Maternal suboptimal selenium intake and low-level lead exposure affect offspring’s microglial immune profile and its reactivity to a subsequent inflammatory hit

Micronutrients such as selenium (Se) are essentials since prenatal life to support brain and cognitive development. Se deficiency, which affects up to 1 billion people worldwide, can interact with common adverse environmental challenges including (Pb), exacerbating their toxic effects. Exploiting our recently validated rat model of maternal Se restriction and developmental low Pb exposure, our aims were to investigate: (i) the early consequences of suboptimal Se intake and low-Pb exposure on neuroinflammation in neonates’ whole brains; (ii) the potential priming effect of suboptimal Se and low-Pb exposure on offspring’s glial reactivity to a further inflammatory hit. To these aims female rats were fed with suboptimal (0.04 mg/kg; Subopt) and optimal (0.15 mg/kg; Opt) Se dietary levels throughout pregnancy and lactation and exposed or not to environmentally relevant Pb dose in drinking water (12.5 µg/mL) since 4 weeks pre-mating. We found an overall higher basal expression of inflammatory markers in neonatal brains, as well as in purified microglia and organotypic hippocampal slice cultures, from the Subopt Se offspring. Subopt/Pb cultures were highly activated than Subopt cultures and showed a higher susceptibility to the inflammatory challenge lipopolysaccharide than cultures from the Opt groups. We demonstrate that even a mild Se deficiency and low-Pb exposure during brain development can influence the neuroinflammatory tone of microglia, exacerbate the toxic effects of Pb and prime microglial reactivity to subsequent inflammatory stimuli. These neuroinflammatory changes may be responsible, at least in part, for adverse neurodevelopmental outcomes.

www.nature.com/scientificreports/body of evidence indicated that exposure to this heavy metal may result in microgliosis and inflammation (for a review see 43 ).This knowledge prompted us to focus on microglia as one of the most likely key cell targets in the maternal programming of offspring's brain in conditions of inadequate Se intake and Pb exposure.
The objectives of the current study were twofold: first, to determine the effects of a prolonged Suboptimal low maternal Se intake on the brain neuroinflammatory profile of the progeny at birth and infancy; second, to test the hypothesis that low maternal Se intake would exacerbate the effects of perinatal Pb exposure and prime microglia to enhanced reactivity to a further inflammatory stimulus in the postnatal period.
To these aims, we used our recently validated in vivo model of Se restriction and low-Pb exposure (i) to analyse whole brains from the offspring at birth for their expression of inflammatory markers and (ii) to establish purified cortical microglial cultures and organotypic hippocampal slice cultures (OHSCs) from postnatal day (PND) 1 and PND 6 pups, respectively, assessing their state of activation and reactivity to the typical inflammatory agent lipopolysaccharide (LPS).
Our data support the notion that adequate Se intake during gestation has a protective role on brain immune properties against the toxic effects of Pb exposure and against further challenges later in life, underlining the importance of adequate Se intake since early developmental phases.

Animals and experimental design
This study is reported in accordance with ARRIVE.Wistar rats were kept under standard animal housing (temperature 20 ± 2 °C; humidity 60-70%) with food and water ad libitum, under a 12 h-12 h light/dark cycle (lights on from 7:00 a.m.till 7:00 p.m.).Following the adaption period and 8 weeks prior to breeding, females (weighing 250 ± 25 g) were randomly allocated to either an optimal or suboptimal dietary Se intake group (Se Opt or Se Subopt, respectively; see "Se-containing diets").After the first 4 weeks under this dietary regimen, female rats from both the Se Opt and Se Subopt groups were further randomly divided into two subgroups based on exposure to Pb or Vehicle (Veh).This ensured that half of the females in each group received water containing 12.5 µg/mL of Pb in their drinking bottles, while the other half received water with no Pb (Veh) throughout pregnancy and lactation until the days of sacrifice (PND 1 or 6).Thus, the resulting four experimental groups were: Se Opt/Veh, Se Opt/Pb, Se Subopt/Veh, and Se Subopt/Pb.
Female rats were mated with males (2:1) for 4 to 5 days to cover the duration of an estrous cycle.The animals were regularly weighed, and their food and water consumption recorded every other day.The timeline of experimental design is reported in Fig. 1.

