Ethanol Induction of Innate Immune Signals Across BV2 Microglia and SH-SY5Y Neuroblastoma Involves Induction of IL-4 and IL-13

Innate immune signaling molecules, such as Toll-like receptors (TLRs), cytokines and transcription factor NFκB, are increased in post-mortem human alcoholic brain and may play roles in alcohol dependence and neurodegeneration. Innate immune signaling involves microglia -neuronal signaling which while poorly understood, may impact learning and memory. To investigate mechanisms of ethanol induction of innate immune signaling within and between brain cells, we studied immortalized BV2 microglia and SH-SY5Y human neuroblastoma to model microglial and neuronal signaling. Cells were treated alone or in co-culture using a Transwell system, which allows transfer of soluble mediators. We determined immune signaling mRNA using real-time polymerase chain reaction. Ethanol induced innate immune genes in both BV2 and SH-SY5Y cultured alone, with co-culture altering gene expression at baseline and following ethanol exposure. Co-culture blunted ethanol-induced high mobility group box protein 1 (HMGB1)-TLR responses, corresponding with reduced ethanol induction of several proinflammatory NFκB target genes. In contrast, co-culture resulted in ethanol upregulation of cytokines IL-4 and IL-13 in BV2 and corresponding receptors, that is, IL-4 and IL-13 receptors, in SH-SY5Y, suggesting induction of a novel signaling pathway. Co-culture reduction in HMGB1-TLR levels occurs in parallel with reduced proinflammatory gene induction and increased IL-4 and IL-13 ligands and receptors. Findings from these immortalized and tumor-derived cell lines could provide insight into microglial-neuronal interactions via release of soluble mediators in vivo.


Introduction:
Alcohol causes a dysregulation of the neuroimmune system in the brain which corresponds to alcohol-induced neurodegeneration and addiction pathology [1][2][3][4]. Alcohol consumption is also correlated with a variety of negative health outcomes, including but not limited to onset of alcohol use disorders (AUD), fetal alcohol spectrum disorders, alcoholic liver disease and cancers. In particular, Toll-like receptor (TLR) signaling has been suggested to be implicated in alcohol pathology. In the periphery, TLRs respond to viral and bacterial components, whereas in brain they respond to endogenous agonists such as high mobility group box protein-1 (HMGB1), which activates multiple TLRs [5][6][7]. Activation of TLRs leads to stimulation of transcription factors like NFκB, which regulates transcription of cytokines, propagating inflammation through their respective receptors [8]. Studies by Consuelo Guerri laid the groundwork for studying TLRs in alcohol pathology, finding a crucial role for TLR4 in ethanol-induced immune responses and neurodegeneration [9][10][11][12]. Multiple components of the TLR pathway have since been found to be upregulated in post-mortem human alcoholic brain; in particular, there is increased expression of TLR3, TLR4, TLR7 [13,14], as well as phosphorylated (activated) NFκB [15] and cytokines such as IL-1β and HMGB1 [13,16,17]. Other components of the TLR signaling, including TLR4 adapter MD-2 and downstream TLR signaling kinase TBK1, have been implicated in alcohol pathology [11,18]. Furthermore, alcohol-potentiated increases in TLR3 and TLR4 increase TLR agonist-induction of cytokines [19,20]. NFκB-related cytokines, such as MCP1 and its receptor, CCR2, have also been implicated in ethanol drinking [21]. These data therefore suggest that TLR pathways are implicated in alcohol pathology but exact mechanisms remain to be elucidated, such as the respective involvement of various cell types as well as cell-to-cell interactions that bring about ethanol-induced innate immune signaling.
