged with a lipopolysaccharide (LPS) injection to stimulate a proinflammatory response, or saline-control, prior to euthanasia and tissue collection. Experiment One (Exp 1) investigated neuroimmune and dopaminergic gene expression, and serum concentrations of cytokines, estradiol, and corticosterone at The University of Texas at Austin. Experiment Two (Exp 2) followed up neuroimmune effects in a separate cohort by quantifying microglial morphology with immunohistochemistry (IHC) at DePaul University, Chicago IL. Selected brain regions of interest were prefrontal cortex, striatum, hypothalamus, and midbrain. These are hormone-sensitive brain regions in which we demonstrated altered neuroimmune and dopamine signaling in PCB exposed rats given LPS in adolescence (Liberman et al. 2020). Current work extended those findings to adulthood to determine the persistence of observed effects, and/or the manifestation of new phenotypes during adult development.

Figure 1: Experimental timelines.

For Exp 1 and 2, following mating on E0 (embryonic day 0), dams were fed wafers containing PCBs (20 μg/kg) or oil-control throughout gestation (E1-E22/23). Pups were born on P0 (postnatal day 0) and raised to adulthood, P90. From each litter, one male and one female were randomly assigned to receive an immune challenge, LPS (lipopolysaccharide, 50 ug/kg, i.p.) or saline control 3.5 hours prior to tissue collection for Exp 1 and 24 hours for Exp 2. Animals from Exp 1 were used to investigate gene expression data, and animals from Exp 2 were used for immunohistochemistry.

2.2. Breeding and husbandry

All animal protocols were approved by the Institutional Animal Care and Use Committee (IACUC) at each respective institution, and experiments were done in accordance with the Guide for the Care and Use of Laboratory Animals and the ARRIVE guidelines (Kilkenny et al. 2010). For both experiments, adult Sprague Dawley rats were purchased from Harlan/Envigo Laboratories, (Houston, Texas for Exp 1; Indianapolis, Indiana for Exp 2) to use as breeders. All animals were housed in a temperature-controlled room (21–23 °C) with a 12-hour light/dark cycle with 2–3 animals per cage. Animals were fed a low phytoestrogen, fishmeal-free Global Diet (Harlan-Teklad 2019, Indianapolis, Indiana) and received DI water from glass bottles with metal sippers ad libitum. Facilities had different caging systems that introduced slight differences in husbandry details. For Exp 1, animals were housed in polycarbonate cages (43 × 21 × 25 cm) with aspen bedding (PJ Murphy Forest Products, Sani-Chip) and were provided with a 5–10 cm long section of polyvinyl chloride pipe for habitat enrichment; for Exp 2, animals were housed in polysulfone cages (47 × 25 × 20 cm) with pine shaving bedding and with irradiated cardboard tubes for enrichment.

In both experiments, rats were acclimated for at least two weeks through daily handling prior to mating. Virgin females (3–4 months old) were paired overnight with a male rat (approximately 6 months old). Each male sired no more than one litter per PCB or oil-control treatment groups. Successful mating was determined via a sperm-positive vaginal smear, after which dams were singly housed, randomly assigned to treatment group in a counterbalanced design, and began receiving oil or PCB treatment, as described in section 2.3. Dams were provided cotton nestlets several days before the expected day of birth and oil and PCB treatment stopped the morning pups were observed, P0 (postnatal day 0). On P1, pups were labeled with a permanent marker and weighed; litter size was standardized to 6–8 pups with equal sex ratio by culling pups with extreme mass and anogenital index measures. Analyses of P1 and P40 siblings from Exp 1 have previously been published (Bell et al. 2018, Liberman et al. 2020). For Exp 2, all animals that remained after P1 cull were raised to ~P90 and used for IHC analysis.

From P7 on, pups were weighed, relabeled, and handled for at least five minutes weekly until adulthood in both experiments. On P21, pups were weaned and housed with 2–3 same-sex littermates per cage. In adulthood (between P84 and 92), offspring were randomly assigned to receive saline control (Sal) or immune challenge (LPS) prior to euthanasia. For Exp 1, 24 total litters (12 PCB litters) were split across three cohorts separated by 1–8 months, and no more than one rat of each sex per litter was used for an endpoint in a Sal or LPS group. For Exp 2, 10 litters (5 PCB litters) were split across two cohorts separated by 2 months; no more than two rats per sex per litter were used for an endpoint within a Sal or LPS group, as in (Bell et al. 2016, Bell et al. 2016). In both experiments, PCB treatment was evenly dispersed across cohorts, and cohort number did not affect gene expression or microglial morphology. The experimenters used coded vials of PCB and LPS solution so as to stay unaware of animal exposure and treatment assignment throughout the duration of both experiments.

2.3. PCB Exposure

PCB exposure was performed as described in (Bell et al 2018 and Liberman et al. 2020). A 1:1:1 ratio of Aroclor 1242, 1248, and 1253 was selected to represent the range of congeners present in the environment and food, predominately noncoplanar congeners with 2–6 chlorine substitutions (Hites et al. 2004, Kostyniak et al. 2005). This mixture includes ‘indicator congeners’ 28, 52, 95, 101, 118, 138, 153, 180 (Frame et al. 1996), including the most commonly detected congeners in human samples, excepting PCB 11 (Sethi et al. 2019, Zota et al. 2018). The dose selected was 20 μg/kg dam body weight (BW), consumed orally to represent typical human exposure routes. While PCB concentrations in tissues were not measured due to logistical limitations, this dose was chosen to mimic that of infants in heavily exposed human populations. This was estimated according to measures of PCBs in human breast milk, adipose tissue, and maternal to infant transmission (Grandjean et al. 1995, Lanting et al. 1998, Stellman et al. 1998, Dewailly et al. 1999, Dekoning and Karmaus 2000) and the relationships between exposure dose and resulting body burden in rats (Kodavanti et al. 1998, Hany et al. 1999).

Aroclors 1242, 1248, and 1254 were purchased from AccuStandard, New Haven, Connecticut: C-242-N-50MG, CAS# 53469–21-9 (Lot # 01141 Exp 1 and 01141-A Exp 2); C-248 N-50MG, CAS# 12672–29-6 (Lot# F-110 Exp 1 and A7080364 Exp 2); C-254 N-50MG, CAS# 11097–69-1 (Lot# 5428 Exp 1 and 24–191-B Exp 2). Of note are the different lots of Aroclors between experiments; while the range of congeners present is likely similar, measures of specific congeners may vary slightly. PCBs were suspended in Vegetable Oil (Crisco, Exp 1) or sunflower seed oil (chosen to match the fatty acid profile of diet, Whole Foods brand, Exp 2), stored in glass vials, and mixed thoroughly prior to use each day. Dams were fed PCB- or oil-treated quarters of ‘Nilla wafers, and were weighed daily to determine appropriate volume of oil-control or PCBs. Exposure to the dam began the morning following mating, as confirmed by sperm presence in the vaginal smear and deemed embryonic day 0 (E0), and continued on weekdays until pups were born (P0), 17–18 total prenatal days. Exposure to the pups continued postnatally via lactation due to residual maternal body burden until rats were weaned (P21) (Hashimoto et al. 1976, Takagi et al. 1986). Litter and developmental effects of PCBs are discussed in Bell et al 2018 and Liberman et al 2020.

2.4. Inflammatory Challenge

In adulthood (P82–92), LPS was used to induce a temporary proinflammatory reaction. Female estrous cycles were tracked via vaginal smears over at least two weeks and the LPS challenge was given on the day when nucleated cells predominated, on proestrus. Rats received intraperitoneal injections of LPS (E. coli 0111:B4, Sigma, L4391) or sterile saline and were immediately returned to their home cage. A low dose of 50 ug/kg was selected to induce a mild response (Shanks and Meaney 1994, Shanks et al. 1995, Nilsson et al. 2002, Hayley et al. 2011) so that any effects of PCBs in potentiating a neuroimmune response could be observed and compared to experiments done in younger animals (Bell et al. 2018, Liberman et al. 2020). Injections occurred during the first four hours of rat’s dark phase in Exp 1 to best observe sickness behaviors, but during mid-light phase in Exp 2 for logistical reasons. Rats were checked for sickness behaviors (piloerection, ptosis, and lethargy (Hart 1988) at 1, 2, and (for Exp 2) 24 hours following injection. For each behavior, scores of 0 (absence of behavior), 1 (mild), or 2 (severe) were recorded independently, as described in (Kentner et al. 2006). Scores of the three behaviors were summed within one time point so that scores could range from 0–6. Scores at 1- and 2-hours post-injection were averaged across time points within an animal; no sickness behaviors were visible at 24 hours post injection in Exp 2.

