LPS-induced histone H3 phospho(Ser10)-acetylation(Lys14) regulates neuronal and microglial neuroinflammatory response

neuroinflammatory


Introduction
Numerous studies on animal models have shown that systemic inflammatory stimuli, such as administration of lipopolysaccharide (LPS), trigger an inflammatory response in the brain, despite the presence of the blood-brain barrier, resulting in microglial cell activation and increased expression and release of inflammatory cytokines that can persist long after peripheral events have subsided (Cunningham, 2013;Qin et al., 2007).In response to a pathogen, activated innate immune cells in the periphery produce pro-inflammatory mediators such as the cytokines interleukin (IL) 1β, IL-6, and Tumour Necrosis Factor (TNF) α and nitric oxide (NO) that can act also within the brain resulting in a strong effect on neuroinflammation (Henry et al., 2009;Sierra et al., 2007).This exaggerated pro-inflammatory response can be neurotoxic, can lead to behavioral alterations and exacerbate pathological processes leading to neurodegenerative and neuropsychiatric diseases (Benatti et al., 2016;Godbout et al., 2005;Layé et al., 1994;Matt et al., 2016;Qin et al., 2007).
Inflammatory mediators seem to act at different levels in the central nervous system (CNS) and determine profound alterations in neurological and endocrine functions, particularly in the areas of the limbic system (Alboni and Maggi, 2015;Godbout et al., 2005;Layé et al., 1994;Qin et al., 2007).
Emerging observations suggest that patterns of gene expression induced by neuroinflammation are activated or silenced through alterations of the chromatin structure (Claycombe et al., 2015;McClung and Nestler, 2008).Epigenetic modifications of both DNA and histone proteins are now accepted as fundamental mechanisms by which neural cells adapt their transcriptional response to developmental and environmental cues (Bagot et al., 2014;Nestler, 2014).Epigenetic regulation may influence gene expression through stable, long-term modifications that involve DNA methylation, or through a more flexible (short-term) way that involves histone modifications, such as phosphorylation and acetylation (Chen et al., 2002;Garden, 2013;Handy et al., 2011;Weaver et al., 2004;Zhou et al., 2011).In particular, the present study was focused on a specific histone modification, the simultaneous phosphorylation on the residue of Serine-10 (S10) and acetylation on the residue of Lysine-14(K14) in the N-terminal of the histone H3.
Phospho(S10)-acetylation(K14) of histone H3 (H3S10phK14ac) is dynamic and reversible and could serve to amplify the readout of upstream signaling pathways causing greater changes in the overall charge density of tails that resulting in greater modification of the chromatin structure of target genes (Colvis et al., 2005;Nowak and Corces, 2000;Rea et al., 2000;Thompson et al., 2001).Further evidence showed an increase of H3 phosphorylation in hippocampal neurons after seizures, that occurs in the c-Fos promoter region determining an elevation of c-Fos gene expression (Clayton et al., 2000;Mori et al., 2013).c-Fos is one of the most widely studied target genes because it is critical for synaptic plasticity and hippocampal memory (Alboni et al., 2017;Fleischmann et al., 2003;He et al., 2002;Tischmeyer and Grimm, 1999).Recent studies demonstrated that neuronal activity induces the formation of DNA double-strand breaks, also within the promoters of early-response genes, such as c-Fos, facilitating their expression (Corbett et al., 2017;Madabhushi et al., 2015;Olsson et al., 2009;Watson and Tsai, 2017).
Little is known about the role of histone modifications in regulating transcription of specific genes involved in neuroinflammation.Recent studies suggest H3S10phK14ac capacity to activate transcriptional response and evidence shows the induction of H3S10phK14ac after administration of LPS in an invertebrate model (Garden, 2013;Ottaviani et al., 2013).It was also demonstrated that LPS-induced phosphorylation (S10) and phospho-acetylation (S10/K14) of H3 histone are crucial for recruiting the transcription factor NF-κB to the promoter of specific cytokines genes, such as IL-12 (Hamon and Cossart, 2008;Saccani et al., 2002).
Based on this knowledge, the purpose of this work was to demonstrate the regulation by H3S10phK14ac of pro-inflammatory gene expression in a rat model of neuroinflammation, focusing the study on the cellular populations more involved in the inflammatory response and epigenetic mechanisms.

Animals and treatment
Adult male rats (Sprague-Dawley, Charles-River Laboratories, Calco (LE), Italy) were used in this study.Animals were housed in polycarbonate cages with ad libitum access to food and water throughout the study and maintained under a 12:12 light-dark cycle in an ambient temperature of 21±3°C with relative humidity controlled.The procedures used in this study were in strict accordance with European legislation on the use and care of laboratory animals (EU directive 2010/63/EU), with the guidelines of the National Institutes of Health on the use and care of laboratory animals (NIH Publications No.8023) and had the approval of the Italian Ministry of Health and of the local Ethical Committee.All efforts were made to minimize animal suffering and to reduce the number of animals used in this study.LPS (830 μg/Kg/mL, Escherichia Coli serotype 0127:B8, Sigma Aldrich L3129) was injected intraperitoneally (i.p.) and an equal volume of vehicle (saline) was used as a control.The dose of LPS was selected on the basis of its ability to induce sickness behavior and cytokine expression in the periphery and in the brain (Dantzer et al., 2008;Dantzer and Kelley, 2007;Lawson et al., 2013;Layé et al., 1994).Animals were sacrificed 2 h or 6 h after injection, hippocampi and hypothalamus were harvested, immediately frozen on dry ice, and stored at -80°C until further molecular analysis.