Se-containing diets
Rat diets were prepared by Envigo by supplementing a selenium deficient diet (Teklad Custom Diet TD.92163), containing < 0.01 mg Se/kg with l-selenomethionine to obtain a diet with optimal (Opt, 0.15 mg Se/kg) or suboptimal (Subopt, 0.04 Se mg/kg) Se content, respectively.

Pb exposure
Pb acetate (Pb(Ac) 2 •3H 2 O) was purchased from Sigma Aldrich (Merck KGaA, Darmstadt, Germany).Test solutions for animal treatment were prepared in soft tap water, slightly acidified (pH 4.6) with acetic acid, and containing no Pb acetate (Veh) or Pb acetate at a concentration 12.5 µg/mL of Pb (as element).The selected medium ensured complete dissolution of the Pb salt and acceptance by the animals was the same as plain tap water.The

Whole brain mRNA extraction
The day of birth was designated as PND 0. At PND 1, 6 pups from 6 different litters/group were sacrificed by decapitation under anesthesia induced by hypothermia.and whole brains rapidly dissected on ice; olfactory bulb and medulla were discarded, and the rest of the tissue was snap frozen and kept at − 80 °C until RNA extraction with Trizol reagent.

Purified microglial culture preparation and stimulation
Each mixed primary glial culture was obtained from the cerebral cortex of 1 rat pup/litter from 4 different litters/group (at PND 1), sacrificed as above.following previously published procedures 44 .Entire removal of meninges ensured negligible blood contaminants.Mixed primary glial cultures were maintained in Basal Eagle's Medium (BME), supplemented with 10% heat-inactivated foetal bovine serum, 2 mM glutamine and 100 μg/ mL gentamicin.Medium was replaced after 3 days and cultures maintained in the same medium as above for additional 8 days.After this time, microglial cells were harvested by mild shaking, resuspended in fresh medium without serum, and seeded on uncoated plastic wells at the density of 1 × 10 5 cells/cm 2 .Cell viability was > 95% in all culture conditions, as tested by Trypan Blue exclusion.The cultures consisted of 96% positive cells for the microglia/macrophage marker CD68 (ED-1; 1:200; Serotec).After 24 h in culture, the cells were stimulated or not with 10 ng/ml lipopolysaccharide (LPS, Sigma), in fresh medium.Following 24 h of incubation, the culture media were collected, centrifuged at 1500 rpm to remove cell debris, and stored at − 20 °C until biochemical analyses.Cell monolayers were harvested in Trizol reagent and stored at − 80 °C until RNA extraction.

Organotypic hippocampal slice cultures (OHSC)
Organotypic hippocampal slice cultures (OHSC) were prepared as described in 45 , with few modifications.In brief, 6 rat pups from 6 different litters/group were decapitated (PND 6), and brains rapidly dissected and placed in ice-cold dissection medium (HBSS 1x; HEPES 20 mM; Pen/Strep100 U/ml; d-glucose 6 mg/ml); the hippocampi were isolated and sectioned into 350-μm transverse slices with a McIlwain tissue chopper (Mickle Laboratory Engineering Co. Ltd, Goose Green, UK).The slices were then carefully separated and transferred onto porous collagenated membrane inserts (four slices per insert; PICMORG50, Millicell, Millipore), in 6-well culture plates.1200 μL culture medium, consisting of Neurobasal A medium, 1 mM l-glutamine, 100 U/mL penicillin and 100 µg/mL streptomycin, were added to the lower compartment of each well.On the next day of culture, the culture media were replaced with fresh medium and, from that time, changed twice a week.The slices were incubated at 35 °C, 5% CO 2 for 7 days, before stimulation or not with LPS (5 μg/ml; Sigma), to allow the inflammatory reaction following the mechanical procedure to subside.OHSCs viability was routinely assessed by visualization of propidium iodide (PI; 5 μg/ml; Sigma) staining and fluorescence microscopy; not vital slices were discarded.Following 24 h of LPS stimulation culture media were collected as described above, and slices transferred in Trizol for RNA extraction.Four pups per experimental group were used in these experiments.