2.5. Tissue collection

To quantify changes in gene expression for Exp 1, rats were euthanized via rapid decapitation 3–4 hours post LPS injection. Brains were quickly removed from the skull, chilled on ice, and sectioned coronally using a rat brain matrix into 1- or 2-mm sections of the prefrontal cortex and hypothalamus, respectively. From these sections, triangular dissections were collected with razor blades and transferred into individual RNase-free microcentrifuge tubes, quickly frozen on dry ice, and stored at −80 °C until use. Trunk blood samples were collected and allowed to clot for 30 min before centrifugation (1500 ×g for 5 min). Sera were collected and stored at −80 °C until use. Adrenals and gonads were also dissected out and weighed to indicate gross organ function.

To quantify microglial morphological activation in Exp 2, rats were euthanized 24 hours after LPS injection, within the standard range of peak morphological change (Hoogland et al. 2015). Animals were deeply anesthetized with an i.p. injection of 2,2,2, tribromoethanol (300 mg/kg) in 2-methyl-2-butanol (%5 in normal sterile saline), as other injectable anesthetics are known to affect immune endpoints (Bette et al. 2004). Suitable analgesia was determined by lack of response to a strong foot pinch. Using a peristaltic pump set to flow 10 ml/minute, animals were rinsed with cold PBS (>150 ml) and then 4% paraformaldehyde (female >200 ml, male >350 ml) to fix tissue. Following perfusion, brains were harvested and stored in 4% paraformaldehyde overnight (4 °C) and transferred to 30% sucrose for storage (4 °C) until use.

2.6. Serum corticosterone and estradiol quantification

In Exp 1, total serum corticosterone was determined via a radioimmunoassay (ImmuChem Double Antibody ¹²⁵I RIA Kit, MP Biomedicals LLC, Orangeburg NY). Samples were run in duplicate 100 μl volumes in a single assay, according to manufacturer directions. Assay limit of detection was 14.4 ng/ml and intra-assay C.V. was 1.99%. Three outliers from different groups were identified via Grubb’s test and so were removed.

Also in Exp 1, total serum estradiol (E2) was determined in male and female rats via radioimmunoassay (UltraSensitive Estradiol RIA, Cat No DSL4800, Beckman Coulter, Pasadena, CA), according to manufacturer directions. Exp 1 females were all in proestrus on day of euthanasia and serum collection. Samples were run in duplicate volumes of 200 μl in a single assay. Assay limit of detection was 2.5 pg/ml and intra-assay C.V. was 2.02%.

2.7. Serum cytokine quantification

In Exp 1, serum samples were thawed on ice and diluted in assay buffer. The Milliplex Cytokine/Chemokine Hormone assay (RECYTMAG-65K) was run according to manufacturer directions. This assay contains interleukin (IL)- 1a, 1b, 4, 6, 10, interferon gamma (IFNγ), and tumor necrosis factor (TNF, also known as TNFα), which were selected based on literature and assay availability. Samples (25 μl) were run in duplicate across two plates and experimental groups were represented evenly between plates. Fewer than 30% of the samples within each group were above limits of detection for IFNγ, IL1a, and IL4 and so were not analyzed further. Variability (%CV) of sample replicates within an assay and quality control values between assays were as follows, respectively: IL1b (2.13%, 8%), IL6 (1.38%, 8%), IL10 (1.15%, 5%), and TNF (6.37%, 11%). Two animals were removed from the ANOVA analysis of the TNF assay because their results were orders of magnitude higher than the rest of their Female PCB Saline and Male PCB LPS groups. Limits of detection for IL1b, IL6, IL10, and TNF were 3.96, 73.41, 3.49, and 3.73 pg/ml, respectively.

2.8. RNA isolation and gene expression quantification

In Exp 1, RNA was isolated from the hypothalamus as previously described (Bell et al. 2018). In brief, tissue was homogenized and extracted using the Qiagen mini RNeasy protocol and treated with DNase while on the column. RNA yield was determined with a NanoDrop^™ Spectrophotometer, and randomly chosen samples were analyzed to confirm RNA integrity was above 9 with a Bioanalyzer 2100 (Agilent Technologies). RNA was reverse transcribed to cDNA using a high-capacity cDNA reverse transcriptase kit with RNase inhibitor (ThermoFisher, Cat # 4374967). 48 genes were quantified using custom designed microfluidic Taqman Low Density Array (TLDA) cards (Applied Biosystems, Cat # 4342253) with Taqman Gene Expression Mastermix (Applied Biosystems, Cat # 4369016). A list of the genes in this Neuroimmune Panel is shown in Table 1. All procedures were completed in consultation with the MIQE guidelines (Bustin et al. 2009) and gene assay details are included in (Bell et al. 2018). The samples were run at 50 °C for 2 min, 95 °C for 10 min, 45 cycles of 95 °C for 15 s, and 60 °C for 1 min using a ViiA 7 qPCR system (Applied Biosystems), which automatically determined the quantification cycle (Cq) of each sample. Gapdh, Rpl3a, and 18s were included as reference genes and their geometric mean was used to normalize sample Cq values to calculate the relative expression of target genes to female oil and saline-control groups (Schmittgen and Livak 2008).

Table 1.

Neuroimmune Panel: Summary of main effects and interactions of sex, PCB exposure, and/or LPS challenge on adult hypothalamic gene expression.

Gene	Transcript	Sexes Combined	Within Females	Within Males
Xenobiotic signaling
AhR	aryl hydrocarbon receptor			* Sal < LPS
Arnt	aryl hydrocarbon receptor nuclear translocator
Arntl	aryl hydrocarbon receptor nuclear translocator-like
Neuroimmune signaling
Ccl22	chemokine (C-C motif) ligand 22	** Sal < LPS	* Sal < LPS	** Sal < LPS
Cxcl9	chemokine (C-X-C motif) ligand 9	** Sal < LPS	** Sal < LPS	** Sal < LPS
Cybb	cytochrome b-245, beta (Nox2)		* Sal < LPS
Ifna1	interferon-alpha 1
Ikbkb	inhibitor of kappa light polypeptide gene enhancer in B-cells, kinase beta
Il1a	interleukin 1 alpha	** Sal > LPS	** Sal > LPS	** Sal > LPS
Il1b	interleukin 1 beta	** Sal < LPS	** Sal < LPS	** Sal < LPS
Il6	interleukin 6	** Sal < LPS		** Sal < LPS
Il7r	interleukin 7 receptor	** Sal < LPS	* Sal < LPS	** Sal < LPS
Itgam	integrin, alpha M (CR3A, CD11b)
Itgb2	integrin, beta 2 (CD18)			* Sal < LPS
Map3k7	mitogen activated protein kinase kinase kinase 7 (TAK1)			* Sal < LPS
Myd88	myeloid differentiation primary response gene 88	** Sal < LPS		** Sal < LPS
Nfkb1	nuclear factor of kappa light polypeptide gene enhancer in B-cells 1 (pre p50)	* Sal < LPS
Ptges	prostaglandin E synthase	** Sal < LPS	** Sal < LPS	** Sal < LPS
Ptgs2	prostaglandin-endoperoxide synthase 2 (COX-2)	** Sal < LPS	** Sal < LPS	** Sal < LPS
Rela	v-rel reticuloendotheliosis viral oncogene homolog A (p65)	** Sal < LPS	** Sal < LPS	** Sal < LPS
Tlr4	toll-like receptor 4	* PCB × LPS ** Sex × PCB × LPS	** PCB × LPS
Tnf	tumor necrosis factor	** Sal < LPS * Sex × PCB	** Sal < LPS	** Sal < LPS
Neuroimmune modulators
Arrb1	arrestin, beta 1
Tgfb2	transforming growth factor, beta 2
Endocrine enzymes and receptors
Ar	androgen receptor
Crh	corticotropin releasing hormone	** Male < Female * Sex × PCB
Cyp19a1	cytochrome P450, family 19, subfamily a, polypeptide 1 (aromatase)	** Female < Male * Sal <LPS	* PCB × LPS	* Sal < LPS
Esr1	estrogen receptor 1	** Male < Female
Esr2	estrogen receptor 2
Dopamine enzymes, receptors, and transporters
Drd1a	dopamine receptor D1A	* Sal < LPS * PCB × LPS	* Sal < LPS
Drd2	dopamine receptor D2	** Sal < LPS	* Sal < LPS
Th	tyrosine hydroxylase
Slc6a3	solute carrier family 6, member 3 (dopamine transporter)
Opioid precursors and receptors
Oprk1	opioid receptor, kappa 1	** Male < Female * Oil > PCB
Oprm1	opioid receptor, mu 1
Pdyn	prodynorphin
Pomc	proopiomelanocortin
Serotonin enzymes, receptors, and transporters
Htr1a	5-hydroxytryptamine receptor 1A	* Female < Male	* Oil > PCB
Htr2a	5-hydroxytryptamine receptor 2A	** Sal < LPS		* Sal < LPS
Tph1	tryptophan hydroxylase 1
Slc6a4	solute carrier, family 6, member 4 (serotonin transporter)

Significant effects (* p < 0.05; ** p < 0.01) are noted in sexes combined and within each sex, with cells containing PCB effects highlighted in gray.

