Suboptimal nutrition in early life affects the inflammatory gene expression profile and behavioral responses to stressors

Accepted Manuscript Suboptimal nutrition in early life affects the inflammatory gene expression profile and behavioral responses to stressors Nicola M. Grissom, Robert George, Teresa M. Reyes PII: DOI: Reference:

S0889-1591(16)30476-7 http://dx.doi.org/10.1016/j.bbi.2016.10.013 YBRBI 2992

To appear in:

Brain, Behavior, and Immunity

Received Date: Revised Date: Accepted Date:

2 September 2016 6 October 2016 14 October 2016

Please cite this article as: Grissom, N.M., George, R., Reyes, T.M., Suboptimal nutrition in early life affects the inflammatory gene expression profile and behavioral responses to stressors, Brain, Behavior, and Immunity (2016), doi: http://dx.doi.org/10.1016/j.bbi.2016.10.013

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Suboptimal nutrition in early life affects the inflammatory gene expression profile and behavioral responses to stressors

Nicola M. Grissom1, Robert George2 and Teresa M. Reyes3

1

University of Minnesota, Department of Psychology, Minneapolis, MN, USA

2

University of Pennsylvania, Department of Pharmacology, Philadelphia, PA, USA

3

University of Cincinnati, Department of Psychiatry and Behavioral Neuroscience, Cincinnati, OH, USA

*Corresponding Author: Teresa M. Reyes University of Cincinnati, College of Medicine Dept of Psychiatry and Behavioral Neuroscience 2120 East Galbraith Road, A-129 Cincinnati, OH 45237-1625 email. [email protected] tel. 513-558-4338 Keywords: stressor; perinatal diet; high fat; low protein; cytokine; chemokine; resilient; sickness behavior

Highlights • • • • •

Poor quality perinatal diet affects adult response to stressors Suboptimal diet leads to a blunted CNS proinflammatory response to LPS Maternal high fat diet exacerbates the proinflammatory response in PFC Low protein offspring show reduced sickness behavior High fat offspring have a potentiated locomotor reduction to LPS

Abstract Nutritional conditions in early life can have a lasting impact on health and disease risk, though the underlying mechanisms are incompletely understood. In the healthy individual, physiological and behavioral responses to stress are coordinated in such a way as to mobilize resources necessary to respond to the stressor and to terminate the stress response at the appropriate time. Induction of proinflammatory gene expression within the brain is one such example that is initiated in response to both physiological and psychological stressors, and is the focus of the current study. We tested the hypothesis that early life nutrition would impact the proinflammatory transcriptional response to a stressor. Pregnant and lactating dams were fed one of three diets; a low-protein diet, a high fat diet, or the control diet through pregnancy and lactation. Adult male offspring were then challenged with either a physiological stressor (acute lipopolysaccharide injection, IP) or a psychological stressor (15 min restraint). Expression of 20 proinflammatory and stress-related genes was evaluated in hypothalamus, prefrontal cortex, amygdala and ventral tegmental area. In a second cohort, behavioral responses (food intake, locomotor activity, metabolic rate) were evaluated. Offspring from low protein fed dams showed a generally reduced transcriptional response, particularly to LPS, and resistance to behavioral changes associated with restraint, while HF offspring showed an exacerbated transcriptional response within the PFC, a reduced transcriptional response in hypothalamus and amygdala, and an exacerbation of the LPS-induced reduction of locomotor activity. The present data identify differential proinflammatory transcriptional responses

throughout the brain driven by perinatal diet as an important variable that may affect risk or resilience to stressors.

Introduction Early life nutrition is a key determinant in proper growth and development. Both deficient [1], as well as excess nutrition [2] (e.g., total calories or specific macro- (fat, protein) or micro- nutrients (iron, vitamins), have been linked to altered growth parameters. Clinically, this can lead to babies being born small-for-gestational age (SGA) or large for gestational age (LGA), respectively. Beyond nutrition, a wide range of pregnancy conditions increase the risk for SGA (maternal infection, hypertension or placental dysfunction [3]) and LGA (gestational diabetes or maternal obesity [4]). Both SGA and LGA increase the risk for perinatal complications and chronic health conditions (hypertension and obesity [5, 6]) later in life and have been associated with an increased risk for adverse neurobehavioral disorders, including schizophrenia, and difficulty with emotional regulation [7]. Therefore, it is important to define how early life nutrition affects the developing brain. The role of proinflammatory responses in the brain are increasingly appreciated as contributors to a number of mental health disorders, including depression, anxiety [8] and schizophrenia [9]. Both physiological stressors such as an immune challenge [1013] or psychological challenges [14, 15] can induce a proinflammatory response throughout the brain, with cortex, hypothalamus, amygdala and hippocampus being the regions most often examined. Additionally, proinflammatory responses in the brain have been shown to contribute to the behavioral changes that accompany both physiological stressors [16, 17], and psychological stressors (e.g., repeated social defeat [18] or chronic restraint [19]).

The goal of the present experiment was to determine whether altered early life nutrition could affect the proinflammatory gene expression profile in the offspring brain. We model SGA and LGA offspring by feeding dams a diet deficient in protein or high in fat, respectively, through breeding and lactation [7, 20-22]. Using this model, we have shown alterations in dopamine and opioid expression in the brain, as well as executive function deficits. The present studies examined a panel of 20 proinflammatory-related genes in response to peripheral LPS administration, a physiological stressor, or 15 min restraint, a psychological stressor [23]. Four brain regions were examined; hypothalamus, a central brain structure in the response to stressors, as well as prefrontal cortex (PFC), amygdala (AMYG), and ventral tegmental area (VTA), structures involved in emotional processing as well as stress responses. The present data support the conclusion that early life nutrition affects proinflammatory gene expression profiles throughout the brain, and these responses differ by brain region, by stressor, and by dietary challenge.

Methods. C57BL/6J females and DBA/2J males were ordered from Charles River (Wilmington, MA) and bred. B6D2F1/J mice were used in all studies as a hybrid background is more similar to the heterogeneity observed in humans as opposed to a pure inbred strain, and hybrid vigor can lead to robustness (e.g., increased litter size, and resiliency [24, 25]). Breeding pairs were randomly assigned to one of three diet conditions, and diets were fed from the onset of breeding, through pregnancy and lactation. Experimental diets were purchased from Test Diet (Richmond, IN, USA);

control: #5755, 4.09 kcal/g with 18% of total energy calories from protein, 22% from fat, and 60% from carbohydrate; low protein: #5769, 4.13 kcal/g with 8.5% of total energy calories from protein, 22% from fat, and 69.5% from carbohydrate; high fat: #58G9, 5.21 kcal/g with 18% of total energy calories from protein, 60% from fat, and 22% from carbohydrate (see Table 1). The source of fat across the diets was the same (although the quantities differed), with an increase in mostly lard, and to a smaller extent corn oil in the high fat diet. The maternal diets do not differentially affect litter size or composition, and, pups born to dams fed the low protein diet have significantly lower birth weights [21], while those born to high fat fed dams are significantly heavier at birth [22]. Male offspring were weaned at 21 days of age and fed standard chow (Lab Diet 5001), and housed in groups of 5 in housing rooms with an ambient temperature of 22-23˚ C. Only one animal per litter was used in any specific experiment. Gene expression experiments were conducted when animals were 8 weeks of age, behavioral experiments were conducted when animals were 4.5 months of age.