RNA extraction, reverse transcription, and real-time PCR analyses
For gene expression experiments, RNA was extracted using Trizol reagent.The quality and concentration of RNA were assessed at NanoDrop spectrophotometer (Thermo Scientific).cDNA was reverse transcribed from 1 µg of RNA using the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Thermo Fisher Scientific).Real-time PCR was performed with a 7500 Real-time PCR System apparatus (Applied Biosystems, Thermo Fisher Scientific) on the reverse transcription products with TaqMan master mix and TaqMan™Gene Expression Assays (Applied Biosystems, Thermo Fisher Scientific).
Annealing temperature was 60 °C for all the primer pairs listed.All samples were run in duplicate, and each PCR well contained 20 μl as a final volume of reaction, including 2 μl of complementary DNA corresponding to approximately 60 ng total RNA, 750 nM of each primer, and 10 μl PCR master mix.Thermal cycling conditions were as follows: 1 cycle at 95 °C for 10 min; 40 cycles at 95 °C for 15 s, and 60 °C for 1 min.The relative expression level of each mRNA in different experimental groups was calculated using the 2 −ΔΔCt method, normalized to HPRT and expressed as fold change relative to the Opt Se group, taken as 1.

ELISA determination of cytokine levels
IL-1β and IL-6 protein levels accumulated in culture supernatants were assessed by specific ELISAs from Diaclone (detection limit for IL-1β: 4 pg/ml and for IL-6: 12 pg/ml), according to manufacturer's instructions.

Nitrite levels determination
Production of nitric oxide (NO) was determined by measuring the content of nitrite, one of the end products of NO oxidation in the media, as previously described 45 .A standard nitrite curve (0.25-50 µM) was generated using a 10 mM solution of NaNO 2 .All chemicals for the NO assay were from Sigma.

Statistical analyses
Data are expressed as means ± standard error of the mean (SEM).Statistical analyses were performed using unpaired Student's T-test for comparison between two groups, and Two-way ANOVA for multiple comparisons, where the effect of "diet" is defined as Se diet and "treatment" is defined as Pb exposure.Tukey's post hoc test on

Effects of maternal Suboptimal Se diet on brain immune-related gene expression in the offspring
We evaluated whether a prolonged maternal Subopt dietary intake of Se could alter-in their offspring-the basal Central Nervous System (CNS) expression of a selected group of markers involved in immune functions, brain homeostasis and development.Brains were collected on PND 1 from offspring of dams fed with Opt or Subopt Se diets, and the expression of the inflammatory cytokines interleukin (IL)-6 and IL-1β, the immunomodulatory cytokine IL-10, the inflammatory/oxidative stress-related enzymes inducible nitric oxide synthase (iNOS) and arginase-1 (Arg-1), and the microglial/macrophage phagocytic receptors cluster of differentiation (CD)11b and Mannose Receptor C-Type 1 (MRC-1), were analyzed by Real Time PCR.
We found increased mRNA levels of IL-6, IL-1β, IL-10, CD11b and MRC-1 in the Subopt group when compared to the Opt group (Fig. 2), suggesting that Subopt Se intake during gestation profoundly alters immune/ plasticity-related functions in the offspring brain.The levels of the inducible enzyme iNOS, responsible for nitric oxide (NO) production under inflammatory conditions, showed only a tendency to increase in the Subopt group, while Arg-1 levels did not change.

Effects of different maternal Se dietary intakes and Pb exposure on offspring's microglia
The effects of low Se, Pb exposure and their combination on the properties of neonatal microglia were assessed in microglial cultures from the cortex of rat pups (PND 1) from each of the four experimental groups i.e.: Se Opt/ Veh, Se Opt/Pb, Se Subopt/Veh, and Se Subopt/Pb, analysing the mRNA expression of the inflammatory markers examined in whole brain extracts (i.e.IL-6, IL-1β, IL-10, iNOS, Arg-1, MRC-1, CD11b; Fig. 3).
IL-10, iNOS, and MRC-1 mRNAs levels (Fig. 3C,D,G, respectively) were higher in microglia from Subopt compared to the Opt group, in line with what observed in whole brain extracts, while IL-1β (Fig. 3B) and CD11b (Fig. 3F) showed only a tendency to increase.In addition, Arg-1 (Fig. 3E) was down-regulated in Subopt microglia.Exposure to Pb during gestation did not further modify the specific expression of these genes in offspring's microglia due to diet (Opt or Supopt Se-diet), whereas it strongly increased that of IL-6 (Fig. 3A) in conditions of Subopt Se intake only (2-way ANOVA results and Tukey's post hoc analyses for all genes are reported in Fig. 3 legend).