RNA was extracted from PFC samples with TRIzol^™ (Cat # 15596026) per manufacturer directions and subsequently treated with TURBO^™ DNase (Cat # AM2238). Expression of Nfkb1, Rela, Ikbkb, and Tlr4, genes previously impacted by PCB exposure in sister experiments (Bell et al. 2018, Liberman et al. 2020) were quantified with Taqman Assays (Cat # 4414095) and Taqman Gene Expression Mastermix (Cat # 4369016) with Gapdh as a reference gene using a Quantstudio 6 qPCR system and analysis as described above.

2.9. Immunohistochemistry (IHC) of Iba1

In Exp 2, brains were sectioned coronally into six series at 30 μm using a cryostat and stored in an ethylene glycol-based cryoprotectant at −20°C until use. Rinses and incubations occurred on a shaker at room temperature unless otherwise noted. One whole 1:6 tissue series from each animal was transferred to individual glass scintillation vials (20 ml, 2 cm diameter base). Tissue was washed (0.5 M PBS) and then permeabilized and blocked for 60 minutes using 0.3% hydrogen peroxide in 40% methanol, 10% Normal Goat Serum (NGS, Pel-Freez Biologicals 32130–5), and 0.3% Triton-X solution. Then tissues were incubated in primary antisera (Wako, Rabbit anti-Iba1, 019–19741, 1:10,000) in 2% NGS and 0.3% Triton-X for 48 hours at 4°C, washed, and incubated with secondary antisera (Biotinylated goat, anti-rabbit H+L Vector, CAT # BA-1000–1.5, 1:500 dilution) in 2% NGS and 0.3% triton X for 60 minutes. The tissues were again washed, treated with Avidin-biotin complex (Vector Labs PK-4000), 60 min, rinsed again, and reacted with Diaminobenzadine chromagen (MP Biomedicals Cat # 980681, 0.02% with 0.01% peroxide). The reaction was monitored to optimize staining intensity, and tissue was stored at 4°C until additional processing. Sections were mounted onto subbed slides, dehydrated with increasing concentrations of ethanol, defatted with xylenes, and coverslipped with Permount (Fisher Scientific, SP15–500).

2.10. Microglial morphological analysis

Microglial morphological analysis was conducted on 2–4 sections of the prelimbic PFC (Bregma 3.72–2.76) and medial preoptic area of the hypothalamus (Bregma −0.24 - −0.60), per (Paxinos and Watson 2009) (Figure 2). Images from the left and right PFC and dorsal and ventral aspects of the preoptic area were collected with a 20x objective using a Leica MC170 HD camera on a Leica DM2000 microscope. Microglia were counted and categorized with ImageJ cell counter plugin. Microglial morphology, including the number and thickness of projections, as well as soma size and shape, was used to categorize microglia as ramified, hyperramified, reactive, and phagocytic, as shown in Figure 2 and discussed in (Ransohoff and Perry 2009, Bollinger et al. 2019). The number and percentage of microglia of the different categories within an image were averaged across all 4–6 images within an animal.

Figure 2:

Iba1-stained tissue was imaged in three sections of A) the bilateral prelimbic PFC and B) the dorsal/ventral unilateral extent of one hemisphere of the media preoptic area. Microglia were categorized according to their morphology. Ramified cells (C) had a moderate number of thin projections that were radially distributed around a small, round soma. Hyperramified cells (D) had longer, bushier projections with slight polarity and more irregularly shaped somas. Reactive microglia (E) had less than three root projections emanating from opposite sides of a large, oblong soma. Phagocytic microglia (F) had large and more spherical somas lacking visible projections. Brain outlines adapted from Paxinos and Watson 6^th ed; microglial schematic drawings by K.A. Walker.

2.11. Analysis and statistics

Individual animals were the unit of analysis, with 9–10 animals produced per group per experiment. Data were combined from both Exp 1 and 2 when available: body weight and sickness behavior. Grubbs tests were used to identify outliers within each group prior to analysis. In the hypothalamus and prefrontal cortex, three and four animals from separate experimental groups had 4 or more genes identified as outliers, and so were removed from further gene expression analysis. Final sample size shown in figures and tables. Graphpad Prism 9, SPSS 25 and R 3.6.2 were used for data analysis depending on availability and investigator expertise. Main effects and interactions are shown as *PCB, ^Sex, and ^#LPS, p < 0.05.

Effects of PCBs on morphometric data were assessed within sex, using a two-way repeated model (PCB × age) for body weight and t-tests for adrenal and gonadal weight at euthanasia. Serum measures, gene expression, and microglial morphological data were first analyzed to determine effects and interactions between PCB exposure, LPS challenge, and sex via a three-way analysis of variance test (ANOVA). Because of our a-priori hypotheses regarding sex-specific effects of PCBs, all outcomes were also analyzed with two-way ANOVAs within sex. Any significant interactions were followed up with an unpaired t-test within specific variables, or non-parametric equivalent tests when appropriate. Levene’s test, visual examination of residual plots, and Shapiro Wilks tests were used to confirm data met parametric assumptions. Data that failed to meet homogeneity of variance assumptions were analyzed with a Welch one-way ANOVA that identified main effects of sex, PCB, or LPS by collapsing across other groups. In this case, if an interaction was indicated in the three- or two-way ANOVA, ‘group’ was used as a factor. Only groups exposed to LPS showed non-normal distributions; because ANOVAs can tolerate violations to this assumption well and LPS effects were large, no additional tests were performed. Datasets on genes (Il6, Ccl22, and Cxcl9) and serum cytokines (IL-1b, IL-6, IL-10) where at least one group had fewer than 25% of samples amplify (CT < 35) or were above limits of detection were analyzed with a Fisher’s exact test to determine percentage of samples detectable.

3. Results

3.1. Morphometric and Sickness Behavior

PCB exposure did not cause statistically significant main effects on male or female body weight (Figure 3). A PCB × Age interaction was detected in males (F_{(4, 284)} = 2.56, p < 0.05). While it appears that PCB exposed males were slightly heavier than controls at specific ages, the source of the interaction could not be determined.

Figure 3:

Effects of PCBs on body weight, presented as mean values ± SEM, *p < 0.05. A significant PCB × Age interaction was found, but specific age at which PCB exposure caused greater body weight could not be determined.

Adult males exposed to PCBs had significantly lower gonadal weights (9.90 g +/− 0.17) compared to oil-exposed animals (10.45 g +/− 0.17) (t₍₃₃₎ = 2.28, p < 0.05). PCB exposure did not alter gonadal weight in females, or adrenal weight in either males or females.

As expected, LPS challenge caused significantly greater display of sickness behaviors within 2 hours after injection (F_{(1, 139)} = 124.4, p < 0.01). LPS animals showed 1–2 mild behaviors (piloerection, lethargy, or ptosis), with avg +/− SE as follows: 1.72 +/− 0.22 and 2.06 +/− 0.20 in females and males, respectively. This is in contrast to 0.19 +/− 0.08 and 0.09 +/− 0.04 in saline treated females and males. No significant main effects of PCBs or sex or interactions between were seen.

3.2. Serum estradiol, corticosterone and cytokine analysis

Serum was analyzed for circulating hormones and cytokines (Figure 4). Serum estradiol concentrations were not affected by PCB exposure but showed main effects of LPS and Sex (Fig. 4a). Animals challenged with LPS showed lower concentrations of estradiol when the sexes were combined across males and females (F_(1,61) = 4.24, p < 0.05), an effect that was not detected within either sex. Females had higher concentrations of serum estradiol than males, independent of PCB or LPS treatment (F_(1,61) = 29.41, p < 0.01).

Figure 4: PCBs did not alter circulating serum hormones or cytokines.