Stress paradigms. For the gene expression experiments, animals (n=5-6) were sacrificed either under basal conditions or 2 hours after the onset of either a physiological stressor (lipopolysaccharide (LPS) administration, IP), or a psychological stressor (15 minute restraint) [26]. Samples were collected between 1200-1400, with lights on at 7am. Because animals were group housed, all animals in a given cage were subjected to the same stress condition. Animals from which basal samples were collected were left undisturbed in their home cage in the vivarium housing room

adjacent to the procedure room until the time of sample collection. Animals in the LPS condition received 10ug LPS (serotype 055:B5, Sigma) IP, and were returned to their homecage in the housing room. Animals in the restraint condition were placed in a clean Decapicone (Braintree Scientific, Braintree, MA) and secured by taping the rear opening closed. Restraint lasted 15 minutes, at which point animals were returned to their cage and the cage was returned to the housing room.

Nucleic acid extraction and gene expression. RNA extraction occurred as previously described [27]. Briefly, at the time of sacrifice, brains were rapidly extracted, placed in RNAlater and stored at -20C. The hypothalamus, prefrontal cortex, amygdala, and ventral tegmental area were dissected as previously described [21, 27, 28]. RNA was extracted using Qiagen AllPrep DNA/RNA Mini kit. 100ug/ul cDNA was synthesized using Applied Biosystems High Capacity Reverse Transcriptase kit.

Gene expression and cluster analysis. Gene expression was assayed (n=5-6) using Taqman primers on custom high-throughput OpenArray real-time PCR plates (Life Technologies) assaying housekeeping genes (GAPDH, ACTB) and 20 genes of interest, including chemokines/receptors (CCL2, CCR2, CXCL10), proinflammatory cytokines/receptors (IL-1, IL-18, IL-1R, IL-6, TNF-α), proinflammatory signaling molecules (COX2, NOS, PGES, SOCS3, TLR4, NFκB, IkBα), and neuron/glia signaling (MR, GR, GFAP, GAD1, ß-adrenergic receptor (ADRB1; see Table 2). In the PFC, one sample was lost in each of the following conditions due to assay failure; SC-

basal, HF-basal, SC-restraint, LP-restraint. Expression of housekeeping genes did not differ between groups. Fold change values are expressed compared to basal control animals. Cluster analysis was performed as previously published [29]. Briefly, normalized gene expression was converted to z-scores, and unsupervised hierarchical clustering was performed (uncentered correlation) of genes and treatment conditions using Cluster 3.0 and JavaTreeview ([30] as described in [28]). Complete linkage creates clusters of approximately equal size, but does not require assumptions as to the number of clusters, allowing us to define broader patterns of gene expression within each brain region. The relative expression level of genes was transformed into a yellow-blue heat map, where yellow indicated higher than the average level of expression for that gene across all samples in the cluster (positive z-scores), and blue indicated lower than average expression of the same gene (negative z-scores). The hypothalamic expression of 5 genes (neuron/glia signaling: ADRB1, NR3C1, NR3C2, GFAP, and GAD1) has been published previously [27] as part of a manuscript that evaluated the HPA response to these stressors. These data are included in the present manuscript to evaluate whether and how this subset clusters across brain regions and in relation to the proinflammatory genes. Because of the large number of genes in the dataset, individual gene targets were depicted graphically on the basis of a significant diet effects, diet x stressor interactions and significant posthoc assessment.

Behavior. Food intake, x-axis locomotor activity, and oxygen consumption (metabolic rate) were measured in metabolic cages (Comprehensive Lab Animal Monitoring System,

Columbus Instruments, Columbus, OH). At least one week prior to testing, animals (n=6) were singly-housed in the CLAMS overnight to acclimate to powdered food and a novel cage. While in the CLAMS cages, animals had ad libitum access to powdered diet and water. For the experiment, animals were housed in the metabolic cages for 48 hrs. The first 24 hr represented the animals baseline state, and the second 24 hrs (which was initiated at the same circadian time, directly following the first 24hrs) were the animals’ response to either the LPS or restraint stressor (two separate groups of animals). With this design, each animal’s response to the stressor is compared to their own baseline. For food intake and locomotor activity, data for the two hours following the stressor is presented, for oxygen consumption the 24 hr average is presented.

Statistics. Two way Analysis of Variance (ANOVA) was used to analyze the data, with maternal diet (control, LP, HF) and stress (basal, LPS, restraint) as factors. Baseline behavioral measures were analyzed with one-way ANOVA. Planned post-hoc comparisons comparing either LP or HF offspring to controls were conducted using Fisher's least significant difference.

Results. As the target genes were chosen for their stress-responsiveness, all genes were affected by stressor exposure, as predicted. Overall, LPS led to larger fold changes as compared to restraint. The specific question to be addressed in the present experiments was whether maternal diet differentially affected the response to

stressors, therefore, the current data presentation and discussion will focus on those genes that were affected by maternal diet. The present data reveal that maternal diet affected inflammatory gene expression in the brain, and this effect differed by brain region. In the hypothalamus (Fig 1), the primary differences were detected in response to adult challenge with LPS, such that offspring from dams fed either the LP or HF diet were hyporesponsive to LPS. For low protein offspring, this effect was seen for CXCL10, IL-18, IL-1R and SOCS3, while for HF offspring the effect was evident for CXCL10, TNF-α, and SOCS3 (see Table 3 for statistics). Additionally, in response to restraint stress, LP offspring showed reduced transcriptional responses of IL-18 and IL-1R. In the amygdala, a similar pattern was observed, such that maternal diet decreased the transcriptional response to LPS in LP or HF offspring. Both LP and HF offspring had reduced levels of CCL2, IL-6, PGES (Fig 2), and GFAP (Fig 6), while additionally LP offspring had lowered responses to IL-1ß, COX-2, and SOCS3. In response to restraint, the only difference was that HF offspring had significantly reduced expression of GFAP. Interestingly, the effects in the PFC were more limited, primarily affecting HF offspring (Fig 3). In contrast to the HYP and AMYG, HF offspring showed potentiated responses to LPS for CXCL10, TNF-α, and CCL2. In response to restraint, both HF and LP offspring showed a decrease in NOS2 expression that was not evident in the control animals.