The innate immune system takes on unique roles outside of traditional inflammation mediation in brain. In particular, cytokines often function as signaling molecules between different cell types in brain. Several of these systems have been previously described for their roles in development and synaptic pruning, such as neuronal fractalkine (CX3CL1) signaling through its microglial receptor (CX3CR1) or the complement system [22,23]. Cytokines released by activated microglia can also cause neurotoxic activity in astrocytes [24]. However, the ability of other cytokine systems to function as microglial-neuronal signaling mechanisms is less well described. In particular, IL-4 and IL-13 signaling is an emerging glia-to-neuron signaling pathway. In the periphery, IL-4 and IL-13 function as anti-inflammatory cytokines produced by Th2 cells [25]. The receptors for these cytokines form a heterodimer, consisting of IL-4Rα1 (IL-4R) and IL-13Rα1 (IL-13R), indicating that these two cytokines have similar signaling pathways. In brain, microglia appear to express low levels of IL-4 and IL-13 basally but can induce expression under certain conditions [26,27]. Interestingly, IL-13 receptor is located on dopaminergic neurons and overactivation of this receptor has been linked to loss of dopaminergic cells [28], prompting studies on the impact of alcohol exposure due to the importance of dopamine for drug-induced reward. Furthermore, IL-4 and IL-13 also have roles in learning and memory [29,30]. Stimulation of this pathway leads to activation of transcription factor STAT6, which is linked to IL-10 and TGFβ expression, cytokines which themselves have roles in neuroprotection and learning/memory [31][32][33][34].
Microglia, the canonical macrophages in brain, are thought to be the main mediator of innate immune signaling in brain and are therefore the focus of many studies on ethanol-induced innate immune signaling. Indeed, Guerri's previous studies have shown that ethanol induces various cytokines in microglial cultures [10,12]. Our lab has reported neuronal localization of TLR3, TLR4 and HMGB1 in rat brain [35], as well as neuronal IL-1β in mouse brain [17]. In addition, we recently found that ethanol increases both TLR3 and TLR7 in SH-SY5Y neuroblastoma and at lower ethanol concentrations than in BV2 microglia, in addition to HMGB1 release in both cell types [36]. These experiments were conducted in single cell types, whereas in brain there is constant and dynamic communication between neurons and microglia [37]. However, it remains unknown how microglial-neuronal interactions via secretion of soluble mediators can affect these ethanol-induced changes in innate immune gene expression. In addition, while IL-13 is increased in serum of cirrhotic alcoholic patients [38], little is known about how alcohol affects IL-4 and IL-13 signaling.
In this study, we examine how co-culture of BV2 microglia and SH-SY5Y neuroblastoma influences ethanol-induced innate immune signaling of these cell lines. Co-culture modules using glial cells and SH-SY5Y have been used previously in multiple disease settings [39,40]. Since SH-SY5Y adopt a dopaminergic phenotype [41] and IL-4 and IL-13 modulate dopaminergic signaling, SH-SY5Y could provide insight into responses of dopaminergic neurons. We extend previous findings by finding that co-culture modifies induction of cytokines and TLRs in BV2 and SH-SY5Y. We next found that ethanol upregulates IL-4 and IL-13 signaling in co-cultured BV2 and SH-SY5Y. Using a primary hippocampal-entorhinal slice culture model we found that IL-4 and IL-13 diminish expression of pro-inflammatory cytokines TNFα and IL-1β. Thus, co-culture of BV2 and SH-SY5Y results in differential immune responses to ethanol than exposure of each cell type to ethanol alone.
In all groups treated with ethanol, a concentration of 100 mM (equivalent to 460 mg/dL or 0.46%) for 24 h was used. This high concentration is consistent with the high (>400 mg/dL) blood alcohol concentrations (BACs) achieved in alcoholics in emergency rooms [42], including awake and conscious patients with a BAC of greater than 500 mg/dL [43]. Binge models in rodents that achieve similarly high BACs have also been used to model heavy drinking [44]. Cells exposed to ethanol were placed into an ethanol-saturated chamber using a beaker with 200 mL of 4% ethanol, allowing the media to remain constant throughout the duration of the experiment as we have reported previously [45]. Control groups were given an equal volume of vehicle (PBS). Following treatment, cells lysates were harvested for real-time polymerase chain reaction (RT-PCR) analysis. For each experiment, 5-6 culture wells were treated per group.