LPS challenge decreases concentrations of circulating estradiol (A) and increases corticosterone (B) and TNFα (C). LPS challenge also increased the number of samples that were above limits of detection for IL-1β (D), IL-6 (E), and IL-10 (F). Females had higher circulating estradiol and corticosterone than males. Data are presented as mean values ± SEM with individual data points superimposed (A-C) or as the number of samples above or below limited of detection per group (D-F). Main effects and interactions are shown as ^Sex, and ^#LPS, p < 0.05.

Serum corticosterone concentrations were, similarly, not affected by PCB exposure but showed main effects of LPS and Sex (Fig. 4b). Animals exposed to LPS had higher concentrations of corticosterone than those exposed to saline across sexes (F_(1,59) = 115.05, p < 0.01), in females (F_(1,29) = 40.73, p < 0.01) and in males (F_(1,30) = 91.56, p < 0.01). Females also had higher concentrations of corticosterone than males, independent of PCB or LPS treatment (F_(1,59) = 43.92, p < 0.01).

Serum cytokine concentrations were also not affected by PCBs but showed main effects of LPS (Fig. 4c–F). Animals challenged with LPS showed greater concentrations of TNFα across sex (F_(1,59) = 20.21, p < 0.01), within females (F_(1,27) = 10.96, P < 0.01), and within males (F_(1,27) = 24.08, p < 0.01). LPS challenge was also associated with a greater number of samples with detectable concentrations of IL-1β in females but not males, and in IL-6 and IL-10 in both males and females, independent of PCB exposure (χ², p < 0.01, for all).

3.2. Gene expression in the hypothalamus

Effects of, and interactions between, PCB exposure, LPS challenge, Sex on gene expression are summarized in Table 1.

PCB exposure altered expression of two of eighteen genes associated with proinflammatory signaling in the hypothalamus (Figure 5a, b). Analysis of Tlr4 expression revealed a Sex x PCB x LPS interaction (F_(1,68) = 8.21, p < 0.01). A PCB x LPS interaction was observed in females (F_(1,33) = 15.47, p < 0.01) but not males. In oil-exposed females, LPS decreased Tlr4 expression (t₍₁₅₎ = 3.27, p < 0.01), but in PCB-exposed females, LPS increased Tlr4 expression (t₍₁₅₎ = 2.25, p < 0.05). Thus, PCB exposure reversed effects of LPS on Tlr4 expression in females. Analysis of Tnf revealed a PCB x Sex interaction (F_(1,68) = 4.17, p < 0.05) independent of LPS treatment. It appears that PCB treatment caused lower Tnf expression in females and greater expression in males, however the source of this interaction could not be resolved statistically. Main effects of LPS were also observed on Tnf expression across sex (F_(1,68) = 49.83, p < 0.01), within females (F_(1,33) = 15.88, p < 0.01) and within males (F_(1,34) = 39.93, p < 0.01), irrespective of PCB exposure.

Figure 5: Effects of PCBs on neuroimmune, neuroendocrine, and neurotransmitter endpoints within the hypothalamus.

Images show tissue dissected from two 1-mm sections. A) PCB exposure reversed the effect of LPS on Tlr4 expression in females but not males. B) A PCB x Sex interaction on Tnf expression was detected across sexes, but the source of the interaction could not be determined. LPS increased expression of Tnf in males and females. C - E) Expression of Crh, Cyp19a1, and was greater in females than males. C) The effect in Crh was qualified by an interaction, such that the sex effect was only present within oil, not-PCB, exposed animals. D) Cyp19a1 expression was greater in rats exposed to LPS as compared to saline control in males and females combined. LPS increased expression in oil-control females, but not in PCB-exposed females; a main effect of LPS was observed in males. PCB exposure F) prevented effects of LPS on Drd1a expression G) decreased expression of Oprk1 across sexes, and H) decreased Htr1a expression in females. Expression of G) Oprk1 was greater in males than females and H) Htr1a was greater in females than males. Data are presented as mean values ± SEM with individual data points superimposed. Main effects and interactions are shown as *PCB, ^Sex, and ^#LPS, p < 0.05, across and/or within a sex.

PCBs altered expression of genes associated with endocrine enzymes and receptors in sex-specific ways in the hypothalamus (Figure 5c, d, e). Analysis of Crh expression revealed an effect of sex (F_(1,68) = 7.33, p < 0.01) that was qualified by a PCB x Sex interaction (Fig 5c, F_(1,68) = 4.07, p < 0.05), independent of LPS treatment: oil-exposed females showed greater expression of Crh than males, but PCB exposure prevented this typical sex difference. No main effects of PCBs or LPS were observed. To further examine potential mechanisms of sexspecific effects, expression of aromatase (Cyp19a1) was examined. Cyp19a1 expression analysis revealed a main effect of sex, such that males had greater expression than females (Fig 5d, F_(1,68) = 56.73, p < 0.01). A main effect of LPS on Cyp19a1 expression was also observed across sex (F_(1,68) = 6.74, p < 0.05) such that LPS-challenged animals showed greater expression than saline controls. This LPS effect was also observed in males (F_(1,34) = 5.12, p < 0.05), irrespective of PCB exposure, but was not observed in females. Within females, an interaction of PCBs and LPS was observed (F_(1,33) = 6.60, p < 0.05); in oil-exposed females, there was an effect of LPS (t₍₁₅₎ = 2.50, p < 0.05), which was absent in PCB-exposed females. Analysis of Esr1 revealed a main effect of sex (Fig 5e, F_(1,68) = 15.46, p < 0.01), where females showed greater expression than males, independent of PCB exposure or LPS challenge. No main effects of PCBs or LPS or interactions on Esr1 expression were observed. No effects of any variable were found on expression of Ar or Esr2.

In the hypothalamus, PCB exposure had only moderate effects on expression of enzymes and receptors for opioid, dopamine and serotonin (Figure 5f, g, h). Analysis of Drd1a revealed a main effect of LPS across sex (Fig 5f, F_(1,68) = 6.21, p < 0.05), such that LPS-challenged animals had greater expression than saline controls. This effect was present within females (F_(1,33) = 6.33, p < 0.05) but not males, and qualified by a PCB x LPS interaction across sex (F_(1,68) = 4.24, p < 0.05). PCB exposure blocked the normal effect of LPS within oil-exposed animals (t₍₃₃₎ = 3.23, p < 0.01) across sex. No other effects of PCBs were observed on other dopaminergic endpoints. Analysis of Oprk1 revealed a main effect of PCB across sex (Fig 5g, F_(1,68) = 4.60, p < 0.05), where animals exposed to PCBs had lower levels of Oprk1 expression, independent of sex or LPS treatment. A main effect of sex was also observed (F_(1,68) = 11.89, p < 0.01) such that females showed greater expression of Oprk1 than males. No effects of PCBs on Oprk1 expression within sex, or other opioid endpoints, were found. Analysis of Htr1a expression revealed a main effect of sex (Fig 5h, F_(1,68) = 4.76, p < 0.05), such that males had greater expression independent of PCB exposure and LPS treatment. Within a sex, PCB exposure also decreased expression of Htr1a in females (F_(1,33) = 4.45, p < 0.05) but not in males. No other effects of PCBs on serotonin endpoints were found.

Independent of PCB exposure (with no PCB x LPS interactions), LPS altered gene expression of various neuroimmune endpoints (Table 2). For six proinflammatory genes (Il1b, Il7r, Ptges, Ptgs2, and Rela), LPS increased expression across and within sex. Additional LPS effects were specific to a sex or only observed across both sexes: LPS increased expression of Nfkb1 across sex; Myd88 across sex and within males; Cybb in females only; and Ahr, Itgb2, and Map3k7 in males only. For one gene (Il1a), LPS decreased expression across and within sex. Additionally, LPS altered the proportion of samples that were quantifiable, including greater percent amplification of Ccl22 and Cxcl9 across and within sex and of Il6 across sex and within males. Like Drd1a discussed above, LPS increased expression of Drd2 across sex and within females, and also increased expression of Htr2a within males.

Table 2:

Significant Effects of LPS on gene expression in hypothalamus, independent of PCB.