In the VTA (Fig 4), again LP offspring were primarily affected, demonstrating a reduced expression of COX-2, PGES, SOCS3 and NOS2 in response to LPS, while HF offspring had a reduced expression of only COX2 in response to LPS. At baseline, there were a few notable differences detected. Most notably, expression of CCR2 was affected at baseline in HYP, PFC and AMYG (Fig 5). In HYP, both LP and HF offspring had increased expression of CCR2, while in PFC, only HF offspring had increased expression. In contrast, in the AMYG, both LP and HF offspring had decreased expression levels of CCR2 at baseline. Further, two additional genes in the AMYG showed altered expression at baseline (Fig 2), PGES and IL-18, such that both LP and HF offspring showed reduced expression. Also of note, was GFAP (Fig 6), as LP offspring had reduced basal expression in PFC, AMYG, and VTA, while HF offspring had reduced levels in the AMYG and VTA. In addition to the analysis of individual genes, a cluster analysis was performed on the expression of all genes across all samples (Fig 7). Within each brain regions, the clustering algorithm was applied to both the gene list (to find out which transcripts were expressed most similarly to each other across all samples) and to the individual sample list (to identify which samples/treatments had the most similar expression pattern to each other across all genes). As we predicted, the most notable gene cluster in all brain regions was formed by those transcripts that were increased in expression in response to LPS (seen as bright yellow, sample IDs shaded in gray). Additionally, in the PFC, AMYG, and VTA, a second cluster was defined by the expression of select genes: CCR2, IL-18, GFAP, MR, GR, ADRB1, IL-18, and TLR4 (not all genes in each brain region). Genes

following this latter pattern have their names highlighted gray in each regional cluster. This cluster tended to show higher levels of expression under baseline (particularly in the PFC and AMYG; sample IDs boxed) and reduced expression following LPS administration, reflected in blue coloration in the heatmap. Behavioral responses to LPS and restraint were evaluated in metabolic cages. Baseline data and response to the stressor (LPS or restraint) were collected 24 hrs apart, and data is shown as percent of baseline values. There were no differences in baseline food intake nor locomotor activity during the light period (Fig 8A, 8C). In control animals, both stressors reduced food intake to approximately 35-40% of baseline food intake (Fig 8B). HF offspring did not differ from controls in this reduction of food intake. However, LP offspring were resistant to the stressor-induced anorexia (main effect for diet; F(2,31)=11.02, p<.0002). Locomotor activity was also decreased in control animals in response to both stressors, and maternal diet significantly impacted the animals’ response to the stressors (significant diet x stressor interaction, F(2,32)=8.26, p=.0013). In response to LPS, both LP and HF offspring showed a potentiated response, with a greater decrease in locomotor activity, while in response to restraint, only the LP offspring were different, again showing resistance to restraintinduced decreases in locomotor activity. Oxygen consumption (VO2) is a measure of metabolic rate, and at baseline, HF offspring had a lower metabolic rate (main effect diet: F(2, 36)=14.77, p<.0001). In response to either stressor, there were no difference due to maternal diet, however, LPS lowered metabolic rate in all animals (main effect stress: F(1,32)=38.66, p<.0001).

Discussion. The present data demonstrate that prenatal nutrition can significantly impact the CNS proinflammatory response to stressors. These differences were most pronounced in response to the LPS challenge, however notable baseline differences were observed, as well as select differential responses to restraint. In general, LP offspring were more affected than HF offspring, and in HYP, AMYG and VTA, the affected animals had decreased transcriptional responses. However, the effect in PFC differed, such that only HF animals were affected, and the effect of prenatal diet was to exacerbate the transcriptional response. Interestingly, pronounced behavioral differences were observed as well, but primarily in response to restraint. In response to a stressor, either one that is physiological in nature, in this case, peripheral immune challenge, or one that is psychological in nature, in this case restraint, proinflammatory gene expression in the brain is induced [10-15, 31]. In the present manuscript, we focused on a panel of immune-related genes that are known to be expressed in the CNS and responsive to stressors, and asked whether maternal diet during pregnancy and lactation could alter the proinflammatory transcriptional response to stressors in the offspring. As hypothesized, perinatal diet profoundly affected the proinflammatory gene expression response in the brain. The differential responses were most pronounced when examining the response to LPS. In HYP, AMYG and VTA, both LP and HF offspring showed a reduction in the transcriptional response to LPS, such that the transcriptional change was substantially reduced in the LP or HF offspring. This reduced response was evident in twice as many genes in the LP offspring (14 genes) versus the HF offspring (7 genes). These data indicate that

the LP and HF offspring are compromised with regard to their proinflammatory response to LPS. In contrast, gene expression profiles were also affected in the PFC, however, in this case, the effect of altered perinatal diet was to increase the transcriptional response of the proinflammatory mediators, and importantly, this effect was seen only in the HF offspring, and not in the LP offspring, which did not differ from the control offspring for any gene. This finding highlights a unique susceptibility of the HFD offspring, as neuroinflammation in the PFC has been associated with a range of adverse mental health outcomes, including depression [32, 33], schizophrenia [34], chronic fatigue [35] and PTSD [36]. These data support the conclusion that a HFD during perinatal development may increase susceptibility to these adverse outcomes, specifically through increased inflammation in the prefrontal cortex. Interestingly, two of these disorders, certain subtypes of depression [37] and chronic fatigue syndrome [38, 39], have also been linked to HPA hypofunction. Previously, we have shown in these animals [27], that at the time of tissue collection (2 hrs post-challenge), both LP and HF offspring have significantly reduced levels of corticosterone in the undisturbed and restraint-exposed animals. Collectively these data suggest that HF offspring, through increased inflammation in the PFC coupled with HPA hypofunction, may be at increased risk for depression and/or chronic fatigue syndrome. Further, given the fact that glucocorticoids can have both pro- and anti-inflammatory responses in the brain [40], understanding how differential HPA activation in LP and HF offspring interacts with the differential gene transcripts presented here remains an important area of future research.