Transwell Co-Culture Model
SH-SY5Y were plated (3.75 × 10 5 cells per well) on 6-well plates. After the differentiation procedure described above, media was refreshed to 10% FBS-containing DMEM and Transwell inserts (Corning CLS3452, Corning, NY, USA) were placed on top of the SH-SY5Y-containing wells and allowed to equilibrate for 30 minutes. Transwells are a well-characterized 3D model of co-culture and have been used extensively for neuronal co-cultures [40,[46][47][48], including with SH-SY5Y and BV2 in particular [49]. BV2 were plated (1.5 × 10 5 cells per well) on top of the Transwell inserts; see Figure 1 for a schematic. Sixteen hours prior to ethanol treatment (30 h following plating of BV2), media was changed to 2% FBS. See Figure 1 for a diagram of the treatment outline. expression of pro-inflammatory cytokines TNFα and IL-1β. Thus, co-culture of BV2 and SH-SY5Y results in differential immune responses to ethanol than exposure of each cell type to ethanol alone.
In all groups treated with ethanol, a concentration of 100 mM (equivalent to 460 mg/dL or 0.46%) for 24 hours was used. This high concentration is consistent with the high (>400 mg/dL) blood alcohol concentrations (BACs) achieved in alcoholics in emergency rooms [42], including awake and conscious patients with a BAC of greater than 500 mg/dL [43]. Binge models in rodents that achieve similarly high BACs have also been used to model heavy drinking [44]. Cells exposed to ethanol were placed into an ethanol-saturated chamber using a beaker with 200 mL of 4% ethanol, allowing the media to remain constant throughout the duration of the experiment as we have reported previously [45]. Control groups were given an equal volume of vehicle (PBS). Following treatment, cells lysates were harvested for real-time polymerase chain reaction (RT-PCR) analysis. For each experiment, 5-6 culture wells were treated per group.

Transwell Co-Culture Model
SH-SY5Y were plated (3.75 × 10 5 cells per well) on 6-well plates. After the differentiation procedure described above, media was refreshed to 10% FBS-containing DMEM and Transwell inserts (Corning CLS3452, Corning, NY, USA) were placed on top of the SH-SY5Y-containing wells and allowed to equilibrate for 30 minutes. Transwells are a well-characterized 3D model of co-culture and have been used extensively for neuronal co-cultures [40,[46][47][48], including with SH-SY5Y and BV2 in particular [49]. BV2 were plated (1.5 × 10 5 cells per well) on top of the Transwell inserts; see Figure  1 for a schematic. Sixteen hours prior to ethanol treatment (30 hours following plating of BV2), media was changed to 2% FBS. See Figure 1 for a diagram of the treatment outline. SH-SY5Y were plated on the bottom of 6-well culture plates and Transwell inserts were placed in the wells. BV2 microglia were plated on top of the Transwells. Cells were treated with ethanol (100 mM) for 24 hours, keeping the ethanol concentration constant by saturated the chamber with ethanol. This was followed by RNA isolation in both BV2 microglia and SH-SY5Y neuron cell lysates. SH-SY5Y were plated on the bottom of 6-well culture plates and Transwell inserts were placed in the wells. BV2 microglia were plated on top of the Transwells. Cells were treated with ethanol (100 mM) for 24 h, keeping the ethanol concentration constant by saturated the chamber with ethanol. This was followed by RNA isolation in both BV2 microglia and SH-SY5Y neuron cell lysates.

Hippocampal Entorhinal (HEC) Slice Culture
All protocols followed in this study were approved by the Institutional Animal Care and Use Committee at UNC and were in accordance with National Institutes of Health regulations for the care and use of animals in research. Organotypic brain slice cultures were prepared as described previously [50]. Briefly, the hippocampal entorhinal region was dissected and sliced transversely (375 µm thick) from postnatal day 7 rat pups. HEC slices were placed onto tissue insert membrane (10 slices/insert) and cultured with medium containing 75% MEM with 25 mM HEPES and Hank's salts, 25% horse serum (HS), 5.5 g/L glucose, 2 mM L-glutamine in a humidified 5% CO2 incubator at 36.5 • C for 7 days in vitro (DIV), followed by 4 DIV in medium containing 12.5% HS and then 3 DIV in serum-free medium supplemented with N2 (Thermo Fisher #17502001, Waltham, MA, USA). The cultures after 14 DIV were used for experiments and drug treatments with serum-free N2supplemented medium. For ethanol exposures, slices were exposed to ethanol (100 mM) for 48 h in a desiccator saturated with an equivalent concentration of ethanol to prevent evaporation from the media. Recombinant rat IL-4 protein (500 ng/mL; R&D Systems, #504-RL-005) and recombinant rat IL-13 protein (1 ug/mL; R&D Systems, #1945-RL-025) were also added to the indicated groups for 48 h.

