Maternal high-fat diet induces sex-specific effects on inflammatory responses to corticosterone and lipopolysaccharide challenge in adult rat offspring

Acute elevations in endogenous CORT caused by stress or exogenous administration potentiate inflammatory gene expression. Maternal obesity through high-fat diets (HFD) has been linked to higher basal levels of neuroinflammation, including increased expression of pro-inflammatory genes such as NFκB and interleukin-6 in the amygdala. These findings suggest that adult rats exposed to maternal HFD may elicit pro-inflammatory responses in the presence of an immune stressor and elevated CORT. To investigate, adult rat offspring exposed to maternal control fat house-chow diet (CFD) or HFD were administered exogenous CORT and lipopolysaccharide (LPS), a component of gram-negative bacteria. Transcript abundance of CORT receptors and downstream inflammatory genes were measured in the amygdala, hippocampus and prefrontal cortex, brain regions that mediate neuroendocrine and behavioural responses to stress. We found sex-specific responses, where HFD female offspring exhibited elevations in anti-inflammatory transcripts, and HFD male offspring responded with a larger pro-inflammatory response to simultaneous CORT and LPS administration. These findings suggest that exposure to maternal HFD leads to sex-specific alterations that could regulate the neural immune response, possibly as an adaptive response to basal inflammation.


INTRODUCTION
Obesity during pregnancy is an early life stressor that is becoming increasingly common in both developed and developing countries (Amugsi et al., 2017;Bhurosy and Jeewon, 2014;Forsyth et al., 2016). Maternal obesity can lead to metabolic and behavioural disorders involved with stress response in offspring (Catalano and deMouzon, 2015;Contu and Hawkes, 2017;King, 2006;Rivera et al., 2015;Van Lieshout et al., 2011;Zambrano et al., 2016). For example, epidemiological studies indicate that maternal obesogenic diets high in saturated fats are linked to increased risk anxiety-like disorders in offspring (Edlow, 2017;Rivera et al., 2015;Rofey et al., 2009).
The mechanisms by which exposure to maternal high-fat diet (HFD) alter inflammatory gene expression in brain regions involved with regulating anxiety behaviour is not well understood. In rodent models of maternal obesity, HFD exposure results in increased anxiety-like behaviour in the Open field and Elevated plus maze tasks at adulthood, accompanied by changes associated with the hypothalamic-pituitary-adrenal (HPA) axis, the neuroendocrine system that regulates stress responses (Bilbo and Tsang, 2010;Peleg-Raibstein et al., 2012;Sasaki et al., 2013;Sullivan et al., 2012). In events of stress, the HPA axis leads to the secretion of corticosterone (CORT) in rodents, which binds to glucocorticoid receptor (GR) and mineralocorticoid receptor (MR) to mediate behavioural and immune stress responses in limbic brain regions including the amygdala, hippocampus (HPC), as well as the prefrontal cortex (PFC) (Munck and Náray-Fejes-Tóth, 1992;Silverman and Sternberg, 2012). In basal conditions, CORT binds to GR and leads to the expression of anti-inflammatory genes including, nuclear factor kappa beta-inhibitor alpha (IBA) and mitogen-activated protein kinase phosphatase 1 (MKP-1) (Munhoz et al. 2010;Sorrells et al. 2009). However, acute elevations in CORT caused by stress potentiate pro-inflammatory gene expression (Munhoz et al. 2006(Munhoz et al. , 2010Sorrells et al. 2009). Systemic stressors such as lipopolysaccharide (LPS), a component of gramnegative bacteria, activate inflammatory processes via NFB, which leads to interleukin 6 (IL6), and cluster of differentiation molecule 11B (CD11B) expression.
The GR-NFB pathway that regulates inflammatory stress is altered in the brain of adult rats exposed to maternal HFD. In basal conditions, there are increases in pro-inflammatory gene expression of NFB and IL6 in the amygdala, along with downregulated serum CORT (Sasaki et al. 2013). These findings suggest that in adult rats exposed to maternal HFD, acute elevations in CORT will potentiate neuroinflammatory effects of LPS to a larger degree than in control animals. To investigate how maternal HFD impacts GR signalling and inflammatory gene expression in conditions of stress, adult female and male rat offspring exposed to maternal control or high-fat diet were administered exogenous CORT to simulate psychological stress, LPS to simulate systemic stress, or simultaneous CORT and LPS challenge. Transcript abundance of CORT receptors and downstream inflammatory pathway genes was measured in the amygdala, HPC, and PFC, limbic brain regions that mediate HPA axis responses to stress.
Females and males were examined separately due to prominent sex differences in body weight and response to HFD, CORT, and LPS exposures as determined in past literature (Ashdown et al., 2007;Bilbo and Tsang, 2010;Chistyakov et al., 2018;Sasaki et al., 2013;Seale et al., 2004) subcutaneous dose of CORT (10 mg/kg of body weight), 2) an intraperitoneal dose of LPS (50 μg/kg of body weight), 3) a simultaneous dose of CORT and LPS (10 mg/kg, 50 μg/kg), or 4) handled controls (n=6 per diet, sex, and treatment). A 10mg/kg dose of CORT was previously shown to lead to heightened anxiety-like behaviour (Mitra and Sapolsky, 2008) and circulating plasma CORT levels similar to that of several hours of acute physiological stress (Stein-Behrens et al., 1994). The 50 μg/kg dose of LPS was shown to activate the HPA axis within 0.5-4 h, as shown by increased CORT levels in whole-blood (Vore et al., 2017) and induce proinflammatory cytokine expression in the hippocampus of offspring exposed to maternal HFD (Bilbo and Tsang, 2010). Animals were sacrificed 3 h post injection by CO2 inhalation followed by rapid decapitation at the mid-point of the light phase (11-3 pm) to control for circadianrelated changes in gene expression. Brains were dissected, flash-frozen in isopentane and dry ice, and stored at -80 C.