Effects of different maternal Se dietary intakes and Pb exposure on offspring's microglial reactivity to LPS
To test whether Suboptimal maternal Se intake and Pb exposure could alter offspring microglia reactivity to a further inflammatory challenge, microglial cultures isolated from the cortex of PND1 rat pups from the four experimental groups as above, were stimulated with 10 ng/ml LPS for 24 h.At the end of treatment, we measured the levels of nitrite (as an index of NO production) and the accumulation of the cytokines IL-1β and IL-6, in microglial supernatants.As expected, after LPS stimulation, the levels of these metabolites were significantly induced in all experimental groups (Fig. 4A-C), while they remained undetectable in unstimulated cultures (not shown).
Higher levels of IL-6, IL-1β and nitrite, as well as reduced levels of Arg-1 were measured in Subopt/Veh compared to Opt/Veh cultures (Fig. 4A-D), suggesting a greater reactivity to LPS in suboptimal Se condition.The co-occurrence of Pb exposure did not modify IL-6 production while we found a strong interaction between diet and Pb in the modulation of IL-1β and nitrite.Indeed, perinatal Pb exposure enhanced nitrite production in response to LPS in Subopt cultures, while increased that of IL-1β only in Opt microglia.2-way ANOVA results and Tukey's post hoc analyses for all metabolites are reported in Fig. 4 legend.
In the cell extracts obtained from the same cultures, we analyzed the mRNA expression of the enzyme Arg-1, and of the receptors CD11b and MRC-1 (Fig. 4D,E).As expected, LPS upregulated Arg-1 and CD11b expression in all experimental groups, while it reduced that of MRC-1.Also in this case, the effects of perinatal Pb exposure were dependent on maternal Se intake, as Arg-1 mRNA levels were reduced by Pb in condition of optimal Se intake and moderately increased in suboptimal condition.In any case, as a net effect, Arg-1 levels in Subopt/Pb cultures were lower than in Opt/Pb.As concerning CD11b, Pb exposure had no effect in optimal Se conditions while it further enhanced its expression in Subopt cultures.
At variance with the other genes analyzed, MRC-1 mRNA levels (Fig. 4F) were comparable in all four experimental groups and remained lower than in control-unstimulated cultures (2-way ANOVA and Tukey's post hoc values for all the genes analyzed are reported in Fig. 4 legend).

Effects of different maternal Se dietary intakes and Pb exposure on hippocampal immune homeostasis and reactivity to LPS in the offspring: organotypic hippocampal slice cultures (OHSCs)
To assess the possible enduring consequences of different dietary Se contents and Pb exposure throughout pregnancy and lactation on hippocampal immune homeostasis in the offspring, we established OHSCs from PND 6 pups.To avoid the confounding effects of the inflammatory reaction consequent to the mechanical manipulation of the slices, OHSCs were cultured for 1 additional week before stimulation.The slices were then either stimulated or not with LPS (5 µg/ml) for 24 h, and the mRNA levels of IL-6, IL-1β, IL-10, iNOS, Arg-1 and CD11b, were evaluated by Real Time PCR (Fig. 5).
As shown in Fig. 5A-C,E, basal OHSCs obtained from the Se Subopt groups had higher mRNA levels of IL-6, IL-1β, IL-10, consistently with a state of higher activation, as observed in whole neonatal brains and neonatal purified cortical microglial cultures.Arg-1 mRNA was upregulated, differently from what observed in whole brain extracts and cortical microglia, suggesting a brain-region specific regulation of this gene.
Perinatal Pb exposure did not change the expression of IL-6 and IL-1β whereas it reduced that of Arg-1, independently from the diet.A strong interaction between diet and Pb was found for IL-10 and iNOS mRNAs (Fig. 5C,D) that were downregulated only in optimal Se condition (2-way ANOVA and Tukey's post hoc results are reported in the figure legend).As for CD11b mRNA levels (Fig. 5F), they were comparable in all groups.
We then analyzed the reactivity to LPS of OHSCs from the different experimental groups, measuring the levels of IL-6, IL-1β, and nitrite in the culture media, following 24 h of stimulation (Fig. 6).An interaction between diet and Pb was found for IL-6 and nitrite that were increased by Pb in the Subopt condition (Fig. 6A,C).IL-1β protein levels (Fig. 6B) were higher in Subopt cultures as a whole, and not modified by Pb exposure (2-way ANOVA and Tukey's post hoc analyses are reported in the figure legend).
RT PCR analysis of the LPS-induced expression of Arg-1 mRNA (Fig. 6D), showed significantly higher levels in the Se Subopt group as a whole compared to the Opt one.In OHSCs from Pb exposed rats, Arg1 mRNA levels showed a tendency to decrease in both dietary groups.CD11b mRNA expression, analyzed in the same cultures, remained unmodified in all groups (Fig. 6E).