Cell cultures and treatments
The mouse clonal hippocampal neuronal cell line HT-22 was a generous gift from Dr. Maher Pamela (The Salk Institute for Biological Studies, La Jolla, San Diego, CA).The murine microglial cells BV-2 were obtained from Professor Samuele Peppoloni (University of Modena and Reggio Emilia, Modena, Italy).Cells were maintained at 37°C and 5% CO 2 in complete medium containing DMEM High glucose for HT-22, or RPMI 1640 for BV-2, supplemented with 10% heat-inactivated (56°C, 30 min) fetal bovine serum (FBS Gibco ® Italy, Milan), 50 U/ml penicillin, 50mg/ml streptomycin and passed by tripsinization.Cells (2x10 6 ) were seeded into Petri dishes and grown for 20-24 h until 70-80% confluent for protein analysis.For gene expression analysis cells were plated at a density of 1x10 6 cells per well on 6-well plates in 2 mL of complete medium and treated the day after plating.Cells were harvested 2, 6 or 24 h after treatment with 100 ng/mL of LPS (Escherichia Coli serotype 0127:B8, Sigma Aldrich L3129), while the control group received phosphate-buffered saline (PBS).
Two independent experiments were performed.
A fraction of the lysate was collected (total extract) and the remaining fraction was centrifuged at 1500 g for 5 min at 4°C.The supernatant was collected (cytoplasm fraction) and the pellets were resuspended in a Low Salt Buffer (20 mM Hepes, 2 mM MgCl 2 , 0.1 M KCl, 25% Glycerol, 5 mM Dithiothreitol (DTT)).The fraction was centrifuged at 1500 g for 5 min at 4°C.The supernatant was discarded, and the pellets were resuspended in a High Salt Buffer (20 mM Hepes, 2mM MgCl 2 , 1M KCl, 1mM EDTA, 25% Glycerol, 5 mM DTT), then vortexed vigorously and kept in ice for 30 min.
Finally, the lysate was centrifuged at 13200 rpm for 30 min at 4°C.The supernatant was collected, and the pellets were resuspended in a Resuspension Buffer (120 mM NaCl, 20 mM Hepes, 0.1 mM EGTA, 0.1 mM DTT, protease and phosphatase inhibitor) (nuclear enriched extract).
Protein concentration was determined using standard protocol Coomassie® reagent (Thermo Fisher Scientific, Waltham, MA USA ).Total, cytosolic and nuclear fractions were stored at -20°C.

Total RNA extraction, reverse transcription, and real time polymerase chain reaction
RNA extraction and DNAse treatment were performed as previously described (Benatti et al., 2014) using GenElute™ Mammalian Total RNA Miniprep Kit and DNASE70-On-Column DNase I Digestion Set (Sigma Aldrich ®, Milan, Italy).Two µg of total RNA were reverse transcribed with High Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific, Waltham, MA USA) and Real Time PCR was performed, as previously described (Benatti et al., 2018), in ABI PRISM 7900 HT (Thermo Fisher Scientific, Waltham, MA USA) using Power SYBR Green mix (Thermo Fisher Scientific, Waltham, MA USA) and specific forward and reverse primers at a final concentration of 150 nM (Table 1 for both in vitro and in vivo experiments).Cycle threshold (Ct) value was determined by the SDS software 2.2.2 (Thermo Fisher Scientific, Waltham, MA USA), mRNA expression was calculated with the ΔΔCt method with glyceraldehydes-3-phosphate dehydrogenase (GAPDH) as endogenous control.