Real-Time (RT)-PCR
RNA was extracted from cell lysates using TRIzol (Invitrogen #15596026, Carlsbad, CA, USA). RNA was extracted from HEC slices using RNeasy Mini Kit (Qiagen #74104, Hilden, Germany). RNA concentration was determined using a Nanodrop (Thermo Fisher) and was reverse-transcribed to cDNA. The SYBR green PCR master mix (Life Technologies #4309155) was used for real-time PCR analysis. The relative differences in expression between groups were expressed using cycle time (Ct) values normalized with β-actin and relative differences between control and treatment groups were calculated and expressed as relative increases setting control as 100%. Primers used are listed in Table 1.  TCA ACT AA  TCA TAA CGA GCC TTG TCA GC  IKKβ  Mouse  GGAGCCTGGGAAATGAAAGAA  GCCAGAGCCCTACCTGATTG  IL-10  Mouse  GGC AGA GAA GCA TGG CCC AGA A  GGC TGA GGC GCT GTC ATC GAT T  IL-13  Mouse  CCTGGCTCTTGCTTGCCTT  GGTCTTGTGTGATGTTGCTCA  IL-13R  Mouse  TCAGCCACCTGTGACGAATTT  TGAGAGTGCAATTTGGACTGG  IL-1β  Mouse  CTG GTG TGT GAC GTT CCC ATT A  CCG ACA GCA CGA GGC TTT  IL-4  Mouse  TGGGTCTCAACCCCCAGCTAGT  TGCATGGCGTCCCTTCTCCTGT  IL-4R  Mouse  TCTGCATCCCGTTGTTTTGC  GCACCTGTGCATCCTGAATG  iNOS  Mouse  GTGGTGACAAGCACATTTGG  CATGGTGAACACGTTCTTGG  MCP1  Mouse  ACT GAA GCC AGC TCT CTC TTC CTC  TTC CTT CTT GGG GTC AGC ACA GAC  MD-2  Mouse  CGC TGC TTT CTC CCA TAT TGA  CCT CAG TCT TAT GCA GGG TTC A  RAGE  Mouse  GAA GGC TCT GTG GGT GAG TC  CCG CTT CCT CTG ACT GAT TC  TBK1  Mouse  ACTGGTGATCTCTATGCTGTCA  TTCTGGAAGTCCATACGCATTG  TGFβ  Mouse  CTCCCGTGGCTTCTAGTGC  GCCTTAGTTTGGACAGGATCTG  TLR3  Mouse  GTG AGA TAC AAC GTA GCT GAC TG  TCC TGC ATC CAA GAT AGC AAG T  TLR4  Mouse  ATG GCA TGG CTT ACA CCA CC  GAG GCC AAT TTT GTC TCC ACA  TLR7  Mouse  ATG TGG ACA CGG AAG AGA CAA  GGT AAG GGT AAG ATT GGT GGT G  TNFα  Mouse  GAC CCT CAC ACT CAG ATC ATC TTC T  CCT CCA CTT GGT GGT TTG CT  TREM1  Mouse  GACTGCTGTGCGTGTTCTTTG  GCCAAGCCTTCTGGCTGTT  TREM2 Mouse Rat ATGTGGAACTGGCAGAGGAG ACGAGCAGGAATGAGAAGAGG

Enzyme-Linked Immunosorbent Assay (Elisa)
Media (1 mL) was collected after ethanol treatment and spun at 500× g for 10 minutes to remove cell debris. HMGB1 protein concentration in media was determined using an HMGB1 ELISA kit (IBL, Hamburg, Germany), per manufacturer's directions. Briefly, samples were diluted 1:3 using the supplied Diluent buffer. The HMGB1 standards (0-10 ng/mL) were prepared using the supplied HMGB1 stock and Diluent Buffer. 50 uL of Diluent buffer was added to each well of the supplied 96 well plate (which came pre-coated with an anti-HMGB1 antibody) using a multi pipette, followed by 50 uL of each standard and sample, both in duplicate. The plate was then incubated at 37 • C for approximately 24 h. The wells were then aspirated and the plate was washed 5 times with the supplied Wash Buffer (1×). 100 uL of the supplied Enzyme conjugate was added to each well and incubated at 25 • C for 2 h. The plate was then washed 5 times with Wash Buffer and 100 uL of the supplied Color solution was added to each well. Following a 30-minute incubation at room temperature, 100 uL of the supplied Stop solution was added into each well. The optical density was measured using a spectrophotometer at 450 nm, with a reference wavelength at 600. To calculate HMGB1 concentration, each duplicate was first averaged followed by subtraction of the blank value (i.e., only Diluent Buffer) to remove background. The OD of the standards (x-axis, log scale) was plotted against the concentration (y-axis, log scale), followed by a linear trendline. The concentration of each sample was determined according to the trendline equation, followed by correction based on the dilution factor.