The altered transcriptional responses were largely limited to the response to LPS. One potential explanation for this may be that LPS was a much more potent stimulus for this panel of immune-related genes, increasing expression of these genes profoundly (20-600 fold), which may make detection of differential responses more likely. LP and HF offspring showed consistent reductions of ~30-50% across genes and brain regions supporting the conclusion that these offspring were hyporesponsive to LPS. How these differential transcriptional responses relate to downstream, more functional differences, remains to be determined. Overall restraint was less effective in inducing transcription of these proinflammatory genes (e.g., CCL2 in BLA: 2 fold, IL-18 in HYP: 2 fold. and IL-1R in HYP: 3 fold). Importantly, in HYP where IL-18 and IL-1R were induced by restraint, LP offspring showed a reduced transcriptional response, consistent with the LPS pattern. It is also important to highlight that using a different panel of genes, one focused on epigenetic and neurotransmitter related molecules, we observed a number of significant differences in the hypothalamic response to restraint in the LP and HF offspring [27], indicating that the lack of differential response to restraint observed in the present experiments was likely a function of the specific panel of genes, and not a general failure to respond to restraint. The changes in CCR2 and GFAP observed at baseline were notable, as only these two genes showed such extensive changes at baseline, and further, with no differential response to the stressors. The GFAP findings suggest that these perinatal dietary manipulations may decrease the overall number of astrocytes, but interestingly not in the hypothalamus, which may be due to the relatively earlier development of the hypothalamus as compared to cortex and/or amygdala [41]. For CCR2, unlike the

response pattern to LPS, where HYP and AMYG tended to respond similarly, this was not the case for CCR2; expression levels were upregulated in LP and HF offspring in the HYP and downregulated in the AMYG. These findings are consistent with a recent report that demonstrated an upregulation of CCR2 in hypothalamic cultures from rats exposed to HFD prenatally from E7 to E19 [42]. Further, expression levels were significantly increased in the PFC, and again, this occurred only in the HF offspring. CCR2 is a chemokine receptor, and the endogenous ligands include CCL2, CCL7, CCL8, CCL12 and CCL13. Chemokines in the brain play a role in the recruitment of immune cells into the CNS. CCR2 deficiency can limit macrophage traffic into the brain and subsequently limit spatial and memory impairments following traumatic brain injury [43]. CCR2 has also been shown to have a protective role in hypoxia-ischemia driven learning impairments [44]. However, these situations involve limiting a deleterious response to injury or challenge, and the implications of altered baseline levels of CCR2 remain unclear. However, in the report mentioned previously which showed an upregulation of CCR2 in hypothalamic cultures exposed to HFD prenatally [42], the authors also reported a differential response to the ligand CCL2 in culture, with decreased migration and orexigenic gene expression induction, suggesting that dysregulation of the CCR2/CCL2 system during development can profoundly affect neuronal development and function. Given the notable differences observed across brain regions in the present study (e.g., up in HYP and PFC, down in AMYG) future studies should examine the developmental impact in each of these regions. Behavioral data revealed that in addition to transcriptional changes, perinatal diet altered behavioral responses to stressor exposure as well. The behavioral data were

collected at the same time of day (during lights on) in which the transcriptional experiment had been completed. At baseline, food intake and locomotor activity did not differ, however oxygen consumption, a measure of metabolic rate, was significantly lower at baseline in HF offspring, consistent with our previously published data [45]. A lower metabolic rate may predispose these animals to weight gain, as LGA is known to increase the risk of later life obesity [46]. Stressor exposure, both physiological and psychological, inhibit food intake [17, 47] and reduce locomotor activity [48, 49], and in the present study these behavioral responses to stress were observed in control animals (~70-80% reduction in food intake and ~40% reduction in locomotor activity). In response to restraint, the LP offspring were strongly resistant to decreases in both food intake and locomotor activity, while HF offspring did not differ from controls. In response to LPS, maternal diet did not affect the food intake response, however, the locomotor response in both LP and HF offspring was significantly potentiated, demonstrating a decrease of ~80% in response to LPS, nearly twice that of the control animals. These patterns of responses suggest that in response to restraint, food intake and locomotor activity may be regulated similarly, as the effect of maternal diet on these two behaviors was similar, while for LPS, maternal diet significantly exacerbated the locomotor suppression, indicating that the circuits that underlie LPS-induced locomotor depression appear vulnerable to dysregulation by maternal diet. The amygdala may be involved in locomotor depression in response to LPS, and indeed a number of genes were affected by maternal diet in the amygdala. LPS induced decrease in social interaction and c-fos in amygdala were both blocked by IL-

1ra or vagotomy [50, 51]. It is interesting that a number of proinflammatory changes in the amygdala were similar between LP and HF offspring, specifically a reduced responsiveness of CCL2, IL-6, and PGES to LPS. One interpretation linking LPS induced amygdalar gene expression and locomotor suppression would suggest that these proinflammatory mediators are important for recovery to normal activity levels, and with reduced expression, a more pronounced hypoactivity is observed. A more precise timecourse would be needed to fully test this hypothesis. Additionally, IL-18 and PGES were also reduced at baseline in the amygdala to a similar extent in both LP and HF offspring, and it is possible that this baseline difference contributed to the differential response to LPS. The hypothalamus has also been suggested to play a role in mediating decreased locomotor activity in response to immune challenge [52]. In the present study, both LP and HF offspring showed a reduced hypothalamic CXCL10, and SOCS3 response to LPS. As with the expression changes observed in the BLA, it is possible that CXCL10 and SOCS3 play a role in restraining the behavioral response to LPS administration, as a reduced expression of these molecules resulted in an exacerbated behavioral phenotype. Consistent with the large fold changes observed in response to LPS, clustering of individual samples revealed that the most notable cluster in all brain regions was the cohort of animals which received treatment with LPS, with little to no distinction between maternal diet conditions, reflecting greater similarity of expression of these transcripts within the LPS condition rather than within maternal diet treatments. Additionally, however, the cluster analysis revealed an interesting set of genes that were similarly regulated in PFC and AMYG, and to some extent VTA (genes

highlighted in grey in Fig 7), but not in hypothalamus. This cluster identified genes more highly expressed at baseline, and that showed a reduced relative expression in response to LPS. Three of the genes, ADRB1 (beta adrenergic receptor) [53], MR [54] and GR [55], are receptors that play a central role in the sympathetic and hormonal responses to stress, while both IL-18 [56] and GAD1 [57] are also known to be stress responsive. The current data identify these genes as being coordinately regulated in these regions, however, a better understanding of how these molecules participate in the stress response within these regions in needed. The focus of the present studies was on perinatal dietary manipulations, including both gestation and lactation. Maternal diet is known to change the nutrient content of breastmilk [58], so inclusion of the lactational time period was important. Maternal protein restriction is a well-validated rodent model for in utero growth restriction, and our maternal LP diet results in pups that are born small for gestational age (SGA) [20, 21]. In addition to poor nutrition, numerous other factors can lead to growth restriction in utero, including total calorie restriction, maternal infection or hypertension, or placental insufficiency [3]. We have characterized maternal high fat diet as a model of large for gestational age (LGA) offspring [22]. Similar to SGA, there are numerous causes of LGA, including maternal obesity, gestational diabetes, and excessive gestational weight gain [4]. The extent to which our findings apply to other (nondietary) models of SGA and LGA more broadly will be the focus of future studies. The current studies did not include female offspring, so the present data apply only to males. Given the well-documented sex differences in inflammation and differential perinatal programming, it will be important for future studies to include females.