	Across Sexes		Within Females		Within Males
Gene	RE (Mean±SE)	Statistics	RE (Mean±SE)	Statistics	RE (Mean±SE)	Statistics
Xenobiotic signaling
Ahr	Sal: 0.94 ± 0.03	ns, F(1,68) = 1.85	Sal: 1.00 ± 0.06	ns, F(1,33) = 0.00	Sal: 0.90 ± 0.04	^F(1,34) = 4.44
	LPS: 1.02 ± 0.04		LPS: 1.00 ± 0.06		LPS: 1.05 ± 0.06
Neuroimmune signaling
Ccl22	Sal: 30.30%	^^ χ2 (1) = 24.81	Sal: 46.67%	^χ2 (1) = 5.40	Sal: 16.67%	^^χ2(1) = 21.13
	LPS: 88.89%		LPS: 84.2%		LPS: 94.12%
Cxcl9	Sal: 45.45%	^^χ2 (1) = 20.07	Sal: 33.33%	^^χ2 (1) = 11.57	Sal: 55.56%	^^χ2 (1) = 9.79
	LPS: 94.44%		LPS: 89.47%		LPS: 100%
Cybb	Sal: 1.00 ± 0.04	ns, F(1,68) = 0.27	Sal: 1.03 ± 0.07	^F(1,33) = 6.31	Sal: 0.98 ± 0.5	ns, F(1,34) = 0.60
	LPS: 0.96 ± 0.07		LPS: 0.84 ± 0.04		LPS: 1.1 ± 0.1
Il1a	Sal: 1.09 ± 0.05	^^F(1,67) = 39.31	Sal: 1.06 ± 0.08	^^F(1,33) = 12.38	Sal: 1.11 ± 0.05	^^F(1,33) = 30.48
	LPS: 0.68 ± 0.05		LPS: 0.72 ± 0.06		LPS: 0.64 ± 0.07
Il1b	Sal: 1.22 ± 0.09	^^F(1,56) = 96.23	Sal: 1.2 ± 0.1	^^F(1,29) = 35.15	Sal: 1.2 ± 0.1	^^F(1,26) = 61.08
	LPS: 7.1 ± 0.5		LPS: 6.1 ± 0.7		LPS: 8.3 ± 0.7
Il6	Sal: 54.55%	^^χ2 (1) = 8.33	Sal: 73.33%	ns, χ2 (1) = 1.503	Sal: 38.89%	^^χ2 (1) = 6.88
	LPS: 86.11%		LPS: 89.47%		LPS: 82.35%
Il7r	Sal: 0.94 ± 0.03	^^F(1,68) = 23.42	Sal: 0.97 ± 0.05	^F(1,33) = 4.82	Sal: 0.94 ± 0.04	^^F(1,34) = 28.24
	LPS: 1.22 ± 0.04		LPS: 1.18 ± 0.07		LPS: 1.28 ± 0.05
Itgb2	Sal: 0.95 ± 0.02	ns, F(1,68) = 3.07	Sal: 0.95 ± 0.04	ns, F(1,33) = 0.10	Sal: 0.94 ± 0.02	^F(1,34) = 6.54
	LPS: 1.01 ± 0.03		LPS: 0.98 ± 0.04		LPS: 1.05 ± 0.04
Myd88	Sal: 0.94 ± 0.02	^^F(1,68) = 13.94	Sal: 0.95 ± 0.04	ns, F(1,33) = 2.58	Sal: 0.93 ± 0.03	^^F(1,34) = 13.38
	LPS: 1.08 ± 0.03		LPS: 1.04 ± 0.04		LPS: 1.13 ± 0.04
Nfkb1	Sal: 1.05 ± 0.03	^F(1,68) = 4.82	Sal: 1.06 ± 0.04	ns, F(1,33) = 3.08	Sal: 1.04 ± 0.03	ns, F(1,34) = 1.82
	LPS: 1.14 ± 0.03		LPS: 1.16 ± 0.04		LPS: 1.12 ± 0.04
Ptges	Sal: 1.04 ± 0.04	^^F(1,68) = 138.77	Sal: 1.06 ± 0.06	^^F(1,33) = 44.62	Sal: 1.02 ± 0.06	^^F(1,34) = 175.27
	LPS: 5.8 ± 0.4		LPS: 6.0 ± 0.6		LPS: 5.5 ± 0.3
Ptgs2	Sal: 1.12 ± 0.05	^^F(1,68) = 147.39	Sal: 1.08 ± 0.08	^^F(1,33) = 49.65	Sal: 1.15 ± 0.08	^^F(1,34) = 123.96
	LPS: 3.0 ± 0.1		LPS: 2.9 ± 0.2		LPS: 3.1 ± 0.2
Rela	Sal: 0.96 ± 0.02	^^F(1,68) = 24.39	Sal: 0.96 ± 0.04	^^F(1,33) = 8.66	Sal: 0.96 ± 0.02	^^F(1,34) = 18.59
	LPS: 1.11 ± 0.02		LPS: 1.10 ± 0.03		LPS: 1.11 ± 0.03
Tnf	Sal: 1.13 ± 0.06	^^F(1,38) = 49.83	Sal: 1.1 ± 0.1	^^F(1,33) = 15.88	Sal: 1.18 ± 0.09	^^F(1,34) = 39.99
	LPS: 1.90 ± 0.09		LPS: 1.8 ± 0.1		LPS: 2.0 ± 0.1
Neuroimmune modulators
Map3k7	Sal: 0.98 ± 0.04	ns, F(1,68) = 2.85	Sal: 1.02 ± 0.08	ns, F(1,33) = 0.38	Sal: 0.96 ± 0.04	^F(1,34) = 4.93
	LPS: 1.09 ± 0.04		LPS: 1.09 ± 0.07		LPS: 1.10 ± 0.05
Endocrine enzymes and receptors
Cyp19a1	Sal: 1.9 ± 0.1	^F(1,68) = 6.74	Sal: 1.18 ± 0.09	ns, F(1,33) = 2.57	Sal: 2.5 ± 0.2	^F(1,34) = 5.12
	LPS: 2.4 ± 0.3		LPS: 1.4 ± 0.1		LPS: 3.4 ± 0.4
Dopamine enzymes, receptors, and transporters
Drd1a	Sal: 1.15 ± 0.08	^F(1,68) = 6.21	Sal: 1.05 ± 0.07	^F(1,33) = 6.33	Sal: 1.2 ± 0.1	ns, F(1,34) = 2.40
	LPS: 1.6 ± 0.2		LPS: 1.5 ± 0.1		LPS: 1.7 ± 0.3
Drd2	Sal: 0.95 ± 0.03	^^F(1,68) = 8.77	Sal: 0.93 ± 0.03	^F(1,33) = 7.25	Sal: 0.97 ± 0.06	ns, F(1,34) = 3.77
	LPS: 1.18 ± 0.06		LPS: 1.14 ± 0.06		LPS: 1.2 ± 0.1
Serotonin enzymes, receptors, and transporters
Htr2a	Sal: 1.02 ± 0.03	^^F(1,68) = 8.69	Sal: 1.04 ± 0.04	ns, F(1,33) = 2.61	Sal: 0.99 ± 0.03	^F(1,34) = 6.55
	LPS: 1.13 ± 0.03		LPS: 1.13 ± 0.03		LPS: 1.13 ± 0.04

Data were analyzed across sex via a three-way ANOVA and within each sex via a two-way ANOVA, or on the percent of samples that amplified (CT < 35) with a χ² test. Because the below effects are independent of PCBs, they are shown collapsed across oil and PCB groups and are presented as mean ± SEM. Significance at ^p < 0.05, ^^p < 0.01 are noted across sex and within sex, with genes where only sex showed a significant effect highlighted in gray.

3.3. Gene expression in prefrontal cortex

Effects of PCB exposure on expression of four neuroimmune genes, selected based on their adolescent sensitivity to PCB effects, were determined in the prefrontal cortex (Figure 6). Main effects of LPS were observed across sex, irrespective of sex and PCB exposure on Ikbkb (Fig 6a, F_(1,66) = 6.42, p < 0.05), Nfkb1 (Fig 6b, F_(1,67) = 28.04, < 0.01), and Rela (Fig 6c, F_(1,67) = 25.08, < 0.01) expression. These LPS effects were also present within each sex, with the exception of expression of Ikbkb in males. Analysis of Rela expression revealed a PCB x Sex interaction (Fig 6c, F_(1,67) = 5.65, p < 0.05), such that PCB exposure increased expression in females (F_(1,33) = 6.49, p < 0.05) but not in males. Expression of Tlr4 was also affected: a main effect of sex was observed (Fig 6d, F_(1,68) = 5.33, p < 0.05) independent of PCB exposure and LPS treatment. In addition, a PCB x LPS interaction was observed across sex (F_(1,68) = 4.38, p < 0.05), as animals exposed to PCBs showed greater expression than oil controls, but only non-LPS challenged animals (F_(1,32) = 5.63, p < 0.05). However, PCB x LPS interactions were not detected within sex.

Figure 6: PCB exposure increased Rela and Tlr4 expression within the PFC.