Statistical Analysis
Two-way ANOVAs were performed on all co-culture experiments (consisting of 5-6 wells in each group) to determine the main effect and interaction of co-culture and ethanol treatment. The means represent the mean of the values from the 5-6 wells assessed for each group. Post-hoc Bonferroni analyses were used to determine significant effects between groups (co-culture control, co-culture ethanol, alone-control and alone-ethanol) [51]. HEC slice culture data was analyzed using a one-way ANOVA followed by post-hoc Bonferroni's test. Standard error of the mean (SEM) bars are shown for the data in order to better show how precisely the data define the mean in each particular group [52]. A p-value of less than 0.05 was considered significant. All data analysis was conducted using Prism Version 7 (GraphPad, La Jolla, CA, USA).
We previously found that HMGB1, an endogenous TLR agonist with cytokine-like activity when released, is released by ethanol treatment in both BV2 microglia and SH-SY5Y [36]. In order to investigate how co-culture alters this ethanol-induced release of HMGB1, we examined HMGB1 levels in the media of ethanol-treated BV2 microglia alone, SH-SY5Y alone, as well as BV2/SH-SY5Y co-cultures. We confirmed our previous findings of ethanol increasing HMGB1 release in both BV2 microglia and SH-SY5Y alone but interestingly, co-culture prevented ethanol-induced HMGB1 release ( Figure 4A). As diagramed in Figure 4B,C, our findings indicate that while ethanol-induced HMGB1 release is correlated with an increase of multiple NFκB-regulated proinflammatory innate immune genes, co-culture blocks both HMGB1 and other proinflammatory genes. Co-culture blocks ethanol-induced HMGB1 release. BV2 microglia alone, SH-SY5Y alone and co-cultured BV2/SH-SY5Y were treated with ethanol (EtOH, 100 mM) for 24 hours. Media was collected and analyzed for high mobility group box protein 1 (HMGB1) protein using enzyme-linked immunosorbent assay (ELISA). (A) HMGB1 was increased in the media of EtOH-treated BV2 microglia and SH-SY5Y but not in co-cultured cells. (B) Schematic diagraming EtOH increasing HMGB1 release in BV2 and SH-SY5Y alone, leading to increased expression of NFκB genes. (C) Schematic diagraming a lack of HMGB1 release in co-cultured BV2 and SH-SY5Y, preventing transcription of NFκB genes. n = 5-6 wells per group; **p < 0.01, ****p < 0.0001. Schematic made using Biorender (biorender.com).