ChIP and PCR analyses
Chopped hippocampus and hypothalamus of LPS (n=6) and saline (n=6) treated rats were crosslinked with 1% formaldehyde for 15 min then the reaction was stopped by adding 0.125M glycine for 5 min at RT.Fixed tissues were washed twice with ice-cold PBS and homogenized by potter (15 strokes at 60 rpm) in 1 mL of PBS in the presence of protease inhibitor Complete.Samples were centrifuged at 5000 rpm for 5 min at 4°C, pellet was resuspended in Cellular Lysis Buffer (10 mM Tris-HCl pH 8.0, 10 mM NaCl, 0.2% Igepal, protease and phosphatase inhibitors) and kept on ice for 15 min.The lysate was centrifuged at 5000 rpm for 5 min at 4°C then the supernatant was collected, and the pellet resuspended in Nuclear Lysis Buffer (50 mM Tris-HCl pH 8.0, 10 mM EDTA, 0.5 % SDS, protease and phosphatase inhibitors) and kept on ice for 30 min.The chromatin was sonicated 4 times (30 sec on/30 sec off) at 270 setting in Labsonic sonicator (B.Braun International®) to produce 500-1000bp fragments, then was centrifuged at 14000 rpm for 10 min.The protein concentration was measured using standard protocol Coomassie® reagent (Thermo Fisher Scientific, Waltham, MA USA).The volume of lysate for 1 μg/μL was diluted to a final volume of 2 mL in a mixture of 9 parts Dilution buffer (1.1% Triton X-100, 1.2 mM EDTA pH 8, 16.7 mM Tris-HCl pH 8 and 167 mM NaCl) and 1 part of Lysis buffer (1% SDS, 10 mM EDTA pH 8, 50 mM Tris-HCl pH 8) and pre-cleared for 2 h, rotating at 4°C, with 10 μL of protein G Agarose beads 50% slurry (KPL, Gaithersburg MD, USA).
Samples were centrifuged at 2000 rpm for 5 min and the supernatant was incubated with 3 μL of the antibody anti-phospho(S10)-acetyl(K14) histone H3 (Millipore®, #07-081) overnight rotating at 4°C.Simultaneously, 10 μL of protein G beads were incubated with a pre-blocked mix consisting of 9:1 (Dilution: Lysis), 100 μg/mL BSA and 500 μg/mL sheared salmon sperm DNA in a final volume of 100 μL for each chromatin and rotated overnight at 4°C.After immuno-precipitation, 10 μL of protein G pre-blocked beads, previously washed, were added to each sample and incubated for 2 h while rotating at 4°C.The unbound material was recovered as input and beads were washed 3 times with Wash Buffer (1% Triton X-100, 0.1% SDS, 150 mM NaCl, 2 mM EDTA pH 8.0, 20 mM Tris-HCl pH 8.0) and once with a Final Wash Buffer (1% Triton X-100, 0.1% SDS, 500 mM NaCl, 2 mM EDTA pH 8.0, 20 mM Tris-HCl pH 8.0).ChIPed material was eluted by two 15 min incubations at RT with 250 μL of Elution Buffer (1% SDS, 0.1 M NaHCO 3 ).Chromatins, including input samples, were cross-linked reverse and DNA was submitted to RNAse and proteinase K digestion by incubating first at 37°C for 1 h, then at 65°C for 4 h minimum, and finally extracted by phenol-chloroform.Immunoprecipitation reactions were performed in duplicate using a no-antibody sample as a non-specific control.Immunoprecipitated DNA and inputs of hypothalamus and hippocampus were analyzed by a gradient PCR analysis using GoTaq® Flexi DNA Polymerase (Promega® Italia, Milan, Italy).
The promoter region of the gene of interest was amplified using specific primers as reported in Table 1.PCR results were analyzed by ImageJ software: each immuno-precipitated chromatin was normalized to the corresponding total chromatin (input).

Sample preparation for histology
For histological processing, animals were anesthetized with chloral hydrate (400 mg/kg, i.p.).
Intracardial perfusion was performed with 4% paraformaldehyde and 0.2% picric acid in Phosphate Buffered Saline (PBS) (100mL/15min) preceded by an infusion of 50 mL of 0.9% NaCl saline containing heparin sodium (5000U/L); then, brains were dissected out.The brains were post-fixed in the same solution for 12 h, rinsed in 15% PBS-sucrose for approximately 12 h and then in a 30% PBS-sucrose for 1 day.The brains were frozen using dry ice, and coronal 50 μm thick sections series were cut at a cryotome, washed three times in cold PBS and stored at -20°C in a glycerol-PBS solution until use.

Statistical analysis
Statistical analysis was performed by means of two-way analysis of variance (ANOVA) using treatment group and time as factors.In case of significant effects, further comparisons between LPS-and vehicle-treated groups were performed by means of one-way ANOVA.p<0.05 was considered as a threshold for significant difference.
For gene expression analysis, the mRNA levels of each target were normalized to the endogenous control, GAPDH.Endogenous control mRNA levels were not affected by any treatment (p>0.05,Oneway ANOVA).For quantitative evaluation of changes the comparative ΔΔCt method was performed, using as calibrator the average levels of expression of vehicle-receiving animals sacrificed 2 h after injection.
For ChIP analysis, quantitative enrichment of LPS sample, compared to control sample, was statistically performed by means of an independent Student's t-test and expressed in fold changes.
No difference was observed in both expression and protein levels of the evaluated targets between vehicle treated animals sacrificed 2 or 6 h after the injection (p>0.05).
Statistical analysis for in vitro results was performed by means of one-way ANOVA followed by Dunnett's post hoc test.
Data are presented as mean ± standard error of the mean (S.E.M.).Analyses were conducted using SPSS for Windows® v.25 (SPSS Inc., Chicago, USA).