Image shows tissue dissected from one 2 mm section. LPS exposure increased expression of A) Ikbkb, B) Nfkb1, and C) Rela, independent of PCB and sex. Expression of C) Rela was greater in PCB exposed females compared to oil-controls, while no effect was observed in males. Expression of D) Tlr4 was also greater in animals exposed to PCBs compared to oil-controls, but only in saline-control animals. Data are presented as mean values ± SEM with individual data points superimposed. Main effects and interactions are shown as *PCB, ^Sex, and ^#LPS, p < 0.05.

3.4. Microglial morphology in the hypothalamus

Effects of PCBs on morphology of Iba1-positive microglia at rest or after LPS challenge in the hypothalamus is shown in Figure 7, including representative images from PCB treated males (Fig 7A–B). The presence of phagocytic cells was very uncommon (average of less than 1 cell per section) and no effects of PCBs or LPS were seen on this outcome, so data are not shown. No main effects of PCB, LPS, or Sex on the total number of microglia or on specific morphologies were found (Fig 7C–F). However, interactions between the factors on the percentages of microglia of distinct morphologies were present, as follows.

Figure 7. PCB exposure altered responses to LPS in male hypothalamus.

Photomicrographs of images representative of Iba1 staining in A) Saline and B) LPS challenged PCB-treated male brains is shown under 20x objective. No significant effects of PCB, LPS, or Sex were seen on the number of C) total, D) ramified, E) hyperramified, F) reactive microglia or G) percentage of ramified microglia. Significant PCB x LPS x Sex interactions were found on the percentage of H) hyperramified and I) reactive cells. PCBs caused a smaller percentage of cells to be hyperramified and a greater percentage to be reactive, but only in LPS-challenged males. This inverse relationship is J) a significant negative correlation across all animals and K) summarized with a schematic drawing. Data are presented as mean values ± SEM with individual data points superimposed. *, # indicate effects of PCBs and LPS, respectively; p < 0.05.

While the percentage of ramified microglia was not affected by PCBs or LPS individually (Fig 7G), the percentage of hyperramified microglia showed a significant PCB x LPS x Sex interaction (Fig 7H F_(1,64) = 4.10, p < 0.05). In males, an interaction between PCB exposure and LPS challenge was detected (F_(1,31) = 5.50, p < 0.05): PCB-exposed (but not saline-control) males challenged with LPS showed a smaller percentage of hyperramified microglia compared to saline-control males (p < 0.05). No significant effects on the percentage of hyperramified microglia were seen in females. The percentage of reactive microglia also showed a significant PCB x LPS x Sex interaction (Fig 7I, F_(1,64) =4.36, p < 0.05). In males, an interaction between PCB exposure and LPS challenge was detected (F_(1,31) = 4.87, p < 0.05); males exposed to PCBs showed a greater percentage of reactive microglia when challenged with LPS compared to saline-control males. No significant effects on the percentage of reactive microglia were seen in females. This pattern is the inverse of that found with hyperramified cells. There is a strong negative correlation between the percentage of cells that are hyperramified and reactive within samples across all groups (Fig 7J, r(70) = −0.93, p < 0.01), with LPS driving a shift from hyperramified to reactive cells in PCB exposed males (Fig 7K).

3.5. Microglial morphology in the prefrontal cortex

Effects of PCBs on morphology of Iba1-positive microglia at basal states or after LPS challenge in the prefrontal cortex is shown in Figure 8, including representative images from LPS challenged females (Fig 8A–B). As in the hypothalamus, phagocytic cells were very uncommon and unaffected by treatment, and so data are not shown. No effects of Sex, PCB or LPS were seen on total number of microglia (Fig 8C), but interactions between these factors were present within distinct morphologies. The number of ramified microglia showed a significant PCB x LPS interaction (Fig 8D, F_1,64 =4.62, p < 0.05) across sexes, however, follow-up tests could not resolve the source of this interaction. There was no effect of PCB or LPS on the percentage of ramified microglia (Fig 8G).

Figure 8. PCB exposure drives atypical response to LPS in female PFC.

Photomicrographs of images representative of Iba1 staining after LPS challenge in A) Control and B) PCB treated female brains is shown under 20x objective. C) No significant effects of Sex, PCB or LPS were seen on overall number of microglia. D) A PCB x LPS interaction of undetermined source was found on the number of ramified cells, G) with no effect on the percent of cells that were ramified. E) PCBs did increase the number of hyperramified cells, but only in females exposed to LPS, H) an effect that was mirrored in the percentage of hyperramified microglia. F) While no effects of PCBs were observed on the number of microglia that were reactive, G) a reduction in the percentage of microglia that were reactive was observed in percentage of microglia that were reactive, only in LPS-challenged females. This inverse relationship is J) a significant negative correlation across all animals and K) summarized with a schematic drawing. Data are presented as mean values ± SEM with individual data points superimposed. *p < 0.05, **p < 0.01

Across sex, there was a significant interaction between PCB exposure and LPS challenge on the number of hyperramified cells (Figure 8E, F_1,64 =5.87, p < 0.05); PCB exposure increased the number of hyperramified cells compared to Oil-controls, but only in LPS-challenged rats (p < 0.05). This effect is specific to females, as only females showed a PCB x LPS interaction (F_1,34 =7.88, p < 0.01), with LPS challenge revealing effects of PCBs (p < 0.01). This was mirrored in the percentage of hyperramified cells (Fig 8H), where a PCB x Sex x LPS interaction was observed (F_1,64 =4.49, p < 0.05). This effect was also specific to females, which showed a significant PCB x LPS interaction on the percentage of microglia that were hyperramified (F_1,34 =4.12, p = 0.05). Specifically, females exposed to PCBs had a larger percentage of reactive cells, but only after LPS challenge (p < 0.05). There was also a significant main effect of sex on the average number of hyperramified cells in the PFC (F_1,64 =5.81, p < 0.05), with females having more than males.

Effects on reactive microglia were inverse to those of hyperramified cells. Across sex, there was a significant interaction between PCB exposure and LPS challenge on the percent of reactive cells (Figure 8G, F_1,64 = 4.93, p < 0.05). Within sex analysis shows that this PCB x LPS effect is present in females (F_1,34 = 7.25, p < 0.05) and not in males, such that females exposed to PCBs had a smaller percentage of reactive microglia compared to oil-controls, but only when challenged by LPS (p < 0.05). No significant effects were found on the number of reactive cells. While less pronounced than in the hypothalamus, there is also a significant inverse correlation between percentages of hyperramified and reactive cells within samples across all groups in the prefrontal cortex (Fig 8J, r(70) = −0.43, p < 0.01), with PCBs driving a shift from reactive to hyperramified in LPS challenged females (Fig 8K).

4. Discussion

This is the final report in a series of three studies designed to identify the effects of perinatal PCB exposure on neuroimmune gene expression in response to a postnatal immune challenge across neonatal, adolescent, and adult life stages (Bell et al. 2018, Liberman et al. 2020) (summarized in Figure 9). The current results demonstrate that impacts of early life exposure persist into adulthood but are not just a continuation or maintenance of earlier effects. More specifically, whereas analysis of hypothalamic tissue sampled earlier in life revealed significant main effects of PCBs, most effects of PCBs in adulthood occurred as interactions with LPS treatment. In the prefrontal cortex, the gene targets affected by PCBs, and the direction of the effects, were consistent between adolescent and adult time points, however they were detected earlier in males than females. Effects on gene expression, while not large in magnitude, are notable because they were relatively low to approximate human exposure and present over 70 days after the last acute exposure to PCBs via lactation. Most effects were sex-specific, and observed in either hypothalamic or prefrontal cortex regions, as discussed below. Previous studies have detected effects of PCBs on neuroimmune gene expression endpoints (Sipka et al. 2008, Miller et al. 2010, Hayley et al. 2011, Rude et al. 2019, Matelski et al. 2020, McCann et al. 2021). However, as many of these genes can be expressed by neurons and astrocytes, the present morphological Iba1 data provide the first demonstration of PCBs affecting microglia in adult brains. If true, this raises a novel target for potential therapeutics for individuals with high PCB exposure.

Figure 9: Effects of early life PCBs that occur independent of LPS (“Oil </> PCB”) or interact with the secondary inflammatory challenge (“PCB x LPS”) across development.

Neonatal and adolescent results are summarized from Bell et al 2018 and Liberman et al 2020. Effects specific to either females or males are shown in pink or blue text; outcomes that were not significant or measured are shown as “ns” or “−“, respectively. Early in life, when PCBs are likely still in circulation from recent acute exposure, PCBs altered serum cytokine responses to inflammatory challenge; no serum effects were found in adulthood. Hypothalamic neuroimmune gene expression also shifted over development. Early in life, PCB exposure generally increased neuroimmune gene expression; in adulthood, PCBs effects were revealed as shifting a response to a secondary inflammatory challenge, with females more affected than males. Hypothalamic neuromodulatory genes also revealed effects of PCBs differentially across development. Frontal cortex neuroimmune responses to LPS were altered by PCBs in both adolescence and adulthood, with the sexes differing in the age at which this disruption is revealed.