Il-4 and IL-13 Expression Is Increased by Ethanol in Co-Cultured BV2 Microglia and SH-SY5Y
In the periphery, cytokines IL-4 and IL-13 commonly modulate anti-inflammatory responses and share receptors that dimerize [53], resulting in complementary signaling pathways. While IL-4 and IL-13 have been discovered in brain [54] and have also been shown to affect dopaminergic neurons [21], little is known about this unconventional glial-neuronal signaling pathway. We examined both IL-4 and IL-13 ligands and receptors in our BV2 microglia/SH-SY5Y co-culture model to further study this pathway following ethanol treatment. In the BV2 microglia, both IL-4 and its receptor, IL-4R, had a main effect of co-culture (IL-4: F [1,18]  Post-hoc analysis found that that ethanol only increased IL-4 (2-fold, p < 0.01) and IL-4 receptor (IL-4R; 2-fold, p < 0.0001) expression in cells co-cultured with SH-SY5Y ( Figure 5A, 5B). Interestingly, IL-13 was not detectable in BV2 . Co-culture blocks ethanol-induced HMGB1 release. BV2 microglia alone, SH-SY5Y alone and co-cultured BV2/SH-SY5Y were treated with ethanol (EtOH, 100 mM) for 24 h. Media was collected and analyzed for high mobility group box protein 1 (HMGB1) protein using enzyme-linked immunosorbent assay (ELISA). (A) HMGB1 was increased in the media of EtOH-treated BV2 microglia and SH-SY5Y but not in co-cultured cells. (B) Schematic diagraming EtOH increasing HMGB1 release in BV2 and SH-SY5Y alone, leading to increased expression of NFκB genes. (C) Schematic diagraming a lack of HMGB1 release in co-cultured BV2 and SH-SY5Y, preventing transcription of NFκB genes. n = 5-6 wells per group; ** p < 0.01, **** p < 0.0001. Schematic made using Biorender (biorender.com).

IL-4 and IL-13 Reduce Ethanol-Induced Tnfα and IL-1β in Hippocampal-Entorhinal Slice Culture
Previous experiments indicate that IL-4 reduces endotoxin-induced TNFα and IL-1β in human macrophages [55,56]; however, it is unclear how IL-4 and IL-13 may alter ethanol induction of these proinflammatory cytokines in brain. In order to investigate effects of IL-4 and IL-13 in an ex vivo setting containing all brain cell types, we treated hippocampal-entorhinal (HEC) brain slices from postnatal day 7 rats with IL-4 and IL-13 recombinant proteins. Both IL-4 and IL-13 reduced TNFα (0.7-fold, p < 0.001 and 0.48-fold, p < 0.0001, respectively) as well as IL-1β (0.51-fold, p < 0.0001 and 0.2-fold, p < 0.0001, respectively) mRNA expression ( Figure 7A). We next treated the HEC slices with EtOH (100 mM, 48 h) either with or without IL-4 and IL-13 proteins. We found that while EtOH increased expression of both TNFα (3.2-fold, p < 0.0001) and IL-1β (1.8-fold, p < 0.0001), both IL-4 and IL-13 reduced this increase ( Figure 7B). These data indicate that IL-4 and IL-13 are capable of reducing ethanol-induced proinflammatory cytokines TNFα and IL-1β in brain.

IL-10 and Tgfβ Are Increased by Ethanol in Co-Cultured BV2 Microglia and SH-SY5Y
IL-10 and TGFβ are both cytokines that are activated downstream of IL-4 and IL-13 signaling [26,57,58] and play roles in learning, memory and neuroprotection [31,34] that may be impacted by alcohol pathology. To determine whether these genes are impacted by neuronal-glial ethanol induced signaling, we examined expression of IL-10 and TGFβ in our co-culture model. In  Figure 8C-D). Overall, these data indicate that genes downstream of IL-4 and IL-13 signaling, TGFβ and IL-10, are induced in neuronal-glial cultures, further indicating a role for the IL-4 and IL-13 pathway induction in these cells.