Systemic LPS induces the expression of the pro-inflammatory cytokines IL-1β, IL-6, and TNF-α,
the enzyme iNOS, and the immediate early gene c-Fos in rat hypothalamus and hippocampus.
In order to ensure the effectiveness of LPS (830 μg/Kg, i.p.) in stimulating the immune system and induce an inflammatory condition in the CNS, the expression levels of the major pro-inflammatory cytokines known to be produced within the brain, such as IL-1β, IL-6 and TNF-α, were measured in the rat hypothalamus and hippocampus 2 and 6 h after a systemic LPS challenge (Dantzer, 2009;Dantzer et al., 2008).
Subsequently, we investigated the mRNA levels of the inducible nitric oxide synthase (iNOS) which plays an important role in the response to pathogens, as well as in tumor growth, and produces NO upon stimulation by pro-inflammatory cytokines such as IL-1 and TNF-α (Carreira et al., 2015;Moraes et al., 2006;Satta et al., 1998).In both the hypothalamus and hippocampus, highly significant increases in iNOS transcript levels were detected at 2 h (about 110 and 50 times, respectively) and were further increased at 6 h (about 800 and 230 times, respectively) after LPS injection (Fig. 1D Finally, we tested the effects of LPS on the expression of the immediate early gene c-Fos, which is considered to be a marker of neuronal activity and shown to be induced after the activation of the peripheral innate immune system in a rapid and transient manner in different brain areas proposed to be involved in the cytokine-induced sickness response.Though less intense than for cytokines and iNOS, a significant increase of c-Fos mRNA levels was detected at both 2 and 6 h (about 3.5 and 3 times, respectively) after LPS injection in the hypothalamus, whereas no significant change was detected in the hippocampus (Fig. 1E
We measured, by western blot, the effects of an inflammatory stimulus on the levels of histone H3 phosphorylation (S10) (H3S10ph) and H3S10phK14ac in the nuclear fractions extracted from the rat hypothalamus and hippocampus 2 and 6 h after the systemic administration of LPS (830 μg/Kg, i.p.).
In both the hypothalamus and hippocampus, a significant increase of H3S10phK14ac levels was detected 2 h after LPS treatment, whereas at 6 h H3S10phK14ac levels were markedly decreased in both regions though a trend for a significant increase was still present in the hypothalamus (Fig. 2A   and B