Discussion
In this study, we identified for the first time that maternal Se dietary restriction, developmental exposure to a low Pb dose, and their combination induce specific molecular changes in the offspring's microglia evident from birth, thus advancing our knowledge of the cellular targets that may be responsible for the adverse neurobehavioral outcomes related to this condition, recently described by our group 30 .First, we show that Se suboptimal diet, per se, increased whole brain levels of the inflammatory cytokines IL-6, IL-1β, and IL-10 and of the microglial phagocytic receptors CD11b and MRC-1 in newborn rats.
In cortical offspring microglia, we confirmed the higher expression of IL-6, IL-10, MRC-1, and a moderate increase of IL-1β and CD11b, in suboptimal Se condition supporting the acquisition of an inflammatory activation state that can influence fundamental physiological brain functions.Moreover, higher mRNA levels of iNOS and reduced levels of Arg-1 were found in suboptimal Se, an effect that was masked when analysing the whole brain.
Higher mRNA levels of IL-6, IL-1β, IL-10, and iNOS were observed also in organotypic hippocampal slices (OHSCs), suggesting a persistent inflammatory polarization occurring in the hippocampus at a more advanced stage of development.Indeed, the maturational stage reached by OHSCs at the time of stimulation (1 week post explant) roughly corresponds to PND 14-15, as hippocampal slices in vitro closely recapitulate the developmental program observed in vivo 46,47 .
The inflammatory molecules analyzed in this study play critical roles in almost every aspect of neural development, including neurogenesis, synaptogenesis, pruning and plasticity [48][49][50] .Dysregulations of cytokines, such as IL-1β, IL-6 and IL-10, and relative signaling pathways, have been implicated as factors involved in neurodevelopmental disorders 51 .
As an example, IL-1β and IL-6 has been found increased in maternal serum, amniotic fluid, and serum of children subsequently diagnosed with autism spectrum disorders (ASD) as well as in the well-known maternal immune activation rodent model 52 .A deregulated expression of the enzymes iNOS and Arg-1, which are mainly produced by microglia/macrophages in the CNS, may compromise important aspects of brain development 53,54 .
The microglial/macrophage-specific marker CD11b likewise displays a key developmental role, mediating the phagocytosis of supernumerary synapses and dendritic spines in the immature brain, thus contributing to neural circuit shaping.An altered complement signaling early in development results in sustained defects in synaptic connectivity, and it might contribute to neurodevelopmental diseases 55,56 .An important developmental function has been suggested also for the other microglial receptor analyzed, MRC-1 57 .Its deregulated expression was associated with abnormal neuronal development in a model of perinatal inflammation 58 .
Based on this evidence, we hypothesize that early microglial alterations described here can contribute, at least in part, to behavioral alterations later in life, as those reported in vivo in the same model of maternal Se restriction, including increased ultrasonic vocalisation emissions (an index of distress and increased emotional arousal), increased wall climbing, and locomotion 30 .
In addition, we provide evidence that optimal Se nutrition protects cortical microglia from the effects of Pb exposure, while some inflammatory changes are induced by Pb in hippocampal organotypic cultures also in optimal Se condition.The heterogeneity of the cell populations that constitute the OHSC model, which includes glial and neuronal cells, could in part account for the different response compared to isolated microglial cultures.In OHSCs, where the complex three-dimensional cell architecture is largely maintained, the basal behavior as well as the response to an inflammatory challenge are indeed the result of the crosstalk among the cell components, which contribute to the overall modulation of gene expression.Among the OHSC glial populations, astrocytesplaying along with microglia a key role in maintaining brain homeostasis and innate immune functions-are likely the main contributors to the different responses of OHSCs vs. primary microglia, through their activation and release of cytokines and other inflammatory mediators.