4.1. PCBs alter hypothalamic gene expression differentially over development.

In the current study, PCBs increased expression in 2 of 18 neuroimmune genes assessed in the hypothalamus. Specifically, perinatal exposure to PCBs caused greater expression of Tlr4 and lower expression of Tnf in females in response to the LPS challenge. TLR4 is the main receptor that detects LPS associated with bacterial infections, which could drive the resulting cytokine response including TNFa expression. The mechanisms of the effects observed within this study, with potentially greater TLR4 mediated sensitivity to LPS co-occurring with a lower TNF cytokine response to LPS, is unknown. We should note that the brain cytokine response to LPS in this study could be mediated by indirect relay of the peripheral inflammatory signal, and not mediated by brain TLR4. However, when seeking to understand the integrated physiological impacts of PCB exposure on neuroimmune processes, we can also consider co-occurring alterations to genes that are affected by or modulate immune processes, as follows.

Inflammatory challenge is typically associated with an increase in expression of genes that code for corticotropic releasing hormone (Crh) (Berkenbosch et al. 1987, Ericsson et al. 1994) and aromatase (Cyp19a1) that may modulate the inflammatory response (Dean et al. 2012, Pedersen et al. 2017). Notably, these responses were observed in oil-control females in the current study but were blunted by PCB exposure. Dopamine, opioid, and serotonin neurotransmitter systems and receptors are also known to help downregulate inflammation (Alicea et al. 1996, Färber et al. 2005, Krabbe et al. 2012, Glebov et al. 2015, Yan et al. 2015, Wu et al. 2019, Broome et al. 2021, Missig et al. 2022). PCBs blocked the increase in Drd1a expression in response to LPS in both males and females in the hypothalamus and caused overall lower expression of Oprk1 and Htr1a. As such, the absence of these responses in PCB exposed animals indicates a dysregulated, and potentially blunted, hypothalamic neuroimmune response to inflammatory challenge.

When compared with neonatal and adolescent observations, these current results demonstrate that perinatal PCB effects can be differentially revealed across developmental periods (Fig 9). One possibility for these life stage-specific effects could be a reduction of direct brain tissue exposure to PCBs over time, after external intake has ceased upon weaning and as enzymatic metabolism and adipose tissue sequestration have continued (McPhail et al. 2016). For example, PCBs altered circulating IL1β and IL6 in neonates and IL1β in adolescents, but no serum cytokines in adulthood (Bell et al. 2018, Liberman et al. 2020), perhaps indicating a reduction in circulating PCBs. Changes in blood brain barrier permeability over development could also be associated with age-specific exposures to both PCBs and circulating LPS or LPS-induced cytokines (Anthony et al. 1997, Stolp et al. 2005). At the same time, brain cells might be differentially sensitive to ongoing effects of PCBs at different ages. Two of several possible mechanisms by which PCBs affect cells include induction of CYP-450 enzyme activity and resulting ROS production, and potentiating ryanodine receptor dependent calcium signaling, reviewed in (Pessah et al. 2019). Critically, ryanodine receptor expression, calcium handling, mitochondrial metabolism, and antioxidant enzyme activity appear to change over time (Mori et al. 2000, Abu-Omar et al. 2018, Serpa et al. 2021), which could drive different responses to similar PCB exposure over time.

In addition, it is also possible that a unique suite of long-term effects occur in response to earlier acute effects and are only revealed in an age-specific way. For example, effects of PCBs on dopaminergic enzymes and transporters were detected in neonates, while receptors were affected in adolescents and adults. This may indicate plasticity-dependent compensation for earlier effects within neural signaling systems over time, as has been indicated in previous studies (Lein et al. 2007). It also raises the possibility that earlier exposure to PCBs programs microglia to respond to LPS differentially in adulthood in subtle ways, as occurs in response to gonadal and stress steroid hormones or immune activity (Cronk et al. 2015, Amit et al. 2016, Villa et al. 2019, VanRyzin et al. 2020). Indeed, microglia experience epigenetic regulation of their activity over development in response to local environments that could be disrupted by environmental challenges (Matcovitch-Natan et al. 2016, Ayata et al. 2018). Results from newborn rats (Bell et al. 2018) show that during the early critical period, PCB exposure increases hypothalamic cytokines and decreases genes related to the production and handling of immune modulatory neurotransmitters. It could be that PCB exposure has altered neuroimmune systems in ways that became developmentally entrenched, or PCBs could be modulating epigenetic machinery directly.

4.2. Early life PCB exposure alters prefrontal cortex gene expression in response to adult inflammatory challenge.

Analysis of the prefrontal cortex focused on a narrower subset of targets than the hypothalamus that had been identified in adolescent studies; two of these four genes studied were affected by PCBs. PCB treated females showed greater expression of Rela than same sex-controls. Given that Rela codes for the p65 subunit of the NFĸB transcription factor, this could indicate the potential for larger cytokine responses to secondary inflammatory stimuli. PCB exposed males and females also showed greater expression of Tlr4 than oil controls, but this effect was not present after LPS challenge. While intraperitoneal LPS challenge can increase brain Tlr4 expression (Wang et al. 2014, Iwasa et al. 2015), many microglial specific experiments show that LPS causes a reduction in Tlr4 expression while expression of inflammatory cytokines increase (Loram et al. 2012, Marinelli et al. 2015, Turano et al. 2017, Lively et al. 2018). As such, gene expression results may indicate a PFC that is more responsive to secondary LPS challenge.

This prefrontal response is unique from the hypothalamic response, as is predicted by the literature (Bollinger et al. 2017, Walter et al. 2017, Reichelt et al. 2021). Recent analysis of mRNA expression and cellular behavior suggests that several microglial functional states exist and are modified by their local environments within specific brain region in rodents (Doorn et al. 2015, Grabert et al. 2016, Furube et al. 2018) and humans, as reviewed in (Tan et al. 2020, Paolicelli et al. 2022). One of many possible mechanisms behind region-specific responses to PCBs in the current dataset is differential modulation of microglia by dopamine systems. Dopamine has both pro- and anti-inflammatory actions (Yan et al. 2015, Nolan et al. 2020, Vidal and Pacheco 2020) and different dopamine systems are present in the prefrontal cortex (A10 terminals) and hypothalamus (A12, A14 cell bodies and local terminals). These systems are known to be affected by PCBs e.g. (Seegal et al. 1986), are differentially susceptible to oxidative stress (Yokoyama et al. 1993), and data from this series of studies demonstrate that they may also be differentially affected by PCBs. Here, PCBs altered dopamine receptor expression in the hypothalamus, as it did in the hypothalamus but not prefrontal cortex during adolescence (Liberman et al. 2020). Therefore, PCBs could be differentially modulating prefrontal or hypothalamic dopamine systems that then alter microglial activity in a region-specific manner.

4.3. PCB exposure alters microglial morphological responses to adult inflammatory challenge.

PCBs did not alter the number of total microglia present or shift morphology independent of secondary challenge. However, previous exposure did cause a change in the distribution of cells across hyperramified and reactive categories after LPS challenge. In the hypothalamus, LPS challenge shifted cells from the hyperramified into the reactive phenotype in PCB, but not Oil, -exposed males. The inverse was found in the prefrontal cortex, where the morphology after LPS was shifted from a reactive towards the hyperramified phenotype in PCB exposed females. Possible mechanisms for sex and brain region specific effects are described herein. While it is not possible to infer cellular activity from morphological state (Paolicelli et al. 2022), the results indicate long-lasting effects of PCB exposure on microglia-specific reactions to adult systemic inflammatory challenge.

Integrating the region-specific results of the gene expression and immunohistochemical experiments raises interesting challenges, as the affected sex differed between outcomes. This reinforces the need for a nuanced understanding of microglial ‘activation’, including multiple neuroimmune processes/phenotypes that may be altered by PCBs independently (Paolicelli et al. 2022). This observation could also be explained by methodological differences between gene expression and immunohistochemical experiments, including the type of plastic in animal cages, bedding, and the lots of Aroclors used. We have been as transparent as possible in the current study to clarify experimental conditions and environment to enable future replication. In addition, gene expression data includes RNA from the whole hypothalamus, but IHC analysis focused specifically on the POA. The differences in microglial populations within the hypothalamus are observed when comparing data from (Schwarz et al. 2012, Lenz et al. 2013), which could potentially explain the seemingly incongruent results. Finally, we must consider the differential timing between LPS challenge and euthanasia in the two experiments. To assay gene expression changes, animals were euthanized after ~2 hours while morphological changes were observed 24 hours after challenge. Given the dynamic nature of microglial function, we may be measuring microglia at different states along their activation and recovery profiles.