Histone H3 phospho(S10)-acetylation(K14) is a mediator of the transcriptional response induced by LPS-challenge.
It has been suggested that the status of a chromatin region is coded by the combination of differentially modified histone tails (Sawicka and Seiser, 2012).Among these modifications, H3S10phK14ac results in transcriptional activation by opening chromatin conformation (Chen et al., 2002;Cheung et al., 2000;Tsankova et al., 2004).Here we hypothesized that LPS-induced transcriptional activation of IL-1β, IL-6, iNOS or c-Fos genes may be regulated by the phosphoacetylation status of histone H3.Therefore, we evaluated the association of H3S10phK14ac to the promoter region of IL-1β, IL-6, iNOS, and c-Fos genes through ChIP experiments.Results showed that H3S10phK14ac was able to regulate the LPS transcriptional response in a time-and regionspecific manner.
Two h after LPS administration, in the hypothalamus, PCR analysis showed a strong increase in H3S10phK14ac association to iNOS and c-Fos promoters compared to vehicle (t = 33.304,p=0.0009 and t=21.638,p=0.002, respectively) (Fig. 4A).A mild increase was observed for IL-6 gene (t=20.684,p=0.002) (Fig. 4A), while no change in H3S10phK14ac association was found for IL-1β (Fig. 4A).On the other hand, ChIP analysis in the hippocampus highlighted an H3S10phK14ac enrichment of over 10 times only to iNOS promoter after LPS administration when compared to vehicle-receiving counterparts (t=53.069,p=0.0004) (Fig. 4B).
Six h post LPS exposure, in the hypothalamus, PCR results showed an increased association of H3S10phK14ac to iNOS gene approximately 6 times greater than what was observed in the vehicle group (t=40.669,p=0.0006) (Fig. 4C).Instead, no changes in binding of H3S10phK14ac was detected for IL-1β, IL-6 and c-Fos gene promoters (Fig. 4C).Similarly, in the hippocampus, no alteration was found in H3S10phK14ac association to the promoters of any of the target genes examined (Fig. 4D).
All samples were compared to total chromatin (input) and to the negative control without antibody (Fig. 4 A-D).
Therefore, we decided to assess what cell types express increased H3S10phK14ac levels upon systemic exposure to LPS.In particular, neurons, astrocytes and microglial cells in rat hypothalamus and hippocampus were analyzed by immunostaining with cell-specific antibodies including the neuron-specific anti-NeuN, the astroglia-specific anti-GFAP, and the microglia-specific anti-Iba-1.Scattered H3S10phK14ac immunoreactive nuclei were detected in several hypothalamic nuclei of control rats (see, Fig. 5A, D, G and Fig. 6A, D, G).A clear increase in nuclear H3S10phK14ac immunoreactivity was observed after LPS administration, though with different patterns of change in the different hypothalamic nuclei.In the magnocellular paraventricular nucleus, positive nuclei markedly increased at 2 h and further increased at 6 h (not shown).In the parvocellular paraventricular (Fig. 5) and periventricular (not shown) nuclei, the marked increase at 2 h (Fig. 5B,   E, H) was followed by a slight decrease at 6 h (Fig. 5C, F, I).Finally, in the arcuate nucleus, the sharp increase of H3S10phK14ac positive signal at 2 h was followed by a marked decrease at 6 h, that did not, however, reach control levels (not shown).A clear increase in H3S10phK14ac immunoreactivity was also observed in ependymal cells of the III ventricle (Fig. 6) both at 2 (Fig. 6B, E, H) and 6 h (Fig. 6C, F, I) after LPS administration as well as in vascular endothelial cells (not shown).
In double immunofluorescence experiments using antibodies against H3S10phK14ac and cellspecific markers, we could identify the cellular identity of the H3S10phK14ac immunoreactive nuclei.The large majority of H3S10phK14ac immunoreactive nuclei co-expressed NeuN immunoreactivity and were therefore identified as neuronal nuclei (Fig. 7A, B).Instead, we could not identify clear-cut instances of co-localization between H3S10phK14ac immunoreactivity and the nuclei of astroglial cells, identified by GFAP immunoreactivity, in hypothalamic nuclei (not shown).
Finally, regarding the nuclei of microglial cells, identified by Iba-1 immunoreactivity, a moderate level of H3S10phK14ac/Iba-1 co-localization was detected at 2 h in some cells of the hypothalamus (Fig. 7C).Instead, several instances of H3S10phK14ac/Iba-1 co-localization were detected at 6h in the hypothalamic nuclei examined (e.g., paraventricular, periventricular, and arcuate nuclei) (Fig. 7D).These latter microglial cells showed an amoeboid/phagocytic form.The intensity of H3S10phK14ac immunoreactivity in Iba-1 positive nuclei was generally lower than the intensity of H3S10phK14ac immunoreactivity in NeuN-positive nuclei, with the exception of a few H3S10phK14ac/Iba-1 double immunoreactive cells with very intense H3S10phK14ac immunoreactivity.Notably, these latter microglial cells were in close proximity to ependymal cells of the ventricles and putative vascular endothelial cells (Fig. 7D).
In the hippocampus, dispersed H3S10phK14ac immunoreactive nuclei were detected in control rats all along the different subregions and layers (not shown).Two h after LPS, a sharp increase in H3S10phK14ac immunoreactive nuclei could be detected both in layers containing projection neurons (pyramidal and granule cell layers) and layers containing interneurons (Fig. 8A-C).At 6 h a decrease in H3S10phK14ac immunoreactivity was detected, in particular in pyramidal and granule cell layers (Fig. 8D-F).In all regions, several instances of H3S10phK14ac immunoreactive vascular endothelial cell nuclei were observed at both 2 and 6 h after LPS administration (not shown).
Double immunofluorescence experiments with cell type-specific markers, delineated a different pattern of staining of the principal cell types of the neural tissue.While only scattered NeuN cells showed H3S10phK14ac immunoreactivity in control rats, H3S10phK14ac/NeuN double immunoreactive cell nuclei markedly increased 2 h after LPS administration (Fig. 8A) and decreased at 6 h (Fig. 8D) without reaching control levels.We further analyzed the expression of H3S10phK14ac immunoreactivity in specific neuronal subpopulations expressing neuropeptide Y (NPY) or glutamic acid decarboxylase (GAD) immunoreactivity.We could detect both H3S10phK14ac/NPY and H3S10phK14ac/GAD double immunoreactive neurons at 2 h, that were no longer present at 6 h.H3S10phK14ac/GFAP double immunoreactive cells were rare in control rats and increased at both 2 and 6 h (Fig. 8G-J).Note that at 6 h H3S10phK14ac/GFAP double immunoreactive cells had a reactive phenotype.Finally, H3S10phK14ac/Iba-1 double immunoreactive cells were not detected in control rats (not shown) and 2 h (Fig. 8G) after LPS administration but were clearly detected at 6 h (Fig. 8I) in some microglial cells with amoeboid/phagocytic form.
Overall, the analysis at cellular level showed a relatively heterogeneous pattern of change in the regions analyzed.Everywhere, however, neuronal cells appeared to be the predominant cell type expressing increased H3S10phK14ac immunoreactivity after LPS.Astroglial cells were absent or modestly positive in most areas (Fig. 8H, J), whereas several microglial cells with an amoeboid/phagocytic form showed H3S10phK14ac immunoreactivity in all the regions analyzed, especially 6 h after LPS administration.