In summary, the present data identify perinatal diet as a critical environmental variable that can alter the behavioral and transcriptional response to both physiological and psychological stressors. Offspring from low protein fed dams showed a generally reduced transcriptional response, particularly to LPS, and resistance to behavioral changes associated with restraint, while HF offspring showed an exacerbated transcriptional response within the PFC, and an exacerbation of the LPS-induced reduction of locomotor activity. Early life nutrition can program later life health or disease and the physiological response to stressors is a key component of a number of disease states, including metabolic and psychiatric conditions. The present data identify differential proinflammatory transcriptional responses throughout the brain as an important variable that may translate into risk or resilience to stressors.

Funding: This work was supported by the National Institutes of Health grant MH087978 (TMR) and a NARSAD Young Investigator award to NMG. The authors have no financial interests or conflicts of interest to disclose.

Figure legends. Figure 1. Gene expression in the hypothalamus. Gene expression levels at baseline (basal), or 2 hrs after LPS or restraint in control (white bars), low protein offspring (gray bars) and HF offspring (black bars). *p<.05, posthoc comparison indicating different from SC control within basal, LPS or restraint.

Figure 2. Gene expression in amygdala . Gene expression levels at baseline (basal), or 2 hrs after LPS or restraint in control (white bars), low protein offspring (gray bars) and HF offspring (black bars). *p<.05, posthoc comparison indicating different from SC control within basal, LPS or restraint.

Figure 3. Gene expression in the PFC. Gene expression levels at baseline (basal), or 2 hrs after LPS or restraint in control (white bars), low protein offspring (gray bars) and HF offspring (black bars). *p<.05, posthoc comparison indicating different from SC control within basal, LPS or restraint.

Figure 4. Gene expression in the VTA. Gene expression levels at baseline (basal), or 2 hrs after LPS or restraint in control (white bars), low protein offspring (gray bars) and HF offspring (black bars). *p<.05, posthoc comparison indicating different from SC control within basal, LPS or restraint.

Figure 5. Expression levels of CCR2 across regions. Gene expression levels at baseline (basal), or 2 hrs after LPS or restraint in control (white bars), low protein

offspring (gray bars) and HF offspring (black bars). *p<.05, posthoc comparison indicating different from SC control.

Figure 6. Expression levels of GFAP across regions. Gene expression levels at baseline (basal), or 2 hrs after LPS or restraint in control (white bars), low protein offspring (gray bars) and HF offspring (black bars). *p<.05, posthoc comparison indicating different from SC control.

Figure 7. Cluster analysis reveals groups of genes that are coordinately regulated in HYP, PFC, AMYG, and VTA. The largest cluster is comprised of genes that are upregulated in response to LPS, seen as bright yellow (LPS sample IDs are marked in gray). A second cluster of genes (marked in gray) are downregulated in response to LPS (seen as blue in the cluster), and relatively upregulated in primarily basal samples in PFC and AMYG (sample IDs are boxed).

Figure 8. Food intake, locomotor activity and metabolic rate (oxygen consumption; VO2). (A) There were no differences in the baseline food intake. (B) Food intake during the 2hr post LPS or restraint, expressed as a percent change from baseline values 24 hr prior to testing. LP offspring had an attenuated response to restraint. *p<.05, different from SC restraint (C) There were no differences in baseline locomotor activity. (D) Locomotor activity during the 2hr post LPS or restraint, expressed as a percent change from baseline values 24 hr prior to testing. LP and HF offspring had a potentiated response to LPS, while LP offspring were resistant to the effect of

restraint. *p<.05, different from SC LPS or SC restraint (E) HF offspring had significantly reduced oxygen consumption at baseline. *p<.05, 1 way ANOVA (F) LPS significantly reduced oxygen consumption 24 hr post stressor exposure. *p<.05, main effect for stress. Control (white bars), low protein (gray bars) and HF (black bars) offspring.

Table 1.

Test Diet Number % energy from Fat % energy from Carbohydrate % energy from Protein Energy (kcal/g)2 Sucrose (% of ingredients)

Control Diet 5755 22.1 59.6 18.3 4.07 15

Low Protein Diet 5769 21.8 70.1 8.1 4.12 27

High Fat Diet 58G9 59.9 21.4 18.6 5.21 6.4

Table 2.

Gene

Assay ID

ACTB ADRB1 CCL2 CCR2 COX2 CXCL10 GAD1 GAPDH GFAP GR IkBalpha IL-18 IL-1ß IL-1R IL-6 MR NFkB (p65) NOS PGES SOCS3 TLR4 TNFα

Mm00607939_s1 Mm00431701_s1 Mm00441242_m1 Mm00438270_m1 Mm00478374_m1 Mm00445235_m1 Mm04207432_g1 Mm99999915_g1 Mm01253033_m1 Mm00433832_m1 Mm00477798_m1 Mm00434225_m1 Mm01336189_m1 Mm00434237_m1 Mm00446190_m1 Mm01241596_m1 Mm00501346_m1 Mm00440502_m1 Mm00452105_m1 Mm00545913_s1 Mm00445273_m1 Mm00443258_m1

Table 3.

HYPOTHALAMUS CXCL10 TNF-α IL-18 IL-1R1 SOCS3 BLA CCL2 IL-1ß IL-6 COX-2 PGES SOCS3 IL-18

Interaction

F(4,37)=2.79; p=.04 F(4,37)=2.67; p=.0469 F(4,37)=4.13; p=.0073

Main Effect: Stress F(2,37)=132.61; p<.0001 F(2,37)=50.63; p<.0001 F(2,37)=13.03; p<.0001 F(2,37)=47.94; p<.0001 F(2,37)=68.13; p<.0001

F(4,36)=4.36; p=.0058

F(2,35)=89.28; p<.0001 F(2,33)=67.07; p<.0001 F(2,34)=81.7; p<.0001 F(2,36)=189.5; p<.0001 F(2,36)=110.46; p<.0001 F(2,36)=179.41; p<.0001 F(2,36)=9.83; p=.0004

PFC CXCL10 TNF-α CCL2 NOS2

F(2,31)=93.33; p<.0001 F(2,24)=34.74; p<.0001 F(2,32)=100.43; p<.0001 F(2,33)=6.53; p=.0041

VTA COX2 PGES SOCS3 NOS2

F(2,38)=107.40; p<.0001 F(2,38)=193.3; p<.0001 F(2,38)=244.2; p<.0001 F(2,38)=133.9; p<.0001

CCR2 HYP PFC AMYG VTA GFAP PFC AMYG VTA

F(4,34)=2.66; p=.0493 F(4,26)=2.88; p=.0422 F(4,33)=6.58; p=.0005

F(2,34)=7.48; p<.002 F(2,26)=4.04; p=.0297 F(2,33)=95.11; p=.0005 F(2,38)=34.66; p<.0001

F(2,32)=4.43, p=.0201 F(2,36)=8.26, p=.0011

Main Effect: Diet

F(2,37)=4.12; p=.0243 F(2,37)=4.7; p=.0152

F(2,36)=5.99; p=.0057 F(2,36)=5.78; p=.0067

F(2,33)=8.03; p=.0014

F(2,32)=3.62, p=.0384 F(2,36)=15.02, p<.0001 F(2,38)=4.34, p=.0201

References

1. 2. 3. 4.