RNA Extraction and cDNA Synthesis
Whole amygdala, dorsal hippocampus, and medial prefrontal cortex were cryo-sectioned using a Leica CM3050 cryostat and stereotaxic coordinates (Paxinos and Watson, 2007). RNA was extracted from the amygdala, dorsal hippocampus, and medial prefrontal cortex using TRIzol reagent (15596026; Invitrogen) in combination with RNeasy Plus Mini Kit (74134; Qiagen) as per the manufacturer's instructions. RNA quantification and quality assessments were done using a Nanodrop Spectrophotometer (ND-2000C; Thermo Scientific). 1 g of total RNA was converted to cDNA using High Capacity cDNA Reverse Transcription Kit (4368814; Applied Biosystems) according to the manufacturer's instructions.

Gene expression Analysis by RT-qPCR
Relative mRNA expression of glucocorticoid receptor (GR), mineralocorticoid receptor (MR), nuclear factor kappa light chain enhancer of activated B cells (NFκB), nuclear factor of kappa light polypeptide gene enhancer in B-cells inhibitor, alpha (IκBα), interleukin 6 (IL6), interleukin 10 (IL10), cluster of differentiation molecule 11B (CD11B), mitogen activated protein kinase phosphatase 1 (MKP1), and insulin-like growth factor 1 (IGF1) in the three brain regions was measured using a StepOne Plus real-time thermocycler with a Fast SYBR Green PCR master mix (4385612; Applied Biosystems). Primers were purchased from Qiagen or Eurofins Genomics and designed according to GenBank sequence information at the National Center for Biotechnology Information (NCBI) ( Table 1).
Relative gene expression was calculated using the quantity mean based on a standard curve of 11 serial dilutions ranging from 500 ng/L to 0.49 ng/L of cDNA. A standard curve was run per plate and per set of comparisons. Quantity means were normalized against the GEOmean of four reference genes, YWAZ, GAPDH, 18s, and Actin B. Reference genes were identified as stable internal controls based on geNORM analysis of stability across experimental groups, brain regions, and sex (Vandesompele et al., 2002).

Statistical Analysis
Body weight of adult offspring was analyzed by 3-way (diet x drug x sex) ANOVA using SPSS (IBM Corp,23). Transcript levels were analyzed using 1-way ANOVA with Scheffe post-hoc analysis using R Statistical Software (R Foundation for Statistical Computing, Vienna, Austria, 3.4.2). Relationships were considered statistically significant at p ≤ 0.05. Table 1. Primer sequences used in RT-qPCR analysis.

Offspring body weight
As reported previously by our group for animals in this cohort, offspring with maternal HFD exposure showed no differences in body weight at birth, however weighed more than CFD  (Sasaki et al., 2014;Sasaki et al., 2013). In adulthood (PND90), male offspring weighed more than female offspring (main effect of sex (F(1,93) = 220.20, p<0.001, Figure 2), and both sexes with maternal HFD exposure weighed more than CFD offspring (females (F(1,46) = 19.71 , p< 0.001, males (F(1,45) = 39.82, p< 0.01, Figure 2). There were no significant differences in body weight between animals assigned to each of the treatment groups (CORT, LPS, CORT+LPS, control handled).
In male offspring, we observed more differences in transcript abundance in response to CORT challenge ( Figure Figure 4K) in the hippocampus.

LPS challenge
In females, LPS challenge led to transcript abundance between HFD and CFD offspring ( Figure   5A, 5B). Both diet groups showed decreased GR in the amygdala (main effect of challenge  Figure 6J).

DISCUSSION
The current study builds upon previous findings that adult offspring exposed to maternal HFD show evidence of sex-specific increases in inflammatory gene expression in the amygdala and the hippocampus (Bilbo and Tsang, 2010;Sasaki et al., 2014;Sasaki et al., 2013). As expected in CFD animals, CORT potentiated pro-inflammatory gene expression induced by LPS in both female and male offspring. This potentiation was also observed in HFD animals, however compared to the CFD animals, there were elevated levels of anti-inflammatory transcript in females, and a larger pro-inflammatory response in males. These findings suggest that maternal HFD-exposed adult rats have sex-specific alterations to the programming of pro-and antiinflammatory genes.