Incubation with LPS elicits different effects on gene expression and nuclear H3S10phK14ac levels in hippocampal neuronal cell lines (HT-22) or murine microglia (BV-2)
We assessed IL-1β, IL-6, TNF-α, iNOS, and c-Fos transcription after a 2, 6, and 24 h exposure to LPS in HT-22 hippocampal neurons and BV-2 microglial cells (Fig. 9).One-way ANOVA revealed a main effect of incubation with LPS on mRNA levels of these targets in both cell lines, with Dunnett's posthoc analysis showing a different time-dependent pattern for BV-2 and HT-22.
In hippocampal neurons IL-1β, IL-6, and TNF-α expression was significantly induced by a 2h exposure to LPS (by about 4, 20, and 14 times, respectively) with respect to untreated cells.No significant difference was found between LPS treated cells and unstimulated controls 6 or 24 h after LPS Also iNOS expression showed a distinct temporal pattern between neurons and microglia: in HT-22 cells iNOS was induced by about 1100 times as early as 2 h after LPS exposure (F(3,35)=99.139,p<0.0001) with respect to untreated cells, while prolonging the incubation resulted in a decrease of this effect at 6 h, that was abolished at the 24 h time point (Fig. 9D).Conversely, in BV-2 cells the induction of iNOS expression by LPS with respect to unstimulated cells became statistically significant only after 6 h of exposure and was still present prolonging the incubation up to 24 h (by about 400 and 120 times, respectively; F(3,17)=1538.767,p<0.0001).
We also measured, by western blot, the effects of a 2, 6 and 24 h incubation with LPS on the levels of H3S10phK14ac in the nuclear fractions of both HT-22 and BV-2 cell lines (Fig. 9F).One-way ANOVA revealed a main effect of LPS exposure on H3S10phK14ac levels in HT-22 (F(3,23)=4.351,p=0.016) and BV-2 (F(3,11)=22.126,p<0.0001).A significant increase of H3S10phK14ac levels was detected in LPS treated hippocampal neurons with respect to unstimulated controls at all the time points considered.When BV-2 murine microglia were exposed to LPS for 2 or 6 h, H3S10phK14ac levels were not different with respect to untreated cells, but a significantly increase (by about 3 times) was observed when the incubation lasted for 24 h.LPS treatment did not modify total H3 protein levels in either HT-22 or BV-2 cells (data not shown).