5. 6. 7.

8.

9. 10.

11. 12.

13.

14.

15. 16. 17.

Triunfo, S. and A. Lanzone, Impact of maternal under nutrition on obstetric outcomes. J Endocrinol Invest, 2015. 38(1): p. 31-8. Li, N., et al., Maternal prepregnancy body mass index and gestational weight gain on pregnancy outcomes. PLoS One, 2013. 8(12): p. e82310. Kramer, M.S., The epidemiology of low birthweight. Nestle Nutr Inst Workshop Ser, 2013. 74: p. 1-10. Yu, Z., et al., Pre-pregnancy body mass index in relation to infant birth weight and offspring overweight/obesity: a systematic review and meta-analysis. PLoS One, 2013. 8(4): p. e61627. Rosebloom, T.J., et al., Effects of prenatal exposure to the Dutch famine on adult disease in later life: an overview. Mol Cell Endocrinol, 2001. 185(1-2): p. 93-98. Ravelli, G.P., Z.A. Stein, and M.W. Susser, Obesity in young men after famine exposure in utero and early infancy. N Engl J Med, 1976. 295(7): p. 349-53. Grissom, N.M. and T.M. Reyes, Gestational overgrowth and undergrowth affect neurodevelopment: similarities and differences from behavior to epigenetics. Int J Dev Neurosci, 2013. 31(6): p. 406-14. Ramirez, K. and J.F. Sheridan, Antidepressant imipramine diminishes stress-induced inflammation in the periphery and central nervous system and related anxiety- and depressive- like behaviors. Brain Behav Immun, 2016. 57: p. 293-303. Esslinger, M., et al., Schizophrenia associated sensory gating deficits develop after adolescent microglia activation. Brain Behav Immun, 2016. Ziko, I., et al., Neonatal overfeeding alters hypothalamic microglial profiles and central responses to immune challenge long-term. Brain Behav Immun, 2014. 41: p. 32-43. Strenn, N., et al., Expression of inflammatory markers in a genetic rodent model of depression. Behav Brain Res, 2015. 281: p. 348-57. Kentner, A.C., et al., The effects of rewarding ventral tegmental area stimulation and environmental enrichment on lipopolysaccharide-induced sickness behavior and cytokine expression in female rats. Brain Res, 2008. 1217: p. 50-61. Doremus-Fitzwater, T.L., et al., Male adolescent rats display blunted cytokine responses in the CNS after acute ethanol or lipopolysaccharide exposure. Physiol Behav, 2015. 148: p. 131-44. Vecchiarelli, H.A., et al., Divergent responses of inflammatory mediators within the amygdala and medial prefrontal cortex to acute psychological stress. Brain Behav Immun, 2016. 51: p. 70-91. Knapp, D.J., et al., Stress and Withdrawal from Chronic Ethanol Induce Selective Changes in Neuroimmune mRNAs in Differing Brain Sites. Brain Sci, 2016. 6(3). Konsman, J.P., P. Parnet, and R. Dantzer, Cytokine-induced sickness behaviour: mechanisms and implications. Trends Neurosci, 2002. 25(3): p. 154-9. Reyes, T.M. and P.E. Sawchenko, Involvement of the Arcuate Nucleus of the Hypothalamus in Interleukin-1-Induced Anorexia. J. Neurosci. , 2002. 22(12): p. 5091-5099.

18.

19. 20.

21. 22. 23.

24. 25. 26.

27.

28.

29.

30. 31.

32.

33. 34.

Wohleb, E.S., et al., Knockdown of interleukin-1 receptor type-1 on endothelial cells attenuated stress-induced neuroinflammation and prevented anxiety-like behavior. J Neurosci, 2014. 34(7): p. 2583-91. Kim, T.K., et al., Local Interleukin-18 System in the Basolateral Amygdala Regulates Susceptibility to Chronic Stress. Mol Neurobiol, 2016. Whitaker, K.W., K. Totoki, and T.M. Reyes, Metabolic adaptations to early life protein restriction differ by offspring sex and post-weaning diet in the mouse. Nutr Metab Cardiovasc Dis, 2012. 22(12): p. 1067-74. Vucetic, Z., et al., Early life protein restriction alters dopamine circuitry. Neuroscience, 2010. 168(2): p. 359-70. Vucetic, Z., et al., Maternal high-fat diet alters methylation and gene expression of dopamine and opioid-related genes. Endocrinology, 2010. 151(10): p. 4756-64. Li, H.Y., A. Ericsson, and P.E. Sawchenko, Distinct mechanisms underlie activation of hypothalamic neurosecretory neurons and their medullary catecholaminergic afferents in categorically different stress paradigms. Proc Natl Acad Sci U S A, 1996. 93(6): p. 2359-64. Davis, J.A. and W.R. Lamberson, Effect of heterosis on performance of mice across three environments. J Anim Sci, 1991. 69(2): p. 543-50. Anisman, H., Task complexity as a factor in eliciting heterosis in mice: aversively motivated behaviors. J Comp Physiol Psychol, 1975. 89(8): p. 976-84. Reyes, T., et al., Categorically distinct acute stressors elicit dissimilar transcriptional profiles in the paraventricular nucleus of the hypothalamus. J Neurosci, 2003. 23(13): p. 5607-16. Grissom, N.M., R. George, and T.M. Reyes, The hypothalamic transcriptional response to stress is severely impaired in offspring exposed to adverse nutrition during gestation. Neuroscience, 2015. doi: 10.1016/j.neuroscience.2015.07.022. Grissom, N.M., et al., Dissociable deficits of executive function caused by gestational adversity are linked to specific transcriptional changes in the prefrontal cortex. Neuropsychopharmacology, 2015. 40(6): p. 1353-63. Carlin, J.L., et al., Voluntary exercise blocks Western diet-induced gene expression of the chemokines CXCL10 and CCL2 in the prefrontal cortex. Brain Behav Immun, 2016. DOI: 10.1016/j.bbi.2016.07.161. Eisen, M.B., et al., Cluster analysis and display of genome-wide expression patterns. Proc Natl Acad Sci U S A, 1998. 95(25): p. 14863-8. Reyes, T.M., et al., Categorically Distinct Acute Stressors Elicit Dissimilar Transcriptional Profiles in the Paraventricular Nucleus of the Hypothalamus. J Neurosci, 2003. 23(13): p. 5607-5616. Shelton, R.C., et al., Altered expression of genes involved in inflammation and apoptosis in frontal cortex in major depression. Mol Psychiatry, 2011. 16(7): p. 75162. Wei, Y.B., et al., Elevation of Il6 is associated with disturbed let-7 biogenesis in a genetic model of depression. Transl Psychiatry, 2016. 6: p. e869. Pillai, A., et al., Association of serum VEGF levels with prefrontal cortex volume in schizophrenia. Mol Psychiatry, 2016. 21(5): p. 686-92.