Enhanced anti-inflammatory responses to CORT challenge in male HFD offspring
CORT is classically associated with anti-inflammatory and immunosuppressive effects (Boumpas et al., 1993;Sorrells and Sapolsky, 2007). The anti-inflammatory actions of CORT include increased expression of IGF-1, IB, and MKP1, which supress the expression of inflammatory mediators. In this study, both diet groups and sexes showed increased expression of IB in the hippocampus in response to CORT challenge (Figures 3-4). However, in males, the increase in IB in the hippocampus in response to CORT was greater among HFD offspring when compared to CFD offspring. Overall, HFD males also showed reduced MR transcript in the hippocampus. It is possible that reduced MR transcript may have potentiated further GR-induced anti-inflammation with CORT administration, since reduced availability of MR and receptor saturation with CORT is associated with increased GR expression (Groeneweg et al., 2011;Smoak and Cidlowski, 2004;Sorrells and Sapolsky, 2007). This would suggest that male offspring exposed to maternal HFD may tolerate acute psychological or endocrine stressors more efficiently than CFD offspring in adulthood in order to initiate protective, anti-inflammatory signalling in limbic brain regions.

Enhanced pro-inflammatory response to LPS challenge in female HFD offspring
LPS binds to toll-like receptor 4, leading to the activation of MAP kinases that support inflammatory gene transcription. Over time, pro-inflammatory activation of the HPA axis leads to a return to homeostasis through GR-mediated expression of anti-inflammatory genes MKP-1 and IB that prevent further accumulation of cytotoxic pro-inflammation (Arenzana-Seisdedos et al., 1997;Huang and Miyamoto, 2001;Lee and Hannink, 2001;Rodriguez et al., 1999;Sachdev et al., 2000;Silverman and Sternberg, 2012).
We found that LPS challenge largely led to increases in both pro-and anti-inflammatory expression in both sexes and diet groups, with a potentiated IL6 response in the hippocampus of HFD females (Figures 5-6). In both sexes, diet groups, and brain regions we found increased anti-inflammatory IB and MKP1 in response to LPS. Both sexes and diet groups also showed increased IL6 in the amygdala, and females showed increased NFB transcript in the hippocampus. However, we found that, in the hippocampus, females exposed to maternal HFD had significantly higher levels of IL6 than CFD controls post-LPS. This discrepancy in IL6 response in females suggests that maternal HFD exposure may potentiate inflammatory mechanisms in conditions of immune stress. Notably, similar results have been reported in the hypothalamus, where chronic HFD consumption led to increased inflammation and neuronal apoptosis (Milanski et al., 2009;Zhang et al., 2008).

Sex-specific alterations to CORT+LPS challenge in HFD offspring
Basal levels of CORT dampen the pro-inflammatory effects of LPS through GC signalling and anti-inflammatory gene expression. Acute elevations in CORT through exogenous administration or psychosocial stress increases pro-and anti-inflammatory transcript expression, however, levels of pro-inflammatory transcript are higher in the presence of exogenous CORT compared to basal CORT levels (Munhoz et al., 2010(Munhoz et al., , 2006) ( Figure 11B).
HFD females had lower pro-/anti-inflammatory IL6/IL10 cytokine ratio relative to CFD females, indicating an enhanced anti-inflammatory transcriptional response in the amygdala ( Figure 7). In the prefrontal cortex, only HFD females displayed an increase in MKP-1 ( Figure   9). These findings suggest that maternal HFD exposure is predominantly associated with increased anti-inflammatory transcriptional response in females. In contrast, CORT+LPS led to a significant increase in NFB in the prefrontal cortex of HFD males (Figure 9). These findings suggest that maternal exposure to HFD in males may exacerbate a neuroinflammatory response to combined CORT+LPS challenge during adulthood.

CONCLUSION
Anxiety behaviour in rodents can be directly associated with inflammatory gene expression and inflammation in brain regions regulating HPA axis response to stress (Dantzer et al., 2008;Rodgers et al., 2012). Adult offspring exposed to maternal HFD exhibit increased anxiety during behavioural tests in the open field and elevated plus maze (Bilbo and Tsang, 2010;Sasaki et al., 2014;Sasaki et al., 2013;Sullivan et al., 2012). Future behavioural work is necessary to assess whether anti-inflammatory responses that are robustly activated upon physiological (CORT) and immune triggers (LPS) during adulthood seen in this study, may attenuate anxiety-like behaviour in offspring with perinatal HFD exposure. Our findings suggest the possibility that HFD-induced HPA axis programming during early life may convey neuroprotective and anti-inflammatory outcomes during adulthood when triggered by acute physiological and/or immune challenges.          CORT diffuses into the cytosol and binds to glucocorticoid receptor (GR). The CORT-GR complex induces the expression of anti-inflammatory molecules including IGF1, MKP1, and IBα, which at basal levels inhibit proinflammatory expression of NFκB, IL6 and CD11B induced by LPS signalling to toll-like receptor 4 (TLR4) and mitogen-activated protein (MAP) kinases. However, with higher levels of CORT, anti-inflammatory effects are reduced. This system is altered in females exposed to maternal HFD, whereby there is increased anti-inflammatory MKP1 and reduced IL6/IL10 ratio. In males exposed to maternal HFD, there is increased proinflammatory NFκB.