Discussion
The present study demonstrated that a systemic (i.p.) administration of LPS affects the hypothalamic and hippocampal transcription of pro-inflammatory targets in an area-specific and time-dependent manner through specific epigenetic mechanisms.In particular, we demonstrated that the histone modification H3S10phK14ac is rapidly induced 2 h after the inflammatory challenge, decreases at 6 h post-LPS and is differentially regulated in cell populations of the areas analyzed.
ChIP results indicated that LPS was able to induce active chromatin remodeling by H3S10phK14ac modification at c-Fos, IL-6, and iNOS gene promoter in a region-and time-dependent manner.This suggests that phospho(S10)-acetylated(K14) histone H3 transcriptional regulation induced by LPS is, at least in part, responsible for the observed very high increase in hypothalamic and hippocampal iNOS transcript and for the lower increase of c-Fos and IL-6 transcripts in the hypothalamus.
As expected, LPS increased the expression of the main pro-inflammatory cytokines in the hippocampus and hypothalamus (Alboni et al., 2014b;Damm et al., 2012;Dantzer, 2009;Layé et al., 1994;Lynch, 2002), however its effects on IL-1β, IL-6, TNF-α mRNA levels showed distinct changes in the two brain areas.In particular, we observed a stronger induction in the hypothalamus than in the hippocampus.We also demonstrated a time dependent massive increase of the levels of mRNA coding for the iNOS gene.The mRNA coding for the enzyme responsible of NO synthesis was enhanced 2 h after LPS injection in both areas, and this effect was still present and further increased 6 h post-injection.Expression of c-Fos, a marker of activated neurons, was markedly induced 2 and 6 h following the LPS challenge in the hypothalamus while no effect was observed in the hippocampus.Transcriptional effects on immediate-early genes suggest distinct temporal patterns of brain reactivity following LPS exposure (S Alboni et al., 2017;da Silva et al., 2014;Frenois et al., 2007;Ryabinin et al., 1997).Consistently with our data, the hypothalamus appears to be recruited in the early phase of the response to LPS, as suggested by the strong induction of c-Fos expression and FosB immunoreactivity occurring within 6 h post-injection observed by Frenois and colleagues (Frenois et al., 2007), while the hippocampus, not recruited in terms of c-Fos expression at any of the time points considered in the present paper, may mediate the delayed effect of LPS, with changes occurring at a later stage.
H3S10phK14ac has been established to be an activating histone modification favouring specific gene transcriptional events (Clayton et al., 2000;Nowak and Corces, 2000).Increased H3S10phK14ac in the promoter regions of c-Fos in the brain has been demonstrated following other stimuli such as electroconvulsive seizures (Tsankova et al., 2004), administration of several neurotransmitter receptor agonists (Crosio, 2003), stress (Chandramohan et al., 2007;Collins et al., 2009).Present and previous evidence indicates that H3S10phK14ac is a crucial, though not necessary (Kumar et al., 2005), component of the epigenetic cascade driving c-Fos expression in the brain though other histone modifications may also contribute to the activatory regulation (Mifsud et al., 2011;Tsankova et al., 2004).
The causal link between H3S10phK14ac levels and target gene activation is reinforced by ChIP experiments that highlight that H3S10phK14ac increase precedes the peak of transcriptional activity of target genes and in parallel their temporal and regional heterogeneity.For instance, both H3S10phK14ac association to iNOS gene and iNOS mRNA levels are higher in hypothalamus than in hippocampus, and H3S10phK14ac association to c-Fos gene and increased c-Fos mRNA levels are only present in the hypothalamus.
In addition, LPS-induced changes in H3S10phK14ac are gene-specific since other genes, such as IL-1β and IL-6, whose mRNA levels are markedly increased by LPS treatment in both hypothalamic nuclei and hippocampus, did not show a significant association with H3S10phK14ac.
These data suggest that the two regions considered use different transcriptional regulation mechanisms in response to LPS stimulation.In hypothalamus the transcriptional activation of iNOS, IL-6, and c-Fos is bound, albeit to a different extent, to H3S10phK14ac levels on the promoter of these genes, whereas, in hippocampus, only iNOS promoter is upregulated after the inflammatory insult, suggesting that in this region the transcriptional response to LPS administration follows different transcriptional dynamics, possibly involving other histone modifications or epigenetic mechanisms.
We observed also a significant increase in H3S10ph after LPS treatment as also demonstrated in other animal models of brain insult, such as kainate or pilocarpine induced-epilepsy (Mori et al., 2013;Taniura et al., 2006), depression (Bagot et al., 2014;Nestler, 2014) and ethanol exposure (McClain and Nixon, 2016;McClung and Nestler, 2008).After LPS challenge, increase in H3S10ph levels appeared less region-and time-dependent than H3S10phK14ac as it was increased by a similar extent in both hypothalamus and hippocampus at 2 and 6 h post-LPS injection.In fact, while H3S10ph in both areas is induced after 2 h and is still elevated even at 6 h, the H3S10phK14ac is attenuated at 6 h.
Notwithstanding the two histone modifications studied here have been shown to occur in a cooperative fashion, with phosphorylation of H3S10 promoting subsequent acetylation of H3K14 (Nowak and Corces, 2000), they are thought to positively regulate transcriptional processes targeting different genes and undergoing different post-stimulus changes (Chandramohan et al., 2007;James et al., 2012;Sawicka and Seiser, 2012).Accumulated data identified the MAPK/ERK cascade via mitogen-and stress-activated protein kinase (MSK) pathway as the mechanism mediating phosphorylation, acetylation and phosphoacetylation of histone H3 (Ciccarelli et al., 2013;Hamon and Cossart, 2008;Saccani et al., 2002;Soloaga et al., 2003;Thomson et al., 1999).
The MAPK/ERK pathway results activated after different stimuli and its blockade, or the use of knockout cells for MSK1 and MSK2, leads to a marked reduction in the histone H3 phosphorylation, acetylation, and phospho-acetylation (Chwang et al., 2006;Soloaga et al., 2003).
Overall, present data suggest that H3S10phK14ac is a principal epigenetic mechanism that organizes the specific response to LPS of a large number of neuronal and non-neuronal cells.Cellular distribution analyses indicate that in both hypothalamic nuclei and hippocampus, neuronal cells appear to be the main cell population responsible for the increase of H3S10phK14ac observed after LPS treatment.LPS administration is known to induce multiple neuroendocrine, autonomic and behavioral consequences, that include sickness behavior (hypolocomotion, social withdrawal, fatigue, anorexia, sleep pattern alterations), learning and memory deficits, and activation of the hypothalamus-pituitary-adrenal axis (Cunningham, 2013;Dantzer et al., 2008;Henry et al., 2009;Qin et al., 2007;Satta et al., 1998;Sens et al., 2017).This complex alteration of neuronal circuits tallies well with present evidence of extensive increased positivity for H3S10phK14ac, and the associated transcription activation, in many classes of neurons of hypothalamic and hippocampal circuits.Non-neuronal cells showed a sharp increase in H3S10phK14ac immunoreactivity even if the number of positive cells was more limited and heterogeneous compared to the number of positive neurons.
H3S10phK14ac immunoreactive microglial cells were visible both in hypothalamus and hippocampus though only a few of them were intensely positive and only at 6 h post-LPS injection.
Notably, these microglial cells with intense positivity for H3S10phK14ac had an amoeboid/phagocytic form, that is typical of activated cells (Cherry et al., 2014).These data are in line with previous evidence on glial response to LPS.At early time points (first h after LPS), microglial activation has been repeatedly shown and is thought to contribute to both the inflammatory responses to LPS and the stimulation of neuron-mediated sickness behavior (Jin et al., 2016).
A few H3S10phK14ac immunoreactive astroglial cells were visible only in the hippocampus at 2 h and, as regards astrocytes with reactive morphology, at 6 h.A late activation of astrocytes tallies well with the evidence that cytokine expression and release by astrocytes peaks 12-24 h after LPS challenge and parallels the resolution of sickness behavior and microglial activation (Norden et al., 2016).
Besides glial cells, in the brain regions analyzed, other types of non-neuronal cells showed increased H3S10phK14ac levels, including ventricular ependymal cells and vascular endothelial cells.In fact, these cell types represent key components of the brain systems that monitor the presence of peripheral innate immune responses and trigger the associated alterations in brain circuits (Cheung et al., 2000;Dantzer et al., 2008).
The cell populations that showed the most intense H3S10phK14ac expression after LPS treatment were neurons and amoeboid, i.e., activated, microglia.Activated microglia become deramified and develop an enlarged cell body with several short, thickened processes.This morphological transformation occurs within hours from the initial activation and coincides with microglial homing and adhesion to damaged neurons (Kloss et al., 1999;Raivich et al., 1999).Microglia plays a fundamental regulatory role on neuronal function and homeostasis under physiological conditions, and it is thought that disruption of such fundamental processes upon an immune challenge or inflammation, may contribute to neuronal or synaptic dysfunction (Frick et al., 2013;von Bernhardi et al., 2015).
The time-and cell-specific pattern of H3S10phK14ac expression in vivo tallies well with our in vitro results showing a different pattern of regulation for H3S10phK14ac following LPS exposure between hippocampal neurons and microglia.While H3S10phK14ac levels were increased in neurons as early as after a 2 h exposure and remained elevated at least until 24 h in the hippocampal cell line HT-22, in microglia the phosphoacetylation of histone H3 was increased only after 24 h.Different transcriptional regulation mechanisms in response to LPS stimulation between BV-2 and HT-22 were confirmed by gene expression results as well.mRNA levels of genes related to the inflammatory response were strongly induced in HT-22 by LPS after a 2 h exposure and returned to basal levels following longer treatment, while for BV-2 cells the transcriptional effect of LPS was detectable even after exposure to LPS for 6 and 24 h.Future studies will be needed to dissect the role of H3S10phK14ac in the transcriptional activation induced by LPS in different cell populations in vivo and in vitro, between HT-22 and BV-2.
iNOS hyperexpression is a principal molecular outcome of microglial activation since high concentrations of the radical molecule NO constitute a CNS cellular defense mechanism aimed at removing cellular debris and combating infection.However, excessive production of NO may induce neuroinflammation-mediated neuronal loss (Yuste et al., 2015).Accordingly, iNOS induction was shown to cause neuronal damage in the hippocampus leading to memory impairments (Abd El-Aleem et al., 2008;Carreira et al., 2015;Chuang et al., 2010;Luo et al., 2007), while in the hypothalamus it can lead to neuroendocrine responses to infection (McCann et al., 1997;Satta et al., 1998), insulin-resistance and consequent alteration in energy homeostasis observed in LPSinduced sickness behavior (Katashima et al., 2017;Ribeiro et al., 2013).Furthermore, in the same experimental condition, NO may activate the hypothalamic-pituitary-adrenal axis (Uribe et al., 1999).Besides microglia, also neurons and endothelial/ependymal cells may contribute to the observed increase in iNOS mRNA expression as already shown after traumatic brain injury, and oxygen and glucose deprivation (Heneka and Feinstein, 2001;Petrov et al., 2000;Rodrigo et al., 2001).In these cells, differently from microglial cells, H3S10phK14ac increased expression occurs rapidly and remains sustained at 6 h post-LPS injection, as confirmed also from our in vitro results.
However, it is worth remembering that data from the literature suggest a different outcome for LPS on microglia activation depending on the route of administration, with centrally administered LPS inducing a clearer microglia activation than systemic LPS (Hoogland et al., 2015;Lawson et al., 2013).
Present data clearly imply an epigenetic molecular mechanism based on increased H3S10phK14ac expression and association with iNOS gene in LPS-induced overexpression of iNOS in different cellular populations of both the hypothalamus and the hippocampus.This evidence adds on previous studies showing the association of iNOS activation with specific chromatin modifications at the iNOS promoter observed in gastric cell after infection (Angrisano et al., 2012).