4.4. Sex differences in effects of PCBs are observed throughout development.

In the current adult study and as neonates, females appeared more affected by perinatal PCBs than males, whereas males were more vulnerable than females in adolescence (Fig 9) (Bell et al. 2018, Liberman et al. 2020). Sex-specific effects of PCBs on DNA methylation, gene expression, neuronal morphology, and behavior are consistently observed in other studies. However, the sex that demonstrates greater vulnerability to PCBs depends on additional factors like specific outcome, brain region, age, congener, and additional challenge (Dickerson et al. 2011, Tian et al. 2011, Bell et al. 2016, Gillette et al. 2017, Keil Stietz et al. 2021, Hilz and Gore 2022, Laufer et al. 2022). There are several possible reasons for sex-specific effects. One explanation could be sex-specific exposure: while PCBs accumulate in the brain similarly between males and females (Miller et al. 2010, Sethi et al. 2021, Li et al. 2022), differences in enzymatic activity between sexes could drive differential clearance or exposure to metabolites (Wu et al. 2013, Stamou et al. 2014, Nagai et al. 2016, Gochfeld 2017). Future research could assess induction of CYP1A, 1B, 2B or 3A enzyme families in response to PCBs to determine sex-dependent metabolism of different congener structures (Grimm et al 2015, Klocke and Lein 2020).

Alternatively, PCB could be affecting neuroendocrine processes that are sexually-differentiated, which later modulate neuroimmune function. Indeed, PCBs 138 and 153, expected to be high in the current mixture, are known to be anti-estrogenic (Hamers et al. 2011, Takeuchi et al. 2017, Pěnčíková et al. 2018). In the current study, PCBs altered Cyp19a1 in female hypothalamus, which could alter aromatase-dependent local production of estradiol not detected in circulating serum (Amateau et al. 2004). As estradiol regulates microglial function indirectly in neonates and directly in adulthood, it is possible that PCBs could be disrupting gonadal hormone regulation of microglial activity in a sex-specific manner (Dodel et al. 1999, Bruce-Keller et al. 2000, Dimayuga et al. 2005, Loram et al. 2012, Lenz et al. 2013, Villa et al. 2016). Similarly, the relationship between glucocorticoid, CRH, and neuroimmune activity may explain some of the current results. PCB 153 is known to be a glucocorticoid receptor antagonist (Takeuchi et al. 2017) and PCBs altered Crh expression in the current study. Glucocorticoid signaling modulates microglial responses (Sierra et al. 2008, Frank et al. 2019, Sugama and Kakinuma 2020) and may involve direct interactions between CRH and TLR4 (Zhang et al. 2016, Balan et al. 2018). The HPA axis and CRH system are sexually differentiated and estradiol sensitive (Deak et al. 2015, Bangasser and Wicks 2017, Wellman et al. 2018). Therefore, differential effects of PCBs on CRH systems could result in sex-specific neuroimmune outcomes.

Finally, sex differences in microglial development could drive differential sensitivity to direct effects of PCBs on microglia and vulnerability to later perturbation. Neuroimmune systems demonstrate sex differences in basal phenotypes and in responses to a range of stimuli that often depend on brain region and microenvironment (Schwarz et al. 2012, Nelson et al. 2017, Fonken et al. 2018, Guneykaya et al. 2018, VanRyzin et al. 2020, Han et al. 2021, Patten et al. 2022). Many of the genes influenced by PCBs in the current study show sex and brain region specific changes over development, including TLR4 and regulators of NFkB activity (Loram et al. 2012, Crain et al. 2013, Iwasa et al. 2015, Lively et al. 2018). Indeed, microglia appear to transition through sexually differentiated distinct maturational stages organized by differential transcriptional regulation (Hanamsagar et al. 2017).

4.5. Limitations

While this is the first demonstration of PCB effects on microglial cells, more work is still required to determine if microglia themselves are directly sensitive to PCBs or if the effects are relayed via nearby regulatory cells. Continued experiments are also required to confirm the physiological relevance of small gene expression effect sizes at a protein activity level, as TLR4 and NFkB activity are dynamically regulated post transcription (Carpenter et al. 2014, Christian et al. 2016). The current experiment was limited to exploring effects of a single dose of a legacy Aroclor mixture; future research could explore dose-dependent effects of non-legacy PCBs such as PCB 11, which would be complemented by assessing PCB body burdens and local concentrations within the brain.

4.6. Implications for human health

These results add to our understanding of the possible mechanisms by which early life PCB exposure has long-lasting effects on brain outcomes in adults (Chu et al. 2019). PCB exposure is associated with greater risk of autism spectrum disorder, depressive symptoms and potentially Parkinson’s disease in aging adults (Fitzgerald et al. 2008, Hatcher-Martin et al. 2012, Bohler et al. 2019, Raffetti et al. 2020, Simhadri et al. 2020, Carlson et al. 2022). While effects of PCBs on dopamine and thyroid hormone processes are possible drivers of these relationships, the current results and previously demonstrated links between microglial dysfunction and neurodevelopmental, neuropsychiatric, and neurodegenerative conditions suggest that PCB-associated microglial dysfunction could also play a role (Perry et al. 2003, Block et al. 2007, Prinz et al. 2019). The sex-specific effects observed in the current study also align with the sex differences in prevalence or symptomology of the above disorders (Hanamsagar et al. 2017, Bekhbat and Neigh 2018). Indeed, circulating PCBs are associated with altered expression of genes related to Parkinson’s Disease in females but not males (Bohler et al. 2019). It also raises the possibility that PCBs could be linked to other disorders with sex-differentiated presentations and developmental origins associated with microglial dysfunction, including vulnerability to substance use disorders (Fonken et al. 2018, McGrath and Briand 2019) and schizophrenia (Meyer et al. 2011, Rodrigues-Neves et al. 2022) that have not been well-explored. It is important to note that members of communities of color and low income communities often carry higher body burdens of PCBs and other environmental contaminants than the population average, and may therefore experience a disproportionately larger disease burden (Schantz et al. 2010, Nguyen et al. 2020, Payne-Sturges et al 2023).

4.7. Conclusion

This study emphasizes the need to test for toxicity in some form of a challenged state more representative of human lives; not doing so risks missing latent effects only revealed by a secondary ‘hit’ over a lifespan (Kraft et al. 2016). This is especially true in immune processes, where priming exists at a molecular level at inflammasomes (Bauernfeind et al. 2009, Haneklaus et al. 2013) and over development to alter health-related outcomes later in life (Shanks et al. 1995, Bilbo and Schwarz 2009, Giovanoli et al. 2013). This study also adds PCBs to the list of environmental contaminants that affect microglia, along with air pollution, bisphenol A, and lead in vivo (Levesque et al. 2013, Kumawat et al. 2014, Luo et al. 2014, Rebuli et al. 2016, Wise et al. 2016, Bolton et al. 2017, Mumaw et al. 2017) and perfluorinated alkylated substances (PFAS) in vitro (Zhu et al. 2015, Liu et al. 2021). While the field of toxicology is recognizing the need to study mixtures of chemicals, these results reinforce the idea that a complete understanding of environmental risks requires considering “mixtures” of other forms of environmental perturbations, especially during early life, on brain health outcomes (Hollander et al. 2020, Burkett and Miller 2021). This ‘exposome’ includes exposure to other factors that drive neuroimmune activity, including drugs of abuse, high sugar/fat diets, physical or social stress, and maternal immune activation; critically, exposure to one challenge could prime or blunt the impacts of another challenge later in life because of shared neuroimmune mechanisms, often in a sex-specific manner (Bilbo and Tsang 2010, McClain et al. 2011, Giovanoli et al. 2013, Schwarz et al. 2013, Thion et al. 2018, Bollinger et al. 2019, Bordeleau et al. 2020, Gildawie et al. 2020).

Highlights.

• Early PCB exposure alters neuroimmune responses to adult inflammatory challenge

• PCBs alter Tlr4 and Nfkb gene expression responses to LPS independent of cytokines

• PCBs alter hypothalamic dopamine, CRH, aromatase gene expression responses to LPS

• PCBs shift microglial morphological responses to LPS in hypothalamus and PFC

• PCB effects differed between sexes, brain regions, and from those at earlier ages