35.

36.

37. 38.

39. 40. 41. 42. 43.

44.

45.

46. 47. 48. 49.

50.

51.

Ifuku, M., et al., Induction of interleukin-1beta by activated microglia is a prerequisite for immunologically induced fatigue. Eur J Neurosci, 2014. 40(8): p. 3253-63. Wilson, C.B., et al., Inflammation and oxidative stress are elevated in the brain, blood, and adrenal glands during the progression of post-traumatic stress disorder in a predator exposure animal model. PLoS One, 2013. 8(10): p. e76146. Struber, N., D. Struber, and G. Roth, Impact of early adversity on glucocorticoid regulation and later mental disorders. Neurosci Biobehav Rev, 2014. 38: p. 17-37. Vangeel, E., et al., Chronic Fatigue Syndrome and DNA Hypomethylation of the Glucocorticoid Receptor Gene Promoter 1F Region: Associations With HPA Axis Hypofunction and Childhood Trauma. Psychosom Med, 2015. 77(8): p. 853-62. Nijhof, S.L., et al., The role of hypocortisolism in chronic fatigue syndrome. Psychoneuroendocrinology, 2014. 42: p. 199-206. Duque Ede, A. and C.D. Munhoz, The Pro-inflammatory Effects of Glucocorticoids in the Brain. Front Endocrinol (Lausanne), 2016. 7: p. 78. Workman, A.D., et al., Modeling transformations of neurodevelopmental sequences across mammalian species. J Neurosci, 2013. 33: p. 7368-83. Poon, K., et al., Prenatal fat-rich diet exposure alters responses of embryonic neurons to the chemokine, CCL2, in the hypothalamus. Neuroscience, 2016. 324: p. 407-19. Hsieh, C.L., et al., CCR2 deficiency impairs macrophage infiltration and improves cognitive function after traumatic brain injury. J Neurotrauma, 2014. 31(20): p. 1677-88. Pimentel-Coelho, P.M., J.P. Michaud, and S. Rivest, C-C chemokine receptor type 2 (CCR2) signaling protects neonatal male mice with hypoxic-ischemic hippocampal damage from developing spatial learning deficits. Behav Brain Res, 2015. 286: p. 146-51. Carlin, J., R. George, and T.M. Reyes, Methyl donor supplementation blocks the adverse effects of maternal high fat diet on offspring physiology. PLoS One, 2013. 85: p. e63549. Johnsson, I.W., et al., A high birth weight is associated with increased risk of type 2 diabetes and obesity. Pediatr Obes, 2015. 10(2): p. 77-83. Smagin, G.N., et al., Prevention of stress-induced weight loss by third ventricle CRF receptor antagonist. Am J Physiol, 1999. 276(5 Pt 2): p. R1461-8. Cabib, S., et al., Different effects of acute and chronic stress on two dopaminemediated behaviors in the mouse. Physiol Behav, 1988. 43(2): p. 223-7. Swiergiel, A.H. and A.J. Dunn, Effects of interleukin-1beta and lipopolysaccharide on behavior of mice in the elevated plus-maze and open field tests. Pharmacol Biochem Behav, 2007. 86(4): p. 651-9. Konsman, J.P., et al., Central nervous action of interleukin-1 mediates activation of limbic structures and behavioural depression in response to peripheral administration of bacterial lipopolysaccharide. Eur J Neurosci, 2008. 28(12): p. 2499-510. Konsman, J.P., et al., The vagus nerve mediates behavioural depression, but not fever, in response to peripheral immune signals; a functional anatomical analysis. Eur J Neurosci, 2000. 12(12): p. 4434-46.

52.

53. 54. 55. 56.

57.

58.

Whitaker, K.W. and T.M. Reyes, Central blockade of melanocortin receptors attenuates the metabolic and locomotor responses to peripheral interleukin-1beta administration. Neuropharmacology, 2008. 54(3): p. 509-20. Berridge, C.W., B.E. Schmeichel, and R.A. Espana, Noradrenergic modulation of wakefulness/arousal. Sleep Med Rev, 2012. 16(2): p. 187-97. de Kloet, E.R., et al., STRESS and DEPRESSION a crucial role of the mineralocorticoid receptor. J Neuroendocrinol, 2016. McKlveen, J.M., et al., Role of prefrontal cortex glucocorticoid receptors in stress and emotion. Biol Psychiatry, 2013. 74(9): p. 672-9. Sahin, C., et al., Agmatine Reverses Sub-Chronic Stress-induced Nod-like Receptor Protein 3 (NLRP3) Activation and Cytokine Response in Rats. Basic Clin Pharmacol Toxicol, 2016. Veeraiah, P., et al., Dysfunctional glutamatergic and gamma-aminobutyric acidergic activities in prefrontal cortex of mice in social defeat model of depression. Biol Psychiatry, 2014. 76(3): p. 231-8. Oosting, A., et al., Rapid and selective manipulation of milk fatty acid composition in mice through the maternal diet during lactation. J Nutr Sci, 2015. 4: p. e19.