Conclusions
In this study we showed that a systemic LPS administration induces a marked region-and cellspecific phospho(S10)-acetylation(K14) of histone H3 (Fig. 10).LPS effects on H3S10phK14ac were cell population specific which emerged from both in vitro and in vivo experiments, even considering the limitations concerning the use of cell lines, as they do not always accurately replicate the features of native cells.
In addition, we propose that the increase in H3S10phK14ac may drive the transcription of some of the principal genes mediating LPS response in the brain, such as c-Fos, IL-6, and iNOS.Future experiments will focus on defining more complete molecular mechanisms which in turn will add to the generalizability of our findings.In particular, attention will also be directed to investigating which intracellular pathway mediates the effect of LPS on H3S10phK14ac and in determining a causal relationship between the transcriptional effects of LPS and histone H3 modifications by targeting in vitro phosphorylation of S10 or acetylation of K14.
The study underlines the importance of time-, region-and cell-specific epigenetic mechanisms that regulate gene transcription sustaining the neuroinflammatory response to an immune challenge.

Declaration of interest: None
Funding Sources: This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.sustained activation of neuronal cells starting as early as after a 2 h exposure and tardive recruitment of microglia in both the hippocampus and hypothalamus (A).This effect was further confirmed in vitro: in the hippocampal cell line HT-22 the phosphoacetylation of histone H3 remained elevated at least until 24 h, while in BV-2 microglial cells an increase was observed starting at 24 h (B).

Figure 3 .
Figure 3. Systemic LPS enhances the phosphorylation(S10) of histone H3 in the rat hypothalamus

Figure 9 .
Figure 9.Effect of incubation with LPS on the expression of the pro-inflammatory cytokines IL-1β,

Figure 10 :
Figure 10: Time-and cell-specific effects of LPS on H3S10phK14ac in vivo (A) and in vitro (B).LPS-

Table 1 :
Transcript and sequence of each primer used in real time PCR and ChIP assays.