Beyond the mentioned differences between the two culture models, their responses are also likely reflecting the different susceptibility of cortex and hippocampus to Pb toxicity and to the effects of Pb interaction with Se status 59,60 .As we previously showed in vivo, optimal Se status during pregnancy counteracted the effects of Pb on somatic growth in pre-weaned pups and on locomotor activity and anxiety at later life stages, while Pb, regardless of Se status, altered glutamatergic receptor subunit trafficking to the post-synapse in the juvenile hippocampus and affected the long-term recognition memory in adult rats 30 .Although the molecular mechanisms underlying Pb neurotoxicity as well as the protective effects of optimal Se need to be further explored, our study suggest the convergence of Pb and Se on common neuro-inflammatory pathways.While few papers reported pro-inflammatory effects of high-level of Pb exposure in the young rodent brain 61,62 , and several studies demonstrated Se antioxidant and anti-inflammatory effects [9][10][11]63 , we provide the first evidence of a possible interaction of maternal dietary Se and low-Pb exposure in the regulation of neuroplasticity-and neuroinflammatory-related genes in the offspring's brain.
Another relevant aspect of our study is the observation that suboptimal Se nutrition, per se, and the combination with Pb exposure prime the developing brain enhancing its reactivity to a subsequent immune challenge, such as LPS.Indeed, both cortical microglia and hippocampal organotypic cultures from Se suboptimal fed-offspring showed a stronger response to LPS than optimal cultures, as indicated by higher levels of IL-6, IL1-β and-in the case of microglia-also of nitrite.The co-occurrence of suboptimal Se diet and Pb exposure exacerbated the LPS-induced production of nitrite, and increased that of CD11b in cortical microglia, as well as increased that of nitrite and IL-6 in OHSCs.These data reveal that the interaction among maternal dietary factors and an adverse environmental exposure since early phases of development set mechanisms of innate immune memory, inducing a long-lasting molecular reprogramming that potentially leads to harmful outcomes for the offspring facing a secondary insult.This so-called immune "training", to use a recently proposed nomenclature 64 , was formerly described in experimental and clinical conditions of neurodegenerative diseases.More recently, similarly to our findings, it was proposed as a mechanism that can be set also in the foetal life as a consequence of maternal stress 65 , maternal immune activation 66 , and maternal high-fat diet 38 , and that was associated to altered trajectories of brain maturation and increased vulnerability to psychopathologies in the offspring 67 .
The sexual dimorphism of microglia reactivity is increasingly viewed as a key factor determining the sex bias observed in many neurodevelopmental and neurological disorders, with males particularly vulnerable to life-long illnesses of neurodevelopmental origin.In this study, we analysed data by combining the results of the two sexes and this could have masked sex-specific effects, which deserve to be explored in future studies.
Despite this limitation, we believe that our results are highly informative on the importance of adequate maternal nutrition for the correct programming of offspring development and protection against neurotoxic substances.
Further studies are needed to understand the influence of cellular crosstalk, by mean of experimental cell co-culture systems, to address the molecular mechanisms underpinning offspring's microglial training by low maternal Se intake and developmental Pb exposure, and to establish to which extent are these effects reversible and amenable of intervention.

Conclusions
Overall, our findings support the notion that optimal maternal Se nutrition during pregnancy and lactation is essential for proper offspring development and that is able to counteract, at least in part, the toxicity of developmental exposure to Pb.These results also show that the interaction among maternal dietary factors and an adverse developmental environment can shape brain immune reactivity since early phases of development in a regionspecific manner, possibly contributing to the risk of neurodevelopmental and neurodegenerative disorders.

Figure 2 .
Figure 2. Relative levels of inflammatory gene mRNAs, as measured by real-time RT-PCR in whole brain extracts (PND 1), expressed as fold change versus the levels found in Se Opt group taken as 1, and normalized to HPRT as the reference gene in each sample.The transcript levels for IL-1β, IL-6, IL-10, CD11b and MRC-1 were significantly higher in brains from the Se Subopt group when compared to the Se Opt group.Data are represented as mean ± SEM (n = 6 per group; unpaired T test: *p < 0.05 vs. Se Opt).