Figure 1

Hypothalamus CXCL10

600

SC LP HF

* *

400 200 3 2 1 0

TNF-alpha 80

Fold change

Fold change

800

60

SC LP HF

20 2 1

Basal

LPS

0

Restraint

Basal

6

*

1.5

Fold change

*

1.0

Restraint

4

SC LP HF

* *

2

0.5 0.0

Basal

LPS

0

Restraint

Basal

SOCS3 25 20

Fold change

Fold change

2.0

SC LP HF

LPS

IL 1R1

IL 18 2.5

*

40

15

SC LP HF

* *

10 5 3 2 1 0

Basal

LPS

Restraint

LPS

Restraint

Figure 2

Amygdala IL-1ß

CCL2

Fold change

400 200

SC LP HF

175 150

* * Fold change

600

3 2

125 100

1 0

SC LP HF

*

75 2.0 1.5 1.0 0.5

Basal

LPS

0.0

Restraint

Basal

LPS

IL6

Fold change

30 20

SC LP HF

COX-2 8 7

*

*

6

Fold change

40

10 2.0 1.5 1.0

SC LP HF

*

5 4 3 2 1

0.5 0.0

Basal

LPS

0

Restraint

Basal

LPS

PGES

1.5 1.0 0.5 0.0

10

*

*

8

* * Basal

Restraint

SOCS3

Fold change

Fold change

2.0

SC LP HF

SC LP HF

*

6 4 2

LPS

0

Restraint

IL 18

1.5

Fold change

2.5

Restraint

1.0

* *

Basal

LPS

SC LP HF

0.5

0.0

Basal

LPS

Restraint

Restraint

Figure 3

PFC CXCL10

Fold change

180 160 140

SC LP HF

12

*

10

Fold change

200

TNF alpha

120 100 3 2 1 0

8 6 4 2.0 1.5 1.0 0.5

Basal

LPS

0.0

Restraint

Basal

50 5 4 3 2 1 0

Basal

Fold change

Fold change

100

LPS

Restraint

SC LP HF

1.5

*

SC LP HF

LPS

NOS2

CCL2 150

*

SC LP HF

Restraint

1.0

* *

0.5

0.0

Basal

LPS

Restraint

Figure 4

VTA COX2 SC LP HF

5

* *

4

Fold change

Fold change

100 80 60 40 20 4 3 2

Fold change

20 15

Basal

SC LP HF

LPS

SOCS3

2.5 2.0

*

10 5 3 2 1 0

*

2

0

Restraint

Fold change

25

3

PGES

1

1 0

SC LP HF

Basal

SC LP HF

LPS

Restraint

NOS2

*

1.5 1.0 0.5

Basal

LPS

Restraint

0.0

Basal

LPS

Restraint

Figure 5

CCR2 HYP

6

*

SC LP HF

*

4 2 0

PFC

10

6 4 2

Basal

LPS

0

Restraint

Basal

BLA SC LP HF

1.0

0.0

* * Basal

LPS

Restraint

LPS

Restraint

VTA

1.5

Fold change

Fold change

1.5

0.5

SC LP HF

*

8

Fold change

Fold change

8

SC LP HF

1.0

0.5

0.0

Basal

LPS

Restraint

Figure 6.

GFAP HYP

1.0

0.5

0.0

Basal

LPS

SC LP HF

1.5

Fold change

Fold change

1.5

PFC

SC LP HF

1.0

0.5

0.0

RESTRAINT

*

Basal

LPS

VTA

AMYGDALA

1.0

SC LP HF

* *

* *

0.5

0.0

Basal

LPS

* RESTRAINT

1.5

Fold change

Fold change

1.5

RESTRAINT

SC LP HF

1.0

*

*

0.5

0.0

Basal

LPS

RESTRAINT

Figure 7

LPS LP LPS LP LPS HF LPS HF LPS HF LPS SC LPS HF LPS LP LPS SC LPS SC LPS SC LPS HF LPS LP LPS LP LPS SC Basal HF Basal LP Basal LP Basal SC Basal SC Basal SC Basal SC Basal LP Basal LP Restraint LP Restraint LP Basal HF Basal HF Basal HF Restraint HF Restraint HF Basal LP Basal SC Restraint SC Basal HF Restraint HF Restraint LP Restraint HF Restraint SC Restraint LP Restraint LP Restraint SC Restraint SC Restraint HF Restraint SC

AMYG

PFC CCR2 Gfap IL-18 Adrb1 Nr3c1 IL-1R1 IL-6 Ccl2 IL-1ß TNF-α CXCL10 IkBα Socs3 COX2 PGES Nos2 NFkB (p65) Gad1 Nr3c2 TLR4

Adrb1 IL18 Nr3c1 Gad1 Gfap IL1R1 Nos2 NFkB (p65) IL-1ß CCL2 IL6 TNFα CXCL10 Socs3 IkBα PGES COX2 CCR2 Nr3c2 TLR4

2.00 1.71 1.43 1.14 0.86 0.57 0.29 0.00 -0.29 -0.57 -0.86 -1.14 -1.43 -1.71 -2.00

LPS HF LPS LP LPS SC LPS HF LPS HF LPS HF LPS HF LPS SC LPS LP LPS LP LPS SC LPS LP LPS SC LPS LP LPS SC Basal HF Basal HF Restraint HF Basal HF Basal SC Basal SC Basal SC Restraint LP Basal HF Basal SC Restraint SC Basal LP Basal LP Basal LP Restraint HF Restraint SC Basal LP Basal LP Restraint SC Restraint HF Restraint LP Restraint HF Restraint HF Restraint LP Restraint LP Restraint SC

VTA

HYP

Basal HF Basal LP Basal HF Basal LP Basal SC Basal SC Basal SC Basal HF Basal SC Basal LP Restraint LP Basal SC Restraint SC Basal HF Basal LP Basal LP Restraint HF Restraint LP Restaint HF Restraint HF Restraint LP Basal HF Restraint LP Restraint HF Restraing LP Restraint LP Restraint SC Restraint SC Restraint SC Restraint HF Restraint SC LPS LP LPS HF LPS HF LPS LP LPS HF LPS HF LPS HF LPS LP LPS SC LPS LP LPS LP LPS SC LPS SC LPS SC LPS SC

Ccr2 Gfap Nr3c1 Adrb1 Nr3c2 IL-18 TLR4 Gad1 IL-1R1 CXCL10 Ccl2 IL-6 IL-1ß Tnf-α IkBα PGES Socs3 Nos2 COX2 NFkB (p65)

Ccr2 Gfap IL-18 Adrb1 Nr3c1 Nr3c2 Gad1 IL-1R1 CXCL10 PGES Ccl2 IL6 IL-1ß IkBα Socs3 COX2 TNF-α Nos2 NfKb (p65) TLR4

Basal LP Restraint SC Basal HF Basal HF Basal LP Restraint HF Restraint LP Basal HF Basal LP Basal HF Restraint SC Basal SC Basal SC Basal SC Basal LP Basal SC Basal HF Restraint LP Restraint SC Restraint SC Basal HF Restraint HF Restraint LP Restraint LP Basal LP Basal LP Restraint HF Restraint HF Restraint LP Basal SC Restraint SC Restraint HF LPS LP LPS LP LPS HF LPS LP LPS HF LPS LP LPS SC LPS HF LPS SC LPS SC LPS HF LPS LP LPS SC LPS HF LPS SC

Figure 8 Baseline food intake

A.

Light period food intake Food intake 120 100

0.020

% baseline

kcals food/BW (g) (g) kcals chow/body weight

0.025

0.015 0.010

60 40 20

SC

LP

0

HF

Baseline locomotor activity

3000

D.

% of baseline

x-axis beam breaks

1000

SC

LP

60 40

0

HF

Baseline oxygen consumption

2000

1000

0

% of baseline

VO2 (ml/kg/hr)

*

3000

* LPS

* RESTRAINT

24 hr mean VO2

F.

4000

*

SC LP HF

80

20

0

RESTRAINT

Locomotor activity

100

2000

LPS

Locomotor activity

120

E.

*

SC LP HF

80

0.005 0.000

C.

B.

Baseline food intake

120

SC

100

*

LP

HF

80 60 40 20

SC

LP

HF

0

LPS

RESTRAINT

Highlights • • • • •

Poor quality perinatal diet affects adult response to stressors Suboptimal diet leads to a blunted CNS proinflammatory response to LPS Maternal high fat diet exacerbates the proinflammatory response in PFC Low protein offspring show reduced sickness behavior High fat offspring have a potentiated locomotor reduction to LPS