Discussion
This study demonstrates for the first time that co-culture of SH-SY5Y and BV2 alters ethanol-induced expression of innate immune signaling molecules. Co-culture with SH-SY5Y prevents ethanol induction of cytokines TNFα and IL-1β in BV2 microglia. Similarly, co-culture with BV2 microglia prevents ethanol-induced increases in TLR3 and TLR7 in SH-SY5Y. This was accompanied by a lack of ethanol induction of HMGB1 in the media of the co-cultured cells. Previous studies have indicated that HMGB1 facilitates the activation of multiple TLRs, including TLR3, TLR4 and TLR7 [5,7]. Therefore, the lack of HMGB1 induction in co-cultured BV2 microglia and SH-SY5Y may play a role in the lack of TLR and NFκB induced genes, including TNFα and IL-1β (see diagram in Figure 9A). Furthermore, we found that co-culture in conjunction with ethanol increased several known microglial-neuronal signaling pathways, such as fractalkine and CD200. Fractalkine (CX3CL1) attenuates lipopolysaccharide-induced increases in TNFα in rat primary microglia [59], suggesting that BV2 microglial CX3CR1 activation may underlie some of the effects seen in our co-culture model. Similarly, in a Aβ model of inflammation, addition of neurons into microglial cultures inhibited induction of IL-1β and TNFα, which was blocked with a CD200 blocking antibody [60]. These studies suggest that future research should examine the possible contribution of fractalkine and CD200 in ethanol-induced innate immune signaling between cell types.
Our findings were surprising in that previous studies indicate ethanol induces HMGB1 as well as various pro-inflammatory cytokines and TLRs in SH-SY5Y cultures alone, hippocampal entorhinal brain slice culture and in brain [13,19,20]. Importantly, while most of our findings consist of mRNA data, our previous studies have indicated a correspondence between mRNA and protein for targets such as TNFα, IL-1 β and TLR7 [14,20]. However, a similar inhibition of innate immune activation in SH-SY5Y due to lipopolysaccharide + interferons or Aβ has been observed when co-cultured with THP-1 monocytes [61]. Our results therefore suggest a unique and specific mechanism of inhibiting HMGB1 and other TLR/cytokine expression that exists as a paracrine signaling mechanism between SH-SY5Y and BV2 microglia. Astrocytes, as well as direct cell-to-cell contact, may modulate this response in brain, leading to increased HMGB1 and pro-inflammatory cytokines observed in slice culture and in vivo. Indeed, certain microglial-neuronal signaling molecules, such as fractalkine, are neuroprotective when in soluble form whereas membrane forms are pro-inflammatory [62]. However, further research is needed to explore the unique modulatory effects of neuronal-microglial co-cultures and whether our findings are found in primary neuronal and microglial cultures. by a lack of ethanol induction of HMGB1 in the media of the co-cultured cells. Previous studies have indicated that HMGB1 facilitates the activation of multiple TLRs, including TLR3, TLR4 and TLR7 [5,7]. Therefore, the lack of HMGB1 induction in co-cultured BV2 microglia and SH-SY5Y may play a role in the lack of TLR and NFκB induced genes, including TNFα and IL-1β (see diagram in Figure  9A). Furthermore, we found that co-culture in conjunction with ethanol increased several known microglial-neuronal signaling pathways, such as fractalkine and CD200. Fractalkine (CX3CL1) attenuates lipopolysaccharide-induced increases in TNFα in rat primary microglia [59], suggesting that BV2 microglial CX3CR1 activation may underlie some of the effects seen in our co-culture model. Similarly, in a Aβ model of inflammation, addition of neurons into microglial cultures inhibited induction of IL-1β and TNFα, which was blocked with a CD200 blocking antibody [60]. These studies suggest that future research should examine the possible contribution of fractalkine and CD200 in ethanol-induced innate immune signaling between cell types. Our findings were surprising in that previous studies indicate ethanol induces HMGB1 as well as various pro-inflammatory cytokines and TLRs in SH-SY5Y cultures alone, hippocampal entorhinal brain slice culture and in brain [13,19,20]. Importantly, while most of our findings consist of mRNA data, our previous studies have indicated a correspondence between mRNA and protein for targets such as TNFα, IL-1 β and TLR7 [14,20]. However, a similar inhibition of innate immune activation in SH-SY5Y due to lipopolysaccharide + interferons or Aβ has been observed when co-cultured with THP-1 monocytes [61]. Our results therefore suggest a unique and specific mechanism of inhibiting HMGB1 and other TLR/cytokine expression that exists as a paracrine signaling mechanism between SH-SY5Y and BV2 microglia. Astrocytes, as well as direct cell-to-cell contact, may modulate this IL-4 and IL-13 are both described as anti-inflammatory cytokines in the periphery and in microglia these cytokines have been found to promote an M2 phenotype [63,64]. However, the impact of ethanol on IL-4 and IL-13 signaling has been poorly understood. We observed for the first time that ethanol induced both IL-4 and IL-13 and their receptors on BV2 microglia only when in the presence of SH-SY5Y. Interestingly, we also observed that IL-13 was not detectable in BV2 alone, paralleling previous studies that found that endotoxin induces IL-13 in primary microglia co-cultured with neurons but not alone or with astrocytes [27]. Interesting new research also finds that IL-4 and IL-13 may play roles in learning and memory [29,30], suggesting a role for neurons in IL-4 and IL-13 signaling. Furthermore, a novel role for neuronal IL-13Rα1 was recently discovered, in which stimulation of this receptor led to dopaminergic cell loss [28]. In our study, we confirmed that SH-SY5Y IL-4 and IL-13 signaling is enhanced following ethanol treatment in BV2/SH-SY5Y microglial co-cultures, with an increased induction of IL-4R and IL-13R by ethanol in co-cultured SH-SY5Y. Our data indicates that interactions between the two cell types induces IL-4 and IL-13 signaling in response to ethanol, a finding which may have relevance in both neuroprotection and learning, although further studies are needed to translate these findings in vivo. Our findings may represent a specific feature of microglial-neuronal signaling that may occur via the secretion of soluble mediators.
IL-4 and IL-13 both reduce endotoxin-induced IL-1β and TNFα in peripheral macrophages [55,56], suggesting that in brain these cytokines may inhibit the expression of pro-inflammatory cytokines. Indeed, we confirmed using hippocampal-entorhinal slice culture that both IL-4 and IL-13 not only decrease expression of IL-1β and TNFα but also inhibit ethanol induction of these cytokines. In brain, excessive expression of IL-1β and TNFα has been linked to disruption of long-term potentiation [65,66], an important mechanism that regulates learning and memory formation. Pro-inflammatory cytokines are also linked to neurodegeneration [67]. Therefore, induction of IL-4 and IL-13 signaling may underlie the lack of induction of pro-inflammatory cytokines in BV2/SH-SY5Y microglial co-culture.
We also found that downstream mediators of IL-4 and IL-13 signaling, TGFβ and IL-10, are induced by ethanol in co-cultured BV2 microglia, with TGFβ being induced in co-cultured SH-SY5Y. Although the exact mechanism of this activity is unknown, IL-4 and IL-13 signaling in peripheral macrophages activates transcription factor STAT6 [68]. IL-4 has also been shown to activate STAT6 in BV2 microglia [69]. This suggests that IL-4 and IL-13 may be signaling in a STAT6 dependent manner to promote expression of TGFβ and IL-10 (see Figure 9B); however further studies are needed to elucidate these responses.
As stated above, it is important to note that this study uses immortalized cell lines for the co-culture experiments. SH-SY5Y is a well-characterized neuronal line used for modeling dopaminergic neurons in Parkinson's disease [70] and BV2 are commonly used to model microglial immune responses, exhibiting a similar response as primary microglia upon lipopolysaccharide stimulation [71]. However, these findings should be investigated in primary neuronal and microglial co-cultures. Our co-culture model provides insight into signals that might occur in vivo via the release of soluble mediators but lacks cell-cell contact and the presence of other cell types. Our findings in rat primary hippocampal-entorhinal slice culture provides further evidence for the anti-inflammatory role of IL-4 and IL-13 in brain. Whether this signaling system is induced by ethanol in vivo should be explored.

Conclusions
Overall, our findings indicate a role for IL-4 and IL-13 in ethanol-induced signaling between BV2 microglia and SH-SY5Y. Interestingly, we also observed an overall effect of co-culture on BV2 microglia and SH-SY5Y gene expression, indicating that signaling molecules between the two cell types even in basal settings influence their activity. Further research is needed to elucidate these signaling mechanisms. We further show that ethanol induction of TLR agonist HMGB1 is blocked by co-cultured neurons and microglia. These data reveal that neuronal-microglial signaling is altered by ethanol and indicate that more research is needed to further explore these cell-to-cell signaling mechanisms and how they play a role in alcohol pathology.